Hoffmann, C. & Bublitz, W. (2017) - Pragmatics of Social Media

Pragmatics of Social Media
HoPs 11
Handbooks of Pragmatics
Editors
Wolfram Bublitz
Andreas H. Jucker
Klaus P. Schneider
Volume 11
De Gruyter Mouton
Pragmatics of Social
Media
Edited by
Christian R. Hoffmann
Wolfram Bublitz
De Gruyter Mouton
ISBN 978-3-11-043969-4
e-ISBN (PDF) 978-3-11-043107-0
e-ISBN (EPUB) 978-3-11-043111-7
Library of Congress Cataloging-in-Publication Data

A CIP catalog record for this book has been applied for at the Library of Congress.
Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available on the Internet at http://dnb.dnb.de.
© 2017 Walter de Gruyter GmbH, Berlin/Boston

Cover image: Efetova/iStock/Getty Images Plus
Typesetting: Dörlemann Satz GmbH & Co. KG, Lemförde
Printing and binding: CPI books GmbH, Leck
♾ Printed on acid-free paper
Printed in Germany
www.degruyter.com
Table of contents
Preface to the handbook series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
1. Log in: Introducing the pragmatics of social media

Christian R. Hoffmann. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
I. The nature of social media
2. Participation as user involvement

Daniela Landert. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3. Participation as audience design

Marta Dynel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4. Publicness and privateness

Birte Bös and Sonja Kleinke. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
II. Social media platforms
5. Message boards
Jenny Arendholz. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
6. Blogs
Theresa Heyd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
7. YouTube
Marjut Johansson. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
8. Twitter
Michele Zappavigna. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
9. Social network sites/Facebook

Volker Eisenlauer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
xii Table of contents
III. Social media and discourse
10. Discourse and organization

Maximiliane Frobenius and Cornelia Gerhardt. . . . . . . . . . . . . . . . . . 245
11. Discourse and topic

Elisabeth Fritz. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
12. Discourse and cohesion

Christoph Schubert. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
13. Discourse and cognition

Andreas Langlotz. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
14. Discourse and ideology

Stephen Pihlaja and Andreas Musolff . . . . . . . . . . . . . . . . . . . . . . . . . 381
IV. Social media and identity
15. Facework and identity

Miriam A. Locher and Brook Bolander . . . . . . . . . . . . . . . . . . . . . . . . 407
16. Evaluation
Michele Zappavigna. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
17. Politeness and impoliteness

Sage Lambert Graham. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
18. Flaming and trolling

Claire Hardaker. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493
19. Narration
Ruth Page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523
20. Fandom
Monika Bednarek. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
Table of contents xiii
V. Social media and functions
21. Getting “liked”

Carmen Maíz-Arévalo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575
22. Conflictual and consensual disagreement

Brook Bolander and Miriam A. Locher . . . . . . . . . . . . . . . . . . . . . . . . 607
23. Compliments and compliment responses

María Elena Placencia and Amanda Lower. . . . . . . . . . . . . . . . . . . . . 633
24. Requesting and advice-giving

Phillip R. Morrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
About the authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691

Name index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
Subject index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721
Preface to the handbook series
Wolfram Bublitz, Andreas H. Jucker and Klaus P. Schneider
The series Handbooks of Pragmatics, which comprises thirteen self-contained

volumes, provides a comprehensive overview of the entire field of pragmatics.
It is meant to reflect the substantial and wide-ranging significance of pragmatics
as a genuinely multi- and transdisciplinary field for nearly all areas of language
description, and also to account for its remarkable and continuously rising popu-
larity in linguistics and adjoining disciplines.
All thirteen handbooks share the same wide understanding of pragmatics as the
scientific study of all aspects of linguistic behaviour. Its purview includes patterns
of linguistic actions, language functions, types of inferences, principles of commu-
nication, frames of knowledge, attitude and belief, as well as organisational prin-
ciples of text and discourse. Pragmatics deals with meaning-in-context, which for
analytical purposes can be viewed from different perspectives (that of the speaker,
the recipient, the analyst, etc.). It bridges the gap between the system side of lan-
guage and the use side, and relates both of them at the same time. Unlike syntax,
semantics, sociolinguistics and other linguistic disciplines, pragmatics is defined
by its point of view more than by its objects of investigation. The former precedes
(actually creates) the latter. Researchers in pragmatics work in all areas of lin-
guistics (and beyond), but from a distinctive perspective that makes their work
pragmatic and leads to new findings and to reinterpretations of old findings. The
focal point of pragmatics (from the Greek prãgma ‘act’) is linguistic action (and
inter-action): it is the hub around which all accounts in these handbooks revolve.
Despite its roots in philosophy, classical rhetorical tradition and stylistics, prag-
matics is a relatively recent discipline within linguistics. C. S. Peirce and C. Mor-
ris introduced pragmatics into semiotics early in the twentieth century. But it was
not until the late 1960s and early 1970s that linguists took note of the term and
began referring to performance phenomena and, subsequently, to ideas developed
and advanced by Wittgenstein, Ryle, Austin and other ordinary language philoso-
phers. Since the ensuing pragmatic turn, pragmatics has developed more rapidly
and diversely than any other linguistic discipline.
The series is characterised by two general objectives. Firstly, it sets out to
reflect the field by presenting in-depth articles covering the central and multifar-
ious theories and methodological approaches as well as core concepts and topics
characteristic of pragmatics as the analysis of language use in social contexts. All
articles are written specifically for this handbook series. They are both state of
the art reviews and critical evaluations of their topic in the light of recent devel-
opments. Secondly, while we accept its extraordinary complexity and diversity
DOI 10.1515/9783110431070-202
Brought to you by | Cambridge University Library

Authenticated
Download Date | 9/14/17 5:16 PM
vi Preface to the handbook series
(which we consider a decided asset), we suggest a definite structure, which gives

coherence to the entire field of pragmatics and provides orientation to the user of
these handbooks. The series specifically pursues the following aims:
– it operates with a wide conception of pragmatics, dealing with approaches that

are traditional and contemporary, linguistic and philosophical, social and cul-
tural, text- and context-based, as well as diachronic and synchronic;
– it views pragmatics from both theoretical and applied perspectives;
– it reflects the state of the art in a comprehensive and coherent way, providing a
systematic overview of past, present and possible future developments;
– it describes theoretical paradigms, methodological accounts and a large num-
ber and variety of topical areas comprehensively yet concisely;
– it is organised in a principled fashion reflecting our understanding of the struc-
ture of the field, with entries appearing in conceptually related groups;
– it serves as a comprehensive, reliable, authoritative guide to the central issues
in pragmatics;
– it is internationally oriented, meeting the needs of the international pragmatic
community;
– it is interdisciplinary, including pragmatically relevant entries from adjacent
fields such as philosophy, anthropology and sociology, neuroscience and psy-
chology, semantics, grammar, discourse and media analysis as well as literary
studies;
– it provides reliable orientational overviews useful both to students and more
advanced scholars and teachers.
The thirteen volumes are arranged according to the following principles. The first
three volumes are dedicated to the foundations of pragmatics with a focus on micro
and macro units: Foundations must be at the beginning (volume 1), followed by
the core concepts in pragmatics, speech actions (micro level in volume 2) and
discourse (macro level in volume 3). The following six volumes provide cognitive
(volume 4), societal (volume 5) and interactional (volume 6) perspectives and
discuss variability from a cultural and contrastive (volume 7), a diachronic (vol-
ume 8) and a medial (volume 9) viewpoint. The remaining four volumes address
methodological (volume 10), sociomedial (volume 11), fictional (volume 12), and
developmental and clinical (volume 13) aspects of pragmatics:
1. Foundations of pragmatics
Wolfram Bublitz and Neal Norrick
2. Pragmatics of speech actions
Marina Sbisá and Ken Turner
3. Pragmatics of discourse
Klaus P. Schneider and Anne Barron

Authenticated
Preface to the handbook series vii
4. Cognitive pragmatics
Hans-Jörg Schmid
5. Pragmatics of society
Gisle Andersen and Karin Aijmer
6. Interpersonal pragmatics
Miriam A. Locher and Sage L. Graham
7. Pragmatics across languages and cultures
Anna Trosborg
8. Historical pragmatics
Andreas H. Jucker and Irma Taavitsainen
9. Pragmatics of computer-mediated communication
Susan Herring, Dieter Stein and Tuija Virtanen
10. Methods in pragmatics
Andreas H. Jucker, Klaus P. Schneider and Wolfram Bublitz
11. Pragmatics of social media
Christian R. Hoffmann and Wolfram Bublitz
12. Pragmatics of fiction
Miriam A. Locher and Andreas H. Jucker
13. Developmental and clinical pragmatics
Klaus P. Schneider and Elly Ifantidou

Authenticated
Authenticated
Acknowledgements
This handbook represents the work of a large number of individuals over a fairly
long period of time. It goes without saying that we owe each of our contributors a
great debt of gratitude for accepting our invitation to contribute, for the time and
patience they have invested in this project and for the excellent results of their
hard work. They have made this handbook what it has eventually become. We
have enjoyed working with you and, in this process, have learnt a lot about the
pragmatics of social media. We would also like to express our gratitude to Barbara
Karlson, Birgit Sievert, Wolfgang Konwitschny and Frauke Schafft of de Gruyter,
who were as supportive and encouraging as they were professional in making sure
that this volume materialized in a timely manner.
Augsburg and Berlin, May 2017
DOI 10.1515/9783110431070-203

Authenticated
Authenticated
1. Log in: Introducing the pragmatics of social
media
Christian R. Hoffmann
Abstract: This introduction to the handbook of the Pragmatics of Social Media

is meant to explain its main structure, purpose and objectives, introduce the most
important features of the collection and define the key concepts of pragmatics and
social media. With research on social media gaining momentum, the introduction
makes a strong case for a pragmatic approach to studying these new platforms
for digital interaction, surfacing new insights into the way we interact, negoti-
ate, evaluate, judge and change propositional, social and interpersonal meanings
online. Considering the handbook’s five integral parts, it also provides short but
comprehensive overviews of all individual chapters, acknowledging their topical
and conceptual links and variances.
1. The aim of the handbook
This handbook provides a comprehensive overview of state-of-the-art pragmatic

research on social media. It pays tribute to the growing influence of social media
as a “new platform for socialization, public debate, and information exchange”
(Quan Haase and Sloan 2017: 2) as well as to the fact that social media have
become one of the primary objects for research in pragmatics over the last decade
(cf. Zappavigna 2012; Page 2013; Seargeant and Tagg 2014; Page et al. 2014; Tagg
2015). The individual contributions of this handbook are meant to help students
and researchers alike to find orientation in the wide range of research on social
media. It enables them to detect and critically appraise pragmatic work on social
media in a coherent and substantive manner. To this effect, all chapters emphasize
the conceptual and methodological similarities and differences between pragmatic
studies on the topic at hand while also evaluating and reflecting on their research
findings and interpretations. Just as the other volumes in the series of Handbooks
of Pragmatics (HoPs) have done, this handbook essentially aims to give structure
and coherence to a novel and still emergent field of pragmatic studies. Accordingly,
we have organized the handbook in ways consistent with this overall purpose.
At the very outset, we must acknowledge that this handbook is the series’ sec-
ond instalment specifically dedicated to pragmatic research on computer-mediated
communication (CMC); a fact which underscores the key relevance of the Internet
for pragmatic research today. While both handbooks, Pragmatics of Computer-
Mediated Communication (HoPs 9) and Pragmatics of Social Media (HoPs 11),
DOI 10.1515/9783110431070-001
In: C. R. Hoffmann and W. Bublitz (eds.). (2017). Pragmatics of Social Media, 1–28. Berlin/Boston: De Gruyter
Mouton. Brought to you by | Cambridge University Library
Authenticated
2 Christian R. Hoffmann
may share a mutual focus on computer-mediated communication, they are differ-

ent in many respects.
First, as opposed to its forerunner, this handbook focusses on a differ-
ent research object. While Pragmatics of Computer-Mediated Communication
(HoPs 9) reviewed pragmatic research on CMC in general, the present handbook
(HoPs 11) takes an exclusive look at second-generation Internet platforms which
foster and facilitate new participatory acts and forms of communication, i.e. social
media. This means that we have deliberately cast aside all pragmatic studies which
concentrate on classic Internet platforms, e.g. websites, Internet chats or email, for
the benefit of research describing the pragmatics of language use in digital plat-
forms such as social-network sites, blogs, microblogs, message boards, etc. In fact,
most of these forms of communication (as well as pragmatic research about them)
had not even fully evolved when HoPs 9 was devised more than ten years ago.
Second, this new handbook reviews linguistic research which captures the
pragmatics of participation, which, as we contend, is native to social media.
Unsurprisingly, most pragmatic research on social media has focussed, in one way
or another, on the participatory nature of social media, i.e. the technological-com-
positional and social-interpersonal factors which generate and drive two-way com-
munication between individual “produsers” (Bruns 2007: 3) online. Most prag-
matic work summarized and classified in this handbook recognizes the versatile
communicative moves through which Internet users reciprocally and collectively
create, share, contribute and negotiate meanings rather than go online to “predom-
inantly […] consume content and information” (Seargeant and Tagg 2014: 5).
In line with this interactional focus, pragmaticists today adapt and apply a
broader range of theoretical frameworks and methodological approaches, drawn
not only from pragmatics per se but also from other neighbouring fields of scien
tific inquiry. Such points of origins may, for instance, include interactional soci-
olinguistics, comparative film and media studies, sociology or social psychol-
ogy. Androutsopoulos (2008) referred to the contrast between earlier and more
recent studies on CMC as a first and a second wave of linguistic research. While
the first wave explored “features and strategies that are (assumed to be) specific
to new media […]”, the second wave investigated CMC’s “situated language
use and linguistic diversity” (Androutsopoulos 2008: 1–2). With respect to linguis-
tic diversity, pragmatic scholars have increasingly accentuated the vast generic
proliferation and dispersion that the Internet has undergone in the last two decades.
As a result, the latter has spawned an astounding number of specialized (hybrid)
genres and sub-genres online, each one designed for a specific set of purposes,
equipped with its very own set of formal and functional features. The increased
granularity of Internet communication, which results from this digital evolution,
has had considerable consequences for pragmatic research in the last decade. As
of now, pragmatic studies on social media are likely to make substantive claims
about the socio-pragmatic contour of specific Internet genres, e.g. political blogs,

Authenticated
Log in: Introducing the pragmatics of social media 3
celebrity Facebook profiles, academic tweets. However, (rather futile) attempts to

sketch out the more general linguistic or communicative shape of Internet plat-
forms, e.g. the language of Twitter, Facebook, message boards, have become rare.
In many ways, Internet pragmatics has by now assumed a more constructive posi-
tion between the two untenable extremes of technological determinism and social
constructivism.
Finally, a third important difference between this handbook and its CMC-
based predecessor (HoPs 9) concerns the organization of the two handbooks. Both
feature a consecutive number of in-depth overviews on various different pragmatic
phenomena in and about CMC. This handbook, however, also includes a thorough
and systematic introduction to its central notion of social media. The first part is
thus dedicated to explaining what social media are, eliciting and describing their
conceptual basis and key features, such as participation, involvement as well as
publicness and privateness. We hope to establish common ground on the forms
which social media comprise over time, show what they have been used for and
which role they now play in our everyday communicative practices. This central,
explanatory part provides a suitable springboard for the subsequent parts of this
handbook, making what comes more approachable and tangible in light of what
has been established before.
We need to remind ourselves that when HoPs 9 (Pragmatics of Computer-Me-
diated Communication) was published in 2013, the editors pointed out that prag-
matic research on computer-mediated discourse was still very much “in flux” (Her-
ring, Stein and Virtanen 2013: 4), which made the identification and organisation
of emerging pragmatic trends such a challenging endeavour. Today, the evolution
of discourse on the Internet (as well as the development of compatible pragmatic
research) is still in its early stage of development.
However, the body of research on social media which pragmatics has accu-
mulated over the last decade alone is substantial, and it is high time to compile
and review the existing pragmatic work to structure this novel field of pragmatic
inquiry. We hope that this handbook will play a part in this effort. It can provide
guidance to those who seek to know how social media have transformed our mean-
ing-making practices, those who wish to probe the interactional order of social
media discourse or those who aim to scrutinize the creative ways in which users
collaborate to share and negotiate meanings, feelings, values and beliefs in differ-
ent online settings. After all, the pragmatics of social media has much to offer not
only to linguists alone but to everyone who shares our interest in the way language
is currently used on social media, regardless of their scholarly provenance.

Authenticated
2. Defining social media
Before we cast our eye on social media, we feel it is necessary to clarify some of
the more conspicuous terms in electronic discourse which have been circulating
in linguistics over the last two decades, with varying meanings and uses. We thus
propose our own taxonomy of terms without, of course, making any claim to its
originality or completeness. We simply wish to illustrate what we intend to mean
when we use the terms and show how they are connected.
On the Internet, communicators use material, immobile (desktop computers)
or mobile devices (laptops, netbooks, tablets or smartphones) to perform network-
ing and socializing practices. In doing so, they rely on predetermined software.
These digital platforms, e.g. weblogs, message boards, social-network sites, etc.,
come equipped with particular technological affordances and provide specific
compositional templates for users. The platforms’ preconfigured text design does
not per se determine, restrict or preclude content, topic or function of any kind,
and it usually gives rise to a plethora of different text genres. As a result, rich
genre ecologies have taken root within the realm of each platform over time. Blog
templates, for instance, have been used for a variety of different purposes, such as
the semi-public revelation of personal experiences (personal blogs), the promo-
tional activities of companies (corporate blogs) or the journalistic documentation
of political events (journalistic blogs).
In the pre-digital era, linguists traditionally used the concept of medium to
relate to the material device or code system which enables human beings to pro-
duce, store, access and engage in communication over time, e.g. paper, voice,
computer screen or spoken and written language. In contrast, the term social media
refers to (the totality of) digitally mediated and Internet-based platforms which are
interactively used (by individual and collective participants) to exchange, share
and edit self- and other-generated textual and audio-visual messages. It is this
rather broad meaning of social media that we would like to adopt for the purpose
of this handbook.
In a sense, most media are, of course, social in that they can be used to provide
and share communicative content and thus to socialize. However, over the last
two decades, the epithet social has acquired a significant extension of its mean-
ing due to considerable technological expansion of the Internet and its concurrent
communicative affordances. As a convenient “buzzword” (Mandiberg 2012: 4), it
has attracted all kinds of narrow definitions depending on the technical, concep-
tual, linguistic, industrial- or business-related interests and needs of its users. As
a result, a number of competing concepts have been emerged over time and con-
tinue to co-exist beside social media, e.g. “Web 2.0” (O’Reilly 2005), “user-gener-
ated content” and “convergence culture” (Jenkins 2006) or “participatory media”
(Rheingold 2008: 97). However, all of these concepts have focussed on slightly
different aspects of social media, such as their potential for commercial exploita-

Authenticated
tion by enriching user experiences (Web 2.0), their general use in promoting busi-
ness/corporate-related objectives (user-generated content), the ethical responsibil-
ities and challenges which come with blurring the author-user divide (convergence
culture) and the social upshots of a collaborative multi-authored Internet (partic-
ipatory media). Despite these different colorations of the concepts, they all seem
to share a common descriptive basis which is perhaps best summarized by Mandi
berg (2012: 1):
These new frameworks have become more and more focused on enabling media crea-
tion, as this so-called amateur media becomes the raison d’être of these very profession-
al media organizations. These sites are pointless without audience participation: from
the audience’s perspective in order to experience the site you have to become a media
producer, and from the organization’s perspective, without audience participation, their
sites will fail.
In contrast to ‘classic’ Internet websites then, social media encourage Internet

users to post, comment, evaluate, link and contribute self-selected (audio-visual)
content. They are deliberately designed to encourage and enable non-expert users
to create, share and disseminate digital content (text, images, photos, videos) on
the platform. Social media thereby characteristically feature new technological,
participatory interfaces, including participatory tools such as opinion polls, com-
ment sections, guest posts, retweets, etc. It is through these participatory options
that social media manage to bridge the old productive divide between ‘active’
authors and ‘passive’ readers, with the self-proclaimed aim to neatly interlace the
processes of writing and reading (not only temporally but also conceptually).
In light of the relative activity and passivity of Internet users, it is useful to dis-
tinguish between the two terms of interaction and participation, frequently evoked
in this context. While the former refers “to the exchange of messages between
participants” (Landert, Ch. 2, this volume), the latter is either used synonymously
with the former, or, more frequently, goes beyond it. The notion of participation,
in other words, reflects the technological and interactional empowerment of Inter-
net users, which is exactly what social media strive to achieve. To this end, previ-
ous scholars (e.g. Seargeant and Tagg 2014: 2) have defined social media rather
broadly as “any digital environments which involves interaction between partici-
pants”. While this is, of course, very much true, we would still propose a slightly
modified description of social media as any digital platform which actively pro-
motes and enables the participation and interaction of Internet users.
In the end, social media want to transform the individual user experience into
a joint, collaborative social undertaking, in which form and content are not only
chosen by the few (authors) but by the many (users). A social medium thus needs
its authors to turn into readers and expects its readers to be authors, contributing
texts, videos, music, links, etc. to the platform. This is why we often experience
social media, e.g. Facebook, blogs, Twitter, etc., as communication hubs of larger

Authenticated
social collectives rather than the single work of an individual. By now, we have
become conscious of the (technologically empowered) potential of social media,
which enables us to actively participate in generating and sharing information with
myriads of other users, as the principal feature shared by all social media plat-
forms. At the core of our own notion of social media are Internet platforms such
as online message boards and discussion fora, blogs, microblogs (Twitter), social
network sites (MySpace, Facebook), media-sharing sites (YouTube, Instagram)
and instant messaging services (SMS, MMS, WhatsApp, Skype). Although this
list is far from being exhaustive, we believe that it captures the most ubiquitous
and widely used social media platforms today and serves as a veritable starting
point for all pragmatic departures into the realm of social media.
3. Defining pragmatics
In our view, pragmatics occupies a unique position among all fields of linguistic
analysis in that it is both an art and a scientific approach. It is the language user’s
art of understanding what is meant beyond what is explicitly said. Simplifying con-
siderably, we contend that the meaning of what is said (dictum) and the meaning of
what is not said but nonetheless meant (implicatum) can be referred to as ‘seman-
tic’ and ‘pragmatic meaning’ respectively. While semantics is concerned with the
dictum, i.e. the literal and pre-contextual meaning of sentences, pragmatics is con-
cerned with the implicatum, i.e. the incommensurable (and frequently figurative)
contextual meaning of utterances that is jointly negotiated and shared in situ by the
participants involved. Pragmatic meaning is the outcome of a complex process of
interpretive enrichment of semantic meaning. As language users, we are masters in
the art of reading between the lines, viz. of ascribing meaning to what is not articu-
lated in so many words. When understanding we cannot help but relating what we
hear to our contextual knowledge as well as to the entrenched conceptual schemas
and frames of our cognitive knowledge. As linguists, moreover, we employ prag-
matics as an approach to describe the principles and mechanisms, the means and
procedures that guide both the conveying speaker and the understanding hearer in
their endeavour to establish and secure successful communication.
Nonetheless, we have to acknowledge that (because of its dual nature as a
science and an art) there is not one generally accepted definition of pragmatics
as a unified and homogeneous field of study. As has been pointed out repeatedly,
most studies tend to prefer a narrow or a broad reading of pragmatics (of which the
former is sometimes allocated to an “Anglo-American” and the latter to a “Con-
tinental [European]” tradition of pragmatics, cf. Huang 2007: xi). According to
the narrow view, pragmatics is a module of language theory on a par with and at
the same time in juxtaposition to semantics. Its core issues include deixis, speech
acts, presuppositions, implicatures and related types of inferences. In this volume,

Authenticated
we thus adopt a much broader point of view and understand pragmatics as the
scientific study of all aspects of linguistic behaviour. In this we follow the general
outline guiding all volumes in the Handbooks of Pragmatics series as it is laid
down in the series’ preface:
All […] handbooks share the same wide understanding of pragmatics as the scientific
study of all aspects of linguistic behaviour. Its purview includes patterns of linguistic
actions, language functions, types of inferences, principles of communication, frames
of knowledge, attitude and belief, as well as organisational principles of text and dis-
course. Pragmatics deals with meaning-in-context, which for analytical purposes can be
viewed from different perspectives (that of the speaker, the recipient, the analyst, etc.).
It bridges the gap between the system side of language and the use side, and relates
both of them at the same time. Unlike syntax, semantics, sociolinguistics and other
linguistic disciplines, pragmatics is defined by its point of view more than by its objects
of investigation. The former precedes (actually creates) the latter. Researchers in prag-
matics work in all areas of linguistics (and beyond), but from a distinctive perspective
that makes their work pragmatic and leads to new findings and to reinterpretations of
old findings. The focal point of pragmatics (from the Greek prãgma ‘act’) is linguistic
action (and inter-action): it is the hub around which all accounts in these handbooks
revolve. (Preface to the handbook series)
As such, pragmatics also takes account of the interactional turn as a characteris-
tic feature of the most recent development of interactive (Web 2.0-based) media
formats, which is at the core of the present volume. In our pragmatic perspective,
we adopt a constructivist point of view which allows for the inclusion of (given
and new) contextual, situational, cognitive and notably media-technological var-
iables. In line with the general politics of our handbook series, we do not set up
any restrictions on methodology over and above the required focus on a pragmatic
perspective. Contributors are free to choose from the established repertoire of ana-
lytical methods to take account of social media, among them quantitative and qual-
itative, empirical and more heuristic approaches.
4. Overview of the handbook
This handbook consists of five main parts. The first part, “The Nature of Social
Media”, sets up the conceptual groundwork as it explores key concepts such as
social media, participation as well as privateness and publicness. The second part,
“Social Media Platforms”, focuses on the pragmatics of the most influential types
of social media, e.g. YouTube, Facebook, Twitter. The third part, “Social Media
and Discourse”, sets out to cover the micro- and macro-level organization of dis-
course on social media, while the fourth part, “Social Media and Identity”, reveals
the multifarious ways in which users collectively (re-)construct and reconfigure
aspects of their online (and offline) identities. The final and fifth part, “Social
Media and Functions”, surveys pragmatic studies on a selective collection of

Authenticated
speech act functions, e.g. disagreeing, complimenting, requesting. Each chapter

comprises a number of in-depth surveys, which are mostly large-scale overviews
or else overview-cum-case-study papers. The following presents the content and
purpose of each major part of this handbook in more detail, shedding more light
on each chapter individually.
4.1. The nature of social media

As indicated, to uncover the theoretical foundations, including the main features,
of our principle object of study, the present handbook opens with an expositional
section on the nature of social media. The authors of this section zone in on the
three characteristics which have come to shape the very essence of social media
communication, i.e. participation as user involvement, participation as audience
design and publicness and privateness.
In the first chapter of this section, Daniela Landert sets the scene by exploring
the conceptual differences between user interaction, participation and involve-
ment, explaining the various ways in which these concepts overlap and the differ-
ent communicative configurations they assume in social media environments. She
contends that while the concept of participation may be broader than the notion of
interaction, the term usually addresses the emotional or evaluative user engage-
ment. The distinctions Landert offers are helpful in contrasting social media’s tech-
nological potential to boost user participation (interactivity) with the actual degree
of reciprocal user engagement and exchange on social media platforms (interac-
tion and involvement). She equally clarifies the gradual communicative divide
between professional and non-professional Internet users (vertical communica-
tion) and the more ‘egalitarian’ communicative relationship which holds between
non-professional users (horizontal communication). While Landert reviews the
ways in which social media have created and promoted new forms (and means)
of participation over time, she likewise recognizes three main strategies which
social media users frequently employ to create user involvement, i.e. interaction,
linguistic immediacy and private content. Using a case study of political commu-
nication on Twitter, Landert exemplifies how the clever and strategic interplay
between language use, accessible private content and participatory technologies
can increase user involvement and boost the popularity of the social media chan-
nel. She also manages to show that these strategies are neither exclusively tied to
social media nor to electronic communicative practices in general but are even
applicable to pre-digital types of communication.
In the second chapter of this section, Marta Dynel describes how the imper-
ative drive of social media to facilitate and increase user-to-user interactions has
led to a noticeable shift in the participation structure of online communication. If
classic mass media have established well-defined production and reception roles
to assess the quality and status of public media communication (e.g. producer,

Authenticated
editor, author, printer vs. critic, addressed reader, unaddressed reader), social
media continuously blur such divisions, clouding the audience design of online
communication. Dynel argues that social media users must cope with “the loss of
control over their messages and the easy transgression of the boundaries typifying
the traditionally separate communicative frames” (Dynel, Ch. 3, this volume). In
this vein, she critically reviews the conceptual evolution of participation frame-
works in pragmatics, resketching their development from Erving Goffman’s the-
oretical insights, subsequent seminal work by Bell (1984), Clark (1996) or Clark
and Schaefer (1989) to more recent amendments by Hoffmann (2012), Eisenlauer
(2014), Dynel (2014) and Chovanec and Dynel (2015).
Decomposing the production format in social media contexts, Dynel draws
on the notion of collective producer, a collective of individual users whose indi-
vidual contribution to the process of publishing social media messages is hard to
fathom. She concedes that the reception format of social media is by no means less
complex. In great detail, she explores the different participation configurations
one encounters in social media, which raises questions of reader ratifiability (is
the author of the message aware of the reader, and do they know them?) as well
as audience design (does the author dovetail her message to a particular (ratified)
target audience?).
Birte Bös and Sonja Kleinke conclude the opening section of this handbook. In
their chapter, they take an inside look at how our general perception of privateness
and publicness has been altered with the rise of social media-based communica-
tion. Skilfully navigating the reader through the often messy web of terms and
definitions, Bös and Kleinke manage to pin down the conceptual pairing of pri-
vateness and publicness with the help of a range of prototypical, often interrelated,
characteristics, i.e. accessibility, visibility, persistence, replicability, scalability
and searchability.
While both authors give a first impression on how social media messages
can be shifted from the private sphere to the public arena (or vice versa), they
also stress the important difference between actual and perceived publicness/pri-
vateness in social media contexts. While the former refers to the actual, objec-
tive degree of technologically-induced publicness/privateness, the latter relates
the degree of publicness/privateness which Internet users personally ascribe to a
given interaction. Following recent research from pragmatics and communication
science (cf. Papacharissi 2010), Bös and Kleinke proceed to add a third component
to the key couplet of publicness and privateness, namely sociality, which serves as
an intermediary concept between the private, characterized as the safety zone of
intimacy and reclusion, and the political public, accessible and visible to anyone
anytime. They claim that sociality is especially relevant to social media interac-
tions in which users construe and negotiate collective norms and shared beliefs
over time while keeping a median level of visibility and accessibility. At the same
time, they often protect their personal interest and cover their local identities. In

Authenticated
what follows, Bös and Kleinke first decompose the related concept of privacy
into three integral sub-categories, i.e. informational privacy, social privacy and
psychological privacy. On this basis, they can explain the notion of the “privacy
paradox” (Barnes 2006), i.e. the contradictory effect that “the privacy concerns of
users do not necessarily have an impact on their actual online behavior” (Bös and
Kleinke, Ch. 4, this volume). They then shift their focus on various privacy man-
agement strategies social media users have applied in online settings.
The final part of their chapter is devoted to a systematic overview of pragmatic
research on the topic at hand, building on, but not restricted to, Koch and Oes-
terreicher’s seminal framework of the language of distance and proximity (Koch
and Oesterreicher 1985), Dürscheid’s improved model of mediality (Dürscheid
2007) or Landert and Jucker’s recent model of participation (Landert and Jucker
2011). In this vein, the authors trace down the evolution of compatible frame-
works, incorporating additional CMC-based approaches.
4.2. Social media platforms

The second part of this handbook covers a number of comprehensive summaries
of recent pragmatic research on the most popular types of social media today,
including chapters on message boards, blogs, video-sharing sites (YouTube),
micro-blogs (Twitter) and social-network sites (Facebook). In the first chapter of
this second part, Jenny Arendholz retraces the technological and communicative
development of message boards from their early beginnings to their contemporary
state. She contrasts various competitive concepts, such as newsgroups and Inter-
net forums, skilfully dismantling their technological, organisational and commu-
nicative differences. While she carefully discusses the most prominent features of
message boards, she does not forget to acknowledge their impact on the interactive
engagement of message board users as well as the communicative organization of
message board topics in so-called threads.
Having established a basic understanding of the technological and composi-
tional infrastructure of message boards, Arendholz then continues to group and
evaluate existing pragmatic research. Although she concedes that pragmatic
studies are still comparatively sparse, she manages to detect three major lines of
research on message boards: (a) (im-)politeness and conflict in message boards,
(b) identity and face(work) in message boards and (c) advice and authority in mes-
sage boards. She concludes her chapter with an additional glance at studies which
“deviate […] from the mainstream trends” outlined before (Arendholz, Ch. 5, this
volume), for instance Virtanen’s investigation (2013) of mock performatives in
message boards or Arendholz and Kirner-Ludwig’s (2015) contrastive analysis of
quoting strategies in message boards and medieval manuscripts.
Similar to Arendholz’s approach, Theresa Heydt begins the second chapter
with an outline of the evolution and relevance of blogging as one of the histor-

Authenticated
ical benchmarks of social media interaction. Describing the difficulties of defin-

ing blogs using a limited and conclusive set of formal and functional parameters,
Heydt resolves the issue by classifying different blog genres, featuring differing
topical, linguistic and functional profiles, e.g. personal blogs, filter blogs, cor-
porate blogs, etc. She draws on these generic distinctions between blog types as
a conceptual point of departure for her ensuing critical reflection on pragmatic
research on blogs in general. To this end, she first exposes more central readings
on the language of blogs, moving toward a critical evaluation of studies which she
allocates to more specialized pragmatic fields of interest. The research panorama
which Heydt thereby unfolds spans from traditional pragmatic desiderata, such
as Gricean pragmatics, relevance theory, politeness and deixis to closely related
sociolinguistic topics such as identity, narrativity and authenticity. While, natu-
rally, some of these themes may be explored in more detail in other parts of this
handbook, Heydt’s chapter certainly proves that blogs are still one of the most
widely researched types of social media, and the summaries she imparts are val-
uable shortcuts for those who wish to peruse the pragmatics of blogs in various
directions.
Marjut Johansson’s chapter discloses the social, technological and interactional
features of video-sharing sites, using the most popular example of the social media
site YouTube as a case in point. To do this, she explores the pragmatic nature of
YouTube from three different vantage points, (a) the sociocultural context, i.e. the
history and socio-technological development of YouTube, (b) the generic context,
i.e. the processes of generic reconfiguration and hybridisation by YouTube vid-
eos and (c) the interactional context, i.e. the impact on user participation through
comments in the construction and negotiation of meanings on YouTube. To this
effect, Johansson starts out by giving a historical overview of the technological
and communicative rise of YouTube over the last decades. She then uncovers the
topical, structural and technological influence of YouTube’s generic predecessors,
such as short films, music videos, humorous videos (pranks, candid camera) or
home video compilations. In addition to the extensive and useful description of
YouTube’s technological and interactional affordances, she exemplifies all three
of the aforementioned perspectives, drawing on YouTube videos by the vlogger
(videoblogger) PewDiePie.
In the final part of her chapter, Johansson takes more recent pragmatic studies
into consideration when she illustrates how identities are negotiated between users
in YouTube videos and comment sections; how Internet users contextualize their
verbal and visual contributions and how they try to address their contributions to
different sections of the blogging audience, blurring or specifying participation
types for their individual purposes.
In the next chapter, Michele Zappavigna introduces the micro-blogging service
Twitter. She starts with a short formal description of micro-blogging, eliciting the
historical roots of this social media service and surfacing its technological con-

Authenticated
straints and options, before turning to interdisciplinary and pragmatic research

on Twitter. Most research sees microblogging as a “semiotic technology” (Zap-
pavigna, Ch. 8, this volume) which enables users to engage in a complex range of
multimodal activities, including the upload of images, photos, links, videos, verbal
messages, etc. To this effect, Zappavigna introduces some essential functions and
Twitter-exclusive moves, e.g. hashtags, retweets and @mentions, to show how
social media users can exploit Twitter for their own interpersonal needs and indi-
vidual identity-management practices.
To illustrate her claims, Zappavigna consistently draws on a number of clear
examples which indicate that twitter users not only attempt to convey and negotiate
conceptual but also interpersonal meanings. She finishes her chapter by proposing
that such personal alignments may not necessarily take on explicit forms but can
equally be realized more implicitly. It is in such acts of “ambient affiliation” that
Zappavigna claims “individuals […] may engage in mass practices such as hash-
tagging in order to participate in particular kinds of ‘belonging’” (Zappavigna, Ch.
8, this volume).
The final chapter which concludes this second part of the handbook is Volker
Eisenlauer’s critical review of pragmatic research on social network sites, probing
the most prominent example of Facebook. Again, Eisenlauer begins his chapter
with a thorough description of social-network sites, narrowing down their cen-
tral technological, compositional and communicative characteristics. He convinc-
ingly couples the descriptive part of his chapter with an in-depth discussion of the
conceptual nature of social-network sites lodged between genre and medium. To
capture social-network sites as representatives of what he calls “meso-media”, he
uses the term “form of communication” which he adopts from Ermert (1979). He
additionally develops a pragmatic categorization scheme for social network sites
to carve out their essential technological and communicative differences. On this
basis, Eisenlauer can then draw up the findings of various recent pragmatic works
on social media that focus especially on the most influential social network site
today, i.e. Facebook. In the last section, Eisenlauer divides studies which investi-
gate the impact and role of context in Facebook communication from pragmatic
research examining the use of speech acts on Facebook. He describes, for instance,
some new forms of speech acts he identifies on Facebook pages, i.e. “creative
text actions”, which are largely controlled by users and “automated text actions”
(Eisenlauer, Ch. 9, this volume). They are mainly produced, sent and distributed
by the software system. As to their structure and purpose, Eisenlauer points out the
crucial ethical and interpersonal effects of using and re-using these action types on
Facebook, arguing for a critical hypertext analysis of social media communication
in general.

Authenticated
4.3. Social media and discourse
The third part of this handbook is concerned with the interactional rules and reg-
ularities which provide structure to the discourse in social media. It is this struc-
ture which users mutually orient to in order to maintain a sufficient degree of
mutual understanding and build up the cohesive texture of audio-visual discourse.
In social media, such a structure is not only created individually by deploying a
network of recurring cohesive devices on the computer screen, but also collabo-
ratively by many different users through the reciprocal use of additions, amend-
ments, quotes or comments.
One of these interactional arenas is the sequential structure of two-way commu-
nication in various social media settings, highlighting the “central organizational
principles of interaction” (Garcia and Jacobs 1999: 339). This is the topic of the
first chapter of this part of the handbook by Maximiliane Frobenius and Cornelia
Gerhardt. Within the broader research framework of conversation analysis, the
authors first establish the identification of the sequential patterns of online interac-
tion as an ethnomethodological practice (Garfinkel 1967; Sacks 1995). They then
address the fact that the technological and communicative exigencies of social
media impose on the written (text-based) organization of online interaction. Since
social media widely enable asynchronous types of communication, the temporal
immediacy of classic face-to-face conduct loses (at least some of) its regulative
interactional force. To account for this communicative shift, the authors track the
relevance of various degrees of (a)synchronicity on the turn-taking structure of
social media interaction. They draw our attention to recent studies which centre
on new turn-allocation strategies conducive to different online environments and
manage to identify a number of compensatory moves, paramount to social media
interactions, e.g. rhythmic content, bursts and breaks, the use of multi-unit texts,
quoting strategies, terms of address (vocative cues), etc.
Frobenius and Gerhardt continue to discuss pragmatic research on repairs,
openings and closings in online interactions, illustrating their findings and conclu-
sions with a large number of useful examples from tweets, chat, instant messages
and emails. Their overview includes useful new concepts from Tanskanen and
Karhukorpi’s “concessive repair” (2008), Greenfield and Subrahmanyam’s “slot-
filler” (2003) to Rintel, Pittam and Mulholland’s novel “automated joining event
(AJE)” (2003) and explains their significance for digital talk-in-interaction against
the background of classic research in conversation analysis and ethnomethodolog-
ical sociology. The chapter closes with a reflection on how interactions in spoken
social media may differ from their written counterparts.
Frobenius and Gerhardt relate to recent work by Sindoni (2014) or Licoppe
and Morel (2012), for instance, to address differences between visually-enhanced
digital telephone calls via skype and classic phone calls. Essentially, they cap-
ture the gradual formation of new interactional orders which emerge on social

Authenticated
media fuelled by the incessantly changing and malleable technological and com-
positional evolution of online communication. As Frobenius’ own work on open-
ing sequences in videoblogs convincingly shows (Frobenius 2014), the growing
diversity of interactional moves and strategies which social media fosters are just
as diversified as the ever-increasing technological and social granularity of online
communication today.
The next chapter by Elisabeth Fritz continues along the path set up by Frobe-
nius and Gerhardt. She focuses on the relationship between social media interac-
tion and topic management and explores this connection on two separate levels.
On a micro level, Fritz explores the individual topic development strategies which
social media users have adopted from face-to-face conversation to establish, nego-
tiate and change topics in online settings. On a macro level, she explains how in
social media, topics can drift away from initial posts, branching off into a myriad
of subordinate threads. Fritz reveals how high numbers of users in message boards
can potentially cloud their message’s relevance, reference or topical orientation in
regard to what has been posted before and afterwards. More specifically, she struc-
tures her review of pragmatic research on topic development around three “prom-
inent characteristics” (Fritz, Ch. 11, this volume) which can be said to impose on
topical coherence in social media, i.e. synchronicity, polyloguicity and monotopi-
cality. While Fritz first leads the reader through classic research on topic manage-
ment, the main part of her chapter discusses the various ways in which users have
been shown to apply compensatory strategies to secure cohesion and thus maintain
a reasonable degree of topical coherence.
The emphasis on topic connectivity, reference and coherence builds a suit-
able bridge to the ensuing chapter by Christoph Schubert. It focusses on means
of (hyper-)cohesion within and between social media sites. Although Schubert
begins his chapter with a conceptual introduction of key terms such as cohesion,
coherence and texture, uncovering their origins in the seminal work by Halliday
and Hasan (1976), he soon shifts his focus to more recent studies on cohesion
in computer-mediated environments. Considering the use of cohesive devices on
social media, he rightly acknowledges cohesion as a social phenomenon next to
its classic lexico-semantic as well as technological dimension. The phenomenon
at hand can thus equally be understood through the interpersonal bonds between
users which it establishes by its textual, semantic or digital links.
In a consecutive manner, Schubert explores the (hyper-)cohesive effects of
multi-linear browsing in online contexts and discusses the cohesive challenges
which typically result from interrupted adjacencies in asynchronic social media
interactions. He explains the semantic/semiotic load in different types of elec-
tronic (hyper-)links and discusses their respective cohesive effects on users’ own
forward-looking planning strategies (cf. Jucker 2005). He takes the discussion of
link types as a starting point to explore various navigation tools which are meant
to establish and maintain user orientation within and across hypertext pages

Authenticated
and within as well as between social media messages. Based on van Leeuwen’s
notion of “multimodal cohesion” (2005: 179), Schubert argues that cohesion on
social media is not only a verbal and technological but also a semiotic concept.
More recent audio-visual representations on YouTube, Twitter and Facebook, for
instance, allow for comprehensive sets of audio-visual, intermodal links which
frame individual communicative parts as contextualization cues. In the remainder
of his chapter, Schubert illustrates the actual use and texture of cohesion in four
different social media settings, i.e. YouTube, Facebook, Twitter and blogs, based
on previous pragmatic research. He successfully reveals that cohesion is not a
consistent feature of all online communication but that it is actually deployed in
new ways, used for contrastive purposes and thus plays a very important social,
interactional and discursive role on different social media platforms.
As Schubert illustrates, cohesion should not only be seen as a purely tex-
tual concept but necessitates and activates the mutual cognitive engagement of
social media users in order to become meaningful. Andreas Langlotz’s chapter
on cognition and discourse takes this insight as a stepping stone to explore the
socio-cognitive dimension of language use on social media. Langlotz organizes his
chapter into three main parts which revolve around (a) the cognitive skills respon-
sible for understanding social media discourse, (b) the cognitive processes which
secure and drive users’ on-line discourse comprehension and (c) social media’s
technological and communicative affordances which affect the sociocognitive
processes of online users. Using this three-fold structure as a basic orientation
scheme, Langlotz provides various sample analyses from Twitter to substantiate
and exemplify his claims and findings. The first section of his chapter centres on
introductory descriptions of the key terms of discourse and cognition. Drawing
extensively on recent work in cognitive studies, he discusses the role of different
cognitive skills in communication, e.g. perception, categorisation and conceptu-
alisation, making inferences and judgments. He also provides useful overviews
of research in cognitive linguistics and pragmatics, e.g. relevance theory, frame
theory, cognitive semantics, etc. All of these approaches have tried to reduce the
complexity of cognitive processing, accounting only for some of its parts or proce-
dures. In the following section, Langlotz sketches the parallel historical evolution
of pragmatic and cognitive research which has led to a “divide between social
practices and individual cognition” (Langlotz, Ch. 13, this volume). He contends
that this gap can be overcome by more recent socio-cognitive approaches which
underscore the inseparable tie between individual cognition and social communi-
cation. To this effect, he outlines Herbert Clark’s theory of language use (Clark
1996), explaining how the systematic and conventional employment of grounding
devices in Tweets (and Twitter Comments) builds up and maintains a consistent
web of mutual understanding between the interlocutors. In the final section of his
chapter, Langlotz shows how the technological and communicative affordances
of social media (in general) and Twitter (in particular) affect the socio-cognitive

Authenticated
activity of grounding. He thus lays out the socio-cognitive foundation of the cohe-
sive texture in social media introduced by Schubert in the previous chapter.
Stephen Pihlaja and Andreas Musolff’s chapter, which rounds off this third part
of the volume, touches upon the ideologies of social media, which surface through
the audio-visual representation or practice indicative of a particular set of social
beliefs or values. The authors provide detailed definitions of the central terms
of ideology and discourse. Focussing more on the discursive than the cognitive
aspects of ideology, Pihlaja and Musolff hold that “[t]he ‘sharing’ of socially eval-
uative beliefs and attitudes is itself the product, not the pre-condition of ideology
in discourse” (Ch. 14, this volume). Next to the ideological bias of specific social
media platforms, e.g. Facebook’s “commodification of interaction”, the authors
of this chapter also discuss ethical issues which are related to studying, selecting
and analysing social media in an academic context. The chapter exemplifies the
use of ideologies with a view to three separate social media platforms, i.e. Internet
discussion forums, blogs and video-streaming platforms. Pragmatic research on
the former has primarily scrutinized power asymmetries between the gatekeeping
forum managers and the regular users. Studies have shown what users actually
talk (or do not talk) about. Pihlaja and Musolff indicate the verbal strategies users
regularly apply to stage, hide, mitigate or boost their opinions, evaluations and
beliefs, positioning the individual user vis-à-vis their interlocutors’ type of action,
feelings and points-of-view. In contrast, recent studies on blogs have revealed that
the latter seek to “cultivate an insider-ethos” (Pihlaja and Musolff, Ch. 14, this
volume). The last part of this chapter addresses critical issues and controversies
concerning the study of ideology on social media platforms, such as anonymity,
free speech, multiple identities and “context collapse”. With their chapter, Pihlaja
and Musolff thus subtly shift our attention away from the organization, negotiation
and comprehension of discourse in social media to a new topic, which shall govern
the chapters of the next part of our volume, i.e. the construction, reproduction and
negotiation of identity in social media contexts.
4.4. Social media and identity

In the fourth part of this handbook, the audio-visual presentation, negotiation
and ratification of user identities takes centre stage. It not only covers pragmatic
research which overtly recognizes and classifies identity-related interactional
moves and patterns, e.g. facework, positioning theory, but also looks at the specific
means and strategies users apply when their identity is at stake, e.g. evaluation,
(im-)politeness, flaming, storytelling. In many ways, the study of online identity
has been gaining momentum in the last decade. It is thus not surprising to find that
this part of the handbook is the longest, boosting a confection of different subsid-
iary topic areas, all of which constitute important research objectives in Internet
pragmatics today.

Authenticated
In their first chapter, Locher and Bolander survey the rich collection of prag-
matic studies devoted to the change of users’ individual or group-related identities
on social media. Their chapter shows the broad range of theoretical and methodi-
cal approaches to identity- and facework in digital environments. Narrowing down
their conceptual roots, the first section retraces the historical evolution of identity
and face, following up pragmatic research from the early 1970s to the present
day. Identity can be viewed as a stable, de-contextualized social variable or as
a transient, discursive, social phenomenon whose function and meaning is often
tightly interwoven with the situational and cultural contexts in which it emerges. It
is arguably the latter dynamic and interactional definition of identity which seems
particularly amenable to the new communicative affordances of social media.
Locher and Bolander proceed to pin down current pragmatic identity (and face)
research, focussing on the most pertinent research themes, i.e. (im-)politeness,
gender-specific language use as well as expertise, authenticity and trust. While
reviewing and classifying central studies and their findings in all of these three
subfields of identity research, the authors repeatedly acknowledge that all of them
are highly interrelated. To illustrate, they substantiate some of the crucial results of
the studies they have presented, using a range of sample excerpts from Facebook
and Twitter.
In the following chapter, Michele Zappavigna shifts our attention from the
general description of how users construct, change and negotiate their online iden-
tities to the semiotic toolkit which sets this process into motion. More specifically,
she explores how linguistic work on appraisal or evaluation can play a vital part
in revealing how users present and manage their identities and how they ascribe
face or identity to others. Zappavigna retraces the historical development of lin-
guistic evaluation theories (Hunston and Thompson 2000; Martin and White 2005;
Bednarek 2006, 2008) and recapitulates the conceptual groundwork of Martin and
White’s “appraisal theory”. She takes us through the various systemic classifica-
tions, illustrating the categorical networks and choices with the help of various
original tweets. In so doing, she manages to link up her conceptual introduction to
a specific theory of evaluation to a successive description and discussion of cur-
rent pragmatic research on user’s concrete evaluation practices on social media.
In a number of consecutive steps, she moves from Knight’s study of evaluative
“couplings” (Knight 2010: 219) to more recent studies on the evaluative nature of
“social tagging”, closing her chapter with the novel concept of “ambient affilia-
tion” (Zappavigna 2011).
If Locher and Bolander already hinted at the fundamental importance of (im-)
politeness in the co-construction of users’ joint online identities, Sage L. Graham
provides a more substantial overview of pragmatic research on said topic in her
chapter. It starts with a number of preliminary reflections on how the technological
and communicative affordances of social media may impose on users’ (im-)polite
behaviour. Closing in on all ‘digital factors’ which make social media interac-

Authenticated
tion feasible, she creates a conceptual background for the following description
of (im-)politeness theories. Her summary of influential (im-)politeness theories is
often interspersed with reflections on how these may be applied to social media
contexts, for instance concerning the difference between appropriate and polite
user actions or how guidelines or FAQs may frame and restrict such behaviour on
a meta-communicative, normative level.
Graham then continues with a number of separate sections devoted to short
overviews of current research trends in different types of social media, e.g. social-
network sites, blogs, message boards. The final section of her chapter applies the
most essential insights she has imparted to a range of sample analyses of the pop-
ular social media site “Twitch”. She reveals the explicit and implicit codes of
conduct which can be retrieved from the site and examines the ways in which vio-
lations of these regulations are sanctioned. In addition, we learn how moderators
use a range of manual and automatic strategies to remove potentially face-aggra-
vating content from their sites and find out about the technical and communica-
tive difficulties such deletions may imply. Put briefly, Graham shows that (im-)
politeness theories are, and continue to be, challenged by pre-programmed impo-
liteness detection software, e.g. BotMods, which need to be trained to identify
the (im)polite nature of contributions and must be able to spot the communicative
context in which specific impolite contributions are likely to emerge. Naturally,
contemporary programs are neither fully capable of fulfilling these tasks, nor are
our theoretical insights into the social nature of (im-)polite behaviour substantial
enough. Graham fittingly ends her chapter with an outlook on future directions in
(im-)politeness research on social media.
In line with Graham’s account of (im-)politeness on social media, Claire
Hardaker’s next chapter elaborates on the much debated phenomena of flam-
ing and trolling, which, as she claims, are “particularly native to social media”
(Ch. 18, this volume). Hardaker offers a critical view on various definitions of
flaming and trolling which have come into usage in pragmatics during the last two
decades. Her conceptual journey into the etymological and theoretical origins of
both terms is informed by an examination of the elementary communicative char-
acteristics of digital communication, e.g. anonymity, pseudonymity, disclosure,
etc. It is believed that these framing conditions of social media interactions do not
only influence users’ rights and obligations, expectations and interactional moves
but also their proclivity to engage in online abuse. The main part of Hardaker’s
chapter describes the central descriptive (and etymological) contours of flaming
and trolling, accompanied by systematic discussions of both concepts in regard to
two exemplary studies. Her discussion of methodological reflections as well as
qualitative research findings does not only shed light on how pragmatic analyses
of flaming and trolling can be conducted but also encompass larger sections deal-
ing with current methodical and interpretative challenges in the field. The chapter
ends with a brief look at future directions in related pragmatic research.

Authenticated
Given the plethora of definitions and approaches to the study of narratives in

pragmatics and discourse analysis, Ruth Page wisely begins chapter 19 with a his-
torical overview of key definitions in the field. She sketches the conceptual change
of narrative definitions from their early beginnings in structuralism and sociolin-
guistics (Labov and Waletzky 1967) to more current, poststructuralist models of
narrativity (Ochs and Capps 2001), which stress the gradual, transient and coop-
erative nature of storytelling. Such an approach lends itself to an elaborate dis-
cussion of the ways in which stories are told in computer-mediated environments,
more specifically how these contexts (and their socio-communicative exigencies)
enhance or constrain the creation and understanding of stories online. With a view
to social media, at least three main aspects come to mind: (a) the option to tell and
categorize stories across time and space (in different posts, across posts and com-
ments, across social media sites, by cross-referencing, linking, tagging stories),
(b) the opportunity to tell stories across different modes of expression (verbal text,
images, photos, videos, etc.), (c) the capacity to engage in collaborative, interac-
tive storytelling, including multiple online users. Page gives due credit to all of
these aspects and more. She discloses the potential “episodic nature” of storytell-
ing on Facebook, distinguishes users’ “transportable” from their local “situated
identities” and exposes the particular “technological formats” users routinely draw
on to stage identities by telling a story online (Ch. 19, this volume). The final
section of this chapter then reviews current trends in narrative research in Internet
pragmatics.
All of the previous chapters in this fourth part of the handbook have either
dealt with identity (or identity-related concepts) directly (see for instance Locher
and Bolander or Graham) or have focussed on particular linguistic strategies which
are commonly used in online identity work (Zappavigna; Hardaker; Page). Monika
Bednarek’s last chapter arguably takes a somewhat different approach to the sub-
ject of identity. In contrast to the preceding chapters, she draws our attention to a
particular group of online users, i.e. fans.
As Bednarek shows, fandom (in what fans represent and what they do) seems
to be particularly visible on social media. Here, fans increasingly meet, commu-
nicate and, either explicitly or implicitly, forge their communal identity through
recurrent exchanges. In many ways, it appears that social media are one of the
main driving forces in the global growth, organization and orientation of fans
today, and Bednarek makes a case for studying the pragmatic dimension of what
they say, do and achieve on Facebook, Twitter or message boards. At the begin-
ning of her chapter, she acknowledges that the concept of ‘fan’ is inherently fuzzy
and in dire need of further intensional and extensional specification. Offering a
number of decisive steps into this direction, she then informs us how fans, in the
last two decades, have increasingly flocked to social media platforms to exchange
their shared affection for television series, film stars or other celebrities. Bednarek
sheds light on the various ways in which social media have changed the practices

Authenticated
of fandom in recent times, which involves inter alia the new electronic options to
search for like-minded aficionados, the international reach of social media as well
as the quick and easy ways to get in touch with other fans.
The main part of the chapter is devoted to studies which have focussed on
fan audience responses to various different television series. Bednarek primarily
addresses two different aspects of audience behaviour which are arguably con-
stitutive of fan behaviour on social media: (a) the types of linguistic actions fans
perform and (b) the stance fans express when talking about their idol or object of
fascination. With a view to the former, Bednarek discusses the possible range and
limits of fan-based action types. She reviews previous studies in narrative and
discourse analysis as well as sociolinguistics, which have attempted to come up
with suitable classification models (e.g. Gregoriou 2012; Androutsopoulos and
Weidenhöffer 2015). In the remainder of her chapter, we learn about the stances
of fans we may connect with individual action types. Bednarek explains that fan
culture on social media is increasingly a global and transnational phenomenon
(Ch. 20, this volume), which enables her to talk of different “styles of fandom”.
Such styles can find expression in the number and type of fan actions but equally
in inscribed or evoked evaluations.
In the end, Bednarek’s chapter could likely be seen as an exemplary applica-
tion of the central insights established in the preceding chapters, since it clearly
illustrates how the identity of a particular group of users (fans) can emerge through
the interplay of self- and other presentation (facework) and the use of audio-visual
strategies of evaluation, storytelling and (im-)politeness.
4.5. Social media and functions

The final part of our handbook picks up the issue of important speech acts and their
use and function on different social media platforms. Naturally, the scope of this
handbook has forced us to constrain the types of speech acts we can consider, and
we have done so in view of two main factors: Our selection of speech acts needed
to be conducive to social media either in terms of their frequency and functional
salience (“Getting ‘liked’”, “Conflictual and consensual disagreement”) or in light
of the amount of pragmatic research devoted to their study in both offline and
online contexts (“Compliments and compliment responses”, “Requests and advice
giving”). We hope that this part of the handbook provides some general orienta-
tion on where these well-known pragmatic functions appear on social media and
explain who uses them, in which numbers and to which end. We thereby know-
ingly continue and extend a growing body on pragmatic research which reaches
beyond pragmatic studies in the digital realm and covers all types of human com-
munication – spoken and written, offline and online, in past and present.
Carmen Maíz-Arévalo’s chapter on “getting ‘liked’” on social media is nat-
urally based on the premise that human beings strive toward becoming socially

Authenticated
accepted and valued members of social groups and play a meaningful part in larger
societies and organizations. Clearly, this view perfectly aligns with Brown and
Levinson’s idea of a positive face every human member claims for themselves
and it equally reverberates in Goffman’s idea of the ways and strategies human
beings choose to assume, reject or ratify lines of identity in ongoing discourse. We
can see how the function of liking and getting liked plays a central role on social
media, whose interactions are by and large oriented to the establishment, main-
tenance or negotiation of users’ social, interpersonal bonds. Moving beyond this
commonplace observation, Maíz-Arévalo dissects the various economic, narcis-
sistic or other strategic reasons which may trigger acts of liking on different social
media platforms. She gives structure to these various usages, reducing the multi-
farious motivations for liking individuals online to four essential “macro-strate-
gies”: self-presentation (aka self-disclosure), social expressivity, implicitness and
humour.
The first of these strategies is introduced with the help of ample examples
and described along four central parameters which define its identity effect on
users. These are (a) the quantity of published (personal) information, (b) the level
of intimacy or privacy of information, (c) the degree of ‘positive’ embellishment
in the way personal information is presented and (d) the level of staged (or per-
ceived) authenticity of information. As the author shows, the second macro-strat-
egy of “social expressivity” is not concerned with self-presentation but with user
activities which cater to the needs and feelings of others, expressing empathy,
compassion and understanding. The third strategy of “implicitness” addresses the
fact that often the (co-)construction of user identities on social media does not rest
on users` explicit verbal maneuvers but on their implicit, more or less accidental
online behavior. The final strategy covers the large field of humorous actions on
social media which mainly contribute to the building and nurturing of friendships
and even larger social networks. Maíz-Arévalo carefully describes each of these
parameters in much detail intermittently pointing to relevant pragmatic research
which has explored the relationship between each of these strategies and the func-
tion of liking in CMC. Her clear and nuanced descriptions as well as her balanced
look at different pragmatic research traditions and findings make Maíz-Arévalo’s
paper an excellent gateway to the chapters to come.
Bolander and Locher’s chapter is devoted to the speech act of disagreeing in
social media contexts, and they account for existing pragmatic research on this topic
from a variety of angles. Their exploration of this fundamental function in social
media begins with the consideration of socio-technological factors which give rise
to either conflictual or consensual types of disagreements in online environments.
This background conceptually frames the ensuing comprehensive overview of
pragmatic literature on both of these subtypes of disagreement in social media.
Three major lines of research, their respective objectives and foci are discussed,
i.e. comparative studies of disagreements in online and offline environments, the

Authenticated
relationship between social and medium factors, the link between disagreements,
the gender-specific use of disagreements, disagreements in dialogic and polylogic
interaction, disagreements in educational/pedagogical contexts and the contrastive
analysis of disagreements across varieties, languages and cultures. While each of
these emergent research traditions merits closer inspection, Bolander and Locher
especially focus on two key issues which have surfaced in pragmatic research
on disagreement in the last two decades, i.e. the context and contextualization of
disagreements as well as the indexing of emotions. The research focus on the con-
textual configurations into which disagreements enter involves the use of various
semiotic modes (text, pictures, film, etc.) from which users can draw to express
themselves, but also different “lines” of communication, i.e. offline vs. online con-
texts (Bolander and Locher, Ch. 22, this volume). The various modal and medial
parameters that constitute the communicative interface through which disagree-
ments are voiced yield specialized “contexts of interaction” (Angouri and Tse-
liga 2010: 66) which affect the overall use as well as interactional placement and
negotiation of disagreements. A different research strand, which has seen much
attention in the last years, concerns the relation between the quantity and quality
of disagreements in social media, their audio-visual display of emotions and the
relational work to which they contribute.
The topic of compliments (and compliment responses) has been at the forefront
of pragmatic research over the last fifty years, from Manes and Wolfson’s semi-
nal work on structural compliment prototypes (1981) and Holmes’ classic stud-
ies on the gender-specific use of particular types of compliments (Holmes 1986,
1988) to more recent work by Jucker and Taavitsainen 2008, Maíz-Arévalo and
García-Gómez 2013 or Placencia and Lower 2013. Compliments have success-
fully retained their popularity as one of the most debated research topics in offline
settings in the age of new media. The chapter by Placencia and Lower in this vol-
ume confirms this success story, critically presenting a growing body of pragmatic
research on the subject in a principled fashion. Besides providing definitions of
compliment types, forms and functions, the authors compile and evaluate existing
research, stressing similarities and differences of the given findings. They also
discuss the various methods of analysis with which scholars have tried to capture,
count, assess and interpret compliments and compliment responses in online set-
tings. Thereby, the useful classification schemes and tables we find in this chapter
provide rich overviews of pragmatic studies, categorized according to the particu-
lar language variety, the social media platform as well as the number and types of
compliments they investigate. The typical features of compliment behaviour are
introduced and discussed in appropriate detail, including compliments’ structural
make-up, topics and functions. The chapter then draws our attention to compli-
ment responses on social media, again accounting for their quantitative, structural
and functional differences in different online settings. The end of their chapter sees
Placencia and Lower reflect on some of the more pressing ethical issues as well

Authenticated
as methodological pitfalls which linguists can expect to encounter when studying

compliments in CMC.
Philipp Morrow’s final chapter of this handbook takes an insight look at the
speech acts of requesting and advice-giving in social media. Morrow sets out to
survey concurrent pragmatic research, including email requests in academic and
other professional settings, requests on social network sites as well as health-re-
lated websites. Comparing and contrasting the results of the different studies in
their different CMC settings, he also looks at research aimed at advice-giving in
various CMC contexts. In this vein, he touches upon issues of pragmatic com-
petence and cultural appropriateness, (in-)directness and (im-)politeness, social
power and interpersonal rapport. His critical review allows us to catch a glimpse
of the formal and functional variability of the two speech acts at hand. We find out
that the way users shape and perform their requests or ask for advice depends on
multifarious contextual details, such as our cultural background, our specific moti-
vation and objective, our interpersonal rapport with the addressee(s), the relative
social power divide between the interlocutors, the socio-technological affordances
of individual social media platforms, the multimodal options for expressing our
views, etc. Morrow’s chapter gives due credit to these different practices, and sum-
marizes and appraises quantitative and qualitative findings in a diligent fashion. It
completes the final part of this handbook, closing off a section which has compiled
and revisited pragmatic research into the most ubiquitous and elementary speech
acts on social media today.
5. Concluding remarks
We hope that this handbook provides structure and coherence to a still young and
growing field of research which – as we strongly believe – is not only trend-
ing but about to take full shape in pragmatics in the years to come. Its twen-
ty-four individual chapters survey what we know about the pragmatics of social
media at the present time. They provide readers with systematic insights into the
socio-communicative nature of social media and describe striking differences in
the way we talk or write about, read, share and negotiate meanings on different
social media platforms. They reveal how we establish, keep, change and challenge
different types of identities as social media users and shed light on how and when
we tend to use and apply particular speech acts on Facebook, Twitter or YouTube.
The different parts, this handbook consists of, are meant to give scholars easy
access to comprehensive overviews on pragmatic research, based on their indi-
vidual conceptual, topical or functional interests. There are, necessarily, a wealth
of conceptual and topical interrelations both between the chapters and between
the parts, and this is not by chance but by design. The cross-references enrich our
understanding of social media communication and invite readers to read ‘across’,

Authenticated
exploring the pragmatics of social media platforms as well as new research topics
and approaches to the study of digital communication. To this effect, we under-
stand that this handbook may not only be useful to linguists but to scholars in other
scientific disciplines as well.
References
Androutsopoulos, Jannis
2008 Potentials and limitations of discourse-centred online ethnography. Language@
internet 5.8.
Androutsopoulos, Jannis and Jessica Weidenhöffer
2015 Zuschauer-Engagement auf Twitter: Handlungskategorien der rezeptions
begleitenden Kommunikation am Beispiel von #tatort. Zeitschrift für ange-
wandte Linguistik 62(1): 23–59.
Angouri, Jo and Theodora Tseliga
2010 “you HAVE NO IDEA WHAT YOU ARE TALKING ABOUT!” From e-dis
agreement to e-impoliteness in two online fora. Journal of Politeness Research
6(1): 57–82.
Arendholz, Jenny and Monika Kirner-Ludwig
2015 In-between cognitively isolated quotes and references: Looking for answers
lurking in textual margins. In: Jenny Arendholz, Wolfram Bublitz and Monika
Kirner-Ludwig (eds.), The Pragmatics of Quoting Now and Then, 319–342.
Berlin/Boston: de Gruyter.
Barnes, Susan B.
2006 A privacy paradox: Social networking in the Unites States. First Monday 11
(9). http://firstmonday.org/article/view/1394/1312 (last accessed 15 February
2017).
Bednarek, Monika
2006 Evaluation in Media Discourse: Analysis of a Newspaper Corpus. London:
Continuum.
Bednarek, Monika
2008 Emotion Talk Across Corpora. London: Palgrave Macmillan.
Bell, Allan
1984 Language style as audience design. Language in Society 13: 145–204.
Brown, Penelope and Stephen C. Levinson
1987 Politeness: Some Universals in Language Usage. Cambridge: Cambridge Uni-
versity Press.
Bruns, Axel
2007 Produsage: Toward a Broader Framework of User-Led Content Creation. Kel-
vin Grove, Queensland: Queensland University of Technology.
Clark, Herbert
1996 Using Language. Cambridge: Cambridge University Press.
Clark, Herbert and Edward Schaefer
1989 Contributing to discourse. Cognitive Science 13: 259–294.

Authenticated
Chovanec, Jan and Marta Dynel

2015 Researching interactional forms and participant structures in public and social
media. In: Marta Dynel and Jan Chovanec (eds.), Participation in Public and
Social Media Interactions, 1–23. Amsterdam/Philadelphia: Benjamins.
Dürscheid, Christa
2007 Private, nicht-öffentliche und öffentliche Kommunikation im Internet. Neue
Beiträge zur Germanistik 6(4): 22–41.
Dynel, Marta
2014 Participation framework underlying YouTube interaction. Journal of Pragmat-
ics 73: 37–52.
Eisenlauer, Volker
2014 Facebook as a third author – (Semi)automated participation framework in
Social Network Sites. Journal of Pragmatics 72: 73–85.
Ermert, Karl
1979 Briefsorten. Untersuchungen zu Theorie und Empirie der Textklassifikation.
Tübingen: Niemeyer.
Frobenius, Maximiliane
2014 The pragmatics of monologue: Interaction in video blogs (Doctoral disserta-
tion). http://scidok.sulb.uni-saarland.de/volltexte/2014/5895/pdf/dissertation_
frobenius_scidok.pdf (last accessed 15 February 2017).
Garcia, Angela C. and Jennifer B. Jacobs
1999 The eyes of the beholder: Understanding the turn-taking system in quasi-
synchronous computer-mediated communication. Research on Language and
Social Interaction 32: 337–367. doi: 10.1207/S15327973rls3204_2.
Garfinkel, Harold
1967 Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.
Greenfield, Patricia and Kaveri Subrahmanyam
2003 Online discourse in a teen chatroom: New codes and new modes of coher-
ence in a visual medium. Journal of Applied Developmental Psychology 24(6):
713–738. doi: 10.1016/j.appdev.2003.09.005.
Gregoriou, Christiana
2012 ‘Times like these, I wish there was a real Dexter’: Unpacking serial murder
ideologies and metaphors from TV’s Dexter internet forum. Language and
Literature 21: 274–285.
Halliday, Michael A. K. and Ruqaiya Hasan
1976 Cohesion in English. London: Longman.
Herring, Susan, Dieter Stein and Tuija Virtanen (eds.)
2013 Pragmatics of Computer-Mediated Communication. (Handbooks of Pragmat-
ics 9.) Berlin/Boston: de Gruyter.
Hoffmann, Christian R.
2012 Meaning and Interaction in Personal Weblogs. Amsterdam/Philadelphia: Ben-
jamins.
Holmes, Janet
1988 Paying compliments: A sex-preferential politeness strategy. Journal of Prag-
matics 12: 445–465.
Holmes, Janet
1986 Compliments and compliment responses in New Zealand English Anthropo-
logical Linguistics 28: 485–508.

Authenticated
Huang, Yan
2007 Pragmatics. Oxford: Oxford University Press.
Hunston, Susan and Geoffrey Thompson
2000 Evaluation in Text: Authorial Stance and the Construction of Discourse.
Oxford: Oxford University Press.
Jenkins, Henry
2006 Convergence Culture: Where Old and New Media Collide. New York: New
York University.
Jucker, Andreas H. and Irma Taavitsainen (eds.)
2008 Speech Acts in the History of English. Amsterdam/Philadelphia: Benjamins.
Jucker, Andreas H.
2005 Hypertext research: Some basic concepts. In: Lilo Moessner and Christa
M. Schmidt (eds.), Anglistentag 2004 Aachen: Proceedings, 285–295. Trier:
Wissenschaftlicher Verlag Trier.
Koch, Peter and Wulf Oesterreicher
1985 Sprache der Nähe – Sprache der Distanz. Mündlichkeit und Schriftlichkeit
im Spannungsfeld von Sprachtheorie und Sprachgeschichte. Romanistisches
Jahrbuch 36: 15–43.
Knight, Naomi
2010 Wrinkling complexity: Concepts of identity and affiliation in humour. In:
Monika Bednarek and James R. Martin (eds.), New Discourse on Language:
Functional Perspectives on Multimodality, Identity, and Affiliation, 35–58.
London/New York: Continuum.
Labov, William and Joshua Waletzky
1967 Narrative analysis. In: June Helm (ed.), Essays on the Verbal and Visual Arts,
12–44. Seattle: University of Washington Press.
Landert, Daniela and Andreas H. Jucker
2011 Private and public in mass media communication: From letters to the editor to
online commentaries. Journal of Pragmatics 43(5): 1422–1434.
Licoppe, Christian and Julien Morel
2012 Video-in-interaction: “Talking heads” and the multimodal organization of
mobile and Skype video calls. Research on Language and Social Interaction
45(4): 399–429.
Maíz-Arévalo, Carmen and Antonio García-Gómez
2013 “You look terrific!” Social evaluation and relationships in online compliments.
Discourse Studies 15: 735–760.
Mandiberg, Michael
2012 Introduction. In: Michael Mandiberg (ed.), The Social Media Reader, 1–12.
New York: New York University Press.
Manes, Joan and Nessa Wolfson
1981 The compliment formula. In: Florian Coulmas (ed.), Conversational Routine:
Explorations in Standardised Communication Situations and Prepatterned
Speech, 115–132. The Hague: Mouton.
Martin, James R. and Peter White
2005 The Language of Evaluation: Appraisal in English. New York: Palgrave Mac-
millan.

Authenticated
Ochs, Lisa and Elinor Capps

2001 Living Narrative. Cambridge: Cambridge University Press.
O’Reilly, Tim
2010 What is Web 2.0? Design patterns and business models for the next generation
of software. In: Helen Donelan, Karen Kear and Magnus Ramage (eds.), Online
Communication and Collaboration. A Reader, 225–236. London: Routledge.
Page, Ruth E.
2013 Stories and Social Media: Identities and Interaction. London: Routledge.
Page, Ruth E., David Barton, Johann W. Unger and Michele Zappavigna (eds.)
2014 Researching Language and Social Media: A Student Guide. London: Routledge.
Papacharissi, Zizi
2010 A Private Sphere. Democracy in a Digital Age. Cambridge/Malden: Polity
Press.
Placencia, María Elena and Amanda Lower
2013 Your kids are stinking cute. Complimenting behavior on Facebook among fam-
ily and friends. Intercultural Pragmatics 10: 617–646.
Quan-Haase, Anabel and Luke Sloan
2017 Introduction to the Handbook of Social Media Research Methods: Goals, chal-
lenges and innovations. In: Luke Sloan and Anabel Quan-Haase (eds.), The
Sage Handbook of Social Media Research Methods, 1–11. London/Thousand
Oaks: Sage.
Rheingold, Howard
2008 Using participatory media and public voice to encourage civic engagement. In:
W. Lance Bennett (ed.), Civic Life Online: Learning How Digital Media Can
Engage Youth, 97–118. Cambridge, MA: MIT Press.
Rintel, Sean E., Jeffrey Pittam and Joan Mulholland
2003 Time will tell: Ambiguous non-responses on Internet Relay Chat. Electronic
Journal of Communication 13(1). http://www.cios.org/EJCPUBLIC/013/1/
01312.HTML (last accessed: 15 Oct. 2016).
Sacks, Harvey
1995 Lectures on Conversation. Ed. by Gail Jefferson. Oxford: Blackwell.
Seargeant, Philip and Caroline Tagg (eds.)
2014 Language and Social Media: Communication and Community Online. Lon-
don: Palgrave.
Sindoni, Maria Grazia
2014 Through the looking glass: A social semiotic and linguistic perspective on the
study of video chats. Text and Talk 34(3): 325–347.
Tagg, Caroline
2015 Exploring Digital Communication: Language in Action. London: Routledge.
Tanskanen, Sanna-Kaisa and Johanna Karhukorpi
2008 Concessive repair and negotiation of affiliation in e-mail discourse. Journal of
Pragmatics 40: 1587–1600.
Van Leeuwen, Theo
2005 Introducing Social Semiotics. London: Routledge.
Virtanen, Tuija
2013 Mock performatives in online discussion boards: Toward a discourse-prag-
matic model of computer-mediated communication. In: Deborah Tannen and

Authenticated
Anna Marie Trester (eds.), Discourse 2.0: Language and New Media, 155–
166. Washington, DC: Georgetown University Press.
Zappavigna, Michele
2011 Ambient affiliation: A linguistic perspective on Twitter. New Media and Soci-
ety 13(5): 788–806.
Zappavigna, Michele
2012 Discourse of Twitter and Social Media: How We Use Language to Create Affil-
iation on the Web. London: Bloomsbury.

Authenticated
2. Participation as user involvement
Daniela Landert
Abstract: The aim of this chapter is to explore the potential of social media for
user participation from different perspectives. This includes technical factors that
determine how much participation is possible on a given platform as well as the
degree and forms of participation that can actually be observed. The chapter also
addresses the relation between user involvement that results from interaction and
other involvement strategies that can be found on social media, such as the pres-
ence of personal content and language of immediacy. It concludes with a case
study that illustrates how different involvement strategies are combined in politi-
cal communication on a social networking site.
1. Introduction: Interaction, participation and involvement
The potential for users to participate in interaction and to contribute content is per-
haps the main defining characteristic of social media (see also Hoffmann, Ch. 1,
this volume). It presents a contrast to traditional mass media communication, in
which communication is typically unidirectional from a professional text producer
to a large anonymous mass audience. On social media, anyone is able to partici-
pate, to exchange messages and to interact with any number of other users. This
chapter explores this potential from different perspectives, focussing on how it
relates to user involvement.
Three concepts are of central relevance when looking at participation as user
involvement, namely interaction, (social) participation and involvement. There is
some degree of overlap between these terms and all of them are used in various ways
in different research traditions. Interaction, as understood in this chapter, refers to
the exchange of messages between participants. In its most basic form, it consists
of A sending a message to B and B being able to react to this message in a way that
can be perceived by A. In other words, communication needs to be (at least poten-
tially) bidirectional in order to qualify as interaction. Following this definition, a
unidirectional communicative framework such as prototypical mass media com-
munication without any opportunities for the addressees to respond to the message
would not qualify as interaction. When looking at online settings it makes sense to
distinguish further between the potential of interaction – i.e. whether the addressee
has the option of responding to a message – and interactions in which this potential
is actually realised – i.e. the addressee responds (see also Section 2 on forms and
degrees of interaction and Section 3.2 on types of interactivity). From the point of
DOI 10.1515/9783110431070-002
Authenticated
32 Daniela Landert
view of pragmatics, interaction is the most accessible of the three concepts, since it
can be studied directly on the basis of observable communicative exchanges.
In some instances, the term participation is used more or less synonymously
to interaction in order to refer to the communicative activities of participants in a
given communicative situation (see also Dynel, Ch. 3, this volume). Social partici-
pation goes beyond mere interaction, though. It involves a certain degree of power,
which means that participants not only have the opportunity to exchange mes-
sages, but that their messages also have an effect or, more precisely, that partici-
pants have influence on social organisation and social processes (Carah and Louw
2015: 231; Stein 2013). This is the reason why increased access to the Internet in
general and social media in particular cannot be equated with democratisation. As
Tredinnick (2008: 124) points out, new forms of communication technology tend
to become subject to the same power relations that used to control older forms. On
a very basic level, online access is not available to everyone. In addition, access
to social media is sufficient for sharing messages with a large public, but whether
these messages are read and whether authors are able to exert influence through
them is far from guaranteed. For pragmaticists, social participation is more diffi-
cult to investigate than interaction. The social effects of individual communicative
exchanges are in most cases not visible immediately. As a consequence, it is hard
to know to what extent specific messages allow their authors to participate in a
given social context in a meaningful way.
Involvement, finally, means that individuals engage with content, typically in
a way that affects them emotionally. Again, uses of the term vary and sometimes
involvement is used more or less synonymously with participation and interaction,
referring to situations in which participants interact actively. However, involve-
ment can also be used more specifically to refer to internal states and emotional
engagement of participants, as well as to characteristics of texts that are associated
with emotional engagement (for a critical discussion of these different uses of the
term, see Caffi and Janney 1994). From a linguistic point of view, the problem
with involvement is that we do not have any direct access to the emotional state
of interactants. If and how readers are affected by the messages they read is a
question that cannot be answered based on the message alone. As a consequence,
linguistic studies of involvement tend to focus on characteristics of texts that
are associated with involvement, such as involving content and involving and/or
involved language. To give an example, involving content may deal with personal
stories and emotions with which the addressee is expected to empathise. Involved
language includes, for instance, emotionally charged vocabulary, expressions that
explicitly refer to the sender and addressee (e.g. first and second person pronouns),
expressions of emphasis such as exclamation marks and capitalisations, the use of
emoticons, and so on (see also Section 4 below).
Interaction, (social) participation and involvement can and do co-occur. For
example, we can picture a situation in which users interact with each other on

Authenticated
Participation as user involvement 33
social media in a personally and emotionally involved way on a political topic that
affects them directly. If such interactions gain sufficient momentum they can con-
tribute to the formation of a larger social movement with the power to affect real
social change. The Arab Spring is often presented in such a way, although views
differ as to how large a role social media actually played for the political move-
ment (see, for instance, Papacharissi and Blasiola 2016; Thurlow 2013: 237). How-
ever, this co-occurrence of involvement strategies is by no means a given in social
media. Interactions on social media are not always tied to emotional involvement
and cases in which they form part of the type of participation that leads to more
substantial social change are the exception rather than the rule. Moreover, even
interaction in the narrow sense, i.e. with communication taking place in more than
one direction, cannot be taken for granted on social media. While this is certainly
the prototypical case, there are also instances in which the potential of interaction
is not realised. Far from ‘going viral’, most messages posted by private individuals
reach only a small audience and not all of them even receive a response.
The relation between social media, new forms of interaction and participation
needs careful consideration at various levels. These levels include whether there is
technical potential of interaction on a given platform, to what extent users are free
to shape their own interaction with others, whether or not users realise the potential
of interaction, what content users publish and what effect this has, i.e. whether the
content reaches an audience and whether it is able to lead to change. Not all these
aspects are equally central to a pragmatic study of social media. Especially the last
question concerning the wider social effects of social media participation is one
that is more suitably dealt with in the context of social and political sciences. The
remaining aspects will be discussed in Sections 2 and 3 of this chapter. Section 4
will deal with strategies of creating audience involvement through content and lan-
guage, and with how these strategies are related to involvement strategies that are
based on interaction. The connection between these different strategies is further
illustrated with a case study on the use of social media in political communication
in Section 5.
2. Forms of interaction and participation
2.1. The technical potential of interaction

Online interaction can take many different forms. One way of classifying these
forms is to order them according to the degree to which user contributions are
pre-shaped; or, put differently, the degree to which they allow users to contribute
their own content. This classification is technical in the sense that it depends on the
infrastructure of the social media platform. At one end of the scale, we find forms
of interaction in which users do not contribute any content at all. An example of

Authenticated
34 Daniela Landert
this is the tracking of user behaviour on websites, which can be used for arranging
content according to what is most popular or liked. The contribution by users is
very small in this case; in the case of most read articles on online news sites, for
instance, users might not even be aware that their behaviour has an impact on
how the content of a website is presented. Still, this immediate technical feedback
is more interactive compared to offline media like books or print newspapers,
where technical feedback about reader behaviour is restricted to overall circulation
or sales figures. At the next level we find interaction of the type of (dis-)liking
content, where users consciously submit their feedback by clicking pre-defined
buttons. In this case, there is a clearer intention to contribute to the presentation
of content (e.g. by increasing the visibility of certain items), but the actual con-
tent is still not changed. User polls are slightly more interactive. They present a
way of collecting new content, but this content is aggregated over all participating
users and participation is restricted to selecting among pre-defined options. User
comments are perhaps the most basic form of individual user-generated content.
Depending on the platform, there may be restrictions with respect to length and
type of content, but the text can be formulated freely and users are free to choose
the topic of their contribution, at least as long as they do not violate the guidelines
of the platform. Even freer forms of participation can still be integrated on insti-
tutional platforms, for instance in the form of blog sections where users can freely
contribute their own content. Social media platforms like Facebook, Twitter or
YouTube, offer a range of different options for users to publish their own content,
for instance in the form of profiles, status updates, messages and media uploads.
Finally, at the opposite end of the scale, users are free to set up their own online
spaces and publish content independent of any restrictions of existing platforms.
Closely related to the purely technical dimension of participation is the ques-
tion of user policy and regulations concerning the control over content. All major
social media platforms have user policy documents that specify the degree and
form of acceptable user participation. Violations of such policies can result in
deletion of user content, blocking of user access and, in extreme cases, can even
lead to legal action. While some common standards have developed over the last
two decades, there are still marked differences across platforms in how acceptable
content is defined and monitored (Buni and Chemaly 2016). In addition, Stein
(2013) shows how policy documents grant users different degrees of control over
their content depending on platform. Comparing YouTube, Facebook and Wiki-
pedia, Stein finds that Wikipedia is the only platform of the three in which users
have dominant control over the platform’s content and over user data.1 In addition,
1
However, this does not mean that every user succeeds in adding the content they want.
Wikipedia has a user-based editorial system that monitors content quite closely, at
least in some cases. Johnstone (2011) compares the representation of different types of

Authenticated
policies are decided “through a process of discussion-based consensus” among the

users of Wikipedia (Stein 2013: 367). In contrast, YouTube and Facebook grant
their users very little influence over content and policies. Their user policy doc-
uments mainly serve to inform users about the company’s practices. Thus, social
media platforms differ with respect to the degree of participation they grant to
users on two levels: the purely technical infrastructure and the policies that regu-
late how this infrastructure can be used.
2.2. Realising the technical potential: Degrees of interaction

The interactive potential of social media platforms is not always fully realised. In
highly interactive settings, we would expect replies to occur very regularly and we
would also expect to be able to observe long interactions between small groups of
participants who develop ideas and arguments over the course of their interaction.
However, most studies that investigate the degree of interaction in social media
come to the conclusion that such interactions are the exception rather than the rule.
The likelihood for messages to receive replies and to lead to interactions varies
across different communication forms and platforms. Jucker and Dürscheid (2012:
43) argue that communicative acts are characterised by different uptake expecta-
tions depending on the context in which they occur. For instance, a status update
on Facebook by a user who posts holiday pictures typically has a very high degree
of uptake expectation; the user posting the message expects friends to react to the
message by liking and perhaps even commenting on it. The uptake expectations
can be further increased when specific friends are addressed in the post, perhaps
even with questions. Likewise, on Twitter, uptake expectations can be increased
through the use of hashtags and the address of other users with @-phrases (see
Jucker and Dürscheid 2012: 48).
There are various ways of measuring the degree of interaction. The overall
number of users and contributions can serve as a first indication. This can be used
to compare different platforms (e.g. number of users on Facebook and on Twitter)
as well as sections within platforms (e.g. number of posts on different topics on a
given message board; the number of followers on Twitter). Calculating the num-
ber of contributions per user can further help identify platforms that have a stable
community of regular participants. Taking this approach one step further, grouping
users by the number of messages they post is a way to identify user groups with
expertise in articles on Pittsburghese across different media, including print newspaper

articles, a website and an email discussion forum. She comes to the conclusion that
Wikipedia has the tightest editorial constraints on what types of sources are considered
acceptable, and it is the only platform in her sample that does not regard personal expe-
rience as a valid source of expertise (2011: 12).

Authenticated
36 Daniela Landert
different levels of participation on a single platform. For instance, McDonald and

Woodward-Kron (2016) analyse data from an online message board on bipolar dis-
order by dividing the posts into sub-corpora according to “user’s post count at the
time of posting” (2016: 162). Posts in the sub-corpus of first posts are contributed
by new members, while posts in the sub-corpora of 220–559th and 560+ posts are
contributed by veteran members. This allows McDonald and Woodward-Kron to
compare the language use of users with different degrees of active participation in
the online community. Their findings show that the different roles in the commu-
nity are directly reflected in different communicative patterns of giving advice, in
metadiscourse, and in the construction of identities in relation to bipolar disorder
(McDonald and Woodward-Kron 2016).
Another way to study the degree of interaction on a given platform is to ana-
lyse the length and branching of message threads. In highly interactive exchanges,
users engage in interactions that span several turns, with later messages replying to
earlier messages. This leads to deep structures with long message threads. In con-
trast, flat structures are indicative of exchanges with a lower degree of interaction,
in which users tend to post messages which start new topics or which only reply
to one starting point message (e.g. a newspaper article, a blog post). Marcoccia
(2004) is an example of a study that investigates the length of message threads.
He studied French-speaking newsgroups in 1997 and found an average length of 5
messages per conversational sequence.
Some approaches to interaction further study how closely messages relate to
each other. Rafaeli and Sudweeks (1997) propose a three-way distinction. In one-
way communication no interaction takes place at all. In reactive two-way com-
munication, the addressee replies to a previous message, taking into account that
previous message, but no messages that preceded it. In contrast, in fully interactive
communication, the reply to a message takes into account the entire sequence
of previous messages and the way in which they were reactive or interactive. In
their early study of three online discussion groups in 1993, Rafaeli and Sudweeks
(1997) found different levels of interaction in the three groups and they relate the
degree of interaction to the degree of user involvement, arguing that high degrees
of interaction create engagement.
Arendholz (2013) develops her own measure of the degree of interaction on
message boards, which is calculated on the basis of the number of messages in a
given thread and the number of users who created the messages. Her results show
that only a small proportion of threads in her sample qualify as highly interactive
according to her measure. Two out of fifty threads contain on average roughly
four messages per user, which is the result with the highest interaction quotient of
her sample. In contrast, in several of the threads the number of messages is only
marginally higher than the number of users, which means that interaction between
users is hardly possible. Repeated replies between users would have resulted in
much higher numbers of messages per user.

Authenticated
Similar results were found in studies of user comment sections on online

news sites by Kleinke (2010) and Neurauter-Kessels (2013). Kleinke (2010: 202)
observes in her data from BBC Talk and the Spiegel Online forum that messages
that refer to previous messages usually do not receive a reply by the user who
posted the earlier message. This results in largely unidirectional interaction, in
which messages sometimes build chains without ambidirectional interaction: User
A may post a reply to user B’s message, which was posted in reply to user C’s mes-
sage, but C does not reply to B and B does not reply to A (Kleinke 2010: 202–205).
A similar observation is made by Neurauter-Kessels (2013), who studies conflic-
tive exchanges in user comments of five British online newspapers. All conflictive
exchanges start with an initial message that contains a personal attack against other
users, journalists or moderators. For four of the five newspapers, the percentage
of such attacks that did not receive any response was very high, between 72 %
and 88 %. Only for one newspaper, the Express Online, a clear majority of attacks
received a response, namely 66 %. This was also the only newspaper for which con-
flictive exchanges regularly reached a length of four messages and more (21 %).
Neurauter-Kessels (2013: 228–229) explains this difference with the smaller num-
ber of users that are active on the Express Online compared to the other news sites,
and the fact that the users who were involved in the conflicting exchanges were all
quite active on the site. In other words, the higher degree of interaction is likely to
be related to a more close-knit community of users (2013: 263).
Likewise, for blogs the general finding is that the interactional potential that
is offered by the technical infrastructure is often not fully exploited by users (e.g.
Bolander 2013; Hoffmann 2012). Studying data collected from ten different blogs,
Hoffmann (2012: 202) observes that “interaction in comments is […] extremely
limited”. In his data, most comments are directed at blog authors and if blog authors
respond to these comments at all, their comments do not receive any further reply
(2012: 202, 211–212). Bolander (2013) made very similar observations in her
study of eight diary blogs. Of the 717 reader comments she analysed, 74 % were
written in response to the blog post (2013: 106), which leaves only about a quarter
of reader comments that can potentially be used for longer interactions. She also
found that 75 % of readers in her data produced only one comment, and that cases
in which readers submitted more than one comment in the same section are rare
(2013: 107). The typical patterns of interaction that can be observed consist of
two to three turns: a post by a blogger, responded to by a reader comment which,
in some cases, receives a reply by the blogger (Bolander 2013: 113). In contrast,
longer sequences and interactions between readers are far less common in her data.
However, Bolander also observes differences across her data. The type of dis-
course move has an impact on the degree of interaction; more precisely, disagree-
ments and criticisms are particularly common in reader comments directed towards
other readers and they seem to be more likely to trigger further reader comments
(Bolander 2013: 128–129, 204; see also Bolander 2012: 1614). In addition, she

Authenticated
38 Daniela Landert
finds differences between the eight blogs in her sample. In one of the blogs, the
largest group of comments are comments by readers in response to a previous
reader comment. On this blog, this group accounts for 41 % of all comments –
compared to less than 10 % for all the other blogs (2013: 114). Bolander relates
this difference to group size and participant relationships, especially with respect
to how well users know each other. Similar to Neurauter-Kessels’ (2013) findings
mentioned above, the higher degree of user interaction in Bolander’s study occurs
in the setting in which the number of active users is smallest. Bolander (2013: 115)
is careful to point out that we cannot assume a simple causal relationship between
these factors; not all small online communities of users who know each other well
show a very high degree of interaction. Still, the available evidence so far indicates
that social factors like group size, familiarity of users, the type of discourse move
and the interactional norms within a given group play a very important role for the
degree of interaction.
In sum, the results so far show that it is certainly not enough to look at the tech-
nical infrastructure of a given platform in order to judge the degree of interaction
between users. While a certain technical infrastructure is required for interaction,
such technical potential is not sufficient to ensure that users actually engage in
interaction.
2.3. Horizontal and vertical communication

When approaching questions of participation, there is another aspect that becomes
important, namely who interacts with whom. Chung (2012: 43) introduces the dis-
tinction between horizontal and vertical communication. Vertical communication
describes interactions between journalists, public figures, and other professional
communicators on the one hand, and users on the other. Typical examples are
online newspapers, where readers can post comments in reaction to news articles,
and social media profiles by politicians and political parties, which allow support-
ers to follow, like, share and comment on content. In contrast, horizontal commu-
nication refers to interaction between (non-professional) users. Typical examples
are social media interactions between friends who also interact with each other
offline and message boards devoted to a topic where users can meet other users
who share their own interests.
These two types of social media interaction create different forms of user par-
ticipation. Vertical communication leads to a form of invited participation in which
the previously largely invisible mass audience is given an opportunity to talk back.
Motivations for inviting such user participation are manifold. Apart from gaining
insights into the opinions and preferences of their audience, social media interac-
tion with and among users can also be used to increase involvement and to create a
sense of community, which can lead to a higher degree of identification with (and
thus loyalty to) products, companies and political parties. In political campaign-

Authenticated
ing, this use of social media has been gaining importance over the last decade to
an extent that social media competence is now viewed as an important factor in
winning elections (see also Section 5). In the case of public service providers, ver-
tical communication can also be a way to fulfil their public service mandate (Enli
2008; Allan and Thorsen 2011: 29). As far as participation is concerned, vertical
communication typically takes place on platforms that are maintained by institu-
tions or professional communicators who consequently can control the degree and
form of participation that is available to users.
Horizontal communication takes place between peers who may or may not
know each other outside of social media contexts. In some cases, social media sim-
ply provide an additional channel of communication alongside face-to-face inter-
action, phone conversations, and other forms of mediated communication. In other
cases, social media enable users to interact with likeminded others with whom
interaction would be unlikely, difficult or even impossible offline. Sometimes
such interaction results in the formation of relatively stable groups or communities
online, and sometimes interaction takes place in more ad-hoc groups, which are,
for instance, based on specific hashtags. This latter option has been characterised
as “ambient affiliation” by Zappavigna (2012, 2014; see also Zappavigna, Ch. 16,
this volume). Social media can also provide spaces for horizontal interaction in
contexts in which free speech is restricted through political control and censor-
ship (see, for instance, Han 2016). While only accounting for a small share of all
interactions, this is the dimension of social media that is most clearly tied to social
participation and democratic processes and which, therefore, is of high symbolic
value. To what extent social media interactions of this kind are actually able to
influence political processes remains controversial, though (see Han 2016; Khia-
bany 2016; Papacharissi and Blasiola 2016; Thurlow 2013: 237).
The distinction between horizontal and vertical communication is analytically
useful, but it is important to note that most social media platforms do not fall neatly
into one or the other category. Platforms which, at first sight, appear to be designed
for vertical communication are sometimes used mainly for horizontal communica-
tion among users. User comment sections on online newspapers are an example.
While some users explicitly address their comments to journalists, it is quite rare
that journalists actually reply to such messages and it is often unclear if the com-
ment is actually read by the intended addressee. In an early study on user comment
sections on the New York Times website, Schultz (2000) found that comments were
only contributed by readers, and most journalists he interviewed said that they did
not even read the user comments. In contrast, interactions between users are more
common, although they also tend to be limited to occasional short sequences of
messages on most websites. At the same time, platforms which are often associated
with interaction among friends, such as Facebook, also have frequent institutional
uses by companies, associations and public figures who address their customers,
supporters and voters. Such institutional uses of social media often show similar

Authenticated
40 Daniela Landert
characteristics as user comment sections on news sites. Politicians may address

their voters in the first post and users may respond to this by liking, sharing and
commenting (vertical communication), but subsequent turns in the interaction in
comment sections often take place between users (horizontal communication).
In sum, horizontal and vertical communication lead to different forms of par-
ticipation. Vertical communication is a form of invited participation between pro-
fessional and non-professional individuals whereas horizontal communication
takes place among equals. This distinction plays an important role for the analysis
of wider social and political effects of social media interaction.
3. New models for new forms of participation
Social media interaction and participation have a number of consequences in dif-

ferent areas. One that has already been mentioned and which will not be further
discussed here is their potential to contribute to democratic processes. Other con-
sequences relate to shifts of roles in the widest sense and are often characterised
as blurring of boundaries of one type or other. A typical characterisation of this
can be found in Chung: “Interactivity fundamentally challenges existing models
of communication and blurs the lines between mass and interpersonal, sender and
receiver, and traditional and new media” (Chung 2012: 37).
It is certainly true that new media have led to various innovations and that they
have had a profound impact on some areas of communication and social innova-
tion. However, it is also true that many of these changes are not quite as categorical
as might seem at first sight. Moreover, in some cases the speaking of a blurring of
lines can block the view for the complexities of specific communicative constella-
tions. Many of the apparently blurred distinctions can be disentangled, sometimes
with the help of new models and concepts that are better suited to representing the
new communicative forms. Recent years have seen a number of suggestions for
such new frameworks, and some of these will be discussed briefly in this section.
3.1. Interpersonal and mass media communication

First, there have been suggestions for new models of mass media communication
and interpersonal communication. A traditional view of the two types of communi-
cation presumes that mass communication is publicly accessible, unidirectional, and
addressed at an anonymous mass audience (Luhmann 1995; McQuail 1987: 31–32),
while interpersonal communication is interactional and takes place between a closed
group of participants. In social media interactions, this binary opposition is clearly
not tenable. However, it is important to note that the clear dichotomy of mass and
interpersonal communication has been challenged long before social media came
into being. As early as 1972, Chaffee talks about the fact that mass media and inter-

Authenticated
personal interaction are interrelated. Mass media try to integrate characteristics of

interpersonal communication, for instance through letters to the editor, and people
rely on knowledge acquired through mass media when engaging in conversations
with others (Chaffee 1972: 95, 114). Referring to media change, he further com-
ments on the fact that the audience also shapes the way in which media develop: “It
seems that the influence of the audience on the medium will be at least as important
in this case as will the more conventional question of the medium’s influence on the
audience” (Chaffee 1972: 115). Still, social media present new challenges and make
it necessary to reconsider mass and interpersonal communication.
Lüders (2008) proposes a model in which she characterises the relationship
along two independent axes. The fist axis distinguishes between institutional or
professional content on the one hand and de-institutional or de-professional con-
tent on the other. The second axis describes whether communication is symmet-
rical interaction or asymmetrical quasi-interaction. Mass media are characterised
as formal/professional content and asymmetrical quasi-interaction. In contrast,
social media – called personal media by Lüders – contain de-institutional/de-pro-
fessional content and can be placed at various points on the interaction axis. While
Lüders’ model helps describe similarities and differences between mass media and
social media, one of its limitations is the fact that it cannot accommodate social
media use by institutional or professional users. Another limitation is that it cannot
describe settings in which more traditional mass media communication is com-
bined with interpersonal communication.
Janoschka’s (2004) model of interactive mass communication deals with this
aspect. She addresses the fact that many online settings provide at the same time
one channel for unidirectional mass communication and other channels for inter-
active communication between sender and addressee. For instance, a corporate
website may include many parts that are unidirectional, providing information
about the company and its products. In addition to this, some interactional chan-
nels might be provided, too, such as the option of submitting comments online, or
of engaging in an online chat with a support agent. In contrast to Lüders (2008),
Janoschka’s model can account for such cases in which mass communication and
interpersonal communication are combined on the same platform without giving
up the analytical distinction between the two.
Janoschka’s model has been further developed for the case of online news
sites, which provide a particularly rich example of combining unidirectional mass
media communication (e.g. publication of news stories) with various forms of user
interaction and user-generated content (Landert 2014a). This revised model takes
into account that not all sections of online news sites provide the same options
for user interaction. Some online news sites have sections that are devoted to user
comments and other forms of user-generated content. In these sections, users can
actively participate and may even have a relatively high degree of control over
their content, even though it is usually moderated by the news site. At the same

Authenticated
42 Daniela Landert
time, online news sites always include sections devoted to editorial content, such
as news articles, where the control over content lies entirely with journalists and
other professional news workers. The relationship between editorial content and
user content – in terms of relative space of each and degree of differentiation
between the two – is one of the aspects in which online news sites differ from each
other quite considerably (Landert 2014a).
3.2. Users and collaborative content

Bublitz (2012a, 2012b) addresses another boundary that is blurred in social media,
the boundary between readers and writers. He argues that the duality principle
of communication is undergoing erosion in computer-mediated communication.
Older forms of communication, such as handwritten letters or manuscripts, clearly
distinguish between the person who writes and the recipient who reads. In social
media communication, these roles are no longer clearly distinguished. Instead,
the new concept of the user replaces both writers and readers; users receive and
read messages, and they also participate in the interaction more actively. Bublitz
(2012b: 155) adopts Eisenlauer and Hoffmann’s (2010: 103) distinction of three
types of interactivity, cognitive, structural and productive. Cognitive interactivity
refers to the cognitive processes of meaning making on the part of the reader which
take place in all forms of communication. Structural interactivity occurs whenever
readers take an active role in deciding in which sequence they read a given text,
for instance by following hyperlinks to read texts in a non-linear fashion. This type
of interactivity is typical of computer-mediated texts but not restricted to them.
Examples of printed texts in which structural interactivity occurs regularly are, for
instance, encyclopedias and gamebooks (Bublitz 2012b: 155). Finally, productive
interactivity refers to the option of modifying and supplementing texts. In many
online settings, the degree of productive interactivity is much higher than in offline
mediated communication. This is the main factor that distinguishes interactivity
in online settings from earlier forms of communication, and it is the reason why
Bublitz (2012b) argues that participants in online interaction are more appropri-
ately characterised as users than as readers and writers.
A related tendency is noted on a slightly different level by Bruns (2008). He
looks at collaborative content creation online and speaks of produsers, i.e. users
who not only consume media products, but who actively contribute to them by
producing their own content. Wikipedia is a prominent example of online content
that is produced by its own users. Other examples of “participatory culture” that
have been analysed include fan fiction (e.g. Barton and Lampley 2014; Jenkins
1992, 2006) and citizen journalism (e.g. Bruns 2005; Papacharissi 2009). Both
forms have had precursors that predate the widespread use of the Internet, but
social media have helped them gain momentum to a very considerable extent.
From the perspective of pragmatics, such new forms of collaborative content

Authenticated
creation have not been fully explored yet. The studies that have been carried out
so far point to the potential of collaborative content to provide new insights on a
number of issues. For instance, Bartlett (2012) presents a case study on the nego-
tiation of the wording of a sentence that caused disagreements on Wikipedia. He
draws attention to different ways in which the editors’ decisions are influenced by
their role as “prosumers” of Wikipedia and points to the need to analyse the negoti-
ation of objectivity and “neutral point of view” in the context of the norms of their
community of practice. Page (2013) focuses on the open-ended and collaborative
narration on Wikipedia. She analyses article versions from two different language
versions of Wikipedia (English and Italian) and two points in time on one topic
(“Murder of Meredith Kercher”) and shows how the relative prominence of differ-
ent narratives varies across cultural contexts and how it changes over time. Like
Bartlett (2012), she emphasises the importance of context for the analysis of the
negotiations of dominant versions of the narrative and she argues that the analysis
should not be restricted to the final product (one version of the article) but take
into account the production process that is characterised by open-endedness and
multiple tellership of the narrative (see also Page, Ch. 19, this volume).
4. Strategies for creating involvement
In the previous sections of this chapter we looked at involvement that relied on

actual user interaction through a social media platform. However, involvement
can also be created in other ways. On the level of content, the topic of a message
and the way in which the topic is presented play an important role. A focus on
private topics and personal stories tends to create more involvement than abstract
and generalised topics. And linguistic immediacy – i.e. language that is typically
associated with informal private conversations – can be used to create involvement
through language (see, for instance, Chafe 1982; Koch and Oesterreicher 1985;
Tannen 1982, 1986). Both these strategies can be used to personalise mass media
communication (see Landert 2014b). They foreground the individuals that partic-
ipate in the interaction and they present them not as professional communicators,
but as private individuals. Despite the fact that the author and the addressee are not
co-present and that they may not even know each other, the communication shows
characteristics that are typical of personal face-to-face interaction between friends,
either in terms of topic or language.2
These alternative strategies for creating user involvement are not restricted to
social media. For instance, tabloid newspapers are well known for using stories
2
For a detailed discussion of how these characteristics relate to orality and literacy, see
Bös and Kleinke, Ch. 4, this volume.

Authenticated
44 Daniela Landert
about the private lives of individuals as well as language that violates the journal-
istic norms of objectivity and detachment in order to involve their readers (see,
for instance, Conboy 2003, 2006; Sparks and Tulloch 2000). Advertisement is an
area in which linguistic strategies of involving the addressee are very prominent.
Hermerén (1999) argues that print advertising uses questions and forms of direct
address as personalisation strategies to “create the impression that the product or
service promoted is tailor-made” for the recipient of the mass media message.
Similar means of “[creating] the impression of personal communication” were
found by Janoschka (2004: 132) in non-interactive forms of web advertising. Like-
wise, Barron (2012) demonstrates how public information messages make use of
various linguistic strategies to create the impression of interaction, even though no
interaction was taking place. The linguistic means that she identifies include, for
instance, first person pronouns, deictic reference, questions, directives, and lexical
references to the target group.
Fairclough ([1989] 2001) coined the term “synthetic personalization” to refer
to such linguistic means of simulating interaction. He describes synthetic person-
alisation as a linguistic strategy of mass media:
Synthetic personalization simulates solidarity: it seems that the more ‘mass’ the media
become, and therefore the less in touch with individuals or particular groupings in their
audiences, the more media workers and ‘personalities’ (including politicians) purport
to relate to members of their audience as individuals who share large areas of common
ground. (Fairclough [1989] 2001: 160)
Thus, mass media use such linguistic strategies to address their anonymous mass
audience as individuals and to simulate interaction linguistically.
The three main strategies for creating involvement – interaction, linguistic
immediacy, and private content – can be used independently. Business emails,
while being interactive, may be formulated in very formal language and often
deal with content that has little potential to involve the addressee emotionally.
In contrast, a death notice published in a newspaper may present involving con-
tent in formal language and with no immediate options for interaction. Linguistic
immediacy is often used in print advertising, where the content is not emotionally
involving and where interaction is not immediately available. These differences
can be visualised with a three-dimensional model, in which each axis stands for
one of the three involvement strategies (see Figure 1, based on Landert 2014b).
Each communicative act can be placed in this model with respect to how much
it makes use of each involvement strategy. A communicative act that uses hardly
any involvement strategies – such as a prototypical article in a print newspaper on
a hard news topic – is placed at position A in the model. A business e-mail with
some degree of interaction, non-private content and no linguistic immediacy can
be placed at position B. A death notice would correspond to position C and an
interactive print advertisement might correspond to position D.

Authenticated
Figure 1. Visualisation of involvement strategies in a three-dimensional model

(based on Landert 2014b: 30)
Despite this independence of the three strategies, they are often used in combi-
nation on social media platforms. Thus, a prototypical Facebook post deals with
a personal story, written in informal and involving language, and posted with the
intention of eliciting a reaction quickly in the form of likes, shares or comments.
This corresponds to position E in Figure 1. The co-occurrence of strategies is
not entirely coincidental. Social media tend to be very supportive of interaction,
the quasi-synchronous nature of communication encourages the use of linguistic
immediacy (Dürscheid 2007: 38), and the fact that social media are often used for
informal interaction between friends and acquaintances means that private content
is quite common. In this sense, the three strategies for creating involvement are
closely related to social media.
However, this does not mean that the technological development of social
media is the sole driving force behind increased levels of linguistic immediacy and
private content. Instead, we can look at the use of social media for involved inter-
action as a technically enabled continuation of the much older aim to simulate sol-
idarity and to relate to members of the audience as individuals described by Fair-
clough ([1989] 2001). A deterministic approach to media change would assume
that technological developments lead to changes in language and communication
in a unidirectional way. Contrary to this, there is ample evidence that the relation
between technological change, on the one hand, and social practices – including

Authenticated
46 Daniela Landert
communication –, on the other, is more complex. Herring (2003) sees technology

as facilitator of change. New forms of technologically-mediated communication
can be “a site for the emergence and evolution of linguistic norms” (Herring 2003:
1). Other supporters of a weak social constructionist perspective argue that devel-
opments in technology and in social practices are interdependent, and that social
factors often influence technological developments (e.g. Bijker 1997; Bolter and
Grusin 1999: 73–77; Bublitz 2012a). Such a perspective sheds new light on the
relation between social media and involvement. Rather than assuming that the new
technological setting of social media has caused users to communicate in more
involved ways, one can argue that the technological innovations enabling social
media platforms were, at least in part, motivated by the aim to relate to the mass
audience in more personalised and involved ways.
Finally, the frequent co-occurrence of involvement strategies is a consequence
of genre conventions that have developed. Many social media platforms are
strongly associated with the private use by individuals who exchange personal
content in informal language. This prototypical use has played an important role in
establishing the norms of interaction on these platforms. Today, many social media
platforms are also used by professional communicators, such as commercial busi-
nesses, organisations, politicians and celebrities (see also Thurlow 2013 for a criti-
cal discussion). For instance, Facebook states that there were more than 50 million
active business pages on their platform by the end of 2015 (Facebook for Business
2015). There is not a great deal of research so far about the differences between the
communicative practices of private individuals and professional communicators
on social media platforms. However, it seems that professional communicators
adjust to the involved style that is the norm on many platforms. To give an exam-
ple, Puschmann (2010) analyses the use of interpersonal pronouns on corporate
weblogs and identifies possible explanations for their strong presence. Similar to
their use in advertising, they fulfil functions that help further the communicative
goals of corporate weblogs by personalising the communication between compa-
nies and their customers. At the same time, Puschmann (2010: 188) emphasises
the role of genre in shaping readers’ expectations about the texts. In other words,
readers of corporate weblogs who are familiar with the norms of personal weblogs
expect to find a certain degree of similarity between the two genres.
5. Political communication on social media
Politicians are a group of professional actors for whom social media have gained
importance over the last decade. A crucial point in this development was Obama’s
first presidential campaign in 2008 and his re-election campaign in 2012, in which
social media are said to have played a decisive role – to the extent that Obama has
been referred to as “the first social media president” (Rutledge 2013). Indeed, his

Authenticated
election was in part attributed to the fact that he was able to mobilise new seg-
ments of voters through the successful use of social media (e.g. Bimber 2014; Enli
and Naper 2016; Kaye 2011; Nagourney 2008). Since then proficiency in social
media campaigning has become a key skill of any election team, even though only
few politicians succeed in drawing attention to their social media presence on a
large scale (Nielsen and Vaccari 2013).
Larsson (2015: 151) points out that most research on social media use by poli-
ticians so far has focused on election campaigns, despite the fact that social media
use has long become part of the everyday practices of many politicians. Everyday
use of social media can serve different purposes. For instance, politicians regularly
use social media to release information directly, without relying on news media as
gatekeepers (Riboni 2015: 260), and sometimes strategic communication on social
media is even used for agenda building and to influence the press (Kreiss 2016:
1475). How widespread the practice of releasing information on social media is
could be seen at the conclusion of the Iran nuclear talks in April 2015. After a series
of meetings over several days, the agreement on Iran’s future nuclear programme
was announced by Iran’s Foreign Minister Javad Zarif via Twitter directly from
the meeting. His tweet “Found solutions. Ready to start drafting immediately.”
was quoted verbatim in a breaking news story on BBC News before the press con-
ference could take place (see Figure 2). This example is particularly compelling if
one takes into account that Twitter was banned in Iran at the time of this tweet (see
Murgia 2016). By releasing the information on his Twitter account, Zarif was able
to draw attention to his own role in the successful negotiations.
Another function of social media use is to keep in touch with their electorate
between election periods, which can make it easier to secure their support during
election periods. Larsson (2015: 150) refers to this aspect as “permanent cam-
paigning”. A successful election campaign can lead to a very strong emotional
response of the supporters (see, for instance, Zappavigna’s (2012) analysis of the
Twitter response to Obama’s first election). The hope is that if voters follow politi-
cians on their social media profiles, then part of this emotion may be kept alive and
could even lead to some sort of community building among supporters, which in
turn would increase affiliation with the politicians and the likelihood of continued
support in the future.
5.1. Case study: The White House on Facebook

Obama was the first US president to communicate through Facebook. Less than
three months after he assumed office, the page “The White House” was launched
on Facebook. This Facebook page is only one of the channels of communication,
together with the White House website, which also includes a blog, a Twitter feed,
and presence in traditional mass media. It is also important to note that the Face-
book page of The White House is not maintained by Obama personally, but by his

Authenticated
48 Daniela Landert
Figure 2. Tweet by Iran’s Foreign Minister Javad Zarif (left) and quotation of Tweet in
breaking news report on BBC News (right)
staff. In addition, Obama has been present on Facebook with his personal profile
since November 2015.
In what follows, I present a very brief case study on the Facebook page of The
White House. The case study is based on a small set of data consisting of 37 posts
that were published by The White House between 8 March and 9 April 2015. The
question that I will deal with is how and to what extent the posts create involve-
ment. More specifically, I will look at how the posts make use of interaction, con-
tent and language to involve followers of The White House.
Undoubtedly, interaction plays a very central role for creating involvement in
the collected posts. The posts offer various opportunities for interaction through
liking, sharing and commenting. All of these options were used very actively for all
of the posts; every post received at least 10,000s of likes, 1,000s of shares and 100s
of comments. The post that received most comments contained Easter greetings
from the Obama family. It received 83,684 comments and more than 2 million likes
within the year following the publication of the post. For technical and ethical rea-
sons, comments to the posts were not collected and analysed separately. However,

Authenticated
a cursory view at the top-listed comments indicates that some of them are addressed
to Obama while others are addressed to the community of followers.3 There is also
interaction between followers, with some of the top-listed comments receiving
more than 100 replies. This high number of comments and replies suggests that the
Facebook page succeeds in creating personal involvement through interaction, even
though there is no indication that Obama or his staff reacted to the comments directly.
In addition to enabling interaction, the communicative setting of the Facebook
page plays a role for creating involvement in two more ways. First, it is relevant
that most non-professional users of Facebook use the platform typically to stay in
touch with friends and acquaintances. As a consequence, when they follow The
White House, political posts are displayed on the users’ news feed in between their
friends’ status updates on private topics. This reception of political posts together
with private content is certainly a very important factor in presenting Obama in a
personal way and in creating involvement.
Second, the platform supports multimodal communication and hyperlinking.
All of the posts in my sample contain either an image or a video, most posts con-
tain hyperlinks, and many contain hashtags. In contrast, there is relatively little text
included in the posts. The number of words (including hyperlinks and hashtags)
ranges from 0 to 185 with an average of 40. Both multimodality and hyperlinking
serve to increase involvement. Hyperlinks are a form of structural interactivity
(see Section 3) inviting readers to select their own reading paths, to interact with
content, and to follow up on stories by visiting other sites. Visual elements are
generally associated with a higher degree of involvement compared to text, which
is sometimes attributed to a more immediate perception of content (Eisenlauer and
Hoffmann 2008: 8) and a stronger reliance on association and higher emotional
response compared to text (Kappas and Müller 2006; Müller and Kappas 2011). In
addition, a study on Obama’s Facebook posts during the 2012 election campaign
found that posts with photos of Barack Obama had a higher likelihood to receive
likes, comments and shares than other posts, and for photos including Michelle
Obama or one of their daughters, the effect was even stronger (Gerodimos and
Justinussen 2015: 125). This suggests that this type of image can further increase
involvement through leading to a higher degree of actual user interaction.
The images and videos that are included in the posts from The White House
further create involvement through their content and style. Out of the 37 images
and videos, 27 showed Barack Obama and in 3 he appeared together with members
of his family. Of the 10 remaining images and videos, 2 showed his handwriting
and one showed Michelle Obama. This strong visual focus on Obama serves to
foreground his person rather than his office. This is even more the case since
3
By default, two “most relevant” comments are displayed directly below each post. How
exactly relevance is assessed by Facebook is not clear.

Authenticated
50 Daniela Landert
Figure 3. Two playful images posted on The White House Facebook page in April 2015
many of the images are quite informal and playful. To give some examples, one
of the images shows Barack Obama with boxing gloves standing next to Michelle
Obama who is holding a dumb-bell. In another picture we see Obama from behind
sitting next to an Easter bunny in front of the Washington monument (see Fig-
ure 3). These are pictures that are designed to show Obama from a private side,
rather than in his professional role as President of the United States. Some other
pictures show him in professional contexts, but in situations in which it appears
that he is not posing for camera. An example of this is a picture that shows him
during a visit of a solar energy plant. Neither he nor the person who is with him
look into the camera and the picture gives the impression of presenting a glance
behind the scenes of the visit. Overall, the images and videos contribute to the cre-
ation of involvement by focusing on the person of the private, backstage Obama.
In terms of topic and content, most posts deal with current affairs (see Table 1).
These posts typically provide updates about ongoing political issues, often with
quotes and sometimes with videos from Obama’s speeches. This group of posts
is least involving, and it shows that despite the more personal and playful nature
of some of the content, the main aim of the Facebook presence is still to transport
political messages. The group of posts classified as ceremonial deals with anni-

Authenticated
versaries and historically significant events, such as the 50-year anniversary of the
Selma to Montgomery marches, the start of a year-long space mission, and St. Pat-
rick’s Day. The posts on these topics include quotes by Obama that acknowledge
the significance of the event, sometimes taken from speeches. Posts dealing with
events at the White House covered the White House student film fest and the five-
year anniversary of the West Wing Week, a weekly behind-the-scenes video show
of events at the White House. These posts do not focus on Obama but promote
events and related web resources. The six Easter posts include Easter wishes by
Obama and his family and Easter events taking place at the White House. They are
among the posts that present Obama from a more personal perspective. Finally, the
three posts classified as “other” include one post on the First Lady’s visit to Cam-
bodia, an update of the cover photo, and Obama’s basketball tournament bracket.
This last post is another example of focusing on Obama in a non-professional role
as someone who is interested in basketball and who makes his own predictions
about the tournament, noted down in handwriting.
Table 1. Content classification of 37 Facebook posts of The White House
Topic Number of posts

Current affairs 15
Ceremonial 7
Events at White House 6
Easter 6
Other 3
Total 37
Concerning linguistic immediacy, 17 of the posts address readers directly by using

second person pronouns, questions and imperatives (see Examples 1 and 2). While
the overall language use is not especially informal, there are a few instances of cre-
ative word play, such as egg-cellent in the context of an Easter post (Example 2)
and Baracketology when referring to Obama’s basketball tournament bracket.
Such interpersonal and playful language use again contributes to involvement by
staging an interaction with readers. In addition 19 of the 37 posts consist mainly of
direct quotes, 16 of them quotes from Obama (e.g. Examples 3 and 4).
(1) Share the good news: President Obama just announced new steps to train more Ameri-
cans and veterans for clean-energy jobs → http://go.wh.gov/YYmtFD #ActOnClimate
(4 April 2015)
(2) Watch some egg-cellent highlights from today’s White House Easter Egg Roll! wh.gov/
EasterEggRoll #GimmeFive (7 April 2015)
(3) “Today, the United States—together with our allies and partners—has reached an his-
toric understanding with Iran.” —President Obama: wh.gov/irandeal #IranDeal (2 April
2015)

Authenticated
52 Daniela Landert
(4) “Good luck, Captain. Make sure to Instagram it. We’re proud of you.” —President
Obama
Congrats to NASA Astronaut Scott Kelly as he takes off for the International Space
Station and his #YearInSpace: go.nasa.gov/1xFtpnu (27 March 2015)
The presence of Obama’s own voice contributes to involvement to some extent.

However, it is important to note that his voice is presented in quoted speech. Out-
side of quotation Obama is referred to in the third person. This is different on
Obama’s own Facebook profile, which was created a few months after the end
of the period from which I collected data, on 9 November 2015. Here, Obama
appears as the first person author, and even comments on this fact in his first post
(see Figure 4).
This new Facebook presence marks a new degree of involvement. Obama
appears as author and speaks about his personal attitudes, opinions and prefer-
ences in the first person. He combines his professional role (“I’m heading to Paris
to meet with world leaders”) with aspects of his private person (“something I try
to do at the end of the day before I head in for dinner”), positioning himself as
“one of us”, so to speak. Readers are addressed directly and invited to share their
own content. This new, even more involved channel of interaction was established
almost seven years after launching the Facebook page of The White House, which
shows that social media use by politicians is still gaining importance. This and
similar newly developing forms of interaction, participation and involvement are
only waiting to be studied.
6. Conclusion
The potential of participation is a characteristic that all forms of social media have
in common. However, this potential is realised to different degrees, in different
forms, and with different effects across social media platforms. Such differences
require careful consideration and critical examination. Perhaps the most important
result of research so far is that there is often a considerable discrepancy between
the degree of participation that is possible and the degree of participation that is
actually realised. Truly collaborative and interactive online environments tend to
be the exception rather than the rule. Nevertheless, such cases provide exciting
new research opportunities for pragmatics.
This chapter also addressed the fact that user involvement can be created in
different ways on social media. Interaction is perhaps the most obvious factor in
creating involvement, but content and language often play an important role, too.
The relation between different strategies of creating user involvement presents
promising avenues for future research, especially in the context of commercial and
political uses of social media.

Authenticated
Figure 4. First post on new Facebook account of President Obama, 9 November 2015

Authenticated
54 Daniela Landert
Acknowledgments
I would like to acknowledge the financial support of the University of Zurich

(Forschungskredit, grant no. FK-14–074).
References
Allan, Stuart and Einar Thorsen

2011 Journalism, Public Service and BBC News Online. In: Graham Meikle and
Guy Redden (eds.), News Online: Transformations and Continuities, 20–37.
Basingstoke: Palgrave Macmillan.
Arendholz, Jenny
2013 (In)appropriate Online Behaviour: A Pragmatic Analysis of Message Board
Relations. (Pragmatics & Beyond New Series 229.) Amsterdam/Philadelphia:
Benjamins.
Barron, Anne
2012 Public Information Messages. A Contrastive Genre Analysis of State-citizen
communication. (Pragmatics & Beyond New Series 222.) Amsterdam/Phila-
delphia: Benjamins.
Bartlett, Tom
2012 Lay metalanguage on grammatical variation and neutrality in Wikipedia’s
entry for Che Guevara. Text and Talk 32(6): 681–701.
Barton, Kristin M. and Jonathan Malcolm Lampley (eds.)
2014 Fan CULTure. Essays on Participatory Fandom in the 21st Century. Jefferson/
London: McFarland & Company.
Bijker, Wiebe E.
1997 Of Bicycles, Bakelites, and Bulbs. Toward a Theory of Sociotechnical Change.
Cambridge, MA: MIT Press.
Bimber, Bruce
2014 Digital media in the Obama campaigns of 2008 and 2012: Adaptation to the
personalized political communication environment. Journal of Information
Technology & Politics 11(2): 130–150.
Bolander, Brook
2012 Disagreements and agreements in personal/diary blogs: A closer look at
responsiveness. Journal of Pragmatics 44(12): 1607–1622.
Bolander, Brook
2013 Language and Power in Blogs. Interaction, Disagreements and Agreements.
(Pragmatics & Beyond New Series 237.) Amsterdam/Philadelphia: Benjamins.
Bolter, Jay David and Richard Grusin
1999 Remediation: Understanding New Media. Cambridge, MA: MIT Press.
Bruns, Axel
2005 Gatewatching. Collaborative Online News Production. New York: Peter Lang.
Bruns, Axel
2008 Blogs, Wikipedia, Second Life, and Beyond. From Production to Produsage.
New York: Peter Lang.

Authenticated
Bublitz, Wolfram
2012a Der duale Internetnutzer: Ansätze einer dissoziativen Kommunikation. In:
Konstanze Marx and Monika Schwarz-Friesel (eds.), Sprache und Kommunika-
tion im technischen Zeitalter. Wieviel Internet (v)erträgt unsere Gesellschaft?,
26–52. Berlin/Boston: de Gruyter.
Bublitz, Wolfram
2012b From speaker and hearer to chatter, blogger and user: The changing metacom-
municative lexicon in computer-mediated communication. In: Ulrich Busse
and Axel Hübler (eds.), Investigations into the Meta-Communicative Lexicon
of English. A Contribution to Historical Pragmatics, 151–176. Amsterdam/
Philadelphia: Benjamins.
Buni, Catherine and Soraya Chemaly
2016 The secrect rules of the Internet. The murky history of moderation, and how it’s
shaping the future of free speech. The Verge. http://www.theverge.com/2016/
4/13/11387934/internet-moderator-history-youtube-facebook-reddit-censor-
ship-free-speech (last accessed 14 April 2016).
Caffi, Claudia and Richard W. Janney
1994 Toward a pragmatics of emotive communication. Journal of Pragmatics 22
(3–4): 325–373.
Carah, Nicholas and Eric Louw
2015 Media and Society. Production, Content and Participation. Los Angeles/Lon-
don: SAGE Publications.
Chafe, Wallace L.
1982 Integration and involvement in speaking, writing, and oral literature. In: Deb-
orah Tannen (ed.), Spoken and Written Language: Exploring Orality and Lit-
eracy, 35–53. Norwood, NJ: Ablex.
Chaffee, Steven H.
1972 The interpersonal context of mass communication. In: F. Gerald Kline and
Phillip J. Tichenor (eds.), Current Perspectives in Mass Communication
Research, 95–120. Beverly Hills/London: Sage Publications.
Chung, Deborah S.
2012 Interactivity: Conceptualizations, effects, and implications. In: Seth M. Noar
and Nancy Grant Harrington (eds.), eHealth applications: Promising Strate-
gies for Behavior Change, 37–55. London: Routledge.
Conboy, Martin
2003 Parochializing the global. Language and the British tabloid press. In: Jean
Aitchison and Diana M. Lewis (eds.), New Media Language, 45–54. London/
New York: Routledge.
Conboy, Martin
2006 Tabloid Britain: Constructing a Community Through Language. London:
Routledge.
Dürscheid, Christa
Eisenlauer, Volker and Christian R. Hoffmann
2008 The metapragmatics of remediated text design. Information Design Journal
16(1): 1–18.

Authenticated
56 Daniela Landert

2010 Once upon a blog … Storytelling in weblogs. In: Christian R. Hoffmann (ed.),
Narrative Revisited. Telling a Story in the Age of New Media, 79–108. Amster-
dam/Philadelphia: Benjamins.
Enli, Gunn and Anja Aaheim Naper
2016 Social media incumbent advantage. Barack Obama’s and Mitt Romney’s
tweets in the 2012 U.S. presidential election campaign. In: Axel Bruns, Gunn
Enli, Eli Skogerbø, Anders Olof Larsson and Christian Christensen (eds.), The
Routledge Companion to Social Media and Politics, 364–377. New York/Lon-
don: Routledge.
Enli, Gunn
2008 Redefining Public Service Broadcasting: Multi-Platform participation. Con-
vergence 14(1): 105–120.
Facebook for Business
2015 New tools for managing communication on your page. https://www.facebook.
com/business/news/new-tools-for-managing-communication-on-your-page?
__mref=message_bubble (last accessed 14 April 2016).
Fairclough, Norman
[1989] 2001 Language and Power. 2nd ed. (Language in Social Life). Harlow: Long-
man.
Gerodimos, Roman and Jákup Justinussen
2015 Obama’s 2012 Facebook campaign: Political communication in the age of the
like button. Journal of Information Technology & Politics 12(2): 113–132.
Han, Rongbin
2016 Cyberactivism in China. Empowerment, control and beyond. In: Axel Bruns,
Gunn Enli, Eli Skogerbø, Anders Olof Larsson and Christian Christensen
(eds.), The Routledge Companion to Social Media and Politics, 268–280. New
York/London: Routledge.
Hermerén, Lars
1999 English for Sale. A Study of the Language of Advertising. (Lund Studies in
English 99.) Lund: Lund University Press.
Herring, Susan C.
2003 Media and language change: Introduction. Journal of Historical Pragmatics
4(1): 1–17.
2012 Cohesive Profiling. Meaning and Interaction in Personal Weblogs. (Pragmat-
ics & Beyond New Series 215.) Amsterdam/Philadelphia: Benjamins.
Janoschka, Anja
2004 Web Advertising. New Forms of Communication on the Internet. (Pragmatics
& Beyond New Series 131.) Amsterdam/Philadelphia: Benjamins.
Jenkins, Henry
1992 Textual Poachers: Television Fans and Participatory Culture. New York:
Routledge.
Jenkins, Henry
2006 Fans, Bloggers, and Gamers. Exploring Participatory Culture. New York:
New York University Press.

Authenticated
Johnstone, Barbara
2011 Making Pittsburghese: Communication technology, expertise, and the discur-
sive construction of a regional dialect. Language and Communication 31(1):
3–15.
Jucker, Andreas H. and Christa Dürscheid
2012 The linguistics of keyboard-to-screen communication. A new terminological
framework. Linguistik Online 56: 35–60.
Kappas, Arvid and Marion G. Müller
2006 Bild und Emotion – ein neues Forschungsfeld, Publizistik 51: 3–23
Kaye, Barbara K.
2011 Between Barack and a net place. Motivations for using social network sites
and blogs for political information. In: Zizi Papacharissi (ed.), A Networked
Self. Identity, Community, and Culture on Social Networking Sites, 208–231.
Khiabany, Gholam
2016 The importance of ‘social’ in social media. Lessons from Iran. In: Axel Bruns,
Gunn Enli, Eli Skogerbø, Anders Olof Larsson and Christian Christensen
(eds.), The Routledge Companion to Social Media and Politics, 223–234. New
York/London: Routledge.
Kleinke, Sonja
2010 Interactive aspects of computer-mediated communication. ‘Disagreement’ in
an English and a German public news group. In: Sanna-Kaisa Tanskanen, Mar-
ja-Liisa Helasvuo, Marjut Johansson and Mia Raitaniemi (eds.), Discourses in
Interaction, 195–222. Amsterdam/Philadelphia: Benjamins.
1985 Sprache der Nähe – Sprache der Distanz: Mündlichkeit und Schriftlichkeit
Kreiss, Daniel
2016 Seizing the moment: The presidential campaigns’ use of Twitter during the
2012 electoral cycle. New Media & Society 18(8): 1473–1490.
Landert, Daniela
2014a Blurring the boundaries of mass media communication? Interaction and
user-generated content on online news sites. In: Jukka Tyrkkö and Sirpa Lep-
pänen (eds.), Texts and Discourses of New Media. (Studies in Variation, Con-
tacts and Change in English 15.) Helsinki: VARIENG. http://www.helsinki.fi/
varieng/series/volumes/15/landert/.
Landert, Daniela
2014b Personalisation in Mass Media Communication: British Online News between
Public and Private. (Pragmatics & Beyond New Series 240.) Amsterdam/Phil-
adelphia: Benjamins.
Larsson, Anders Olof
2015 The EU parliament on Twitter – Assessing the permanent online practices of
parliamentarians. Journal of Information Technology & Politics 12(2): 149–
166.
Lüders, Marika
2008 Conceptualizing personal media. New Media & Society 10(5): 683–702.

Authenticated
58 Daniela Landert
Luhmann, Niklas
1995 Die Realität der Massenmedien. Opladen: Westdeutscher Verlag.
Marcoccia, Michel
2004 On-line polylogues: conversation structure and participation framework in
internet newsgroups. Journal of Pragmatics 36(1): 115–145.
McDonald, Daniel and Robyn Woodward-Kron
2016 Member roles and identities in online support groups: Perspectives from cor-
pus and systemic functional linguistics. Discourse & Communication 10(2):
157–175.
McQuail, Denis
1987 Mass Communication Theory. An Introduction. 2nd ed. London: Sage Publica-
tions.
Müller, Marion G. and Arvid Kappas
2011 Visual emotions – emotional visuals. Emotions, pathos formulae, and their
relevance for communication research. In: Katrin Döveling, Christian von
Scheve and Elly A. Konijn (eds.), The Routledge Handbook of Emotions and
Mass Media, 310–331. (Routledge International Handbooks.) London/New
York: Routledge.
Murgia, Madhumita
2016 Twitter and YouTube unblocked in Iran for some users after sanctions lifted.
Telegraph. http://www.telegraph.co.uk/technology/social-media/12105631/
Twitter-and-YouTube-unblocked-in-Iran-for-some-users-after-sanctions-
lifted.html (last accessed 15 April 2016).
Nagourney, Adam
2008 The ’08 campaign: Sea change for politics as we know it. New York Times,
November 3, 2008. http://www.nytimes.com/2008/11/04/us/politics/04memo.
html (last accessed 2 May 2016).
Neurauter-Kessels, Manuela
2013 Impoliteness in Cyberspace: Personally Abusive Reader Responses in Online
News Media. Zurich: University of Zurich Doctoral Dissertation.
Nielsen, Rasmus Kleis and Cristian Vaccari
2013 Do people ‘like’ politicians on Facebook? Not really. Large-scale direct candi-
date-to-voter online communication as an outlier phenomenon. International
Journal of Communication 7: 2333–2356.
Page, Ruth
2013 Counter narratives and controversial crimes: The Wikipedia article for the
‘Murder of Meredith Kercher.’ Language and Literature 23(1): 61–76.
Papacharissi, Zizi (ed.)
2009 Journalism and Citizenship. New Agendas in Communication. New York:
Routledge.
Papacharissi, Zizi and Stacy Blasiola
2016 Structures of feeling, storytelling, and social media. The case of #Egypt. In:
Axel Bruns, Gunn Enli, Eli Skogerbø, Anders Olof Larsson and Christian
Christensen (eds.), The Routledge Companion to Social Media and Politics,
211–222. New York/London: Routledge.
Puschmann, Cornelius
2010 ‘Thank you for thinking we could’. Use and function of interpersonal pro-

Authenticated
nouns in corporate web logs. In: Heidrun Dorgeloh and Anja Wanner (eds.),
Approaches to Syntactic Variation and Genre, 167–194. Berlin/New York: de
Gruyter.
Rafaeli, Sheizaf and Fay Sudweeks
1997 Networked interactivity. Journal of Computer-Mediated Communication
2(4). http://onlinelibrary.wiley.com/enhanced/doi/10.1111/j.1083–6101.1997.
tb00201.x (last accessed 3 April 2016).
Riboni, Giorgia
2015 Enhancing citizen engagement. Political weblogs and participatory democ-
racy. In: Marta Dynel and Jan Chovanec (eds.), Participation in Public and
Rutledge, Pamela
2013 How Obama won the social media battle in the 2012 presidential campaign.
The Media Psychology Blog. http://mprcenter.org/blog/2013/01/how-obama-
won-the-social-media-battle-in-the-2012-presidential-campaign/ (last accessed
10 April 2015).
Schultz, Tanjev
2000 Mass media and the concept of interactivity: an exploratory study of online
forums and reader email. Media, Culture & Society 22(2): 205–221.
Sparks, Colin and John Tulloch (eds.)
2000 Tabloid Tales. Global Debates over Media Standards. Lanham: Rowman &
Littlefield.
Stein, Laura
2013 Policy and participation on social media. The cases of YouTube, Facebook,
and Wikipedia. Communication, Culture & Critique 6: 353–371.
Tannen, Deborah
1982 The oral/literate continuum in discourse. In: Deborah Tannen (ed.), Spoken
and Written Language: Exploring Orality and Literacy, 1–16. Norwood, NJ:
Ablex.
Tannen, Deborah
1986 Introducing constructed dialogue in Greek and American conversational and
literary narrative. In: Florian Coulmas (ed.), Direct and Indirect Speech, 311–
332. Berlin: de Gruyter.
Thurlow, Crispin
2013 Fakebook. Synthetic media, pseudo-sociality, and the rhetorics of web 2.0. In:
Deborah Tannen and Anna Marie Trester (eds.), Discourse 2.0. Language and
New Media, 225–249. Washington, DC: Georgetown University Press.
Tredinnick, Luke
2008 Digital Information Culture the Individual and Society in the Digital Age.
Oxford: Chandos.
Zappavigna, Michele
2012 Discourse of Twitter and Social Media. How We Use Language to Create Affil-
iation on the Web. London/New York: Continuum.
Zappavigna, Michele
2014 Ambient affiliation in microblogging. bonding around the quotidian. Media
International Australia 151: 97–103.

Authenticated
Authenticated
3. Participation as audience design
Marta Dynel
Abstract: This chapter presents a new look at the concepts of participation struc-
ture and receiver-oriented design of messages in social media, where the tradi-
tional dyadic model of communication is largely inapplicable. This fact necessi-
tates new frameworks of participation in order to capture the nature of multi-party
interactions on various social media platforms, which show many characteristics
different from offline communication. However, the essential features of inter-
action hold also for the diversified social media formats. This chapter critically
examines a selection of problems concerning the multifold reception end and pro-
duction end in social media (in comparison to traditional media formats). Also,
the issue of designing contributions in multi-party interactions among, frequently
anonymous, participants is addressed.
1. Introduction
There is no novelty in the observation that Web 2.0, the central concomitant of
social media, has established various new forms of communication (understood
in a commonsensical manner as verbal and non-verbal transmission of messages).
These new forms promote different configurations of participants, i.e. partici-
pant structure, in online interactions (traditionally defined as all manner of mutual
social action, here the focus being on those serving communication).1 These
changes are consequent upon interactivity, i.e. active interaction with the medium
that guarantees decision-based reception of messages (see Jucker 2003), and all
interactants’ active involvement (see Landert, Ch. 2, this volume). As Johansson
(2014: 43) puts it, “[p]articipation in Web 2.0, and thus the possibility for every-
body to interact with each other and generate content, is ‘a defining principle of
digital culture’ (Deuze 2006: 67)”. Social media are characterised by two features
that directly impact on the different participation frameworks: the loss of control
over one’s messages, and the easy transgression of the boundaries typifying the
traditionally separate communicative frames (Chovanec and Dynel 2015). Even
seemingly private exchanges via instant communicators between two participants
can mutate into publically available discourse (see Bös and Kleinke, Ch. 4, this
1
Participation, as defined by Goffman (1963), concerns individuals’, i.e. participants’,
involvement in an interaction.
DOI 10.1515/9783110431070-003
Authenticated
62 Marta Dynel
volume). Once communicated, messages can be transmitted further, whether in the

original form or modified and recontextualised. All this has a bearing on the par-
ticipant structure, which usually transcends the prototypical communicative dyad.
The classical models of communication (e.g. Bühler [1934] 1965; de Saus-
sure [1916] 1974; Jakobson 1960) depict human interactions as dyadic exchanges,
whilst only tacitly allowing also for interactions of more than two participants.
The standard dyadic framework involves the producer/sender at the production
end and the receiver/recipient at the reception end, technically labelled speaker
and hearer in linguistic research (regardless of the different modes and channels
of communication and the human faculties involved). Moreover, a statement can
be ventured that this prototypical constellation of two interactants lies at the heart
of most linguistic theories that concern the production and reception of discourse
and/or specific linguistic phenomena. However, natural face-to-face interactions
very frequently cannot be reduced to exchanges of merely two participants, neces-
sitating more complex frameworks (e.g. Goffman 1981; Levinson 1988; Clark
and Carlson 1982; Schober and Clark 1989; Clark and Schaefer 1987, 1992; Clark
1996; Verschueren 1999; Dynel 2010, 2011a, 2014a). Further complications can
be observed in interactions that arise thanks to traditional media epitomised by
programmes and films broadcast to wide audiences. Such interactions rely on
two levels of communication (on-screen interactions, which typically resemble
non-televised interactions, and the level including the viewer) and are only inter-
mittently restricted to two interactional roles, namely the speaker and the (col-
lective role of) the viewer (see e.g. Scannell 1991; O’Keeffe 2006; Lorenzo-Dus
2009; Dynel 2011a, 2011c, 2011d and references therein). Nowhere else are the
participatory frameworks more complex than in the diversified forms of social
media interactions consisting in computer-mediated communication,2 where the
standard communicative dyad is a rarity.
Most formats of interaction in the various forms of social media inherently
involve more than two participants. These are usually separated spatially and tem-
porally, participating in interactions asynchronously (and less frequently synchro-
nously). These features are not intrinsic solely to interactions in social media. For
example, the traditional, and nowadays obsolescent, form of communication by
letter writing evinces both these features, too. Also, a great proportion of commu-
nications in social media enjoy indeterminate, and frequently infinite, numbers
of anonymous participants at the reception end. Because of these characteristics,
the communication in social media has been dubbed ‘mediated quasi-interaction’
(see Chovanec 2010, 2011 and references therein). Essentially, this term refers to
2
This well-entrenched term may now be considered a misnomer insofar as contemporary
Internet users tend to use equipment other than computers, such as smartphones or tab-
lets (Chovanec and Dynel 2015).

Authenticated
Participation as audience design 63
a range of strategies that simulate the impression of direct contact between partic-
ipants who are removed in space and time.
However, social media interactions should not be regarded as feigned interac-
tions or pseudo-interactions despite of their indeterminate participation structure.
The frequently indeterminate hearership, physical distance and asynchronicity
should not be treated as reasons for rendering the notion of interaction inapplica-
ble in social media contexts. They do display the intrinsic features of interaction:
the sender’s production and the receiver’s reception of a turn, which is a minimal
contribution to an interaction, whether written or spoken (e.g. Goodwin 1981; see
Dynel 2010 and references therein). Thus, any message (however complex) sent
via the Internet,3 such as posting a comment, may be compared to taking a turn in
an ordinary conversation (Herring 2010; Dynel 2014b; Boyd 2014), even though
this turn-taking may not be a matter of occupying time slots (with much commu-
nication being asynchronous) but space on the screen after a message/turn has
been sent. Essentially, the users “experience CMC in fundamentally similar ways
to spoken conversation, despite CMC being produced and received by written
means” (Herring 2010: 2). An important complication is that turns in social media
are very frequently multi-modal, involving images, videos, and sounds, next to
verbalisations, whether typed or spoken, which amount to Internet discourse (see
Hoffmann 2012),4 the focus of pragmatic analysis (see Herring 2013; Herring,
Stein and Virtanen 2013). This questions the adequacy of the labels ‘speaker’ and
‘hearer’, unless their basic semantic meanings are marginalised, and the two labels
are treated as technical terms. In the present discussion, however, the labels ‘pro-
ducer’ and ‘receiver’ will be used as the basic generic terms.
The thrust of all this is that social media interactions show the characteris-
tics of both everyday and mass-media communication, and their new forms never
cease to challenge extant theory (cf. Poole and Jackson 1993). Although interac-
tions in social media are frequently asynchronous and often enjoy large audiences,
and the turns even in one interaction may manifest different formats (a video, an
image, a long text, a post, or a hyperlink), these interactions otherwise bear strong
similarity to everyday exchanges.
The present chapter addresses the issue of participation framework, under-
stood as the configuration of diversified production and reception roles. Also,
some thought is given to the idea of how the producers of messages facilitate
the understanding of receivers, indeterminate though they may be. The chapter is
3
As already signalled, one online message may be developed upon, or incorporate, pre-
vious messages. Social media allow infinite embedded messages, which have previ-
ously constituted independent turns, as one complex turn.
4
Essentially, discourse is the product of verbal (and non-verbal) communication in a
given context.

Authenticated
64 Marta Dynel
organised as follows. Section 2 addresses the production-reception frameworks in

social media based on the numbers of participants in an interaction, proposing a
simple dichotomy between two formats based on turn-taking. Section 3 is devoted
to the frameworks focussing on the types of producers and receivers. The point
of departure for this discussion is Goffman’s (1981) seminal framework, which
is critically examined and duly applied to the salient forms of communication in
social media as discussed in the existing scholarship. A number of general obser-
vations on the fundamental participatory roles will be made. Also, the construct of
“audience” is examined, and a few arguments are provided in favour of a view that
it is not a receiver category, in a technical sense. In Section 4, attention is paid to
the strategies enabling multiple (anonymous) receivers’ understanding in the light
of a number of proposals that concern the design of utterances given the prospec-
tive receiver’s needs.
2. Number of participants
One of the prevalent assumptions reverberating across multidisciplinary research

on Internet communication is that it is centred on one-to-many communication,
with one message (such as an advertisement or online article) being addressed
to vast audiences (see Webster and Lin 2002), or many-to-many communication,
which places emphasis on interactivity (e.g. Venkatesh, Dholakia and Dholakia
1996). However, in linguistic studies that focus specifically on the forms of inter-
action, alternative number-based participation structures are proposed.
Adopting the criteria of (a)synchronicity and the number of individuals
involved, Morris and Ogan (1996) differentiate between one-to-one asynchronous
communication (e.g. e-mail); many-to-many asynchronous communication (e.g.
Usenet or electronic bulletin boards); synchronous communication that can be
one-to-one, one-to-few, or one-to-many (e.g. chat rooms); and asynchronous com-
munication involving many-to-one, one-to-one, or one-to-many source-receiver
relationships (e.g. website browsing). The categories of one-to-one, one-to-few,
or one-to-many, many-to-one and many-to-many as forms of communication may
give rise to several misgivings. First of all, the distinction between ‘many’ and
‘few’ appears to be elusive, with no caesura between the two being determined.
Secondly, and more importantly, the types of frameworks do not appear to be
premised on the notion of turn-taking, as evidenced by ‘many’ at the produc-
tion end. This may mean many contributors making contributions independently
(which the authors most likely mean) and typically asynchronously, or it may
denote a number of individuals communicating a joint message, with one individ-
ual actually publicising it online. The truth of the matter is that each message/turn
shows one production end, however complex it may be (e.g. a YouTube film may
be constructed by a group of people but is ultimately submitted originally from

Authenticated
one account), and however many independent producers may be involved in an

asynchronous interaction (e.g. a topical strand on a forum).
Referring to notions previously mentioned by other researchers (Baron
1998, 2010; Herring 1996, 2007; Yates 2000), Bou-Franch, Lorenzo-Dus and
Garcés-Conejos Blitvich (2012) propose that participation in CMC can be divided
into three types: one-to-one communication, one-to-many interaction, and inter-
group discussions. Apart from indicating different numbers of participants at the
reception and production ends, these concepts appear to signal the nature of the
communicative activity at hand. Nonetheless, technically, there do not seem to be
any clear boundaries between communication, interaction and discussions, which
is perhaps a narrower term, compared to the first two. Like the previous one, this
proposal does not account for the turn-taking and changing participatory roles as an
interaction unfolds. Intergroup discussions are predicated on constant turn-taking
between a producer and the receivers, with each turn being communicated on a one-
to-many basis (e.g. message boards, see Arendholz, Ch. 5, this volume). Also, in the
‘one-to-many interactions’, one turn is a (multi-modal) message (e.g. a review of a
film on a personal website, a film on YouTube, see Johansson, Ch. 7, this volume;
or a textual message on Twitter, see Zappavigna, Ch. 8, this volume) submitted
for the consideration of receivers. Very rarely does an interaction terminate there,
though. Typically, receivers may duly take the floor as producers, for instance by
posting their comments under a video or a text, also in a one-to-many manner,
possibly giving rise to an ‘intergroup discussion’. Therefore, irrespective of the
seeming differences between them, both intergroup discussions and one-to-many
interactions can be captured under one umbrella term: multi-party interactions.
Essentially, most communication in social media can be seen as multi-party
interactions based on one-to-many messages, even though one-to-one interactions
are also possible (for instance, dyadic exchanges held on Facebook thanks to the
private messaging function, see Eisenlauer, Ch. 9, this volume). Admittedly, not
very much diversified, one-to-one interactions basically coincide with the classic
dyadic exchanges involving the speaker-hearer/sender-receiver dyad. In contrast,
multi-party interactions comprised of one-to-many messages involve many pro-
ducers of turns, with one user (or more in the case of joint production) taking the
floor at a time, and many individuals participating (asynchronously or synchro-
nously) at the reception end. The receivers may perform the same or different
participatory roles. Thus, the different genres of social media display potential for
multifarious participation frameworks.
An important note on the terminology is in order at this stage. Whilst many
authors resort to the labels ‘monologue’, ‘dialogue’ and ‘polylogue’ in reference
to different forms of online and offline interactions depending on the number of
participants, these terms are not endorsed here.
Monologue (originally from a Greek word that means ‘speaking alone’), which
is a useful term in literary studies to denote a longer stretch of text produced by a

Authenticated
66 Marta Dynel
character, may not be an adequate term in participation research, insofar as it fails

to bring out the necessary recipientship. As Bakhtin (1981: 279) rightly avers, “[t]
he dialogic orientation of discourse is a phenomenon that is of course a property
of any discourse”. Elsewhere, he states that “an essential (constitutive) marker of
the utterance is its quality of being directed at someone, its addressivity” (Bakhtin
1986: 333). Similarly, Duranti (1986: 243) states that “[t]alk, in fact, does not need
to be exchanged between parties for us to say that communication was coopera-
tively achieved. The mere presence of an audience socially constitutes and ratifies
the nature of a speech event (e.g. a sermon, a play, a class lecture, a story telling)”.
Importantly, the addressee/audience need not be physically present when a mes-
sage is actually produced, and may only be expected or imagined. This is saliently
the case with the ‘monologues’ in vlogging (e.g. Frobenius 2011, 2014), where
speakers typically talk to the camera eye, as if addressing the receivers, whom
the speakers expect to take part in the interaction asynchronously as viewers. All
messages in social media will qualify as the basis for an interaction between the
sender and at least one receiver even if the latter should not make any contribution.
Additionally, in practice, there are hardly any social media genres based on the
unidirectional communication of only one sender to the receiver(s), with replies
being possible. For example, (video) bloggers (see Heyd, Ch. 6, this volume) do
allow for comments and if at least one does appear in reference to one entry, the
single-sender communication is terminated.
Similar to ‘monologue’, the recent coinage ‘polylogue’ (Kerbrat-Orecchioni
2004; Marcoccia 2004; Garcés-Conejos Blitvich 2010; Lorenzo-Dus, Garcés-Cone-
jos Blitvich and Bou-Franch 2011; Bou-Franch, Lorenzo-Dus and Garcés-Conejos
Blitvich 2012) used in reference to a multi-party interaction may be considered
otiose. This is because it was originally proposed as the antonym of ‘dialogue’
understood as a dyadic interaction (Kerbrat-Orecchioni 2004). However, as the
etymology of the term ‘dialogue’ indicates (‘dia’ means ‘across’), it does cover
both dyadic interactions and those which involve more than two individuals (see
Chovanec 2011). As already indicated, this is the most salient type of participation
in new media interactions.
A pending query is how these different individuals in multi-party interactions
can be positioned in terms of their specific roles.
3. Types of participants
Goffman’s 1981 collection of previously published essays marks a watershed in

the studies of interaction across many disciplines. His writings are used as a depar-
ture point for studies of communication forms that Goffman cannot have envis-
aged, notably social media communication. One of the salient issues that Goffman
(1981) addresses in his various essays is a selection of participatory problems

Authenticated
and participant types representing the production and reception networks, which
have been frequently revisited (e.g. Levinson 1988; Dynel 2010, 2014a). When
his remarks scattered across his various essays are collated, Goffman may even
be regarded as the author of the first fully-fledged participation framework that
encompasses both ends of a multi-party interaction (see Dynel 2011a). The central
tenets deserve to be briefly revisited here so that the more recent postulates con-
cerning participation in social media can be critically examined.
3.1. Goffman’s participant types

Goffman advocates a division of participants into the production format (1981:
145, 146, 226) and the participation network (1981: 226), also called participation
framework (1981: 146). The latter label is somewhat misleading, since speakers
are also participants, which Goffman does acknowledge.
The production format decomposes the speaker into three distinct roles, viz.
the animator, the author and the principal. The animator is somebody who artic-
ulates utterances, in Goffman’s words, “the talking machine, a body engaged in
acoustic activity” (Goffman 1981: 144) or “the sounding box” (1981: 226). For
the purposes of written, not spoken, verbalisations, such as those in social media,
these metaphorical explanations might need to be broadened. Secondly, the author
is “someone who has selected the sentiments that are being expressed and the
words in which they are encoded” (Goffman 1981: 144) or “the agent who scripts
the lines” (Goffman 1981: 226). Thirdly, the principal is the party whose “position
is established by the words” (Goffman 1981: 144) or “the party to whose posi-
tion the words attest” (Goffman 1981: 226). Goffman (1981) points out that many
forms of talk can be mapped onto several persons, entertaining different (albeit
frequently meshing) roles. For instance, in fresh talk (i.e. when a person is pro-
ducing an utterance spontaneously, presenting his/her own ideas), the three roles
prototypically coincide, but when reading aloud from a script, the animator need
not simultaneously perform the other two roles that the production format encom-
passes. Neat as this tripartite division may appear at a glance, more complexities
loom large (see Levinson 1988; Verschueren 1999).
As originally conceptualised, the reception end encompasses several listen-
ers, hearers or recipients. These are the three labels which Goffman (1981) uses
interchangeably. Most importantly, hearers are divided into ratified and unratified
types (1981: 10, 226), which show subtypes.
Bystanders are unratified participants “whose access to the encounter, however
minimal, is itself perceivable by the official participants” (Goffman 1981: 132)
and who should “politely disavail” themselves of any opportunities to listen (1981:
132). Technically, therefore, bystanders who do not engage in an act of listening
are not participants. Nonetheless, a bystander may develop into an “inadvertent”
overhearer or an “engineered” eavesdropper (Goffman 1981: 9, 1981: 132). Over-

Authenticated
68 Marta Dynel
hearers “temporarily follow the talk, or catch bits and pieces of it, all without
much effort or intent” (1981: 132), whereas eavesdroppers “surreptitiously exploit
the accessibility” (1981: 132). This distinction is a bit vague, inasmuch is it not
based on one criterion. In contemporary research, the distinction between over-
hearers and eavesdroppers rests on the speakers’ (lack of) awareness of the unrati-
fied activity. Unlike overhearing, eavesdropping is surreptitious in the light of the
speaker’s obliviousness to an unratified hearer, let alone his/her listening in on
the talk (Dynel 2010, 2011a, 2011d; cf. Bell 1984, 1991; Clark and Carlson 1982;
Schober and Clark 1989; Clark and Schaefer 1987, 1992; Clark 1996).
As regards the ratified hearer types, Goffman discriminates between the
addressed recipient, i.e. “the one to whom the speaker addresses his visual atten-
tion and to whom, incidentally, he expects to turn over his speaking role” (1981:
133) and the rest of the official hearers who are not addressed (1981: 133). Whilst
the differentiation between addressed vs. not addressed ratified hearers is an
important one, Goffman’s conditions do not always hold. For example, direct eye
contact is quite frequently absent and the addressee need not be expected to reply,
both for instance in the case of the collective addressee (see Dynel 2010, 2014a
for discussion). While discussing these roles in face-to-face contexts, Goffman
(1981: 138, 144) rightly notes that talk manifests itself both in everyday conversa-
tions and in other communicative contexts, such as podium talk (a lecture, comedy
routines, poetic readings) or media broadcast. Therefore, listeners to talk need
not be fellow conversationalists but audiences, either addressed (lectures or TV
news) or unaddressed (staged plays) (Goffman 1981: 137–140). This claim indi-
cates audiences’ inherent ratified status, even though they do not normally take the
floor. Nevertheless, sometimes members of addressed audiences have the right to
take the floor (e.g. question time during rallies, or phone-in programmes). Also, as
already signalled, addressees need not (be obliged to) take the floor either, whilst
other ‘official hearers’ may (see Dynel 2010, 2011b, 2014a). Altogether, some
doubt may be cast on whether an audience is a distinct hearer type if it evinces the
different characteristics of the other roles, having a few subtypes. The audiences of
non-mass-mediated genres could potentially be conceptualised as addressees (and
in some contexts, secondary audiences may be official hearers). The unaddressed
audiences of mass-mediated genres need to be tackled independently.
In the context of programmes on television and on the radio, Goffman (1981)
talks about the audience that bifurcates into remote and live (studio) types. Goff-
man (1981: 138) dubs the former “imagined recipients”, based on the assumption
that broadcasters must talk to someone they cannot see, adopting a “conversa-
tional mode of address” (Goffman 1981: 138). Whether remote or live, the audi-
ence is “treated as if it were a ratified participant, albeit one that cannot assume
the speaking role” (Goffman 1981: 234). On the other hand, Goffman (1981: 83)
also suggests that audiences are “eavesdroppers” listening in on an actor who is
delivering a soliloquy in a theatrical performance in accordance with a script.

Authenticated
Words uttered by a character “are eternally sealed off from the audience, belong-
ing entirely to a self-enclosed, make-believe realm” (Goffman 1981: 139), even
if an actor performing on stage can appreciate the presence of an audience. While
the soliloquiser is “really talking to self when no one is around; we members of
the audience are supernatural out-of-frame eavesdroppers” (Goffman 1981: 83).
Goffman’s observation that audiences are eavesdroppers rests on the (misguided)
combination of the characters’ layer with the underlying production layer rendered
by the film crew, who construct characters’ conversations especially for viewers’
benefit. If the two layers are kept separate, the audience cannot be perceived as
being unratified hearers (see Dynel 2011d and references therein and the discus-
sion in Section 3.3).
Taking stock, ‘audience’ appears not to be a participatory role per se but a
folk notion that needs to be classified under specific hearer categories. However,
‘audience’ is also a prevalent construct in the literature on communication via the
Internet.
3.2. Audience in social media

Communication in social media, like any other form of communication, necessi-
tates the production and reception ends (see Herring 2007), which may be captured
under different labels. For example, Morris and Ogan (1996) differentiate between
producers and audiences. Generally, the prevalent term in the multi-disciplinary
literature, as indicated also by the title of this chapter, is audience (Roscoe 1999;
Webster and Lin 2002; Livingstone 2004), which seems to indicate the ‘many’
format at the reception end. Also, researchers tend to emphasise the multiplicity,
anonymity and asynchronous participation of receivers by using labels such as
“imagined ‘mass’ of ordinary users” (Burgess and Green 2008: 8), or “imagined
audience” (e.g. boyd 2010; Marwick and boyd 2010; Brake 2012; Litt 2012). As
Bell (1984, 1991) observes, in mass communication (on television), it is impossible
to determine all members of the target audience (in his parlance, the “addressees”).
The same applies to social media users at the reception end. Thanks to subscrip-
tions and the follower status of chosen accounts and fan pages, a single message
can reach countless receivers around the world. These numbers are further boosted
by the possibility of reposting. Consequently, “the imagined audience” (i.e. the
envisaged audience) “might be entirely different from the actual readers of a pro-
file, blog post, or tweet” (Marwick and boyd 2010: 97).
However, only those individuals who do access a given message can be con-
ceptualised as receivers, that is participants at the reception end. This is one of
the reasons why it may be wrong to conclude that the folk label ‘audience’ can
be considered a proper receiver type in social media (or in other communicative
contexts). When it is considered to be such, researchers end up making a seem-
ingly paradoxical claim that audiences can be producers, which is typical only of

Authenticated
70 Marta Dynel
social media interactions. Essentially, the same is possible in offline interactions

as audience members take the floor and become speakers, the producers of turns
(e.g. by asking questions during political rallies or sending letters to editors with a
view towards a prospective publication).
In this vein, a distinction is also sometimes made between active and passive
participants (lurkers), depending on whether or not they make use of the commu-
nicative tools a social media service (such as YouTube) affords in order to mark
their presence (Moor, Heuvelman and Verleur 2010; Bou-Franch, Lorenzo-Dus
and Garcés-Conejos Blitvich 2012). The passive vs. active distinction seems to
correspond to whether or not participants ever join the production format in an
interaction by contributing any messages, which they are entitled to do (typically,
having registered). A query one may raise is whether these roles should really be
assigned permanently, and thus having posted one comment, a user qualifies as an
active participant. Participatory roles ought to be ascribed to individuals for each
contribution/turn, not as a matter of a permanent property. In any case, the distinc-
tion tacitly indicates the importance of distinguishing between the production and
reception ends.
On the other hand, it is sometimes claimed that thanks to online interactivity,
the distinction between producer and receiver is blurred, and the standard commu-
nication process involving the production and reception ends is a thing of the past,
resulting in the “amalgamation of writer/author and reader/recipient into the fuzzy
concept of the user” (Bublitz 2012: 26; see also Eisenlauer 2013, 2014). However,
this is not to suggest that the different roles are actually conflated into one as
long as the process of turn-taking is accounted for. Social media users can swiftly
change the roles from receivers to producers and vice versa (just as participants
in everyday conversations do). That being said, the production and reception pro-
cesses, as well as the mechanics of turn-taking, are much more complex in social
media than in offline interactions.
The audience, that is technically the receivers of an original message (for exam-
ple an article, a video, or another multi-modal message, such as an Internet meme),
can communicate either with each other or with the producers or other receivers
(cf. Chovanec 2015; Adami 2015), thereby reversing the standard one-way pattern
of communication (Morris and Ogan 1996; Jucker 2003) typical of the traditional
media.5 Thus, two types of basic participation frameworks dependent on an audi-
ence’s response to a message might be distinguished: receiver-to-receiver (e.g.
two YouTube users having an exchange originally inspired by a YouTube video);
or receiver-to-producer (e.g. a YouTube user commenting on a film), which may
5
However, viewers’ contributions to some televised programmes are nowadays possible
thanks to phone calls on air, as well as texting and e-mails which are duly shown on the
screen.

Authenticated
further lead to the producer’s reply, possibly resulting in the receiver’s attaining
the status of a (co-)author. In any case, as they take the floor, the individuals pre-
viously occupying the reception end should no longer be considered receivers in
a technical sense but producers of the new messages (see Bublitz 2012), whatever
forms these may take.
The tendency to talk about audiences that contribute to interactions may stem
from the fact that some proportion of the original senders starting an interaction
have an institutional character (e.g. a TV channel’s fanpage on Facebook), or a
celebrity status acknowledged in the public sphere also offline (e.g. YouTube or
Twitter accounts of singers and bands) or developed from scratch in social media,
sometimes only within very small circles (as is the case of fashion bloggers or
YouTube presenters). However, from an interactional pragmatic perspective, in
social media, these widely recognised produces have essentially the same sta-
tus as ‘audiences’, the ordinary people, if they take the floor in order to interact
with them, usually by posting comments.6 Additionally, the dichotomy between
producers and audiences may be fostered by the message format, namely longer
texts or articles (e.g. on blogs), pictures or combinations thereof (e.g. memes on
humour websites), as well as videos, all of which are meant to attract many users,
as opposed to brief textual posts, which stand little chance of being so widely
recognised. By contrast, the notion of audiences is inapplicable, or at least less
likely applicable, in the cases when numerous users, whether anonymous (hiding
under nicknames) or not (providing what seem to be their real names and sur-
names) exchange numerous messages in the same format (for example posts on
discussion forums). Even if a great proportion of participants never take the floor
on such platforms, it is more difficult to conceive of them as audiences. Finally,
when messages, whatever their format may be, are not made widely available,
being restricted to chosen non-anonymous individuals, no matter how many (e.g.
Facebook friends), these individuals are not typically considered audiences.
On the strength of all these arguments, it is argued here, technical receiver
labels need to be proposed for the different individuals at the reception end (and
production end as well), taking Goffman’s lead.
3.3. Reception and production roles in social media

Various participatory configurations and frames of communication obtaining
in different social media formats have been proposed, for instance, with regard
to weblogs (Hoffmann 2012), YouTube (Dynel 2014b; Boyd 2014), Facebook
6
Needless to say, in practice, it is dubious whether addressing a Facebook message to a
world-famous person means that the message reaches the addressee and that the inter-
action envisaged by the sender does take place.

Authenticated
72 Marta Dynel
(Eisenlauer 2013, 2014), discussion boards (Haugh and Chang 2015), and online
broadcasts (Chovanec 2015). These frameworks are very complex and typically
capture only part of the possible communicative processes, which keep changing
as the social media platforms never cease evolving and allowing users new com-
municative options. It is impossible to propose an overarching framework for all
media platforms, given their complexity and diversity. Therefore, only general
observations are offered on how specific participatory roles can be assigned to
producers and receivers.
As regards the production format, one message may show different configura-
tions of the three production roles discerned by Goffman (1981). Echoing Goffman
(1981), Marcoccia (2004) provides a tripartite division of the individuals at the pro-
duction end in newsgroups: the transmitter (or animator), who is the physical source
of the message; the writer (or author), who formulates the message; and the enuncia-
tor (or principal), the participant to whose position the message attests. These roles
do not always coincide in one person. By the same token, discussing the production
side of weblogs, Hoffmann (2012: 62–63) argues that a blogger need not be the ani-
mator per se even though he/she “decides when, how and why to update her weblog’s
content (principal) and writes – and links – the actual content of her posts (author)”.
Further complications are possible in practically all social media genres. First
of all, the collective producer may be involved. The ultimate message (e.g. a You-
Tube video) is the product of a number of individuals, who share the three Goffma-
nian roles in different configurations. In addition, one message may be modified
and reposted in the same or a different social media service, thereby showing mul-
tiple animators and authors. Alternatively, users may be not the standard producers
but carriers of messages produced by others (Ayaß 2012: 6), thus being animators
but not authors.
As regards the conceptualisation of reception roles, Marcoccia (2004: 115)
lists “favored recipient”, “addressed recipient”, and the “eavesdropper” or “simple
reader”.7 This tripartite division does not appear to be well-founded. The explana-
tion provided for the first distinction is as follows: “[w]hen a message is addressed
[to someone other than the host], hosts are non-addressed ratified recipients, and
when a message is not addressed, hosts are main addressees or ‘favored recipi-
ents’” (Marcoccia 2004: 142). This seems to mirror Goffman’s distinction between
the addressee and non-addressed official hearer. Nonetheless, lack of overt forms
of address does not mean that an act of addressing has not taken place. In tune
with Goffman’s postulate, if there is just one type of hearer at the reception end,
it must be the addressee (see Dynel 2010, 2011a, 2014a). In the case mentioned
7
Apart from proposing the production and reception formats, Marcoccia (2004) lists
independently three participant types in newsgroups: simple readers (or eavesdrop-
pers), casual senders, and hosts.

Authenticated
by Marcoccia (2004), the host will be the addressee, which renders the ‘favoured
recipient’ category redundant. On the other hand, the conceptualisation of a ‘sim-
ple reader’ as an eavesdropper flies in the face of Goffman’s commonsensical defi-
nition of the eavesdropper as an unratified hearer. Marcoccia (2004) defines the
eavesdropper as an individual whose identity the speaker does not know, an anon-
ymous ratified participant that the sender of a message “knows to be listening”
(2004: 140). This seems to be a tenuous claim in the light of how this hearer type is
typically defined in interactional studies, namely: a hearer of whose presence the
speaker is oblivious, whom the speaker does not know to be listening, and whom
the speaker does not intend to listen (Dynel 2010, 2011a; and references therein).
The producer of any online message must be mindful of the fact that it will be
widely available and must ratify either selected individuals (e.g. followers, friends
or subscribers) or an infinite number of (imagined/expected) anonymous Inter-
net users (Burgess and Green 2008; Chovanec and Dynel 2015). The prospec-
tive receivers may be envisaged only to an extent depending on the content of
a message and its purpose. Anybody who does receive an online message made
available to him/her becomes a ratified participant, any personal, social or legal
restrictions notwithstanding. By contrast, those who do not access the message,
whether or not they are ratified, are not participants at all. This corresponds to the
widely supported claim that participatory roles are negotiated by the parties at the
production and reception ends.
Ratified receivers can be divided into addressees and non-addressed ratified
receivers. Importantly, forms of address will manifest themselves differently in
Internet communication (see de Oliveira 2013), compared to face-to-face conver-
sation. The addressee of a comment is determined in the light of a number of
criteria, most importantly: the content of the message (i.e. its relevance to another
user’s message), terms of address (e.g. the use of names and nicknames or second
person pronouns), as well as the position of the message in the structure of an
exchange, or simply its topical alignment to the previous posts. Some services
allow choosing an option ‘reply’ under each post, whereby the previous speaker
is addressed and one topical strand can develop with two or more participants
contributing to it by clicking the reply button (see Fritz, Ch. 11, this volume). As
any message is sent and addressed to a chosen individual, the addressee, or to cho-
sen individuals, the collective addressee (i.e. distributed addressees performing
the same function), other receivers of the comment (those who have access to it
and do process it) enjoy the status of non-addressed ratified participants, dubbed
collective third party, for instance. On the other hand, there may be no indication
whatsoever that the addressee role is granted only to select individuals. In such
cases, all receivers are ratified as the collective addressee. Also, visitors to any
website, blog, or participants in the different social networks who get their news-
feed or look for content made available to them are ratified participants who can
be conceived as the collective addressee.

Authenticated
74 Marta Dynel
Among the ratified participants in social media, a separate role might need to
be allocated to the interactants who, at a given point in an asynchronous interac-
tion, play the role of viewers of videos on YouTube or other video-sharing sites (as
opposed to being receivers of written texts). Needless to say, whether or not they
duly take the floor by posting comments or adding reply-to videos, their interac-
tional role in online interaction changes the moment they stop watching a video
(Dynel 2014b). Although it allows for purely textual interaction as the users sub-
mit their comments, YouTube has been dubbed a form of ‘post-television’ (Lister
et al. 2009; Tolson 2010), so it may be reasonable to define viewers watching You-
Tube videos (which may actually be previously televised programmes or screened
films) as the same receiver type as traditional media viewers are (Bou-Franch,
Lorenzo-Dus and Garcés-Conejos Blitvich 2012). The relevant postulates on the
participatory role of television viewers must be revisited in order to determine the
characteristics of video viewers in social media.
Regarding the hearer-related status of the television/film viewer, the claim pre-
vailing in the literature on traditional media talk is that audiences are overhearers or
an overhearing audience (see Dynel 2011d for references and criticism). As several
authors (Scannell 1991; Hutchby 2006; O’Keeffe 2006; Lorenzo-Dus 2009) have
rightly observed, the label ‘overhearer’ used with regard to ratified media audi-
ences is inaccurate, if not misleading. In this vein, Lorenzo-Dus, Garcés-Conejos
Blitvich and Bou-Franch (2011) are averse to the conceptualisation of (YouTube)
viewers as overhearers, insofar as the discourse is produced precisely to be broad-
cast to them. Viewers should then regarded as ratified hearers and, in tune with
Goffman’s (1981) ‘imagined recipients’ and Hutchby’s (2006) ‘distributed recipi-
ents’, they may be technically dubbed recipients, which is pertinent to the viewers
of both television broadcast and films, series and serials (Dynel 2011b, 2011c,
2011d), to which videos in social media can be compared. Therefore, engaged in
an act of watching YouTube videos, regardless of their character (e.g. vlogs, ama-
teur recordings, reposted broadcast or film extracts), YouTube viewers can also be
classified as recipients, i.e. ratified hearers external to the multimodal video mes-
sage. Such recipients may take the floor on a different level of interaction by post-
ing their comments (Herring 2011; Lorenzo-Dus, Garcés-Conejos Blitvich and
Bou-Franch 2011; Bou-Franch, Lorenzo-Dus and Garcés-Conejos Blitvich 2012;
Dynel 2014b), and even videos in reply (see Adami 2015).
On the whole, unratified receivers in social media are relatively rare. Publicis-
ing a turn, a speaker must be mindful of the fact that it will be widely available and
may potentially reach even those to whom the speaker does not specifically intend
to communicate a given message, that is whom the speaker does not wish to ratify
(albeit not explicitly making them unratified, either). What the producer actually
does by making his/her message publically available is then tacitly ratify potential
receivers. However, ratification can be altogether denied to users, whereby they
gain the status of non-participants to a given turn, or interaction taken as a whole.

Authenticated
This pertains both to potential receivers (when access to a message is restricted to

select users) and to senders whose messages may be blocked or deleted by other
users (e.g. an owner of a YouTube channel) or authorised bodies who are not par-
ticipants in the same frame of interaction (e.g. website owners or moderators).
Nevertheless, unratified receivers may be individuals who get access to a pass-
word-protected service (e.g. a Facebook account) and take part in an interaction,
assuming the role of an eavesdropper on each turn. Another possibility of unrati-
fied recipientship arises due to the fact that Internet communication allows seem-
ingly private messages to be easily replicated and transformed into public ones
via reposting (Chovanec and Dynel 2015). This happens when receivers take the
role of transmitters, technically animators, reproducing the messages authored by
others beyond the original communicative frame. This may involve the embedding
of interactional frames (Goffman 1981; Chovanec 2015), whereby a message first
communicated in one interaction becomes part of another interaction, whether on
the same or on a different social media platform. Through the multiplication effect,
a chain of several links of receivers is established as the message snowballs through
social media. From the perspective of the original/previous producer, the new
receivers ratified by the secondary producer, are indeed unratified. Such receivers
are either eavesdroppers if the original produces are unaware of the recipientship,
or overhearers, of whose receiver activity the producers are cognisant.
The wide availability of a large proportion of messages in social media and
the indeterminacy of the ratified recipientship are crucial for the formulation and
comprehensibility of the messages.
4. Being understood in social media
Speakers typically construct their utterances so as to be understood by their hear-

ers. Sacks, Schegloff and Jefferson (1974) are credited for having first proposed
the notion of recipient design in reference to how meanings are constructed for
hearers in everyday conversations. By “recipient design”, that is utterance design
inspired by the receiver’s needs, the authors mean “a multitude of respects in
which the talk by a party of a conversation is constructed or designed in ways
which display an orientation or sensitivity to the particular other(s) who are the
co-participants” (Sacks, Schegloff and Jefferson 1974: 727). Similarly, the notion
of audience design (Clark and Carlson 1982; Clark and Murphy 1983; Clark and
Schaefer 1992) refers to how speakers construct their utterances with the intention
of being understood by particular receivers. Clark and Murphy (1983) define audi-
ence design as the attitude the speaker has in tailoring his/her spoken utterances
for particular listeners, with whom the speaker shares momentary thoughts and
beliefs. For his part, Bell (1984, 1991) proposes the notion of audience design for
news media, according to which speakers adjust their speech styles to match the

Authenticated
76 Marta Dynel
audience’s needs. Adjusting procedures are compatible with the degree to which
the speaker recognises and ratifies the audience, in the context of the speaker’s lin-
guistic choices, and sociolinguistic variables. In this vein, much has been written
about recipient design for film discourse (e.g. Bubel 2008, Dynel 2011d).
Participants tend to assume that they share a vast amount of common ground,
which encompasses mutual knowledge, mutual beliefs, and mutual assumptions
(Clark and Brennan 1991; see also Lewis 1969; Clark and Marshall 1981; Clark
and Carlson 1982). On a different axis, common ground is comprised of com-
munal common ground held by members of a given society and personal com-
mon ground developed by individuals in their interactions, as well as the current
common ground, i.e. that generated by consecutive utterances within a particular
communicative exchange (Clark and Marshall 1981; Clark and Schaefer 1992;
Clark 1996). It is also claimed that common ground cannot be properly updated
without grounding (see Clark and Schaefer 1987, 1989; Clark and Brennan 1991).
According to Clark and Brennan (1991: 129), grounding “is the collective process
by which the participants try to reach” a mutual belief that “the partners [hearers]
have understood what the contributor meant”.
In the light of this review of theoretical literature, an important query arises
about how audience design, common ground and grounding relate to communica-
tion in social media. A statement can be ventured that these theoretical proposals
will apply to social media interactions just as they apply to the other domains
of human communication, whilst lack of understanding or misunderstandings
(whether or not discovered post factum) can happen in any form of communica-
tion.
Producers of messages in different social media genres have a variety of tools
at their disposal to get their meanings across, even if in many cases they know
very little about the receivers. What helps achieve this aim is various verbal
and non-verbal devices. For example, following in Clark and Carlson’s (1982)
footsteps, Frobenius (2014) lists devices that fall into five categories (linguistic
devices, conversational history, physical arrangement, gaze/gesture, and manner
of speaking) that help vloggers allocate participant statuses, some peculiar to vlog-
ging, and other ones possible also in other forms of interactions. In written for-
mats, producers can make use of emoticons, for instance. Disregarding such cues,
whether familiar with each other or anonymous, social media users capitalise on
common ground available to them (see Hoffmann 2012).
Although a great proportion of messages, whether multimodal or purely textual,
in social media are communicated to anonymous, unknown receivers, and may be
regarded as relying on a ‘leap of faith’ type of grounding, they are successfully
communicated based on communal common ground, as is the case also with media
discourse (e.g. programmes and films targeted at wide audiences). Needless to say,
the availability of chosen messages, and thus the numbers of the ‘imagined audi-
ence’ and the actual receivers will depend on the character of a given message.

Authenticated
Some messages, apart from depending on topical interests, rely heavily on various
forms of expertise in different areas (e.g. specialist knowledge or familiarity with
cultural or socio-political events), or knowledge of a jargon. Thus, users of cho-
sen services or websites (e.g. those devoted to humour and entertainment) tend to
develop their argot (abounding in neologisms, semantic shifts, or abbreviations,
among other things), which newbies visiting these websites may find practically
incomprehensible, thereby proving that they are not members of the ‘imagined
audience’ even though they are essentially ratified participants.
Acknowledgements
I would like to thank Wolfram Bublitz and Christian Hoffmann for their careful
reading of the previous version of this chapter and for their detailed comments and
suggestions.
References
Adami, Elisabetta
2015 What I can (re)make out of it: Incoherence, non-cohesion, and re-interpreta-
tion in YouTube video responses. In: Marta Dynel and Jan Chovanec (eds.),
Participation in Public and Social Media Interactions, 233–257. Amsterdam/
Ayaß, Ruth
2012 Introduction: media appropriation and everyday life. In: Ruth Ayaß and Cor-
nelia Gerhardt (eds.), The Appropriation of Media in Everyday Life, 1–15.
Amsterdam/Philadelphia: Benjamins.
Bakhtin, Mikhail
1981 The Dialogic Imagination: Four Essays. Ed. by Michael Holquist, transl. by
Caryl Emerson and Michael Holquist. Austin, TX: University of Texas Press.
Bakhtin, Mikhail
1986 Speech Genres and Other Late Essays. Transl. by Vern W. McGee. Austin, TX:
University of Texas Press.
Baron, Naomi
1998 Letters by phone or speech by other means: The linguistics of email. Language
and Communication 18: 133–170.
Baron, Naomi
2010 Discourse structures in instant messaging: The case of utterance breaks. Lan-
guage@Internet 7.
Bell, Allan
Bell, Allan
1991 The Language of News Media. Oxford: Blackwell.

Authenticated
78 Marta Dynel
Bou-Franch, Patricia, Nuria Lorenzo-Dus and Pilar Garcés-Conejos Blitvich,

2012 Social interaction in YouTube text-based polylogues: a study of coherence.
Journal of Computer-Mediated Communication 17: 501–521.
boyd, danah
2010 Social Network Sites as networked publics: affordances, dynamics, and impli-
cations. In: Zizi Papacharissi (ed.), Networked Self: Identity, Community, and
Culture on Social Network Sites. 39–58. New York: Routledge.
Boyd, Michael
2014 (New) participatory framework on YouTube? Commenter interaction in US
political speeches. Journal of Pragmatics 72: 46–58.
Brake, David
2012 Who do they think they’re talking to? Framings of the audience by social
media users. International Journal of Communication 6: 1056–1076.
Bubel, Claudia
2008 Film audiences as overhearers. Journal of Pragmatics 40: 55–71.
Bublitz, Wolfram
2012 Der ‘duale’ Internetnutzer: Ansätze einer dissoziativen Kommunikation. In:
Konstanze Marx and Monika Schwarz-Friesel (eds.), Sprache und Kommunika-
tion im technischen Zeitalter. Wieviel Technik (v)erträgt unsere Gesellschaft?,
Bühler, Karl
[1934] 1965 Sprachtheorie. Die Darstellungsfunktion der Sprache. Stuttgart: Gustav
Fischer. Transl. by Donald Fraser Goodwin 2011 Theory of Language: The
Representational Function of Language. Amsterdam/Philadelphia: Benjamins.
Burgess, Jean and Joshua Green
2008 Agency and controversy in the YouTube community. In: Proceedings IR 9.0:
Rethinking Communities, Rethinking Place Association of Internet Research-
ers (AoIR) Conference. Copenhagen: IT University of Copenhagen. http://
eprints.qut.edu.au/ 15383/1/15383.pdf.
Chovanec, Jan
2010 Online discussion and interaction: The case of live text commentary. In: Leon-
ard Shedletsky and Joan Aitken (eds.), Cases on Online Discussion and Inter-
action: Experiences and Outcomes, 234–251. Hershey, PA: IGI Global.
Chovanec, Jan
2011 Humour in quasi-conversations: Constructing fun in online sports journalism.
In: Marta Dynel (ed.), The Pragmatics of Humour across Discourse Domains,
243–264. Amsterdam/Philadelphia: Benjamins.
media. In: Marta Dynel and Jan Chovanec (eds.), Participation in Public and
Clark, Herbert
Clark, Herbert and Susan Brennan
1991 Grounding in communication. Perspectives on socially shared cognition. In:
Lauren B. Resnick, John M. Levine, and Stephanie D. Teasley (eds.), 127–
149. Washington: American Psychological Association.

Authenticated
Clark, Herbert and Thomas Carlson

1982 Hearers and speech acts. Language 58: 332–372.
Clark, Herbert and Catherine Marshall
1981 Definite reference and mutual knowledge. In: Aravind K. Joshi, Bonnie Web-
ber, and Ivan Sag (eds.), Elements of Discourse Understanding, 10–63. Cam-
bridge: Cambridge University Press.
Clark, Herbert and Gregory Murphy
1983 Audience design in meaning and reference. In: J. F. LeNy and Walter Kintsch
(eds.), Language and Comprehension, 287–299. Amsterdam: North-Holland
Publishing.
1987 Concealing one’s meaning from overhearers. Journal of Memory and Lan-
guage 26: 209–225.
1989 Contributing to discourse. Cognitive Science 13: 259–294.
1992 Dealing with overhearers. In: Herbert Clark, Arenas of Language Use, 248–
273. Chicago: University of Chicago Press.
de Oliveira, Sandi Michele
2013 Address in computer-mediated communication. In: Susan C. Herring; Dieter
Stein; Tuija Virtanen (eds.), Pragmatics of Computer-Mediated Communica-
tion, 291–313. Berlin/Boston: de Gruyter.
de Saussure, Ferdinand
[1916] 1974 Course in General Linguistics. Trans. by Wade Baskin. London: Fontana/
Collins.
Deuze, Mark
2006 Participation, remediation, bricolage: considering principal components of a
digital culture. Information Society 22: 63–75.
Duranti, Alessandro
1986 The audience as co-author: An introduction. Text 6: 239–247.
Dynel, Marta
2010 Not hearing things. Hearer/listener categories in polylogues. mediAzioni 9. Avail-
able at: http://www.mediazioni.sitlec.unibo.it/images/stories/PDF_folder/
document-pdf/2010/dynel_2010.pdf
Dynel, Marta
2011a Revisiting Goffman’s postulates on participant statuses in verbal interaction.
Language and Linguistics Compass. Sociolinguistics 5/7: 454–465.
Dynel, Marta
2011b Entertaining and enraging: The functions of verbal violence in broadcast
political debates. In: Villy Tsakona and Diana Popa (eds), Studies in Political
Humour, 109–133. Amsterdam/Philadelphia: Benjamins.
Dynel, Marta
2011c Two communicative levels and twofold illocutionary force in televised politi-
cal debates. Poznan Studies in Contemporary Linguistics 47(2): 283–307.
Dynel, Marta
2011d ‘You talking to me?’ The viewer as a ratified listener to film discourse. Journal
of Pragmatics 43: 1628–1644.

Authenticated
80 Marta Dynel
Dynel, Marta
2014a On the part of ratified participants: Ratified listeners in multi-party interaction.
Brno Studies in English 40: 27–44.
Dynel, Marta
2014b Participation framework underlying YouTube interaction. Journal of Pragmat-
ics 73: 37–52.
Eisenlauer, Volker
2013 A Critical Hypertext Analysis of Social Media.
The True Colours of Facebook. London/New York: Continuum.
Eisenlauer, Volker
2014 Facebook as a third author – (Semi)automated participation framework in
2011 Beginning a monologue: The opening sequence of video blogs. Journal of
Pragmatics 43(3): 814–827.
2014 Audience design in monologues: How vloggers involve their viewers. Journal
Garcés-Conejos Blitvich, Pilar
2010 The YouTubification of politics, impoliteness and polarization. In: Rotimi
Taiwo (ed.), Handbook of Research on Discourse Behavior and Digital Com-
munication: Language Structures and Social Interaction, 540–563. Hershey,
PA: IGI Global.
Goffman, Erving
1963 Behavior in Public Places: Notes on the Social Organization of Gatherings.
New York: The Free Press.
Goffman, Erving
1981 Forms of Talk. Philadelphia: University of Philadelphia Press.
Goodwin, Charles
1981 Conversational Organization: Interaction Between Speakers and Hearers.
New York: Academic Press.
Haugh, Michael and Wei-Lin Melody Chang
2015 Troubles talk, (dis)affiliation and the participation order in Taiwanese-Chinese
online discussion boards. In: Marta Dynel and Jan Chovanec (eds.), Partici-
pation in Public and Social Media Interactions, 99–133. Amsterdam/Philadel-
phia: Benjamins.
Herring, Susan
1996 Two variants of an electronic message schema. In: Susan Herring (ed.), Com-
puter-Mediated Communication: Linguistic, Social and Cross-Cultural Per-
spectives, 81–106. Amsterdam/Philadelphia: Benjamins.
Herring, Susan
2007 A faceted classification scheme for computer-mediated discourse. Language@
Internet 4.
Herring, Susan
2010 Who’s got the floor in computer-mediated conversation? Edelsky’s gender
patterns revisited.” Language@Internet 7.

Authenticated
Herring, Susan
2011 Computer-mediated conversation, Part II: Introduction and overview. Lan-
guage@Internet 8.
Herring, Susan
2013 Discourse in web 2.0: familiar, reconfigured and emergent” In: Deborah
Tannen and Anna Marie Trester (eds.), Discourse Web 2.0: Language in the
Media, 1–26. Washington, DC: Georgetown University Press.
Herring, Susan, Dieter Stein and Tuija Virtanen
2013 Pragmatics of Computer-Mediated Communication. Berlin/Boston: de Gruyter.
Hoffmann, Christian
2012 Cohesive Profiling. Meaning and Interaction in Personal Weblogs. Amster-
Hutchby, Ian
2006 Media Talk: Conversation Analysis and the Study of Broadcasting. Glasgow:
Open University Press.
Jakobson, Roman
1960 Closing statement: linguistics and poetics. In: Thomas Sebeok (ed.), Style in
Language, 350–377. Cambridge, MA: MIT Press.
Johansson, Marjut
2014 Reading digital news: Participation roles, activities, and positionings. Journal
Jucker, Andreas
2003 Mass media communication at the beginning of the twenty-first century.
Dimensions of change. Journal of Historical Pragmatics 4: 129–148.
Kerbrat-Orecchioni, Catherine
2004 Introducing polylogue. Journal of Pragmatics 36(1): 1–24.
Levinson, Stephen
1988 Putting linguistics on a proper footing: Explorations in Goffman’s participa-
tion framework. In: Paul Drew and Anthony Wootton (eds.), Goffman: Explor-
ing the Interaction Order, 161–227. Oxford: Polity Press.
Lewis, David
1969 Convention. A Philosophical Study. Cambrige MA: Harvard University Press.
Lister, Martin, Jon Dovey, Seth Giddings, Iain Grant and Kelly Kieran
2009 New Media: A Critical Introduction. London: Routledge.
Litt, Eden
2012 Knock, Knock. Who’s there? The imagined audience. Journal Broadcasting &
Electronic Media 56(3): 330–345.
Livingstone, Sonia
2004 The challenge of changing audiences: or, what is the audience researcher to
do in the age of the internet? European Journal of Communication 19(1):
75–86.
Lorenzo-Dus, Nuria
2009 Television Discourse. Analysing Language in the Media. Basingstoke/New
York: Palgrave Macmillan.
Lorenzo-Dus, Nuria, Pilar Garcés-Conejos Blitvich and Patricia Bou-Franch
2011 On-line polylogues and impoliteness: The case of postings sent in response to
the Obama Reggaeton YouTube video. Journal of Pragmatics 43: 2578–2593.

Authenticated
82 Marta Dynel
Marcoccia, Michel
internet newsgroups. Journal of Pragmatics 36: 115–145.
Marwick, Alice and danah boyd
2010 I tweet honestly, I tweet passionately: Twitter users, context collapse, and the
imagined audience. New Media & Society 13: 96–113.
Moor, Peter, Ard Heuvelman and Ria Verleur
2010 Flaming on YouTube. Computers in Human Behavior 26: 1536–1546.
Morris, Merrill and Christine Ogan
1996 The Internet as a mass medium. Journal of Computer-Mediated Communica-
tion 1(4).
O’Keeffe, Anne
2006 Investigating Media Discourse. London: Routledge.
Poole, Marshall and Michele Jackson
1993 Communication theory and group support systems. In: Leonard Jessup and
Joseph Valacich (eds), Group Support Systems: New Perspectives, 281–293.
New York: Macmillan.
Roscoe, Timothy
1999 The construction of the World Wide Web audience. Media, Culture & Society
21(5): 673–684.
Sacks, Harvey, Emanuel Schegloff and Gail Jefferson
1974 A simplest systematics for the organisation of turn-taking for conversation.
Language 50: 696–735.
Scannell, Paddy
1991 Introduction: The relevance of talk. In: Paddy Scannell (ed.), Broadcast Talk,
1–13. London: Sage.
Schober, Michael and Herbert Clark
1989 Understanding by addressees and overhearers. Cognitive Psychology 21: 211–
232.
Tolson, Andrew
2010 A new authenticity? Communicative practices on YouTube. Critical Discourse
Studies 7(4): 277–289.
Venkatesh, Alladi, Ruby Roy Dholakia and Nikhilesh Dholakia
1996 New visions of information technology and postmodernism: Implications
for advertising and marketing communications. In: Walter Brenner and Lutz
Kolbe (eds.), The Information Superhighway and Private Households, 319–
338. Heidelberg: Physica-Verlag.
Verschueren, Jef
1999 Understanding Pragmatics. London: Hodder Arnold.
Webster, James G. and Shu-Fang Lin
2002 The internet audience: Web use as mass behavior. Journal of Broadcasting and
Electronic Media 46: 1–12.
Yates, Simeon
2000 Computer-mediated communication: The future of the letter? In: David Barton
and Nigel Hall (eds.), Letter Writing as a Social Practice, 233–251. Amster-

Authenticated
4. Publicness and privateness
Birte Bös and Sonja Kleinke
Abstract: This contribution deals with the question of how social media have
shifted and blurred the boundaries between public and private. It outlines attempts
to tackle the notoriously fuzzy notions of publicness and privateness from a social
media perspective and brings together models from communication science and
pragmatics, thus shedding light on the different dimensions of publicness and pri-
vateness in social media contexts. These notions necessarily need to be concep-
tualized in a flexible way in order to take account of the fluid perceptions and
practices users have developed of these issues.
1. Introduction
As do the other contributors to this handbook, we adopt a broad perspective in the

definition of social media, or “the social Web”, as it is sometimes also called (e.g.
Thelwall 2011: 151). Hidden behind these terms are “rich genre ecologies” (Hoff-
mann, Ch. 1, this volume) which display an incredibly high degree of complexity.
What they have in common is their (technical) potential for interactivity which
allows users to create and share content online. The effects of the communicative
opportunities opened up by the new technologies have been discussed by layper-
sons, media representatives and researchers alike. Baym and boyd declare:
Old practices and patterns continue to thrive in new media. However, social media
blur boundaries between presence and absence, time and space, control and freedom,
personal and mass communication, private and public, and virtual and real, affecting
how old patterns should be understood and raising new challenges and opportunities for
people engaging others through new technologies (2012: 320).
There is a broad agreement that established (idealized) public/private dichotomies

no longer hold true, although certainly not everybody would agree with Facebook
founder Mark Zuckerberg’s assertion that “[t]he rise of social networking online
means that people no longer have an expectation of privacy” and that “privacy [is]
no longer a ‘social norm’” (Johnson 2010). Indeed, privacy concerns are a ques-
tion of high societal relevance, and the media provide ample evidence of the anxi-
eties and controversial debates triggered by the social media. Yet, this is a phenom-
enon often observed at the introduction of new technologies, which – according to
Jarvis (2011) and Tannen (2013: 101), among others – can be traced back as far as
the introduction of printing in the 15th century. Such technologies essentially force
DOI 10.1515/9783110431070-004
Authenticated
84 Birte Bös and Sonja Kleinke
“human society to redefine the boundaries between the public and private and to
re-conceptualize the concept of personal privacy” (Yao 2011: 111).
Public and private and their related terms publicity/publicness and privacy/
privateness cover extremely fuzzy and complex notions, both in everyday lan-
guage and as termini technici. Even their basic semantic definitions, as provided
by the Oxford English Dictionary (OED), already display a range of meanings. In
the case of public, they include, for example, openness to observation and general
accessibility, a relation to the community or nation, and a focus on official posi-
tions (OED, s.v. public, adj). In the case of private, semantic components such
as restricted access, the involvement of one particular person or group of peo-
ple, and the non-official capacities of persons, are highlighted (OED, s.v. private,
adj). Derived from the adjectives, publicity/publicness and privacy/privateness
are often treated as synonyms with certain discipline-related preferences, privacy
being more common in communication science (CS) and privateness being more
often found in linguistic contexts.1
Of course, it is not just the terms themselves that differ; they also lack consis
tent definitions (Papacharissi 2010: 35) and their conceptualizations are often per-
ceived as “contested and messy” (boyd 2010: 40, in relation to ‘public’), resulting
from different “theoretical languages or universes of discourse, each with its own
complex historical cargo of assumptions and connotations” (Weintraub 1997: 3).
As the basic semantic definitions show, publicness and privateness can only be
conceptualized in relation to each other. Yet, we also have to accept that we are
not dealing with “a single paired opposition, but a complex family of them, nei-
ther mutually reducible nor wholly unrelated” (Weintraub and Kumar 1997: xii).
Thus, in view of the wide range of social media formats and interactions and the
diverse perceptions of their users, a flexible approach to issues of publicness and
privateness is required.
For a start, we will outline some major characteristics of social media from
the public/private perspective. With roughly 40 % of the world population having
internet access today (http://www.internetlivestats.com/internet-users/), technical
accessibility to social media – and therefore to “new public spaces” (Bateman,
Pike and Butler 2010: 81) – is generally high. The access points to this “virtual
public” (2010: 81) are located in “a private media environment […] within the
individual’s personal and private space” (Papacharissi 2010: 306). However, with
more and more users accessing the internet via their smartphones, the private is
increasingly removed from the domestic setting in a form of “mobile privatiza-
tion” (Papacharissi 2010: 133, in reference to Raymond Williams’ (1974) term).
Instead, usage patterns are determined by sociodemographic aspects such as gen-
1
This contribution adopts privateness/publicness as flexible umbrella terms, but it will
follow the usage of the respective authors, where necessary.

Authenticated
Publicness and privateness 85
der, age, socioeconomic and ethnic background, and factors such as experience
and skills, self-efficacy and motivation, which we will not go into here.2
Accessibility clearly goes beyond technical aspects. Social media, in a very
basic conceptualization, provide a floor for the consumption and distribution of
public and private information (see Table 1). On the one hand, they allow individ-
uals to gain unprecedented access to public information of any sort at any time.
Beyond the mere informational function, the Web, with its public discussion spaces,
simultaneously invites users to participate in public discourse in ways that were
not open to them before (cf. Androutsoupolous 2013: 49, see also Landert, Ch. 2,
this volume), providing them with a ‘new visibility’ in the public sphere (Thomp-
son 2005).3 On the other hand, by distributing private, often personal information,
individuals gain visibility in online networks (of various sizes and densities) and
with that increase their chances to access private information from other users (see
also Arendholz 2013: 111 on “interactional accessibility” and Androutsopoulos
2013: 48 on “low entry-requirements”).
Table 1. Access and visibility in the social media
Social media
Individuals gaining access Individuals gaining visibility
Access to public information Participation in public discourse
Access to private information Distribution of private information
The ever-growing number of users essentially magnifies the impacts of accessibil-

ity and visibility. While some social media platforms rely on “exclusive member-
ship”, e.g. ASmallWorld, which requires peer recommendation, blogs with access
restrictions in corporate and organizational spaces (cf. Puschmann 2013: 83) or
online medical support groups (cf. Kleinke 2015), many allow for free access and
aim for as many users as possible. In fact, some virtual communities have attracted
millions of users, among them well-known Western representatives like Facebook,
WhatsApp, Instagram and Youtube, but also Asian platforms like Tencent QQ or
WeChat, which can boast more than a billion registered users.4
2
But see e.g. boyd (2007), Hargittai (2008) and Perrin (2015) for findings and further
references on this subject.
3
But see Papacharissi (2010: 19) on emancipation in private rather than public spaces
and Johansson (in press) for a critical account of the liberating effects of the internet.
Thurlow (2013: 243–244) speaks of “synthetic media” and “pseudo-sociality”.
4
See https://en.wikipedia.org/wiki/List_of_virtual_communities_with_more_than_100_
million_active_users (last accessed 5 September 2016).

Authenticated
With regard to social media content, the following properties (originally

defined for Social networking sites) have important impacts on publicness/pri-
vateness considerations:
Persistence: Digital expressions are automatically recorded and archived.
Replicability: Digital content is easily duplicated.
Scalability: The potential visibility of digital content is great.
Searchability: Digital content is often accessible through search engines.
(boyd and Marwick 2011: 9)
Clearly, these characteristics pose a challenge to users’ control over (potentially

private) data, but they do not grant large-scale public audiences per se. As Baym
and boyd (2012: 322) point out, social media have public and quasi-public quali-
ties, as the potential for visibility is not necessarily fully exploited. Much of what
users communicate online is, in fact, only noticed by a small number of people,
and it is hard to determine “who is out there and when” (Baym and boyd 2012:
323). Additionally, there might be different levels of “perceived publicness”, i.e.
“the degree to which users believe that others have unrestricted access to their
information” (Bateman, Pike and Butler 2010: 81–82; cf. also Puschmann 2013:
94, who observes a “discrepancy between popular perception and everyday use” in
blogging). The potential conflicts arising from these communicative constellations
have been much debated. Quite generally, Petronio states that “individuals must
balance the usefulness of privacy with the utility of openness” (2000: 37). From
the social media perspective, the conflict is that “users desire social interaction and
connectivity and disclosing information plays an essential role; yet users may not
wish to have their information made publicly accessible to an unknown audience”
(Bateman, Pike and Butler 2010: 79).
One of the best available ways to explore how actual users balance private-
ness and publicness is through an analysis of language use. With their complex
technical constellations and participation structure, social media combine “ele-
ments of interpersonal and mass communication” (Mendelson and Papacharissi
2010: 252), again blurring boundaries. In fact, language use in social media has
often been associated with the immediacy and conceptual orality of casual face-
to-face communication. However, Crystal (2001, 2006) – though still adhering to
an overarching notion of Netspeak – already discussed scalar phenomena on an
oral/written axis more than a decade ago. More recent studies on specific social
media formats (e.g. Hoffmann 2012 on conceptual writtenness in personal blogs
and Sindoni 2013 on the multimodality of videochats, blogs and YouTube com-
ments) provide a far more differentiated picture. It is therefore vital to reconsider
the notions of orality and literacy and of immediacy and distance in the light of
publicness and privateness.
Given their prominence, the special dynamics of their communicative con-
stellations and practices, and their diversity and multi-modal complexity, it is not

Authenticated
surprising that social media have attracted research in a broad range of academic
fields and inspired interdisciplinary perspectives. This contribution will focus on
two major, well-established research strands: linguistic, (micro-)pragmatic inves-
tigations and studies in the field of communication science (CS).5 As Sections 2
and 3 show, these research strands have provided different contributions to the
modelling and analysis of privacy and publicness in social media. While up to
now, pragmatic conceptualizations have essentially been text- and context-based,
communicative science shapes its models with a strong user- and context-orienta-
tion, including psychological, social scientific and marketing-oriented aspects of
privacy and publicness. Section 4 demonstrates where the models and concepts of
the two disciplines meet and benefit from each other. It summarizes findings from
selected case studies which illustrate attempts to link CS and pragmatic findings as
well as briefly discussing authenticity, reliability and research ethics at the inter-
face of both strands.
2. Publicness and privateness through the lens of communication

science
2.1. Extending the dichotomy

Notions of public and private – and the supposed opposition between them – can
be encountered “in contexts that run from the most abstract theorizing to the most
practical and immediate arenas of life” (Weintraub and Kumar 1997: xi). Our goal
here is to explore interpersonal aspects of the construction of privacy and public-
ness in social media contexts.6
The categories of ‘public’ and ‘private’ play an essential role in the way com-
munication shapes interpersonal relations, as they serve to “qualify and distin-
guish manifested choices in everyday life” (Papacharissi 2010: 25). Following
Weintraub (1997: 5), they can be defined by the two basic criteria of visibility
(hidden, withdrawn, i.e. private vs. open, accessible, i.e. public) and collectiv-
ity (pertaining to an individual, i.e. private vs. a collectivity of individuals, i.e.
public), which were already displayed in the OED definitions (see Section 1) and
reflect entrenchment in language use. Publicness and privateness are best under-
stood as overlapping, scalar notions, which are shaped by cultural, historical and
5
We understand communication science as the social study of communication, which
investigates the fundamental processes, codes, functions, and contexts of human com-
munication (cf. Berger, Roloff and Roskoss-Ewoldsen 2010).
6
For the purposes of this contribution, we necessarily have to take a limited view. For
discussions about public and private spaces based on Habermas’ influential ideas see
Schmidt (2011a: 97), Cooke (2011: 300) and Papacharissi (2014: 152–155).

Authenticated
linguistic contexts (Papacharissi 2010: 25, 35, see also boyd and Marwick 2011:
25). Their conceptualization as scalar is also useful from a linguistic perspective.
Given the two poles, Papacharissi (2010, drawing on Wolfe 1997: 196) suggests a
further addition, which breaks up the dichotomy and extends it into a trichotomy
with a third component of sociality, “a realm of distinct publics”, defined in the
following way:
These publics are partially collective in that they share norms and common goals, but
the basis of this collectivity is social, not political. They are also partially private, in that
they may shield the individual from the public realm in ways that enable the develop-
ment of self-interest and private identity. (Papacharissi 2010: 49–50)
This extension proves fruitful from a social media perspective, where it surfaces
as one of the special characteristics of “networked publics”, defined not only as
“(1) the space constructed through networked technologies”, but also as “(2) the
imagined community that emerges as a result of the intersection of people, tech-
nology, and practice” (boyd and Marwick 2011: 7),7 with potential audiences fixed
in users’ profiles as “privileged publics” (boyd 2010: 43). In this context, Lange
(2008), investigating how YouTube users manipulate technical features and social
mechanisms in negotiating membership in social networks, provides a further dif-
ferentiation of “publicly private” and “privately public” behaviour. “Publicly pri-
vate” behaviour is displayed in YouTubers’ exposure of private identities hardly
accessed by other users, whereas “privately public” behaviour relates to limited
private information which is widely accessible (Lange 2008: 361).
Sociality furthermore finds a counterpart in the notion of communities of prac-
tice, a concept widely used in sociolinguistics and pragmatics, which manifests
itself in research on computer-mediated discourse (CMD) in the notion of Virtual
Communities (first introduced by Rheingold 1993):8
A community of practice is a collection of people who engage on an ongoing basis
in some common endeavor. Communities of practice emerge in response to common
interest or position, and play an important role in forming their members’ participation
in, and orientation to, the world around them. They provide an accountable link, there-
fore, between the individual, the group, and place in the broader social order, and they
provide a setting in which linguistic practice emerges as a function of this link. (Eckert
2006)
The shared idea of the social motivation and “the semi-private and semi-public
social needs of these multiple private publics” (Papacharissi 2010: 50) thus pro-
7
For an outline of the development and examples of internet communities cf. Johnston
(2014: 20−28).
8
For an empirical study on the potential construction of virtual communities in two
online contexts and a critical discussion see also Herring (2004: 5−6).

Authenticated
vides a bridge between the two approaches under scrutiny in this contribution.
Social media have not simply catapulted their users into a post-privacy era, as it
is sometimes decried, but have shaped new hybrid spaces “neither convention-
ally public nor entirely private” (Papacharissi and Gibson 2011: 75). These spaces
of sociality require giving up privacy to some extent (Papacharissi 2010: 133),9
but fulfil users’ needs to connect and interact, in that they replace former “public
spaces of political expression” with “private spaces” from which they broadcast
“individual opinions on public and private affairs” (cf. Papacharissi 2010: 38 on
new forms of civic engagement, and also Johansson 2015, in press).
2.2. Dimensions of privacy

In CS research (just as in laypersons’ debates), the concept of privacy, as “a basic
human need” which “leverages well-being” (Trepte and Reinecke 2011a: v), has
clearly attracted much more attention than its counterpart.10
Trepte and Reinecke (2011b) outline three dimensions of privacy which prove
particularly relevant from a social media perspective: informational privacy, social
privacy and psychological privacy (cf. also Debatin 2011: 48). To date, informa-
tional privacy has been the primary focus in social media research. It relates to
individuals’ control over the kind and amount of information about them available
to others online, which – given the properties of persistence, replicability, scal-
ability and searchability – are strongly influenced by technical affordances and
users’ ability and willingness to handle these efficiently. Social privacy covers
users’ possibilities to manage their interpersonal relationships online. Psycholog-
ical privacy refers to users’ perceived potential of sharing contents of their choice
(thoughts, feelings, etc.) without being censored (Trepte and Reinecke 2011b: 61).
Additionally, a physical, spatial dimension is postulated in several models (cf.
Loosen 2011: 206), which also proves relevant from a social media perspective:
a spatial meaning of ‘private’, which takes account of the fact that social media
communication, both thematically private and public, is typically created in a pri-
vate space and returns to the private space (Papacharissi 2010: 37–38).
This multi-dimensional conceptualization of privacy sheds light on a phe-
nomenon described by Barnes (2006) as the “privacy paradox” of social media.
9
Drawing on Zhang et al. (2010), Papacharissi and Gibson (2011: 80) report a “newer
paradigm for sociality” equating “disclosure with being social”.
10
But cf. Schmidt (2011a: 107−133) on “personal publics” as essential characteristics of
social media, which can be distinguished from “traditional publics” at three levels of
how information is handled: “selection by personal relevance”, distribution to explicit
network ties and communication in a “conversational mode”. They prove especially
relevant for Twitter interaction (Schmidt 2014: 4).

Authenticated
It relates to the apparent contradiction that the privacy concerns of users do not
necessarily have an impact on their actual online behaviour.11 Based on a range of
empirical studies, Trepte and Reinecke assume that users may indeed perceive a
lack of control regarding informational privacy. However, “the Social Web seems
to foster feelings of psychological and social privacy” (2011b: 65), the subjective
feelings of anonymity and intimacy lower users’ inhibitions, and thus private dis-
closure is often encouraged in such online contexts.
2.3. (Self-)disclosure, privacy risks and privacy management

The disclosure of private information in social media is described by Debatin as “a
kind of bargaining process” (2011: 56), with the major motivation – “the expected
benefits of social networking” (2011: 56) – often outweighing potential risks
involved in optional self-disclosure. Social networking sites (SNSs) are generally
considered to have greatest potential for conflicts in this regard (Thelwall 2011:
255) and will therefore be our focus here. On SNSs, dynamic disclosure by active
participation is vital to enhance the social capital gained, whereas mere consump-
tion of other contributions leads to social isolation (Ellison et al. 2011: 24). The
main loci of privacy management on SNSs are users’ profiles.
The information users are willing to reveal on SNSs can be characterized with
regard to its amount, depth and duration (Bateman, Pike and Butler 2010: 80). As
regards depth, Bateman, Pike and Butler (2010: 90) identify three layers of disclo-
sure: peripheral disclosure (e.g. name, age and gender of the user), intermediate
disclosure (e.g. religious views, political orientation) and core disclosure (includ-
ing emotions, feelings, values, etc.). Self-disclosure is often encouraged by site
settings (Joinson et al. 2011: 34–35) and participants often provide identifiable,
detailed information in their personal profiles (Ellison et al. 2010: 130), on shared
past referents (e.g. schools attended), their interests (favourite books, films, etc.),
and contact options (even including their offline location) (Lampe, Ellison and
Steinfield 2007: 438). Users post and share photographs depicting themselves and
other people, provide intimate information in their status updates and even dis-
cuss personal relationships in semi-public environments such as walls or include
reports about illegal activities (Joinson et al. 2011: 34−35). Lampe, Ellison and
Steinfield’s study of Facebook profiles reveals that the kind and amount of infor-
mation shared have an impact on users’ connectedness, as reflected in their friends
11
But cf. Dienlin and Trepte (2015), who confirm these findings but discover a link
between more specified privacy attitudes and privacy behaviour on Social networking
sites using a more refined methodology which includes informational, social and psy-
chological privacy and their impact on participants’ online behaviour.

Authenticated
lists. Providing more information helps to reduce “the costs of connecting” and
promotes larger networks (Lampe, Ellison and Steinfield 2007: 437).
On the other hand, the kinds of network connections have an impact on the
management of private information. Petronio, in her description of a general
rule-based privacy management system, points out that “[s]trong linkages reflect
a higher level of responsibility for managing disclosed information” (2000: 41).
This might help to further explain why the loose networks often found in social
media sometimes promote disclosure more than face-to-face communication in
high-density networks. Still,
[t]he relation between privacy and a person’s social network is multi-faceted. In certain
occasions we want information about ourselves to be known only by a small circle of
close friends, and not by strangers. In other instances, we are willing to reveal personal
information to anonymous strangers, but not to those who know us better. (Gross and
Acquisti 2005: 2)
Simply withdrawing from interaction by refraining from frequenting SNSs is obvi-

ously not an option for users who want to profit from the social benefits of SNSs.
Thus, they need to rely on other mechanisms of privacy management (cf. Debatin
2011: 49), negotiating ethical self-regulation, i.e. “a context-appropriate flow of
personal information” (2011: 50), within the legal framework and technical set-
tings of SNSs. For that purpose, users might resort to self-censorship, “perform-
ing for the lowest common denominator so as to produce a performance that will
comply with the expectations of the broadest possible audience” or to “polysemic
performances, presentations of the self that contain layers of meaning, signifying
different impressions to various audiences” (Papacharissi and Gibson 2011: 81).
In addition, privacy management in social media is not just a matter of deciding
which information users would like to make visible to others. While contributions
by other users might often be welcome (boyd 2010: 43), the possibilities of co-cre-
ation – an essential characteristic of Web 2.0 – have sometimes been considered
a “greater risk than disclosure by users themselves”, as social media encourage
“unfettered sharing of personal information that intrudes upon other users” (Join-
son et al. 2011: 35−36).
Furthermore, the technical affordances of most SNSs allow for the fine-tun-
ing of individual privacy settings, with options ranging from “truly public” to
“semi-public” (boyd 2010: 43). However, research shows that a majority of users
keeps the default settings (cf. e.g. Light and McGrath 2010: 301 or Zimmer and
Proferes 2014: 173 for Twitter). This might be due to a perception of the envi-
ronment as “safe and closed worlds” (Light and McGrath 2010: 302). However,
it might also be related to the fact that users need to acquire the necessary tech-
nical expertise themselves (Yao 2011: 115), which is further impeded by the fre-
quent changes in interface design, making it increasingly more difficult to even

Authenticated
locate the options for privacy settings.12 Peer pressure is another influencing factor
(Gross and Acquisti 2005: 2).
The situation is rendered even more confusing for users by the inconsistent
privacy policies and practices on the different SNSs as well as modifications
over time, which have an impact on their perceived publicness (Bateman, Pike
and Butler 2010: 89) and stir insecurities and sometimes also anger among their
users. This was the case, for example, when it became evident in August 2016 that
the messaging service WhatsApp was supposed to pass on personal information,
including users’ names and phone numbers, to its parent company Facebook as
part of plans to allow businesses to send messages to users (e.g. reported in Tynan
2016). Such data harvesting for economic purposes constitutes another component
of unwanted outside access and intrusion (cf. Joinson et al. 2011: 35; Thelwall
2011: 256) – one not generating social, but economic capital (Andrejevic 2010:
84). Yet, there is also another side of the coin, as many private users have started
earning money by click-baiting via (self-)disclosure in the social media.
Section 2 has shown that, drawing on neighbouring disciplines, CS research
mainly focuses on informational, social, psychological and physical components
of users’ social media interaction, providing valuable insights into the gains and
risks of (self-)disclosure, the kinds of information revealed by users as well as the
interplay of their on- and offline aspects of identity. However, actual linguistic
practices are largely neglected in CS. Section 3 now turns to the pragmatic per-
spective on privateness and publicness.
3. Publicness and privateness through the lens of pragmatics
While both CS and pragmatics stress the scalar and often fuzzy nature of public-
ness and privateness, their respective treatments of the concepts tend to highlight
different aspects. Whereas CS research mainly foregrounds the levels of access
and content management (e.g. in the sense of privacy management), pragmatic
research focuses on the linguistic characteristics of digital discourse (DD)13 and
their potential for constructing private and public aspects of communicated con-
tent within the broader framework of pragmatic meaning, including users’ vir-
tual identities. Still, some CS approaches already hint at pragmatically relevant
12
See Light and McGrath’s documentation of changes in the Facebook interface (2010:
302–303).
13
Thurlow and Mroczek’s (2011) term ‘digital discourse’ is used in this contribution to
broadly cover the diverse channels and modes (i.e. computer-mediated communication
systems and their social and cultural practices, cf. Herring 2007: 3) of electronically
mediated communication. (But see Herring 2013: 6 for a discussion of varying termi-
nologies and their limits.)

Authenticated
phenomena. For example, Papacharissi and Gibson acknowledge the performative

character of privacy management and evoke a link to the pragmatic notion of face
wants, defining privacy as “the right to be left alone”, distinct from “the desire to
be left alone” (2011: 78). The authors also link up digital literacy and privacy, sug-
gesting “that an advanced form of digital literacy can enable individuals to redact
performances of the self online, so as to navigate public and private boundaries
fluently” (2011: 76, their emphasis). The idea that identities are created on the
keyboard (cf. boyd and Ellison 2007: 211) along with notions of conceptual orality
and literacy and thus authenticity is also taken up by Mallan and Giardina (2009)
and Trepte and Reinecke (2011b).
Pragmatic social media research approaches the public/private domain mainly
from four different angles, rather than providing a comprehensive and coherent
model:
(1) the framing conditions of online interaction, especially with reference to accessibility
by the public as well as mass audiences and patterns of participation such as lurking
(e.g. Crystal 2006: 31 for DD in general, and more specifically, e.g. for forum discus-
sions, Gruber 2013; Bös and Kleinke 2015);
(2) the oral–written/immediacy–distance paradigm relating to private and non-private uses
in early linguistic internet research (e.g. Crystal 2001, 2006; Danet 2001) and their
extension into enriched models of immediacy and distance which differentiate the fac-
ets of access and content (e.g. Dürscheid 2007; Landert and Jucker 2011);
(3) multimodal aspects of publicness and privateness, investigating multimodality at the
interface of the oral/written and public/private axes (e.g. Yus 2011; Sindoni 2013);
(4) complex models of computer-mediated discourse analysis (CMDA) and facetted classi-
fication schemes of CMD/DD, which incorporate the public/private dimension in vari-
ous facets and at various levels of analysis (e.g. Herring 2004, 2007 and 2013).
Empirical pragmatic research on social media discourse relevant from the public-
ness/privateness perspective has mainly focussed on the three broad perspectives
already outlined above: immediacy and distance, multimodality and the framing
conditions of specific applications. Some examples of research are provided in
Table 2, which also indicates the types of application studied.
Furthermore, recent studies include issues of visibility and access (cf. our
Table 1) into the analysis of collaborative text-production, see e.g. Herring (2013:
15) on Wikipedia as “massively multi-authored by internet users who usually do
not know one another”, which relates privacy concerns to issues of authentic-
ity and reliability.14 Last but not least, empirical studies on specific (ranges of)
applications scrutinize privateness issues within complex patchwork approaches,
which increasingly also use corpus-linguistic means of content analyses in order
to reveal language practices related to self-disclosure (e.g. Sindoni 2013; Pusch
14
See also Johansson (in press) on participation and access in online newspapers.

Authenticated
mann 2013; Herring et al. 2006). To cover the relevant perspectives, this chapter,
unlike Herring et al. (2013), draws on a “wide understanding of pragmatics as the
scientific study of all aspects of linguistic behaviour” (Bublitz, Jucker and Schnei
der, Preface to the handbook series, this volume: v), which includes structural
and socio-pragmatic as well as increasingly socio-semiotic (Page et al. 2014: 18)
patterns of language use.
Table 2. Examples and surveys of empirical pragmatic research perspectives
Feature Author and year, type of application

Application-related elements as framing conditions
Patterns of participation, Arendholz 2010, Kleinke 2010, Arendholz 2013,
accessibility Bös and Kleinke 2015, Kleinke and Bös 2015:
Forum discussions
Locher 2013: Advice fora
Kleinke 2015: Advice fora, medical self-support
groups, Twitter
Orality/Literacy – Immediacy/Distance
Genre Pérez-Sabater 2012: Facebook
Genre and structure Bieswanger 2013: CMD (various applications)
Gruber 2013: Mailing list communication
(Sub-)Genre and content/topic/ Baron 2013: Instant messaging
audience Sindoni 2013: Blogs, YouTube
Arendholz 2010: Forum discussions/message boards
Eisenlauer and Hoffmann 2010: Blogs
Hoffmann 2012: Blogs
Puschmann 2013: Blogs (personal and corporate)
Genre and narrativity Jucker 2010: Life-ticker
Arendholz 2010: Message boards
Eisenlauer and Hoffmann 2010: Blogs
Hoffmann 2012: Blogs
Authenticity and identity Leppänen et al. 2015: Facebook, YouTube, Twitter,
forum discussions, blogs
Multimodality
Authenticity/content/ Sindoni 2013: Videochat, Personal Blogs, YouTube
accessibility/virality and Vlogs
Lange 2008: YouTube
Complex semiotic practices
Accessibility, collaborative story- Eisenlauer and Hoffmann 2010: Blogs
telling, personal and public narra- Page 2012: Forum discussions, Twitter, blogs,
tive spaces SNSs, Wikis

Authenticated
Perspective (1) has often merely been mentioned in passing, explaining the
general context conditions of online interaction. Perspectives (2) − (4) indicate the
pragmatic complexity of the private/public dimension and its interfaces with CS
research and will therefore be in the focus of our discussion here.
3.1. Early accounts of publicness and privateness in the immediacy/distance

paradigm
Focusing on the textual make-up of various modes of online applications, early
linguistic research on CMD approached the concepts of publicness and private-
ness rather indirectly. Crystal (2001, 2006), in his discussion of Netspeak, uses the
terms public and private only with regard to the accessibility of early applications
of DD (e.g. e-mails and chats). Yet, he contributes to our understanding of more
comprehensive notions of publicness and privateness by positioning Netspeak on
a continuum of speech and writing15 (which, he acknowledges (2001: 25), is nec-
essarily an oversimplification) and by linking its diverse realizations to scales of
personal/non-personal subject matter and formal/informal style.
Applying prototypical spoken and written language criteria to the different
applications of Netspeak, Crystal (2001: 42–43)16 offers a rough categorization
based on bundles of characteristics such as time- and space-boundedness, sponta-
neity, information density, interactiveness and revisability. He furthermore links
micro-structural characteristics to their socio-pragmatic dimensions, which form
fluent continua on scales of (non)personal and (in)formal characteristics, indicat-
ing different potentials for immediacy and distance (cf. Table 3).
Crystal’s approach reflects the majority of the prototypical communicative
conditions and linguistic features already outlined in Koch and Oesterreicher’s
(1985) seminal work on conceptual orality and literacy. Koch and Oesterreicher
correlate the communicative conditions and linguistic elements with the aspect
of accessibility (“not public”/“public”), and associate them with the two poles
of communicative immediacy and communicative distance. Welcoming this dis-
tinction of “public”/“not public” (instead of “private”), Dürscheid (2007: 30−31)
15
The hybrid nature of DD and its perceived intermediate but independent position on a
scale of speech and writing is also manifested in other terms suggested, e.g. ‘interactive
written discourse’, ‘electronic language’, ‘computer-mediated communication’ or ‘key-
board-to screen communication’ (KSC), a term coined by Jucker and Dürscheid (2012)
focusing on users’ activities. See Danet (2010: 146) for an overview of earlier uses and
Crystal (2011: 2–3); Herring, Stein and Virtanen (2013: 5) and Page et al. (2014: 17) for
a critical discussion of these terms, which allude to different analytical levels.
16
In his second edition (2006: 45, 47), Crystal adds blogging and instant messaging, yet,
as it works mainly with ‘yes/no/variable’ distinctions, his characterization still remains
relatively general.

Authenticated
Table 3. Crystal’s (2001) characterization of Netspeak (our summary)
Immediacy Distance
(Non)personal characteristics
– Discussion of sensitive (personal, reli- – Discussion of safe topics
gious or political) topics
– Certain knowledge of each other’s vir- – Addressee(s) often unknown
tual (sometimes also real) identities
– Direct reference to other participants – Little direct reference to other partici-
possible pants
– Personal letter formats (terms of
address, salutation) – Official announcements
– Some phatic communication
(In)formal characteristics
– Informal lexis, nonsense words – Formal lexis
– Less complex grammatical structures – Complex grammatical structures
– Often highly liberal spelling conven- – Formal terms of address, formal saluta-
tions (sometimes in-group conventions) tions
– Use of emoticons to compensate for loss – Academic style-features (e.g. numerical
of visual contact structuring of texts, scientific terms)
points out that communicative contents do not necessarily coincide with social
closeness (as might be suggested by Koch and Oesterreicher’s idealized scheme).
She therefore differentiates “public”/“partially public”/“non-public” (relating to
accessibility) from “private”/“non-private” (relating to communicative content and
relationship of communicative partners) and establishes a category of “secondary
intimacy” (topics perceived as intimate in a culture, which are spread into the
public realm, e.g. on a private homepage or blog). This differentiation is already
suggested in earlier works (e.g. Schütte 2000: 149; Dürscheid 2003: 1517) and
taken up by Landert and Jucker (2011) in their development of a more complex
model (see below).
From a genre perspective, Crystal’s early, relatively rough categorization of
Netspeak has invited some justified criticism (cf. e.g. Dürscheid 2004). However,
his major observations have been verified by empirical studies on selected appli-
cations (e.g. Danet 2010; Jucker 2010; see Sindoni 2013: 39–40 for a detailed
discussion). Thus, Crystal’s parameters, next to general patternings of orality/lit-
eracy and immediacy/distance (e.g. Chafe 1982; Koch and Oesterreicher 1985)
17
However, in line with her differentiation of the private/non-private and public/non-pub-
lic dimensions, Dürscheid (2003) argues for a fine-grained classification of online
applications on a scale of orality/literacy rather than immediacy/distance.

Authenticated
still prove relevant and have found their way, in an adjusted form, into more recent
and complex approaches and models accounting for privateness and publicness
issues in DD, which also have to cope with the new phenomena raised by Web
2.0. Further input has been provided by developments in corpus-based research
on registers, e.g. the pioneering work by Biber (1988), originally focussed on off-
line registers, but increasingly adopted for the linguistic-pragmatic descriptions of
social media applications18 (cf. also Sindoni 2013 for a comprehensive overview
of research on English language data).
3.2. Enriched models

More recent linguistic approaches incorporate privateness and publicness explic-
itly in their models, in line with a more general trend of linguistic(-pragmatic)
research from text-oriented to activity-oriented approaches, as demanded, e.g.,
by Thurlow and Mroczek (2011: xx−xxvii), Page et al. (2014: 108) and reflected
in Herring, Stein and Virtanen (2013: 9), among others. The first comprehen-
sive, three-dimensional linguistic model of privateness and publicness is offered
by Landert and Jucker (2011). It is inspired by Koch and Oesterreicher’s (1985)
model, which contributes the axis of language with communicative immediacy
and communicative distance as its poles. Additionally, Landert and Jucker inte-
grate two further dimensions into their model by drawing on Dürscheid’s (2007)
differentiation of accessibility (public/non-public) and content (private/non-pri-
vate), which was also postulated in CS approaches (see Weintraub’s 1997 dimen-
sions of visibility and collectivity and Lange’s 2008 notions of “publicly private”
and “privately public” behaviour described in Section 2).
Landert and Jucker’s model still focuses primarily on the text itself (as the
product of the interaction to be analyzed) and views accessibility, immediacy and
content as characteristics of the text itself rather than of users’ activities in the pro-
duction of the text (cf. e.g. Jucker’s 2010 analysis of live-ticker commentaries as
real-time narratives). Despite this text-as-result orientation, their enriched model
is suited to coping with issues such as the much discussed topical phenomenon
described by Dürscheid (2007: 30−31) as “secondary intimacy”, i.e. topics usually
considered in a particular culture as private but leaving the realm of non-public
communication in DD and especially in social media.
In her study on personalization in the mass media, Landert (2014: 29) modifies
the model, replacing the dimension of accessibility (despite public access being a
defining criterion of mass media communication) with an interpersonal dimension
of involving vs. non-involving communicative settings (see also Landert, Ch. 2,
18
For attempts at using Biber’s model more systematically for the analysis of online
forum communication, see e.g. Collot and Belmore (1996).

Authenticated
this volume for a graphic representation of the model). Involving settings are those
“which invite their audience to interact with text producers and to contribute their
own content”, often in multimodal ways (Landert 2014: 29). The non-involving
pole of the scale denotes non-interactive settings. By changing Landert and Juck-
er’s (2011) primarily text-oriented model into one that can account for users’ par-
ticipation practices, Landert’s adjusted model allows for a more systematic inves-
tigation of linguistic output in relation to participation patterns. However, in order
to account for the public/private dimension of social media use beyond participa-
tion, the level of access is central and indispensable for the social and psycholog-
ical dimensions of privacy discussed in CS.
The incredibly complex dimension of content is still somewhat fuzzy in
Landert and Jucker’s model. Eisenlauer and Hoffmann show that even one of the
most traditional applications, blogs, have to be considered as “intermediary genres
in fusing both internal and external information, private and public topics” (2010:
82). For example, they observe that in blogs, stories from the news media are
occasionally retold as “news narratives”. This is personal in that it reveals some-
thing about the attitudes and the identity of the blogger (2010: 85–86), and yet,
the content of what is being retold can be considered as entirely non-private (in
Landert and Jucker’s sense of “(traditionally) … suitable for discussion in public
contexts”, 2011: 1427). The same holds true for other social media applications.
YouTube postings, to name yet another example, can combine the public con-
tent of an artist’s performance in a video sequence with the personal comments
launched by the original poster and other users. At a deeper (ontological) level,
public and private contents also tend to blur from the perspective of users’ digital,
often anonymous or invented identities. On the one hand, digital identities are con-
structed to enhance users’ visibility and their ability to act in public. On the other
hand, it is precisely their practiced anonymity that disguises users’ non-virtual
private identities and disembodies private content, which is thereby hidden from
the public eye and resists access.
Allowing for a pragmatic perspective on DD that goes beyond the roots of
research on conceptual spokenness/writtenness, Landert and Jucker’s (2011)
model can be placed in the context of two further developments central for prag-
matic DD-research on publicness/privateness: Firstly, the emergence of new mul-
timodal practices in social media to be accounted for, and secondly, the formation
of multi-facetted classification models for DD.
3.3. Publicness, privateness and multimodality

Technological developments and the advent of Web 2.0 have encouraged both
massive media convergence and multimodality in the past two decades. Whereas
people used to initiate social media contact deliberately, by logging into their user
accounts, now many users take it for granted that everyone is permanently logged

Authenticated
in. Multimodal practices are deeply entrenched in users’ social media behaviour;
they usually handle the complex semiotic practices of social media very easily, and
the choice of the medium “becomes communicative intent” (Madianou and Miller
2013:169; cf. also Yus 2011: 113).
It is thus important that constant user engagement and multimodality are
accounted for in new media studies (cf. Thurlow and Mroczek 2011: xxv−xxvi),
even if their complexity defies the creation of lean, streamlined and comparatively
stable models of description. Yet, while Crystal (2001: 202) already hinted at the
features of multimodality in early applications (e.g. the use of hypertext links to
enhance interactivity), this aspect was long neglected in pragmatic social media
research. Landert and Jucker’s (2011) model, for example, disregards the dimen-
sion of channel, simply pointing out that its focus is on language realized graph-
ically and not phonically (and thus implicitly supporting a dichotomous under-
standing of spoken vs. written, 2011: 1426) – clearly, this is a practical decision
(based on the nature of the data analyzed there).
Whereas researchers agree that multimodality in contemporary social media
exceeds the anchoring of spoken and written discourse by far, the relation between
different semiotic resources is less clear. In their KSC-model, Jucker and Dürscheid
(2012) postulate a quadro-modality consisting of two main modalities with two
realizations each (pictures, still/animated, and language, graphic/phonic), within
which, however, the “boundaries between these codes are clear-cut”.19 Sindoni
(2013: 2) includes a broader range of semiotic resources (“speech, writing, gesture,
movement, gaze […] in spontaneous web-based interactions”), but characterizes
them as integrated in “unprecedented ways enacting new interactional patterns and
new systems of interpretation among web users”, see also Stæhr (2015). These
modes pose new questions regarding the nature of their interaction and integration
and (degrees of) users’ responsiveness to verbal and non-verbal resources. From
an analytical perspective, the question “at what level […] integrative analyses of
the meanings and functions of the complex whole take place, and what theories
and models exist to guide it?” (Herring 2013: 19) requires clarification.
Multimodality is obviously not a characteristic of new media applications per
se, but Web 2.0 has tremendously facilitated the combination of different semiotic
resources, especially the use of pictures, thus making “complex semiotically hybrid
textures” possible (cf. Eisenlauer and Hoffmann 2010: 83 on vlogs, fotologs and
audioblogs). Thurlow and Mroczek attribute “the potential in new media for inven-
tion and creativity” to “[a]ll texts, all communicative events [which] are always
19
But see e.g. Dürscheid and Frick (2014: 173–174) on pictorial writing as a form of
iconographic communication in WhatsApp, for which the application provides the nec-
essary technical prerequisites, or Page et al. (2014: 16) on visual icons in Facebook
update templates.

Authenticated
achieved by means of multiple semiotic resources, even so-called text-based new

media like instant- and text-messaging” (2011: xxvi). Multimodality can thus be
considered an outstanding, but not isolated phenomenon of social media, which
has to be analyzed against the background of other new media principles such as
multi-linearity, fragmentation and interaction (see Eisenlauer and Hoffmann 2010:
93–100).
Which semiotic modes prevail is strongly influenced by the different social
media platforms (see Page et al. 2014: 16–17 for a detailed discussion). However,
this is not just a question of technical affordances; it is also a matter of “the ways
in which people take-up and create the possibilities which they perceive” (Page et
al. 2014: 18, drawing on Barton and Lee 2013, see also Lee 2011: 120−121). Thus,
users’ choices regarding the opportunities an application offers must be considered
in a dynamic, user-centred perspective on social media genres that goes beyond
text-based approaches.
The multimodal and linguistic choices users make are meaningful in the pro-
cess of balancing privacy and sociality in social media environments. For example,
research by Sindoni (2013) on videochats shows that the written channel is often
preferred in cases of secrecy (e.g., when potential overhearers are in the vicinity),
and when the perceived intimacy of the contents is high (informants often reported
that “written chat proved a ‘safer’ way to express confidential or private feelings”,
2013: 66–67). Systematic considerations regarding users’ choices thus have to be
embedded in a multi-layered and fluid notion of context framing users’ handling
of multimodal affordances. Page et al. (2014: 33) subsume six levels of context,
which all may have an impact on users’ pragmatic privacy choices: participants
(people involved in the interaction “and their relationship to others in the group”),
imagined context (“projected context created cognitively by participants on the
basis of their world knowledge and the cues provided in CMC” – possibly includ-
ing projected audiences and assumed communities and community membership),
extra-situational context (“offline social practices”, including macro- and meso-
level cultural values), behavioural context (i.e. the physical situation in which the
participants interact), textual context (co-text) and generic context (i.e. the social
media site in which the interaction takes place, including its “stated purpose, rules
and norms for conduct”).
Current research discusses the impact of multimodal interaction on privateness
and publicness primarily as a challenge to the traditional oral/written and imme-
diacy/distance paradigm (e.g. Sindoni 2013). Additionally, linguistic research –
just like CS research – has acknowledged another aspect of multimodal devices
as significant with regard to the privateness or publicness of an interaction: the
capacity of multiple semiotic resources to enhance the construction of new types
of contents, such as status updates and profile information (Herring 2013: 5).
From a relevance-theoretical perspective, the multiple semiotic resources typi-
cal of social media applications provide the context for the construction and inter-

Authenticated
pretation of pragmatic meaning. This includes, for example, visually transmitted

private information at the content level. As Yus observes, pictures, videos and
links to further web-applications provide “more and more options for contextual-
ization” (2011: 31) and thus create more assumptions that are mutually manifest
for the participants in an interaction. Contextualizing items can be placed in differ-
ent slots in SNS profiles such as the profile picture, personal information, friends,
name and status or the contents of individual comments (Yus 2011: 113).
Taking a Gricean perspective in her analysis of YouTube comments, Sindoni
postulates a “multimodal relevance maxim” which “assumes that relevance needs
to encompass all semiotic resources that are deployed in web-based interactions”
(2013: 205, 211). Thus, successful privacy management has at least a twofold
impact on pragmatic meaning construction: On the one hand, private contents
serve as contextualization cues enhancing the socio-pragmatic construction of
meaning. On the other hand, the way a user handles their private information (e.g.
the impact of their photos and texts on other users on a SNS) helps to construct and
index their identity and status (Yus 2011: 36).
At a third level, again quite in line with CS findings, the multimodal resources
users exploit in their privacy management of SNSs also have an impact on authen-
ticity and reliability of social media interaction. Pictures, videos or links to further
web-applications play a part in contextualizing users within their “hybrid (physi-
cal-virtual) personal networks of interaction” (Yus 2011: 26) with “offline identi-
ties very much carry[ing] over to online behaviour” (Hargittai 2008: 277).
3.4. Publicness and privateness in multifeature classification schemes for DD

Clearly, multi-facetted, fine-grained classifications are necessary to account for
the complexities of social media, including multimodal practices, new criteria for
media textuality, such as hypertextuality, and complex patterns of narrativity as,
for example, outlined in Bublitz (2008), Eisenlauer and Hoffmann (2010) and Page
(2012). The gradually growing development of research paradigms in this vein
was pioneered by Herring’s multi-facetted classification of CMD (2007), an early
attempt to do justice to the complexity of social media. It exemplifies the develop-
ment from primarily text-centered to primarily user-centered approaches by incor-
porating technological (medium-) as well as social context-factors as relevant facets
of CMD-genres20 systematically (for a detailed review see Page et al. 2014: 10–11).
20
See Herring (2001: 612) on her preference of ‘computer-mediated discourse’ (CMD) as
the term which focuses on language and language use in computer-networked environ-
ments, analyzed by methods of Discourse Analysis to address that focus, and Herring
(2013: 6) on why she continues to use the term despite other more recent notations such
as DD used in our paper.

Authenticated
Herring’s scheme offers a suitable starting point of analysis for current social
media privacy practices. Already the self-contained facet privacy settings in her
system links up with the dimension of access (“public”, “semi-public”, “semi-pri-
vate”, “private” contexts). Further aspects of privacy discussed in CS are anchored
in several other medium-factors suggested in Herring’s (2007) model: message
transmission (“one-to-one”, “one-to-many”); persistence of transcript (“ephem-
eral” vs. “archived”); channels of communication (words, image, sound, video);
and anonymity (extent to which the participants’ identities are represented within
a site). In addition, privacy issues link up to all situation factors (though to var-
ying degrees): participation structure (number of participants involved and their
patterns of active/passive and collaborative participation); participant character-
istics (stated and assumed demographic and ideological characteristics, alluding
to the problem of whom to tell what); purpose (goals of interaction as a volitional
attitude concerning a (group of) user(s)); topic (adhering to the contents of what
is posted); tone (“formal” or “informal”, illuminating whether the user perceives
the interaction as private or non-private); norms (e.g. accepted (privacy) practices
established by the group); and code (the language variety a user chooses, which
may be interpreted as an indirect and possibly unintended form of self-disclosure)
(Herring 2013: 621; Arendholz 2013: 111).
Herring’s (2007) classification scheme still lacks the explicit linguistic dimen-
sion outlined in Landert and Jucker (2011); however, it has to be considered against
the background of Herring’s (2004) model of CMD analysis (CMDA), a “set of
methods (a toolkit) grounded in linguistic discourse analysis […] organized around
four levels (structure, meaning, interaction management and social phenomena)”21
(Herring 2013: 4). It includes orality, formality, complexity, efficiency, expressiv-
ity as well as genre characteristics to be dealt with at the level of structure, but
without relating these factors explicitly to issues of privateness and publicness.
Consequently, for Web 2.0 applications, Herring suggests two extensions of the
2007 model and the CMDA toolkit: language structure as a third group of fac-
etted dimensions (though like Landert and Jucker 2011 without going into detail
about any structures related to privateness), and the inclusion of multimodality as
one more level of systematic analysis in CMDA (2013: 20). However, adopting
a privateness/perspective, it is useful to adjust the model slightly by carving out
facets related to the public/private dimension, linking these up more explicitly to
CS-findings and enriching the model by elements drawn from empirical research
as outlined in section 4.
21
To this set, Herring adds a further level of participation patterns (2004: 3, 2013: 4).

Authenticated
4. Pragmatics and Communication Science as input spaces for models

of publicness and privateness
By now, the “inadequacy of any single or dichotomous model of the public/private

distinction to capture the institutional and cultural complexity of modern socie-
ties” (Weintraub and Kumar 1997: xiii) has become abundantly clear. In this vein,
authors such as Herring (2013: 5, 20), Page et al. (2014: 14, 48–50), Puschmann
(2013: 85–86) and Androutsoupolous (2013: 47) argue strongly for inter-, intra-
and partly also transdisciplinary approaches and for “incorporating findings from
neighbouring disciplines such as communication, ethnography, sociology, and
social psychology, […] to contextualize language-based approaches” (Puschmann
2013: 102). Some of these contextualizations can be associated with the differ-
ent disciplines relatively straightforwardly (see Table 4). What all these different
disciplines (apart from computer science) have in common is their focus on the
user rather than the digital text as the output of users’ online practices (even in the
broadest sense of electronic textuality and multimodality discussed in Section 3).
Linguistic pragmatics and CS both have a long tradition of including the find-
ings of these neighbouring disciplines into their research paradigms. However,
incorporating their “potential of researching how language is used in social media
contexts is not always easy to handle” (Page et al. 2014: 48). As Herring (2013: 4)
already observes for her model of CMDA, adequate models usually require input
from a wide range of studies revealing potentially relevant dimensions of descrip-
tion “in principle from any relevant discipline that analyses discourse”. However,
even with a much narrower focus on just one facet of the linguistic description of
privateness/publicness (spoken and written discourse and multimodality in online
interaction), Sindoni feels that the “textual web-galaxy” with its ever-changing
strategies of personal expression and performance does “not allow for stable theo-
ries or for the elaboration of models of analysis belonging to one specific field of
studies” (2013: 171).
Thus, what currently looks like patchwork approaches to privateness and pub-
licness, determined by shifting descriptive interests, requires re-evaluation from a
linguistic perspective, (re-)adjusting the linguistic toolkit to a fluid, highly com-
plex online environment shaped by flexible user practices. Herring (2013: 21),
for example, exemplifies different foci in analyses of more traditional Web 2.0
applications vs. discourse-focused approaches to contemporary convergent media
web phenomena, pointing out that “each is a lens with a central focus and a periph-
ery, and certain phenomena are outside the scope of each”. Whereas the former
tend to concentrate on user-created content, user control of content, collaborative
processes and folksonomies to distribute public contents, the core areas of conver-
gent media analysis comprise descriptions of media coactivity anywhere online
(including mobile phones), communication, social interaction and the language
used therein.

Authenticated
Table 4. Contributions of various disciplines to public/private conceptualizations
Disciplines Accounting for

Computer science medium factors potential of constantly emerging new tech-
nical affordances
Socio-semiotics multimodal privacy practices
Management and business economic and social functions of privacy management
studies (users as customers exploited by or benefitting from mas-
sive data mining practices)
Information studies, users’ gaining access to public and private information as
media studies, journalism, well as visibility through participation in public discourse
political sciences and the distribution of private information therein
Sociology and (social) (group-related) aspects of privacy management in virtual
psychology communities
Psychology psychological mechanisms motivating users’ balancing of
risks and gains of online disclosure
Anthropology ethnographically motivated privacy practices (e.g. gender-
or age-related preferences)
Education and pedagogy privacy practices in educational contexts; acquisition of
culturally determined privacy norms; degree of computer
literacy needed for successful privacy management
4.1. Converging the input spaces
The complexity of privateness and publicness in fast-changing social media applica-

tions and the fluid, overlapping and rapidly changing research interests thus require
models that provide a framework that possesses exactly this flexibility. We there-
fore suggest that categories for the analysis of social media in CS and pragmatics,
whether already established or newly emerging, should be conceived of as flexible,
dynamic input spaces, which researchers can resort to as well as add to, depending
on their particular research questions. The notion of input spaces is inspired by the
flexibility of cognitive linguistic models as outlined in conceptual-blending theory,
which brings together elements from two mental spaces by various cross-space
mappings, linking incoming contextual information and more permanent cognitive
models “as online conceptual representations” (cf. Ungerer and Schmid 2006: 259).
Figure 1 illustrates the categories and facets of description for the input spaces
deduced from models from both strands (CS and pragmatics), enriched by catego-
ries employed in empirical work on privacy/privateness and publicness, which may
serve as a starting point for the further elaboration of both input spaces. In addition
it depicts selected cross-space mappings already revealed by empirical research.

Authenticated
Figure 1. Pragmatic and CS input spaces for the description of publicness and
privateness/privacy

Authenticated
Input space 1 is informed by a wide notion of pragmatics and takes into account
the relation between the two main dimensions of language and context of interac-
tion, including a socio-semiotic dimension with medium- and situation-factors as
meaning-constructing entities. It builds on Landert and Jucker’s (2011) and Her-
ring’s models (2004, 2007, 2013), adjusting the latter to the private/public dimen-
sion and adding elements from empirical research.
The dimension of language is based on two of the elements outlined in Landert
and Jucker’s (2011) model: content/topic (related to Herring’s 2007 facet of ‘sub-
ject matter’) and language of immediacy/distance (including e.g. Herring’s fac-
ets of ‘tone’ and ‘code’, structural elements of CMDA such as orality, formality
and complexity, as well as the categories of deixis and addressivity (Puschmann
2013)).
Context of interaction, the second major element of the pragmatic input space,
includes user-related and application-related elements. The user-related elements
comprise Herring’s participant/participation-related facets (e.g. participant struc-
ture and characteristics) and “demographic and ideological factors” (Page et al.
2014: 10), which may be deduced or merely inferred from users’ on- and offline
activities. It furthermore comprises activity type, including Herring’s ‘purpose/
goal of interaction’, but slightly broader in the sense of dynamic interactional
genre characteristics and hence also relating to ‘norms’, i.e. “conventional prac-
tices within the computer-mediated environment” (Herring 2007: 21).
Application-related elements include Herring’s ‘message format’, i.e. how
contents and user activity are organized on the screen, the display and accessibil-
ity of affordances, templates and archiving principles, cf. also Page et al. (2014:
17−18). Furthermore, Landert and Jucker’s (2011) third dimension of publicness/
privateness, ‘access’, is related to Herring’s ‘privacy settings’, her facet ‘per-
sistence of transcript’ (2007), extended by social media characteristics such as
persistence, replicability, scalability, searchability,22 and her facet ‘channels of
interaction/multimodality’ (2007, 2013). This last facet involves various semiotic
resources and “media richness” (i.e. the combination of resources such as text,
image, video), which opens up different choices in terms of amount and depth of
personal disclosure invited by different types of social media (Kaplan and Haen-
lein 2009: 62; see also Page et al. 2014: 15–16).
Of course, this list is by no means comprehensive and even though the elements
are presented here as separate independent entities, they are closely intertwined.
Categories may merge into each other, also depending on the research question,
as is the case with the elements content/topic and language (of immediacy and
22
Persistence, replicability, scalability and searchability have also been discussed from
a CS-perspective, but as they are inherently application-related we prefer to deal with
them in the pragmatics input space.

Authenticated
distance), with overlapping semantic aspects of meaning revealed in lexical con-

tent analyses (cf. e.g. Sindoni 2013 for blogs and YouTube comments). Also, the
user- and application-related elements cannot be seen as completely clear-cut
dimensions. In order to become pragmatically relevant for the construction of pri-
vateness and publicness, application-related elements such as ‘privacy settings’
depend on users’ ability and willingness to handle them successfully.
Input space 2 is inspired by the interdisciplinary approaches in CS and includes
aspects of privacy/privateness and publicness rooted in psychological, sociologi-
cal, media and communication categories. Drawing on the vast CS research out-
put, including empirical studies, we have identified seven elements that feed into a
pragmatic interpretation of social media content from a privacy angle (the list is, of
course, open). Unlike those common in linguistic-pragmatic approaches, all these
levels are primarily user-oriented, as they either focus on users’ social and psy-
chological conditions or foreground users’ handling of the different affordances.
As outlined in Section 2, dimensions of privacy and privacy concerns include
informational privacy and social, psychological and physical privacy. They link up
to the pragmatic input elements ‘access’ and ‘content/topic’, but go beyond these
in specifying individual elaboration sites of the abstract concept of privacy and
in offering a quantitative level, ‘amount’, which relates to Kaplan and Haenlein’s
(2009) notion of “media richness”.
Layers of disclosure ranks contents and topics of self-disclosure from a psy-
chological, user-oriented (rather than data-protection) perspective as ‘peripheral’
(name, address, friends, etc.), ‘intermediate’ (political views, past-time prefer-
ences, etc.) and ‘core disclosure’ (emotions and attitudes).
Connectedness may be instantiated in social media applications in many ways,
including ‘lists of friends and followers’, which, according to Jucker and Dürscheid
(2012), may also lead to asymmetrical (privateness) constellations. Connectedness
also becomes manifest as ‘virality’, i.e. the uncontrolled, perhaps also unintended
popularity of a particular social media contribution (Sindoni 2013: 181).
Privacy management includes ethical self-regulation/self-censorship, i.e. pri-
vacy protection strategies (often dependent on internet literacy) such as the use of
initials, nicknames and social roles to protect others (Schmidt 2011b: 167–168).
Further strategies include providing fake or inaccurate information, excluding per-
sonal information, sending private messages within an application instead of post-
ing messages to a friend’s wall, blocking former contacts, providing contacts with
access to limited profiles, changing default privacy settings activated by an appli-
cation, untagging from images and/or videos posted by contacts for self-protection
(Young and Quan-Haase 2009: 271) and polysemic performances (constructing
different virtual identities for different types of audience on different – or maybe
even within the same – social media application).
Another element of the CS input space is audience design, which relates to
Bell’s (1984) model. It illuminates how speakers’ expectations concerning their

Authenticated
audience shape their language style (cf. also Puschmann 2013: 93–96 and Page et
al. 2014: 23–24) and links up to the idea of users constructing privileged publics.
In contrast to ‘purpose of interaction’ as one aspect of the activity type/genre in
the pragmatic input space, (large-scale) interactive goals comprise users’ com-
municative intentions related to ‘self-expression/gaining visibility’ and the social
capital they may gain thereby (cf. e.g. Puschmann 2013: 102 for blogs). Using
multimodal resources, both goals are easily pursued across different affordances in
one and the same application or even across different social media (“cross-media
practice”, cf. Lee 2011: 120–121 for Facebook, blogs and MSN). Such cross-over
phenomena link up to the element context collapse, i.e. uniting large numbers of
users in the context of one social media application, who in offline, face-to-face
environments would never get in touch with each other. This is, for example, the
case in contexts such as forum discussions and comment sections on online news-
papers, blogs or YouTube, where users often interact anonymously and reveal little
personal information.
The input space model (similar to models used in cognitive semantics) offers
not just flexibility, but also the bonus of providing an additional explicit slot, cross-
space mappings, which can account for both links between elements of spaces on
a single level, such as those connecting pragmatic and CS input spaces, as well as
vertical links connecting independent sub-spaces on different levels. After all, as
Yus (2011: 288) argues from a relevance theoretical perspective, humans have a
“biologically rooted […] cognitive principle of relevance”. Thus, just as in offline
interaction, social media users’ practices and choices of interpretation draw on
the entire social-cultural context available, including elements from the CS-input
space, which as part of “the context and the qualities of the medium […] can influ-
ence the estimation of relevance and the very choice of an interpretation” (Yus
2011: 288) and thus the mutual construction of meaning. Therefore, also extra-sit-
uational and general context-factors, discussed by Page et al. (2014) and spelled
out by Puschmann for blogs, such as the “motivation to blog, the role blogging
plays for a community, and the reflection of practitioners on their practice” (2013:
87), are part of the context of the medium also in a relevance-theoretical sense and
“call for the addition of ethnographic, sociological, and psychological research
approaches” (2013: 87) to exclusively linguistic approaches.
Privacy has been an independent research issue in CS for quite some time, with
empirical studies covering literally all elements and dimensions of input space 2,
sometimes also including user- and application-related aspects of the context of
interaction from the pragmatic research paradigm (see e.g. Trepte and Reinecke
2011 and Papacharissi 2011 for an outline). In empirical linguistic-pragmatic
research, however, the privateness/publicness dimension has mainly been treated
rather indirectly and as secondary to the construction of interpersonal meaning
(see e.g. Gruber 2013: 60 with reference to users’ awareness of the public nature
of the ‘communicative situation’ in mailing lists and the survey in Table 2 above).

Authenticated
Figure 1 exemplifies cross-space mappings between the pragmatic and the CS

input spaces from empirical research by Puschmann (2013) and Sindoni (2013).
Puschmann (2013) provides examples of how elements from both the pragmatic
and the CS input spaces can be linked in the analysis of privacy-practices. For
example, profile information (such as title, text, author, time of publication and
previous postings) provides deictic centres which enhance deictic language (e.g.
first-person pronouns, temporal and spatial adverbs, etc.) as one linguistic char-
acteristic of blogs (2013: 91−92). Audience design is shown to correlate with lan-
guage choice in terms of addressivity/orality in an example of a goal-oriented blog
in corporate communication (2013: 93−96, see also Puschmann 2010). (Large
scale) communicative goals such as self-expression and gaining visibility typi-
cal of the “controlled discourse environment” of blogs (2013: 98) are interlinked
with the context-related conditions of access-control in a “topic-centric style”
or “author-centric style” with their respective linguistic manifestations (2013:
98−102).
Sindoni (2013) correlates elements from both input spaces, e.g. by analysing
the impact of privacy concerns (and other factors such as internet literacy, techni-
cal problems) on users’ swapping between the written and oral mode in spontane-
ous web-based video interactions and its effects on face-protection, an important
aspect of the psychological dimension of privacy (2013: 64−68). Furthermore, she
scrutinizes the virality of a video on YouTube against the background of manag-
ing private content from a multimodal perspective including elements from input
space 1, i.e. text, gaze, and visual and audio elements in remakes (2013: 181−184,
see also Lee 2011 for managing private contents on Facebook, in blog entries and
in MSN).
Puschmann’s and Sindoni’s work already attest to the utility of an input-space
model correlating elements. Relations between other elements from both input-
spaces such as ‘content/depth of disclosure’, ‘participation patterns’ and ‘(expected)
audiences’; ‘anonymity’, ‘persistence, replicability, scalability, searchability’, and
‘multimodality’ have not been studied systematically and are very likely to yield
interesting results.
4.2. Authenticity, reliability and research ethics

Our discussion has repeatedly touched on authenticity and reliability as aspects of
privateness and publicness in the social media, which has turned into “a core value
in the production of discourse for a personal public” (Androutsopoulos 2015: 74).
Social media authenticity is created in multiple ways, and as users have devel-
oped a high sense of normativity and reflexivity, it is also subject to metaprag-
matic negotiation, as Kytölä and Westinen (2015) observe in Finnish Twitter data.
Pragmatic research has just started to understand the complexities of authenticity
and identity construction, i.e. “in what particular ways and under what conditions

Authenticated
authenticity is made locally meaningful, and how it is oriented to, indexed and
communicated to in linguistic and semiotic action and interaction” (Leppänen et
al. 2015: 1).
From a pragmatic perspective on privateness, all the major elements outlined
in input space 1 prove relevant. Relating to the dimension “language”, pragmatic
studies (not just) of social media have shown, for example, that linguistic fea-
tures of immediacy enhance authenticity, as users appear to present themselves in
the unfiltered, instantaneous ways typically associated with personal face-to-face
communication. For example, authenticity is created by cross-over phenomena,
including e.g. the use of vernacular language, and other phenomena of linking up
virtual to offline identities (e.g. Arendholz 2010: 118; Eisenlauer and Hoffmann
2010: 84). At the content level, the discussion of personal topics with substan-
tial personal detail also enhances authenticity and obviously reduces privateness.
Application-related features like multimodality can further boost authenticity, yet,
at the same risk. Also visual elements such as photos and videos provide proof and
enrich contextualization (cf. Yus 2011: 31).
This is also acknowledged in CS studies (e.g. Mendelson and Papacha-
rissi 2010: 267) and, hence, “authenticity” provides a suitable candidate for the
cross-mapping space of our model: From a CS perspective, the elements outlined
in input space 2 are important as well. CS research has shown that authenticity,
e.g. via core disclosure, raises social and economic capital, but obviously impedes
privacy. Just like in pragmatic research, topic-dependence has been emphasized in
CS, with authentic information “more likely to occur in user-centric than in pri-
marily interest-centric SNSs” (Ziegele and Quiring 2011: 181). For example, “the
idea of personal authenticity and subjectivity” belongs to the central norms and
expectations in blogs (Schmidt 2011b: 167). As far as connectedness and large-
scale communicative goals are concerned, offline identities have been found to
transmit to the online world much more than originally assumed (Hargittai 2008:
277), and SNSs are often preferred as means of maintaining offline networks
instead of making new, online acquaintances (2008: 290).
Still, there is, of course, also “impression management” going on (Walther
2011: 7). The information provided via a range of semiotic resources is usually
carefully filtered and thus highly subjective (cf. Mendelson and Papacharissi
2010: 252). This might cause (unintended) effects of perceived unauthenticity,
which – as Walther (2011: 7) points out – are more easily forgiven among close
friends than with less well-known acquaintances, who might be suspected of
hypocrisy. Yet, intentional deceitfulness is quite rare in SNS profiles, which tend
to be if not honest, then playful or ironic (Young and Quan-Haase 2009: 271).
Both pragmatics and CS have been more critical of the reliability of data users
distribute when participating in public discourse (e.g. Papacharissi 2010; Herring
2013: 19 or Arendholz 2013: 112 on degrees of reliability in users’ self-portrayal;
and Johansson in press).

Authenticated
The easy accessibility of social media data, their persistence and replicability
might appear appealing to researchers, yet, as Eysenbach and Till already pointed
out in 2001, “[i]nternet communities’ members do not expect to be research sub-
jects” (2001: 1103), at least not that of research other than marketing studies.
Social media research thus requires careful ethical consideration, even more so
nowadays when people are constantly logged in and search engines have become
much more powerful. As summarized by Rosenberg, social media researchers tend
to negotiate the nature of their data between two major positions: “online phenom-
ena can be considered public either (1) if publicly accessible or (2) if perceived as
public by participants” (2010: 24). That the de- and recontextualization of social
media can be perceived as a violation of privacy has not only been documented in
CS studies (cf. e.g. boyd and Marwick 2011: 6, reporting a case where teachers, in
order to discuss privacy issues, had shown extracts from the Facebook profiles of
US teenagers and thus evoked their anger).
It is certainly not easy to find the balance between the two positions outlined
above. Eysenbach and Till (2001: 1105) provide a summary of issues to be con-
sidered by researchers: intrusiveness, perceived privacy, vulnerability, potential
harm, informed consent, confidentiality, intellectual property rights. That the con-
sideration of user perspectives is a necessary step “into the right direction” is also
emphasized by Rosenberg (2010: 24). In fact, this step certainly also proves bene-
ficial for researchers’ insights into the fluid notions of privateness and publicness –
just like the first-order perspectives recently suggested in other fields of pragmatic
research (e.g. on (im-)politeness phenomena).
5. Conclusion
Privateness and publicness are incredibly complex notions affected by cultural,

socio-historical and linguistic contexts and highly user-dependent. Their conceptu-
alization thus needs to be flexible and multi-layered. As our discussion has shown,
it is essential to consider different disciplinary perspectives. Drawing particularly
on insights from CS and pragmatics, the input-space model brings together user-
and text-focused approaches and allows for a more comprehensive understanding
of publicness and privateness in the social media.
CS, with its predominant user-orientation, helps us to capture the multi-di-
mensional nature of privacy, including informational, social, psychological and
physical components. It offers insights into the gains and risks of (self-)disclosure,
the kinds of information revealed by users, the way online and offline aspects of
users’ identities are mingled, and the authenticity and reliability of their (self-)pres-
entation. It also provides comprehensive accounts of strategies of privacy manage-
ment in social media. Yet, actual linguistic realizations are largely neglected in this
research strand. In this regard, our understanding of publicness and privateness

Authenticated
benefits from linguistic (micro-)pragmatic analyses, which are largely – though

not exclusively – text-oriented. Pragmatic approaches have considered the impact
of the framing conditions of social media discourse. They outline access, content
and language dimensions, focusing particularly on linguistic patterns of orality/
writtenness and immediacy/distance. They also consider multimodal aspects of
publicness and privateness. Their description is enriched by multifeature classi-
fication schemes, incorporating socio-technical facets and thus accounting for the
complexity of DD. Overlaps in the conceptualizations of both approaches are made
visible in the cross-mapping space of our model, which indicates the proximity of
the two research strands under scrutiny and their beneficial correlation. Insights
from neighbouring disciplines can further advance the conceptualization of public-
ness and privateness and thus might demand still more attention in future research.
The same is true for the notion of publicness, which so far has often been conceptu-
alized in a rather parasitic way, as the opposite end of the privateness/privacy scale,
and certainly deserves more systematic treatment in linguistic pragmatic studies.
References
Andrejevic, Mark
2010 Social network exploitation. In: Zizi Papacharissi (ed.), A Networked Self:
Identity, Community and Culture on Social Network Sites, 82–101. New York:
Routledge.
Androutsopoulos, Jannis K.
2013 Participatory culture and metalinguistic discourse: Performing and negotiating
German dialects on YouTube. In: Deborah Tannen and Anna Marie Trester
(eds.), Discourse 2.0. Language and New Media, 47–71. Washington D.C.:
Georgetown University Press.
2015 Negotiating authenticities in mediatized times. Discourse, Context and Media
8: 74–77.
Arendholz, Jenny
2010 Need to put this out there (my story). Narratives in message boards. In: Chris-
tian R. Hoffmann (ed.), Narrative Revisited. Telling a Story in the Age of New
Media, 109–142. Amsterdam/Philadelphia: Benjamins.
Arendholz, Jenny
2013 (In)appropriate Online Behavior: A Pragmatic Analysis of Message Board
Relations. Amsterdam/Philadelphia: Benjamins.
Barnes, Susan B.
2006 A privacy paradox: Social networking in the Unites States. First Monday 11
(9). http://firstmonday.org/article/view/1394/1312 (last accessed 15 Septem-
ber 2016).
Baron, Naomi S.
2013 Instant messaging. In: Susan C. Herring, Dieter Stein and Tuija Virtanen (eds.),
Pragmatics of Computer-Mediated Communication, 135−161. Berlin/Boston:
de Gruyter.

Authenticated
Barton, David and Carmen Lee

2013 Language Online. Investigating Digital Texts and Practices. London/New
York: Routledge.
Bateman, Patrick, Jacqueline C. Pike and Brian S. Butler
2010 To disclose or not: Publicness in social networking sites. Information Technol-
ogy and People 24(1): 78–100.
Baym, Nancy K. and danah boyd
2012 Socially mediated publicness: An introduction. Journal of Broadcasting and
Bell, Allan
Berger, Charles R., Michael E. Roloff and David R. Roskos-Ewoldsen (eds.)
2010 The Handbook of Communication Science, 2nd ed., Los Angeles: Sage.
Biber, Douglas
1988 Variation across Speech and Writing. Cambridge: Cambridge University Press.
Bieswanger, Markus
2013 Micro-linguistic structural features of computer-mediated communication. In:
Susan C. Herring, Dieter Stein and Tuija Virtanen (eds.), Pragmatics of Com-
puter-Mediated Communication, 463−485. Berlin/Boston: de Gruyter.
Bös, Birte and Sonja Kleinke
2015 The complexities of thread-internal quoting in English and German online dis-
cussion fora. In: Jenny Arendholz, Wolfram Bublitz and Monika Kirner-Lud-
wig (eds.), The Pragmatics of Quoting Now and Then, 71–96. Berlin/Boston:
de Gruyter.
boyd, danah
2007 Why youth (heart) social network sites: The role of networked publics. In:
David Buckingham (ed.), Youth, Identity and Digital Media, 119–142. Cam-
bridge, MA: MIT Press.
boyd, danah
2010 Social network sites as networked publics: Affordances, dynamics and impli-
cations. In: Zizi Papacharissi (ed.), A Networked Self: Identity, Community,
and Culture on Social Network Sites, 39–58. New York: Routledge.
boyd, danah and Nicole B. Ellison
2007 Social network sites: Definition, history, and scholarship. Journal of Computer-
Mediated Communication 13: 210–230.
boyd, danah and Alice E. Marwick
2011 Social privacy in networked publics: Teens’ attitudes, practices, and strategies.
http://ssrn.com/abstract=1925128 (last accessed 9 September 2016).
Bublitz, Wolfram
2008 ‘Sailing the islands or watching from the dock’: The treacherous simplicity of
a metaphor: How we handle ‘new (electronic) hypertext’ versus ‘old (printed)
text’. In: Andrea Gerbig and Oliver Mason (eds.), Language, People, Numbers:
Corpus Linguistics and Society, 249−273. Amsterdam: Rodopi.
Chafe, Wallace L.
1982 Integration and involvement in speaking, writing, and oral literature. In: Deb-
orah Tannen (ed.), Spoken and Written Language: Exploring Orality and Lit-
eracy, 35–53. Norwood, NJ: Ablex.

Authenticated
Collot, Milena and Nancy Belmore

1996 Electronic language: A new variety of English. In: Susan C. Herring (ed.),
Computer-Mediated Communication. Linguistic, Social and Cross-Cultural
Perspectives, 13–28. Amsterdam/Philadelphia: Benjamins.
Cooke, Maeve
2011 Pragmatics in Habermas’ Critical Social Theory. In: Wolfram Bublitz and
Neal Norrick (eds.), Foundations of Pragmatics, 289−312. Berlin/Boston: de
Gruyter.
Crystal, David
2001 Language and the Internet. Cambridge: Cambridge University Press.
Crystal, David
2006 Language and the Internet, 2nd ed. Cambridge: Cambridge University Press.
Crystal, David
2011 Internet Linguistics. A Student Guide. London/New York: Routledge.
Danet, Brenda
2001 Cyberpl@y: Communicating Online. Oxford: Berg.
Danet, Brenda
2010 Computer-mediated English. In: Janet Maybin and Joan Swann (eds.), The
Routledge Companion to English Language Studies, 146–156. London/New
York: Routledge.
Debatin, Bernhard
2011 Ethics, privacy, and self-restraint in social networking. In: Sabine Trepte
and Leonard Reinecke (eds.), Privacy Online. Perspectives on Privacy and
Self-Disclosure in the Social Web, 47–60. Heidelberg/Dordrecht/London/New
York: Springer.
Dienlin, Tobias and Sabine Trepte
2015 Is the privacy paradox a relic of the past? An in-depth analysis of privacy
attitudes and privacy behaviors. European Journal of Social Psychology 45:
285–297.
Dürscheid, Christa
2003 Medienkommunikation im Kontinuum von Mündlichkeit und Schriftlichkeit.
Theoretische und empirische Probleme. Zeitschrift für angewandte Linguistik
38: 37–56.
Dürscheid, Christa
2004 Netzsprache – ein neuer Mythos. In: Michael Beißwenger, Ludger Hoffmann
and Angelika Storrer (eds.), Internetbasierte Kommunikation, 141–157. Osna-
brücker Beiträge zur Sprachtheorie 68. Osnabrück: Gilles und Francke.
Dürscheid, Christa
Dürscheid, Christa and Karina Frick
2014 Keyboard-to-Screen-Kommunikation gestern und heute: SMS und WhatsApp
im Vergleich. Networx 64: 149–181.
Eckert, Penelope
2006 Community of Practice. In: Encyclopedia of Language and Linguistics. http://
web.stanford.edu/~eckert/PDF/eckert2006.pdf (last accessed 15 September
2016).

Authenticated

2010 Once upon a blog … Storytelling in weblogs. In: Christian R. Hoffmann (ed.),
Narrative Revisited. Telling a Story in the Age of New Media, 79–108. Amster-
Ellison, Nicole B., Cliff Lampe, Charles Steinfield and Jessica Vitak
2010 With a little help from my friends: How social network sites affect social
capital processes. In: Zizi Papacharissi (ed.), The Networked Self: Identity,
Community, and Culture on Social Network Sites, 124–145. New York: Rout-
ledge.
Ellison, Nicole B., Jessica Vitak, Charles Steinfield, Rebecca Gray and Cliff Lampe
2011 Negotiating privacy concerns and social capital needs in a social media envi-
ronment. In: Sabine Trepte and Leonard Reinecke (eds.), Privacy Online. Per-
spectives on Privacy and Self-Disclosure in the Social Web, 19–32. Heidel-
berg/Dordrecht/London/New York: Springer.
Eysenbach, Gunther and James E. Till
2001 Ethical issues in qualitative research on internet communities. BMJ 323:
1103–1105.
Gross, Ralph and Alessandro Acquisti
2005 Information revelation and privacy in online social networks (The Facebook
case). Pre-proceedings version. ACM Workshop on Privacy in the Electronic
Society (WPES). https://www.heinz.cmu.edu/~acquisti/papers/privacy-face-
book-gross-acquisti.pdf (last accessed 16 September 2016).
Gruber, Helmut
2013 Mailing list communication. In: Susan C. Herring, Dieter Stein and Tuija Vir-
tanen (eds.), Pragmatics of Computer-Mediated Communication, 55–82. Ber-
lin/Boston: de Gruyter.
Hargittai, Eszter
2008 Whose space? Differences among users and non-users of social network sites.
Herring, Susan C.
2001 Computer-mediated discourse. In: Deborah Schiffrin, Deborah Tannen and
Heidi E. Hamilton (eds.), Handbook of Discourse Analysis, 612−634. Oxford:
Blackwell.
Herring, Susan C.
2004 Computer-mediated discourse analysis: An approach to researching online
behavior. In: Sasha A. Barab, Rob Kling and James H. Gray (eds.), Designing
for Virtual Communities in the Service of Learning, 338–376. New York: Cam-
bridge University Press.
Herring, Susan C.
Internet 4, article 1. http://www.languageatinternet.org/articles/2007/761 (last
accessed 10 August 2016).
Herring, Susan C.
2013 Discourse in Web 2.0: Familiar, reconfigured, and emergent. In: Deborah Tan-
nen and Anna Marie Trester (eds.), Discourse 2.0. Language and New Media,
1–25. Washington D.C.: Georgetown University Press.

Authenticated
Herring, Susan C., Louise Scheidt, Inna Kouper and Elija Wright
2006 A longitudinal content analysis of weblogs: 2003–2004. In: Mark Tremayne
(ed.), Blogging, Citizenship, and the Future of Media, 3–20. London/New
York: Routledge.
Herring, Susan C., Dieter Stein and Tuija Virtanen
2013 Introduction to the pragmatics of computer-mediated communication. In:
puter-Mediated Communication, 3−32. Berlin/Boston: de Gruyter.
internet live stats.
http://www.internetlivestats.com/internet-users/ (last accessed 10 September
2016).
Jarvis, Jeff
2011 Privacy, publicness, and the Web: A manifesto. IEEE SPECTRUM. http://spec-
trum.ieee.org/telecom/internet/privacy-publicness-and-the-web-a-manifesto
(last accessed 5 September 2016).
Johansson, Marjut
2015 Bravo for this editorial! Users’ comments in discussion forums. In: Elda Weiz-
man and Anita Fetzer (eds.), Follow-ups in Political Discourse. Explorations
across Contexts and Discourse Domains, 83–107. Amsterdam/Philadelphia:
Benjamins.
Johansson, Marjut
in press Everyday opinions in news discussion forums: Public vernacular discourse. In
Marjut Johansson, Sonja Kleinke and Lotta Lehti (eds.), The Digital Agora of
Social Media. Special Issue in Discourse, Context & Media.
Johnson, Bobbie
2010 Privacy no longer a social norm, says Facebook founder. The Guardian, 11
January 2010. https://www.theguardian.com/technology/2010/jan/11/face-
book-privacy (last accessed 13 September 2016).
Johnston, Angus
2014 Community and social media. In: Jeremy Hunsinger and Theresa Senft (eds.),
The Social Media Handbook, 18–29. New York/London: Routledge.
Joinson, Adam N., David J. Houghton, Asimina Vasalou and Ben L. Marder
2011 Digital crowding: Privacy, self-disclosure, and technology. In: Sabine Trepte
York: Springer.
Jucker, Andreas H.
2010 “Audacious, brilliant!! What a strike!” – Live-text commentaries on the Inter-
net as real-time narratives. In: Christian R. Hoffmann (ed.), Narrative Revis-
ited. Telling a Story in the Age of New Media, 79–108. Amsterdam/Philadel-
phia: Benjamins.
framework. Linguistik Online 56(6): 39–64.

Authenticated
Kaplan, Andreas and Michael Haenlein

2009 Users of the world unite! The challenges and opportunities of social media.
Business Horizons 53: 59−68.
Kleinke, Sonja
2010 Interactive aspects of computer-mediated communication – ‘Disagreement’ in
an English and a German public news group. In: Sanna-Kaisa Tanskanen, Marja-
Liisa Helasvuo, Marjut Johansson and Mia Raitaniemi (eds.), Discourses in
Kleinke, Sonja
2015 Internetforen: Laiendiskurs Gesundheit. In: Albert v. Busch and Thomas
Spranz-Fogasy (eds.), Handbuch Sprache in der Medizin, 405–422. Berlin/
Boston: de Gruyter.
Kleinke, Sonja and Birte Bös
2015 Intergroup rudeness and the metapragmatics of its negotiation in online dis-
cussion fora. Pragmatics 25(1): 47–71.
Kytölä, Samu and Elina Westinen
2015 ‘I be da reel gansta’ – A Finnish footballer’s Twitter writing and metaprag-
matic evaluations of authenticity. Discourse, Context and Media 8: 6–19.
Lampe, Cliff, Nicole Ellison and Charles Steinfield
2007 A familiar face(book): Profile elements as signals in an online social network.
In: Proceedings of the SIGCHI Conference on Human Factors in Computing
Systems, 435–444. New York: ACM. https://msu.edu/~steinfie/CHI_manu-
script.pdf.
Landert, Daniela
2014 Personalisation in Mass Media Communication. Amsterdam/Philadelphia:
Benjamins.
Lange, Patricia
2008 Publicly private and privately public: Social networking on YouTube. Journal
of Computer-Mediated Communication 13: 361–380.
Lee, Carmen
2011 Texts and practices of micro-blogging: Status updates on Facebook. In: Crispin
Thurlow and Kristine Mroczek (eds.), Digital Discourse. Language in the New
Media, 110–128. Oxford/New York: Oxford University Press.
Leppänen, Sirpa, Janus Spindler Møller, Thomas Rørbeck Nørreby, Andreas Stæhr and
Samu Kytölä
2015 Authenticity, normativity and social media. Discourse, Context and Media 8:
1−5.
Light, Ben and Kathy McGrath
2010 Ethics and social networking sites: A disclosive analysis of Facebook. Infor-
mation Technology and People 23(4): 290–311.

Authenticated
Locher, Miriam A.
2013 Internet advice. In: Susan C. Herring, Dieter Stein and Tuija Virtanen (eds.),
Pragmatics of Computer-Mediated Communication, 339–362. Berlin/Boston:
de Gruyter.
Loosen, Wiebke
2011 Online privacy as a news factor in journalism. In: Sabine Trepte and Leonard
Reinecke (eds.), Privacy Online. Perspectives on Privacy and Self-Disclosure in
the Social Web, 205–218. Heidelberg/Dordrecht/London/New York: Springer.
Madianou, Mirca and Daniel Miller
2013 Polymedia: Towards a new theory of digital media in interpersonal communi-
cation. International Journal of Cultural Studies 16(2): 169–187.
Mallan, Kerry and Natasha Giardina
2009 Wikidentities: Young people collaborating on virtual identities in social net-
work sites. First Monday 14 (6). http://firstmonday.org/ojs/index.php/fm/
article/view/2445/2213 (last accessed 17 September 2016).
Mendelson, Andrew I. and Zizi Papacharissi
2010 Look at us: Collective narcissism in college student Facebook photo galleries.
In: Zizi Papacharissi (ed.), A Networked Self: Identity, Community and Culture
on Social Network Sites, 251–273. New York: Routledge.
Oxford English Dictionary. www.oed.com (last accessed 19 September 2016).
Page, Ruth
2012 Stories and Social Media: Identities in Interaction. London: Routledge.
Page, Ruth, David Barton, Johann W. Unger and Michele Zappavigna (eds.)
2014 Language and Social Media. A Student Guide. London/New York: Routledge.
Papacharissi, Zizi
2010 A Private Sphere. Democracy in a Digital Age. Cambridge/Malden: Polity
Press.
Papacharissi, Zizi
2011 A Networked Self: Identity, Community, and Culture on Social Network Sites.
Papacharissi, Zizi
2014 On networked publics and private spheres in social media. In: Jeremy Hunsin
ger and Theresa Senft (eds.), The Social Media Handbook, 144–158. London/
Papacharissi, Zizi and Paige L. Gibson
2011 Fifteen minutes of privacy: Privacy, sociality, and publicity on social network
sites. In: Sabine Trepte and Leonard Reinecke (eds.), Privacy Online. Perspec-
tives on Privacy and Self-Disclosure in the Social Web, 75–89. Heidelberg/
Dordrecht/London/New York: Springer.
Pérez-Sabater, Carmen
2012 The Linguistics of social networking: A study of writing conventions on Face-
book. Linguistik online. 56(6). https://bop.unibe.ch/linguistik-online/article/
view/257/347 (last accessed 5 September 2016).
Perrin, Andrew
2015 Social media usage 2005–2015. Pew Research Center. http://www.pewinter-
net.org/2015/10/08/2015/Social-Networking-Usage-2005–2015 (last accessed
29 September 2016).

Authenticated
Petronio, Sandra
2000 The boundaries of privacy: Praxis of everyday life. In: Sandra Petronio (ed.),
Balancing the Secrets of Private Disclosures, 37–49. Mahwah, NJ: Lawrence
Erlbaum.
2010 “Thank you for thinking we could”: Use and function of interpersonal pro-
nouns in corporate web logs. In: Heidrun Dorgeloh and Anja Wanner (eds.),
Syntactic Variation and Genre, 167−194. Berlin/New York: de Gruyter.
2013 Blogging. In: Susan C. Herring, Dieter Stein and Tuija Virtanen (eds.), Prag-
matics of Computer-Mediated Communication, 83–108. Berlin/Boston: de
Gruyter.
Rheingold, Howard
1993 The Virtual Community. Reading, MA: Addison-Wesley.
Rosenberg, Åsa
2010 Virtual world research ethics and the private/public distinction. International
Journal of Internet Research Ethics 3: 23–37.
Schmidt, Jan-Hinrik
2011a Das neue Netz. Merkmale, Praktiken und Folgen des Web 2.0. 2nd revised ed.
Konstanz: UVK Verlagsgesellschaft.
Schmidt, Jan-Hinrik
2011b (Micro)blogs: Practices of privacy management. In: Sabine Trepte and Leon-
ard Reinecke (eds.), Privacy Online. Perspectives on Privacy and Self-Disclo-
sure in the Social Web, 159–173. Heidelberg/Dordrecht/London/New York:
Springer.
Schmidt, Jan-Hinrik
2014 Twitter and the rise of personal publics. In: Katrin Weller, Axel Bruns, Jean
Burgess, Merja Mahrt and Cornelius Puschmann (eds.), Twitter and Society,
3–14. New York: Peter Lang.
Schütte, Wilfried
2000 Sprache und Kommunikationsformen in Newsgroups und Mailinglisten. In:
Wer ner Kallmeyer (ed.), Sprache und neue Medien, 142–178. Berlin: de
Gruyter.
2013 Spoken and Written Discourse in Online Interactions. A Multimodal Approach.
New York/London: Routledge.
Stæhr, Andreas
2015 Reflexivity in Facebook interaction – Enregisterment across written and spo-
ken language practices. Discourse, Context and Media 8: 30–45.
Tannen, Deborah
2013 The medium is the metamessage: Conversational style in New Media inter-
action. In: Deborah Tannen and Anna Marie Trester (eds.), Discourse 2.0.
Language and New Media, 99–117. Washington D.C.: Georgetown University
Press.
Thelwall, Mike
2011 Privacy and gender in the Social Web. In: Sabine Trepte and Leonard Reinecke
(eds.), Privacy Online. Perspectives on Privacy and Self-Disclosure in the
Social Web, 251–265. Heidelberg/Dordrecht/London/New York: Springer.

Authenticated
Thompson, John B.
2005 The new visibility. Theory, Culture and Society 22(6): 31–51.
Thurlow, Crispin
2013 Fakebook: Synthetic media, pseudo-sociality, and the rhetorics of Web 2.0. In:
Deborah Tannen and Anna Marie Trester (eds.), Discourse 2.0. Language and
New Media, 225–249. Washington D.C.: Georgetown University Press.
Thurlow, Crispin and Kristine Mroczek
2011 Introduction: Fresh perspectives on New Media Sociolinguistics. In: Crispin
Thurlow and Kristine Mroczek (eds.), Digital Discourse. Language in the New
Media, xix–xliv. Oxford/New York: Oxford University Press.
Trepte, Sabine and Leonard Reinecke (eds.)
2011 Privacy Online. Perspectives on Privacy and Self-Disclosure in the Social
Web. Heidelberg/Dordrecht/London/New York: Springer.
Trepte, Sabine and Leonard Reinecke
2011a Preface. In: Sabine Trepte and Leonard Reinecke (eds.), Privacy Online. Per-
spectives on Privacy and Self-Disclosure in the Social Web, v–vi. Heidelberg/
Dordrecht/London/New York: Springer.
Trepte, Sabine and Leonard Reinecke
2011b The Social Web as a shelter for privacy and authentic living. In: Sabine Trepte
York: Springer.
Tynan, Dan
2016 WhatsApp privacy backlash: Facebook angers users by harvesting their data.
The Guardian, 26 August 2016. https://www.theguardian.com/technology/
2016/aug/25/whatsapp-backlash-facebook-data-privacy-users (last accessed
13 September 2016).
Ungerer, Friedrich and Hans-Jörg Schmid
2006 An Introduction to Cognitive Linguistics, 2nd ed. London/New York: Pearson
Longman.
Walther, Joseph B.
2011 Introduction to privacy online. In: Sabine Trepte and Leonard Reinecke (eds.),
Privacy Online. Perspectives on Privacy and Self-Disclosure in the Social
Web, 3–8. Heidelberg/Dordrecht/London/New York: Springer.
Weintraub, Jeff
1997 The theory and politics of the public/private distinction. In: Jeff Weintraub and
Krishan Kumar (eds.), Public and Private in Thought and Practice. Perspec-
tives on a Grand Dichotomy, 1–42. Chicago/London: University of Chicago
Press.
Weintraub, Jeff and Krishan Kumar (eds.)
1997 Public and Private in Thought and Practice. Perspectives on a Grand Dichot-
omy. Chicago/London: University of Chicago Press.
Wikipedia
List of virtual communities with more than 100 million active users. https://
en.wikipedia.org/wiki/List_of_virtual_communities_with_more_than_100_
million_active_users (last accessed 5 September 2016).
Williams, Raimond
1974 Television, Technology and Cultural Form. London: Routledge.

Authenticated
Wolfe, Alan
1997 Public and private in theory and practice: Some implications of an uncertain
boundary. In: Jeff Weintraub and Krishan Kumar (eds.), Public and Private
in Thought and Practice. Perspectives on a Grand Dichotomy, 182–203. Chi-
cago/London: University of Chicago Press.
Yao, Mike Z.
2011 Self-Protection of online privacy: A behavioral approach. In: Sabine Trepte
Self-Disclosure in the Social Web, 111–125. Heidelberg/Dordrecht/London/
New York: Springer.
Young, Alyson L. and Anabel Quan-Haase
2009 Information revelation and Internet privacy concerns on social network sites:
A case study of Facebook. http://iisi.de/fileadmin/IISI/upload/2009/p265.pdf
(last accessed 17 September 2016).
Yus, Francisco
2011 Cyberpragmatics. Internet-mediated Communication in Context. Amsterdam/
Zhang, Chi, Jinyuan Sun, Xiaoyan Zhu and Yuguang Fang
2010 Privacy and security for online social networks: Challenges and opportunities.
IEEE Network 24(4): 13–18.
Ziegele, Marc and Oliver Quiring
2011 Privacy in social network sites. In: Sabine Trepte and Leonard Reinecke (eds.),
Privacy Online. Perspectives on Privacy and Self-Disclosure in the Social
Web, 175–189. Heidelberg/Dordrecht/London/New York: Springer.
Zimmer, Michael and Nicholas Proferes
2014 Privacy on Twitter, Twitter on privacy. In: Katrin Weller, Axel Bruns, Jean
Burgess, Merja Mahrt and Cornelius Puschmann (eds.), Twitter and Society.

Authenticated
Authenticated
5. Message boards
Jenny Arendholz
Abstract: This chapter provides readers with an introduction to the origins, tech-
nological exigencies and communicative challenges of message boards from a
pragmatic point of view. Furthermore, it offers an in-depth state of the art report
about literature regarding various pragmatic investigations of message boards (and
related online platforms), spanning the period of their inception in the mid 1990s
until today. Most studies cited in this chapter focus on interpersonal pragmatics,
including analyses of interpersonal relations, identity construction and advice giv-
ing or seeking. There are only a handful of papers that adopt a different perspec-
tive, which shows that in contrast to other forms of computer-mediated commu-
nication (CMC) such as emails or blogs, scholarly research on message boards is
still limited, both topically as well as in terms of figures. Message boards still need
to gain a fixed place in standard literature about CMC.
1. Introduction
Over the past two decades, websites inviting discussions have become an integral
part of the Internet. Even pages with a different agenda, such as online newspa-
pers and certain businesses, have set up sections on their website to bring together
like-minded people with similar interests. The topics which are being discussed
are as diverse as the names users give to these kinds of discussion platforms: To
that effect, you can contribute to Internet, Web, online or discussion forums/fora,1
but also message or discussion boards, bulletin boards and (electronic) discussion
groups, or, as Crystal (2006: 134–156) likes to call them, asynchronous chatgroups.
Unfortunately, there is only very little scholarly research and literature (see part 3
of this chapter) dedicated to investigating, let alone defining these different notions.
Consequently, this type of social media is still to large parts uncharted territory. To
guarantee legibility, this chapter will give preference to the notion message boards,
which is among the prevalent terms used in the relevant literature.
The aim of this chapter is twofold: providing readers with a technological
introduction to the social media platform message boards from a pragmatic point
of view and offering an in-depth state of the art report about literature regarding
1
Note that there is no agreement on the formation of the plural among message board
users, let alone in scholarly research, as both terms seem to be used in equal measure.
DOI 10.1515/9783110431070-005
In: C. R. Hoffmann and W. Bublitz (eds.). (2017). Pragmatics of Social Media, 125–149. Berlin/Boston:
De Gruyter Mouton. Brought to you by | Cambridge University Library
Authenticated
126 Jenny Arendholz
various pragmatic investigations of message boards (and related online platforms),

spanning the period of their inception in the mid 1990s until today. In detail, Sec-
tion 2.1 sets out to give an overview of the historic roots. The technological pre-
requisites of message boards and their implications for studies in pragmatics are
depicted in Section 2.2, which also outlines users’ options to partake in message
board interaction in order to exchange information, establish relationships and
present themselves. Section 3 is dedicated to the documentation of the (surpris-
ingly slim) array of pragmatic research on message boards. After some general
remarks about the status of message boards within the comparatively vast amount
of literature dealing with CMC in general (3.1), the following sections bear wit-
ness to the fact that interpersonal pragmatics takes pride of place in the investiga-
tion of online message boards. In particular, Section 3.2 summarizes a relatively
large quantity of monographs and papers which explore facework, politeness,
impoliteness, flaming, conflict and codes of conduct, in short the negotiation of
interpersonal relations, from various angles. Most of them base their analyses
on standard literature such as Goffman (1967), Brown and Levinson (1978) and
Locher and Watts (2005). Section 3.3 recapitulates or references to several studies
about identity construction. As this topic is closely related to interpersonal rela-
tions, a clear-cut differentiation between the two cannot always be effectuated.
The same holds true for Section 3.4, in which literature about advice and authority
are summed up. Finally, Section 3.5 is a multifaceted compilation of studies which
go beyond mainstream (i.e. interpersonal) pragmatic topoi. In contrast to previ-
ously mentioned studies, the ones cited in 3.5 vary also in that message boards
do not necessarily take center stage. Last but not least, Section 4 rounds off this
chapter by offering some concluding remarks. Thus the reader will be introduced
to the technological exigencies and communicative challenges of this type of
social media while, at the same time, gaining a substantial overview of pragmatic
research about this particular Internet platform.
2. Some technological essentials reviewed from a pragmatic perspective
2.1. Origins and early developments

The roots of message boards, as we know them today, can be traced back to the
Usenet, a network of host computers which enables users to “post messages to
newsgroups that can be read (and responded to) by anyone who has access to
the system through a newsreader” (http://harmsen.net/ahrc/news.htm, 02 February
2016). The Usenet, which was originally set up as an experimental bulletin board
system at the University of North Carolina in 1979 (Jasper 1997: 12), comprised
the ever-expanding newsgroups, now hosted by Google and covering every con-
ceivable topic. The current status of these newsgroups, which reached their prime

Authenticated
Message boards 127
in the mid 1990s (Crystal 2006: 135), is somewhat obscure. Some users insist on
their unbroken existence, others perceive them in decline (cf. Yus 2011: 223) or
even as electronic ghost towns. A newsgroup should not be confused with contem-
porary message boards and is defined by Marcoccia (2004: 217) as follows:
A hierarchically-organized forum open to users interested in a specific topic. Discus-

sions take the form of electronic messages (e-mails), which are sent to the forum by
users and filed on internet sites. Users can read the messages filed on the site and also
post new messages. Newsgroups are, in fact, a hybrid of interpersonal and mass com-
munication. The newsgroup’s usenet address indicates its content (for example, fr.rec.
arts.litterature) and defines the forum’s frame, which is essentially thematic.
Although the latter aspect of newsgroups is also shared by message boards, an

important difference between newsgroups and present-day message boards is the
need for a newsreader (cf. MS Outlook Express©, NewsAgent© etc.), a program
to decipher the contents of the newsgroups (cf. Schütte 2000: 150). Since the year
1996, however, web-based versions of this application gained popularity, render-
ing the need for newsreaders obsolete. Another forefather of message boards can
be seen in mailing lists, or a Listserv, which is “a collective repository of e-mails.
Users need to register on a list, and then they will receive the messages sent by
any of its members to its electronic address” (Yus 2011: 224). Present-day message
boards and their predecessors are thus platforms of communication for self-se-
lected participants who collect, categorize and offer contents relevant for other
members, thereby shaping their very own communicative environment both topi-
cally as well as interpersonally.
2.2. Message boards today: Forms and functions

Nowadays, message boards are web based applications which “serve the same
purposes as a face-to-face club meeting, where participants share information and
enthusiasm” (Biber and Conrad 2009: 190). Most of them have administrators
(also known as moderators), i.e. experienced users who monitor ongoing discus-
sions and check the content and form of contributions with regard to their appro-
priateness. The existence of administrators and their extended set of opportunities
of action creates a power differential between users, which can find expression in
constant processes of users’ negotiations of desired form and contents of posts.
Even in cases in which those processes are not argued out publicly and are there-
fore not part of the ongoing discussion, their mere existence and results can be
witnessed: Some users get a negative reputation (which might even be publically
displayed, see below), some posts read “content deleted” (or similar) and other
posts vanish altogether. Naturally, all of this can lead to severe disturbances of
topical coherence, topic shifts or even the temporary or permanent end of the com-
municative exchange (cf. Fritz, Ch. 11, this volume).

Authenticated
128 Jenny Arendholz
Some but not all message boards require the user to register before actively
participating in online discussions. In some cases, this registration is even neces-
sary just to read other users’ contributions. This leads to highly divergent degrees
of commitment, participation and also willingness to open up. Then again mes-
sage boards which do not ask for any kind of registration allow for users to never
indicate their presence and merely read posts without ever becoming active them-
selves, thereby remaining lurkers. The number of (registered) users is therefore
hard to tell and varies greatly from board to board and from topic to topic. For that
reason, users can never be sure who and how many people they actually address.
Despite the strong tendency towards a geographically dispersed, ever-changing
crowd of participants, “the research to date shows that intimate relationships and
development of community are possible online, and that online forums can be
productive and sustainable” (Fayard and DeSanctis 2005).
All message boards share a common goal: They help working professionals,
scholars, researchers, students or, in general, any community of practice (Lave and
Wenger 1991) that is based on shared (sets of) interests to get and stay in touch
with each other, to exchange information and to present themselves. Still, they dif-
fer considerably in layout and structure. Besides a myriad of layout options which
follow the tastes of host institutions and support the purposes of every individual
message board, there are two prototypical, classic structures that can generally be
differentiated, as illustrated in Figure 1 and 2:
As can be seen in Figure 1, the tree structure discloses a chronological and
referential order of the ongoing discussion and thus visualizes the argumentative
maze. This means that all chronologically ordered chains of contributions per-
taining to one particular topic, the so-called threads, can be perceived at once.
Owing to the tree structure, message board users can establish reference relatively
quickly and can recognize at a glance which entry refers to which. Most of these
types of message boards allow for readers to know immediately who started a
thread or commented on it (see the nicknames in italics such as Pat and Wendy in
the first bullet point in Figure 1) as well as how fast other users picked up on that
topic. These pieces of information, nicknames and timestamps, reveal not only
how many users are involved in a discussion but indicate also how urgent and
heated the topic is. This kind of contextual information (see also below, Figure 2)
can prove valuable and illuminating for the interpretation of discourse set in a
communicative format which is otherwise rather context-deprived.
Furthermore, this schematic representation of headings (standing for complete
entries) perfectly visualizes the nature of the ongoing online discussion, which can
be flat at times, for example, when all users respond to one entry (see sixth bullet
point in the lower half of Figure 1), or rather complex, as in the first bullet point.
In the latter case, users do not only respond to the original post itself, but take com-
ments and responses to the first post as a springboard for new contributions. This
convenient technique of structuring polylogues, viz. visualizing the conversational

Authenticated
Message boards 129
Figure 1. Message board with chronological tree-structure, http://www.network54.com/

Forum/358339/page-14, 02 February 2016.
nodes, can also become a source of misunderstandings when inexperienced users

choose the wrong place for their contributions, thus creating unintended links
between unrelated posts and disturbing the conversational coherence.
To see the full text version of the entries, users just have to click on these head-
ings. If one page is filled, a new one is opened automatically with new opening
posts always positioned on top of the latest page. Older contributions usually do
not vanish but are meticulously stored, as they are relegated to other pages pos-
terior to the first page or even to an archive. Discourse produced on such pages
should always be retrievable, and is thus less fleeting than its face to face counter-
part. In this regard, oral communication and discourse produced in message boards
can be seen as two opposite poles of a scale, on which (hand- and type-)written
communication as well as that produced on a computer keyboard take middle posi-
tions. Discourse generated in hand- and type-written documents is certainly less
fleeting then the spoken word and can be read as often as necessary. It is, how-
ever, not stored automatically but instead dependent on individuals’ manual filing
systems. This cannot be said about discourse created with the help of computers.
Although not fleeting either, computer documents can rely on semi-automated,
computer-assisted storing possibilities. After all, someone has to click on “Save”
for the document to be neatly and (hopefully) permanently stored on the hard
drive. In this respect at least, discourse produced with word processing comes
closest to message board interaction.
The second prototypical message board structure offers their users a thematic

Authenticated
130 Jenny Arendholz
approach: Choosing from overarching categories first (e.g. sports, university,

health etc.) and subordinate topics second (e.g. football, spring break, headache
etc.), users are directly led to the full content of the thread discussion. The opening
post can always be found on top of the page with comments being numbered (see
numbers on the upper right of each post in Figure 2) and positioned strictly chron-
ologically (see timestamp) underneath it. Accordingly, and in contrast to the tree
structure, the latest contribution of a long discussion is shown at the bottom of a
page or even on the next page, which might take a little bit to get used to. In fact, it
is important for users to understand this conversational structure and layout, as it is
the prerequisite for their keeping track of the conversation and producing topically
relevant contributions.
In order to (re-)establish topical coherence, users can click on the reply button,
thus creating a quoting box with the original content (see post number 5 in Fig-
ure 2). Users even have the liberty to shorten the original post, if they wish to com-
ment only on a certain part of it. Tailor-made quotes can also be used strategically
in order to twist somebody’s words or to support one’s own argumentative point by
highlighting some and neglecting other parts of someone else’s thoughts. Making
use of the quote button in general also helps to address a fellow user explicitly and
thus to avoid communicative aberrations. After all, posts are by default addressed
to anyone on the board if not marked in any way.
What is also shown in Figure 2 is personal information about the contributors
(see left hand side of each post), which tries to compensate for the lack of social
context cues always present in face to face encounters: Besides nicknames, users
can – at least in some message boards – add a profile picture (also known as the
avatar) and even more intimate – or fake – details about themselves in a profile
section. Profile sections can be used to disclose nationality, current location, study
subjects, join date, and star sign, to name but a few. Depending on the number of
contributions (which is counted automatically and is sometimes even openly dis-
played) and “likes”, contributors can also earn reputation gems or diamonds (the
actual forms depend on the board), indicating not only their social status in the
group but also whether their posts are generally appreciated or not. Some systems
even distribute negative credits (e.g. red gems) to warn others of aggressive users.
Another option to further personalize a post, which is available in a lot of message
boards, is individualized signatures. According to Kollock and Smith this template
is “a combination of business card and bumper stickers that members use to dis-
play their interests, opinions, and occupations” (1999: 10). Signatures are filled
creatively, set to be automatically inserted in every post and appear underneath the
entry. Contextual clues of this nature can serve as a basis for investigations into
aspects of identity construction, self-presentation and face (cf. Goffman 1967).
It should go without saying that the number and shape of these templates is
highly dependent on the respective message board and needs to be considered indi-
vidually for every single investigation. Message boards are a highly chameleonic,

Authenticated
Message boards 131
Figure 2. Message board with topic-driven structure, http://www.thestudentroom.co.uk/,

22 February 2016.
multifunctional and alterable type of CMC and escape generalizations and classi-
fications. One common denominator, however, is the asynchronicity of communi-
cation (cf. Crystal 2006: 135). Users are not pressured by time constraints when
composing or commenting on entries. Consequently, discussions can be continued
with or without interruptions over days, months or even years, a fact that users
must know in order to avoid wrong expectations. Interlocutors can but do not nec-
essarily have to answer immediately, within the next hour or day. In marked con-
trast to spoken conversation, users usually cannot know when or even if (or from
whom) to expect an answer. What is more, this temporal freedom puts users – at
least theoretically – in the comfortable position of composing and editing as well
as interpreting and reflecting on complex and well-wrought entries while taking
their time (Tanskanen 2007: 89, cf. Suler 2005).
Another feature shared by a lot of message boards is sections that help new-
comers (so-called newbies) to find their way around. “About”, “Help” or “FAQs”
sections gather information about technical details of the board and suggest guide-
lines for appropriate behavior (the netiquette) in order to avoid alienating the more
experienced crowd and to uphold a certain social norm (cf. Spiegel and Klein-
berger Günther 2011; Schütte 2000, 2002 for newsgroups and mailing lists; for

Authenticated
132 Jenny Arendholz
netiquette in the context of the Usenet see McLaughlin, Osborne, and Smith 1995).
Among other things, people are advised to avoid flaming (i.e. hostile behavior
towards others, cf. Hardaker, Ch. 18, this volumne) and to stay on topic. In case of
topic drifts, it is recommended to open a new thread rather than to lead an existing
thread into a (probably) unintended direction. Should a user violate the rules of the
game time and again, administrators have the right to ban that user, be it tempo-
rarily or permanently. As the two examples taken from the netiquette show, online
behavior – all differences in the communicative setup aside – should not differ
from social norms prevalent in face-to-face conversations, which users take as a
blueprint both in creating as well as comprehending contributions.
Last but not least, posts are basically phrased with the help of users’ keyboards,
although most of the message boards also allow for multimodality, i.e. the inser-
tion of audio-visual elements, such as pictures, graphics, sound bites, videos, links
etc. Verbal input is thus complemented with stimuli for other senses, which clearly
enrich communicative exchanges via message boards, widen the possibilities for
individuals’ needs for expression and thus go far beyond options offered by spoken
or written texts. Since links to all kinds of websites can be embedded, users can
effortlessly choose from a wide range of prefabricated contents on the Internet,
all of which lend themselves to making contributions more vivid and diversified.
3. Pragmatic studies of message boards
3.1. General observations

Despite the fact that message boards have been enjoying great popularity ever
since, they still take a back seat to the ‘classic’ types of CMC such as chat, web-
sites and email. These are the Internet platforms which are discussed in most detail
and oftentimes even in separate chapters in introductory handbooks and edited
volumes (e.g. Runkehl, Schlobinski and Siever 1998; Crystal 2001; Thurlow, Len-
gel and Tomic 2004; Beißwenger, Hoffmann and Storrer 2004; Androutsopoulos
et al. 2006; Herring, Stein and Virtanen 2013). Although this is not the case for
message boards and their namesakes, they still seem to be gaining momentum in
terms of scholarly perception, at least to a certain degree. This can be seen when
looking, for example, at David Crystal’s publications: While there was no refer-
ence to message boards in 2001, the second edition of the same introduction men-
tions newsgroups and bulletin boards and gives an introduction to asynchronous
chatgroups (Crystal 2006: 134–156). In his latest publication, a student guide from
2011, message boards are again only mentioned in passing. There are, however,
other introductory volumes dedicating at least some space and thought to mes-
sage boards, such as Herring (1996), who mentions bulletin board systems, discus-
sion lists, electronic forums and the Usenet and who presents several studies that

Authenticated
Message boards 133
include forums. Misoch (2006) cites a list from the Internet (but does not comment
on it) summarizing aspects of newsgroup netiquette. Page et al. (2014) also use
examples from message boards, among other types of CMC.
A systematic and comprehensive introduction to the language and linguistics, let
alone pragmatics of message boards, is still missing, though. We do, however, find
very detailed and highly practical, yet mostly outdated introductions to the Usenet
and newsgroups (Bins and Piwinger 1997; Jasper 1997; Fabrot 2001). In addition to
that, very few (online) publications deal with practical aspects of message boards,
i.e. advice and support services on how to access, join or create a message board
(e.g. Donnerhacke 1996; Münz 2002). A great part of that literature focuses on the
technical side of online communication, mostly stressing computer programming.
On the one hand, there are surprisingly few monographs but quite a few papers
which investigate message boards in its own right as a stand-alone subject. On the
other hand, there are also publications that make use of this type of CMC – most
of the time along with others (e.g. email or blogs) – to illustrate their respective
object of investigation with the help of online examples, sometimes contrasting
them with examples from oral and/or written contexts. Both of these two broad
categories have one thing in common, though: the distinct focus on interpersonal
pragmatics, mirrored by studies about facework, (im)politeness and conflict (see
Section 3.2) as well as about identity construction (see Section 3.3). The prefer-
ence for this field of research is also expressed by a special issue of Pragmatics,
edited by Locher, Bolander and Höhn (2015), whose introduction reviews liter-
ature on (im)politeness and relational work in the context of CMC. Publications
addressing other branches of pragmatics are in the minority and will be reviewed
separately (see Sections 3.4 and 3.5).
3.2. Facework, (im)politeness and conflict in message boards

Baym (2000) was the first to adopt an interpersonally pragmatic perspective when
looking at message boards – without even using the terms interpersonal, prag-
matics or message boards, though. Her longitudinal study pioneered in present-
ing the ethnography of an Internet soap opera fan group which interacted via the
female-dominated soap newsgroup rec.arts.tv.soaps, or r.a.t.s. for short. She offers
an in-depth description of the inner workings of this particular online community
and focuses on functions of posts (such as informative and interpretive as well
as their interplay), interpersonal relationships (including creating friendship and
managing conflict), group norms and the development of individual identity (cf.
precursory studies by Baym from 1995a, 1997, 1998). Since Baym’s study is not a
pragmatic analysis per se, introductions to basic analytical frameworks one might
expect in this context (e.g. Goffman 1967; Brown and Levinson 1978) or termi-
nological discussions about central notions such as identity or norm are evidently
missing.

Authenticated
134 Jenny Arendholz
These can, however, be found in Maricic (2005), who, when comparing three
discussions each of a moderated “Musiclassical” mailing list and the non-moder-
ated alt.news-media newsgroup, focuses on facework, (im)politeness and conflict
(cf. also, respectively, Locher and Bolander, Ch. 15, Graham, Ch. 17 and Bolander
and Locher, Ch. 22, this volume). After a short introduction to CMC and discourse,
Maricic reviews the development of the complex notion politeness, a core con-
cept in linguistic pragmatics, with reference to classic models proposed by Lakoff
(1973), Grice (1975) and Leech (1983), to name only three. She presents Brown
and Levinson’s (1978) face-saving model as the theoretical foundation of her anal-
ysis and discusses verbal conflict and flaming, i.e. hostile language in the broadest
sense. Her examination of the two discussion groups shows fifteen different types
of facework, which she subclassifies as confrontational, cooperative and evasive –
categories which serve as complements to Brown and Levinson’s original model.
Linguistic resources used to execute these different types of facework include
terms of address and salutations for cooperative politeness strategies and indef-
inite pronouns and passive constructions for evasive expressive behavior (2005:
202). Her in-depth analysis of the six threads further shows different functions of
facework as well as various conflict management styles. In conclusion, Maricic
disagrees with “the popular belief according to which predominantly textual com-
puter-mediated interaction […] is a ‘faceless’ kind of interaction” (2005: 213).
Instead, she shows that “the participants’ self-, other-, or mutual face concerns
did transpire quite clearly in their messages” (2005: 213), leading her to argue for
increased facework due to the absence of non-verbal cues.
Strategies for controversies in two competing message board threads about
historical arms and armor, both regulated by moderators, are central for Khoras-
ani’s (2009) investigation. He tries to find out “to what extent the new medium
of the Internet has changed the nature of […] controversies” (2009: 12) from the
early modern period to present-day online discussions, with “pamphlets as an early
medium for carrying out controversies” (2009: 17) being the object of comparison.
So far, this is the only study which has been authored by a moderator of an online
discussion group dedicated to that particular topic. After introductions to social
networks, the socio-cultural context of message boards and the object of investiga-
tion on the one hand and to the topic of conflict and controversies on the other, “a
framework to explain the nature of human conflict and its different forms” (2009:
17) is provided. Khorasani also shows communication principles for both time
frames and explains how they changed over time. His main goal is, however, iden-
tifying and analyzing (as well as comparing) individual moves and strategies in
(pamphlets and) message boards, among them “opening moves, attacking moves,
defensive moves, counterattacks, deflecting moves, accusations regarding the vio-
lation of certain principles, neutral moves, the usage of rhetorical questions, and
the usage of direct questions” (2009: 181). Khorasani comes to the conclusion that
“online discussion forums seem to be a modern substitute for pamphlets to carry

Authenticated
Message boards 135
out controversies” (2009: 280), although the communicative principles did change
over time, which is partly due to medium-dependent structures (2009: 278).
On a related note, there are quite a few studies which are interested in (the
verbal expression of) (im)politeness and/or rudeness, disagreement and conflict,
among them Angouri and Tseliga (2010), who analyze “impolite talk” produced
deliberately in two communities of practice, i.e. Greek students and professional
academics, and Kleinke and Bös (2015), who want to know “how participants use
intergroup rudeness as a means of in- and outgroup construction [and] how inter-
group rudeness is metapragmatically negotiated” (2015: 47, cf. LeBlanc 2010). In
her pioneering and oft-cited paper, Herring (1994) puts special emphasis on the
difference between men and women with regard to (im)politeness and flaming,
i.e. “hostile and abusive message content” (1994: 278). She proves the claim that
“men and women have different ideas of what constitutes appropriate and inap-
propriate behavior on the net” (1994: 278), concedes, however, that these gender
differences cannot be explained by politeness alone (1994: 279).
In order to avoid conflict in the first place, users have to negotiate their code
of conduct. Schütte (2000, 2002) takes a closer look at meta-communication in
newsgroups and mailing lists which attempts to set norms among the users in the
form of netiquettes and FAQs. Although a reasonable and necessary step in online
communication, Schütte also shows that merely addressing diverging concepts of
normative behaviour can in itself create new potential for conflict (2000: 144).
The same holds for quotes. Despite the fact that they are a convenient means for
establishing (mostly) unambiguous reference between utterances, they also bear
the risk of polarizing positions between interlocutors. Similarly, Kleinke (2012b,
cf. Arendholz 2015) examines the interface between interpersonal/intergroup
functions (including criticism), quotations and the construction of coherence.
Two sample threads from an English and a German message board accompanying
non-tabloid broadcast media (cf. Kleinke 2007 and 2012a) were subjected to a
qualitative study in order to determine how posts are coherently linked via various
cases of cognitive metonymy to be found in specific patterns of quotations.
In Arendholz (2013), I set out to validate the commonplace remark that com-
munication via the Internet is hostile rather than helpful, let alone interpersonally
appropriate (cf. Zywiczynski 2007). I present a detailed introduction to the tech-
nical and social background of message boards and to classic and more recent
approaches to the whole range of interpersonal relations, including face(work),
(mock-)politeness, (mock-)impoliteness and flaming. Instead of choosing between
the existing models of politeness, I pursue an integrative approach, incorporating
both Brown and Levinson’s seminal model (1978) as well as Goffman’s approach
to face and facework (1967) in one framework. I thus attempt to benefit from both
models’ advantages regarding their view of interpersonal relations. In addition
to that, I also take Locher and Watts’ (2005) model as a valuable blueprint for a
classification of different kinds of interpersonal relations, among them negatively/

Authenticated
136 Jenny Arendholz
positively marked and unmarked behaviour, all of which allow for their evalua-
tion as either appropriate or inappropriate. To account for the vital importance of
context in the evaluation of interpersonal behavior, I shed light on contextual fac-
tors in general and refer to their concrete realizations in the message board envi-
ronment under investigation, i.e. one of the largest British message boards, The
Student Room. My quantitative analysis of 50 threads (ca. 3300 posts, amounting
to ca. 285,000 words) shows that 92.3 % of all behavior is indeed appropriate (of
which 85.6 % is unmarked behavior) and only 7.7 % is inappropriate – as estimated
by fellow message board interlocutors present in the discussions, who serve as the
basis for the evaluation of their co-users’ behavior. In the qualitative part of the
analysis, I also detail the various kinds of facework with ample examples from my
corpus. In follow-up studies based on the same message board, Arendholz (2014)
details processes of face negotiation, whereas Arendholz (2015) investigates the
interface between interpersonal relations and quoting (cf. Kleinke 2012b) by ask-
ing how the latter can serve as a tool for participants to redefine their relation-
ships.
The latest monograph about facework in message boards is presented by Fröh-
lich (2015). In terms of overall research question as well as structure and content
of the first part of the book, this publication is quite similar to Arendholz (2013),
as it (also) offers a detailed overview of the field of linguistic politeness and face-
work on the one hand and the ins and outs of communication in message boards
on the other. However, Fröhlich’s study (written in German) puts special emphasis
on “Multicodalität”, i.e. the interplay of different codes, such as written texts and
pictures, by detailing her reading of code as opposed to mode (cf. e.g. Stöckl 2004;
Kress and van Leeuwen 2001) and by discussing features and relevant approaches
to images, including the pictorial turn (cf. e.g. Sachs-Hombach 2003, 2009). Also,
her analysis is based on eight Spanish-speaking threads, amounting to 1003 post-
ings, which deal with the topics cinema, film and TV-series. The computer-as-
sisted qualitative and multicodal data analysis conducted by the software program
MAXQDA shows that users are rather concerned with the protection of their own
face, i.e. self-face (as opposed to other-face, 2015: 331). They do so verbally as
well as non-verbally (2015: 337). Then again the topic shared in these threads pro-
motes what Fröhlich calls “collective-face” (2015: 331). The author also details
the use of different codes in that particular online environment and stresses that
facework is a continuum, which is why politeness should never be investigated as
an isolated phenomenon but as one component of a larger framework, including
also impoliteness (2015: 343).
A contrastive field study by Kleinke (2012a) supports the idea that, similar to
face-to-face communication, interpersonal relations are also negotiated among the
participants in message board discussions (cf. Launspach 2000). She compares
900 postings to 22 German and 22 English message boards, paying attention to
the special framing conditions of message boards. In investigating frequencies

Authenticated
Message boards 137
and patterns of responses to rhetorical questions, the author illustrates differences

and similarities between the two groups and distinguishes between ‘true answer’
and ‘replies’, both of which were given not only by the addressees but also by the
addressers. In an earlier study of the same German message board, Kleinke (2007)
focuses on hurtful comments to express rejection of the contents of interlocutors’
posts. Kleinke’s survey from 2008, which can be allocated a peripheral place in
the field of interpersonal communication, investigates how emotional commitment
is articulated in political discourse among laypersons found in 180 posts in three
political message boards. She sets out to cover the continuum between emotional
attachment and emotional detachment and draws comparisons to related registers,
i.e. private conversations and newspaper registers.
There are also studies that pay attention to only one particular pragmatic
marker, among them Landone (2012), who looks at discourse markers imported
from oral conversations in order to regulate politeness in a Spanish message board.
She concludes that the technological setup and the participation framework create
medium-specific discourse markers (2012: 1799). Grozeva (2015) is interested in
hedges as a verbal expression of politeness, fuzziness and vagueness in political
discourse found in selected blogs and message boards.
3.3. Identity construction in message boards

Identity construction in social media in general is investigated in Locher and
Bolander (Ch. 15, this volume). But there are a number more specific studies which
discuss identity construction in message boards. Hanna and De Nooy (2009), for
instance, focus on public Internet discussion boards attached to media websites,
which they regard as “a neglected genre” despite their intriguing “possibilities for
authentic language and intercultural contact” (2009: 1). As both authors are lectur-
ers in French, they are interested in the pedagogical potential of message boards
(cf. Nguyen and Kellogg 2005; Farrell Whitworth 2009 for pragmatic awareness
in the context of L2). They try to find out, “[h]ow […] intercultural communica-
tion [is] negotiated in online discussion” (2009: 2). By analyzing English (The
Guardian) and French (Le Monde) websites, they explore cultural differences
and draw up “some guidelines for teaching with and for the use of such sites”
(2009: 2). An outline of the interplay between culture and the Internet – in gen-
eral and with a focus on language learners in particular – is followed by a short
discussion of identity construction, which, in accordance with Goffman (1967), is
viewed as “a discursive practice” (2009: 124). Since identities are not “the expres-
sion of a pre-existing, stable, unified, authentic self” (2009: 124), Hannah and
De Nooy give examples for creating oneself online. In a case study, they touch
upon Goffman’s concept of footing (1981: 128), which they define as “a stance
or an alignment […] with other participants through [contributors’] projection of
a self in their message” (Hannah and De Nooy 2009: 148). Among other things,

Authenticated
138 Jenny Arendholz
the authors conclude that the renegotiation of footing is constant in public Internet
discussions and can be ascribed to the complex participation framework in this
particular communicative environment (2009: 148).
Chandrasegaran and Kong (2006, cf. Chandrasegaran 2006) also focus on stance
in online discussions, their target group being, however, high-school students with
non-native English-speaking backgrounds. The authors learn that “stance-taking
and stance-support are a part of the average high school students’ repertoire of
social interaction skills” (2006: 389) and find evidence for stance-projection acts
such as the evaluation of issues and the expression of attitudinal meanings, to
name only two.
Identity construction in message boards also plays a role in Benwell and
Stokoe’s (2006) monograph about discourse and identity. Different theories and
types of identities (cf. Taylor 1989) are exemplified with the help of three con-
texts, among them virtual identities. Following the discussion of central notions
such as virtual, virtual identity as opposed to real identity, the authors set out
to “explore how the absence of face-to-face interaction impacts on identity con-
struction” (2006: 13), leading to the description of unique linguistic characteristics
and defining features to be found in their message board data. They also outline,
among others, selected pragmatic dimensions by touching on Grice’s Cooperative
Principle (1975) and its implications for this particular context and offer a case
study about newbies’ identity constructions as opposed to those by regular users
in various kinds of message boards. Finally, Benwell and Stokoe briefly discuss
the newbie identity within the two classic frameworks to politeness (Brown and
Levinson 1978 and Goffman 1967).
In addition to that, we find lots of loosely connected case studies. Matthews
(2010), for instance, explores effects of group identity among computer profes-
sionals in contrast to a group of New Age supporters. She concentrates on the
types of exchange in order to find out whether the content of the boards matches
the communicative intentions (e.g. information exchange or a mere chat). Haugh,
Chang and Kádár (2015) analyze deference (cf. Brown and Levinson 1978) as a
relational practice and as a means of identity construction in a Chinese message
board. In doing so, and focussing on participation order and preferred responses
to troubles talk in Taiwanese-Chinese parenting discussion boards, they provide
support for the position that relational practices may intersect with the emergence
of identities. In his qualitative and quantitative analysis of health identity con-
struction, Marko (2012) examines discussion forums on headache and migraines,
perceiving those forums as “a potential site for challenging roles defined by med-
icine and institutional healthcare” (Marko 2012: 245). Sawyer (2013) presents a
microanalysis of linguistic cues in message board contributions which are used by
one individual poster to construct identity. According to Sawyer, “identity is not
a personal projection of beliefs, values and stances but is instead a collaborative
effort between self and others” (2013: 194). Burkhalter (1999) takes as a spring-

Authenticated
Message boards 139
board the fact that physical cues are lacking in online communication, which is,
however, not an obstacle for the creation of racial identity as can be found in
Usenet discussions. The same underlying assumption is expressed in a very influ-
ential paper by Donath (1999). She also has a look at discussion groups at Usenet
but pays attention to the special case of identity deception, i.e. the deliberate act of
misguiding other users by pretending to be someone else, also called trolling (cf.
Herring et al. 2002, also Hardaker, Ch. 18, this volume). Baym (1995b) explores
the performance of humor in this type of CMC, disproving the claim that “the
medium is inhospitable to humor” (1995b). Her analysis of five exemplary humor-
ous messages (taken from r.a.t.s, see Section 3.2) shows that humor can even be
seen as a key component to the creation of social bonds and “facilitates self pres-
entations, common understandings, group solidarity and identity and discussion
of problematic aspects of the social world” (1995b). For that reason, Baym labels
her analysis “a response to constructions of computer-mediated communication
as a socially-impoverished domain” (1995b). Fayard and DeSanctis (2005) apply
Wittgenstein’s (1953) language game framework to document the three dimen-
sions roles, social identity, and linguistic style in a case study on an Indian knowl-
edge management message board as a gathering place for information systems
professionals. Paying special attention to gendered practices, Herrmann (2007) is
interested in finding out about the creation and maintenance of identity and com-
munity in a discussion board populated by stockholders. His analysis of 18 threads
reveals three main identities taken up by users, viz. mentor, intelligent investor and
admirer in predominantly masculine discourse. Eisenchlas (2012) adopts a similar
perspective when exploring the impact of CMC in the construction of gendered
identities in the context of advice giving in an Argentinean message board (cf.
Section 3.4) and examines how men and women produce and are given advice in
terms of directness, politeness considerations and affect display (cf. Section 3.2).
3.4. Advice and authority in message boards

Another branch of interpersonal pragmatics that caught the interest of quite a few
scholars is advice giving (cf. Locher 2006, also Morrow, Ch. 24, this volume).
Harvey and Koteyko (2013), for instance, explore health communication in spoken,
written and computer-mediated contexts, as represented by doctor-patient encoun-
ters, patient records and information leaflets as well as online support groups respec-
tively. The latter ones, subdivided into peer-to-peer communication and contacts
with health professionals, focus on “the types of advice-seeking routines […], how
peers provide emotional support and advice […], how individuals communicate
problems to health professionals [and on] the expressive potential of electronic dis-
course” (2013: 3). Rather marginally, Harvey and Koteyko’s analysis also touches
upon notions such as politeness and credibility, concluding that self-introductions
are not just lists of facts about users’ medical conditions but acts of storytelling

Authenticated
140 Jenny Arendholz
(cf. Arendholz 2010), including different kinds of facework (2013: 186). The
authors also emphasize the fact that in peer-to-peer advice giving, predefined hier-
archies to draw on with regard to authority and expertise do not exist (2013: 187).
Discursive moves in a Japanese online advice forum about divorce take center
stage in Morrow’s study (2012, see also Morrow 2006 about British users in threads
about depression). These moves are analysed in terms of their frequency, sequenc-
ing and syntactic forms (e.g. requests and suggestions but not imperatives) in order
to identify “distinctive characteristics of advice-giving in the context of the Jap-
anese internet discussion forum” (2012: 275). The most frequent move found in
the data was assessment, followed by advice. Morrow points out that his findings
are reverse to Locher’s (2006) findings for an American Internet advice column.
These contradictory results are, however, ascribed to differences in the classifica-
tion procedure but also to the fact that Morrow investigated peer-advice, whereas
Locher focused on expert advice-givers (2012: 275). This may also explain why
Locher did find imperatives in her 2006 study. Japanese cultural values also serve
as a backdrop for the investigation of the establishment of relationships and of the
expression of stance such as showing empathy and presenting oneself as a fellow
sufferer (2012: 276).
An additional intercultural momentum to this topic can be perceived in Ruble’s
(2011) analysis of English-speaking language assistants in French schools. Among
other communicative functions to be found in these message boards exchanges,
the author lists gaining necessary information about life and work, ranting about
difficult circumstances, commiserating with others in similar situations, and devel-
oping relationships, so-called weak ties, which would not exist, if it were not for
the online contact (2011: 415–416). Ruble also points out that weak-tie support (as
opposed to strong ties between interactants who know each other really well) tends
to be superficial and runs the risk of being inaccurate.
3.5. Message boards beyond mainstream pragmatic topoi

With very few exceptions, the remainder of this chapter summarizes basically iso-
lated studies which investigate various topics that display a pragmatic perspective
that deviates from the mainstream trends outlined in the previous sections. What
these studies have in common, though, is their (at times marginal) discussion of
message board examples.
Virtanen (2013) proposes the term mock performatives, i.e. the doing of an act
by typing it, for utterances such as “I hereby grant you permission to blow cash on
how you look” (2013: 155) and investigates their forms and discourse-pragmatic
functions with the help of data from a discussion board on beauty and fashion. In
this regard, she also explores discourse transformers, a term she coined to account
for “the work that mock performatives do in signaling a shift to a play mode”
(2013: 156).

Authenticated
Message boards 141
Arendholz and Kirner-Ludwig (2015) contrast present-day message board sig-

natures with medieval manuscripts in view of (what looks like) quotes to be found
in both text types. They use examples from both periods to illustrate that classify-
ing, let alone interpreting the communicative functions of these quotes turns out
to be quite complicated, since the presence or even the absence of formal conven-
tions such as verba dicendi or inverted commas is not always a reliable indicator.
Thus, the term cognitively isolated quote/reference seems adequate to account for
that phenomenon. For instance, one example taken from a message board signa-
ture reads “If you strike me down, I shall become more powerful than you could
possibly imagine” (2015: 325). Although characterized by the absolute absence
of formal markers, this statement will easily be classified as a quote by members
of one particular group, i.e. Star Wars connoisseurs, but remain a – cognitively
isolated – mystery to outsiders.
The fact that message boards can also be used as a worthwhile object of study
for investigating narratives is shown by Arendholz (2010) and Page (2012). In
successfully applying two frameworks originally designed for narrative structures
in spoken exchanges, Labov and Waletzky (1967) and Ochs and Capps (2001),
my case study shows that, at least in this regard, message boards should be con-
sidered a hybrid genre between spoken and written language. After all, four mes-
sage board posts of varying lengths (ranging between 28 and 2497 tokens), all
taken from www.oprah.com, clearly exhibit features of oral narratives as proposed
by Labov and Waletzky’s model (e.g. orientation, complication, resolution, coda
and evaluations), thus serving as valid examples for online narratives. Interest-
ingly enough, applying Ochs and Capps’ model led to different results, position-
ing the sample posts in the vicinity of “default narratives”, which are normally
encountered in monologic, static and above all written contexts (Arendholz 2010:
140). Page (2012) dedicates one chapter of her anthology about stories in social
media to second stories told in message boards. Second stories follow a preced-
ing narrative and “establish clear parallels with the first story” (2012: 30). The
author reports about how the collaborative potential of message boards reshapes
the narrative dimensions of tellership, focusing particularly on sequenced story-
telling created in dialogic structures, which leads to multiple tellers. Page also
considers second stories in the context of face, politeness and relational work (cf.
Section 3.2). She interfaces them with giving or seeking advice, face-enhanc-
ing and face-threatening acts, and in the context of creating identity (cf. Section
3.3).
Text samples from message boards also serve as examples for discussions of
the notions register, genre and style by Biber and Conrad (2009). They dedicate
ten pages of their monograph to the investigation of the situational characteristics
(i.e. the communicative setting) and linguistic features of what they call “E-forum
postings” (cf. Brock and Seidel 2002 for a discourse analytic study about the text
type of message boards).

Authenticated
142 Jenny Arendholz
Last but not least, Tanskanen (2007) presents a study about metapragmatic
utterances in asynchronous CMC, viz. in a selection of mailing lists and mes-
sage board discussions, asking how and for what purposes participants use meta
pragmatic utterances. In the end, the author suggests a collaborative purpose in
that users anticipate potential problems, try to avoid misinterpretations and work
towards successful communicative exchanges (2007: 103–104).
4. Conclusion
This chapter about message boards consists of two parts. It explores their his-
toric technological and communicative development over the last two decades and
assembles as well as documents the pragmatic research dedicated to the study
of this particular type of CMC. A vast amount of different shapes and forms of
message boards in terms of layout, topical orientation, but also pragmatic func-
tions is attested in both of these parts. As a consequence, it does not come as a big
surprise that message boards enjoy great popularity and are mushrooming all over
the Internet. What is really surprising, however, is the relatively limited amount
of pragmatic literature which covers message board communication. Most stud-
ies cited in this chapter focus on interpersonal pragmatics, including analyses of
interpersonal relations, identity construction and advice giving or seeking. There
are only a handful of papers that adopt a different perspective, which shows that in
contrast to other forms of CMC such as email or blogs, scholarly research on mes-
sage boards is still limited, both topically as well as in terms of figures. A whole
range of different names referring to basically the same type of online platform as
well as the inconsistent use of the plural in the word forum is a case in point for
the fact that the establishment of a homogeneous research tradition for message
boards is still underway.
References
Androutsopoulos, Jannis K., Jens Runkehl, Peter Schlobinski and Torsten Siever (eds.)
2006 Neuere Entwicklungen in der linguistischen Internetforschung. Hildesheim:
Olms.
2010 ‘you HAVE NO IDEA WHAT YOU ARE TALKING ABOUT!’ From e-dis
1: 57–82.
Arendholz, Jenny
2010 ‘Neeed to put this out there (My Story)’ – Narratives in message boards. In:
Christian R. Hoffmann (ed.), Narrative Revisited: Telling a Story in the Age of
New Media, 109–142. Amsterdam/Philadelphia: Benjamins.

Authenticated
Message boards 143
Arendholz, Jenny
2013 (In)Appropriate Online Behavior. A Pragmatic Analysis of Message Board
Arendholz, Jenny
2014 ’You sound very talented’ – Negotiating face in online message boards. In:
Christiane Maaß, Gudrun Held and Kristina Bedijs (eds.), Face Work and
Social Media, 213–234. Berlin: LIT Verlag.
Arendholz, Jenny
2015 Quoting in online message boards. An interpersonal perspective. In: Jenny
Arendholz, Wolfram Bublitz and Monika Kirner-Ludwig (eds.), The Pragmat-
ics of Quoting Now and Then, 53–69. Berlin/Boston: de Gruyter.
Arendholz, Jenny and Monika Kirner-Ludwig
2015 In-between cognitively isolated quotes and references: Looking for answers
lurking in textual margins. In: Jenny Arendholz, Wolfram Bublitz and Monika
Baym, Nancy K.
1995a The emergence of community in computer-mediated communication. In: Ste-
ven G. Jones (ed.), CyberSociety: Computer-Mediated Communication and
Community, 138–163. Thousand Oaks/CA: Sage.
Baym Nancy K.
1995b The performance of humor in computer-mediated communication. Journal
of Computer-Mediated Communication 1 (2). http://onlinelibrary.wiley.com/
doi/10.1111/j.1083-6101.1995.tb00327.x/full, 12 January 2016.
Baym, Nancy K.
1997 Interpreting soap operas and creating community: Inside an electronic fan
culture. In: Sara Kiesler (ed.), Culture of the Internet, 103–120. Mahwah/NJ:
Lawrence Erlbaum.
Baym, Nancy K.
1998 The emergence of online community. In: Steven G. Jones (ed.), CyberSociety
2.0: Revisiting Computer-Mediated Communication and Community, 35–68.
Thousand Oaks/CA: Sage.
Baym, Nancy K.
2000 Tune In, Log On: Soaps, Fandom, and Online Community. Thousand Oaks/
CA: Sage.
Beißwenger, Michael, Ludger Hoffmann and Angelika Storrer (eds.)
2004 Internetbasierte Kommunikation. Special issue, Osnabrücker Beiträge zur
Sprachtheorie 68.
Benwell, Bethan and Elizabeth Stokoe
2006 Discourse and Identity. Edinburgh: Edinburgh University Press.
Biber, Douglas and Susan Conrad
2009 Register, Genre, and Style. Cambridge: Cambridge University Press.
Bins, Elmar and Boris-A. Piwinger
1997 Newsgroups. Weltweit diskutieren. Bonn: International Thomson Publishing.
Brock, Alexander and Beate Seidel
2002 Text types in computer-mediated communication. The case of discussion
forums. In: Christian Todenhagen (ed.), Text – Text Structure – Text Type,
13–28. Tübingen: Stauffenburg-Verlag.

Authenticated
144 Jenny Arendholz

1978 Universals in language usage: Politeness phenomena. In: Esther N. Goody
(ed.), Questions and Politeness, 56–289. Cambridge: Cambridge University
Press.
Burkhalter, Byron
1999 Reading race online: discovering racial identity in Usenet discussions. In:
Marc A. Smith and Peter Kollock (eds.), Communities in Cyberspace, 60–75.
London: Routledge.
Chandrasegaran, Antonia
2006 Secondary school students’ stance-support strategies in online discussion:
implications for the composition classroom. English in Education 40(2):
22–39.
Chandrasegaran, Antonia and Kah Mun Clara Kong
2006 Stance-taking and stance-support in students’ online forum discussion. Lin-
guistics and Education 17(4): 374–390.
Crystal, David
Crystal, David
2006 Language and the Internet. Second edition. Cambridge: Cambridge University
Press.
Crystal, David
2011 Internet Linguistics. A Student Guide. London/New York: Routledge.
Donath, Judith S.
1999 Identity and deception in the virtual community. In: Marc A. Smith and Peter
Kollock (eds.), Communities in Cyberspace, 29–59. London: Routledge.
Donnerhacke, Lutz
1996 Usenet: Die Einrichtung von Diskussionsforen. In: Martin Rost (ed.), Die
Netz-Revolution: Auf dem Weg in die Weltgesellschaft, 70–75. Frankfurt/Main:
Eichborn.
Eisenchlas, Susana A.
2012 Gendered discursive practices on-line. Journal of Pragmatics 44(4): 335–345.
Fabrot, Bernard
2001 Newsgroups: The.little.Internet.guide. London: Cassell Illustrated.
Farrell Whitwort, Kathleen
2009 The discussion forum as a locus for developing L2 pragmatic awareness. In:
Lee B. Abraham and Lawrence Williams (eds.), Electronic Discourse in Lan-
guage Learning and Language Teaching, 291–316. Amsterdam/Philadelphia:
Benjamins.
Fayard, Anne-Laure and Gerardine DeSanctis
2005 Evolution of an online forum for knowledge management professionals: A lan-
guage game analysis. Journal of Computer-Mediated Communication 10 (4).
http://onlinelibrary.wiley.com/doi/10.1111/j.1083-6101.2005.tb00265.x/full,
2 November 2016.
Fröhlich, Uta
2015 Facework in multicodaler spanischer Foren-Kommunikation. Berlin/Boston:
de Gruyter.

Authenticated
Message boards 145
Goffman, Erving (ed.)

1967 Interaction Ritual: Essays on Face-to-face Behavior. Garden City, NY: Anchor
Books.
Goffman, Erving
1981 Forms of Talk. Oxford: Blackwell.
Grice, H. Paul
1975 Logic and conversation. In: Peter Cole and Jerry L. Morgan (eds.), Syntax and
Semantics, Volume 3: Speech Acts, 41–58. New York: Academic Press.
Grozeva, Maria
2015 Linguistische ‘Hecken’ im Internetforum. In: József Tóth (ed.), Die Sprache
und ihre Wissenschaft zwischen Tradition und Innovation, 353–366. Frankfurt/
Main: Lang.
Hanna, Barbara and Juliana De Nooy
2009 Learning Language and Culture via Public Internet Discussion Forums. Bas-
ingstoke: Palgrave Macmillan.
Harvey, Kevin and Nelya Koteyko
2013 Exploring Health Communication: Language in Action. London: Routledge.
phia: Benjamins.
Haugh, Michael, Wei-Lin Melody Chang and Dániel Z. Kádár
2015 ‘Doing deference’: Identities and relational practices in Chinese online discus-
sion boards. Pragmatics 25(1): 73–98.
Herring, Susan
1994 Politeness in computer culture: Why women thank and men flame: In: Mary
Bucholtz, A. C. Liang, Laurel A. Sutton and Caitlin Hines (eds.), Communi-
cating in, through, and across Cultures: Proceedings of the Third Berkeley
Women and Language Conference, 278–294. Berkeley: Berkeley Women and
Language Group.
Herring, Susan (ed.)
1996 Computer-Mediated Communication. Linguistic, Social and Cross-cultural
Perspectives. Amsterdam/Philadelphia: Benjamins.
Herring, Susan, Kirk Job-Sluder, Rebecca Scheckler and Sasha Barab
2002 Searching for safety online: Managing ‘trolling’ in a feminist forum. The
Information Society 18(5): 371–384.
2013 Pragmatics of Computer-Mediated Communication. Berlin/Boston: de Gruyter.
Herrmann, Andrew F.
2007 ‘People get emotional about their money’. Performing masculinity in a financial
discussion board. Journal of Computer-Mediated Communication 12: 499—
522. http://onlinelibrary.wiley.com/doi/10.1111/j.1083-6101.2007.00335.x/
full, 15 November 2016.
Jasper, Dirk
1997 Internet Newsgroups: Suchen, Anzapfen, Nutzen, Diskutieren. Düsseldorf:
Econ.

Authenticated
146 Jenny Arendholz
Khorasani, Manouchehr Moshtagh

2009 The Development of Controversies: From the Early Modern Period to Online
Discussion Forums. Bern: Peter Lang.
Kleinke, Sonja
2007 Sprachliche Strategien verbaler Ablehnung in öffentlichen Diskussionsforen
im Internet. In: Steffen K. Herrmann (ed.), Verletzende Worte. Die Grammatik
sprachlicher Missachtung, 311–336. Bielefeld: Transcript-Verlag.
Kleinke, Sonja
2008 Emotional commitment in public political Internet message boards. Journal of
Language and Social Psychology 27(4): 409–421.
Kleinke, Sonja
2012a Responses to rhetorical questions in English and German Internet public news
groups. Functions of Language 19(2): 174–200.
Kleinke, Sonja
2012b Metonymic inferencing and metonymic elaboration in quotations: creating
coherence in a public Internet discussion forum. In: Sonja Kleinke, Zoltán
Kövecses, Andreas Musolff and Veronika Szelid (eds.), Cognition and Cul-
ture: the role of metaphor and metonymy, 87–98. Budapest: ELTE Eötvös
Kiadó.
Kleinke, Sonja und Birte Bös
Kress, Gunther R. and Theo van Leeuwen
2001 Reading Images. The Grammar of Visual Design. London/New York: Rout-
ledge.
Kollock, Peter and Marc A Smith
1999 Communities in cyberspace. In: Marc A. Smith and Peter Kollock (eds.), Com-
munities in Cyberspace, 3–25. London: Routledge.
1967 Narrative analysis: Oral versions of personal experience. In: June Helm (ed.),
Essays on the Verbal and Visual Arts: Proceedings of the 1966 Annual Spring
Meeting of the American Ethnological Society, 12–44. Seattle: University of
Washington Press.
Lakoff, Robin T.
1973 The logic of politeness: Or minding your p’s and q’s. Papers from the Ninth
Regional Meeting of the Chicago Linguistic Society: 292–305.
Landone, Elena
2012 Discourse markers and politeness in a digital forum in Spanish. Journal of
Pragmatics 44(13): 1799–1820.
Launspach, Sonja
2000 Literal or loose talk: The negotiation of meaning on an Internet discussion list.
In: Lyn Pemberton and Simon Shurville (eds.), Words on the Web: Computer
Mediated Communication, 87–95. Exeter/Portland: Intellect Books.
Lave, Jean and Etienne Wenger
1991 Situated Learning: Legitimate Peripheral Participation. Cambridge: Cam-

Authenticated
Message boards 147
LeBlanc, Tracy
2010 Impoliteness as a model for virtual speech community building. In: Rotimi
Taiwo (ed.), Handbook of Research on Discourse Behavior and Digital Com-
munication: Language Structures and Social Interaction. Volume 1, 523–539.
Hershey: Information Science Reference.
Leech, Geoffrey
1983 Principles of Pragmatics. London: Longman.
Locher, Miriam A.
2006 Advice Online: Advice-giving in an American Internet Health Column. Amster-
Locher, Miriam A., Brook Bolander and Nicole Höhn
2015 Introducing relational work in Facebook and discussion boards. Pragmatics 25
(1): 1–21.
Locher, Miriam A. and Richard J. Watts
2005 Politeness theory and relational work. Journal of Politeness Research 1(1):
9–33.
Marcoccia, Michael
2004 On-line polylogues: Conversation structure and participation frameworks in
Internet newsgroups. Journal of Pragmatics 36: 115–145. http://www.science-
direct.com/science/article/pii/S0378216603000389, 09 March 2016.
Maricic, Ibolya
2005 Face in Cyberspace: Facework, (Im)Politeness and Conflict in English Dis-
cussion Groups. Växjö: Växjö University Press.
Marko, Georg
2012 My painful self: Health identity construction in discussion forums on head-
ache and migraines. AAA, Arbeiten aus Anglistik und Amerikanistik 37(2):
245–272.
Matthews, Heather
2010 Effects of group identity on discussions in public on-line fora. In: Lyn Pem-
berton and Simon Shurville (eds.), Words on the Web: Computer Mediated
Communication, 79–86. Exeter/Portland: Intellect Books.
McLaughlin, Margaret L., Kerry K. Osborne and Christine B. Smith
1995 Standards of conduct on Usenet. In: Steven G. Jones (ed.), CyberSociety: Com-
puter-Mediated Communication and Community, 90–111. Thousand Oaks/
CA: Sage.
Misoch, Sabina
2006 Online-Kommunikation. Stuttgart: UVK-Verlagsgesellschaft.
Morrow, Phillip R.
2006 Telling about problems and giving advice in an Internet discussion forum.
Discourse Studies 8(4): 531–548.
Morrow, Phillip R.
2012 Online advice in Japanese: Giving advice in an Internet discussion forum. In:
Holger Limberg and Miriam A. Locher (eds.), Advice in Discourse, 255–280.
Münz, Stefan
2002 Foren und Boards. http://aktuell.de.selfhtml.org/artikel/gedanken/foren-
boards/, 09 March 2016.

Authenticated
148 Jenny Arendholz
Nguyen, Hanh Thi and Guy Kellogg

2005 Emergent identities in online discussions for second language learning. The
Canadian Modern Language Review 62: 111–136.
Ochs, Elinor and Lisa Capps
2001 Living Narratives. Creating Lives in Everyday Storytelling. Cambridge/MA:
Harvard University Press.
Page, Ruth E.
2012 Stories and Social Media. New York: Routledge.
Page, Ruth E., David Barton, Johann W. Unger, and Michele Zappavigna
2014 Researching Language and Social Media. A Student Guide. London: Rout-
ledge.
Ruble, Racheal A.
2011 The communication of advice on an online message board for language assis-
tants in France. Journal of Language and Social Psychology 30(4): 396–420.
Runkehl, Jens, Peter Schlobinski and Torsten Siever
1998 Sprache und Kommunikation im Internet. Überblick und Analysen. Opladen:
Westdeutscher. Verlag.
Sachs-Hombach, Klaus
2003 Das Bild als kommunikatives Medium. Elemente einer allgemeinem Bildwis-
senschaft. Köln: Herbert von Halem.
Sachs-Hombach, Klaus
2009 Bildtheorien. Anthropologische und kulturelle Grundlagen des Visualistic
Turn. Frankfurt/Main: Suhrkamp.
Sawyer, Sheryl
2013 Exploring identity construction: a case study from an online forum. Tekst I
dyskurs 6: 241–261.
Schütte, Wilfried
2000 Sprache und Kommunikationsformen in Newsgroups und Mailinglisten. In:
Werner Kallenmeyer (ed.), Sprache und neue Medien, 142–178. Berlin/Bos-
ton: de Gruyter.
Schütte, Wilfried
2002 Normen und Leitvorstellungen im Internet: Wie Teilnehmer/-innen in News-
groups und Mailinglisten den angemessenen Stil aushandeln. In: Inken Keim
and Wilfried Schütte (eds.), Soziale Welten und kommunikative Stile: Fest-
schrift für Werner Kallmeyer zum 60. Geburtstag, 339–362. Tübingen: Narr.
Spiegel, Carmen and Ulla Kleinberger Günther
2011 Höflichkeitsformen und Höflichkeitsnormen in Internetforen und E-Mails.
Der Deutschunterricht 63(2): 34–43.
Stöckl, Hartmut
2004 Die Sprache im Bild, das Bild in der Sprache. Zur Verknüpfung von Sprache
und Bild im massenmedialen Text: Konzepte, Theorien, Analysemethoden.
Berlin/New York: Walter de Gruyter.
Suler, John R.
2005 The basic psychological features of cyberspace. Elements of a cyberpsy-
chology model. http://www-usr.rider.edu/~suler/psycyber/basicfeat.html, 09
March 2016.

Authenticated
Message boards 149
Tanskanen, Sanna-Kaisa
2007 Metapragamtic utterances in computer-mediated interaction. In: Wolfram
Bublitz and Axel Hübler (eds.), Metapragmatics in Use, 87–106. Amsterdam/
Taylor, Charles
1989 Sources of the Self: The Making of Modern Identity. Cambridge, MA: Harvard
University Press.
Thurlow, Crispin, Laura Lengel, and Alice Tomic
2004 Computer-Mediated Communication: Social Interaction and the Internet. Los
Angeles: Sage.
Virtanen, Tuija
2013 Mock performatives in online discussion boards: Toward a discourse-prag-
matic model of computer-mediated communication. In: Deborah Tannen and
Anna Marie Trester (eds.), Discourse 2.0: Language and New Media, 155–
166. Washington, DC: Georgetown University Press.
Wittgenstein, Ludwig
1953 Philosophical Investigations. New York: MacMillan.
Yus, Francisco
2011 Cyberpragmatics. Internet-Mediated Communication in Context. Amsterdam/
Zywiczynski, Przemyslaw
2007 Politeness and aggression: a study of message board communication. In: Bar-
bara Lewandowska-Tomaszczyk, Tomasz Płudowski and Dolores V. Tanno
(eds.), The Media and International Communication, 361–380. Frankfurt/
Main: Lang.
Internet sources:
http://www.thestudentroom.co.uk/, 22 February 2016.
http://harmsen.net/ahrc/news.htm, 02 February 2016.
http://www.network54.com/Forum/358339/page-14, 02 February 2016.
http://www.oprah.com, 15 November 2016.

Authenticated
Authenticated
6. Blogs
Theresa Heyd
Abstract: This chapter provides an overview on blogging, both as a digital practice

and as an object of research in computer-mediated discourse analysis. It sets out
by providing a historical overview on the emergence of weblogs and their typical
forms of occurrence, topical reach, sociotechnical constraints, and communicative
characteristics. It also retraces how blogging has been understood and analyzed,
both from a scholarly perspective and in wider public understanding. The article
then moves on to identify key topics which are central for pragmatic analyses of
weblogs as part of the social media canon: performance of identity and a narrated
self; patterns of interaction and participation; the management of civility and con-
flict; finally, the pragmatic implications of monetization and commodification.
1. Blogging: traditions of practice, traditions of research
Weblogs or blogs as a genre1 and blogging as a digital linguistic practice can be

considered among the most long-standing forms of social media communication.
Indeed, their emergence can be dated to the late 1990s and early 2000s (Blood
2000) and thus precedes the Web 2.0 era and its conventionalized understanding
of the emergence of social media, as described by Hoffmann (Ch. 1, this volume).
Thus after an initial phase where blogs were largely found among users from tech
communities, the practice soon gained social traction and became available to the
popular imagination as a quintessence of ‘innovative’ digital communication. This
public visibility of the genre probably contributed to its digital “enregisterment”
(Squires 2010) as a kind of digital diary. The metaphor of the diary was fueled by
certain structural aspects of blogging such as its counter-chronological ordering,
and reverberated in the naming of early-day blogging platforms such as LiveJour-
nal. This notion of blogging as an online journal informed many early academic
explorations of the practice, and a critical engagement with folk-linguistic notions
such as “the diary on the Internet” (McNeill 2005) has constituted a substantial
part of pragmatic analyses of blogging.
Indeed, the academic exploration of blogging started relatively early and
includes not just approaches from media and technology studies, but an attention
1
This account is following approaches such as Herring et al. (2005) in categorizing blog-
ging as a digital genre.
DOI 10.1515/9783110431070-006
Authenticated
152 Theresa Heyd
to discourse-linguistic and pragmatic aspects. As early as 2004, Kumar et al. treat

blogs as “a remarkable artifact of the Web” (2004: 35) and thus as a practice that
was well entrenched in the scholarly field of inquiry by that time. In that vein, they
describe blogs as
Web pages with reverse chronological sequences of dated entries, usually with sidebars
of profile information and usually maintained and published with the help of a popular
blog authoring tool. They tend to be quirky, highly personal, typically read by repeat
visitors, and interwoven into a network of tight-knit but active communities. (Kumar
et al. 2004: 35)
In contrast to other similarly longstanding sociotechnical modes, in particular web

forums and message boards as described by Arendholz (in Ch. 5, this volume),
blogging is thus very strongly associated with the notion of social media prac-
tice: From the earliest references to Web 2.0, social networking, social media and
related popular terms for participatory digital linguistic practice, blogs have been
perceived as one of the substantial pillars of such forms of interaction and partic-
ipatory “produsage” (Bruns 2007). In this sense, blogging is a persistent commu-
nicative practice that precedes the age of social media, yet has robustly survived
the advent of later sociotechnical modes such as microblogging and audiovisual
formats. The notion of blogging as a “bridging” genre, proposed by Herring et al.
(2005), was thus prescient. In their genre analysis of blogging, the authors propose
that the blog is neither fundamentally new nor unique, but that it – along with other
emergent genres expressed through interactive web technologies – occupies a new posi-
tion in the internet genre ecology. Specifically, it forms a de facto bridge between mul-
timedia HTML documents and text-based computer-mediated communication, blurring
the traditional distinction between these two dominant internet paradigms, and poten-
tially contributing to its future breakdown. (Herring et al. 2005: 143)
The concept of blogging as a bridging, or hybrid, genre will therefore be taken as a

cue and starting point here in order to summarize some of the central insights from
research on blogs and their pragmatics within a social media context. Section 2
sets out by providing a concise overview on blogging as a social media practice,
including facts and figures about usage and distribution, as well as some major
approaches to defining blogging, and a brief review on important discourse-lin-
guistic contributions to the analysis of blogs. After that, Section 3 zooms in on a
number of specific angles for the pragmatic analysis of blogging, including perfor-
mance of identity and a narrated self; patterns of interaction conflict; and finally,
the issue of blogging and discursive authenticity, including controversial topics
such as monetization strategies of bloggers.

Authenticated
Blogs 153
2. Usage, distribution, definitions
2.1. Basic figures
As noted above, blogging has emerged as an identifiable and labeled practice since
the late 1990s (Blood 2000; see also the timeline provided in Myers 2010: 16).
In these early instantiations, blogging was understood primarily as a practice of
sharing and filtering existing material, in particular link sharing. In this scene,
blogs constituted an early form of informal and personal feeds, rather than sites
centered on the creation of content. While the strong embedding of links, and
the redistribution of material remains a central feature of later forms of blogging,
the primary purpose of link-sharing migrated on to other social media practices,
including social networking sites, folksonomies, and visually driven platforms
such as Pinterest.
The quick and pervasive evolution of blogging from this niche practice to a
digital mass phenomenon in the following years is strongly tied to sociotechni-
cal factors. Specifically, the early 2000s saw a quick increase of widely availa-
ble software and digital platforms that provided easily usable templates to enable
blogging. Such early platforms included OpenDiary (founded in 1997, defunct
as of 2014) and LiveJournal (founded in 1999); these early sites, whose names
carry overt references to writing as a journaling activity, were first and foremost
conceived as communities focused on the primary practice of life-writing. In their
wake, blog hosting services such as Blogger (launched in 1999, later acquired
by Google) and WordPress (launched in 2003) became a motor of blogging as a
widespread and everyday practice. It can be assumed that the combined socio-
technical features of such services were crucial to this development, offering eas-
ily usable templates that made blogging available to a wide audience, while also
providing an inbuilt online community through their social network aspect. This
had profound and lasting effects on blogging as a digital genre: despite the contin-
ued importance of multimodal practice and the embedding of pictures, videos and
external material (see Section 2.3 below), blogging has become a practice that is
first and foremost conceived of as a genre of text production, so that a prototypical
understanding of a blog is a collection of (self-)authored text, and blogging is seen
as a writing activity.
Since these early days, blogging has remained one of the staples of social media
practice, and has endured through the rise and spread of competing sociotechnical
modes such as microblogging and the central role of Facebook in the social media
landscape. It has been notoriously difficult to provide reliable statistics on the
number and distributions of blogs, because blogging is distributed between many
different hosting contexts, ranging from major blog hosters such as the ones men-
tioned above, to smaller sites, as well as many individual and decentralized blogs.
In this sense, blogging as a global practice is much harder to measure than social

Authenticated
154 Theresa Heyd
media genres geared toward single-site usage such as microblogging through

Twitter. With this caveat in mind, widely quoted estimates give a number of blogs
between 150 million and 200 million. For example, a statistic provided by Nielsen
(NM Incite 2011) indicates a total of 173 million blogs worldwide as of October
2011, and suggests an almost linear rise through the years from 35 million blogs
in October 2006.
As Puschmann (2013: 79) stresses, the relevance of such numbers is limited
not just because of their lack of precision, but because the sheer existence of a
blog says little about its vigour as a site of active linguistic practice. Thus many
blogs are quickly abandoned, dormant or entirely fossilized. By consequence, a
more meaningful metric may concern the number of blog posts written daily: here,
frequently quoted estimations range around 2 million blog posts per day. However,
this number may well be underreported, given that the platform WordPress alone
self-reports a figure of 65 million new posts per month on its own web-hosting
platform (https://wordpress.com/activity/posting).
Finally, such figures centered on production do not factor in the reach and
reception of blogging. From a pragmatic point of view, the consumption of blogs
by readers and the redistribution of texts through sharing practices on social media
also needs to be taken into account. Indeed, wider psychological factors have been
applied to blogging as a discursive practice. Studies such as Guadagno, Okdie and
Eno (2008) align blogger identities with personality trait paradigms such as the
Big Five and suggest “that people who are high in openness to new experience and
high in neuroticism are likely to be bloggers” (Guadagno, Okdie and Eno 2008:
1993). While such psychological categorizations are almost certainly painted with
too broad a brush, it seems crucial to keep in mind that blogging does play an
important role in identity construction and negotiation; this issue is taken up again
in Section 3.
2.2. Definitions
Attempts at defining blogs as a genre, and blogging as a digital practice, vary
according to different research perspectives, as Puschmann (2013: 77–79) has
noted. Thus blogging may be understood through its form and sociotechnical con-
straints, its genre particularities, its (pre-digital) genre antecedents and (digital)
genre neighbors, its prototypical domains of content,2 its role as a social practice
or through its socio-pragmatic effects and regularities.
2
While specific types of content are not perceived as a particularly meaningful or salient
predictor for genre status by many genre theorists, it seems likely that at least certain
topic-based subgenres of blogging (such as fashion blog, travel blog, corporate blog
etc.) become relevant in the perception of users.

Authenticated
Blogs 155
As many others, an early definition proposed by Herring et al. (2005) is strongly

focused on the technical givens of blogging, describing blogs as “frequently modi-
fied web pages in which dated entries are listed in reverse chronological sequence”
(Herring et al. 2005: 142). This minimalistic definition highlights the reverse
chronological display of blog entries that is part of the template which blogging
software provides; it likens blogging to web pages as a neighboring genre; and it
touches on the social practice of blogging, noting the frequent modification that is
part of prototypical blogs.
A slightly later definition by Schmidt (2007) starts out with a remarkably sim-
ilar approach, yet includes a much stronger focus on the social and particaptory
features afforded through blogging, including digital-performative acts such as
commenting, referencing and sharing blog posts:
Weblogs […] are frequently updated websites where content (text, pictures, sound files,
etc.) is posted on a regular basis and displayed in reverse chronological order. Readers
often have the option to comment on any individual posting, which is identified by a
unique URL. Through such comments and references to other online sources in the
postings, as well as through links to favorite blogs in the sidebar (the “blogroll”), blogs
form a clustered network of interconnected texts: the “blogosphere”. (Schmidt 2007:
1409)
As noted above, many theorists have sought to define blogging with reference to
antecedent or neighboring genres. The early, folk-linguistic notion of blogging
as a form of diary-keeping, as outlined in Section 1, can be found in many early
discourse-linguistic approaches (see e.g. Yus 2007:15, who describes blogging as
“mostly a kind of diary”). In later treatments, this homogeneous notion has been
replaced with more nuanced models that attend to the vast variation in audience
design that can be found among bloggers. Thus Puschmann (2009) proposes a dif-
ferentiation between “ego blogs” and “topic blogs”, which may vary dramatically
in audience design, reach, and a continuum of intimacy vs. openness. Other genre
analyses of blogs have likened them, amongst other forms of textual practice, from
journalism and letters to the editor, to websites (see above) and online commu-
nities/newsgroups (see Herring et al. 2005 for an overview). The genre analysis
conducted by Herring et al. (2005) ultimately situates blogs as a “bridging genre”
that is not only generally hybrid in different ways, but which specifically unites
properties of static HTML documents (such as personal homepages) and the lively
interaction and easy handling of text-based online communities. This assessment,
delivered just before the dawn of social media as the new central paradigm of
research into digital linguistic practice, still seems to captivate some of the most
relevant genre features of blogging.
Definitions have also attended to the typical domains or fields of content that
blogging has covered, a topic that is closely connected to the different types of
audience design mentioned above. In keeping with the notion of a “digital diary”,
many attempts at definition have focused on personal blogs as a seminal content

Authenticated
156 Theresa Heyd
category. Herring et al. (2005), based on structural analysis of a random sample of

publicly available blogs, found over 70 % to be such personal, or “diary” blogs. As
Scheidt (2006) highlights, these personal blogs are characterized by “their ability
to blend personal narrative with performance characteristics”. Within this personal
domain, variation in terms of topic choice and content is considerable, and the per-
sonal blogosphere is accordingly differentiated into a complex network of digital
communities of practice. These can cover topics including personal feelings and
family matters (see e.g. Scheidt 2006 on adolescent blogs; Friedman and Calixte
2009 on “mommy” blogs), sports and leisure (e.g. Cox and Blake 2011 on food
blogs), popular culture, health and sickness (e.g. Page 2008 on narratives of illness
on personal blogs), lifestyle and fashion (e.g. Rocamora 2012), career and work-
life, and many others. As Scheidt (2006) points out, it can be a feature of such per-
sonal blogs that some or all entries may not be open access but password-protected
or entirely private.
While the emphasis placed on personal blogs in existing research is justified
through the importance that the practice holds, other forms of blogging also exist.
In particular, this concerns blogs that are communal efforts, written by a team
of contributors or in the style of a (relatively) open platform. Communal blog-
ging was a focus of early research (e.g. Krishnamurthy’s 2002 analysis of content
aggregation blogs such as Metafilter); it is particularly relevant to those domains
of blogging which intersect with online journalism (Wall 2005; Gillmor 2006) and
open science on the web (Shema, Bar-Ilan and Thelwall 2012). Finally, corporate
blogs have attracted considerable research (Puschmann 2010): in this case, blog-
ging serves a company’s self-representation, and pursues organizational goals, in
contrast with the individuated domains outlined above. The communicative design
and implications of corporate blogs crucially touch on issues of monetization and
commodification of blogging as a genre, which will be taken up for further inquiry
in Section 3.
Finally, blogs can be described and defined in terms of the semiotic resources
they make use of and that make up the bulk of their material. A duality can be
noted here: on the one hand, many definitions have emphasized the almost inher-
ently multimodal and heterogeneous potential of blogging. Thus the sine qua non
condition of enriching blog posts with links to extraneous material is stressed rou-
tinely in descriptive accounts of blogging (e.g. “verbal-visual discourses”, Yus
2007: 18). Indeed, some approaches focus on subforms of blogging that are char-
acterized by the reliance on a specific audio/visual mode. This concerns in par-
ticular the realm of video blogging or “vlogging” (see studies by Frobenius, e.g.
2010; 2014), but also subforms of blogging such as photo blogs (e.g. Heyd 2014)
or generally “visual blogs” (Badger 2004). In this sense, multimodality and the
incorporation of heterogeneous sources and semiotic codes can be considered one
key feature of blogging as a genre, even more so as visually oriented blogging
platforms such as Tumblr are gaining in relevance. At the same time, it is perhaps

Authenticated
Blogs 157
not surprising that the bulk of linguistically oriented, and in particular pragmat-
ically oriented, research has been directed toward the other end of the modality
spectrum, namely toward blogging as discourse and discursive practice in a nar-
rower sense. Within this framework, blogs are primarily understood as a social
media practice that is shaped by textual entities. These textual entities – first and
foremost blog entries, but also metadata and commenting sections, sidebars and
blogrolls, and other discursive elements – are seen as interacting building blocks
which, taken together, constitute blogs as a complex text-oriented genre. As an
example of such text-centric approaches, Grieve et al. (2010: 303) formulate that
“[blog] postings tend to consist primarily of raw text, but may also contain hyper-
links and other media, including picture, video and sound files”.
As a consequence, much of the discourse-linguistic literature on blogging has
oriented toward the analysis of such text-based features by examining them with
the tools available to discourse linguistics: through concepts such as genre and
narrativity; through pragmatic accounts of deixis, speech acts and cooperation;
through analysis of interaction, intertextuality and conflict. Myers (2010: 4) epit-
omizes this tendency by noting that “despite all the possibilities open to bloggers
for inclusion of pictures, sound and video, written language remains central to
most blogs”. Discourse-centric approaches to blogging, understood in this vein,
thus form the basis for the remainder of the overview provided here.
2.3. Key texts in discourse-linguistic research on blogging

Blogging can be considered, overall, as a well-studied phenomenon within the tab-
leau of social media studies. With blogging as a digital linguistic tradition well in its
second decade, a number of book-length studies have been published. Not includ-
ing the myriad practical guides and other purely popular books (which address
topics such as how to design, improve or monetize a personal blog), scholarly pub-
lications have often taken a broad perspective that also includes accounts coming
out of the community of practice, and addressing a wide audience. This is the case
in We’ve got blog: How weblogs are changing our culture, edited by Rodzvilla
(2002), which probably constitutes the earliest book-length treatment of the sub-
ject, but also the edited volume by Friedman and Calixte (2009) on Mothering and
blogging. Rettberg’s (2014/2008) monograph Blogging gives an overview from a
broad communication and digital studies perspective, and thus contains a number
of observations that are relevant for discourse-linguistic and pragmatic research-
ers. Finally, Myers (2010) on The discourse of blogs and wikis is an explicitly dis-
course-oriented monograph which includes pragmatic and sociolinguistic analyses
of blogging; because the book is conceived primarily as a coursebook for students,
it gives a sweeping overview rather than in-depth analyses.
Apart from these broad perspective monographs, there is a growing number
of book-length studies that adopt a more specific approach to blogging. In many

Authenticated
158 Theresa Heyd
cases, these monographs are the result of PhD work and similar project-based
studies. Amongst others, some of the influential topic-based monographs here
include Bolander (2013) on Language and power in Blogs; Frobenius (2014) on
The pragmatics of monologue: interaction in video blogs; Hoffmann (2012) on
Meaning and interaction in personal weblogs; Puschmann (2010) on The corpo-
rate blog as an emerging genre of computer-mediated communication; or Scheidt
(2016) on Scholarly and producer perceptions of genre in female and male ado-
lescent weblogs.
Regarding blogging as it is represented in handbooks, the contribution by
Puschmann (2013) to the handbook of Pragmatics of computer-mediated com-
munication stands out, in that it provides the most concise yet complete overview
on blogging from a broad pragmatic perspective. Handbook entries are also found
in the volumes edited by St. Amant and Kelsey (2008) and Georgakopoulou and
Spilioti (2015).
Finally, individual articles on blogging that adopt a discourse-linguistic or
pragmatic perspective have become so numerous that they are not easily con-
densed into an overview. Many of these individual papers are referenced and/or
discussed in Sections 2 and 3 of this chapter. To single out two seminal works that
did much to set the scene for ensuing discourse-linguistic work on blogging, men-
tion must be made here of the two coordinated studies by Herring et al. (2005) on
“Weblogs as a bridging genre”, and Herring and Paolillo (2006) on “Gender and
genre variation in weblogs”. These two papers, with their respective foci on more
pragmatically oriented genre analyses and more sociolinguistically oriented vari-
ation studies, remain foundational works in assessing the language and discourse
of blogs.
3. Analyzing blogging: pragmatic and discourse-analytical approaches
The following overview presents an aggregation of research into blogging from a

(more general) discourse-linguistic and (more focused) pragmatic perspective. In
this sense, the overview begins with those contributions which examine blogging
within the paradigm of ‘core’ pragmatic reasoning, i.e. Gricean and post-Gricean
theories. This includes the exploration of blogging in the framework of speech
act theory and cooperation, relevance theory and politeness, as well as deixis.
The overview then moves on to approaches which have a wider and more general
overlap with pragmatic research, including analyses of interaction and conflict,
genre and narrativity. Finally, by considering ongoing dynamics in blogging such
as monetization and commodification, the section concludes with an account of
neighboring sociolinguistic approaches to blogging.

Authenticated
Blogs 159
3.1. Blogging and Gricean pragmatics
3.1.1. Speech Act Theory

While no fully-flegged speech act-centered account of blogging has so far been
proposed, the analysis of speech acts and linguistic performativity in general has
been incorporated into pragmatic analyses of blogging. This seems in particular
relevant for specific topic-oriented types of blogging such as journalistic and polit-
ical blogging. Thus Chiluwa (2011) analyzes the online platform Nolitics as a
Nigerian blogger community engaged in political and social argument; in this con-
text, he comes to the conclusion that
[b]ecause Nolitics involves a form of social mobilization, the post writers adopt per-
suasive strategies that tend to ‘direct’ or ‘plead with’ people to perform their civic
responsibilities […]. This form of discursive strategy is also evident in the ‘suggestion/
recommendation’ category where post writers suggest how Nigeria’s social and eco-
nomic problems might be solved. (Chiluwa 2011: 88)
In a methodical paper aimed at extracting speech acts from large blogging corpora,
Anand et al. (2011) come to a similar conclusion, noting that “there is likely to be
practical value in detecting attempts to change belief or attitude using blog posts”
(2011: 3). In other words, where speech act theory is used to assess blogging prac-
tices, the focus is likely to be on the “megaphone” character (Puschmann 2009) of
blog texts, e.g. on their capability of reaching out and activating communicative
networks.
3.1.2. Cooperation and Maxims of conversation

Gricean and post-Gricean frameworks have been used to model communicative
moves in blogging. Thus Puschmann (2009) examines bloggers’ styles accord-
ing to the paradigm of cooperation. He argues that, because of the asynchronous
nature of blogging, the users here have an option of conforming to the cooperative
principle, or of choosing “non-cooperation”:
In a blog, where interaction (via comments or trackbacks) is technically possible but
not formally required and the possibility to inform is clearly given but not mandatory,
the blogger is faced with a choice between cooperation and non-cooperation with his
[conceptualized audience]. (Puschmann 2009: 4)
In his analysis, Puschmann loosely aligns these diverging approaches to coopera-

tive behavior with ego blogging vs. topic blogging, suggesting that topic bloggers,
eager to get their message out, are more invested in establishing common grounds
with their conceptualized audience, and thus more likely to adhere to conversa-
tional maxims. Petersen (2011), exploring the notion of blogs as “conversations”,
goes even further in this vein of reasoning. Referring to Crystal (2006), he sug-

Authenticated
160 Theresa Heyd
gests that “webloggers routinely disregard Grice’s maxims concerning the quality,
relevance, quantity, and manner of conversational speech” (Petersen 2011: 14). It
should be noted, however, that Petersen does not offer any further interpretation
with regard to possible underlying reasons or patterns of interpretations for such
potential non-cooperative behavior, other than noting the spontaneous and some-
what unregulated nature of early-days weblogs. In this sense, it may be critically
questioned for both accounts sketched here whether blogs indeed routinely can
and do show such radical divergence from the cooperative principle. If we under-
stand the Gricean framework in its traditional sense, it is to be interpreted as a very
general and fundamental heuristics of all human communication, so that clear and
on-record disregard of its maxims usually constitutes the exception rather than the
norm in communicative practice.
By contrast, Myers (2010: 42–45) uses a more nuanced and specific approach
to incorporating the Gricean framework into blog analysis. Specifically, he consid-
ers links, as they are habitually embedded in blog posts, through the paradigm of
conversation maxims. As Myers rightfully points out, violations of Grice’s frame-
work are always interpreted by the reader as purpose-driven; as a consequence,
his analysis focuses on practices of flouting and implicature that are strategi-
cally achieved by bloggers. Myers then links this to aesthetic and stylistic effects
achieved through nonstandard linking practices in blogs and reaches the conclu-
sion that “(m)uch of the wit and personality of blogs comes in the flouts of the
maxims of quality, quantity, relation and manner” (Myers 2010: 43).
3.1.3. Relevance Theory

Even more so than the traditional Gricean framework, a more specifically rele-
vance-theoretical approach has informed pragmatic research into blogging. While
relevance-theoretical approaches share many of the assumptions of the Gricean
framework, and in this sense show a considerable conceptual overlap, they are
more explicitly oriented toward cognitive foundations of relevance and its impli-
cations for intentions and cognitive affordances of bloggers and their audiences.
Yus (2007), in his programmatic paper “Towards a pragmatics of weblogs”, lays
out such a cognitively oriented work program:
What can relevance-theoretic pragmatics do to explain the weblog phenomenon? Firstly,

it should be able to trace the [blogger’s] intentions when including information in the
weblog. For that purpose, the weblog itself should be used only as a blueprint for the ac-
tual intended interpretation. […] Secondly, it should be able to analyse what processing
effort is demanded and what hypothetical interest is eventually provided by the informa-
tion coded in the weblog, basically related to what will be called the “contextualisation
potential” of the weblog, that is, the balance of interest and mental effort that the specific
verbal-visual qualities of the weblog produce in the readers. […] Finally, relevance-the-
oretic pragmatics should also be able to decide whether the information coded (and the

Authenticated
Blogs 161
way this information is coded) will eventually be interesting, worth the processing effort
(in other words, relevant), basically weighing the balance of the hypothetical reward
from reading the weblog in exchange for the effort it demands. (Yus 2007: 16)
As Yus demonstrates in his ensuing essay, this approach can be helpful in mode-
ling certain communicative aspects about blogging as a communicative practice,
such as the intentionality behind blogging efforts, or the “relevance-oriented infer-
ences” (2007: 26) that blog readers make.
In a similar vein, blog discourse has been used as a test case to explore more
specific aspects of relevance and its mechanism. For example, Greenall (2009)
proposes a new conceptualization of flouting that is focused on Alfred Schutz’s
notion of communicative relevance. While Greenall’s interest is geared toward
theoretical reasoning rather than empirical analysis, his treatment nevertheless
reveals interesting insights into flouting and other pragmatic mechanisms in blog-
ging practice. He also remarks on a specific pattern of stance-taking that he iden-
tifies in many blogs, namely
how bloggers typically introduce themselves in terms of making excuses for being ex-
cessively wordy, i.e. making sure the reader knows that by entering into the activity of
blogging, the writer is opting out of the first maxim of Quantity. It may be the case that
when the activity type of blogging has become more established, the first maxim of
Quantity will disappear from the frame that constitutes this particular activity type, thus
eradicating the current felt need to opt out. (Greenall 2009: 2306)
In this sense, it can be summarized that both Gricean and more general approaches
to pragmatic principles of relevance can act as a window onto certain mechanisms
found in blogging – mechanisms that may either be very generizable, or that man-
ifest themselves in highly genre-specific discursive moves.
3.1.4. Politeness
Politeness can be seen as a further pragmatic approach that connects the Gricean
maxim of manner with an interest for facework and relational aspects of commu-
nication. As previous research has pointed out, the notion of (im-)politeness in
blogging is not an unproblematic one, given the dual involved-yet-detached com-
municative status of blogs. Puschmann (2013: 89) spells this out in detail:
While a blog can serve as an effective tool for active face work from the perspective of
the blogger, the capacity for [face-threatening acts] is somewhat constrained by the fact
that blogger and audience are relatively removed from each other. Inside the space of
her own post, the blogger is in charge – she cannot be interrupted, pressured, threatened
or interrogated except via comments, over which she has a high degree of control. […]
Thus face threat understood in a way that presumes parties who are aware of each oth-
er’s presence does not really apply to blog posts, which may serve to explain their pop-
ularity as a personal space that is both visible to others and at the same time protected
from undesired intervention.

Authenticated
162 Theresa Heyd
Despite this notion of blogs as a “safe space” that affords the blogger with con-
siderable liberties in terms of his or her identity work, analyses have found that
politeness can be a significant factor in understanding the pragmatic dynamics of
blogging. However, in accordance with the above statement, such issues of face-
work are often deferred to marginal or residual discourse features of blogging,
rather than the nucleus of text-based blog posts. For example, Kavanagh (2010;
2016) provides a comparative analysis of facework done through emoticons in
American and Japanese blogs, and Luzón (2011) analyzes social presence and
antisocial behavior in academic weblogs. What unites both approaches is that they
focus on the comments sections of the analyzed blogs, rather than the original con-
tributions. In other words, facework and face threats become much more salient in
those discursive elements of blogging which are explicitly framed as participatory,
and where interaction and conflict between a blogger and his or her audience is
possible and, sometimes, even encouraged. The same principle holds for works
such as Kouper’s (2010) analyses of advice-seeking on LiveJournal: here, the rela-
tional work done by bloggers is not so much concentrated on the blogging sections
of the platform, but rather on its dual functionality as an online community and
network.
In this sense, we can conclude that politeness research can make an impor-
tant, and indeed highly relevant, pragmatic contribution to the understanding of
blogging (cf. also Graham, Ch. 17, this volume). However, this is crucially tied
to aspects of participation and interaction, and thus pertains highly to the social
media nature of blogging.
3.1.5. Deixis
Finally, deictic elements and strategies of discourse are often considered as part
of the core concerns of pragmatic research. Unsurprisingly, the issue of deixis
and deictic anchoring of blogs has inspired pragmatic and discourse-linguistic
research. This is in part due to an established notion that digital communication
in general is ‘displaced’, detached and separate from the physical world, and that
it takes place in a more or less abstract ‘cyberspace’ that bears little or no con-
nection to our physically grounded and anchored everyday lives (see e.g. Heyd
and Honkanen 2015 for a discussion of this research tradition). While this view
from early digital research has given way to a more convergent approach through
the rise of mobile computing, the notion of computer-mediated communication
(CMC) as “placeless” still informs many texts. Myers’ analysis of deixis in blog-
ging thus begins by positing that
[b]logs […] are placeless by default; they have to do something to signal place or we
don’t think about it. […] When I say blogs are placeless, I mean it in two senses: (1) the
writing is not situated in a particular location and (2) they have the global vision of the
blogosphere. (Myers 2010: 48–49)

Authenticated
Blogs 163
Despite such conceptualisations of blogging, discourse-analytical readings of

blogging do find that deictic anchoring plays a crucial role both in the discursive
design of blog postings, and in the way that readers perceive and process them.
As Puschmann (2009: 2) points out, the very “physics of blogging” prompts the
inscription of deictic information into the discursive fabric of the texts. Because
typical blogging software packages add a timestamp and concomitant metadata to
any blog posting, the deictic origo of a given blog text is not only available, but
made explicit through the very act of posting. As Puschmann expounds:
Author, time of publication and the location of the post differ from the other fields [such
as title and text] by constituting extra-textual […] information that is automatically
associated with the situation and not freely assigned by the blogger. This information
makes deictic language possible, i.e. use of the first person pronoun (always referring
to the blogger-publisher credited with the post), use of temporal adverbs (relating to
post publication date as the point of reference) and spatial adverbs (either relating to the
blogger’s location at the time of publishing or conceptualizing the blog or the Internet
as a whole as a space). (Puschmann 2009: 2)
The idea that anchoring through metadata “makes deictic language possible” has
been explored in qualitative studies. Thus Myers (2010), after the initial cave-
ats delineated above, provides numerous examples of very explicit and highly
engaged place-making in blogs that takes place through features of place, but also
person and temporal deixis. Beyond this deictic anchoring of blog posts to the
material world and its place, time and person constraints, it may also be noted
that blog features such as links and blogrolls are analyzable as a dense network of
discourse-deictic pointers.
Finally, multimodal forms of blogging have other and more complex deic-
tic exigencies. While the focus of this contribution is on text-centered blogging,
research into more audiovisual forms has reavealed intricate deictic strategies.
Thus, Frobenius (2013) provides a detailed account of gestures and the deictic
implications of pointing in video blogs; as her study points out, such deictic fea-
tures make even more tangible the physically situated nature of blogging, because
the deictic space here can and does include pointing to real-world objects and
frames of reference.
3.2. Blogging and discourse-linguistic analysis
3.2.1. Blogging as interaction

Within more broadly discourse-linguistic frameworks, the general notion of
interaction stands out. It is noticeable that, despite the folk-linguistic conception
of blogging as journal-writing or ego-blogging, many studies have nonetheless
focused on the interactional pragmatic strategies that blogging affords. In this

Authenticated
164 Theresa Heyd
regard, the notion of blogging as interpersonal communication, as a means of

reaching out and engaging with others, takes center stage. Some of the studies in
this domain intersect with approaches already described, e.g. the facework that
ensues in acts of advice-giving (see Section 3.1.4).
Within this paradigm, there are numerous approaches, some of which come to
differing conclusions. For example, Petersen (2011) applies a Jakobsonian analy-
sis of communicative functions to the entire genre of blogging, in order to explore
the popular notion that blogs are conceptually close to conversation. He concludes,
however, that while blogging may have a certain conversational potential, this is
by no means pervasive or constitutive. Thus Petersen points out that
[i]n weblogs, addressees determine the extent of their participation as readers in what
can range from close and regular reading to distracted and intermittent scanning. […]
Readers and writers maintain this non-linear, discontinuous, and asymmetrical contact
in weblogs because they find such interaction meaningful. (Petersen 2011: 9)
Nevertheless, Petersen does conclude that blog authors and readers engage and
interact in manifold ways. In an empirical study, Stefanone and Jang (2008) focus
on the notion of using blogs for relationship maintenance. Based on survey data by
a sample of bloggers, they show that this interpersonal and interactional purpose
of blogging is particularly strong where the blog authors display personality traits
prone to extroversion and self-disclosure, and have social networks with strong
ties. Stefanone and Jang conclude that “blogs have been adopted as a mode of
communication for strong tie network contacts, similar to email” (2008: 135).
Conceiving of blogging as interactional and interpersonal communication,
however, should not automatically be equated with constructive and civil dis-
course. In this sense, reseach on interactional pragmatics in blogging has also
focused on issues of conflict and disruption that can accompany blog discourse in
interpersonal communication. In particular, the studies by Bolander on language
and power in blogs (2013) and forms of responsiveness in disagreements (2012)
highlight this issue. As noted in the earlier discussion on politeness, research on
conflict in blogs gravitates toward those discourse features which are naturally
prone to interactional episodes, namely the comments sections (cf. also Bolander
and Locher, Ch. 22, this volume).
3.2.2. Blogging as identity work

If blogging can be conceptualized through the lens of interpersonal work, as out-
lined in Section 3.2.1 above, the inverse perspective is also possible and has gen-
erated a body of discourse-linguistic research: namely the notion that entertaining
and writing a blog is crucially tied to matters of identity work (cf. also Locher
and Bolander, Ch. 15, this volume). In other words, the linguistic output that a
blog constitutes can be seen as a way of establishing and performing a voice, and

Authenticated
Blogs 165
thereby doing identity work, by a blogger. Indeed, this discursive aspect of blog-
ging appears to be so central that it is more or less tacitly assumed in many analyt-
ical approaches and has found its way into many models and general overviews of
blogs as a genre (see e.g. Puschmann’s notion of “ego blogging”, Section 2 above).
Those accounts of blogging which explicitly focus on the concept of (discur-
sive) identity do so via a range of different approaches. Thus, Siles (2012), in
a qualitative study based on semi-structured interviews with bloggers, explores
identity formation in the emergence period of blogging in the late 1990s. Within
this media-historical approach, he distinguishes clearly between “online diaries”
and “blogging” as two historically distinct forms of identity constitution. Never-
theless, in his explicitly Foucauldian framework, he describes both online diaries
and blogs as “web technologies of the self” (Siles 2012).
By contrast, Schmidt’s (2007) “analytical framework” of blogging takes a
strikingly different approach to identity in blogging. In this account inspired by
sociological structuration theory, Schmidt (2007) identifies a set of rules, relations
and codes that govern the process of blogging. In this framework, the notion of
blogger identity plays a central role in the context of “publication rules” (Schmidt
2007: 1412–1413). Schmidt points to “general strategies of identity management
in a mediated environment” while pointing out that “publishing a blog is a way of
self-presentation that has to meet certain expectations about personal authenticity
while maintaining a balance between staying private and being public” (2007:
1413). In this sense, identity work is described not so much as a basic commu-
nicative need, but rather as strategic option for optimizing a blog’s attainment of
certain felicity conditions (and, by extension, success on the blogging market).
Finally, there is a range of discourse-linguistic studies which focus on more or
less specific identities that can be performed through blogging. Researchers have
studied identity work performed through blogging by teenagers (e.g. Scheidt 2006;
Huffaker and Calvert 2005) and mothers (Friedman and Calixte 2009), politicians
(Nilsson 2012) and academics (Kirkup 2010), and others, thus spanning both ‘pro-
fessional’ as well as ‘private’ social roles.
3.2.3. Blogging and narrativity

As many digital genres have become the focus of narrative analysis, it has been
noted that story-telling as an anthropological universal resurfaces strongly in online
environments (see. e.g. Eisenlauer and Hoffmann 2010; Nünning et al. 2012; Day-
ter and Mühleisen 2016 for overviews). In this vein, it is not surprising that blogs
have been subjected to narrative analysis (indeed from early on – see e.g. Huffaker
2004), but in the case of blogging as a digital genre, this connection may be par-
ticularly obvious. As Rettberg remarks, there is (at least) a twofold connection
between blogs and storytelling, “both in terms of narrative structure and in terms
of the uneasy relationships between fiction, self-representation and authenticity”

Authenticated
166 Theresa Heyd
(Rettberg 2014: 3). Thus in structural terms, the sequential nature of postings that
preimposes a certain timeline to the discursive structure can be said to work in
conjunction with the text-centric nature of conventional blogging. Put simply, a
genre that is organized around large chunks of text with temporal sequencing can
be said to be inherently prone to narrativity, much more so than genres into which
narrativity is inserted in spite of their sociotechnical constraints (such as Twitter
or Instant Messaging). Rettberg’s second point sketches aspects of self-representa-
tion through narrative on blogs. Here, discourse-linguistic examination of blogs
may be focused on the concept of particularly authentic and unmediated discourse;
for example, Burgess (2006: 208) emphasizes the importance of vernacular voices
and “the decentralised, accretive and networked, but equally ‘ordinary’ kinds of
storytelling made possible by personal weblogs”. At the same time, this conceptual
feature also raises issues with regards to veridicality and fabrication in the kind of
narrativity found in blogs.
The most comprehensive account of narrativity in blogs is put forward by
Eisenlauer and Hoffmann (2010). Echoing the perspective sketched above, they
argue that “[in] weblogs, narration seems to be particularly crucial, simply because
‘bloggers’ […] usually draw on short stories to report the latest news to an (often
undefined) virtual audience” (Eisenlauer and Hoffmann 2010: 79). Based on the
four dimensions of multi-linearity, fragmentation, interaction and multimodality
as textual “principles of new media” (2010: 87), they explore narrative episodes
in blogging both through a Labovian structuralist framework and more recent con-
structivist approaches.
3.2.4. Blogging and authenticity

Section 3.2.3 above makes mention of authenticity as a category that is perceived
to be important for blogging. As a final analytical focus, notions of authenticity
and inauthenticity have played a role in analyzing the discourse of blogs, and it
can be hypothesized that this topic may gain in relevance in years to come. In the
section on narrativity, authenticity in blogging was conceptualized as a putative
quality of a blogger’s “voice”. This relates, in a pragmatic framework, to issues of
veridicality and trustworthiness, and encapsulates the assumption that producing
a blog is based on sincere, transparent and ultimately cooperative illocutionary
acts. In a more sociolinguistic framework, authenticity might be understood as an
idealized stance that readers expect bloggers to take: a stance that is vernacular
and ‘real’, and, by extension, grounded and legitimized within a community. In
a larger sense, such assumptions have often been tied to more general hopes and
tropes associated with social media, as well as their participatory and democratiz-
ing effect. In particular in early days, political blogs and early journalism through
blogging were sometimes framed as counter-hegemonic discourses that might take
an opposing stance toward existing print formats. As blogs have become absorbed

Authenticated
Blogs 167
into the converged field of media consumption, this perspective becomes less rel-
evant.
One enduring topic related to authenticity and inauthenticity in blogging is
that of commodification processes, specifically of the monetization options that
blogs afford. As Rettberg (2014: 135) remarks, “[i]f you search a bookstore for
books on blogging, you’ll quickly find that most of them are about how to make
money from blogging”. The underlying idea that blogs can become a marketable
and sellable good, and that the voices and stances of a blogger persona can ulti-
mately be worked into a branded image, are of central concern for many blogging
communities. These insights are not recent: as early as 2005, Marwick discussed
the implications of “online identity in the age of a commodified Internet”, with
explicit reference to blogging (Marwick 2005). Analyses of corporate blogging
have particularly highlighted the institutionalized dimension of such commodi-
fication strategies (see analyses by Puschmann, in particular Puschmann 2010).
However, it is in the domain of personal blogging that the discrepancy between
authenticity expectations associated with the genre on the one hand, and mon-
etization strategies which commodify it for strategic purposes, are perceived as
the most striking form of instrumentalization. In returning to the starting point of
this overview, however, it may be noted that a certain level of commodification
and artificiality has been presented almost from the earliest days of blogging. As
Marwick (2005: 163) points out, “many of the structures that we think of as ‘inde-
pendent’, such as blogging software, are actually owned by larger corporations”.
In this sense, blogs will remain a substantial and in some ways typical genre of the
social media tableau: participatory yet motivated and shaped by the stances of the
blogger; creative and malleable yet firmly couched in the socio-technical givens
of the respective software template; perceived as highly individuated yet part of
small and large communities of practice.
4. Outlook
This chapter has given an overview on blogging as part of the social media canon.
Moving from a general perspective on traditions of practice and traditions of
research to more technical facts and figures about usage, demographics and defi-
nitions, the chapter has attempted to capture the discourse-linguistic essence of
blogging. Based on this, Section 3 has highlighted some of the key research tradi-
tions that have emerged from the scholarly community with a focus on pragmatic
and discourse-linguistic interests.
As with all writing on topics of digital communication, overviews on spe-
cific genres or practices afford no more than a snapshot of the status quo at a
given moment. Yet whatever direction blogging will take in the coming years,
this discursive practice has proven robust over almost two decades, and has not

Authenticated
168 Theresa Heyd
been vitally threatened by other socio-techncial modes such as social networks,

microblogging or podcasting. It seems likely, then, that blogging will remain a sta-
ple of the social media spectrum; its future evolution in terms of structure, spread
and reach will provide new research sites for pragmatics and discourse linguistics.
References
Anand, Pranav, Joseph King, Jordan Boyd-Graber, Earl Wagner, Craig Martell, Douglas
W. Oard and Philip Resnik
2011 Believe me: We can do this! Annotating persuasive acts in blog text. The AAAI
2011 Workshop on Computational Models of Natural Argument. 1–5. http://
www.umiacs.umd.edu/~jbg/docs/persuasion.pdf.
Badger, Meredith
2004 Visual blogs. University of Minnesota. Retrieved from the University of Min-
nesota Digital Conservancy, http://hdl.handle.net/11299/172839.
Blood, Rebecca
2000 Weblogs: A history and perspective. Rebecca’s Pocket. Retrieved from http://
www.rebeccablood.net/essays/weblog_history.html.
Bolander, Brook
Bolander, Brook
2013 Language and Power in Blogs: Interaction, Disagreements and Agreements.
Bruns, Axel
2007 Produsage: Toward a Broader Framework of User-Led Content Creation. Kel-
vin Grove, Queensland: Queensland University of Technology.
Burgess, Jean
2006 Hearing ordinary voices: Cultural studies, vernacular creativity and digital sto-
rytelling. Continuum: Journal of Media and Cultural Studies 20(2): 201–214.
Chiluwa, Innocent
2011 On political participation: Discursive pragmatics and social interaction in pol-
itics. Studies in Literature and Language 2(2): 80–92.
Cox, Andrew and Megan Blake
2011 Information and food blogging as serious leisure. Aslib Proceedings 63: 204 –
220.
Crystal, David
Dayter, Daria and Susanne Mühleisen
2016 Personal Narrative Online. Topical Issue of Open Linguistics. https://www.
degruyter.com/page/1252.
Eisenlauer, Volker and Christian Hoffmann
2010 Once upon a blog. Storytelling in weblogs. In: Christian Hoffmann (ed.), Nar-
rative Revisited. Telling a Story in the Age of New Media, 79–108. Amsterdam/

Authenticated
Blogs 169
Friedman, May and Shana Calixte (eds.)

2009 Mothering and Blogging: The Radical Art of the MommyBlog. Toronto: Dem-
eter Press.
2013 Pointing gestures in video blogs. Text and Talk 33(1): 1–24.
2014 The Pragmatics of Monologue: Interaction in Video Blogs. Dissertation; Saar-
brücken: Universität des Saarlandes. http://scidok.sulb.uni-saarland.de/front-
door.php?source_opus=5895.
Georgakopoulou, Alexandra and Tereza Spilioti (eds.)
2015 The Routledge Handbook of Language and Digital Communication. London:
Routledge.
Gillmor, Dan
2006 We the Media: Grassroots Journalism By the People, For the People. Sebas-
topol: O’Reilly Media.
Greenall, Annjo
2009 Towards a new theory of flouting. Journal of Pragmatics 41(11): 2295–2311.
Grieve, Jack, Douglas Biber, Eric Friginal and Tanya Nakrasova
2010 Variation among blogs: A multi-dimensional analysis. In: Alexander Mehler,
Serge Sharoff and Marina Santini (eds), Genres on the Web. Computational
Models and Empirical Studies, 303–322. Berlin: Springer.
Guadagno, Rosanna, Bradley Okdie and Cassie Eno
2008 Who blogs? Pesonality predictors of blogging. Computers in Human Behavior
24(5): 1993–2004.
Herring, Susan C., Lois A. Scheidt, Sabrina Bonus and Elijah L Wright
2005 Weblogs as a bridging genre. Information, Technology and People 18(2): 142–
171.
Herring, Susan C. and John Paolillo
2006 Gender and genre variation in weblogs. Journal of Sociolinguistics 10(4):
439–459.
Heyd, Theresa and Mirka Honkanen
2015 From Naija to Chitown: The New African Diaspora and digital representations
of place. Discourse, Context and Media 9: 14–23.
Heyd, Theresa
2014 Folk-linguistic landscapes. The visual semiotics of digital enregisterment.
Language in Society 43(5): 489–514.
Hoffmann, Christian
Huffaker, David and Sandra Calvert
2005 Gender, identity, and language use in teenage blogs. Journal of Computer-Me-
diated Communication 10 (no pagination).
Huffaker, David
2004 Spinning yarns around the digital fire: Storytelling and dialogue among youth
on the Internet. First Monday 9(1) (no pagination).

Authenticated
170 Theresa Heyd
Kavanagh, Barry
2016 Emoticons as a medium for channeling politeness within American and Japa-
nese online blogging communities. Language and Communication 48: 53–65.
Kavanagh, Barry
2010 A cross-cultural analysis of Japanese and English non-verbal online communi-
cation: The use of emoticons in weblogs. Inter-Cultural Communication Stud-
ies 19: 65–80.
Kirkup, Gill
2010 Academic blogging: Academic practice and academic identity. London Review
of Education 8(1): 75–84.
Kouper, Inna
2010 The pragmatics of peer advice in a LiveJournal community. language@inter-
net 7: 1–21.
Krishnamurthy, Sandeep
2002 The multidimensionality of blog conversations: The virtual enactment of Sep-
tember 11. Internet Research 3 (no pagination).
Kumar, Ravi, Jasmine Novak, Prabhakar Raghavan and Andrew Tomkins
2004 Structure and evolution of blogsphere. Communications of the ACM 47(12):
35–39.
Luzón, María José
2011 ‘Interesting post, but I disagree’: Social presence and antisocial behaviour in
academic weblogs. Applied Linguistics 32(5): 517–540.
Marwick, Alice
2005 Selling Your Self: Online Identity In the Age of a Commodified Internet. MA
thesis. University of Washington. http://www.academia.edu/421101/Selling_
Your_Self_Online_Identity_In_the_Age_of_a_Commodified_Internet.
McNeill, Laurie
2005 Genre under construction: The diary on the Internet. language@internet 2:
1–11.
Myers, Greg
2010 Discourse of Blogs and Wikis. London: Continuum.
Nielsen (NM Incite). n.d. Number of blogs worldwide from 2006 to 2011 (in millions). In:
Statista – The Statistics Portal. Retrieved November 18, 2016 from https://
www.statista.com/statistics/278527/number-of-blogs-worldwide/.
Nilsson, Bo
2012 Politicians’ blogs: Strategic self-presentations and identities. Identity 12(3):
247–265.
Nünning, Ansgar, Jan Rupp, Rebecca Hagelmoser and Jonas Ivo Meyer (eds.)
2012 Narrative Genres im Internet. Theoretische Bezugsrahmen, Mediengattungs
typologie und Funktionen. Trier: WVT.
Page, Ruth
2008 Gender and genre revisited: Narratives of illness on personal blogs. Genre 41:
149–175.
Petersen, Eric
2011 How conversational are weblogs? language@internet 8: 1–18.
2009 Diary or megaphone? The pragmatic mode of weblogs. http://www.cbpusch
mann.net/pubs/diarymegaphone.pdf.

Authenticated
Blogs 171
2010 The Corporate Blog as an Emerging Genre of Computer-Mediated Communi-
cation: Features, Constraints, Discourse Situation. Göttingen: Universitäts
verlag Göttingen.
2013 Blogging. In: Susan Herring, Dieter Stein and Tuija Virtanen (eds.), Prag-
matics of Computer-Mediated Communication. Handbooks of Pragmatics 9,
Rettberg, Jill Walker
2014 Blogging. 2nd. ed. New York: Wiley.
Rocamora, Agnès
2012 Hypertextuality and remediation in the fashion media: The case of fashion
blogs. Journalism Practice 6(1): 92–106.
Rodzvilla, John (ed.)
2002 We’ve Got Blog: How Weblogs Are Changing Our Culture. London: Basic
Books.
Scheidt, Lois Ann
2006 Adolescent diary weblogs and the unseen audience. In: David Buckingham
and Rebekah Willett (eds.), Digital Generations: Children, Young people, and
New Media, 193–210. Mahwah, NJ: Lawrence Erlbaum.
Scheidt, Lois Ann
2016 Scholarly and Producer Perceptions of Genre in Female and Male Adolescent
Weblogs. Ann Arbor: ProQuest Dissertations Publishing.
Schmidt, Jan
2007 Blogging practices: An analytical framework. Journal of Computer-Mediated
Communication 12: 1409–1427.
Shema, Hadas, Judit Bar-Ilan and Mike Thelwall
2012 Research blogs and the discussion of scholarly information. PLoS ONE 7(5):
e35869. doi: 10.1371/journal.pone.0035869.
Siles, Ignacio
2012 Web technologies of the self: The arising of the ‘blogger’ identity. Journal of
Computer-Mediated Communication 17: 408–421.
Squires, Lauren
2010 Enregistering internet language. Language in Society 39(4): 457–492.
St.Amant, Kirk and Sigrid Kelsey (eds.)
2008 Computer-Mediated Communication Across Cultures: International Interac-
tions in Online Environments. Hershey: Information Science Reference.
Stefanone, Michael and Chyng-Yang Jang
2007 Writing for friends and family: The interpersonal nature of blogs. Journal of
Wall, Melissa
2005 ‘Blogs of war’: Weblogs as news. Journalism 6(2): 153–172.
Yus, Francisco
2007 Towards a pragmatics of weblogs. Quaderns de Filologia. Estudis Lingüístics
12: 15–33.

Authenticated
Authenticated
7. YouTube
Marjut Johansson
Abstract: In this chapter I will first give an overview of the timeline of YouTube
and its predecessors in the film and media history, of popular phenomena on You-
Tube and of its technological affordances in order to pave the way to the pragmat-
ics of YouTube. I will discuss pragmatic studies of YouTube videos by focusing on
their categorization as communicative genres, their relation to micro-celebrity and
fame, and their function as a representation of the cultural ‘other’. Finally, I will
consider YouTube videos from the perspectives of networked audiences, participa-
tion frameworks, comment discussions, conflicts and commentary videos.
1. Introduction
YouTube is one of the most popular social media platforms in contemporary dig-
ital culture. Although it is usually associated with user-generated content, it also
carries professionally produced content such as music, news, and commercial
videos. Its vast and ever-growing collection provides a rich source of material for
linguists, as many of the videos portray ordinary communication situations. These
can be studied from a great variety of perspectives, from spoken language and
interaction to multilingual uses and language variation in everyday life. YouTube
offers instant access to new phenomena in popular culture, in professional media,
and user-generated content. Many of these phenomena, such as bullying and hate
speech, are important research topics in pragmatics.
Many social media platforms have been studied in pragmatics and related
fields, but YouTube videos, user discussions, and debates in video comment sec-
tions have been rather rare objects of study. Several studies have focused on the
written forms and text-based interaction of social media, such as blogs, chat rooms,
Twitter, wikis and Facebook (Giltrow and Stein 2009a and 2009b; Myers 2010;
Yus 2011; Page 2012; Eisenlauer 2013; Marcoccia 2016; cf. also Arendholz, Ch. 5,
Heyd, Ch. 6, Zappavigna, Ch. 8 and Eisenlauer, Ch. 9, this volume). Online spo-
ken interaction has been considered, for instance, in research on videoconferenc-
ing and Skype calls (Develotte, Kern and Lamy 2011; Jenks and Firth 2013). But
YouTube videos and videos in general have received little attention and are mostly
only mentioned in passing in review articles or in articles intended to describe the
field of social media research in pragmatics (Jucker and Dürscheid 2012; Her-
ring, Stein and Virtanen 2013a). There are, however, a couple of handbook arti-
cles which consider aspects such as YouTube participation framework, discourse
DOI 10.1515/9783110431070-007
Authenticated
174 Marjut Johansson
structure, multimodal recontextualization practices (Androutsopoulos and Tereick

2015) and identity performance (Leppänen et al. 2014). Additional studies in vari-
ous other domains, such as digital literacy and video games, multimodality, visual
analysis, and analysis of news videos (Gee 2014; O’Halloran 2004; Kress and
van Leeuwen 2006; Jewitt 2009; Kress 2010; Johansson 2012), have also paved
the way for a growing interest in research on videos. From a pragmatic, discourse
analytic or sociolinguistic perspective, researchers use approaches such as com-
puter-mediated communication (CMC), multimodal discourse analysis (MMDA)
and digital discourse analysis (DDA) to explore interactivity and multimodality
as characteristic features of YouTube, videos as communicative genres and texts,
online identities, the use of different languages and the activity of commenting. In
the following sections, I will furthermore consider studies that describe theoretical
notions such as communicative settings, genre, representations, and ideological
discourse.
Digital culture has transformed our everyday lives and continues to do so (Gere
2002). Media convergence (Jenkins 2006) has affected communication practices
in such a way that they are intertwined with technological practices (Cardoso
2008; Lievrouw 2009; Thurlow and Mroczek 2011). The logic of social media
has been widely examined in media and digital culture studies, particularly as
regards its ability to change contemporary communication and social interaction
(Cardoso 2008; Lievrouw 2009; Livingstone 2009; Kaplan and Haenlein 2010;
Couldry 2012; Couldry and Hepp 2013; Dijck and Poell 2013). One could consider
YouTube – as well as many other forms of social media and other digital forms
of communication – as a complex, hybrid combination of mass communication
on the one hand, and interpersonal and personal social interaction on the other
(Helasvuo, Johansson and Tanskanen 2014).
This chapter focuses mainly on presenting communication, interaction and dis-
course on YouTube using a pragmatic approach which comprises three different
yet interrelated perspectives.
The first perspective is that of sociocultural context. For pragmatic studies, it
is important to understand not only the history and publishing logic of this social
media platform, but also the most common trends in YouTube’s popular content
world-wide. These factors affect the video productions of both professional and
amateur users. The latter reflect currently popular material and adapt, imitate and
comment on what they have recently watched (cf. Section 2).
The second perspective concentrates on videos as communicative genres. Prag-
matic research has not fully addressed YouTube videos from this viewpoint, even
though genre is one of the central notions in pragmatic studies on CMC (for exam-
ple, Herring 2007; Giltrow and Stein 2009a; Herring, Stein and Virtanen 2013b).
This perspective helps to understand video content because video genres may be
familiar, emergent or reconfigured (Herring 2013) when they appear on YouTube.
When videos are studied, their interaction, storytelling and multimodal features

Authenticated
YouTube 175
are as important as the way in which they are composed. I will also pay attention to
online identities, self-representations and discourses as well as to micro-celebrities
and famous YouTubers (Section 3).
The third perspective centres on YouTube’s user comments. YouTube com-
menting can create networks of audiences and different kinds of communities,
depending on the topics covered by the respective videos as well as on what users
as members of various communities actually do with their comments. I will dis-
cuss commenting as sequential and investigate the ways in which discussions turn
into debates (Section 4).
2. YouTube: from “broadcast yourself” to entertaining yourself
2.1. Timeline
The timeline is well-known: The three big social media platforms were founded
over a short period, with YouTube starting in 2005 and Facebook and Twitter in
2006 (cf. boyd and Ellison 2007: 212). Since YouTube, which started as an alterna-
tive to television, was purchased by Google in 2006, it has grown to be part of the
global media entertainment industry (Dijck 2013: 127–130). YouTube’s novelty,
its unique selling point, was user-generated multimodal content. New and easy
access to the Internet enabled individuals and groups to upload content in private
spaces, and online video-sharing spread very fast (Dijck 2013: 110–113). This was
displayed in the company’s early logo: Broadcast yourself (2006–2012). Today,
the logo simply contains the company name YouTube.
Almost from the start, however, users did not confine themselves to their own
productions. Very early on, they copied material from television and were accused
of uploading content originating from media companies, which led to ethical and
legal controversies about this type of practice (Burgess and Green 2009: 31–32;
Dijck 2013: 118; Suominen et al. 2013: 89). Apparently unable to prevent such
piracy, media companies and television channels simply adopted YouTube as one
of their means of distribution and promotion of commercial material (Dijck 2013:
127; Suominen et al. 2013). In a similar vein, other institutional actors adopted
YouTube and other social media sites to promote their products, whether polit-
ical, religious, educational or commercial (Androutsopoulos and Tereick 2015:
342).
The main YouTube categories are related to entertainment and news: Music,
Sports, Gaming, Movies, News, Live, and the latest addition, 360° Video (launched
in 2015) are all offered on the YouTube home page. More targeted sections are
Home, Trending and History, in which the content is proposed to users (YouTube
Statistics). The technology is “steered by search engines and ranking algorithms”
that filter and promote content based on its popularity on YouTube, as well as on

Authenticated
the user’s viewing experience (Dijck 2013: 113). Users may also refer to their His-
tory, which lists the videos that they have watched previously as long as they were
signed in at the time of watching.
YouTube user activities and user-generated content are part of produsage,
a term Bruns (2008) proposed to refer to collaborative work and the sharing of
knowledge through social media sites. Even though YouTube is part of its users’
daily life all over the world, most users are no more than consumers of the video
content. YouTube has therefore launched new services targeted at these users. You-
Tube Red, a subscription-based service in the United States of America, started in
2014 to allow users to watch videos, including reality TV and YouTube celebrities,
without advertisements (YouTube Red). Other services, such as YouTube Gaming
and YouTube Kids, which began to take effect in 2015, allow parents to control
and restrict content.
The huge popularity of YouTube is overwhelming, with more than one billion
users. Most are in the 18–34 age group but many are older. YouTube is present in
88 countries in 76 languages (YouTube Statistics). Its popularity is based on its
flexibility and the many videos of diverse content created by ordinary users for
other users to share and follow. Its professionally-produced content makes it even
more popular.
In the next section, I will explore the history and predecessors of YouTube. To
understand why the main features of YouTube are popular today, it is necessary to
understand the timeline because much of the current content is deeply rooted in
film and television history. In other words, video genres have migrated to social
media from other media spaces (cf. Giltrow and Stein 2009b: 9).
2.2. YouTube predecessors, professional and user-generated content

YouTube is an online video-sharing platform that contains all types of video for-
mats ranging from clips, trailers, and mash-ups to live feeds, entire films, and
many others. Short videos are not novel, but have had several predecessors (see
Burgess and Green 2009). Short forms that portray everyday life and funny inci-
dents have been popular in TV and film since the dawn of the moving picture.
When Louis and Auguste Lumière first showed a train arriving at a station and
workers leaving a factory (Lumière 1895a, 1895b), the audience was shocked,
because moving images resembling real life appeared for the first time on the
screen. The Lumière brothers also filmed the first pranks, such as The Sprinkler
Sprinkled (1896). These resemble scenes of everyday life and give the impression
of not being staged productions.
Several different kinds of developments paved the way for YouTube videos.
The first is linked to humor, which has a long media history. For example, Can-
did Camera has been on and off the air in the United States since the early years
of television. Humorous videos moved online and became popular on YouTube:

Authenticated
YouTube 177
babies, animals, weddings, pranks and fails, i.e., videos of somebody failing to do
something and creating a humorous situation.
The second development arose from the first. The TV format of compilations
of home movies and videos, although extant from an earlier date, became hugely
popular in the United States in the 1990s. These shows opened the door to ordinary
people, as their privately-shot media portrayals could now be viewed by a public
audience. The success of TV shows based on ‘ordinariness’ and the presence of
ordinary people in reality TV programs, confessional talk shows, and so on, has
produced the shift that Turner (2010: 12–13) calls the “demotic turn”. In addition,
the celebrity culture of ordinary people in television talent and reality TV pro-
grams such as American Idol gave rise to the social media celebrity culture that
has created competitions of its own (Dijck 2013: 117).
The third development, which created one of YouTube’s most popular features,
came from professionally produced music videos, which also have a long history.
They had their roots in early cinema musicals as well as in the pop music that
emerged in the 1960s and was in need of promotion. Filming music scenes and
promoting pop singles were the precursors of television, especially the video chan-
nel MTV. The first music video was Video Killed the Radio Star by The Buggles
(1979), followed by several videos that were firsts in their genre, both musically
and visually: Ashes to Ashes by David Bowie (1980), Thriller by Michael Jack-
son (1982), and Sledgehammer by Peter Gabriel (1986). Music videos opened up
different ways to combine performance and storytelling with new types of visual
expression. Their addition to visual culture paved the way for the ordinary user.
These developments have underpinned the numerous ways in which the ordinary
and professional users construct their video blogs and other types of videos, such
as parodies and imitations.
The next section describes the first videos and discusses the most popular vid-
eos, as well as the popularity ranking that is at the core of YouTube. Knowledge
of these factors will help us to understand audience choices and the way in which
YouTube videos and channels develop.
2.3. Popular videos and popularity ranking on YouTube

The first video ever uploaded on YouTube was Me at the Zoo (April 23, 2005). In
its 18-second run time, a young man stands in front of the fence at the elephant
exhibit and says that the cool thing about these animals is their very long trunks.
This very short clip shows one of the reasons that users upload videos: to tell about
things and events in their life. In pragmatic terms, this is a mediated digital inter-
action. The young man faces the camera directly, turning his head from the camera
to the elephant as he speaks. He creates in these couple of utterances a fragment of
a narrative in which he situates himself in a place (a zoo) and expresses an evalu-
ative stance (Du Bois 2009; Page 2015: 331, 338).

Authenticated
One of YouTube’s most popular features is its viral videos, pieces that gain
instant success worldwide, supported by stories in newspapers, magazines and
other media. The first, a rap song called Lazy Sunday, was viewed by more than
one million people in 2005 before it was removed due to copyright issues, as
the material came from a TV show (Burgess and Green 2009: 2–3). Among the
most widespread viral videos are the Evolution of Dance, Sneezing Panda, David
after Dentist, and parodies of the film Hitler’s Downfall (Time Magazine). These
videos represent some of the most popular topics on YouTube, such as talented
performances, cute animals, incidents in daily life, and parodies of films used to
comment on all types of issues.
Today, the most popular user-generated video that is not a music video is Char-
lie Bit Me (2009), with more than 840 million views (September 2016). Despite
this huge popularity, it is not even in the top 30 most-viewed videos of all time.
Two small boys, Harry and Charlie, sit in a chair, Charlie in Harry’s lap. The baby
brother Charlie bites Harry’s finger twice, the second time on purpose as Harry
pushes his finger into Charlie’s mouth. Charlie giggles in a rascally manner, while
Harry complains “Charlie bit me and that really hurts Charlie and it’s still hurt-
ing”. Sindoni (2013: 182) points out the multimodal interaction that takes place:
Harry makes eye contact with their father who is filming the scene, while Charlie
switches his focus between his brother, the camera, and an unspecified target.
Sindoni (2013: 182) discusses the video through comments that doubt the video’s
authenticity.
Home and family are major themes in YouTube user-generated videos. Among
other subjects, family members, teenagers, babies, cats and dogs are popular
(Strangelove 2010: 40–63) and users often post footage for their families and
friends to watch (Lange 2007: 365–376). YouTube videos represent all aspects of
human life, ranging from hobbies such as archery to comments on video games,
from health-related videos of surgical operations to tips for a healthy lifestyle.
They can also depict the dark side, such as terrorism and war. Participation on
YouTube and similar networking sites has generated a vernacular creativity and
do-it-yourself (DIY) culture that has several predecessors in sociocultural life in
general, not just in broadcast media (Burgess and Green 2009: 3, 23, 26). YouTube
is a site for communities of all sizes and for participatory culture of all kinds (Jen-
kins 2008: 3; Burgess and Green 2009: 10).
Today, the popularity of YouTube is measured by several ratings and rankings.
Popularity lists abound, including Most responded and Most discussed as well as
What is trending on YouTube at this very moment (see Burgess and Green 2009).
The list of Most viewed videos of all time consists mostly of professionally-pro-
duced music and children’s music videos from 2010 onward. At the top of the list
is the music video Gangnam Style by Korean singer Psy, with 2.70 billion views
since its upload in December 2012 (December 2016). On YouTube, the content
of a particular video can be viewed and discussed across diverse periods – from a

Authenticated
YouTube 179
week or less, to several years (Burgess and Green 2009: 39). Comments can keep
on building long after a video’s original upload.
The List of the most subscribed users on YouTube differs somewhat, as it lists
smaller scale independent producers as well as the professionals in the media and
music industries. This list of 25 users currently includes nine video game com-
mentators, vloggers, and sketch comedy producers among the professional pop
stars and other entertainment-related producers. These may have started as ordi-
nary amateur users, but successful videos and growing audience members have
turned them into professional video-makers. Video gamer PewDiePie tops the list
with more than 49 million subscribers (December 2016). Originally from Sweden,
this YouTube celebrity talks in English, as do almost all the top-ranked vloggers,
though the Chilean Germán Garmendia, who has two channels on the list – one
vlog, one commentary on video games –, broadcasts in Spanish.
I will next move on to users’ main YouTube activities, all catered for on the
site. These include language-related and other semiotic activities undertaken when
viewing or producing videos, posting and reading comments, and interacting in
other ways with the site.
2.4. YouTube site activities and technological affordances

YouTube offers users several simultaneous activities. They can engage in lan-
guage-related and other semiotic activities: viewing videos, commenting and
interacting (Gibson 1979; Hutchby 2001; Eisenlauer 2013; Androutsopoulos and
Tereick 2015).
YouTube’s main activity consists of viewing and producing videos that can con-
tain multisemiotic material: visual, audio and written (cf. Dynel 2014: 39; Androut-
sopoulos and Tereick 2015: 342). A smaller number of users produce, on their own
or with others. Dynel (2014: 40) points out that a video may document real life or
create a fictional world. As videos are the results of various productive activities,
such as scriptwriting, acting, directing, camerawork and editing, they usually con-
tain several layers of cinematic storytelling (Dynel 2014: 43). This is especially true
if the videos are produced professionally, but ordinary users are quite capable of
filming very carefully scripted videos; they can also edit, mix or create mash-ups.
The second activity concerns the posting and reading of comments. A YouTube
discussion is a written, text-based activity that depends on the participant having
viewed the video (cf. Johansson forthcoming). For example, Adami (2009) con-
siders a video to be a prompt to which users react when posting comments. The
comments are organized either in sequential order according to the newest or the
oldest comment or according to the Top comments, those most frequently viewed.
There are several other activities, grouped as various technologically based
affordances, “functional and relational aspects which frame” the possibilities for
activities and actions on the platform (Hutchby 2001: 444). Technological affor-

Authenticated
Screen capture 1. PewDiePie video and associated information (June 2016)
dances allow users to enhance their experience and to express their opinions. For
instance, when they click the like or dislike buttons, they are commenting by algo-
rithm-based means and sharing their opinion with other users. An example of this
form of commenting appears below in PewDiePie’s YouTube site.
Screen capture 1 is a still from a video produced by PewDiePie, I React to My
Old Videos, posted in July 2014. On this page, users can see the video channel and
subscribe to it. They can also see the basic presentation: publishing date, title, link
to the next video, and playlist. This information and the Subscribe button enhance
the viewer’s experience.
Screen capture 1 also shows user activity in regard to this particular video.
The total numbers of views, likes and dislikes are listed below the still image of

Authenticated
YouTube 181
the video. PewDiePie has more than 46 million subscribers. The video has been
viewed more than 22 million times, liked more than 600,000 times and disliked
by more than 9,000 viewers (June 2016). The comment section is still active two
years after the publication of the video and holds more than 51,000 comments.
Although these figures are very high, only 27 % of all viewers have chosen to click
the like button and only 2.3 % have posted a comment to the video. Thus, it appears
that the majority of viewers have done no more than watch the video.
When users upload a video, YouTube proposes a preset category for them to
choose from, thereby designating the subject of the video. This will depend on the
basic information about the downloaded video. They are Music, Entertainment,
Gaming, People and Blogs, Sports, Comedy, Film and Animation, How to and
Style, News and Politics, Cars and Vehicles, Science and Technology, Education,
Pets, and Animals. This particular PewDiePie video is categorized as People and
Blogs. PewDiePie uses several other categories as well and he will be discussed as
a YouTuber later on (Section 3.2).
The following main Section 3 looks at YouTube videos as communicative gen-
res (several video genres are described in terms of their particular features) and
will also cover self-presentation, other forms of presentation, the popularity phe-
nomenon, micro-celebrity and fame on YouTube.
3. Videos
3.1. Videos as a digital communicative genre

Video is a technical format that does not represent a particular communicative
genre – there is no unique video genre. There are many video genres, although
none has been much studied from the perspective of communicative genre. I pro-
pose below that it is possible to consider video game commentaries, battles, paro-
dies, question-answer and clickbait videos as communicative genres.
In many different theoretical approaches, genre has been understood as a form
of social action (e.g., Miller 1984); it is one of the basic approaches in pragmatics
and in CMC. Giltrow and Stein (2009b) present different conceptions about genre,
from corpus-based linguistics to rhetorical and literary theories. They address the
issues of genre change and the effect on genres when the Internet becomes their
communicative setting (Giltrow and Stein 2009b: 8–9). In Herring, Stein and Vir-
tanen (2013a, 2013b: 4), CMC genres (or modes) form part of the basic question-
ing of the language on the Internet. Both these edited volumes consider mainly
written or text-based CMC genres (or modes) and their characteristics (cf. Myers
2010).
In this chapter, the notion of genre is understood as a basic analytic construct
of an activity type. A communicative genre is defined, first, as a situated activity

Authenticated
type that contains spoken and written language use (e.g., Levinson 1992: 69, 71;
Bergmann and Luckmann 1995; Linell 2009: 198–203). According to Gumperz
(1992: 43–45), activities organize discourse in a specific way and can be described
in relational and indexical terms. Second, the communicative genre informs us
about the situation, the participants and frames of social interaction. The social
interaction in a situation depends on the communicative goals, contributions and
the type of verbal and other semiotic means of expression (Linell 2009: 198–203).
The notion of a digital genre refers to the context in which a genre is situated. Dig-
ital genres are realized in digital texts that are mediated, multimodal, and contain
technological affordances. Within a specific genre, there can be variation in the
way it is realized, and a genre can have several sub-genres. It is possible to con-
sider this variation as a generic feature.
How can video genres be recognized? In YouTube, several videos align with
television genres. In this sense, they have migrated from mass media to this social
media platform (Giltrow and Stein 2009: 8–9; Herring 2013). This is particularly
true for most professionally produced news videos which are same as or similar
to television news genres, such as news interviews but can be shorter, for instance
when presenting sound bites of press conferences (Johansson 2012). Slightly
reconfigured video genres are based on the ideas presented in TV programs, such
as Mythbusters. In the video channel Smarter Every Day, the presenter Destin
introduces everyday physical phenomena and tests them in ordinary settings. His
videos range from The Backwards Brain Bicycle to Poop Splash Elimination. The
bicycle video starts with a personal narrative of riding a bike as a youngster and
continues with a challenge to ride a bike with reversed steering. In the other video
he drops playdough, which looks like feces, into water and explains the physics of
splash formation, while offering a way to prevent it.
One popular genre on YouTube are video blogs, or vlogs, the audio-visual off-
spring of text-based blogs (Giltrow and Stein 2009), which can be presented in
several different ways. The communication is commonly organized according to
the main topic, which includes preset categories, for example lifestyle, beauty and
travel. When users reify these categories by producing video clips, their outputs
depend on the type of self-presentation they adopt, the way they address their
viewers, and their position toward the objects they are discussing. Bloggers and
vloggers can topicalize their private life and be oriented toward a restricted audi-
ence of friends and family. Presenters of this “Dear Diary” type concentrate on
personal and identity construction (Griffith and Papacharissi 2009).
Frobenius (2011) conducted one of the rare analyses of ordinary vlogs. She
studied the opening sequences of 41 monologic vlogs which were all filmed
indoors (Frobenius 2011: 818). Using a conversation analytical framework, she
focused on the interactional activities between vlogger and user. More specifi-
cally, she elicited the discursive strategies vloggers employ to address and involve
their online audiences. She investigated greetings, terms of address, self-identi-

Authenticated
YouTube 183
fication and linguistic markers of discourse particles that draw attention to the
monologue`s opening (Frobenius 2011: 819–820).
Vlogs can be produced professionally by vloggers trying to earn a living.
These vlogs tend to relate exciting travels or contain promotional material aimed
at consumers. In Driving a Chevy Caprice Police Car in the video channel Regular
Cars, a group of young men test a car, discussing its model history and function-
alities while driving it. Their interaction provides consumer information as they
review the car and talk about their personal experiences with it. This contrasts
with the focus on driving featured in Haddington’s (2010) study on turn-taking
and embodied action while driving and arriving at a junction where the object of
talk is which way to take or turn. In another vlog, 3 Years of Travel in 3 Minutes
(Lessons and Adventures), the vlogger presents amazing travel scenes from all
over the world. He interrupts its background music to give advice on living: “You
can choose to be in a place of attraction and of abundance”. This video also uses
the presenter’s personal experience to promote tourist sites to potential consumers
with a mash-up type visual narrative.
Social media have also affected news videos, which often use production
techniques that are typical of digital culture, such as mash-ups and remixes (see
Androutsopoulos and Tereick 2015). The French télézapping videos are built on
short digital clips from different video news and put together into a new video.
One such video is based on a TV interview of the president of France, Nicolas
Sarkozy (Télézapping 2011). Such a remix has at least two frames of interpretation
(see Johansson 2013). Firstly, it resembles the traditional news genre by showing
and informing the viewer about the political comments that have been made on
president Sarkozy’s interview. Secondly, it involves the ordering of clips to build
up a new story in which the political opponents of the president express their
opinion forming a constructed dialogue between parties who have not discussed
together. This mash-up video is an example of the bricolage of digital texts using
copy-paste techniques (cf. Deuze 2006; Adami 2012: 28; Johansson 2013). It is
part of the popular video culture of recontextualization.
In linguistics and pragmatics, the notion of recontextualization is based on the
notions of context and contextualization. Context has been a central concept in
pragmatics (Levinson 1983: 5, 9; Goodwin and Duranti 1992). Fetzer (2004: 4)
considers context to include not only linguistic, but cognitive, social and socio-
cultural contexts. As for contextualization, it is a dynamic process by which the
participants make relevant or otherwise show the meaning of the context in their
actions and situated interactions (Gumperz 1982: 130–152; Gumperz 1992; Auer
1992: 4). In his definition, Linell (1998: 154) considers recontextualization as
“a dynamic transfer-and-transformation of something from one discourse/text-
in-context (the context being in reality a matrix or field of contexts) to another.
Recontextualization involves the extrication of some part or aspect from a text or
discourse, or from a genre of texts and discourses, and the fitting of this part or

Authenticated
aspect into another context”. In recontextualization, the meaning of the recontex-

tualized item or discourse inevitably changes as it is placed in another context.
This has been referred to as resemiotization (Iedema 2003). In other words, the
process gives the video a multi-layered new meaning. Recontextualization and
resemiotization are similar in several other cases, as will be shown below. (See
Leppänen et al. 2014; Androutsopoulos and Tereick 2015: 344–346).
A widespread video genre is, of course, humor videos, which are familiar
from their TV and film history. Their dominant features, in addition to humor,
are playfulness and parody. These core factors create the success of viral videos,
such as Charlie Bit Me, Hitler’s Downfall or David after Dentist (see Section 2.3).
Very often, the humor in YouTube videos is derived from an unexpected incident,
caught on camera, caused by the subject’s failure to achieve a goal (Shifman 2012:
196). Where the event is caused by a deliberate practical joke or prank, the vic-
tim’s reaction provides the source of laughter.
Microphone gaffes also provide humor, but in a slightly different way. These
media situations occur when someone is next to a live microphone but believes
that it is off and makes a private comment. Chovanec (2016) analyzed cases in
which sports commentators committed such gaffes. In his paper, Chovanec (2016:
99) explained that the humor derives from the fact that the media audience, nor-
mally a ratified listener, has become an eavesdropper. He based his explanation
on Goffman’s (1981) participation framework and on the way a media situation
is organized. The speakers’ footing changes from public frontstage performance
to private backstage talk in which the audience is neither addressed nor ratified
(Chovanec 2016: 99). The humor is a result of talk that deviates from the norm of
a public media situation (2016: 99).
In a series of different parodies of the film Hitler’s Downfall (2004), the basic
scene is the one in which Hitler realizes his defeat when the Russian army closes in
on Berlin. The parodies are a result of combining the film scene, depicting Hitler’s
extreme emotional reaction, with a description of a recent social or other event,
creating a remix. In the video channel Hitler Rant Parodies, the topics range from
political parodies, for example imagining Hitler’s reaction to Donald Trump’s
election victory (2016), to parodies of hit songs.
The last genre that I cover in this section is the hoax, familiar from text-
based genres: Heyd (2009: 247) mentions virus hoaxes, giveaway hoaxes, char-
ity hoaxes, urban legends and hoaxed hoaxes. This last sub-genre is a parody of
hoaxes (Heyd 2009: 247). The video example of The Hoax Hotel belongs to none
of these categories, but is a hoax of the hoaxers. In these videos, one Peter Green
calls persons who themselves are trying to cheat others. Visually, the videos are
lively, very carefully constructed stories, usually illustrating the parties on the tele-
phone and providing subtitles to enhance the audio. In the telephone interaction,
the caller plays along with the scam up to the point at which the cheat realizes that
they have been exposed. The verbal interaction and visual presentation are cut by

Authenticated
YouTube 185
written comments that recontextualize the thoughts of the caller or show how the
tentative scam is progressing. The video Finally Baiting a Ridiculous Electric Bill
Scammer! ends with the advice “If you receive a sketchy phone call demanding
an immediate payment over the phone to avoid a power disconnection within the
hour, it’s a scam”.
Videos that have built ‘ordinary’ people into star presenters will be the topic of
the following section. Researchers have used the terms authenticity and micro-ce-
lebrity to describe the rise of individual performers on YouTube.
3.2. From micro-celebrity to “Internet famous”

Discussing one’s life and personal thoughts about private and current societal mat-
ters in video blogs is one of the main ways in which the self-presentation of ordi-
nary persons is performed on YouTube. According to Turner (2004: 8), ordinary
people and their lifeworlds have become a main topic of interest in the last decade
because of the influence of reality TV, a format that made the “private self” the
“privileged object of revelation”. Turner (2010: 3) coined the term demotic turn
to describe the rise in fame of such users on social media, reality TV, talk radio
and websites. In this light, common people emerged as the center of public interest
in television shows, becoming active performers and presenters of their everyday
lives in more or less staged, scripted and edited scenes.
In media studies, authenticity refers to publicizing a private life, providing
an apparently truthful disclosure of oneself while doing so (cf. Marwick 2013:
248). This type of performance includes the creation of an interesting persona, one
that appeals to the public and potential fans (Marwick 2013: 114). When ordinary
individuals post videos, they keep their presentation close to their everyday per-
sonalities and create videos that are different from professionally-produced videos
(Marwick 2013: 119).
Users’ identities on social media have been one of the main topics of research
on videos and YouTube. In one of the first media studies analyses, Senft (2008)
examined the home webcams of women who opened up their private lives to make
friends. They adopted a very intimate tone and confessional practices about their
ordinary life stories that would usually remain in the private sphere (Senft 2008).
A famous example was Lonelygirl15’s vlog with its confessional and authentic
style. It was followed by a large audience, but was later revealed to be a perfor-
mance by actors (Senft 2008: 28–31). However, this particular vlog demonstrates
the use of videos for any user desirous of fame (Senft 2008; Burgess and Green
2009: 28–29).
The rise in popularity of YouTubers, or YouTube celebrities, is the basis of
Internet famous or becoming a micro-celebrity. According to Marwick (2013:
115), micro-celebrity is “a way of thinking of oneself as a celebrity, and treat-
ing others accordingly”. She considered professional video gamers, wannabe pop

Authenticated
stars, beauty bloggers and political activists as current examples of micro-celeb-

rities (Marwick 2013: 115). However, gaining fame on social media is not neces-
sarily the result of a conscious act; it can even be unwanted (Marwick 2013: 115).
One example of a micro-celebrity is Saara, formerly known as Smokahontas,
who has so far produced 59 videos mainly in English on her video channel. She is a
young Finnish YouTuber who started posting in 2014, and one of her very first vid-
eos, What Languages Sound Like to Foreigners, went instantly viral. In this video,
she imitates 14 different languages in a mash-up type video. Her other videos
include, first, professional type of funny and talented performances of role play,
and since she has tried to pursue a musical career, she has music videos as well.
Second, she has confessional videos where she tells about her health issues and
weight loss. Third, she has interactional Q-A videos, in which she replies to snap
chat messages sent by her fans. She builds up her public identity through these dif-
ferent types of videos. In addition, the micro-celebrity is built through polymedia
presence (Madianou and Miller 2013) as she tries to reach her audience through
other accounts – Twitter, Facebook and Instagram – and in a broadcast media con-
text, such as her appearance as a talent-show judge (Östman and Johansson 2016).
The game vlogger PewDiePie has gained the highest level of popularity on
YouTube (see Section 2.3). Why is he so popular? There are several reasons. His
main productions are video games commentaries; he makes his comments while
playing. In addition, he produces several types of videos that represent different
popular video genres. Their topics can be categorized into commentary, humor
and parody, explanations of social media phenomena, responses to fans and vlogs
about his daily life. His commentary videos document his reactions to a variety
of subjects, including hate videos and his own previous videos. When he explains
social media phenomena in his Clickbait videos, he reveals the hidden strate-
gies by which YouTubers attract users to click on their videos. In fan videos, he
addresses his followers and fans directly, reads their comments, performs IQ tests,
etc. When he vlogs about his personal life, he describes how he was kicked out
of his apartment and talks with his girlfriend. His wide range of different types of
videos enables him to speak to many different audiences. He has created an offline
community, mostly young boys and gamers, that he calls his “Broarmy”. While
presenting, he can be humorous, parodic, aggressive and satirical. In other words,
his style-switching performance recalls many comedy shows that rely on vernac-
ular humor. He frequently crosses lines of appropriateness and often curses. This
creates a kind of public ‘off-stage’ talk, using vernacular language in public (cf.
Johansson forthcoming). The huge Internet fame has brought about a lot of video
commentaries and parodies about PewDiePie himself.
Two studies on the representation of the other, which adopt the double frames
of presentation in portraying persons in a transcultural world, will be examined
next.

Authenticated
YouTube 187
3.3. Representations of the other
One type of work presented in video performances is based on the discursive con-
struction of the transcultural representation of the other, especially in the work
by Chun and Walters (2011) and Leppänen and Häkkinen (2012). In this type of
video, humor and parody are the main means of representing a ‘distant other’ as
stereotypical or in a simplified version. These presentations bring out an essential-
ist representation of the other, but through the use of polyphonic voices, the videos
can force the audience to reflect on their own conceptions of the performer or of
the other put on the stage.
A representation of the other is given in Chun and Walters’ (2011) article on
Arab Orientalism in stand-up comedian Wonho Chung’s videos. The videos are
recordings of live performances by a comedian who plays with complex indexical
positionings. His background is Korean and Vietnamese, he has perfect knowl-
edge of the local Arab culture, and he presents himself as a Filipino (he says he
is constantly mistaken for one). One of these videos is Wonho Chung Performs at
Friday Night Live, in which he constructs his complex identity. This allows him
to present humorously ideological alignments with what these researchers call an
Oriental figure (Chun and Walters 2011: 251, 261). When explaining the parody,
Chun and Walters refer to Goffman’s (1974) frame analysis and Bakhtin’s (1984)
polyphony. According to both authors, the video parody is based on a play frame
of a double voice, in which the performer puts himself and the parodied figure on
the stage while acknowledging the ideological alignment of the Arab audience
(Chun and Walters 2011: 251, 254). The humor results from two figures that do not
coincide, as the expressed stances are close to racist discourse (Chun and Walters
2011: 254–255, 261). However, the play frame also includes a satirical stance, and
the performer’s stylized performance reduces the risk of his being judged as racist
(Chun and Walters 2011: 261). Some of the comments that this video performance
has received raise the issue of the problematic nature of stereotypical representa-
tion; in others, the users react positively to the act (Chun and Walters 2011: 262).
In the user comments, the affective stance is very strong, and they convey their
surprise based on the imbalance between their own knowledge of Arab culture and
the performer’s physical appearance, which differs from the “typical” appearance
(Chun and Walters 2011: 264–266).
Another type of analysis was conducted by Leppänen and Häkkinen (2012) on
“buffalaxed videos”. The latter are defined as “parody versions of snippets of motion
pictures, TV broadcasts or musical performances which are originally broadcast in
a language incomprehensible to most westerners and which feature such Others as
Bollywood characters or oriental pop singers” (Leppänen and Häkkinen 2012: 17).
One example of this type of video is a song, Kalluri Vaanil, presented in Tamil by
an Indian actor and performer named Prabhu Deva and remixed into a video par-
ody. It was renamed Benny Lava, and contained subtitles in a completely different

Authenticated
language, English. The subtitles do not portray the original meaning, but ‘translate’
the phonetic meaning heard in the lyrics, thus creating a parodic effect. They are
imitations of the original words but are spelled out in the video’s subtitles in the
language of the target audience, thus producing a new meaning (Leppänen and
Häkkinen 2012: 17–18). In their analysis, the authors pointed out that these types of
videos focus on social and cultural groups to create humor based on hierarchies of
value and humor that is not politically correct (Leppänen and Häkkinen 2012: 20).
They considered that the videos represent stereotypes containing features that can
be culturally racist, but are confusing because they can be interpreted in multiple
ways (Leppänen and Häkkinen 2012: 20). This explanation is similar to the one
given by Chun and Walters (2011) about the polyphonic play frame.
In the following main Section 4, I will examine how videos are commented on
in the light of several research studies, covering networked audience and partici-
pation framework, as well as the methods of text-based and video-based comment
production.
4. Commenting on YouTube videos
4.1. Networked audiences and participation frameworks

Different social media platforms construct interaction between users in different
ways, and even an individual platform has to cater for different audiences, as Mar-
wick and boyd (2011) and Page (2012) have shown with Twitter. Twitter forms a
networked audience that can be conceptualized as a fan base, but audience mem-
bers can participate in promoting users’ (professional) identities and marketing
goods in celebrity- or company-related Twitter accounts (Page 2012). Between
these kinds of audience members, there is a simulated dialogue instead of a true
participatory discussion (cf. Page 2012: 194). However, audiences might turn into
counter-audiences that dislike or attack the video in comment posts. They also
remain unfamiliar to each other and engage in discussions without knowing each
other. This is called context collapse by Marwick and boyd (2011: 124), who point
out that “[s]ocial media collapse diverse social contexts into one, making it diffi-
cult for people to engage in the complex negotiations needed to vary identity pres-
entation, manage impressions, and save face”. In sum, the audiences are situated.
They can respond to the content, communication acts, and other users’ comments,
but also to all the contextual information that the situation carries with it.
When studying participation frameworks, Dynel (2014) and Boyd (2014) both
separated production and reception into different levels based on Goffman’s par-
ticipation framework. They considered viewers to be ratified participants. Dynel
(2014: 42) acknowledged some of the problems of applying different participation
roles, such as the inadaptability of eavesdroppers within YouTube communica-

Authenticated
YouTube 189
tion. However, Chovanec (2016) demonstrated that participants can be assigned an

eavesdropper’s role, in his case of microphone gaffes, where the users are put in
the position of eavesdroppers (see Section 3.1). YouTubers can assume a writer’s
role if they post a comment in the video comment section (Dynel 2014: 46). How-
ever, if they post a video, they would instead adopt a speaking role – video com-
ments were possible until recently (see Section 4.3). In his article, Boyd (2014)
expanded the description of participation roles according to the comment’s func-
tion in the interaction. He distinguished, among others, disruptive roles such as
spammer and troll.
I will now turn to YouTube comment discussions and their threads. They will
be considered as debates and conflicts (research in pragmatics has shown the
issues that give rise to the evolution and development of the latter). I also describe
the sequentiality of video comments.
4.2. Comment discussions and conflict as a sequential phenomenon

YouTube comment discussions are online multiparty discussions (cf. Marcoccia
2004). They are determined by the video’s genre, its topic and its social inter-
action. They have been studied as text-based comments and massive polylogues
(Bou-Franch, Lorenzo-Dus and Garcés-Conejos Blitvich 2012). They can have
different types of referential targets: the video itself, other users’ comments, or a
target that is neither of these, but is outside of the situation (cf. Johansson forth-
coming; Section 4.3). These kinds of comments can be about the vlogger’s former
videos, video culture, or the commenter’s own videos. Comment discussions also
vary in their scope: they can concentrate on local issues or more global ones. Pih-
laja (2014a) analyzed one comment discussion within a community of users and
their membership categorization that turned into a conflict.
Benson (2015: 85) considered YouTube interaction in the framework of multi-
modal discourse analysis. His analysis is based on the model of analysis of spoken
interaction developed in the Birmingham school by John Sinclair and Malcolm
Coulthard (1975). Benson investigated turns, response moves, and patterns of
interactions in data on the Cantonese Word of the Week YouTube series, which
presents discussions around the use of this language (Benson 2015: 87).
Several studies have focused on controversial interactions, impoliteness, inco-
herence and conflict management (Lorenzo-Dus Garcés-Conejos Blitvich and
Bou-Franch 2011; Bou-Franch, Lorenzo-Dus and Garcés-Conejos Blitvich 2012;
Bou-Franch and Garcés-Conejos Blitvich 2014) or on metaphor use in antagonism
(Pihlaja 2014b). What these studies have in common is that they focus on digital
interaction based on the kinds of topics that deal with questions of sociocultural
values, morals, and ethics, such as homosexuality, politics, and religion.
Bou-Franch and Garcés-Conejos Blitvich (2014) examined the unfolding
of a conflict in an online discussion on YouTube. They studied the conflict as a

Authenticated
sequential phenomenon, paying attention to how the conflict began, developed,

and ended in a discussion that centered on teen homosexuality portrayed in a pub-
lic announcement by a Spanish LGBT association. The scholars’ objective was
to develop a new method of analysis that would be based on this type of mate-
rial, replacing the methods originating in studies of face-to-face models and pre-
vious CMC-based analyses on flaming. The authors’ decision to focus on conflict
was based on research into CMC, discourse analysis and politeness theory. Here,
conflict is related to social practices and individuals with differing worldviews
on certain topics. Anonymity and deindividuation are the key factors leading to
polarization in online discussions (Bou-Franch and Garcés-Conejos Blitvich 2014:
20–21). According to this study, a conflict arose in a second move that opposed an
initial argument (2014: 24). Conflicts contain views, stated aggressively, that sup-
port or deny the ideological positions presented in the video and in its comments
(2014: 24–26). Multifunctional utterances in the posting contributed to the analy-
sis; the three stages, beginning, middle and end, were shown as reified constructs
in which the conflict remained unresolved (2014: 33). The two authors proposed,
therefore, that the conflict did not unfold as in face-to-face situations but was con-
centrated in the ongoing struggle of the “middle” (2014: 33).
Section 4.3 describes comment threads posted in video format. The practice is
no longer available, but does provide information about the types of replies that
users post to an initial video.
4.3. Video blogs and their video threads

Since YouTube’s launch, users have been able to respond to videos by upload-
ing another video, a practice called video interaction that ended recently. When
commenting in this way, a series of videos builds up, forming video threads. A
number of studies investigated this particular phenomenon, among them Adami
(2009, 2014, 2015). Her approach to video interaction is based on Kress’s (2010:
33) view of multimodal communication, in which a prompt, in this case the video,
catches the attention of a participant, who frames and interprets the prompt in his
or her response. Adami’s data consisted of videos called Where Do YouTube? by
ChangeDaChannel and The Best Video EVER by Chris Crocker. Both have been
among the most downloaded videos of all time. In Where Do YouTube?, video
blogger ChangeDaChannel invited users to share information about their location.
He showed a written note about his own location in California, thus providing an
example of the kind of reply he anticipated (Adami 2009: 386). He received 550
responses in one month, and those responses were followed by their own responses
(2009: 381). According to Adami (2009: 383), the video thread that followed con-
sisted of responses directed at the initial video, because the users did not engage
in interactions. The answers came from 59 countries, most from the United States
(Adami 2009: 387). In the responses, the users oriented themselves to the prompt

Authenticated
YouTube 191
(the initial video) but displayed their distinctiveness in different multimodal ways,
by creating a themed performance or giving a playful answer to the prompt (2009:
388–392). In her 2014 and 2015 articles, Adami studied the relatedness of the
video responses to Chris Crocker’s The Best video EVER. In this extremely short
video, the only thing Crocker does is blink twice. Adami distinguished several
types of relatedness (or its absence). She asserted that the main way in which
a video response is related to the initial video is adjacency. In these cases, the
response video followed the same action as in the initial video: the users blinked,
as Crocker did in the initial video. They did it in either a positive or negative man-
ner, and some added other non-verbal actions, such as drinking or eating (Adami
2015: 240–241). In this sense, the response videos are related to the original one.
Video responses can also show a deviation from the initial video action. Here,
Adami (2015: 242–243) considered videos in which users commented on the
action they witnessed in the initial video. For example, they might comment on the
video’s ‘bestness’ by speaking to the camera. The video responses, however, could
deviate from the main topic of the initial video by adopting another topic with-
out any explanation (2015: 245). According to Adami (2015: 246–249), the initial
video can be used for creative responses or for transformative remixes and par-
odies. Responsiveness can be implicit, for example, when a video response does
not have an explicit link to the initial video but refers to a background element or
a character or has no connection whatsoever to the initial video (2015: 250–254).
5. Conclusion
For linguists, YouTube is an extremely important and largely unexplored terrain,

interesting for several reasons and offering a variety of possible research subjects.
First, videos are digital multimodal texts that are interesting from the perspective
of visual and verbal expression. The monological expression or interactional situ-
ations in video blogs open up new ways in which to look at language use from the
pragmatic perspective. Indeed, vloggers also present exciting material for study:
for instance, how they talk about the world and how they address audiences. The
video channels of different YouTubers reflect different domains of everyday life,
too. They may be professionally produced, aiming at a specific field in entertain-
ment, such as in the case of comedy-based YouTubers. Common phenomena on
YouTube, such as micro-celebrity and viral videos, can be approached in prag-
matic terms. In addition, videos merit study from the perspective of communica-
tive genre: identifying genres could help in identifying genre ecology and other
social media-related phenomena.
Today, video culture is spreading far beyond YouTube. Several sites, such as
Vimeo or Ted Talks, are devoted to videos. Videos are important in online news
sites and are gaining momentum as an important means of professional communi-

Authenticated
cation. For example, research nights and company events are streamed for larger
audiences. After the events, videos such as EU Science and Innovation, a video
channel that presents science news and gives information about research funding,
make them accessible to larger audiences. Another example is Slush, an annual
international event to promote start-up companies and their investors; in the vid-
eos, it uses all kinds of contemporary video genres and narratives to promote their
business.
Videos are rather independent forms on YouTube, where they form video chan-
nels and are commented on. On other social media sites, they are integrated with
other content; they are, e.g., fused with tweets and Facebook posts or digital news
texts (Johansson 2012). The material that is remediated or shared with means of
expression typically found in other social media can be studied from the perspec-
tive of polymedia (Madianou and Miller 2013). The line between professional
and ordinary users’ videos could also be considered as the relation between online
and offline language situations. Overall, both theoretically and methodologically,
research should cross more borders and adopt interdisciplinary perspectives that
have the potential to open up new research objects and new explanatory models in
digital communication studies.
Videos and video channels
3 Years of Travel in 3 Minutes (Lessons and Adventures). Channel High on Life.

https://www.youtube.com/watch?v=wJF5NXygL4k.
Charlie bit my finger
2007 https://www.youtube.com/watch?v=_OBlgSz8sSM.
Chung, Wonho
2010 Wonho Chung performs at Friday Night Live. https://www.youtube.com/
watch?v=-BbJoJFu1Vc.
David after dentist
2009 https://www.youtube.com/watch?v=txqiwrbYGrs.
EU Science and Innovation. https://www.youtube.com/channel/UC1lhGQ0C_OOlaS1rbx-
lXM5Q.
Evolution of Dance
2006 https://www.youtube.com/watch?v=dMH0bHeiRNg.
Hitler Rant Parodies
2009–2016 https://www.youtube.com/user/hitlerrantsparodies/videos?sort=p&view=
0&flow=grid.
Lava, Benny
2008 Crazy Indian Video, Buffalaxed. https://www.youtube.com/watch?v=
pzd6CgrNcU0.
Lumière, Louis and Auguste
1895a La sortie de l’usine Lumière à Lyon [Workers Leaving the Lumière Factory].
https://www.youtube.com/watch?v=NwRAUniWJPY.

Authenticated
YouTube 193

1895b L’arrivée du train à la gare de Ciotat [The Arrival of the Train]. https://www.
youtube.com/watch?v=b9MoAQJFn_8.
1896 Arroseur arrosé [The Sprinkler Sprinkled]. https://www.youtube.com/watch?v
=UlbiNuT7EDI.
Me at the zoo
2005 https://www.youtube.com/watch?v=jNQXAC9IVRw
PewDiePie
2014 I react to my old videos. (Fridays With PewDiePie – Part 80). https://www.
youtube.com/watch?v=0mJiPcKybzU.
Regular cars
https://www.youtube.com/user/RegularCars.
Saara
https://www.youtube.com/user/SmoukahontasOfficial.
Slush
https://www.youtube.com/user/KickNetwork.
Sneezing Panda
2006 https://www.google.fi/webhp?sourceid=chrome-instant&ion=1&espv=
2&ie=UTF-8#q=Sneezing+panda.
Smarter Every Day
https://www.youtube.com/user/destinws2/videos?sort=p&view=0&
flow=grid.
Télezapping
2011 Sarkozy attaque, ses adversaires répliquent [Sarkozy attacks, his a dversaries
reply]. http://www.dailymotion.com/video/xwt97g_telezapping-sarkozy-
attaque-ses-adversaires-repliquent_news.
The Hoax Hotel
https://www.youtube.com/channel/UCnNlJNSRxa3PF8XrKHOEPug.
YouTube and popularity rankings
List of most viewed YouTube videos. Wikipedia. https://en.wikipedia.org/wiki/List_of_

most_viewed_YouTube_videos. Accessed May 6, 2016.
List of the most subscribed users on YouTube. Wikipedia. https://en.wikipedia.org/wiki/
List_of_the_most_subscribed_users_on_YouTube. Accessed September 16, 2016.
Music videos. Wikipedia. https://en.wikipedia.org/wiki/Music_video#1926.E2.80.931959:_
Talkies.2C_soundies.2C_and_shorts. Accessed September 27, 2016.
Time Magazine s. a. YouTube’s 50 Best Videos. Accessed May 6, 2016.
YouTube Red. https://www.youtube.com/red. Accessed September 16, 2016.
YouTube Statistics. https://www.youtube.com/yt/press/statistics.html. Accessed June 24,
2016.

Authenticated
References
Adami, Elisabetta
2009 “We/YouTube”: Exploring sign-making in video-interaction. Visual Commu-
nication 8: 379–99.
Adami, Elisabetta
2012 The rhetoric of the implicit and the politics of representation in the age of
copy-and-paste. Learning, Media and Technology 37(2): 131–144.
Adami, Elisabetta
2014 “Why did dinosaurs evolve from water?”: (In)coherent relatedness in YouTube
video interaction. Text & Talk 34: 239–259.
Adami, Elisabetta
Androutsopoulos, Jannis and Jana Tereick
2015 Language and discourse practices in participatory culture. In: Alexandra Geor-
gakopoulou and Tereza Spilioti (eds.), The Routledge Handbook of Language
and Communication, 341–356. Oxon/New York: Routledge.
Auer, Peter
1992 Introduction: John Gumperz’ approach to contextualization. In: Peter Auer and
Aldo di Luzio (eds.), The Contextualization of Language, 1–38. Amsterdam/
Bakhtin, Michail M.
1984 Problems of Dostoevsky’s Poetics. Edited and translated by Caryl Emerson.
Minneapolis: University of Minnesota Press.
Benson, Phil
2015 YouTube as text. In: Rodney H. Jones, Alice Chik and Christoph A. Hafner
(eds.), Discourse and Digital Practices. Doing Discourse Analysis in the Dig-
ital Age, 81–96. Oxon/New York: Routledge.
Bergmann, Jörg R. and Thomas Luckmann
1995 Reconstructive genres of everyday communication. In: Uta M. Quasthoff (ed.),
Aspects of Oral Communication, 289–304. Berlin/New York: de Gruyter.
Bou-Franch, Patricia and Pilar Garcés-Conejos Blitvich
2014 Conflict management in massive polylogues: A case study from YouTube.
Journal of Pragmatics 73: 19–36.
Bou-Franch, Patricia, Nuria Lorenzo-Dus and Pilar Garcés-Conejos Blitvich
2012 Social interaction in YouTube text-based polylogues: A study of coherence.
Boyd, Michael S.
2014 (New) participatory framework on YouTube? Commenter interaction in US
political speeches. Journal of Pragmatics 72: 46–58.
boyd, danah m. and Nicole B. Ellison
2007 Social network sites: Definition, history, and scholarship. Journal of Computer-

Authenticated
YouTube 195
Bruns, Axel
2008 Blogs, Wikipedia, Second Life, and Beyond: From Production to Produsage.
New York: Peter Lang.
Burgess, Jean and Joshua Green
2009 YouTube. Online Video and Participatory Culture. Cambridge, UK/Malden:
Polity.
Cardoso, Gustavo
2008 From mass to networked communication: Communicational models and the
informational society. International Journal of Communication 2: 587–630.
Chovanec, Jan
2016 Eavesdropping on media talk: Microphone gaffes and unintended humour in
sports broadcasts. Journal of Pragmatics 95: 93–106.
Chun, Elaine and Keith Walters
2011 Orienting to Arab Orientalism: Language, race, and humor in a YouTube
video. In: Crispin Thurlow and Kristine R. Mroczek (eds.), Digital Discourse.
Language in the New Media, 251–273. New York: Oxford University Press.
Couldry, Nick
2012 Media, Society, World: Social Theory and Digital Media Practice. Cambridge,
UK/Malden: Polity.
Couldry, Nick and Andreas Hepp
2013 Conceptualizing mediatization: Contexts, traditions, arguments. Communica-
tion Theory 23(3): 191–202.
Deuze, Mark
2006 Participation, remediation, bricolage: Considering principal components of a
digital culture. The Information Society 22: 63–75.
Develotte, Christine, Richard Kern and Marie-Noëlle Lamy
2011 Décrire la conversation en ligne: le face à face distanciel. Lyon: ENS Éditions.
Dijck, José van
2013 The Culture of Connectivity. A Critical History of Social Media. Oxford: Ox
ford University Press.
Dijck, José van and Thomas Poell
2013 Understanding social media logic. Media and Communication 1: 2–14.
Du Bois, John
2007 The stance triangle. In: Robert Englbretson (ed.), Stancetaking in Discourse.
Subjectivity, Evaluation and Interaction, 139–182. Amsterdam/Philadelphia:
Benjamins.
Dynel, Marta
ics 73: 37–52.
Eisenlauer, Volker
2013 A Critical Hypertext Analysis of Social Media. The True Colours of Facebook.
London/New York: Bloomsbury Academic.
Fetzer, Anita
2004 Recontextualizing Context. Amsterdam/Philadelphia: Benjamins.

Authenticated
Gee, James Paul

2014 What Video Games Have to Teach Us about Learning and Literacy. New York:
Palgrave Macmillan.
Gere, Charlie
2002 Digital Culture. London: Reaktion Books.
Gibson, James
1979 The Ecological Approach to Visual Perception. Boston: Houghton Mifflin.
Giltrow, Janet and Dieter Stein
2009a Genres in the Internet. Issues in the Theory of Genre Amsterdam/Philadelphia:
Benjamins.
Giltrow, Janet and Dieter Stein
2009b Genres in the Internet. Innovation, evolution, and genre theory. In: Janet Gil-
trow and Dieter Stein (eds.), Genres in the Internet, 1–25. Amsterdam/Phila-
delphia: Benjamins.
Goffman, Erving
1974 Frame Analysis. An Essay on the Organization of Experience. Cambridge,
MA: Harvard University Press.
Goffman, Erving
1981 Forms of Talk. London: Basil Blackwell.
Goodwin, Charles and Alessandro Duranti
1992 Rethinking context: An introduction. In: Alessandro Duranti and Charles
Goodwin (eds.), Rethinking Context: Language as an Interactive Phenome-
non, 1–42. Cambridge: Cambridge University Press.
Gumperz, John
1982 Discourse Strategies. Cambridge: Cambridge University Press.
Gumperz, John
1992 Contextualization revisited. In: Peter Auer and Aldo di Luzio (eds.), The Con-
textualization of Language, 39–54. Amsterdam/Philadelphia: Benjamins.
Griffith, Maggie and Zizi Papacharissi
2009 Looking for you: An analysis of video blogs. First Monday 15. <http://first-
monday.org/ojs/index.php/fm/article/view/2769>. Accessed: 13 Oct. 2016.
doi: 10.5210/fm.v15i1.2769.
Haddington, Pentti
2010 Turn-taking for turn-taking: Mobility, time, and action in the sequential organ-
ization of junction negotiations in cars. Research on Language and Social
Interaction 43(4): 372–400.
Helasvuo, Marja-Liisa, Marjut Johansson and Sanna-Kaisa Tanskanen
2014 Johdatus digitaaliseen vuorovaikutukseen. [Introduction to digital interac-
tion]. In: Marja-Liisa Helasvuo, Marjut Johansson and Sanna-Kaisa Tan-
skanen (eds.), Kieli verkossa. Näkökulmia digitaaliseen vuorovaikutukseen
[Language in the Net. Perspective on to Digital Interaction], 9–28. Helsinki:
Suomalaisen kirjallisuuden seura.
Herring, Susan
Internet. http://www.languageatinternet.org/articles/2007/761.
Herring, Susan
2013 Discourse in Web 2.0: familiar, reconfigured, and emergent. In: Deborah Tan-

Authenticated
YouTube 197
nen and Anna M. Tester (eds.), Discourse 2.0. Language and New Media,
1–24. Washington, DC. Georgetown University Press.
2013a Pragmatics of Computer-Mediated Communication. Berlin/Boston: de
Gruyter.
2013b Introduction to the Pragmatics of Computer-mediated Communication. In:
Susan Herring, Dieter Stein and Tuija Virtanen (eds.), Pragmatics of Computer-
Mediated Communication, 3–32. Berlin/Boston: de Gruyter.
Heyd, Theresa
2009 A model for describing “new” and “old” properties of CMC genres. In: Janet
Giltrow and Dieter Stein (eds.), Genres in the Internet, 239–262. Amsterdam/
Hutchby, Ian
2001 Technologies, texts and affordances. Sociology 35: 441–456.
Iedema, Rick
2003 Multimodality, resemiotization: Extending the analysis of discourse as mul-
ti-semiotic practice. Visual Communication 2(1): 29–57.
Jenkins, Henry
2006 Convergence Culture: Where New and Old Media Collide. New York: New
York University Press.
Jenks, Christopher and Alan Firth
2013 Synchronous voice-based computer-mediated communication. In: Susan Her-
ring, Dieter Stein and Tuija Virtanen (eds.), Pragmatics of Computer-Medi-
ated Communication, 217–241. Amsterdam/Philadelphia: Benjamins.
Jewitt, Carey
2009 The Routledge Handbook of Multimodal Analysis. Abingdon: Routledge.
Johansson, Marjut
2012 Political videos in digital news discourse. In: Anita Fetzer and Lawrence
N. Berlin (eds.), Dialogue in Politics, 43–67. Amsterdam/Philadelphia: Benja-
mins.
Johansson, Marjut
2013 Remixing opinions: Video quotations and stances in video press reviews.
Recherches en Communication 39: 95–115.
Johansson, Marjut
forthcoming Everyday opinions in news discussion forums: Vernacular public dis-
course. Discourse, Context and Media.
Jucker, Andreas and Christa Dürscheid
framework. Linguistik Online 56. http://www.linguistik-online.de/56_12/juck-
erDuerscheid.html. Accessed: 13 Oct. 2016.
Kaplan, Andreas M. and Michael Haenlein
2010 Users of the world, unite! The challenges and opportunities of social media.
Business Horizons 53: 59–68.
Kress, Gunther
2010 Multimodality. A Social Semiotic Approach to Contemporary Communication.
London/New York: Routledge.

Authenticated
Kress, Gunther and Theo van Leeuwen

2006 Reading Images: The Grammar of Visual Design. London: Routledge.
Lange, Patricia G.
of Computer-Mediated Communication 13: 361–80.
Leppänen, Sirpa and Ari Häkkinen
2012 Buffalaxed superdiversity: Representations of the other on YouTube. Diversities 2:
17–33. http://www.mmg.mpg.de/fileadmin/user_upload/Subsites/Diversities/
Journals_2012/2012_14-02.pdf#page=19. Accessed: 13 Oct. 2016.
Leppänen, Sirpa, Kytölä Samu, Henna Jousmäki, Saija Peuronen and Elina Westinen
2014 Entextualization and resemiotization as resources for identification in social
media. In: Philip Seargeant and Caroline Tagg (eds.), The Language of Social
Media, 112–136. Basingstoke: Palgrave Macmillan.
Levinson, Stephen C.
1983 Pragmatics. Cambridge: Cambridge University Press.
Levinson, Stephen
1992 Activity type and language. In: Paul Drew and John Heritage (eds.), Talk at
Work. Interaction in Institutional Settings, 66–100. Cambridge: Cambridge
University Press.
Lievrouw, Leah A.
2009 New media, mediation, and communication study. Information, Communica-
tion, and Society 12: 303–325.
Linell, Per
1998 Approaching Dialogue. Talk, interaction and contexts in dialogical perspec-
tive. Amsterdam/Philadelphia: Benjamins.
Linell, Per
2009 Rethinking Language, Mind and World Dialogically. Interactional and Con-
textual Theories of Human Sense-making. Charlotte, NC: Information Age
Publishing.
Livingstone, Sonia
2009 On the mediation of everything. Journal of Communication 59: 1–18.
Lorenzo-Dus, Nuria, Pilar Garces-Conejos Blitvich and Patricia Bou-Franch
2011 On-line polylogues and impoliteness: The case of postings sent in response to
the Obama reggaeton YouTube video. Journal of Pragmatics 4: 2578–2593.
Madianou, Mirca and Daniel Miller
2013 Polymedia: Towards a new theory of digital media in interpersonal communi-
cation. International Journal of Cultural Studies 16(2): 169–187.
Marcoccia, Michel
2004 On-line polylogues: Conversation structure and participation framework in
Internet newsgroups. Journal of Pragmatics 36: 115–145.
Marcoccia, Michel
2016 Analyser la Communication Numérique Écrite. Paris: Armand Colin.
Marwick, Alice E.
2013 Status Update. Celebrity, Publicity and Branding in the Social Media Age.
New Haven/London: Yale University Press.
Marwick, Alice E. and danah boyd
imagined audience. New Media & Society 1: 114–133.

Authenticated
YouTube 199
Miller, Carolyn R.
1984 Genre as social action. Quarterly Journal of Speech 70: 151–167.
Myers, Greg
2010 The Discourse of Blogs and Wikis. London/New York: Continuum.
O’Halloran, Kay
2004 Multimodal Discourse Analysis: Systemic Functional Perspectives. London/
New York: Continuum.
Östman, Sari and Marjut Johansson
2016 Making of a social media celebrity: Web sociability and negotiation of an online
relationship. Paper presented at the Celebrity Studies Conference, Amsterdam,
June 2016. https://celebritystudiesconference.com/conference-programme/.
Page, Ruth
2012 The linguistics of self-branding and micro-celebrity in Twitter: The role of
hashtags. Discourse & Communication 6: 181–201.
Page, Ruth
2015 The narrative dimensions of social media storytelling. Options for linearity
and tellership. In: Anna D. Fina and Alexandra Georgakopoulou (eds.), Black-
well Handbooks in Linguistics: Handbook of Narrative Analysis, 329–347.
Somerset, US: Blackwell.
Pihlaja, Stephen
2014a “Christians” and “bad Christians”: Categorization in atheist user talk on You-
Tube. Text & Talk 34: 623–639.
Pihlaja, Stephen
2014b Antagonism on YouTube: Metaphor in Online Discourse. London: Blooms-
bury.
Senft, Theresa M.
2008 Camgirls: Celebrity and Community in the Age of Social Networks. New York:
Peter Lang.
Shifman, Limor
2012 An anatomy of a YouTube meme. New Media & Society 14(2): 187–203.
Sinclair, John and Malcolm Coulthard
1975 Toward an Analysis of Discourse. Oxford: Oxford University Press.
Sindoni, Maria G.
2013 Spoken and Written Discourse in Online Interactions: A Multimodal Approach.
New York/London: Routledge.
Strangelove, Michael
2010 Watching YouTube: Extraordinary Videos by Ordinary People. Toronto: Uni-
versity of Toronto Press.
Suominen, Jaakko, Sari Östman, Petri Saarikoski and Riikka Turtiainen
2013 Sosiaalisen median lyhyt historia [A Short History of Social Media]. Helsinki:
Gaudeamus.
Thurlow, Crispin and Kristine R. Mroczek
2011 Digital Discourse: Language in the New Media. New York: Oxford University
Press.
Turner, Graeme
2004 Understanding Celebrity. London: Sage.

Authenticated
Turner, Graeme
2010 Ordinary People and the Media. Thousand Oaks, CA: Sage.
Yus, Francisco
2011 Cyberpragmatics. Internet-Mediated Communication in Context. Amsterdam/

Authenticated
8. Twitter
Michele Zappavigna
Abstract: This chapter explores the language of microblogging by focusing on dis-

course produced via Twitter, a popular microblogging service. In the first section
I consider microblogging as a semiotic practice, trace its historical development
and investigate the interdisciplinary research into this form of communication. My
focus will be on linguistic work in the areas of pragmatics and discourse analy-
sis which explores the dominant communicative conventions that have arisen via
Twitter, such as retweeting and hashtagging. I will then turn to the social concept
of what Zappavigna (2012) terms ‘searchable talk’, i.e. discourse which renders
opinion and sentiment readily findable through resources such as social tagging.
The chapter concludes by suggesting the important role that ambient affiliation
plays in microblogging by forging communities through negotiating values, an
issue to which I will return in my chapter on evaluation in social media (cf. Ch.
16, this volume).
1. Introduction: Microblogging as a semiotic practice
The advent of social media has seen a proliferation of semiotic resources for con-
struing experience, negotiating values, and enacting identities and communities
online. Microblogging is a short-form social media technology, involving the
posting of small, typically episodic, messages aimed at internet-mediated audi-
ences. These texts appear on social media services, such as Twitter and Weibo, as
chronologically unfolding streams of posts associated with a user’s social profile.
Microblogging services are “specifically designed to broadcast short but regular
bursts of content to particularly large audiences well beyond a user’s direct social
network” (Murthy 2013: 12). Users subscribe to the feeds of different accounts,
and search posts in the public stream with native or third party applications. An
example of a post is the following, taken from a corpus of tweets about coffee
(Zappavigna 2013):1
1
Examples used in this chapter are drawn from this corpus unless otherwise noted. This
corpus of ‘coffee tweets’ has been used to investigate how people use language to forge
social bonds online, here by aligning around shared feelings about the positive effects
of morning coffee. This has been explored as a shared process of ‘ambient affiliation’
that we will return to at the end of this chapter.
DOI 10.1515/9783110431070-008
Authenticated
202 Michele Zappavigna
Figure 1. An example of a microblogging post
As Figure 1 suggests, different kinds of textual elements augment the ‘body’ of a

post, including social tags (e.g. #needcoffee in the above), various types of meta-
data (e.g. timestamps), and material relating to the author’s profile such as an icon
and username.
At the time of writing in 2016, the range of genres deployed via microblogging
is vast, and correspondingly challenging to map them exhaustively. By deploying
a particular instantiation of a genre “an individual user establishes a contract with a
particular social group or a segment of social media users” (Artemeva 2015: 282).
These genres span most domains of social life from the personal and domestic, to
the political and national, incorporating a range of linguistic functions, from ide-
ational broadcasting of content, to interpersonal sharing of feelings (Zappavigna
2015). Descriptions of microblogging usually imply that these genres are conver-
sational, and that microblogging involves some kind of ‘conversational exchange’
(Honeycutt and Herring 2009). The practice has been described as “lightweight
chat” (Starbird et al. 2010: 242), as “prompting opportunistic conversations” (Zhao
and Rosson 2009: 251), as “a specific social dialect, in which individual users are
clearly singled out and engaged in a conversation” (Holotescu and Grosseck 2009:
163), and as constituted by “dialogue acts” (Ritter, Cherry and Dolan 2010: 172).
This chapter begins by broadly surveying interdisciplinary research into
microblogging and the kinds of features/dimensions of communication as well as
research methods that have been applied in this domain. It then historically con-
textualises microblogging as a semiotic practice, introducing Twitter, one of the
most popular platforms, as a semiotic technology. Some of the main communica-
tive conventions that have arisen on Twitter (such as @mentions, retweets, and
hashtags) are then introduced, alongside linguistic studies incorporating these fea-
tures. Since the social function of social media communication is, broadly speak-
ing, about forging relationships (across a large variety of often quite niche-like
domains), linguistic studies of social media are necessarily bound by the need to
locate theory and analytical methods that are able to account for how such social
affiliation is enacted in discourse. The final section of this chapter thus consid-
ers attempts to understand how people use language to forge (and disrupt) social
bonds, i.e. issues that are currently at stake in social media research.

Authenticated
Twitter 203
2. Interdisciplinary research into microblogging
Research into microblogging is a multidisciplinary arena, spanning domains out-

side linguistics such as psychology, media studies, marketing, and many others.
It often focuses on snapshots (Popescu and Pennacchiotti 2010) of ‘real-time’
microblogging communication at particular temporal or contextual moments, for
instance during natural disasters (Shaw et al. 2013), elections, or as a backchannel
during live entertainment (Highfield, Harrington and Bruns 2013). The diverse
and extensive range of naturally-occurring discourse available on Twitter has also
meant that work has been conducted on this platform across most major fields in
linguistics, ranging from pragmatics, sociolinguistics, corpus linguistics to com-
putational linguistics, and systemic functional linguistics. These studies explore
communication about every conceivable topic from the nature of people’s discus-
sion of e-cigarettes (Cole-Lewis et al. 2015) to the interactions of athletes (Ham-
brick et al. 2010). Much research has investigated a wide range of pragmatic issues
such as the nature of Twitter corporate apologies (Page 2014), politeness in tweets
(Sifianou 2015), the speech act of self-praise (Dayter 2014), linguistic features
associated with sexual aggression in tweets, e.g. the language patterns seen in rape
threats, (Hardaker and McGlashan 2016), gender representation, and issues of per-
ceived linguistic authenticity (Kytölä and Westinen 2015).
Public access to Twitter’s Application Programming Interfaces (API)2 means
that tweets are relatively easy to collect and hence a vast array of domain-spe-
cific research into particular niche issues has built up around Twitter.3 Thus it is
unsurprising that the dominant research strategy has been to analyse (both quanti-
tatively and qualitatively) social media corpora (see for example Baker and McEn-
ery 2015; Page 2012a, 2012b; Seargeant and Tagg 2014; Zappavigna 2012). Some
of this work adopts a sociolinguistic interest in exploring variations in tweeting
practices across variables such as gender, ethnicity, and language variety (Adnan,
Longley and Khan 2014; Bamman, Eisenstein and Schnoebelen 2014; Neubig and
Duh 2013) and individual variation (Kelsey and Bennett 2014). There has also
been some interest in exploring social media texts in terms of narrative theory that
2
The API is the language that software tools use to communicate with Twitter’s back-end
database in order to assemble subsets of posts from the public feed. The ‘garden hose’
is the freely available access provided by Twitter to its data, and is a random sample of
posts (collected according to Twitter’s particular algorithm defining randomness) and
a subset of ‘fire hose’ access to all publically available tweets. Most Twitter corpora
are produced by creating what Popescu and Pennacchiotti (2010) refer to as a Twitter
‘snapshot’, featuring some entity, a time period, and a selection of tweets.
3
Payment, however, is obligatory if an exhaustive set of tweets is required, beyond the
‘random’ sample offered by the ‘garden hose’ feed of tweets that Twitter makes freely
available.

Authenticated
has developed out of early research into narrative such as Labov and Waletzky’s
(1967). For example, in accord with Page’s work, Dayter (2015) approaches tweets
as instances of “small stories” and suggests that these stories can span multiple
tweets and be analysed using the orientation, complication, evaluation and reso-
lution genre structure regularly used to characterise narrative. She also identifies
two types of stories that she claims are typical of an ‘eyewitness microgenre’. The
first is a ‘delayed resolution narrative’ in the form of live commentary which can
only be seen as a single coherent narrative once completed. The second is a ‘tiny
story’ which is a fragmented account of an everyday activity in which the narrative
stance is not clear. A detailed survey of research methods used to explore social
media is provided by the textbook Researching the Language of Social Media
(Page et al. 2014).
Despite the accumulation of linguistic research into social media, the major-
ity of studies investigating Twitter do so from a communication theory or media
studies perspective without applying a particular theory of language to the com-
municative patterns observed. The detailed linguistic analysis of particular types
of social media texts is a complex undertaking from which a wide range of meth-
odological challenges can arise. These include ethical issues due to the intimate
material revealed in posts, problems related to the multimodal nature of the data,
and the complexity of accounting for the relationship between online and offline
contexts (Bolander and Locher 2014). An ongoing and important consideration for
linguistic projects exploring social media is how “bespoke” the analytical instru-
ment to be applied should be, given the fast pace at which social media practices
and platforms change (Giles et al. 2014: 49). Tools that are designed specifically
for analyzing particular types of social media communication, rather than for
exploring general shifts in meaning-making, risk redundancy. This is particularly
perilous in view of the effort required to develop methods to capture the multi-
modal nature of social media texts.
3. Tracing the historical development of microblogging
3.1. Twitter and older media practices

The attempt to identify its historical antecedents is a characteristic feature of the
initial phases of research into new fields. This also applies to the new technol-
ogy of microblogging and leads to the question how it evolved from older media
practices (see for example Murthy 2013). Commentators have been keen to his-
toricise Twitter in terms of older restricted forms of communication prevalent in
the eighteenth and nineteenth century. These include the telegraph, and forms of
record keeping, such as diaries (Murthy 2013). Unravelling the multidimensional
interaction between the evolution of communicative potential (the meanings we

Authenticated
Twitter 205
can make within a particular culture) and changes in technological platforms sup-
porting communication is no easy task. Historically, communication has always
been intimately connected to the material and semiotic particularities of different
media. This means that, on the one hand, it may be productive to think about the
shifts in meaning potential over time that are afforded by each new reading or
writing technology. However, on the other hand, this line of thinking can tend to
background the fact that the interaction between technical platforms and com-
municative practices is a complex feedback loop, evolving in tandem with other
contextual factors.
Microblogging has now been around long enough that it features a broad range
of human expression, involving a wide range of communicative genres and con-
texts, as noted earlier. Because it spans areas as disparate as personal feelings
about daily routine (Zappavigna, 2014) and discourse about elections (Bruns et
al. 2015), microblogging has been seen as blurring the public and private realm
in novel ways (Baym and boyd 2012). However, public and private domains have
regularly intersected throughout history to differing degrees. Communicative
channels that we might characterise, with modern eyes, as highly personal, such as
diaries, in fact have emerged out of traditions in which diaries were often public
documents intended to be shared (Humphreys et al. 2013). A comparison of tweets
with eighteenth and nineteenth century diaries suggested that both condense per-
sonal experience into small texts intended to be distributed in the public realm,
and thus form part of the history of “personal writing for public consumption”
(Humphreys et al. 2013: 414). What is most distinctive about microblogging, if we
are to situate it within the history of communication, is the level of interactivity
enabled with potential audiences for any given text, and the potential reach of that
text to an audience of millions.
Facebook’s ‘status update’ feature is one of the earliest forms of microblogging
that was available to a large cohort of users. This feature invited users to respond
to the prompt question (sometimes referred to as a tagline) “What are you doing
right now?”. Prior to 2008, responses were restricted to a template featuring the
verb form ‘is’, constraining the type of updates that could be fashioned to posts.
This structure created a relationship of attribution, linking the account profile to
the content of the post. The examples below are posts conforming to this structure
by the author from 2007:
is listening to the rain
is having her morning coffee
And, due to the limitations imposed by the template, ungrammatical constructions

occurred, as in:
is just ate BBQ pineapple

Authenticated
Later, the verb form was abandoned, and the prompt shifted to the command “Share
what’s on your mind”, which allowed any type of grammatical construction in the
response. At the time of writing, the prompt is the question “What’s on your mind?”
The choices in response made available by these “invocations to participation”
(Burgess 2014: 283) offer insight into the types of semiotic behaviour a particular
microblogging service is constructing as normative, as well as the way the service
is hoping to brand itself as a media platform. The restrictions imposed are not nec-
essarily static and “[w]hen they do change, they are often accompanied by public
relations materials alerting us to shifts in business logic” (Burgess 2014: 283).
3.2. The advent of Twitter as a microblogging service

Shortly after status updating on Facebook had begun to develop as a practice,
Twitter, a social media service designed purely for microblogging, was released.
It would emerge as “a key player in the colonization of the internet by corporate
social media” (Sharma 2013: 49). Moving from “a niche service to a mass phe-
nomenon” (Weller et al. 2013: xxix), Twitter has integrated itself into important
domains of social life such as journalism, public communication, politics, and
activism, as well as corporate domains such as market research, advertising, and
branding. It has become largely synonymous in most countries with the broader
practice of microblogging, having outlived other nascent services, with the excep-
tion of China where Sina Weibo is the dominant platform (Weller et al. 2013).
Figure 2 below shows founder Jack Dorsey’s notes relating to the initial design
of Twitter. These are an interesting historical document in terms of the develop-
ment of Twitter as a semiotic platform.
As this image suggests, the initial design concept for Twitter was focused on
the notion that a user has a ‘status’ that they wished to share. The reference to the
chat protocol, jabber, suggests that this concept had been drawn from instant mes-
saging, where a status indicated a user’s conversational availability and was part
of the maintenance of online ‘social presence’ (Baron at al. 2005). For example an
‘away message’, indicating that a user is unavailable might be classed as informa-
tional or expressive (Nastri, Peña and Hancock 2006: 1028):
At the library [informational]
I hate this weather [expressive]
These short texts featured a range of speech functions, with assertives being the
most common type (Nastri, Peña and Hancock 2006). A quick search of Twitter
at the time of writing easily retrieves similar posts to the above away messages,
indicating that this early function remains:
been at the library for 4 hours …
I hate this weather

Authenticated
Twitter 207
Figure 2. Twitter design notes produced by Jack Dorsey in 2006. Retrieved from
https://www.flickr.com/photos/jackdorsey/182613360/in/photostream/
Twitter posts, which came to be known as ‘tweets’, were constrained in Dorsey’s

design to 140 characters, in order to allow messages to conform to the 160 char-
acter limit set for SMS on most phones in 2006 (factoring in 20 characters for the
account username). Figure 3 is the first tweet by one of Twitter’s co-founders. The
tweet features a brief message, “Ok we are in the car”, which has been retweeted 8
times, and ‘liked’ 5 times. At this stage of development there was no technical or
communicative convention for replying to other users or aggregating tweets into
types using social tags.
Figure 3. An example of a tweet. Retrieved from https://twitter.com/biz/statuses/10345

Authenticated
Many early tweets functioned as experiential status updates directly related to

a user’s activities, for example, this tweet from 2006 by Twitter founder Biz Stone:
walking the dog
However, just as the Facebook prompt question evolved, Twitter’s prompt shifted
from “What are you doing?” to “What’s happening?” Users had begun to post
about topical events and a wide range of phenomena not limited to their personal
activities (Tagg 2015). At the same time the business model underlying Twitter,
and also Facebook, began to increasingly rely on forms of data mining used to
collect material relevant to corporations and brands (Burgess 2014).
Figure 4 details the basic functionality of a tweet as made available via the
Twitter website in 2016. Tweets incorporate the user’s account name, username,
and a timestamp indicating when the message was posted, as well as a range of
interactive elements providing options to subscribe to the user’s stream of posts,
or to interact with it in some way. This functionality will look different depending
on how the user is accessing the service, for instance either via the web interface
or by a mobile device, or by using a native or third party application. In addition,
it should be noted that Twitter is a moving target in terms of research, since social
media platforms are constantly evolving alongside shifts in contexts of use (Hogan
and Quan-Haase 2010). While this is essentially the case with any form of commu-
Figure 4. The anatomy of a tweet.

Authenticated
Twitter 209
nication, it is highlighted by social media discourse since constant change seems

to be part of maintaining the popularity or social relevance of the media employed.
4. Twitter communicative conventions
While first generation microblogging services were not designed to directly sup-
port ‘conversational’ exchanges between users, perhaps with the exception of
Jaiku’s (another microblogging service) message threading capabilities, the abil-
ity to reference and address posts to particular users has now become a funda-
mental feature of microblogging. This is part of a more general social need for
microblogging to manage discursive heteroglossia (Kristeva 1980), in other words
to allow users to engage with other voices, opinions, and information available
in the social stream. Bruns and Moe (2014) suggest that there are three struc-
tural layers of communication possible with Twitter, corresponding to the types of
information exchange and user interaction: micro (follower-followee networks),
meso (hashtagged exchanges), and macro (@reply conversations). This section
reviews research into key conventions used in Twitter communication that have
been developed both by the system designers and organically through community
use. As Zappavigna (2012) notes, these conventions center around three linguistic
markers: addressing and referencing other users with @, republishing other tweets
with RT, and labelling topics with #.
4.1. Addressivity and @mentions

The first convention marks address with the @ character when a user wishes
to explicitly direct their micropost at another user. In these instances @ will be
deployed as a deictic marker, as in the following example:
@User1 there will always be time for #coffee #beer and #wine ☺
Used in this way, the @ character indicates that the username which it precedes
is directly addressed in the tweet.4 As such it functions to mark a vocative, often
occupying initial position in a clause, though it also can occur in medial or final
position. When not in initial position, @+username is more likely to indicate a
reference to a particular user rather than to explicitly inscribe a direct address. For
example:
Tried to make coffee this am..complete user malfunction – coffee ALL over the kitchen
counter..leaving that job to @User2 from now on
4
All usernames in this chapter have been anonymized.

Authenticated
This tweet is not directly addressing ‘User2’ and instead indirectly refers to this
user with what is termed a ‘mention’.
Mentioning a user with the @ character in this way is a kind of ‘amplified’
reference and potential tool for self-promotion since, depending on privacy set-
tings and the evolving functionality of Twitter, other users who follow this user
may view the mention. The @mention is also amplified in the sense that the @
character is searchable. Mentions can be aggregated and other users can search
for particular instances. It is possible to retrieve all instances of @mentions to a
given user (within a particular time window) with the Twitter search interface or
using metadata and the Twitter API. Twitter is, however, continually modifying
how it deals with @mentions in terms of who will see a conversational exchange
in their feed (e.g. only direct followers of one of the participants), presumably in
an attempt to predict the interests/attention span of its users.
Honeycutt and Herring (2009: 4) provide an overview of the various uses of
the @ character throughout digital communication (examples added):
Addressivity
@User3 I wish I could bring you bagels and coffee. Thinking of you.
Reference
Coffee and pseudo-dinner with @User4 lol. Work talk, makes you think. Change can
go both ways.
Emoticons
Will have a cup of coffee and cookies for breakfast then I’ll be fixing myself. Lazy
bum. @_@
Email
user@email.com
Locational ‘at’
Bought pear balsamic @ the Oilerie. Eating @ door county coffee co. Eyeing up hub-
by’s Reuben.
Non-locational ‘at’
i kno my co-wrkr jst bodied a red bull, a mountain dew, aaand a coffee all @ once …
he goes hard
Other
Forgot all about instant coffee, JEBUS how do they make that cr@! ?
Taking into account current Twitter usage, the @ character seems to be trans-
formed into an increasingly interpersonal resource. This follows a general trend in
the evolution of punctuation identified by Knox (2009), namely an evolution from
textual functions toward more interpersonal functions, and toward the service of
social affiliation (Zappavigna 2015).

Authenticated
Twitter 211
4.2. Retweeting
Another way of bringing external voices into a tweet is to republish another user’s
tweet within your own tweet. This is known as ‘retweeting’ and is usually marked
by the initialism, RT, to indicate that the body of the tweet is quoted text. In other
words, RT marks grammatical projection, economically standing for ‘User X has
posted the following’. In most instances RT will be followed by the @ character to
attribute the retweeted text to its original author, @User1 in the below:
RT @User1: Packed up and heading to Lake Tahoe. Need. Coffee.
Sometimes retweets are used in a similar way to the reply function:

Of course NOT I’m not expecting miracles LOL! RT @UserA: @UserB what you
kicked the coffee habit ? lol
Ha ha yes come to San Diego and I’ll pay for it :] RT @ddlovato: Power through,
power through, power though.. coffee anybody?! haha
The way that Twitter handles retweeting has evolved with the functionality of
the service, from manual retweets (RT @User, as in the above tweet), to native
retweeting by clicking an icon present at the bottom of each tweet (when accessed
via the web interface) (see Figure 4).
Retweeting “allows members to relay or forward a tweet through their net-
work” (Nagarajan, Purohit and Sheth 2010: 295), marking the quoted text as nota-
ble and effectively recommending it to their followers. It can significantly amplify
the reach of a tweet, particularly when a user with a large body of followers, such
as a celebrity, chooses to retweet something. Beyond rebroadcasting, retweeting
“contributes to a conversational ecology in which conversations are composed of a
public interplay of voices that give rise to an emotional sense of shared conversa-
tional context” (boyd, Golder and Lotan 2010: 1). The convention marks a tweet as
worth of the attention within this conversational context and allows the retweeter
to display a stance toward the retweeted text and project it as valuable to the com-
munity, whether in terms of merit or notoriety. For example, the underlined text in
the following has been appended to the original tweet by the retweeter:
Love this! RT @User: Just paid for the couple behind us coffee at starbucks. I encour-
age you all to pay it forward too.
This kind of evaluative appendage has also been noted by Page (2012b) who,
approaching media from the perspective of narrative, has suggested the role of
retweeting in new ‘co-tellership practices’ and noted the tendency of celebrities to
append evaluative assessment to their retweets as a means of aligning with their
audience.
Some studies have explored how often posts are retweeted, for instance as
a measure of success in activism (Potts et al. 2014: 66). While retweeting is an

Authenticated
important way to expand the reach of a tweet to new audiences, it also can be
part of establishing and maintaining social relationships (Boyd, Golder and Lotan
2010). Puschmann (2015) claims that community structure can, to varying extents,
be predicted by the social use of quotation. His study identifies four functions of
retweeting: passing on information, commenting or responding, presenting one’s
own interests, and building social capital (Puschmann 2015).
4.3. Hashtagging and ‘searchable talk’

The drive to make our opinions and emotions readily findable has become an
important social preoccupation, realized in social media as what Zappavigna
(2012; 2015) refers to as ‘searchable talk’, discourse that relies on forms of social
tagging to create alignments with potential audiences. Social tagging is the prac-
tice of appending user-generated metadata (often referred to as social metadata)
to social media texts. An early example is Flickr tags for annotating photos (Bar-
ton 2015). Social tagging has been described as “conversational tagging” (Huang,
Thornton and Efthimiadis 2010) since, in microblogging environments at least,
tags often have a function beyond classification through taxonomies or folkson-
omies5 and facilitate forms of mass ‘conversation’ since “the ‘globally public by
default’ nature of tweets lends itself to the development of means for automatically
organising discussions of specific topics through shared conversation markers”
(Bruns 2012: 1324). Tags can be conversational in the sense that they help users
to engage with topics receiving broad interest, and also, more locally, to produce
interpersonal meanings through forms of metacommentary and linguistic play, as
we will see later in this section.
A dominant form of social metadata is the hashtag, which has received exten-
sive attention in social media research, both due to the interesting new commu-
nicative affordances it has offered users, and because it provides researchers with
5
The kind of collaborative tagging evolving with community use in social media is often
referred to as a practice of ‘folksonomy’ (Vander Wal 2007) or social tagging very dif-
ferent to the top-down hierarchical approaches of traditional document classification.
Whereas document classification involves experts, social tagging engages communities
of general users. For example, it is used heavily on photo sharing sites such as Flickr
where it functions as a cooperative form of verbal indexing involving a ‘bottom-up’
approach to the kind of classification previously achieved by reference librarians.
Indeed, hashtags have been likened to the concept “better known to librarians as a
subject heading” (Ovadia 2009: 203). The tags assigned provide “access to the reader’s
view of aboutness in a way which was previously possible only on a small scale through
elicitation experiments” (Kehoe and Gee 2011).

Authenticated
Twitter 213
an efficient means6 of collecting discourse about a particular topic, person or

event.7 This also includes studies of the veracity and reliability of information
annotated with hashtags in relation to the problems that can be caused by various
kinds of inconsistencies in formats and conventions, the tendency of tags to prolif-
erate, become redundant, or be misused and hijacked. Linguistic studies of hash-
tags have been studied in relation to a wide range of contexts including transla-
tion (Carter, Tsagkias and Weerkamp 2011), inferential processes in reading (Scott
2015), self-branding and microcelebrity (Page 2012b), bullying discourse (Calvin
et al. 2015), digital libraries (Schlesselman-Tarango 2013), and political memes
(Zhu 2015). Zappavigna (2015) explores the different linguistic functions that can
be employed using a hashtag, noting their linguistic flexibility and tendency to
be employed interpersonally in the service of social affiliation through forms of
metacommentary. Similarly, adopting a pragmatic perspective, Scott (2015) and
Wikström (2014) have noted that hashtags perform significant functions beyond
facilitating search, and “have been appropriated by users to perform other roles in
the communicative process” (Scott 2015: 19).
As the name implies, a hashtag is prefixed with a hash symbol marking the
label appended to the tweet. This symbol may be followed by a keyword or con-
catenated phrase or clause, as in the following examples:
I have a feeling about today, it is either going to take lots of #coffee or #beer ?
French Press coffee is the jam. #ilovecoffee
The hashtag emerged through community-use on Twitter and the concept may
derive from Internet Relay Chat (IRC) conventions for naming channels (#chan-
nelname), where a channel is the essential mechanism that people use to communi-
cate with each other during an IRC session. Hashtagging is not restricted to Twitter
and occurs across a range of social media platforms, including image-focused ser-
vices such as Instagram (Highfield and Leaver 2014), notorious for the tendency
for users to employ excessive numbers of tags, for instance a picture of a cup of
coffee accompanied by the following tags:
6
A limitation of such hashtag-based research is that it does not necessarily capture all
the communication surrounding the primary media, since there will be posts that do not
use the hashtag, and, in particular, reply-posts may omit the tag. However, hashtags do,
at the very least, make obtaining a snapshot of particular discourse at a particular time
achievable.
7
For example, see work on election hashtags such as studies of the Australian #auspol
and #ausvotes hashtags (Bruns, Burgess and Highfield 2014; Sauter and Bruns 2015;
Zappavigna 2014).

Authenticated
#coffee #mocha #soymocha #skittlelane #skittlelanecoffee #coffeetime #morningcoffee

#coffeelovers #coffeelove #cafe #sydneycafe #sydneycafes #sydneyfoodie #sgfoodies
#coffeeadventure #cafehopping #mytaveldiary #mytravelgram #visualdiary
While metadata has a long history in the domain of information management,
this is the first historical period where we see it so closely tied to enacting social
relations, having extended its semiotic reach as an information-organising tool to a
social resource for building relationships and communities (Zappavigna 2015). As
Barton (2015: 64) has noted, “[t]here are many things going on in tagging spaces
and it is not just about taxonomies, nor just about folksonomies”. Annotating posts
with hashtags is a form of linguistic innovation (Cunha et al. 2011) and is unfold-
ing as a means of expressing more than just the ‘aboutness’ of a social media text
(Kehoe and Gee 2011). In a study comparing celebrity and ordinary users, Page
(2012b: 187) classifies hashtags into two functional types: topic hashtags cate-
gorising a tweet “in the manner of folksonomic tagging” and evaluative hashtags
“expressing an evaluative sentiment”. Evaluative hashtags are associated with the
broader practice of using social tags for metacommentary, often for humorous
effect as the contemporary equivalent of postmodern quotation marks for signal-
ling ironic or self-conscious distance (Scheible 2015).
An important function which hashtags have evolved to perform, beyond clas-
sification, is construing forms of metacommentary, particularly evaluative meta-
commentary. Some studies have drawn on the pragmatic insights offered by speech
act theory, initially developed to understand face to face communication:
[…] meta-comment tags may be understood in terms of hedging, disclaiming and man-
aging face, through the exploitation or flouting of maxims. The parenthetical explana-
tions are analyzed as providing background information which is sometimes crucial to
clarifying utterance force, but other times supplemental. The emotive and emphatic tags
are analyzed mostly in terms of how they strengthen or change the illocutionary force of
utterances, often in a manner reminiscent of the work done by non-verbal cues in face-
to-face conversation. The humorous and playful uses of hashtags can be understood in
terms of maxim-flouting and the exploitation of background knowledge. (Wikström
2014: 149–50)
Adopting this type of perspective, Wikström (2014: 130) argues that hashtags can
be grouped into a range of types (examples added):
– topic tags, where the topic is designated by the hashtag:
I hate being ‘twired’ where I am awake because of stimulants, but actually tired. #coffee
#sleep
– hashtag games, similar to topic hashtags in terms of categorising function, but
with the aim of participating in a social game:
When I say I’m great at customer service, I mean I can fake a great smile and only
scream on the inside. #BaristaProblems

Authenticated
Twitter 215
I hate every last one of you. Every. last. One. #baristaproblems

Stop ordering fraps. #baristaproblems
– meta-comments, where the hashtag makes a comment on the content of the

tweet itself rather than creating an association with other tweets:
yeah i order a small coffee at the daily drip but grab a large cup, so what? #idont-
giveafuck
– parenthetical explanations/additions, where the hashtag adds information
explaining the tweet:
I’m in a pub at 10:20 on a Friday am. Today will be a good day haha #drinkingcoffee
– emotive usage, where the hashtag supports emotional expression that might
otherwise be realised through some form or paralinguistic cue:
@4james When I worked in offices, I felt like I was the ONLY ONE who ever made
coffee. #sigh #ifeelyourpainbrother
– emphatic usage, where the hashtag realises some form of intensification:
WTF! The coffee machine is broken! I can’t work under these circumstances. #wtfwtf-
wtfwtfwtfwtfwtfwtfwtfwtf
– humorous and playful usage, including hashtags that support some form of
joke structure, hyperbole, or self-conscious humourous self-reference, for
example through excessive hashtagging:
Guys my coffee cup is empty WHAT SHOULD I DO #ifiwasonaforum #thiswould-
beanentirewebcomic #iamnotkidding #thishappens #ohno #theinternet
– memes and popular culture references, where knowledge of a particular cul-
tural trend is needed to interpret the tag. An example explored by Zappavigna
(2012) is employing the fail meme as a kind of humourous self-deprecation
no water nor coffee, me thirsty and sleepy #fail
Some studies have noted that hashtags have the additional function of forming
communities (Yang et al. 2012) or ‘publics’ (Bruns and Burgess 2011), often
through supporting visibility and participation (Page 2012b). Lin et al. (2013: 1–2)
propose that hashtags “are more than labels for contextualizing statements, objects
for bookmarking, or channels for sharing information, but they are active virtual
sites for constructing communities”, where they are able to be used to monitor
“on-going conversations” and “communicate non-verbal cues such as irony”. Just
how to explore the linguistic realization of ambient communities is a key and
emerging area of current research.

Authenticated
5. Conclusion: towards understanding ambient affiliation
This chapter has provided an overview of linguistic research into microblogging

and Twitter. The studies surveyed have explored how experience is construed in
microblogging texts, drawing on a range of linguistic perspectives. However, this
is only part of the picture. While most studies make some reference to the cen-
tral function that forging community plays in microblogging, systematic linguis-
tic models of affiliation are still in development. There has been a long-standing
interest both within and outside linguists in the shared communicative strategies
that might characterize communities of language users, variously theorized as
‘discourse communities’ (Bizzell 1992) and ‘communities of practice’ (Wenger
1998).8 Language is resplendent with resources for negotiating community, such
as naming practices, slang, and all kinds of variation across phonology, grammar,
and semantics that characterise different personae and groups (Martin 2010). In
Chapter 16 (this volume) on evaluation, I expand some of the perspectives intro-
duced here to account for the important role that evaluative language has in shar-
ing feeling and forging social alignments in social media discourse.
The challenge for social media researchers is to model how microbloggers
discursively negotiate their communal identites through sharing ‘bonds’ that con-
situte the value sets of communities and cultures (Knight 2010). In other words
this is a challenge of understanding what Zappavigna (2011, 2012, 2014a, 2014b)
terms ‘ambient affiliation’: social bonding where microbloggers as individuals do
not necessarily have to interact directly, but may engage in mass practices such as
hashtagging in order to participate in particular kinds of ‘belonging’. As an object
of study, microblogging offers a rare opportunity to the linguist to track the par-
ticular configurations of shared social bonds as they emerge and unfold. In part,
this is because the brevity enforced by the medium goes some very small way to
reducing complexity of the endeavour, and because the affordances of the channel
mean that we can collect large volumes of textual interactions that give us insight
into both individual and communal linguistic disposition.9 This gives us some
ability to gain a window on what Firth has referred to as the “general language
of the community” that governs each person’s “command of a constellation of
restricted languages” (Firth 1968: 207) (see also Zappavigna, Ch. 16, this volume).
8
For an overview of these concepts see Prior (2003).
9
This disposition is informed by a persona’s particular semiotic ‘repertoire’ that arises
out of the potential semiotic ‘reservoir’ available via their membership in a given com-
munity (Bernstein 2000).

Authenticated
Twitter 217
References
Adnan, Muhammad, Paul Longley and Shariq M. Khan

2014 Social dynamics of Twitter usage in London, Paris, and New York City. First
Monday 19(5). Available at http://firstmonday.org/ojs/index.php/fm/article/
view/4820/4083.
2014 Moments of sharing: Entextualization and linguistic repertoires in social net-
working. Journal of Pragmatics 73: 4–18.
Artemeva, Natasha
2015 Genre and identity in social media. In: Natasha Artemeva and Aviva Freed-
man (eds.), Genres Studies Around the Globe: Beyond the Three Traditions,
275–298. Edmonton, AB: inkshed Publications.
Baker, Paul and Anthony McEnery
2015 Who benefits when discourse gets gemocratised? Analysing a Twitter cor-
pus around the British Benefits Street debate. In: Paul Baker and Anthony
McEnery (eds.), Corpora and Discourse Studies: Integrating Discourse and
Corpora, 244–265. London: Palgrave Macmillan. http://doi.org/10.1057/
9781137431738_12.
Bernstein, Basil
2000 Pedagogy, Symbolic Control and Identity: Theory, Research, Critique. Lon-
don: Taylor and Francis.
Bizzell, Patricia
1992 What is a Discourse Community. In: Patricia Bizzell, Academic Discourse and
Critical Consciousness, 222–237. Pittsburgh: University of Pittsburgh Press.
Baron, Naomi., Lauren Squires, Sara Tench and Marshall Thompson
2005 Tethered or mobile? Use of away messages in instant messaging by American
college students. In: Rich Ling and Per Pedersen (eds.), Mobile Communica-
tions: Re-negotiation of the Social Sphere, 293–311. London: Springer.
Barton, David
2015 Tagging on Flickr as a social practice. In: Rodney H. Jones, Alice Chik and
Christoph A. Hafner (eds.), Discourse and Digital Practices: Doing Discourse
Analysis in the Digital Age, 48–65. London: Routledge.
Bamman, David, Jacob Eisenstein and Tyler Schnoebelen
2014 Gender identity and lexical variation in social media. Journal of Sociolinguis-
tics 18(2): 135–160.
Baym, Nancy and danah boyd
2012 Socially mediated publicness: An introduction. Journal of Broadcasting and
Bolander, Brook and Miriam Locher
2014 Doing sociolinguistic research on computer-mediated data: A review of four
methodological issues. Discourse, Context and Media 3(1): 14–26.
boyd, danah
2011 Social network sites as networked publics: Affordances, dynamics, and impli-
cations. In: Zizi Papacharissi (ed.), A Networked Self: Identity, Community,
and Culture on Social Network Sites, 39–58. Oxford: Routledge.

Authenticated
boyd, danah, Scott Golder and Gilad Lotan

2010 Tweet, tweet, retweet: Conversational aspects of retweeting on Twitter. Pro-
ceedings of the 43rd Annual Hawaii International Conference on System
Sciences, 1–10. Available at http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnum-
ber=5428313.
Bruns, Axel
2012 How long is a tweet? Mapping dynamic conversation networks in Twitter using
Gawk and Gephi. Information, Communication and Society 15(9): 1323–1351.
Bruns, Axel and Jean Burgess
2011 The use of Twitter hashtags in the formation of ad hoc publics. Paper pre-
sented at the European Consortium for Political Research Conference, Rey-
kjavik, 25–27. Available at http://eprints.qut.edu.au/46515/
Bruns, Axel, Jean Burgess and Tim Highfield
2014 A “Big Data” approach to mapping the Australian twittersphere. In: Paul
Longley Arthur and Katherine Bode (eds.), Advancing Digital Humanities:
Research, Methods, Theories, 113–129. London: Palgrave Macmillan.
Bruns, Axel, Gunn Enli, Eli Skogerbo, Anders Olof Larsson and Christian Christensen
(eds.)
2015 The Routledge Companion to Social Media and Politics. New York/London:
Routledge.
Bruns, Axel and Hallvard Moe
2014 Structural layers of communication on Twitter. In: Katrin Weller, Axel Bruns,
Jean Burgess, Merja Mahrt and Cornelius Puschmann (eds.), Twitter and Soci-
ety, 15–28. New York: Peter Lang.
Burgess, Jean
2014 From “Broadcast yourself” to “Follow your interests”: Making over social
media. International Journal of Cultural Studies 18(3): 281–285.
Calvin, Angela, Amy Bellmore, Jun-Ming Xu and Xiaojin Zhu
2015 #bully: Uses of hashtags in posts about bullying on Twitter. Journal of School
Violence 14(1): 133–153.
Carter, Simon, Manos Tsagkias and Wouter Weerkamp
2011 Twitter hashtags : Joint translation and clustering. Proceedings of the ACM
WebSci’11, June 14–17 2011, Koblenz, Germany, 1–3. Available at http://
www.websci11.org/fileadmin/websci/Posters/125_paper.pdf.
Cole-Lewis, Heather, Jillian Pugatch, Amy Sanders, Arun Varghese, Susana Posada, Chris-
topher Yun, Mary Schwarz and Erik Augustson
2015 Social listening: A content analysis of e-cigarette discussions on Twitter. Jour-
nal of Medical Internet Research 17(10). doi: 10.2196/jmir.4969.
Cunha, Evandro, Gagno Magno, Giovanni Comarela, Virgilio Almeida, Marcos André
Gonçalves and Fabricio Benevenuto
2011 Analyzing the dynamic evolution of hashtags on Twitter: A language-based
approach. Proceedings of the Workshop on Language in Social Media LSM
2011, 58–65. Stroudsburg, PA: Association for Computational Linguistics.
Dayter, Daria
2014 Self-praise in microblogging. Journal of Pragmatics 61: 91–102.
Dayter, Daria
2015 Small stories and extended narratives on Twitter. Discourse, Context and
Media 10: 19–26.

Authenticated
Twitter 219
Firth, John Rupert

1968 Selected Papers of J. R. Firth, 1952–1959. Bloomington: Indiana University
Press.
Giles, David, Wyke Stommel, Trena Paulus, Jessica Lester, and Darren Reed
2014 Microanalysis of online data: The methodological development of “Digital
CA.” Discourse, Context and Media 7: 45–51.
Holotescu, Carmen and Gabriela Grosseck
2009 Indicators for the analysis of learning and practice communities from the per-
spective of microblogging as a provocative sociolect in virtual space. Con-
ference Proceedings of eLearning and Software for Education (eLSE) 01:
163–174. Bucharest: Universitatea Nationala de Aparare Carol I.
Hambrick, Marion, Jason Simmons, Greg Greenhalgh and Christopher Greenwell
2010 Understanding professional athletes’ use of Twitter: A content analysis of ath-
lete tweets. International Journal of Sport Communication 3(4): 454–471.
Hardaker, Claire and Mark McGlashan
2016 “Real men don’t hate women”: Twitter rape threats and group identity. Journal
Highfield, Tim
2015 Tweeted joke lifespans and appropriated punch lines: Practices around topi-
cal humor on social media. International Journal of Communication 9(22):
2713–2734.
Highfield, Tim and Tama Leaver
2014 A methodology for mapping Instagram hashtags. First Monday 20(1).
Available at http://www.firstmonday.dk/ojs/index.php/fm/article/view/5563/
4195.
Highfield, Tim and Tama Leaver
2016 Instagrammatics and digital methods: Studying visual social media, from
selfies and GIFs to memes and emoji. Communication Research and Practice
2(1): 47–62.
Highfield, Tim, Stephen Harrington and Axel Bruns
2013 Twitter as a technology for audiencing and fandom: The #Eurovision phenom-
enon. Information, Communication and Society 16(3): 315–339.
Hogan, Bernie and Anabel Quan-Haase
2010 Persistence and change in social media. Bulletin of Science, Technology and
Society 30 (5): 309–315.
Honeycutt, Courtenay and Susan Herring
2009 Beyond microblogging: Conversation and collaboration via Twitter. Proceed-
ings of the 42nd Annual Hawai’i International Conference on System Sciences,
1–10. doi: 10.1109/HICSS. 2009.89.
Huang, Jeff, Katherine Thornton and Efthimis Efthimiadis
2010 Conversational tagging in Twitter. Proceedings of the 21st ACM Conference
on Hypertext and Hypermedia – HT ’10 ACM: 73–178. doi: 10.1145/1810617.
1810647.
Humphreys, Lee, Phillipa Gill, Balachander Krishnamurthy and Elizabeth Newbury
2013 Historicizing new media: A content analysis of Twitter. Journal of Communi-
cation 63(3): 413–431.

Authenticated
Java, Ashkay, Xiaodan Song, Tim Finin and Belle Tseng

2009 Why we twitter: An analysis of a microblogging community. Lecture Notes in
Computer Science (Including Subseries Lecture Notes in Artificial Intelligence
and Lecture Notes in Bioinformatics) Vol. 5439 LNAI: 118–138. Berlin/Hei-
delberg: Springer. doi: 10.1007/978-3-642-00528-2_7.
Kehoe, Andrew and Matt Gee
2011 Social tagging: A new perspective on textual ‘aboutness’. In: Paul Rayson,
Sebastian Hoffmann and Geoffrey Leech (eds.), Studies in Variation, Contacts
and Change in English. Volume 6: Methodological and Historical Dimensions
of Corpus Linguistics, Helsinki: University of Helsinki. Available at http://
www.helsinki.fi/varieng/series/volumes/06/kehoe_gee/.
Kelsey, Darren and Lucy Bennett
2014 Discipline and resistance on social media: Discourse, power and context in
the Paul Chambers “Twitter Joke Trial”. Discourse, Context and Media 3(1):
37–45.
Knight, Naomi
2010 Wrinkling complexity: Concepts of identity and affiliation in humour. In:
Knox, John
2009 Punctuating the home page: Image as language in an online newspaper. Dis-
course and Communication 3(2): 19–53.
Kristeva, Julia
1980 The bounded text. In: Leon S. Roudiez (ed.), Desire in Language: A Semiotic
Approach to Literature and Art, 36–63. New York: Columbia University Press.
Kytölä, Samu and Elina Westinen
2015 “I be da reel gansta” – A Finnish footballer’s Twitter writing and metaprag-
matic evaluations of authenticity. Discourse, Context and Media 8: 6–19.
1967 Narrative analysis. In: June Helm (ed.), Essays on the Verbal and Visual Arts,
12–44. Seattle: University of Washington Press.
Lee, Carmen
2011 Micro-blogging and status updates on Facebook: Texts and practices. In:
Crispin Thurlow and Kristine Mroczek (eds.), Digital Discourse: Language in
the New Media, 110–128. Oxford: Oxford University Press.
Lin, Yu-Ru, Drew Margolin, Brian Keegan, Andrea Baronchelli and David Lazer
2013 #Bigbirds never die: Understanding social dynamics of emergent hashtag.
Proceedings of the 7th International AAAI Conference on Weblogs and Social
Media (ICWSM 2013). Available at https://arxiv.org/abs/1303.7144.
Martin, James
2010 Semantic variation: Modelling system, text and affiliation in social semiosis.
In: Monika Bednarek and James Martin (eds.), New Discourse on Language:
London: Continuum.
Martin, James, Michele Zappavigna, Paul Dwyer and Chris Cléirigh
2013 Users in uses of language: Embodied identity in youth justice conferencing.
Text and Talk 33(4–5): 467–496.

Authenticated
Twitter 221
Miller, Vincent
2008 New media, networking and phatic culture. Convergence: The International
Journal of Research into New Media Technologies 14 (4): 387–400.
Murthy, Dhiraj
2013 Twitter: Social Communication in the Twitter Age. Cambridge: Polity Press.
Nagarajan, Meenakshi, Hemant Purohit and Amit Sheth
2010 A qualitative examination of topical tweet and retweet practices. Paper pre-
sented at the Fourth International AAAI Conference on Weblogs and Social
Media, Washington, D.C. Available at http://www.aaai.org/ocs/index.php/
ICWSM/ICWSM10/paper/download/1484/1880/.
Nastri, Jacqueline, Jorge Peña and Jeffrey Hancock
2006 The construction of away messages: A speech act analysis. Journal of Comput-
er-Mediated Communication 11(4): 1025–1045.
Neubig, Graham and Kevin Duh
2013 How much is said in a tweet? A multilingual, information-theoretic perspec-
tive. AAAI Spring Symposium: Analyzing Microtext, 32–39. Available at http://
www.phontron.com/paper/neubig13sam.pdf.
Ovadia, Steven
2009 Exploring the potential of Twitter as a research tool. Behavioral and Social
Sciences Librarian 28(4): 202–205.
Page, Ruth
2010 Re-examining narrativity: Small stories in status updates. Text and Talk 30(4):
423–444.
Page, Ruth
2012a Stories and Social Media: Identities and Interaction. London: Routledge.
Page, Ruth
2012b The linguistics of self-branding and micro-celebrity in Twitter: The role of
hashtags. Discourse and Communication 6(2): 181–201.
Page, Ruth
2014 Saying ‘sorry’: Corporate apologies posted on Twitter. Journal of Pragmatics
62: 30–45.
Page, Ruth, David Barton, Johann Wolfgang Unger and Michele Zappavigna
2014 Researching Language and Social Media: A Student Guide. London: Routledge.
Popescu, Ana-Maria and Marco Pennacchiotti
2010 Detecting controversial events from Twitter. Proceedings of the 19th ACM
International Conference on Information and Knowledge Management –
CIKM ’10, 1873–1876. doi: 10.1145/1871437.1871751.
Potts, Amanda, Will Simm, Jon Whittle and Johann Wolfgang Unger
2014 Exploring “success” in digitally augmented activism: A triangulated approach
to analyzing UK activist Twitter use. Discourse, Context and Media 6: 65–76.
Potts, Liza, Joyce Seitzinger, Dave Jones and Angela Harrison
2011 Tweeting disaster: Hashtag constructions and collisions. Proceedings of the
29th Annual International ACM SIGIR Conference on Research and Develop-
ment in Information Retrieval, 235–240. doi: 10.1145/2038476.2038522.
Prior, Paul
2003 Are communities of practice really an alternative to discourse communities.
Paper Presented at the Meeting of the American Association of Applied Lin-

Authenticated
guistics, Arlington, VA. Available at http://www.semremtoo.org/Prior/home/

PriorAAAL03.pdf.
2015 The form and function of quoting in digital media. Discourse, Context and
Media 7: 28–36.
Ritter, Alan, Colin Cherry and Bill Dolan
2010 Unsupervised modeling of Twitter conversations. In: Human Language Tech-
nologies: The 2010 Annual Conference of the North American Chapter of the
Association for Computational Linguistics, Los Angeles, California.
Sauter, Theresa and Axel Bruns
2015 #auspol: The hashtag as community, event, and material object for engaging
with Australian politics. In: Nathan Rambukkana (ed.), Hashtag Publics: The
Power and Politics of Discursive Networks, 47–60. New York: Peter Lang.
Scheible, Jeff
2015 Digital Shift: the Cultural Logic of Punctuation. Minneapolis: University of
Minnesota Press.
Schlesselman-Tarango, Gina
2013 Searchable signatures: Context and the struggle for recognition. Information
Technology and Libraries 32(3): 5–19.
Schmidt, Jan-Hinrik
2014 Twitter and the rise of personal publics. In: Katrin Weller, Axel Bruns, Jean
Burgess, Merja Mahrt and Cornelius Puschmann (eds.), Twitter and Society,
Scott, Kate
2015 The pragmatics of hashtags: Inference and conversational style on Twitter.
Seargeant, Philip and Caroline Tagg (eds.)
2014 Language and Social Media: Communication and Community Online. Lon-
don: Palgrave.
Sharma, Sanjay
2013 Black Twitter?: Racial hashtags, networks and contagion. New Formations: A
Journal of Culture/Theory/Politics 78(1): 46–64.
Shaw, Frances, Jean Burgess, Kate Crawford and Axel Bruns
2013 Sharing news, making sense, saying thanks: Patterns of talk on Twitter dur-
ing the Queensland floods. Australian Journal of Communication.40 (1): 23–
40.
Sifianou, Maria
2015 Conceptualizing politeness in Greek: Evidence from Twitter corpora. Journal
Starbird, Kate, Leysia Palen, Amanda Hughes and Sarah Vieweg
2010 Chatter on the red: What hazards threat reveals about the social life of
microblogged information. Proceedings of the 2010 ACM Conference on Com-
puter-supported Cooperative Work, 241–250. doi: 10.1145/1718918.1718965.
Tagg, Caroline
Vander Wal, Thomas
2007 Folksonomy. Available at http://vanderwal.net/folksonomy.html.

Authenticated
Twitter 223
Weller, Katrin, Axel Bruns, Jean Burgess, Merja Mahrt and Cornelius Puschmann
2013 Twitter and society: An introduction. In: Katrin Weller, Axel Bruns, Jean Bur-
gess, Merja Mahrt and Cornelius Puschmann (eds.), Twitter and Society, xxix–
xxxviii. New York: Peter Lang.
Wenger, Etienne
1998 Communities of Practice: Learning, Meaning, and Identity. Cambridge: Cam-
West, Laura
2013 Facebook sharing: A sociolinguistic analysis of computer-mediated storytell-
ing. Discourse, Context and Media 2(1): 1–13.
Wikström, Peter
2013 & she was like “O_O”: Animation of reported speech on Twitter. Nordic Jour-
nal of English Studies 13(3): 83–111.
Wikström, Peter
2014 #srynotfunny : Communicative functions of hashtags on Twitter. SKY Journal
of Linguistics 27: 127–152.
Yang, Lei, Tao Sun, Ming Zhang and Qiaozhu Mei
2012 We know what@ you# tag: Does the dual role affect hashtag adoption? Pro-
ceedings of the 21st International Conference on World Wide Web, Lyon,
16–20 April 2012, 261–270. doi: 10.1145/2187836.2187872.
Zappavigna, Michele
ety 13(5): 788–806.
Zappavigna, Michele
Zappavigna, Michele
2013 Coffee tweets: Bonding around the bean on Twitter. In: Philip Seargeant and
Caroline Tagg (eds.), The Language of Social Media: Communication and
Community on the Internet, 137–160. London: Palgrave Macmillan.
Zappavigna, Michele
2014a Ambient affiliation in microblogging: Bonding around the quotidian. Media
Zappavigna, Michele
2014b Enacting identity in microblogging through ambient affiliation. Discourse and
Communication 8(2): 209–228.
Zappavigna, Michele
2014c Enjoy your snags Australia … oh and the voting thing too #ausvotes #auspol:
Iconisation and affiliation in electoral microblogging. Global Media Journal:
Australian Edition 8(2): 1–18.
Zappavigna, Michele
2015 Searchable talk: The linguistic functions of hashtags. Social Semiotics 25(3):
274–291.
Zhao, Dejin and Mary Beth Rosson
2009 How and why people Twitter: The role that micro-blogging plays in informal
communication at work. Proceedings of the ACM 2009 International Con-
ference on Supporting Group Work, 243–252. Sanibel Island, FL: ACM. doi:
10.1145/1531674.1531710.

Authenticated
Zhu, Hongqiang
2015 Searchable talk as discourse practice on the Internet : The case of “#binders-
fullofwomen”. Discourse, Context and Media 12: 87–98

Authenticated
9. Social Network Sites/Facebook
Volker Eisenlauer
Abstract: The omnipresence of computers and mobile devices in people’s daily

lives together with the persistent access to Social Network Sites provide ample
opportunities for users to communicate and interact in new ways. This chapter
provides a critical review of the most important pragmatic perspectives on Social
Network Sites/Facebook. Drawing on the concept of form of communication as
well as on established approaches in Social Media research, I will propose a cate-
gorisation scheme for assessing different types of Social Network Sites. The appli-
cation of fundamental concepts in pragmatics and discourse analysis to Facebook
settings (such as context and speech act theory) provides access to the complex
interlacing between text creation and text automation practices that constitute and
constrain member profiles.
1. Introducing Social Network Sites
Social Network Sites (SNS) are increasingly integrated into the daily practices of
users around the world. Supporting people in all sorts of off- and online activities
(such as planning everyday lives, job hunting, traveling, content sharing) these
services facilitate various communicative practices while supporting the forma-
tion and maintenance of social ties of all kinds. The emergence of SNS as a novel
online service dates back to the late 1990s and originates in the combination of
two preceding services, i.e. dating websites and instant messaging. Just like dating
websites, SNS afford the setup of a personal profile, and, comparable to instant
messaging services, SNS support users in building up interpersonal connections to
those they already know. According to boyd and Elison (2007), SixDegrees.com
was one of the first services that combined these affordances. Founded in 1997
and shut down in 2001, SixDegrees.com offered many operational features that
have by now become essential properties of cutting-edge SNS. It was followed by
numerous other services which linked community building to a large variety of
social and discursive practices, for example data sharing (YouTube), professional
networking (LinkedIn), and cultivating friendships (Facebook).
The emergence of the Internet was, from the very beginning, connected to
the exploitation of new social and communicative spaces. Long before the rise of
social media and SNS, users have been exchanging information with the help of
computer technologies while forming, maintaining, and displaying social connec-
tions. As early as 1985, the Whole Earth ’Lectronic Link (WELL) offered early
DOI 10.1515/9783110431070-009
Authenticated
226 Volker Eisenlauer
Internet adopters a virtual space for exchanging ideas emerging in one of the first
online communities (Ebersbach, Glaser and Heigl 2008: 22). Henceforward, peo-
ple have been utilizing the Internet and its various services, i.e. bulletin board
systems, mailing lists, and news groups, to form all kinds of online communities
around a variety of subjects and interests.
From the mid-1990s onwards, new technologies and services, such as multi-
user online games, chat rooms, or message boards, provided new means for online
self-presentation and networking with other users. Recognizing the user’s desire
for identity construction in online worlds, the web-hosting service GeoCities intro-
duced the idea of free server space, as well as website construction kits to the
online community. From 1995 onwards, their platform was one of the first Internet
services that enabled technologically inexperienced users to create and publish
private homepages on the Internet.
Although GeoCities never enforced specific content, its preset modules and
layouts facilitated the presentation of topics in a consistent and compatible style
and structure. Similar to today’s SNS, the free web-hosting services of the mid-
1990s proved to be an easy-to-use, practical tool for self-presentation and social
connectivity. Their creation of a new technological and communicative niche on
the Internet soon turned free web-hosting services into a veritable economic suc-
cess story, with an estimated value of millions of dollars (Knoke 2009). Their
considerable initial market value notwithstanding, the services would not generate
large revenues, unlike other major SNS such as Facebook, Myspace or Google+.
SNS frequently reflect and display social links of previously existing offline
communities (such as family, friends, neighbors and colleagues) in a much greater
variety than message boards, chat rooms, discussion fora and other early online
community tools. This is largely because in SNS, participants “are primarily com-
municating with people who are already part of their extended social network”
(boyd and Ellison 2007). While there are also specific services that are particularly
designed for meeting new people (cf. Meetup), the majority of SNS are primarily
used to manage and maintain contacts with people who already know each other.
Accordingly, Yus (2011: 115–116) argues that one of the key functions of SNS is
“the maintenance of interpersonal relationships that already exist in physical set-
tings, that is, a role of extension of physical relationships in to the virtual realm”.
With online identities being increasingly anchored in offline contexts, the ways
in which online communities are negotiated and perceived have changed dramati-
cally in recent years. Users increasingly interact with offline acquaintances, while
anonymous online activities seem to be on the decline. Thus, online communities
usually reproduce existing offline communities and are firstly built around previ-
ously existing offline connections and only secondly around interests. In addition,
users increasingly understand the Internet “as a tool for presenting and promoting
their ‘real selves’ rather than for taking anonymous action” (Jones and Hafner
2012: 145).

Authenticated
Social Network Sites/Facebook 227
Although SNS constitute a rather specific object of study, online services have
gained considerable research interest across various disciplines. Several studies
are interested in different aspects of (discursive) online practices and explore the
formation of social ties via and within SNS from various research perspectives
and backgrounds, such as communication studies (Baym 2007), anthropology
(boyd 2008), sociology (Androutsopoulos et al. 2013), economics (Heidemann et
al. 2011), political science (Robertson, Vatrapu and Medina 2010), and computer
science (Zhang, Liu and Xu 2015). Depending on the specific research questions
and the kind of data used, these studies emphasize different criteria that constitute
SNS, among them expert search and context awareness (Richter and Koch 2007),
personalizable spaces (Selwyn 2009) or product reviews, and professional profil-
ing (Adedoyin-Olowe, Gaber and Stahl 2013).
Two of the most central practices afforded by most SNS are setting up personal
profiles and forming network contacts. Additionally, different types of SNS pro-
vide readymade solutions for a range of different communicative problems, such as
managing friendship ties (e.g. Facebook), building and maintaining career-related
networks (e.g. LinkedIn), or marketing and sharing research (e.g. ResearchGate).
Being a form of social media, SNS make use of software-supported text-genera-
tion tools that introduce non-expert users to online publishing and have brought
forward a variety of new action patterns and online genres.
Drawing on the concept of form of communication as well as on established
approaches in social media research (Section 2), this article first proposes a cate-
gorization scheme for assessing different types of SNS (Section 3). It then inves-
tigates language use on Facebook from a pragmatic perspective. Reviewing the
application of fundamental concepts in pragmatics and discourse analysis to Face-
book settings (such as context and speech act theory) will provide access to the
complex interlacing between text creation and text automation practices that con-
stitute and constrain member profiles (Section 4).
2. No longer media, not yet genres
Emerging generic and linguistic features particular to new online genres evolve
partly from the particular design of the sites but are likewise affected by the dom-
inant communities who colonize these new online spaces. Platform owners and
software developers commonly plan and implement a variety of features to target
a specific group of users. As we have seen in Section 1, SNS communities fre-
quently mirror offline relationships and engage in some type of common behavior
that may or may not correspond with the targeted practices of the developers.
For instance, launched in 2003, MySpace initially aimed to compete with exist-
ing friendship networks at that time, such as Friendster or MiGente.com. Within
a short time, unsigned rock bands used MySpace to connect and communicate

Authenticated
with their fans and turned the platform into one of the largest global music net-
works. Some years later, MySpace continued to grow as a popular “social network
among young teenagers until being overtaken by Facebook in 2008 in terms of
volume and popularity” (Thuraisingham et al. 2016: 18). It follows that SNS can
only be really understood by taking account of the space between the medium and
the participants. In the broadest sense, a medium can be understood “as a system
that makes a certain type of communication possible” (Posner 1986 in Threadgold
1997: 393). Unlike media serving as a technological means for sign production/
reception, SNS aim at and reinforce particular patterns of social actions. However,
SNS cannot be considered as genres in the sense of recurrent configurations of
meanings (Martin 2009; Hiippala 2014). Though SNS aim at specific discourse
communities and facilitate the emergence of particular rhetorical patterns, they
only provide the settings for individual networking practices. As shown for MyS-
pace, recurrent rhetorical actions have only partly evolved based on the particular
design of the SNS setting; they are likewise affected by whoever colonizes the pre-
set templates. In SNS settings, typified ways of interacting and their conventional
semiotic forms emerge in large parts from the discourse community of which the
users claim membership. To this effect, Lomborg (2011) contends that
[the] interposition of software between media platform and genre […] arguably adds a
new dimension to the adjustment and development of genres. Changes and adjustments
at the software level […] contribute to (re)shaping the communicative functions and
social purposes of a given genre […]. (Lomborg 2011: 60)
While SNS contribute to the production and processing of conventional and repeat-
able patterns of language use, they cannot be equated with genres but emerge in
so-called “forms of communication”. Coined by Ermert (1979), the term form of
communication (“Kommunikationsform”) highlights the situational and medial
conditions that precede any generic characteristics.1 Following Holly (2011), dif-
ferent forms of communication give rise to significant clusters of communicative
conditions and can be distinguished along three different contextual parameters:
(1) Modes and codes, which relate to the physical/ontological basis of the sign and
the sensory channels addressed (e.g. visual, auditory, olfactory, tactile, haptic);
(2) communicative space, which is characterized by different degrees of commu-
nicative copresence, reciprocity, and directionality/directedness; (3) temporality,
which accounts for the durability of signs and can be subdivided in transmitting
and storing forms of communication. According to Holly, “[we] can speak of forms
of communication whenever we can communicate particular signs in a particular
direction (one-way or two-way), with a particular range (public or private) and
1
“Forms of communication […] are not necessarily associated within one single medium.
A monological, written text may, for instance, be realized as a book, an inscription on a
stone, or as an electronic text” (Gruber 2008: 364).

Authenticated
a particular durability (transmitting or storing) in their specific patterns/arrange-

ments (Holly 2004: 2)”.2
In a similar manner, Dürscheid (2005) identifies forms of communication as
media constellations that configure the communicative setting and surface exclu-
sively in conditions external to the text itself. Dürscheid’s (2005) criteria for clas-
sifying forms of communication include the number of participants, the type of
signs, the direction of communication, and its spatio-temporal embedding.
The application of the aforementioned parameters to SNS helps to approximate
and conceptualize the ambivalent status of SNS, standing halfway between media
and user/text actions. In terms of modes and codes, SNS support the creation
and display of multimodal content. Most SNS request members to upload profile
photos. Depending on their scope and focus, some SNS offer enhanced options
for users to incorporate and share different semiotic resources by, for example,
promoting and recording sounds on SoundCloud, sharing video clips on Vine, or
embedding multimodal content stored on external servers on Facebook. Moreover,
SNS encourage the use of semiotic resources that are at least partially generated
by text-automation tools. When distributing content via and within SNS, most
posts are automatically extended by user names, profile photos, and time stamps.
As opposed to offline contexts in which production roles are exclusively adopted
by human agents (Goffmann 1981: 144), in SNS environments, text production
can be significantly supported, altered, outsourced, or even supplanted by soft-
ware-based tools (Eisenlauer 2014: 76). Consequently, any pragmatic theory of
SNS must examine not only the semiotic basis of user discourse but will ultimately
have to account for the human and/or nonhuman actants (Latour 2007: 123) that
are involved in selecting and aligning propositions.
The communicative space of SNS is being constituted by a vast array of net-
work-specific operations that promote bidirectional exchanges among network
members. As opposed to single-tool platforms, such as chat, discussion fora, or
blogs, SNS can be considered “multiple-tool platforms” (Jucker and Dürscheid
2012: 43) that offer a vast array of different services with specific medium-de-
pendent practices. For example, Facebook’s messenger service supports (quasi-)
synchronous and asynchronous user interactions among any number of friends
at one time, while the status update service promotes asynchronous network
exchanges and the notification service listed on Facebook results in a one-way
form of communication.
2
Translated from: „Bestimmte Kommunikationsformen liegen vor, wenn wir bestimmte
dieser Zeichen in bestimmter Direktionalität (einseitig ausstrahlend oder wechselsei-
tig), mit bestimmter Reichweite (privat oder öffentlich), mit bestimmter Haltbarkeit
(übertragend oder speichernd) in den jeweiligen Arrangements kommunizieren kön-
nen” (Holly 2004: 2).

Authenticated
With respect to temporality, SNS possess both transmitting and storing capac-
ities. They make possible communicative interaction and archiving by default
posts, comments, photos, and other user data. Holly’s (2004) binary classifica-
tion into forms of communication with private and public reach falls short of the
audience size that can be reached by SNS. The latter are forms of communica-
tion that support the formation of network sizes, ranging from a few members to
several thousand. They thus blur the distinction between private and mass media
(see “meso medium” in Zerdick et al. 2004). This intermediate position makes it
difficult for users to keep track of who and how many contacts are pursuing their
networking activities.
Even though the services are technically equipped to form and maintain large
networks, users face cognitive limits in maintaining and processing stable social
relationships. As findings in psychology show, humans are capable of sustaining
no more than 150 meaningful social relationships (Dunbar 1993). Users may com-
monly maintain connections to larger networks via and within SNS, such as Face-
book, but they often count far fewer members as actual friends (Ellison, Steinfield
and Lampe 2011). The software services make it easier for members to leverage
weak ties; the individuals whom members have met or with whom they share a
connection can be easily traced and recorded on SNS. Moreover, the various com-
municative tools on SNS lower users’ inhibition threshold when they contact loose
acquaintances (Losh 2008: 347).
To conclude, SNS are forms of communication that are both transmitting and
storing at the same time. They enable spatially separated participants to connect in
synchronous and asynchronous ways in semipublic contexts, and they promote the
performance of social actions in written form while triggering the software-sup-
ported creation, integration, and display of multimodal data.
3. A categorization scheme for social network sites
As we have seen, the concept of ‘form of communication’ unravels the subtle ways
in which SNS platforms emerge in the contextual embedding of user text actions.
Still, the notion is too broad to capture the distinctive characteristics that distin-
guish one SNS service from another. At present, the web offers the appropriate
SNS for every conceivable purpose, supporting all kinds of social relations and
practices. The omnipresence of smartphones has made SNS even more ubiquitous
and paramount in our daily lives and has favored the rise of novel mobile social
networking apps that are, in turn, based on specific mobile interaction patterns.
These apps provide users with constant and instant access to their SNS profiles
from their mobile devices. The large number of ever-growing SNS resources
together with the fluidity of their underlying software environments makes any
classification an open-ended venture.

Authenticated
Previous researchers have recognized the importance of categorizing the diverse

types of SNS and proposed different categorization schemes highlighting different
features that are specific to SNS. According to Fraser and Dutta (2008), SNS can
be framed along the nature and kind of networks that are created and maintained
via and within the different platform environments; they distinguish five broad
categories: 1) Egocentric networks, such as Facebook or Myspace, which serve
as platforms for identity construction and support the creation and administration
of friendship ties. 2) Community networks, such as Classmates or MyCatholicVil-
lage, which replicate existing offline communities online. Its members display
“strong identity linkages based on nation, race, religion, class, sexual orientation
and so on” (Fraser and Dutta 2008: 5). 3) Opportunistic networks, such as Xing
or LinkedIn, which support members in establishing online connections that serve
rational interests as, for example, in business connections. 4) Passion-centric net-
works, such as Dogster or CarDomain, which provide online spaces for people
who share interests, hobbies, and leisure time activities. 5) Media-sharing sites,
such as YouTube or Vimeo, that are used primarily to consume or distribute music,
films, and other media content. These online spaces are not defined “by their mem-
bership, but rather by their content” (Fraser and Dutta 2008: 6).
A different classification scheme is proposed by Heidemann (2009), who dis-
tinguishes open SNS, which deploy only minor access restrictions, from closed
SNS, which are limited to particular user communities. According to Gross and
Acquisti (2005), different services can be distinguished by the ways and extents
to which they support users in revealing personal information. Some SNS types
encourage the use of real names, while others seek to protect members’ personal
identity by supporting the use of nicknames. Different services vary dramatically
in regard to the type and extent of information elicited. While major SNS, among
them Facebook, ask for more general user information, such as hobbies and per-
sonal interests, niche SNS aim at specialized information on, for example, addic-
tion-related behavior on sites supporting people in drug addiction recovery (cf.
myRecovery.com). Depending on individual access restrictions, the visibility of
information can differ drastically across different services.
Certain SNS offer more or less unrestricted access to member profiles, while
others offer refined access-tuning controls that limit profile information access to
a restricted community. With reference to the classification criteria just discussed,
Table 1 below proposes a categorization scheme for distinguishing different types
of SNS and applies it to the 15 most popular SNS, as identified by Alexa Global
Traffic Rank.3
3
Cf. The top 500 sites on the web, http://www.alexa.com/topsites, here accessed 1 Sep-
tember 2015.

Authenticated
The function assigned to the particular SNS rests on the interplay of the plat-
form’s communicative exigencies and the ways in which individual users adapt to
these affordances in different ways. As shown, the functions that users associate
with a particular service may or may not correspond with the initial functions of
the developers. However, at the heart of any SNS is the display of user-generated
content in relation to a community. Depending on the individual platform tools,
the community or the content aspect of the service can be more pronounced. Ser-
vices such as Facebook, LinkedIn, and Meetup excel through a range of options
from which users can draw to create, display, and manage their connections with
others. In contrast, the principal aim of media-sharing and (micro-)blogging net-
works is directed toward content creation and distribution.
These services may well support the creation of social ties, but network com-
munities emerge only as a side effect of the primary function of content sharing.
The type of information members disclose about themselves stretches from general
information on their demographics, hobbies, interests, likes, and dislikes to more
specific private, interest-, or business-related details. While egocentric networks,
such as Facebook or VK, facilitate members’ portrayal and sharing of the differ-
ent facets of their personalities, community, opportunistic, and media-sharing net-
works restrict self-presentation to particular aspect(s) of user identities.
Offline anchorage relates to the degree to which provided data can be con-
nected to offline identities and members can be located in their physical and social
surroundings. Services with low offline anchorage seek to protect the members’
personal identities, while high offline anchorage services elicit and display user
data that are locatable in offline contexts, such as name, residence, occupation,
institutions, partners, and friends (Zhao, Grasmuck and Martin 2008).
Mobile networking accounts for the increasing availability of mobile technol-
ogies for SNS purposes. Most SNS offer mobile applications of their services that
provide users with instant access to their profiles and integrate mobile-specific
affordances (such as swiping, taking photos, and location tagging) with SNS activ-
ities.
It is interesting to note that out of the 15 most popular SNS, only six services
are primarily concerned with the formation and administration of an online com-
munity, while nine services are directed toward the creation and dispersion of
content. Although previous researchers agree that “[s]elf-disclosure is, perhaps,
one of the most important reasons for uploading information on SNS” (Yus 2011:
124), various services restrict self-presentation to the discursive representation of
a few particular identity facets of their users. As multidimensional platforms, SNS
combine commonly existing forms of communication, such as chat, microblog-
ging and e-mail, and enable the integration of different modes, such as voice,
images, and music. Due to the various ways of interactions afforded by the ser-
vices, self-presentation on SNS is never secluded and finished but can only be
described as a highly dynamic concept that is in constant flux.

Authenticated
Table 1. A classification scheme for social network sites
Function Type of in- Offline Mobile Afforded

formation Anchorage networking network
com- content general specific high low native auxil-
munity iary
Facebook egocentric
YouTube media-sharing
Twitter egocentric
LinkedIn opportunistic
Pinterest media-sharing
Google + egocentric
Tumblr egocentric
Instagram media-sharing
Reddit media-sharing
VK egocentric
Flickr media-sharing
Vine media-sharing
Meetup passion-centric
Ask.fm media-sharing
Classmates community
In relation to the concept of offline anchorage, out of 15 services, nine exhibit only
weak linkages to offline contexts. However, member profiles without any explicit
linkage to a user’s personal details, such as residence, employer, or friends, can
still be anchored in more subtle ways to identifiable and locatable offline individu-
als, e.g. in cases in which a user provides subtle information on individual likes or
dislikes or on events he or she is going to visit. With mobile technologies becom-
ing more and more popular, the line between web-based and mobile social net-
working is becoming more and more blurred. In regard to this, it is striking that the
media-sharing network Instagram had originally been developed as a mobile app
before the web version was introduced (Sengupta, Perlroth and Wortham 2012).

Authenticated
4. Language use in Facebook from a pragmatic perspective
On average, Facebook currently hosts more than 1.7 billion active users and has
turned into the world’s largest SNS (cf. statista 2016). Just like other services sup-
porting the realization and administration of egocentric and community networks,4
two fundamental practices are at the heart of Facebook-enabled practices: “the
presentation of self […] and the building and maintenance of networked relation-
ships” (Seargant and Tagg 2014:5). However, self-presentation and community
building via and within Facebook is inevitably bound to standardized templates
and text automation tools that gradually define and standardize user actions. As
Yus (2011) shows, “the repetition of a unique interface for profiles generates a
conventionalization of the SNS genre that reduces, at least initially, the reader’s
effort (to locate information)” (2011: 21). Likewise, when producing texts via and
within Facebook, members gradually outsource text creation and distribution pro-
cesses to preset templates and software-based actants (Latour 2008; Eisenlauer
2013). In order to disclose the gradual impact of the software on the discursive acts
of self and other presentation, it is, therefore, necessary to specify the service’s
key affordances, i.e. the “fundamental properties that determine just how the thing
could possibly be used” (Norman 1988: 9).
Boyd highlights four key qualities of SNS. Profiles enable members to “write
themselves into being” (boyd 2008: 121) by describing different facets of their
personality. Friends lists give an indication of the intended audience, as they pro-
vide information on a member’s individual network connections. Public comment-
ing tools support members in posting texts on their own and befriended members’
profiles and effect in the performance of social connections in front of broader
audiences. Within Facebook, posting texts can be gradually outsourced to text
automatisms, as, for example, when clicking the Like button in response to a post
of a befriended user. Stream-based updates re-display user-generated content,
such as status updates, photos, and comments, in the newsfeed and give members
an impression of those around them. Such automatic updates are at the commu-
nicative heart of everyday Facebook interactions: Any post, comment, or other
user action is automatically displayed in befriended members’ newsfeed streams
and can be instantly commented on, liked, or shared.
From a pragmatic point of view, the medial framing of user actions performed
via and within the social network site Facebook calls for a revision of long-held
beliefs and fundamental concepts in pragmatics and discourse analysis, in particu-
lar the notions of context and speech acts.
4
This is, for instance, the case in Russia’s largest service VK, Google’s Facebook rival
Google+, or the Alumni network Classmates.com.

Authenticated
4.1. Facebook and context
One of the basic tenets in pragmatic theory is the notion that meaning is always
context-dependent. Any production and interpretation of an utterance constitutes
the context of the relevant components of the environments. Accordingly, context
realizes only a subset of meanings offered by the environment and is created col-
laboratively and continually by the participants. When constituting the context,
participants do not only cope with the environmentally situated utterance, but also
rely on declarative, procedural, and episodic knowledge, as well as on knowledge
that relates to earlier, similar contexts (Schegloff 1992; Bublitz and Hoffmann
2012; Meibauer 2012).
This interactive and dynamic concept of context applies equally to language
use via and within Facebook. However, the service facilitates the ongoing integra-
tion of on- and offline worlds. Both worlds contribute to meaning formation, and
thus, the interpretation of the relevant environments can become more and more
complex. A vivid example for how a member’s physical location expands directly
into Facebook discourse is the location service. Supported by location data pro-
vided by their mobile devices, members may share information on their current
whereabouts by choosing from a list of places near them. Such cues on the phys-
ical environment in which a status update, comment or other text was being pro-
duced are integrated with various contextual cues arising from constantly chang-
ing online landscapes in which the text is presented. Meaning environments that
surround communicative interaction in Facebook are, to a large degree, affected
by elaborate software algorithms: A prototypical example for this is Facebook’s
newsfeed service that detects recently updated content by design and aggregates
texts from many different profiles into a member’s individual stream.
The merging of on- and offline environments, the automatic dispersion of user
updates around the network together with the service’s status as a meso medium5
may lead to a communicative setting that has been labeled “context collapse” (Mar-
wick and boyd 2011; Gohl and Schilling 2013): A member’s network of friends
commonly encompasses parents, colleagues, ex-lovers and best friends, and other
discourse communities. By default, all of them are updated about the participant’s
individual network actions, though these might be directed at a particular audi-
ence. It follows that SNS “collapse diverse social contexts into one, making it
difficult for people to engage in the complex negotiations needed to vary identity
presentation, manage impressions, and save face” (Marwick and boyd 2011: 123).
Figure 1 illustrates a case where a member had great difficulties assessing the
heterogenic nature of the potential audience. In the example, the profile owner’s
5
Meso media support the formation of network sizes of up to several thousand members
(see Section 2).

Authenticated
Figure 1. Context collapse in a status update
post closely interrelates with the particular social frames of the intended audience,
to which the commenter obviously did not belong.
4.2. Facebook and Speech Act Theory

Within Facebook, the individual platform options, such as writing status updates,
sharing events, or commenting on or liking befriended user posts, enable and
restrict particular user actions while imposing on the semiotic shape of the indi-
vidual actions. In other words, performing text actions via and within Facebook is
inevitably bound to the platform’s functional affordances. As Yus (2011) shows,
user actions may be customized to varying degrees ranging from ‘zero’ to ‘full per-
sonalization’. The former applies to settings in which “users have to follow strictly
the rules and the default interfaces for profile creation by entering personal data
on successive forms” (Yus 2011: 121); the latter relates to environments that offer
possibilities for individually adapting the profile interface according to the user’s
needs. Facebook and its various communicative features represent a system of par-
adigmatic options that give members great freedom in that they enable all kinds of
identity performances. When presenting themselves and others in status updates,
posts, and comments, members are free in the individual socio-stylistic choices of
their text actions. At the same time, Facebook’s (semi-)automated texts6 result in a
strong standardization and homogenization of user actions. Previous research has
studied both self-presentation practices evolving from automated texts and iden-
tity actions performed in and through user-generated texts. In this light, Bolander
and Locher (2010, 2015) investigated Facebook-enabled identity acts, using posi-
tioning theory (Davies and Harré 1990). More specifically, the authors are inter-
ested in two types of identity performances: first, the identity claims supported
by preset categories of the profile interface and, second, the discursive identity
6
This is initiated, among others, by the Like, the Add Friend or the Going to an Event
button.

Authenticated
construction via status updates. As they show, standardized profile categories sup-
port users in explicit self-labelling practices, i.e., selecting from options in the
relationship status category. Moreover, members employ profile categories for
indirect self-presentation through enumeratives when listing information on their
activities, hobbies, and interests. In relation to an identity construction via status
updates, Bolander and Locher highlight assertives as the most common illocution-
ary type, followed by expressives and commissives.
In Bolander and Locher’s (2010) sample of text actions performed by ten Swiss
Facebook members, status updates are most frequently employed in order to pro-
vide information on members’ current states of mind and/or in order to tell other
members what they are currently doing. Put briefly, the authors provide evidence
that the self-presentation in Facebook stands out through the use of implicit rather
than explicit identity claims.
In a similar vein, Eisenlauer (2013, 2014) highlights implicit identity claims as
a characteristic feature of Facebook discourse. In his study of Facebook afforded
literacy practices, he reflects on the role of electronic agents for Facebook dis-
course, disclosing the role of automated text actions for mitigating users’ identity
claims. According to Eisenlauer (2014), Facebook-bound text actions principally
fall into the two categories of creative text actions and automated text actions. In
cases in which members use keyboards to write status updates, posts, comments,
etc., they perform creative text actions. This means that they have a free choice in
expressing specific utterances and their respective illocutions. In other words, they
remain in charge of animating their propositional choices with the help of blank
text templates. In contrast, in automated text actions, the performance of the prop-
ositional and the utterance act is gradually delegated to various software-based
text-generation processes. When clicking on the Like button, adding a friend func-
tion, or responding to an event invitation, a member’s semiotic choice is reduced
to making predetermined binary decisions, i.e. to generating the text or not. Auto-
matic text-generation processes will have trade-offs and both positive and negative
consequences for user text actions. The employment of software-generated texts
is less time-consuming than self-authored texts and enables users to be present
and in contact with each other in new ways. In relation to identity performances,
automated text actions let members disclose information about themselves and/
or other members in a more indirect fashion. As Eisenlauer points out: “Whereas
self-authored texts on individual attributes and likes would always run the risk of
being interpreted as rather straightforward and blunt identity performances, the
employment of software generated texts enables a member to claim various iden-
tity aspects on her-/himself and others in a more indirect way” (Eisenlauer 2014:
314). On the other, hand-automated text generation in Facebook may lead to cases
where members cannot foresee and/or intend the communicative outcome of their
performances. Figure 2 illustrates a case of a member seeking advice on a discus-
sion forum after having accidently hit Facebook’s Add Friend button.

Authenticated
Figure 2. Seeking advice on Friend request sent by mistake
5. Limitations and outlook
The reliability and efficiency of the pragmatic frameworks developed for studying
SNS/Facebook discourses is obviously affected by updates and newer versions of
the software environments. In fact, since its introduction in 2004, Facebook has
undergone nine extensive relaunches (cf. Wikipedia.org), each introducing various
new functions and capabilities that shape and are being shaped by emerging online
literacy practices. While previous relaunches of the Facebook platform intro-
duced, for example, mobile social networking and hashtagging, the current trend
goes towards moving profile pictures, the use of animated gifs and the integration
of 360 degree videos (cf. Süddeutsche Zeitung). Thus, pragmatic research on SNS
communication needs to come up with theoretic frameworks that are broad enough
to be applied to newer versions and at the same time specific enough to capture the
individual characteristics of SNS-afforded literacy practices. Moreover, it is essen-
tial for pragmatic studies of social media to identify the data collection periods and
software versions under focus.
A further challenge for researchers studying SNS discourse is to strike a bal-
ance between empirical and theoretical research. Previous research has inter alia
studied what members actually do via and within SNS (Gohl and Schilling 2013;
Selwyn 2009), how they enact politeness (Losh 2008; Theodoropoulou 2015), or
in what ways member actions are affected by text automation properties of the
platforms (Eisenlauer 2014). While some of the findings strikingly demonstrated
the need for revisiting pragmatic key theorems and adapting them to the media
constellations that configure the communicative setting in SNS, the evidence is
often based on small samples or particular case studies. Findings in pragmatic
research on SNS communication need to be backed up by more empirical data
on a diverse range of discourse communities and discursive practices. A goal for
future researchers may be therefore to combine pragmatic studies of social media
with data collection and research methods in sociolinguistics. An empirical explo-
ration of large sets of language data could, for example, investigate the ways how

Authenticated
members of specific discourse communities employ Facebook’s text automation

properties to claim and impute community membership.
References
Adedoyin-Olowe, Mariam, Mohamed Gaber and Frederic Stahl

2013 A survey of data mining techniques for social network analysis. Journal of
Data Mining & Digital Humanities 18. http://jdmdh.episciences.org/18.
Androutsopoulos, Jannis, Yin Feng Hsieh, Joanna Kouzina and Reyhan Şahin
2013 Vernetzte Mehrsprachigkeit auf Facebook: Drei Hamburger Fallstudien. In:
Angelika Redder, Julia Pauli, Roland Kießling, Kristin Bührig, Bernhard Bre-
hmer, Ingrid Breckner and Jannis Androutsopoulos (eds.), Mehrsprachige
Kommunikation in der Stadt: Das Beispiel Hamburg, 161–197. Münster, Wax-
mann.
Arendholz, Jenny
Baym, Nancy
2007 The new shape of online community: The example of Swedish independent
music fandom. First Monday 12 (8). http://firstmonday.org/ojs/index.php/fm/
article/view/1978.
Bolander, Brook and Miriam Locher
2010 Constructing identity on Facebook: Report on a pilot study. In: Junod, Karen
and Didier Maillat (eds.), Performing the Self, 165–185. (Swiss Papers in Eng-
lish Language and Literature Vol. 24). Tübingen: Narr.
boyd, danah
2008. Taken out of context. American teen sociality in networked publics. Ph.D.
dissertation, Berkeley: University of California.
boyd, danah and Nicole Ellison
2007 Social Network Sites: Definition, history, and scholarship. Journal of Com-
puter-Mediated Communication, 13 (1). http://onlinelibrary.wiley.com/
doi/10.1111/j.1083-6101.2007.00393.x/full.
Bublitz, Wolfram and Christian Hoffmann
2012 Text and context. In: Martin Middeke, Timo Müller, Christina Wald and
Hubert Zapf (eds), English and American Studies. Theory and Practice, 435–
455. Stuttgart/Weimar: Metzler.
Davies, Bronwyn and Rom Harré
1990 Positioning: The discursive production of selves. Journal for the Theory of
Social Behaviour, 20(1): 43–63.
Dürscheid, Christa
2005 Medien, Kommunikationsformen, kommunikative Gattungen. Linguistik
online, 22 (1/05). http://www.schreibkompetenz.uzh.ch/Publikationen.html.
Dunbar, Robin
1993 Coevolution of neocortical size, group size and language in humans. Behavio-
ral and Brain Sciences 16: 681–694.

Authenticated
Ebersbach, Anja, Markus Glaser and Richard Heigl

2008 Social Web. Konstanz: UVK.
Eisenlauer, Volker
Eisenlauer, Volker
2014 Facebook as a third author (Semi)automated participation frameworks in
Ellison, Nicole, Charles Steinfield and Cliff Lampe
2011 Connection Strategies: Social capital implications of Facebook-enabled com-
munication practices. New Media & Society 13: 873–892.
Ermert, Karl
1979 Briefsorten. Untersuchungen zu Theorie und Empirie der Textklassifikation.
Tübingen: Niemeyer.
Facebook
n. d. Timeline of Facebook. https://en.wikipedia.org/wiki/Timeline_of_Facebook.
Accessed 28 September 2016.
Fraser, Matthew and Soumitra Dutta
2008 Throwing Sheep in the Boardroom: How Online Social Networking Will Trans-
form Your Life, Work and World. London: Wiley.
Goffman, Erving
1981 Forms of Talk. Philadelphia: Pennsylvania University Press.
Gohl, Fabian and Caroline Schilling
2013 What’s on Your Mind? and Who Do You Want to Tell? Negotiating collapsed
contexts and multiple audiences in Facebook status updates. In: Anne Ammer-
mann, Alexander Brock, Jana Pflaeging and Peter Schildhauer (eds), Facets
of Linguistics. Proceedings of the 14th Norddeutsches Linguistisches Kollo-
quium in Halle Saale, 75–87. Frankfurt am Main: Peter Lang.
Gross, Ralph and Alessandro Acquisti
2005 Information revelation and privacy in online Social Networks. Proceedings of
the 2005 ACM Workshop on Privacy in the Electronic Society, 71–80. Alexan-
dria, VA: Association of Computing Machinery.
Gruber, Helmut
2008 Specific genre features of new mass media. In: Ruth Wodak and Veronika
Koller (eds), Handbook of Communication in the Public Sphere, 363–381.
Heidemann, Julia
2009 Online Social Networks. Ein sozialer und technischer Überblick. Informa-
tik-Spektrum 33(3): 262–271.
Heidemann, Julia, Mathias Klier, Andrea Landherr and Florian Probst
2011 Soziale Netzwerke im Web. Chancen und Risiken im Customer Relationship
Management von Unternehmen. Wirtschaftsinformatik & Management 3 (3):
40–45.
Hiippala, Tuomo
2014 Multimodal genre analysis. In: Sigrid Norris and Carmen Maier (eds.), Inter-
actions, Images and Texts: A Reader in Multimodality, 111–123. Berlin/Bos-
ton: de Gruyter.

Authenticated
Holly, Werner
2004 Fernsehen. Tübingen: Niemeyer.
Holly, Werner
2011 Medien, Kommunikationsformen, Textsortenfamilien. In: Stephan Habscheid
(ed.), Textsorten, Handlungsmuster, Oberflächen. Linguistische Typologien
der Kommunikation, 144–163. Berlin/Boston: de Gruyter.
Jones, Rodney and Christoph Hafner
2012 Understanding Digital Literacies: A Practical Introduction. London/New
York: Routledge.
Jucker, Andreas and Dürscheid, Christa
framework. Linguistik online 56 (6): 39–64.
Knoke, Felix
2009 Abschied von der Internet-Mumie. Spiegel Online. April 24, 2009. http://
www.spiegel.de/netzwelt/web/0,1518,620913,00.html.
Latour, Bruno
2007 Eine neue Soziologie für eine neue Gesellschaft. Einführung in die Akteur-
Netzwerk-Theorie. Frankfurt: Suhrkamp.
Lomborg, Stine
2011 Social Media as communicative genres. MedieKultur 27 (51): 48–64.
Losh, Elizabeth
2008 In polite company: Rules of play in five Facebook games. ACE ’08 Proceed-
ings of the 2008 International Conference on Advances in Computer Enter-
tainment Technology, 345–351. New York, NY: ACM.
Martin, James
2009 Genre and language learning: a social semiotic perspective. Linguistics and
Education 20(1): 10–21.
imagined audience. New Media & Society 13(1): 114–133.
Meibauer, Jörg
2012 What is context? Theoretical and empirical evidence. In: Rita Finkbeiner,
Jörg Meibauer and Petra Schumacher (eds.), What is context? Linguistic
Approaches and Challenges, 11–32. Amsterdam/Philadelphia: Benjamins.
Norman, Donald
1988 The Design of Everyday Things. New York: Doubleday.
Posner, Roland
1986 Zur Systematik der Beschreibung verbaler und nonverbaler Kommunikation.
Semiotik als Propädeutik der Medienanalyse. In: Hans-Georg Bosshardt (ed.),
Perspektiven auf Sprache. (Interdisziplinäre Beiträge zum Gedenken an Hans
Hörmann.), 293–297. Berlin: Walter de Gruyter.
Rheingold, Howard
1993 The Virtual Community: Homesteading on the Electronic Frontier. Reading,
MA: Addison-Wesley.
Richter, Alexander and Michael Koch
2007 Social Software. Status quo und Zukunft. Technischer Bericht 2007/01. 243.
München: Fakultät für Informatik, Universität der Bundeswehr München.
https://dokumente.unibw.de/pub/bscw.cgi/d1696185/2007-01.pdf.

Authenticated
Robertson, Scott, Ravi Vatrapu and Richard Medina

2010 Off the wall political discourse: Facebook use in the 2008 U.S. Presidential
Election. Information Polity 15 (1–2): 11–31.
Schegloff, Emanuel
1992 In another context. In: Alessandro Duranti and Charles Goodwin (eds.),
Rethinking Context: Language as an Interactive Phenomenon, 193–227. Cam-
Seargeant, Philip and Tagg, Caroline (eds.)
2014 The Language of Social Media: Identity and Community on the Internet. Bas-
ingstoke: Palgrave Macmillan.
Selwyn, Neil
2009 Faceworking: exploring student’s education-related use of Facebook. Learn-
ing Media And Technology 34 (2): 157–174.
Sengupta, Somini, Nicole Perlroth and Jenna Wortham
April 13, 2012 “Behind Instagram’s Success, Networking the Old Way”. The New York
Times. Accessed 27 September 2016.
Statista
2016 Facebook users worldwide. https://www.statista.com/statistics/264810/num-
ber-of-monthly-active-facebook-users-worldwide/. Accessed 27 September
2016.
Süddeutsche Zeitung
2015 Facebook unterstützt 360-Grad-Videos. Süddeutsche Zeitung. http://www.
sueddeutsche.de/news/service/internet-facebook-unterstuetzt-360-grad-vide-
os-dpa.urn-newsml-dpa-com-20090101-150924-99-03605. Accessed 24 Sep-
tember 2015.
Theodoropoulou, Irene
2015 Politeness on Facebook: the case of Greek birthday wishes. International
Pragmatics Association 25(1): 23–45.
Threadgold, Terry
1997 Social media of semiosis. In: Roland Posner, Klaus Robering and Thomas
Sebeok (eds.), Semiotik. Ein Handbuch zu den zeichentheoretischen Grundla-
gen von Natur und Kultur, 384–404. Berlin/New York: Walter de Gruyter.
Thuraisingham, Bhavani, Satyen Abrol, Raymond Heatherly, Murat Kantarcioglu, Vaibhav
Khadilkar and Latifur Khan
2016 Analyzing and Securing Social Networks. New York: Taylor & Francis.
Yus, Francisco
2011 Cyberpragmatics: Internet-mediated Communication in Context. Amsterdam/
Zerdick, Axel, Valerie Feldman, Dominik K. Heger and Carolin Wolff (eds.)
2004 E-merging Media. Communication and the Media Economy of the Future. Ber-
lin: Springer.
Zhang, Xing-Zhou, Jing-Jie Liu and Zhi-Wei Xu
2015 Tencent and Facebook data validate Metcalfe’s law. Journal of Computer Sci-
ence and Technology 30(2): 246–251.
Zhao, Shanyang, Sherri Grasmuck and Jason Martin
2008 Identity construction on Facebook: Digital empowerment in anchored rela-
tionships. Computers in Human Behavior 24(5): 1816–1836.

Authenticated
10. Discourse and organization
Maximiliane Frobenius and Cornelia Gerhardt
Abstract: Interaction on social media follows certain organizing principles and

patterns, as does face-to-face oral conversation. Research has begun to apply and
adapt methods developed from and for offline interaction, such as Conversation
Analysis and Discourse Analysis, to the sites of interaction that have emerged with
the advent of the Internet and mobile phone technology. Our chapter traces the
results of this research, encompassing Internet Relay Chats, text-messaging forum
interactions, emails, Facebook posts and comments, Twitter, Skype calls and video
blogs. This varied range of both spoken and written data has received attention
regarding the sequential organization of the users’ contributions in terms of both
time and space, and it has drawn interest to its implementation of classic motifs in
Conversation Analysis, that is, repair, openings and closings. The chapter presents
an overview and synthesis of the results of these strands of research.
1. Introduction
This chapter gives an overview of the scholarly efforts made to illuminate the
“central organizing principles of interaction” (Garcia and Jacobs 1999: 339) in
computer-mediated communication (CMC), and, more specifically, social media.
It reviews and critically evaluates work that applies notions first developed in the
Conversation Analytic (CA) tradition to interaction, be it spoken or written, that
is mediated through a digital device. Further, it reviews work situated in the Dis-
course Analytic tradition with a specific focus on topics that touch on the organi-
zational features of CMC.
CA as a discipline is rooted in ethnomethodological sociology (Garfinkel 1967;
Sacks 1995), and as such it is interested in the organization of human behavior
with a focus on spoken conversation. The basic premise is that human interaction
follows a social order whereby interactants co-construct meaningful exchanges
and successfully make sense of one another’s contributions. In other words, con-
versation is a site of orderly, that is, recognizably organized, human behavior. Its
organizing principles are manifestly oriented to by the participants in a conversa-
tion and can be observed and described (e.g. by a researcher). Thus, conversation
is not a random or chaotic conglomerate of utterances, but a well-formed compo-
sition of contributions that are relevant to and dependent on one another.
The observations of orderly interactional behavior are made on the basis of
what is referred to as “members’ method” (Schegloff 1995: xxx). This approach
DOI 10.1515/9783110431070-010
Authenticated
246 Maximiliane Frobenius and Cornelia Gerhardt
requires the researcher to adopt a participant’s understanding of the situation rather

than an outside observer’s. Each turn-at-talk is understood both as a meaningful
and appropriate reaction to the previous turn, as well as an indication of what
the next turn could be like to be understood as an appropriate contribution. This
method allows a researcher to show the interactants’ preferences in the co-ordina-
tion and sequential organization of their contributions.
Turn-taking, the metaphor used to represent the instantiation of sequential
organization, is a central concept in the CA paradigm. It encompasses and makes
relevant the notion of accountability that underlies the design and interpretation
of social actions. Seedhouse (2004: 10) exemplifies accountability through the
template of the greeting exchange, a greeting-greeting adjacency pair: “when one
social actor greets another, a greeting response is the norm or has seen but unno-
ticed status. Failure to respond in this case, however, may be noticeable, account-
able, and sanctionable”. In other words, because interactants hold each other mor-
ally accountable on the basis of established norms, the design of any contribution
in the sequence of turns is constrained by virtue of its relation to the previous turn.
In the case of adjacency pairs, the normative expectation that the first element
necessitates the second is referred to as conditional relevance.
The sequential organization of certain accomplishments in human interaction,
such as conversational openings and closings, repair sequences, interruptions and
so forth, have traditionally been researched in the domain of spoken conversa-
tion. More recently, they have come to attract academic attention with regard to
how they are co-constructed in various online settings, where the contextual affor-
dances of the website or service in question may differ significantly from those of
face-to-face conversation. In this chapter, we revisit and contextualize these efforts
alongside a number of topics that recur in them: Salient issues are the blurring of
boundaries between seeming dichotomies, such as written language and spoken
language and synchronous and asynchronous communication. These issues give
rise to a reconsideration of time as a dimension along which communication is
organized. The research reviewed here recognizes space as a dimension that, in
predominantly visually mediated interaction, also bears significance for the joint
human accomplishment of accountable actions. Accordingly, the notion of turn
must necessarily undergo a revision in light of the contexts under consideration.
A contextual factor of CMC that researchers increasingly attend to is its inherent
multimodality, which offers various affordances that interactants can make use of.
After exploring the written/spoken cline and synchronicity as necessary notions in
this context, we first focus on predominantly written genres, on repairs, on open-
ings and closings and conclude with spoken CMC.

Authenticated
Discourse and organization 247
2. Written communication and spoken communication in social media
In the investigation of language, researchers in the past have found it intuitive

to attend to spoken language first and foremost rather than to written language.
Linell (2005: 27) cites Bloomfield (1933), Saussure (1964), Lyons (1981) and
Hockett (1958) as expressing the primacy of speech as the object of modern lin-
guistic attention over written language. Following Vachek (1949), Linell neverthe-
less rejects the primacy of speech and gives both speech and writing the status as
situated meaning-making activities (see also Watson 2009). Language use online
has predominantly been written language use. As Schandorf (2012: 319) argues,
“despite early techno-utopian visions of virtually embodied interactions in vir-
tual worlds, the wide accessibility of mobile phones and VoIP technologies, the
popularization of video conferencing (e.g. Skype) and the more recent availabil-
ity of mobile video communications (e.g. Apple’s Facetime), most digital media
communication remains firmly text-based”. However, this finding may soon be
outdated given the dynamics of mediated communication and the development of
mobile applications.
Similarly, Herring states that “text-based computer-mediated communication
(CMC) enjoys historical precedence, and it remains more popular than VoIP [Voice
over Internet Protocol]” (Herring 2010a). In other words, while early CMC was
mainly performed as written communication due to reduced access to data transfer
technology, the successive spread of high-speed internet and the development of
audio- and video-based channels had not overturned the dominance of written lan-
guage on the web, at least when Herring wrote her statement in 2010. One might
argue that in the first decade of the 21st century, the internet still provided a perva-
sive domain for written rather than spoken interaction, possibly unlike much of our
offline communication. Hence, for chronological reasons, the demanding nature of
spoken multimodal data, and because of the dominance of writing, there appears to
be more research on written online communication than on spoken.1 This predom-
inance also holds for research on the organization of interaction in CMC.
One aim of research related to the writing/speaking dichotomy is to determine
the speech-like character of online language (Baron 2008, 2010; Crystal 2011;
Markman 2013) or the extent to which written exchanges can be considered con-
versational (see the Special Issue of the Journal of Computer-Mediated Commu-
nication ed. by Herring 2010b). Jucker and Dürscheid (2012: 44) observe a termi-
nological inaccuracy that leads to the confusion of what they term ‘codes’ (speech
1
Cf. discussions on the various terms in addition to ‘CMC’ introduced to denote the
object of study and the field of study, which often cover the written code rather than the
spoken (Jucker and Dürscheid 2012).

Authenticated
or writing),2 which are discrete entities, and the degree of distance and immediacy,
which is a function of the message style.
Going beyond the written/spoken binary, one can understand any situated inter-
action as a convergence of multiple modes, that is, different materialities in which
meaningful elements can be represented (Bublitz 2017). Spoken language is there-
fore represented in the phonic mode and written utterances in the graphic mode.
Also, the notion of ‘text-only’ is in need of revision in an online context. Any text
is situated in a system of affordances (Hutchby 2001: 447–449): the black on white
appearance of printed text on paper is superseded by a pixelated representation on
a screen that is movable, clickable and changeable. This has considerable implica-
tions for the semiotic system that writing is a part of, and both designers of social
media and their users adopt the new functions that are available.
CA, as the approach most known for its interest in uncovering the workings of
the sequential organization of turns in spoken language, seems to be predestined to
be applied to analyses in a largely two-sided, dialogic or polylogic context. How-
ever, Have (1999) explains that central concepts of CA such as gap and overlap are
not readily applicable to written language. With regard to the fundamental analytic
strategy that uses a subsequent contribution to identify the participants’ jointly
negotiated understanding of the meaning or function of a turn (next-turn proof
procedure), he states: “In some forms of CMC, one could use a similar strategy,
in that, for instance, later contributions to a ‘thread’ in a ‘news group’ or ‘discus-
sion list’ can be used to inspect at least some members’ analysis of previous post-
ings” (Have 1999: 276). Thereby, he provides a justification for the extension of
an approach developed for the analysis of spoken language to written interaction
(also see Gibson 2009 for a comparison of written interaction to spoken conversa-
tion in online forum discussions).
3. Synchronicity
In addition to the spoken-written dichotomy, which has been challenged in a

CMC context, in technologically-mediated environments synchronicity needs
to be taken into consideration when analyzing the interactional organization of
these discourses. As previously mentioned, types of CMC are often classified as
synchronous or asynchronous (Baron 2010; Crystal 2006), where synchronous
communication refers to interaction in “real-time” (Crystal 2006: 135) and asyn-
chronous communication to “postponed time” (Crystal 2006: 134). Herring (2007)
explains:
2
Note that we will use the term ‘mode’ in the following when referring to the materiality
of messages (cf. Bublitz 2017).

Authenticated
Asynchronous systems do not require that users be logged on at the same time in or-
der to send and receive messages; rather, messages are stored at the addressee’s site
until they can be read. Email is an example of this type. In synchronous systems, in
contrast, sender and addressee(s) must be logged on simultaneously; various modes of
“real-time” chat are the most common forms of synchronous CMC.
Baron (2010: 1) lists chat, Instant Messaging (IM) and computer conferencing
as synchronous, and email, text-messaging, bulletin boards, blogs and social net-
working sites (SNS) as asynchronous, but not without remarking that depending
on the interactants’ practices, the boundaries can blur. Email, for example, can be
used synchronously if interactants respond immediately, and IM can be used asyn-
chronously if the response is delayed (Baron 2010: 1, 2004; Darics 2014; Herring
2007).
Garcia and Jacobs (1999: 339) use the term ‘quasi-synchronous’ (QS-CMC)
to differentiate further: “[A]lthough posted messages are available synchronously
to participants, the message production process is available only to the person
composing the message. Thus the process of message transmission (posting) in
QS-CMC is not synchronous with message production.” This distinction between
posting and production of a message is significant for the understanding of the
organization of the unfolding interaction, as “information about real-time turn
development is not revealed” (Jones 2013: 490). Thus, the concept of the transi-
tion-relevance place as found in face-to-face conversation (Sacks, Schegloff and
Jefferson 1974) is “more nebulous in chat”, as Markman (2013: 543) points out.
Overlap and interruption in the traditional sense are not possible (Herring 1999;
Markman 2005; Jones 2013). Likewise, the interpretation of pauses in the interac-
tion is complicated by the lack of access to an interlocutor’s actions (Rintel, Pittam
and Mulholland 2003).
As a logical consequence of this more refined notion of synchronicity, that is,
the inclusion of the ‘quasi-synchronous concept’, some studies used as data for
their analysis not just the chat-log or record of messages exchanged, which would
mean analyzing the finished product alone, but they also included information on
the production process, for example through screen-capture technology or vid-
eo-recordings (see Garcia and Jacobs 1999; Markman 2005, 2009, 2013; Jacobs
and Garcia 2013; Jones 2013; Meredith and Stokoe 2014; Reeves and Brown
2016). In these cases, researchers make use of information that is not available
to both (or all) parties of the interaction, but only to the respective producer of a
contribution. Thus, they access information on a core difference between medi-
ated and unmediated (face-to-face) interaction, the separate production process.
This represents an asymmetry of information when comparing the status of the
producer of the message and the researcher to that of the recipient of the message.
It is worth considering a brief theoretical aside: A CA approach to the organ-
ization of talk-in-interaction focusses on only the bare transcript of the interac-
tion, leaving aside contextual information that the participants do not accountably

Authenticated
orient to. Sacks, Schegloff and Jefferson (1974: 729) explain that “the display of
[interlocutors’] understandings in the talk in subsequent turns affords a resource
for the analysis of prior turns and a proof procedure for professional analyses of
prior turns, resources intrinsic to the data themselves”. Any editorial work on a
chat interactant’s part that comes before the transmission, such as typing, eras-
ing, re-typing and so forth, is unavailable to the receiver(s) of the message, mak-
ing it inaccessible as an interpretative resource for them. Therefore, using it as
an interpretative resource for research would then represent a deviation from the
members’ method as established in Ethnomethodology and CA (Garfinkel 1967;
Schegloff 1995; for similar reasoning, cf. Tudini 2014). However, as Meredith
and Stokoe (2014: 186) argue by drawing on Drew, Walker and Ogden (2013), in
the case of self-corrections, editorial work in chat contributions is indicative of an
interlocutor’s sense of what is appropriate in the sequential position to be filled.
The analytical advantage gained from this demonstrates that such an expansion of
data is a useful approach to quasi-synchronous, mediated interaction.3 Reeves and
Brown (2016) argue that investigating data beyond the interaction unfolding on
the screen, that is, the embeddedness of CMC in off-line communication, allows
insights into the integration of SNS into day-to-day life.
Multi-channel online services, such as Google Wave,4 which encompassed
collaborative work on documents, email, instant messaging et cetera, enable par-
ticipants to see when their interlocutor is typing, an affordance called “typing indi-
cator” (Auerbach 2014). One implementation involves a “keystroke-by-keystroke”
(Trapani 2010) transmission, where any keyboard action is instantly visible to the
parties involved.5 In other formats, the prospective receiver of the message sees an
automated system message indicating that the interlocutor is currently involved in
some keyboard action.6 A typing indicator grants all participants in the interaction
access to the same information, thereby eliminating the asymmetry of information.
An analysis of such data compared to data collected through screen capture or vid-
eo-recording would demonstrate how users treat this affordance when both author
and receiver of a message perceive its construction simultaneously. This would
enable the observation whether such devices enhance the projectability of utter-
ances in written CMC, in other words, whether the contributions were temporally
more fine-tuned with smaller pauses between contributions.
3
See also Beißwenger (2008a, 2008b) and Jones (2004, 2013) for discussions of what
contextual dimensions to consider in CMC research.
4
Note that Google Wave was introduced in May 2009 and terminated on April 30, 2012,
thus it is no longer in service (cf. Google Help 2015).
5
E.g. in Google Wave, VAX system Phone Utility (Anderson, Beard and Walther 2010)
and Unix talk (Auerbach 2014).
6
E.g. in WhatsApp (Dürscheid and Frick 2014: 167), Gchat, Google Talk, iChat (Crair
2014), Skype chat, NatWest web chat.

Authenticated
With the development and spread of new affordances in written CMC, the con-
cept of synchronicity must be viewed carefully, especially with regard to attempts
to understand the sequential unfolding of an interaction. After all, ‘no gap, no over-
lap’ as a basic preference in oral conversation refers to a temporal domain, in that
the sounds through which this communication is transmitted are predominantly sit-
uated on a temporal plane. However, writing also contains a visual, that is, spatial,
component. In fact, our conceptual understanding of the coherence achieved in
chat and IM interactions would benefit from building on the arguments made for a
more comprehensive collection of data, thereby acknowledging the significance of
spatial constraints and affordances for the sequential organization of written data.
4. Turn-taking in predominantly written CMC: Chats, Instant

Messaging, blogs/vlogs and comments, Twitter, Facebook
With regard to both quasi-synchronous and asynchronous written multi-partici-

pant CMC, Herring (2010a) writes that “normal face-to-face (F2F) patterns of
turn-taking are disrupted […], resulting in disrupted turn adjacency and overlap-
ping exchanges” (also cf. Garcia and Jacobs 1998; Herring 1999; Panyametheekul
and Herring 2003; Anderson, Beard and Walther 2010).
(1)
1 <ashna> hi jatt
2 *** Signoff: puja (EOF From client)
3 <Dave-G> kally i was only joking around
4 <Jatt> ashna: hello?
5 <kally> dave-g it was funny
(Herring 1999: 3)
In this interlaced exchange the greeting from line 1 is only returned in line 4, while
line 3 is responded to in line 5. In other words, interactional contributions that are
situated next to each other on the screen may not be designed by the interlocu-
tors to be interactionally coherent, and vice-versa, contributions designed to refer-
ence one another may not be spatially adjacent. Delays (lag) caused by technical
parameters may result in the display of contributions in a different order than they
were produced in. Greenfield and Subrahmanyam (2003: 718) explain that “chat
requires strategies for identifying relevant utterances that is, responses that follow
earlier utterances in a coherent conversational thread”.
Interactants use terms of address to make a directed contribution and thereby
nominate the “intended next ‘speaker’” (Panyametheekul and Herring 2003) as a
way of compensating for the lack of audio-visual cues, in comparison to face-to-
face communication (Werry 1996; see also Greenfield and Subrahmanyam 2003:
729 for their observations on “vocative cues”). This strategy of addressing an

Authenticated
interlocutor directly and thereby selecting next ‘speaker’ proved to be the most
frequent turn-allocation strategy in Panyametheekul and Herring’s (2003) data set
from a Thai chat room. Hence, the authors posit a similarity in the turn-alloca-
tion behaviors of interactants in oral conversation and in chat room interaction. In
newsgroups, however, where messages are also generally public, “the newsgroup
system’s infrastructure requires that users specify their addressees and the posi-
tions of their messages in the sequential organization of the discussion” (Marcoc-
cia 2004: 116). In other words, the affordances in Marcoccia’s dataset leave the
user no choice but to specify the cohesive ties of their contributions in a certain
format, whereas in chat rooms the users have more freedom in this respect.
For synchronous (“keystroke-by-keystroke”) written multi-participant CMC,
Anderson, Beard and Walther (2010: 15) observe that the participants in their
study tend to self-select. Their interaction contains both phases of overlap, where
several participants typed simultaneously, and long pauses where the previously
posted messages were strategically decoded.
(2)
23 C: a stop: ::::(4)we need to store some of this info.....(23.7)
24 do we know how?
(Anderson, Beard and Walther 2010: 8)
In Example 2, C self-selects after a pause of 23.7 seconds. Interruption, in the sense

of preventing an interlocutor from finishing their contribution, proves impossible
in this environment. This represents a clear deviation from the ‘no gap, no overlap’
ideal in spoken conversation.
Baron’s (2010: 12) research on IM conversations between college students
reveals that they can be “fairly lengthy” on average, and that they can contain
extended pauses of up to several minutes, displaying a significant difference from
face-to-face communication regarding temporality (2010: 24). Baron explains that
while oral conversation is typically a foregrounded activity, IM interactions are not,
that is, they may be just one of several activities requiring an interactant’s attention.
Regarding the comparability of research on IM (and other quasi-synchronous CMC),
the interactant’s focus in a ‘real-world’ setting might differ from that in more exper-
imental settings (Baron 2010: 24), which relates back to the discussion of the con-
textual information that researchers should take into account. Closings in Baron’s
data appear similar to those of oral conversation. Darics (2014) presents research
on IM as a communicative tool in the workplace, where team-members who are
based in different locations interact. She, too, characterizes IM interaction as inter-
mittent, but in her data, openings and closings seem less prolonged than in Baron’s.
Darics points out that norms may not yet have been established for these practices,
or they simply are of a variable nature depending on a range of factors (2014: 17).
Berglund (2009) highlights the significance of sequencing in IM as a tool for
the creation and maintenance of coherence. On the one hand, she identifies adja-

Authenticated
cency pairs whose first and second pair parts are not spatially adjacent but never-
theless recognizable as two parts of the same unit, and on the other hand, first pair
part structures can be repeated (e.g., several questions) before the second pair parts
are provided (i.e., several answers) (2009: 23).
(3)
Line Mode Time From To Message
1 MSN 17:40:18 BEAY FeliciaSmith hey baby, are u goining home with simon today
2 MSN 17:40:28 BEAY FeliciaSmith i ′mean.. if u wll be in school
3 MSN 17:40:34 BEAY FeliciaSmith can i cook for u guys?
4 MSN 17:40:37 FeliciaSmith BEAY like home to austria you mean?
5 MSN 17:40:38 BEAY FeliciaSmith u and him
6 MSN 17:40:50 BEAY BEAY mm. i think we were going out to eat
7 MSN 17:40:53 BEAY FeliciaSmith Ookie
(Berglund 2009: 13)
In Example 3, a first pair part (line 1) is visibly oriented to by the repetition of

‘home’ in the response line 4.
For data from a gay chat room, Jones (2013: 500) identifies a particular rhythm
in the interactions, namely “bursts and breaks”. He argues that interactants move
through a pattern of topics, thereby negotiating a potential meeting, where longer
pauses between the topics occur because interactants may be involved in several
interactions at the same time, and because pauses can have different implications,
such as lack of interest or display of a “high market value” (2013: 506) (cf. also
Fritz, Ch. 11, this volume). Again, as in other data discussed so far, we find a devi-
ation from the ‘no gap, no overlap’ ideal. Rintel, Pittam and Mulholland (2003), on
the other hand, approach pauses or “non-responses” in internet relay chat (IRC) as
morally and interactionally implicative, regardless as to whether they are caused
by the transmission system (e.g. lag) or by the interactants themselves. The follow-
ing example from their study illustrates the conditional relevance of, for example,
greetings and hence moral implicativeness, in that a user is held accountable for
not responding.
(4)
427. [sammi] ran you there

448. [sammi] RAN you there
503. [sammi] Ran: HELLO!
505. [Ran] Sammi: ya, I’m here, I was just backtracking to see what I missed, I can’t keep up with
this
(Rintel, Pittam and Mulholland 2003)

Authenticated
Markman (2005), whose data includes information on the production process of

individual messages in quasi-synchronous chat, shows that the reversal of the floor
claiming and message production order leads to phantom adjacency pairs, where
question and answer pairs that are seemingly coherent only appear so by chance,
as the production of the second pair part precedes the production of the first pair
part.
Skovholt and Svennevig (2013) have conducted research on email communi-
cation in the work place. They focus on the conditional relevance of messages con-
taining speech acts that in face-to-face conversation regularly require more than a
minimal response, for example a question or a greeting. Their study concludes that
emails that contain a request for information or for action make a response con-
ditionally relevant, that is, if the addressee(s) do(es) not respond, a ‘second pair
part’ is noticeably absent. However, there is a tolerance for delay in the response
(Skovholt and Svennevig 2013: 598). In the case where the request is for com-
ments or corrections to a proposal, a missing answer is read as acceptance of the
proposal. Hence, a response is not conditionally relevant in this case, as is also the
case for emails that do not contain requests or questions (Skovholt and Svennevig
2013: 598).
Text messaging, as a form of CMC that is not genuinely web-based but depend-
ent on a mobile device such as a mobile phone or a smart-phone, has been shown to
follow interactional norms regarding its sequential organization and especially the
conditional relevance of replies to certain types of messages (Laursen 2005, 2012).
Laursen, investigating interactional practices of 14-year-old Danish adolescents,
identifies initiating messages as inducing conditional relevance, making a reply
necessary: “among adolescent mobile users, there is a dominating norm that dic-
tates that an SMS receives a response” (2005: 53). Only if the initiating message
is a “night-time message” or a forwarded chain message, the interactants do not
orient to the initiating message as requiring a reaction (2005: 71). The time delay
between initiating message and reply that the interactants orient to as acceptable
depends on the relationship of the interactants, ranging from merely a few minutes
for best friends and couples, to a few hours for more distant relationships (2005:
71). Failure to respond within the acceptable time frame can result in reminders
being sent; these are usually framed to reflect a candidate interactional problem
along four dimensions: transmission problems, recipient problems, content prob-
lems, and relationship problems (2005: 71). For instance, by resending the mes-
sage, the interactants imply that there might have been a transmission problem. A
candidate relationship problem is exemplified in the following exchange:
(5)
Michael 13:03 pm What’s up beautiful are you coming over to my place by my mother.
morten and I can’t be bothered to go to training. Michael
Susanne – –

Authenticated
Michael 16:28 pm Heydi/hey there. are you doing something on friday. now we haten’t
seen each other in a long time.. I thourght you had plenty time when
we started going out but –><– I was maybe wrong … I don’t know
can you answer me. miss you michael
Susanne 16:59 Hey Mick! I just got back from school … Did you think I had more
time?? Im a busy train train girl … Hey r you upset about it? What
are you doing? *Your Girlfriend*
(Laursen 2005: 61)
Laursen further explains that text message dialogues rarely contain formal opening
or closing procedures, and that in case of a non-initiating message or one that does
not constitute a first-pair part, it is the receiver’s decision whether to continue or
discontinue the dialogue. In other words, text message exchange closings are not
(always) interactionally negotiated, but can be unilaterally implemented (Laursen
2005: 65).
Text-message dialogues do not necessarily occur in interactions that involve
text-messages only, but contributions to the interaction may well be enacted
through, for example, a telephone call (Laursen 2005, 2012; Hutchby and Tanna
2008). While the change from text-message to telephone call does not require
any additional interactional work (Laursen 2012: 83), it can be initiated through
a directive in the last text-message (“call me!”), or the recipient may choose to
switch to phone call without such a prompt (2005: 65–67).
Rather than focus on the sequential unfolding of several text messages,
Hutchby and Tanna (2008) demonstrate that there is a detectable orderliness within
text messages. Text messages can be simple or complex in structure, where “in
‘complex’ formats, more than one identifiable action is bundled up in one text”
(2008: 147). Package-texts contain multiple units fulfilling multiple interactional
functions, such as in the following text message where we find greeting, inquiry,
announcement and question/invitation (one action per arrow).
(6)
1→→ morning u, hows the
2 pampering going?:-)
3→ stu’s tryin 2 organise
4 something next
5→ friday..can u come
6 out?
(Hutchby and Tanna 2008: 153)
Hutchby and Tanna identify as an affordance of the technology the option to com-
pose such multi-unit texts, which in spoken language would result in the pro-
duction of several transition-relevance places “since a recipient only has access
to a message once it is fully composed, sent, and received by their handset, the

Authenticated
opportunity for response only exists at that point rather than at the completion of
any individual unit” (2008: 154). When package-texts are responded to, a signif-
icant difference to oral conversation manifests itself regarding the order in which
the individual elements are taken up. Whereas in oral communication there is a
preference for contiguity, that is, the last element in the previous turn is responded
to first, texters seem to respond to elements in the order found in the previous text
message (2008: 157).
Dürscheid and Frick (2014) differentiate between text-messages and WhatsApp
communication, where the latter can be further distinguished into SMS-like mes-
sages and chat function. The chat function requires “co-presence of the interloc-
utors at the keyboard” (2014: 170), in other words communication is quasi-syn-
chronous. They argue, since WhatsApp indicates whether an interlocutor is online
or not, this knowledge creates the expectation of timely responses (2014: 171). At
the same time, messages that cohere through their propositional content may not
appear next to each other on the screen (2004: 170; cf. Herring’s 1999 disrupted
adjacency).
Honeycutt and Herring (2009) identify Twitter as a noisy environment with a
large number of tweets posted at a high speed. Therefore, disrupted turn-adjacency
is to be expected, making the @-sign as a marker of addressivity a useful and fre-
quently used tool to enable coherent exchanges. Besides the @-sign, the authors
mention the use of third-party clients and individual Twitter homepages as aides
in using Twitter for coherent interaction (2009: 9; cf. also Zappavigna, Ch. 8, this
volume).
Bublitz and Hoffmann (2011) identify quoting as a source of coherence across
several types of CMC: online discussion forums, blogs and their comments, and
social networking sites. Depending on the affordances of the site in question, users
have at their disposal semi-automated if not fully automated acts (2011: 438–440),
whereby a user, through clicking on the ‘like button’ on Facebook, or the ‘Quote’
hyperlink in a forum, activates functions that create or relocate text passages (cf.
also Eisenlauer, Ch. 9, this volume). Greenfield and Subrahmanyam (2003: 721),
in a similar vein, describe repetition as a strategy to create coherence in a mul-
ti-party chat environment.
With regard to contributions on Facebook, Androutsopoulos (2015) explains
that spatial adjacency on a given wall may not necessarily coincide with sequential
coherence: the multimodal configuration and the multiple authors of these spaces
makes a prediction of shape and content of any next contribution impossible.
Frobenius and Harper’s research (2015) on written contributions and comments
in status update sections differentiates between various norms that interactants
orient to. In their data, Facebook users show a tendency to treat spatial immediacy
(rather than temporal adjacency, as is the case for spoken interaction) as a central
norm users adhere to in the creation of meaning and sense, whereby they manipu-
late their language use to make up for a failure to produce spatial immediacy (e.g.,

Authenticated
by employing addressivity). Facebook users have been shown to orient to what

the authors tentatively termed ‘personal immediacy’ when using the personal pro-
file of an interlocutor to indicate relevance structures. This research treats space
as a dimension of its own, rather than as a mere representation of the temporal
unfolding of contributions to an interaction through time. In fact, space provides
resources that allow interactants to disregard notions of temporal adjacency to a
degree.
Finally, gender has an impact on the practices in CMC: Herring’s many studies
illustrate such differences in that men receive fewer responses than women in syn-
chronous CMC and, vice versa, women fewer responses than men in asynchronous
CMC (for an overview, see Panyametheekul and Herring 2003). In a study on a
Thai chat room, similar turn allocation was found in that women make “greater
use of the preferred strategy to create coherence and orient to their conversational
partners, and males initiating more turns independently, as well as initiating more
flirtatious exchanges” (Panyametheekul and Herring 2003).
Turn-taking in CMC is shaped by two forces that are somewhat different in their
nature. First, the affordances of the technological systems7 may enable, impede or
even make obsolete certain mechanisms governing face-to-face talk-in-interac-
tion. While the lack of gaze and body posture in written CMC often requires extra
work by the users to single out addressees, for instance, newsgroup systems have
an inbuilt mechanism that allows and requires the specification of the address-
ee(s). The rule of not having gaps or overlap may also be waived when there is no
competition for the floor because, for example, simultaneous typing is possible.
Space in CMC may not always be a reflection of time, but it comes with its own
affordances depending on the specific technology used.
On the other hand, users can draw on the richness of language, on a reper-
toire of forms, to understand the coherent nature of the unfolding CMC text and
to create cohesive contributions themselves (cf. Schubert, Ch. 12, this volume).
Just as a prolific language user can use different linguistic forms or strategies to
convey similar ideas in a pub or at work, they can also adapt to different techno-
logical environments. Linguistic forms used in CMC contexts include terms of
address, lexical repetition, lexical substitution or quotations. Furthermore, just like
in non-computer-mediated contexts, language allows people to adapt to different
exogenous circumstances or to recreate them in their discourses. For instance,
whether some behavior is seen as morally implicative can not only depend on the
interlocutors’ relationship, but also the context of use (e.g. institutional vs. non-in-
stitutional settings) or the gender of the participants which may be relevant or be
made relevant by the users.
7
These may in turn be shaped by the users’ preferences during their development.

Authenticated
5. Repair
Just like in face-to-face conversation (Schegloff, Jefferson and Sacks 1977), there
is a need to correct one’s own and others’ mistakes or errors, or to solve other
communicative problems such as word searches in communication involving tech-
nologies. Schönfeldt and Golato argue that in multi-party chat interaction, “partic-
ipants take practices from ordinary conversation and apply them to their interac-
tion within this new form of communication” (2003: 275). Specifically, they show
that in chat as in spoken interaction, there is a preference for self-initiation and
self-correction in repair (2003: 272).
(7)
(288) Kaetzchen_F: Okeeeee … muss mal wieder looohooooos!
okaaaay … must once again goooooooo!
okay i have to leave again!
(336) Bebbi: Winke mal zum Ketzchen
wave once to the ((name))
wave to Ketzchen
=> (349) Bebbi: Winke mal zum Kaetzchen
wave once to the ((name=little cat))
wave to Kaetzchen
(Schönfeldt and Golato 2003: 257–258)
Here Bebbi self-initiates a self-repair by correcting a misspelled name. However,
Schönfeldt and Golato point out that they as researchers, as well as the other chat
participants, have no access to the message production process, which means they
cannot comment on self-repair that occurs before the completion and sending of the
contribution. Thus, the first repair initiation position they can identify in the data
“is the next possible turn (message) after the turn containing the trouble source”
(2003: 273). Garcia and Jacobs (1999) and Meredith and Stokoe (2014), whose data
also encompasses the production process, found self-initiated, self-completed same
turn repairs (for instance, the editing of messages while typing). Finally, chatters
use repair to gain access to an ongoing interaction between other chat participants,
as do conversationalists in spoken interaction (Schönfeldt and Golato 2003: 272).
(8)
(323) cousine1: calv: moment …sicher bin net sauer aber total ent-
täuscht hab gedacht ich wärs *heul*
calv: moment …certainly am not sour but totally
disappointed have thought i was+it *cry*
calv: wait … i am certainly not mad but totally
disappointed i thought i was the one *sob*
=> (354) sonnenblume72: cousine: was bist du? *neugierigfrag*
cousine: what are you? *curiously ask*
cousine: what are you? *curiously asking*
(Schönfeldt and Golato 2003: 270)

Authenticated
Here Sonnenblume72 enters the conversation by performing an other-initiated

repair.
Meredith and Stokoe (2014) investigate self-repair in quasi-synchronous Face-
book chat. Within the category of self-initiated self-repair they identify two types:
visible repair and message construction repair. In their data, the former is per-
formed in a contribution immediately following the trouble source or in third turn
position. The latter occurs during the typing of a message and before sending it,
making it unavailable to the recipient(s). Thus, the authors reveal that message
construction repair constitutes a type of repair unavailable to interlocutors in spo-
ken interaction. The functions that message construction repair has in Facebook
chat resemble those that it has in spoken interaction, namely repairs on stance,
prosody, and action formation (2014: 202). Visible repair, on the other hand,
mainly functions as error correction of spelling or typing (2014: 202).
Jacobs and Garcia (2013), whose data consists of video tapes of multi-party
intranet chat, also find self-repair in the construction of messages. In their data,
the placement of a contribution in relation to the one it responds to is a much more
salient issue, as several interactants potentially type at the same time and thus
compete for the same slot on the screen. In other words, failure to place a message
adjacent to the one it relates to can be seen as a trouble source.
Lazaraton (2014) reports on repair in weblogs as a resource for humor and
language play. In her data, commenters playfully orient to the passive voice as
standard language use, and failure to meet this standard regularly results in self- or
other-initiated repair, where other-initiation/repair is significantly more frequent
than in spoken interaction. Unlike Meredith and Stokoe (2014), Lazaraton (2014)
analyzes the finished contributions to the interaction, data that is available to all
participants in the interaction, rather than its production process. The compara-
bly small number of instances of self-repair in relation to other-repair might be
explained by the blog (comment) authors’ option to check contributions before
posting to eliminate potential trouble sources in this asynchronous type of inter-
action (Lazaraton 2014: 114). Using block-quotes to cite passages from another
poster’s contributions containing trouble sources is a repair strategy which initia-
tors use to create coherence. It is a consequence of the blog software’s affordances
because they allow texts to be positioned sequentially adjacent despite temporal
gaps (2014: 114). Collister (2011) presents evidence for the existence of a mor-
pheme which is specific to written online language which initiates repair, namely,
the asterisk (*). Chatters in the online game World of Warcraft use it both for self-
and for other-initiation in the repair of typos and production errors (2011: 919),
and the author further identifies the ^-symbol as an apparent alternative, which has
the same functions.
Harrison (2003) justifies the use of CA as an approach to analyze written data:
drawing on Have (2000), she argues that online communication is increasing and
that this interaction which takes place in a virtual environment can be seen as an

Authenticated
extension of the original object of study for CA, spoken conversation. She then
applies the CA-model of repair to data from email discussion lists, revealing that
all combinations of self- and other-repair and completion occur, with email spe-
cific trouble sources such as missing attachments or blank emails, but also mis-
spellings and personal attacks (flaming) that are disguised as repair.
Tanskanen and Karhukorpi (2008) report on a specific type of repair in asyn-
chronous email use, that is, concessive repair. This practice entails an “overstate-
ment and a successive repair sequence” (2008: 1591). In emails, concessive repair
only occurs in one position, the actual location of the trouble source itself, that
is, same turn (2008: 1598). Therefore, the authors argue, this type of repair has
implications for the interactive nature of these exchanges in that it demonstrates the
author’s adopting the readers’ perspective and pre-empting their potential criticism.
(9)
For the central Europeans it is perfectly normal that their mother stayed at home until they
went to school, when we in Finland are used to the fact that both of the parents work, basi-
cally from the day we are born (well, not exactly but you know what I mean).
(Tanskanen and Karhukorpi 2008: 1592)
In Example 9, the early return to work of Finnish parents is first exaggerated and
then directly repaired to anticipate potential criticism from other users.
Again it can be seen that strategies used in spoken face-to-face talk-in-inter-
action are being reproduced in CMC and also how the specific affordances of the
technologies cause an adaptation of such practices. A preference for self-repair
(even though not always visible to the other interlocutors) remains, implying simi-
lar ideas about politeness in these data as in face-to-face interaction. CMC specific
practices exploiting technological opportunities emerge, such as the use of block-
quotes to draw attention to a trouble source, or asterisks or other symbols as repair
markers. New trouble sources also appear such as one’s typing speed, which may
have direct consequences for the placing of messages in the ongoing conversation
and hence for the coherence of the unfolding text.
6. Openings and closings
As with any other communicative exchange, people have to enter and leave situa-
tions unless they assume an “open state of talk” where “participants have the right,
but not the obligation to initiate a little flurry of talk, then relapse back into silence,
all this with no apparent ritual marking” (Goffman 1979: 10).
Rintel and Pittam (1997) investigate IRC openings and closings, comparing
them to casual group face-to-face interaction. They find similar patterns such as
greetings on the one hand, but also quite different ones such as choice of nicks
(nicknames with which to be identified on the platform) on the other. Practices

Authenticated
identified in either setting serve similar functions (1997: 527), for example, the use
of exclamation marks or capitalization is compared to gestures, facial expressions
or tone of voice (1997: 514–515). Some features of embodied interaction may
even be transferred into written action descriptions, as in the example below.
(10)
6. 344 [DISP-#penpals][ACTION] metal shakes Jacstra’s hand
7. 349 [DISP-#penpals][ACTIONl Bobby *hugaz* his big sis!!!!!!!!!!!!!!!!!!!!!!!
8. 481 [DISP-#penpals][ACTION] melba hugs her lil broother Bobby
9. 553 [DISP-#penpals][ACTION] Megasta *huggggggggaz* his big sister
10. 86 [DISP-#australia][ACTION] JaKe waves to all the new folks …
(Rintel and Pittham 1997: 515)
For both openings and closings, Rintel and Pittam’s data reveals that there are no
fixed sequences, but that the order of elements is fluid. In fact, some elements may
not occur at all, which, apparently, is not oriented to as a flaw in the sequence by
the interactants. Rintel and Pittam (1997: 527–528) identify a number of stages that
IRC interactants pass through when entering an exchange: (1) a server announce-
ment, (2) a greeting to someone or no-one in particular or a question or state-
ment in lieu of a greeting, (3) an optional exchange of verbal representations of
non-verbal gestures of recognition and greeting as commonly used in face-to-face
interaction (see example above), (4) transition to the next phase in the interaction.
Greenfield and Subrahmanyam (2003: 734) find a strategy in teen chat rooms
they call “slot-filler”, which has the “functional properties of adjacency pairs from
oral conversation (Schegloff and Sacks: 1973), but without the formal property
of adjacency”. Slot-fillers are information requests about a potential interactant’s
identity that follow the fixed format a/s/l (age/sex/location). While the authors
demonstrate that this strategy is oriented to as an opening device in chat inter-
actions, they do not discuss whether or not the property of conditional relevance
holds for slot-fillers as much as for what is traditionally considered a first pair
part of an adjacency pair. Thus, the analysis suggests that slot-fillers are a fail-safe
option for interactants to secure a chat partner’s response, which seems to contra-
dict the findings discussed in Rintel and Pittam (1997) and in the research below.
Rintel, Pittam and Mulholland (2003) investigate IRC openings more closely.
Their investigation highlights the central role that server messages play in an
“automated joining event (AJE)” (2003). Drawing a comparison to telephone call
openings, the authors state that the automated joining announcement (JA) is sig-
nificantly different from the ringing of the phone in that it does not entail condi-
tional relevance. Therefore, the JA is not oriented to as a summons that requires
an answer from another party. The authors explain that if a JA possessed the same
conditional relevance the ringing of the phone has, “every JA would have to have
been followed by a greeting from every channel member, leaving little room for
any further interaction! Such a rigid summons-answer situation is clearly an unten-
able option in a group medium” (2003).

Authenticated
(11)
1987. [SERVER] NASA!xx@xx.xx.xx.xx has joined this channel
1988. [ACTION] NASA – hi to everyone
2040. [SERVER] Clown!xx@xx.xx.xx.xx has joined this channel
2044. [ACTION] NASA – is here anyone who would like to talk with me?
2089. [Clown] hello there everyone
2099. [Clown] Hello?
2107. [Clown] Be sociable!
2111. [ACTION] NASA – talk to me too?
2120. [Clown] NASA: hello.
2131. [NASA] Clown-at least one who … umm, hi …:)
2145. [SERVER] NASA has left this channel
2148. [Clown] NASA: hi what’s going down?
(Rintel, Pittam and Mulholland 2003)
As in the example above, the data under consideration reveal that a considera-
ble number of initiating attempts and greeting exchanges did not lead to further
exchanges, demonstrating a demand for a multitude of interaction attempts to
secure one (or several parallel) strands of interaction (see Have 2000 for an explo-
ration of the use of membership categorization devices (MCD) in the chat partner
selection phase).
Closings, similar to openings, are enacted through a number of stages. (1) The
closing phase is initiated; (2) as a medium-specific option, the IRC “/action” com-
mand is employed for a textual rendering of actions; (3) phatic communion tokens
are exchanged; (4) parting gestures are expressed in text form; (5) automated (and
optionally adapted) server message appears (Rintel and Pittam 1997: 529).
Markman, in her research on chat data of virtual team meetings, describes
a two stage process involving an “opening move” and an “agenda setting turn”
that interactants go through in the openings of their team chats (2009: 155).
She found interactional practices in these openings that resemble those of face-
to-face encounters.8 Closings were, again, enacted through a two-stage process,
where the first consisted of an “explicit closing remark/summary statement” and
a “turn projecting future action” (2009: 161). Both opening and closing sequences
as described in these chat data were subject to delays caused by intrusive turns,
owing to the “disjointed temporality” (2009: 161), which is one of the parameters
of this type of interaction.
For two-party IM interactions, Raclaw (2008) identifies two main patterns
which closings regularly follow, the expanded archetype closing and the partially
8
Note that these observations were not made on the basis of the actual first interactional
contributions, but the process of moving from non-task related talk to meeting talk, i.e.
transitional moments. Hence the term ‘opening’ is not entirely congruent with its usual
use in CA research here.

Authenticated
automated sequence. Whereas the former resembles conversational closings in that

there are pre-closing and terminal exchange adjacency pairs, it can also include a
medium specific element, the post-closing. This is a message posted by the soft-
ware as a consequence of a user’s signing out of the program or setting their status
to unavailable, and it may appear below a previously finished interaction. The
partially automated sequence, on the other hand, necessarily includes a message
that is automatically posted by the software to complete the closing sequence.
Interactants negotiate a pre-closing either through extensive pauses between con-
tributions or by giving reasons for leaving the computer/the interaction, which is
then followed by an interlocutors’ automated away-message.
(12)
1 metonym: so i should like, probably start writing my paper (11.0)
2 pudding: yeah i should probably go to bed (8.0)
3 metonym: so i will talk to you tomorrow, jah [yes]? (7.0)
4 pudding: jah [yes] (6.0)
5 pudding: good luck writing!!! (2.0)
6 metonym: thanks! (2.0)
7 pudding: latahz [i’ll talk to you later] (3.0)
8 pudding: haha, bye (9.0)
9 metonym is away
(Raclaw 2008: 9)
In this example, a pre-closing in the form of an exchange of accounts (lines 1

and 2) is followed by an arrangement for a future meeting (lines 3 and 4), a ref-
erence back to the accounts (lines 5 and 6), the terminal exchange (lines 7 and 8)
and finally the post-closing in the form of an automated message by the system
(Raclaw 2008: 10).
With regard to text messaging, Spilioti (2011), for her Greek data, describes
closings as part of politeness. She analyses when participants consider the use
of closings appropriate and which formulae they use. Closings taken over from
face-to-face interaction or from written or other mediated genres were used to
signal degrees of immediacy and proximity. This is in line with earlier research on
closings in text messaging where the social presence of the interlocutors was also
signaled in this way (Laursen 2005; Ito and Okabe 2005; Spagnolli and Gamberini
2007).
Generally it appears that ritual opening and closing sequences are less expected
in technological environments than outside of CMC, which may be a reflection
of a more ‘open state of talk’ or of a lesser commitment to the social encounter.
However, it also transpires that there are local rules in different formats depending
not only on the technology used, but also on sociocultural factors such as age or
gender. Comparing the wealth of studies on embodied spoken language and paper-
based written genres to the number of studies on current CMC based interactions,
in the main, one has to conclude that much more research is necessary to capture

Authenticated
the exact factors that may or may not make a closing or opening section expected
elements in a given environment. One feature of closing and opening phases in
CMC stands out, namely automated system messages that cannot be sidestepped.
In multiparty face-to-face interaction, there may be degrees of participation (Goff-
man 1979) and people use an array of bodily, vocal and verbal means to signal
their participation status. There is no one obligatory single signal for the co-inter-
locutors which convey that he or she is there/he or she has gone.
7. Spoken CMC
With regard to spoken CMC, one may expect a smaller impact from the technolo-
gies used. For instance, comparing Skype calls to classic telephone calls, one can
assume that because of the added visual information, skyping might be even closer
to face-to-face talk-in-interaction. Barron and Black (2015) contribute research
on Skype call openings between native speakers and non-native speakers using
English. For example, their analysis of listener behavior in the form of back-chan-
nels is less indicative of particular practices pertaining to the Skype context, but
linked to individual learners’ proficiency in their L2. Considering that Skype
offers simultaneous audio-visual and text-based interaction between users, further
investigation of these interactional events would be particularly desirable with
respect to the use of multi-modal features for the organization of such interaction.
This would include switching between audio-visual and spoken modes when, for
example, there are interactional troubles caused by a poor internet connection;
but possibly more importantly, the sequential (and not only in a temporal sense)
unfolding of the simultaneous use of these channels should be explored in terms of
the norms that are then oriented to by the interactants.
In this vein, Licoppe and Morel have identified a norm that interactants reg-
ularly orient to in Skype calls: the current speaker should be seen on screen. The
“talking-heads” (2012: 399) arrangement or “‘simplest systematics’ to the organi-
zation of video calls” (Licoppe and Morel 2014: 137–138) is the default on Skype.
This means that anything else that comes to be shown in these calls – and an
increasing mobility through the use of mobile devices allows for objects, or even
buildings or sites to be shown – is under scrutiny for its relevance to the ongoing
interaction. In other words, a camera movement towards an object or the surround-
ings is an accountable action (2012: 405).
Licoppe and Morel (2014) identify these showing sequences as elements that
require joint interactional work in that they have to be prefaced by the shower or
requested by the viewer and then ratified by the other participant. This prepara-
tory collaboration functions as a suspension of talking-heads norm and allows
for potentially irrelevant images to be shown until a first relevant view can be
produced and identified as such. The co-ordination of the audio-visual stream in

Authenticated
Skype calls is therefore a “collaborative and joint interactional process” (2014:

158).
In addition to research into one-on-one Skype conversations, there is also
research on interaction involving audio-visual communication between more than
two parties. Sindoni explores data from Camfrog, a software that allows “mul-
tiparty web-based video chat” (2014: 327) where several users meet in a virtual
room and can use both spoken language and typed messages to communicate with
one another. However, unlike face-to-face interaction, Camfrog allows only one
person to speak at a time. In other words, speakers have no option to create overlap
and are thus tied to this part of the ‘no gap, no overlap’ ideal. Sindoni remarks that
the software allows for another type of overlap conditioned by the use of written
chat message to open ‘parallel floors’, where the speakers turn into authors and
engage in another interaction while, possibly but not necessarily, listening to the
one who is speaking (2014: 327). This study gives ample space to considerations
of multimodal issues in the sequential organization of contributions in multi-party
video chat, a focus that is well worth pursuing in CMC research in an effort to
better understand what impact the contextual setting offered by a web space has
on human interaction in general (for studies of floor management issues in mul-
ti-party video interaction using experimental data, see Sellen 1992 and Heath and
Luff 2000).
Frobenius (2011) investigates openings in video blogs, which represent inter-
esting data with regard to the co-construction of interaction in that the spoken part
is clearly asynchronous and monologic. Her data set shows two clear tendencies,
the first being a unilateral mirroring of sequential conversational openings with
greetings and terms of address, and the second being the lack of both greeting and
term of address. Other more established monologic media such as TV news, radio
shows, or answering machine messages also influence language use on vlogs, with
vloggers using the available editing options to provide written opening credits,
for example. For closings, Frobenius (2014) finds that many vloggers replicate
the conversational pattern of pre-closing and terminal exchange adjacency pairs,
by providing only the first pair parts of each. Clearly, vloggers draw on conver-
sational patterns in the organization of their monologues, despite the fact that a
synchronous collaborative unfolding is impossible in this contextual setting.
Of particular interest in video blogs posted on video-hosting platforms such
as YouTube that are characterized by an inherent multimodality, is how interact-
ants manage to switch from one mode to another without a loss of coherence. On
the one hand, YouTube offers what they call a ‘deep link’. This affordance lets
commenters insert hyperlinks in their written texts which, when clicked on, takes
the reader to a specific point in the video and thereby allows a commenter to link
their comment to a specific point in the audio-visual material. On the other hand,
vloggers have adopted practices to incorporate written material from their view-
ers’ comments into their audio-visual recording: they can read out loud comments

Authenticated
and reply to them in speaking, and they show screenshots of comments while
producing a reaction to them in a voice-over (Frobenius 2014). YouTube clearly
represents a complex contextual setting where interactants’ creativity in the joint
production of meaning and sense is instantiated in novel ways.
With the increase of spoken interaction in CMC, the studies discussed here, in
the main, make manifest a desire for more, empirically based, thorough analyses
of the practices or norms that evolve as new technologies are used as resources for
communication, identity construction, or community building by their users. The
cases previously discussed indicate that a full picture of the organization of spoken
discourse in mediated settings needs to have a broader database as a foundation.
At this point, it mainly transpires that people are well aware of the affordances
of different technologies as enabling or constraining factors for the social actions
they wish to engage in.
8. Outlook
There is a progression in our understanding of the dimension along which types

of CMC are organized. We leave behind binary distinctions like spoken-written
and especially synchronous/asynchronous to enrich our view through a more fine-
grained model that distinguishes production, transmission and reception. This pro-
gression is taken further to include space as a dimension that interactants orient to,
where, for example, screen space produced and seen by the interactants feeds into
a spatio-temporal organizational feature. Additionally, the multimodal contextual
configuration of interaction sites is beginning to be understood as yet another com-
ponent of this organizational conglomerate. While research into the interactional
organization of social media indicates a move away from the time-centric ‘no gap,
no overlap’ ideal postulated for face-to-face conversation, there is still a great deal
to explore with regard to what interactants actually orient to when faced with mul-
titudes of ways to communicate – be it different modes, different platforms, and so
forth – to create a coherent exchange between users. This process of streamlining
multimodal online interaction is necessarily heavily influenced by website fea-
tures and design, making the contextual factors or affordances a starting point for
any pragmatic analysis. With another fundamental shift in online contextualization
options in the near future, namely augmented reality in the form of, for exam-
ple, three-dimensional projections (see devices such as Google Glasses or Oculus
Rift), we will be faced with a range of novel interactional sites and practices.
With the constant development of said affordances, it is to be expected that
interactants will either incorporate these into their orderly, organized communi-
cation processes, or not use them at all. An obvious candidate for pragmatists’
research agenda is to continue to create an inventory of such practices in any set-
ting that is accessible; beyond this ongoing task lies a more holistic (and perhaps

Authenticated
elusive) complex of questions: what regularities do we find across all settings?

How can we successfully adapt the Conversational Analytic approach with its
powerful but restricted reach to the online setting(s) we want to research? With CA
as sociological endeavor, it would possibly be fruitful to focus less on individual
technologies and their organisation, and more on the social fabric that is created
with the help of technologically-mediated communication. Some of the questions
that are raised by current research such as the status of openings may well be fur-
thered by taking groups or networks of people and their use of technology across
platforms as a starting point.
CMC as well as social media are very much a feature of the globalised world
and a necessity for many of its inhabitants. The question therefore arises in how
far the organisation of communication might be more universal than in more tra-
ditional genres. It would be interesting for future research to compare the use of
social media between cultures and languages to answer the question as to whether,
or to what degree, there are local social media practices which have developed
against the backdrop of the specific behaviour patterns from other genres (cf. for
instance Panyametheekul and Herring 2003). This could make the necessary and
minute descriptions of the use of different devices or tools in social media and in
different languages and cultures more relevant for the over-riding questions in
linguistics in general.
References
Anderson, Jeffrey F., Fred K. Beard and Joseph B. Walther

2010 Turn-taking and the local management of conversation in a highly simultane-
ous computer-mediated communication system. Language@Internet 7: article
7. http://www.languageatinternet.org/articles/2010/2804 (last access 15 Oct.
2016).
2015 Networked multilingualism: Some language practices on Facebook and their
implications. International Journal of Bilingualism 19(2): 185–205. doi: 10.1177/
1367006913489198.
Auerbach, David
2014 I built that “so-and-so is typing” feature in chat. And I’m not sorry. http://www.
slate.com/articles/technology/bitwise/2014/02/typing_indicator_in_chat_i_
built_it_and_i_m_not_sorry.html (last access 15 Oct. 2016).
Baron, Naomi S.
2008 Always On: Language in an Online and Mobile World. New York: Oxford
University Press.
Baron, Naomi S.
2010 Discourse structures in instant messaging: The case of utterance breaks. Language
@Internet 7: article 4. http://www.languageatinternet.org/articles/2010/2651
(last access 15 Oct. 2016).

Authenticated
Barron, Anne and Emily Black

2015 Constructing small talk in learner-native speaker voice-based telecollabora-
tion: A focus on topic management and backchanneling. System 48: 112–128.
doi: 10.1016/j.system.2014.09.009.
Beißwenger, Michael
2008a Multimodale Analyse von Chat-Kommunikation. In: Karin Birkner and Anja
Stukenbrock (eds.), Arbeit mit Transkripten in der Praxis: Forschung, Lehre
und Fortbildung, 117–143. Mannheim: Verlag für Gesprächsforschung.
2008b Empirische Untersuchungen zur Produktion von Chat-Beiträgen. In: Tilmann
Sutter and Alexander Mehler (eds.), Medienwandel als Wandel von Inter
aktionsformen—von frühen Medienkulturen zum Web 2.0, 47–81. Wiesbaden:
VS Verlag für Sozialwissenschaften.
Berglund, Therese Örnberg
2009 Disrupted turn adjacency and coherence maintenance in instant messaging
conversations. Language@Internet 6: article 2. www.languageatinternet.org/
articles/2009/2106 (last access 15 Oct. 2016).
Bloomfield, Leonard
1933 Language. New York: Holt.
Bublitz, Wolfram
2017 Oral features in fiction. In: Miriam A. Locher and Andreas H. Jucker (eds.),
Pragmatics of Fiction, 235–263. Handbooks of Pragmatics 12. Berlin/Boston:
de Gruyter.
2011 ‘Three men using our toilet all day without flushing – This may be one of the
worst sentences I’ve ever read’: Quoting in CMC. In: Joachim Frenk and Lena
Steveker (eds.), Anglistentag Saarbrücken 2010 Proceedings, 433–448. Trier:
Collister, Lauren
2011 *-Repair in online discourse. Journal of Pragmatics 23(3): 918–921. doi:
10.1016/j.pragma.2010.09.025.
Crair, Ben
2014 I can see you typing: The most awkward part of online chat. https://newrepublic.
com/article/116268/gchat-typing-indicator-most-awkward-feature-online-
chat (last access 15 Oct. 2016).
Crystal, David
2006 Language and the Internet, 2nd edition. Cambridge: Cambridge University
Press.
Crystal, David
2011 Internet Linguistics: A Student Guide. London: Routledge.
Darics, Erika
2014 The blurring boundaries between synchronicity and asynchronicity: New com-
municative situations in work-related Instant Messaging. International Journal
of Business Communication 51(4): 1–22. doi: 10.1177/2329488414525440.
Drew, Paul, Traci Walker and Richard Ogden
2013 Self-repair and action construction. In: Makoto Hayashi, Geoff Raymond and
Jack Sidnell (eds.), Conversational Repair and Human Understanding, 71–94.
Cambridge: Cambridge University Press.

Authenticated
Dürscheid, Christa and Karina Frick

2014 Keyboard-to-Screen-Kommunikation gestern und heute: SMS und WhatsApp
im Vergleich. In: Alexa Mathias, Jens Runkehl and Torsten Siever (eds.),
Sprachen? Vielfalt! Sprache und Kommunikation in der Gesellschaft und den
Medien. Eine Online-Festschrift zum Jubiläum von Peter Schlobinski, 149–
181. (Networx 64.). http://www.mediensprache.net/networx/networx-64.pdf
(last access 15 Oct. 2016).
Pragmatics 43: 814–827. doi: 10.1016/j.pragma.2010.09.018.
2014 The pragmatics of monologue: Interaction in video blogs (Doctoral disserta-
tion). http://scidok.sulb.uni-saarland.de/volltexte/2014/5895/pdf/dissertation_
frobenius_scidok.pdf (last access 15 Oct. 2016).
Frobenius, Maximiliane and Richard Harper
2015 Tying in comment sections: The production of meaning and sense on Face-
book. Semiotica 204: 121–143. doi: 10.1515/sem-2014-0081.
1998 The interactional organization of computer mediated communication in the
college classroom. Qualitative Sociology 21(3): 299–317. doi: 10.1023/A:
1022146620473.
1999 The eyes of the beholder: Understanding the turn-taking system in quasi-syn-
chronous computer-mediated communication. Research on Language and
Social Interaction 32: 337–367. doi: 10.1207/S15327973rls3204_2.
Garfinkel, Harold
1967 Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.
Gibson, Will
2009 Intercultural communication online: Conversation analysis and the investiga-
tion of asynchronous written discourse. Forum: Qualitative Sozialforschung/
Forum: Qualitative Social Research 10(1): article 49. http://www.qualita-
tive-research.net/index.php/fqs/article/view/1253/2731 (last access 15 Oct.
2016).
Goffman, Erving
1979 Footing. Semiotica 25 (1/2): 1–30. doi: 10.1515/semi.1979.25.1-2.1.
Google Help
2015 https://support.google.com/answer/1083134?hl=en (last access 31 Dec. 2015).
Greenfield, Patricia and Kaveri Subrahmanyam
2003 Online discourse in a teen chatroom: New codes and new modes of coher-
ence in a visual medium. Journal of Applied Developmental Psychology 24(6):
713–738. doi: 10.1016/j.appdev.2003.09.005.
Harrison, Sandra
2003 Repair in email discussions. Discourse Analysis Online http://extra.shu.ac.uk/
daol/articles/closed/2003/004/harrison2003004-01.html (last access: 15 Oct.
2016).
Have, Paul ten
1999 Structuring writing for reading: Hypertext and the reading body. Human Stud-
ies 22: 273–298. doi: 10.1023/A:1005496619385.

Authenticated
Have, Paul ten

2000 Computer-mediated chat: Ways of finding chat partners. M/C: A Journal of
Media and Culture 3(4). http://journal.media-culture.org.au/0008/partners.
php (last access: 15 Oct. 2016).
Heath, Christian and Paul Luff
2000 Technology in Action. Cambridge: Cambridge University Press.
Herring, Susan
1999 Interactional coherence in CMC. Journal of Computer Mediated Communica-
tion. doi: 10.1111/j.1083-6101.1999.tb00106.x.
Herring, Susan C.
Internet 4: 1–37. http://www.languageatinternet.org/articles/2007/761 (last
access 15 Oct. 2016).
Herring, Susan C.
2010a Computer-mediated conversation: Introduction and overview. Language@
internet 7. http://www.languageatinternet.org/articles/2010/2801/ (last access
15 Oct. 2016).
Herring, Susan C. (ed.)
2010b Computer-mediated conversation. Special issue of Language@Internet, 7.
http://www.languageatinternet.org/articles/2010. (last access 15 Oct. 2016).
Hockett, Charles
1958 A Course in Modern Linguistics. New York: Macmillan.
Honeycutt, Courtenay and Susan Herring
2009 Beyond microblogging: Conversation and collaboration via Twitter. In: Pro-
ceedings of the Forty-Second Hawai’i International Conference on System
Sciences. Los Alamitos, CA: IEEE Press.
Hutchby, Ian
2001 Technologies, texts and affordances. Sociology 35(2): 441–456. doi: 10.1177/
S0038038501000219.
Hutchby, Ian and Vanita Tanna
2008 Aspects of sequential organization in text message exchange. Discourse and
Communication 2: 143–164. doi: 10.1177/1750481307088481.
Ito, Mizuko and Daisuke Okabe
2005 Technosocial situations: Emergent restricting of mobile e-mail use. In: Mizuko
Ito, Misa Matsuda and Daisuke Okabe (eds.), Personal, Portable, Pedestrian:
Mobile Phones in Japanese Life, 257–276. Cambridge, MA: Massachusetts
Institute of Technology Press.
Jacobs, Jennifer and Angela Garcia
2013 Repair in chat room interactions. In: Susan Herring, Dieter Stein and Tuija
Virtanen (eds.), Pragmatics of Computer-Mediated Communication, 565–587.
(Handbooks of Pragmatics 9.) Berlin/Boston: de Gruyter.
Jones, Rodney H.
2004 The problem of context in computer mediated communication. In: Philip
LeVine and Ron Scollon (eds.), Discourse and Technology: Multimodal
Discourse Analysis, 20–33. Washington, D.C.: Georgetown University
Press.

Authenticated
Jones, Rodney H.
2013 Rhythm and timing in chat room interaction. In: Susan Herring, Dieter Stein
and Tuija Virtanen (eds.), Pragmatics of Computer-Mediated Communication,
489–513. Handbooks of Pragmatics 9. Berlin/Boston: de Gruyter.
2012 The linguistics of keyboard-to-screen communication. A new terminologi-
cal framework. Linguistik online 56(6): 39–64. http://www.linguistik-online.
org/56_12/juckerDuerscheid.html (last access 15 Oct. 2016).
Laursen, Ditte
2005 Please reply! The replying norm in adolescent SMS communication. In: Rich-
ard Harper, Leysia Palen and A. Taylor (eds.), The Inside Text: Social, Cultural
and Design Perspectives on SMS, 53–73. Dordrecht: Springer.
Laursen, Ditte
2012 Sequential organization of text messages and mobile phone calls in intercon-
nected communication sequences. Discourse & Communication 6(1): 83–99.
doi:1 0.1177/1750481311432517.
Lazaraton, Anne
2014 Aaaaack! The active voice was used! Language play, technology, and repair
in the Daily Kos weblog. Journal of Pragmatics 64: 102–116. doi: 10.1016/j.
pragma.2014.02.002.
2012 Video-in-interaction: “Talking heads” and the multimodal organization of
mobile and Skype video calls. Research on Language and Social Interaction
45(4): 399–429. doi: 10.1080/08351813.2012.724996.
2014 Mundane video directors in interaction. Showing one’s environment in skype
and mobile video calls. In: Mathias Broth, Eric Laurier and Lorenza Mondada
(eds.), Studies of Video Practices. Video at Work, 135–160. Abingdon: Rout-
ledge.
Linell, Per
2005 The Written Language Bias in Linguistics: Its Nature, Origin and Transforma-
tion. London: Routledge.
Lyons, John
1981 Language and Linguistics. Cambridge: Cambridge University Press.
Marcoccia, Michel
internet newsgroups. Journal of Pragmatics 36, 115–145. doi: 10.1016/S0378-
2166(03)00038-9.
Markman, Kris
2005 To send or not to send: Turn construction in computer-mediated chat. In:
Twelfth Annual Symposium about Language and Society, 115–124. Austin:
Texas Linguistic Forum.
Markman, Kris
2009 ‘So what shall we talk about’: Openings and closings in chat-based virtual
meetings. Journal of Business Communication 46(1): 150–170. doi: 10.1177/
0021943608325751.

Authenticated
Markman, Kris
2013 Conversational coherence in small group chat. In: Susan Herring, Dieter Stein
and Tuija Virtanen (eds.), Pragmatics of Computer-mediated Communication,
539–564. Handbooks of Pragmatics 9. Berlin/Boston: de Gruyter.
Meredith, Joanne and Elizabeth Stokoe
2014 Repair: Comparing Facebook ‘chat’ with spoken interaction. Discourse &
Communication 8(2): 181–207. doi: 10.1177/1750481313510815.
Panyametheekul, Siriporn and Susan C. Herring
2003 Gender and turn allocation in a Thai chat room. Journal of Computer-Medi-
ated Communication 9(1). doi: 10.1111/j.1083-6101.2003.tb00362.x.
Raclaw, Joshua
2008 Two patterns for conversational closings in instant message discourse. Colo-
rado Research in Linguistics 21: 1–21.
Reeves, Stuart and Barry Brown
2016 Embeddedness and sequentiality in social media. In: Proceedings of the 19th
ACM Conference on Computer Supported Cooperative Work and Social Com-
puting, 1052–1064. San Francisco: Association for Computing Machinery.
Rintel, Sean E. and Jeffrey Pittam
1997 Strangers in a strange land: Interaction management on Internet Relay Chat.
Human Communication Research 23(4): 507–534. doi: 10.1111/j.1468-
2958.1997.tb00408.x.
Rintel, Sean E., Jeffrey Pittam and Joan Mulholland
2003 Time will tell: Ambiguous non-responses on Internet Relay Chat. Elec-
tronic Journal of Communication 13(1). http://www.cios.org/EJCPUB-
LIC/013/1/01312.HTML (last access: 15 Oct. 2016).
Sacks, Harvey
1995 Lectures on Conversation. Ed. by Gail Jefferson. Oxford: Blackwell.
Sacks, Harvey, Emanuel Schegloff and Gail Jefferson
1974 A simplest systematics for the organization of turn taking for conversation.
Language 50(4): 696–735. doi: 10.2307/412243.
Saussure, Ferdinand de
1964 Cours de Linguistique Générale. Paris: Payot.
Schandorf, Michael
2012 Mediated gesture: Paralinguistic communication and phatic text. Conver-
gence: The International Journal of Research into New Media Technologies
19(3): 319–344. doi:1 0.1177/1354856512439501.
Schegloff, Emanuel
1995 Introduction. In: Harvey Sacks, Lectures on Conversation. Ed. by Gail Jeffer-
son, ix–lxii. Oxford: Blackwell.
Schegloff, Emanuel, Gail Jefferson and Harvey Sacks
1977 The preference for self-correction in the organization of repair in conversation.
Language 53: 361–382. doi: 10.2307/413107.
Schegloff, Emanuel and Harvey Sacks
1973 Opening up closings. Semiotica 8(4): 289–327. doi: 10.1515/semi.1973.8.4.289.
Schönfeldt, Juliane and Andrea Golato
2003 Repair in chats: A conversation analytic approach. Research on Language and
Social Interaction 36(3): 241–284. doi: 10.1207/S15327973RLSI3603_02.

Authenticated
Seedhouse, Paul
2004 Conversation analysis methodology. Language Learning 54:1–54. doi:
10.1111/j.1467-9922.2004.00268.x.
Sellen, Abigail
1992 Speech patterns in video‒mediated conversations. In: Proceedings of the ACM
CHI 92 Conference on Human Factors in Computing Systems Conference,
49–59. New York: Association for Computing Machinery.
2014 Through the looking glass: A social semiotic and linguistic perspective on the
study of video chats. Text and Talk 34(3): 325–347. doi:1 0.1515/text-2014-
0006.
Skovholt, Karianne and Jan Svennevig
2013 Responses and non-responses in workplace emails. In: Susan Herring, Dieter
Stein and Tuija Virtanen (eds.), Pragmatics of Computer-mediated Communi-
cation, 581–603. Handbooks of Pragmatics 9. Berlin/Boston: de Gruyter.
Spagnolli, Anna and Luciano Gamberini
2007 Interaction via SMS: Practices of social closeness and reciprocation. British
Journal of Social Psychology 46: 343–364. doi: 10.1348/014466606X120482.
Spilioti, Tereza
2011 Beyond genre: Closings and relational work in text messaging. In: Crispin
Thurlow and Kristine Mroczek (eds.), Digital Discourse: Language in the
New Media, 67–85. (Oxford Studies in Sociolinguistics) Oxford: Oxford Uni-
versity Press.
Pragmatics 40: 1587–1600. doi: 10.1016/j.pragma.2008.04.018.
Trapani, Gina
2010 What is Google Wave? http://www.macworld.com/article/1146555/business/
whatisgooglewave.html (last access 15 Oct. 2016).
Tudini, Vincenza
2014 Conversation analysis of computer-mediated interactions. In: Carol
A. Chapelle (eds.), The Encyclopedia of Applied Linguistics, 1–7. Hoboken,
NJ: John Wiley & Sons. doi: 10.1002/9781405198431.wbeal1456.
Vachek, Josef
1949 Some remarks on writing and phonetic transcription. Acta Linguistica 5:
86–93.
Watson, Rod
2009 Analysing Practical and Professional Texts: A Naturalistic Approach. Farn-
ham: Ashgate.
Werry, Christopher C.
1996 Linguistic and interactional features of Internet Relay Chat. In: Susan Herring
(ed.), Computer-Mediated Communication: Linguistic, Social and Cross-Cul-
tural Perspectives, 47–63. Amsterdam/Philadelphia: Benjamins.

Authenticated
Authenticated
11. Discourse and topic
Elisabeth Fritz
Abstract: This chapter focuses on the question of how the characteristics of

social media interactions influence topic management. These influences will be
described on two levels: First, local strategies of topic development are appro-
priated by the users of social media from face-to-face practices to the particular
online environment. Second, with a view to whole exchanges, participants develop
larger patterns of divergence and convergence of sub-threads on topics which
exhibit various degrees of relevance to the superordinate topics of the particular
exchanges. The chapter starts with providing a preliminary survey of linguististic
approaches to discourse topics and topic development. This is followed up by a
discussion of the most prominent characteristics of social media interactions (viz.,
(a)synchronicity, polyloguicity and monotopicality), which challenge frameworks
of interactional topic development drawn up for traditional (i.e., mostly dyadic and
spontaneously evolving) face-to-face conversations. The discussion closes with an
overview of existing research concerning strategies of topic initiations and local
topic development as well as more global patterns across different forms of syn-
chronous and asynchronous discourse.
1. Introduction
Topicality or topical coherence plays a fundamental role in the construction and

perception of coherence in discourse (Brown and Yule 1983: 68; for coherence
and cohesion in social media cf. Schubert, Ch. 12, this volume). It is rooted in our
everyday understanding that any stretch of discourse is generally ‘about some-
thing’, and interactants are usually able to give topic summaries such as We were
talking about our plans for the weekend.
Discourse topics pertain to the ideational level of discourse. Nonetheless, they
are essentially understood to be a product of interpretation and collaborative nego-
tiation among participants – “an achievement of conversationalists” (Maynard
1980: 263). As a rule, topics need not be stated explicitly and are not mechanically
derivable from linguistic surface structures, which obviously constitutes a major
difficulty for linguistic analysis and has led to a variety of frameworks and topic
definitions (see Section 2). Still, speakers (and writers) do guide topical under-
standing by linguistic means, such as prominent referents, means of lexical and
grammatical cohesion, discourse markers and other kinds of metadiscursive items
to indicate topical organization, e.g. to come back to what I was saying earlier (cf.,
DOI 10.1515/9783110431070-011
Authenticated
276 Elisabeth Fritz
e.g., Lenk 1998; Keller 1979; Tiittula 1993). Interactants develop topics dynam-
ically and often spontaneously. As recipients, they display their topical orienta-
tion by continuously checking new contributions against their understanding of
the topical framework at hand. In the roles of speakers, they construct their own
contributions so that they are tied and made relevant to previously established
topics.
Discourse topics can be identified and described for different stretches of dis-
course, which brings forth complex interplays of local and global topics of various
scopes, levels of specificity and interrelations. At a local level, topics rarely stay
the same over a stretch of discourse but are continuously expanded, shifted and
changed. Focusing on content relations between individual topics and the com-
municative strategies of topical organization, linguistic research on face-to-face
interactions has described general patterns of local topic development, such as
topic introductions, closings, shifts and changes. From a more global perspective,
however, patterns of topic development can differ widely and depend very much
on the type of discourse at hand, and it may or may not be possible to subsume
different local topics under a global topic.
Studying topic management in interaction is thus both an important as well as
challenging task. However, probably due to the inherent difficulty in identifying
topics and topical boundaries, topic-focused linguistic studies are relatively scarce.
Most of the existing studies concentrate on particular types of interactions taking
place in a face-to-face setting. While they are thus able to identify and describe
strategies of topic constitution, shifts, changes etc., these strategies are necessarily
shaped by their respective communicative setting, and there is little research con-
centrating on differences across different types of face-to-face discourse.
Social media interactions pose a further challenge both to the negotiation of
topical orientation for the participants as well as to the linguistic frameworks aim-
ing to describe them. The impact of the computer-mediated setting on the organi-
zation of discourse has been researched extensively and approached from a large
number of angles. Yet very few studies concentrate specifically on describing the
patterns and strategies of interactive topic management in online interactions and
how they have been adapted from face-to-face practices to fit the respective tech-
nological environments of the various platforms. Those that do (cf. Section 4) are
(necessarily) limited in focus, both in terms of their operationalization of discourse
topics as well as their application to one specific platform and its characteristic
communicative setting.
This chapter will provide an overview of specific patterns in local and global
topic management across the various forms of text-based multiparticipant online
interactions. To this end, it will both draw on the findings from existing topic-cen-
tred frameworks as well as demonstrate how some of the technological functions
specific to particular types of platforms are regularly appropriated by users in the
service of topic management. Before doing so, however, I will first give a sur-

Authenticated
Discourse and topic 277
vey of general directions and foci of topic research, which saw its heyday in the
1970s and 1980s, i.e., before the rise of computer-mediated communication. The
subsequent discussion of prominent characteristics of social media interactions –
subsumed as synchronicity, polyloguicity and monotopicality – will point out
how each of these factors has considerable impact on their topical organization.
Although multiparticipant settings and monotopical orientation are not in them-
selves specific to online communication, it is the clustering of the three factors that
characterizes and shapes the predominant part of public social media interactions
and thus poses fundamental challenges to the study of topic management. None-
theless, some of the concepts and approaches that have been developed to tackle
such complexities in multiparticipant and/or monotopical face-to-face interactions
may also prove useful for the description of topical organization online.
2. Perspectives on discourse topics and topic management in

traditional forms of communication
Linguistic research on discourse topics can be roughly divided into two groups
of approaches, which adopt distinct perspectives and methodologies: Structural
approaches, on the one hand, take on a product-oriented perspective and under-
stand topics as a structuring unit of discourse, which can be described for stretches
of discourse and put into identifiable relations with each other. On the other hand,
dynamic approaches adopt a procedural view on discourse. Rather than character-
izing individual topics, they concentrate on describing the procedures and activi-
ties interactants engage in when they manage topics in conversation (cf. Brinker
and Hagemann 2001: 1252).
Structural topic approaches are typically found in structuralist text and dis-
course analysis, and most prominently in research with a focus on text and dis-
course grammars, cognitive planning or artificial intelligence1 (e.g., van Dijk 1977;
1980; Brinker, Cölfen and Pappert 2014; Gardner 1987; Schank 1977; Reichman
1978, 1990; Hobbs 1990). While they do concur that discourse topics are not static,
independent entities that can be directly derived from the linguistic surface struc-
ture, structural topic analysis feeds on semantic and/or referential links between
topical units, so that the description of cohesive ties between discourse segments
constitutes a basic part in their approach to topical organization. Discourse topics
themselves tend to be explicitly defined as a representation of semantic content.
Some models adopt very distinct standpoints on whether discourse topics should
1
Cf. Sirois and Dorval (1988) for an extensive overview of different theories of dis-
course and their respective resulting views on discourse topics.

Authenticated
278 Elisabeth Fritz
be conceptualized as a proposition (e.g., van Dijk 1977, 1980; Ochs Keenan and
Schieffelin 1976; Hobbs 1990) or a prominent referent (e.g., Fritz 1982; Schank
1977). Propositional theories (particularly van Dijk 1977, 1980) are often supple-
mented with an intricate framework of more or less rigid inference mechanisms.
Example (1) shows van Dijk’s formal representation of his deletion rule:
(1) Ken Holland saw a blonde. She was wearing a white summer frock … DELETION
=> (entails) Ken Holland saw a blonde
(adapted from van Dijk 1980: 78)
DELETION
Given a sequence E of propositions (Pi, Pi+1, …, Pk) or a text T, satisfying the normal
linear coherence constraints, substitute E by a sequence E’ such that each Pi+je E that
is not an interpretation condition (presupposition) for at least one proposition of T does
not occur in E’, whereas E and E’ are further identical.
(van Dijk 1980: 82)
However, such strict topic rules and formulations have been repeatedly criticized
for neglecting the inherently interpretive nature of topics (cf. Brown and Yule
1983: 114–116). Fritz (1982, 1994) and Bublitz (1988), among others, plead for
a freer formulation of discourse topics, arguing that it should be adaptable to the
communicative purpose of a topic summary and the intended degree of specific-
ity and granularity (cf. Bublitz 1988: 20–22). In this vein, in example (2), both
(a) and (b) can count as adequate topic summaries, with the latter being the most
general formulation and as such applicable to any sort of discourse (cf. Fritz 1994:
192):
(2) (a) We have been considering which films to go and see and in that
connection d has told us an amusing story about a visit to the Biograph and then
A and I have been complaining about our local cinema.
(b) We have been talking about films.
(adapted from Bublitz 1988: 21–22)
Topic frameworks are (necessarily) closely tied to the kind of data they aim to
describe. (Prototypical) written discourse is planned and monological; spoken dis-
course evolves spontaneously and in interaction. The central implication for topic
analysis is that clear topic trajectories are unlikely to be found in discourse with
a certain degree of spontaneity and interactive involvement. Consequently, it is
those models which are developed with a view to analyzing written monological
texts that tend to suggest mechanisms and rules for extracting topics from the text
and combine them into hierarchical structures (most prominently van Dijk 1977,
1980; Brinker, Cölfen and Pappert 2014).
In contrast, frameworks describing topical structures in dia- and polylogal dis-
course must allow for unplanned, and in fact unpredictable, topic development.
In the course of the interaction, topics may be brought up which are anchored

Authenticated
in the participants’ common ground,2 including their shared physical surround-

ings (cf. Adato 1980), or utterances may be spontaneously re-interpreted to create
connections to new topics. Nevertheless, local (chains of) topic transitions can
be described in terms of semantic or referential links, and structural approaches
focused on topic development in spoken interactions draw on these connections to
identify topical units and describe transition patterns between them (e.g. Hobbs’
(1990) coherence relations).
It is particularly in this context that dynamic approaches complement a struc-
tural perspective. Dynamic approaches to topic management heavily feed on eth-
nomethodological principles advocated by conversation analysis, which is con-
cerned with describing formally identifiable techniques and procedures by which
interactants “exhibit […] attention to topic” (Sacks 1995: 755) rather than defining
topics in semantic terms. A central distinction is that between topically bounded
and unbounded stretches of talk,3 with stretches of bounded talk exhibiting more or
less discernible disturbances in the smooth flow of speaker change as well as in the
continuity of referents (Maynard 1980: 263–264). Specifically with regard to topi-
cal organization, conversation analysts have employed a multitude of terms to refer
to various patterns of disjunctive topical transitions across clearly disparate (i.e.,
bounded) episodes and more subtle topic transitions that do not display clear dis-
ruptions or disfluencies, e.g.: stepwise moves or stepwise topic transition vs. topical
shift (Jefferson 1984, 1983), topic shading vs. topic bounding (Schegloff and Sacks
1973), pivotal topic transitions (Holt and Drew 2005, cf. also Drew and Holt 1998),
topic change (Maynard 1980), or topic generation (Button and Casey 1984, 1985).
The two perspectives with their different foci and methods are not incompat-
ible and have been fruitfully combined in a number of frameworks (e.g., Bublitz
1988; Korolija and Linell 1996; Linell and Korolija 1997; Linell 1998 for conver-
sational data).4 These frameworks account for topic development both in terms of
2
Common ground is here understood in very broad terms as “a mental representation of
knowledge and attitudes assumed to be shared” (Bublitz 2006: 370) among interactants
(cf. also Lee 2001: 27).
3
The distinction refers to the presence or absence of signals with which participants
separate more or less distinct stretches of talk. Such signals usually include disruptions
to the referential chains (Sack’s tying structures, cf. Sacks 1995: 540) and/or to the
interactants’ negotiation of turns. The typical bounding techniques introducing a possi-
ble boundary are pre-closings (Weeell, so, ok, etc.), i.e., turns, which “occupy the floor
for a speaker’s turn without using it to produce either a topically coherent utterance or
the initiation of a new topic” (Schegloff and Sacks 1973: 304), aphoristic formulations,
or final assessments (Schegloff and Sacks 1973; Holt and Drew 2005; Drew and Holt
1998; Jefferson 1983; West and Garcia 1988).
4
Note, however, that while the concept of topic bounding overlaps with the more seman-
tic distinction between topic shifts and changes (cf. below), they feed on different cri-
teria and are not fully congruent.

Authenticated
280 Elisabeth Fritz
identifying units of talk on a topic and describing their local and global relations
semantically as well as in terms of interactional topic management manifest in for-
mal procedures of collective topical organization. Taxonomies tend to differentiate
between patterns of topic development at local and global levels and of various
degrees of complexity, such as topic (re-)introductions, closings, shifts, changes,
digressions and drifts, with most research focusing on a section of these patterns.
In broad terms, Bublitz’ (1988) classification offers an overarching frame-
work of the (increasingly complex) topical action patterns introduction, closure,
shift, change and digression, focusing on the strategies with which interactants
collaboratively negotiate attention to topics. Based on propositional links as well
as strategies of sequential organization of spontaneous conversation among inti-
mate couples, Crow (1983) identifies similar patterns (maintaining, coherent and
noncoherent shifts, renewals and inserts), which are, however, categorized dif-
ferently.5 Focusing more strongly on the logical/semantic relationships between
the individual topics, Ochs Keenan and Schieffelin’s (1976) basic distinction
between continuous (with collaborating and incorporating discourse topics) and
discontinuous discourse (encompassing re-introducing and introducing discourse
topics) is well-known and taken up repeatedly (see Section 4). Much more fine-
grained frameworks such as Korolija and Linell (1996; also Linell and Korolija
1997) and Reichman (1978, 1990) distinguish according to the nature of the links
holding between consecutive topical units. While the former’s classification is
based on where the new topic referents are anchored (i.e., in the local co-text,
the participants’ shared physical situation or ongoing activity, etc.), Reichman’s
approach is distinctly hierarchical, and many of her categories (e.g., generaliza-
tion, illustrative and restatement, subissue and joining or total shift relations) are
reminiscent of van Dijk’s (1977, 1980) macro-rules. With a more restricted focus
5
Maintaining refers to topical continuity as well as Bublitz’ topic closures; Crow’s coher-
ent shifts are “[t]he structurally preferred mechanism for topic change within a conver-
sation” (Crow 1983: 141) in the sense that they “will not occasion a request for topical
coherence clarification” (Crow 1983: 143). They are subdivided into initiations and
topic shading, which correspond to Bublitz’ “regular” topic changes and shifts respec-
tively as well as the classic topic bounding and topic shading transitions described
by conversation analysts. Renewals are reintroductions of previously discussed topics.
They involve linguistic tokens such as anyway, getting back to the subject etc., and
are consistent with the last move in Bublitz’ larger digression pattern. A noncoherent
shift is an “abrupt [and propositionally completely unconnected] shift that succeeds
in getting the topical floor” (Crow 1983: 146). This matches Bublitz’ description of
topic changes involving the break-off of the prior topic, which, however, seems to be
so marked in spontaneous conversations amongs friends and acquaintances that this
case remains purely hypothetical in his study (cf. Bublitz 1988: 134). Last, the insert
describes abrupt shifts which do not succeed in getting collaboratively established as
topics.

Authenticated
on semantically related transitions, local topic shift rules or coherence relations

(see Section 4) are described by Schank (1977) and Hobbs (1990) respectively. In
Hobbs’ framework, such locally-shifted segments combine to larger segments with
coordinated or subordinated topical constituents (labelled parallel and chained
association respectively). The re-establishment of a previous topic in the wake of
such associations is then called a topic return or a topic pop; however, if such a
topic is not re-established after a series of shifts, this is labelled a topic drift.
Dascal and Katriel (1979) and Pons Bordería and Estellés Arguedas (2009)
particularly focus on the complex pattern of digressions, describing the nature of
the intervening topics and approaching the question of whether or not digressive
patterns differ from a “simple” return to a prior topic after intervening topical
talk. Digressions are often also implicitly or explicitly touched upon in discussions
that declare to target the discontinuous topic changes, and particularly those to a
prior topic. In this way, Sirois and Dorval (1988) differentiate returns according
to the “distance” of this topic, while Planalp and Tracy (1980) classify changes
according to where the newly-established topics are drawn from (conversation vs.
nonconversation). Adato (1979) describes linguistic signals accompanying antici-
pated and unanticipated topic returns.
Strict conversation analytic approaches generally focus on the description of
the more local patterns. Terminologically, they do not always distinguish between
topic shifts and changes, but identify and describe topic transitions in talk by means
of the sort of bounding signals involved in achieving that transition. In this way,
Maynard (1980) distinctly speaks of topic changes in the case of topic initiations
after failed speaker change, while West and Garcia (1988) classify a broader range
of transitions, with their topic transitions after prior topic extinction correspond-
ing to Maynard’s topic changes. Their collaborative topic transitions, on the other
hand, describe what Jefferson (1984) calls stepwise topic transitions, i.e., transi-
tions via aphoristic conclusions, arrangements and other topic-bounding activi-
ties, which would be labelled shifts rather than changes based on the similarity of
successive topics. Employing an even narrower focus on the individual strategies
of introducing new topics, Button and Casey (1984, 1985) have brought forth a
detailed classification of topic generating sequences, which occur at places where
the topical floor is open (either because the previous topic has been exhausted or
because the first topic of the interaction has not yet been established). Speakers
can then either nominate or elicit a new topic. Topic closing procedures are most
frequently described in the context of disjunctive, or bounded, topic transitions
rather than as patterns in their own right. Understood as initiating steps of these
transitions, they are indicative of the participants’ endeavour to get ready to move
on from the present topic towards the next. Schegloff and Sacks (1973) and Button
(1987) explicitly describe closing patterns that (at least potentially, i.e., if partici-
pants do not choose to move out of the closings, cf. Button 1987) open up the slot
for producing final interaction-terminating sequences.

Authenticated
282 Elisabeth Fritz
Acknowledging that interactants orient not only to what they talk about but
also to what they do by talking about it, most approaches to topic development
more or less explicitly supplement their semantic/referential topic analysis with
a pragmatic perspective on participants’ activities and associated speech actions.6
This helps to account for the fact that a progression of local topics may often be
controlled or guided by the ongoing communicative activity and its action trajec-
tory rather than by any evident relation between individual topics (e.g., topically
disparate question-answer pairs in a quiz show, distinct points on a topical agenda
in a business meeting, etc.).
Interactive topic development is to a considerable degree shaped by situative
factors such as the communicative activity or purpose of the exchange, existence
of a pre-determined topic, number of participants and participation structure, and
the setting in which the exchange takes place. Within “traditional” spoken interac-
tion in a face-to-face setting or over the telephone, there is room for considerable
variation: Discourse topics and their development tend to play a different role in
topically-oriented institutional talk than in spontaneously evolving informal con-
versations. In the former case, a larger topical agenda restricts the topical range
of the interaction and acts as a reference point for both participants and observers
to assess the relevance of locally-evolving topics; in the latter, the lack of such an
overarching topical framework may facilitate moments of faltering topic transi-
tions and generally direct the interactants’ attention more towards local topic pro-
gression and management. Institutionalized discourse tends to feature a clearer dis-
tinction of participant roles as, e.g. teachers, students or moderators, who assume
different responsibilities in negotiating topic progression. Additionally, the degree
of explicitness with which interactants themselves state or announce topics and
signal transitioning moves, e.g., by means of discourse markers, explicit meta-
communicative comments, can differ considerably along these lines. Most studies
on topic management in spoken discourse concentrate on one particular type of
interaction and often restrict themselves to the analysis of individual aspects and
patterns of topic development, which are, in turn, often adapted to fit the material
at hand.
3. Major factors governing topic management in social media
Thus far, studies concentrating specifically on strategies and patterns of topic man-
agement in computer-mediated communication are relatively scarce, which may
be due to their inherently interpretive nature and the accompanying difficulties
6
This is probably most noticeable in Bublitz’ (1988) approach, who even incorporates
this aspect in his definition of topical actions.

Authenticated
concerning their operationalization for analysis. While this in itself is a source of

considerable debate in linguistic research on traditional forms of communication
(see Section 2), the various forms of interaction on social media platforms exhibit
various new combinations of technological and situational factors (Herring 2007:
13–23). These present unprecedented challenges both for participants to negotiate
topical coherence in interaction as well as for the analyst to identify and describe
reconfigured (cf. Herring 2013b) discursive strategies of topical organization post
festum.
There is little debate nowadays that computer-mediated discourse in general
is stylistically heterogeneous and exhibits characteristics from both conceptual
orality and literacy.7 As discourse on social media is interactively constructed
by a number of different participants, research on its inner workings usually take
face-to-face conversations as a frame of reference rather than (conceptually and
modally) written discourse.8
The most distinctive factors shaping the organization of topical discourse
across social media platforms are 1) the (quasi-)synchronicity or asynchronicity of
the exchange, 2) the potentially high number of participants interacting with each
other, facilitating complex patterns in terms of participatory structures, and 3) the
7
The distinction of conceptual orality and literacy originates from Koch and Oesterrei
cher (1985). Disentangling the various readings and implications of the commonplace
opposition of spoken vs. written language, they propose a model of communicative
immediacy and distance. At the heart of this lies the conceptual as well as terminological
differentiation between the realization of language and its conceptualization. In terms
of the former, they differentiate between the graphic and the phonic mode, i.e. the writ-
ten vs. spoken format of representation (cf. Bublitz 2017 for this definition of mode).
Conceptually, they propose a continuum running from the poles of conceptually spo-
ken language (or the language of immediacy) to that of conceptually written language
(or the language of distance). This conceptual dimension encompasses both situative
communicative features of the communicative act (such as, e.g., dialogue/monologue,
familiarity of the interactants, temporal and spatial co-presence/separation, presence/
absence of ad hoc speaker change, degree of involvement, free/pre-determined choice
of topic) as well as the various strategies of linguistic expression related to these param-
eters (e.g., provisional/definitive character, degree of information density, compact-
ness, planning). Especially in the context of newly-evolved forms of communication,
this model has been repeatedly criticized and extended (e.g., Dürscheid 2003; Landert
and Jucker 2011; Landert 2014). In order to sharpen the terminological distinction,
Bublitz (2017) proposes to differentiate between conceptual orality vs. literacy on the
conceptual dimension, and orality vs. scripturality at the level of realization.
8
A notable exception to this is Largier (2002), who explicitly regards her German and
French asynchronous forum postings as realizations of conceptually written discourse.
She also observes that the frequent occurrence of a message-initial and final greeting
section is clearly reminiscent of norms of conceptual literacy (Largier 2002: 292; cf.
also Herring’s (1996) electronic message schema).

Authenticated
284 Elisabeth Fritz
fact that acceptable topic development is frequently delimited by pre-determined

global topics.
This section will give an overview of each of these factors’ impact on global
discourse organization, showing how their interplay creates new sets of techno-
logical and situational conditions that render traditional approaches of interaction
management partially unsuitable and in need of modification (cf. Dürscheid and
Brommer 2009: 17; Bou-Franch, Lorenzo-Dus and Garcés-Conejos Blitvich 2012;
Marcoccia 2004).
3.1. Synchronicity
A key aspect in differentiating computer-mediated communication from face-to-
face interactions is that interactants do not need to be physically co-present. This
spatial separation, combined with the early predominance of text-only (or at least
text-based) messages, has been in the focus of linguistic research since its begin-
ning and led to detailed investigations into emerging phenomena and strategies
such as, e.g., emoticons, non-standard spellings, or use of capitalization to com-
pensate for the lack of paralinguistic and non-verbal cues (e.g., Thurlow, Lengel
and Tomic 2007: 125; Baym 2015: 68–69).
One of the most prominent distinctions across the various forms of comput-
er-mediated communication is concerned with synchronicity, i.e. the temporal
dimension of online interactions. In synchronous exchanges, participants are only
spatially separated; they are logged on into the system (e.g., a chat client) simul-
taneously and are thus temporally co-present, which facilitates quick exchanges in
real-time (or at least with minimal delays). In contrast, in asynchronous comput-
er-mediated communication participants do not need to be online at the same time
and can retrieve and react to messages at a later point.
In its simplest form, this distinction is understood as a technological prop-
erty and used to differentiate between synchronous and asynchronous systems,
such as chat or email, discussion forums, etc. respectively (e.g., Herring 2007:
13; Beißwenger 2007: 17–31). This, however, has been criticized and further dif-
ferentiated with regard to matters of temporal organization in the actual ensuing
discourse. On the one hand, chat platforms require participants to be logged on
at the same time. In this light, Beißwenger (2007: 37) and Jucker and Dürscheid
(2012: 39; cf. also Cherny 1999: 154) have argued that true simultaneity of pro-
duction and reception is only facilitated in a 2-way transmission format, when
users’ contributions are transmitted keystroke by keystroke instead of being sent as
a whole. In such an understanding, most chat systems commonly called synchro-
nous only facilitate quasi-synchronous exchanges. On the other hand, so-called
asynchronous platforms such as discussion forums, mailing lists etc., may well
be designed to afford asynchronous communication but can also accommodate
quasi-synchronous exchanges of rapidly sent consecutive messages between users

Authenticated
who are simultaneously oriented to the ongoing interaction (Dürscheid and Brom-
mer 2009: 6; Beißwenger 2007: 22; Baym 2006: 523; Jucker and Dürscheid 2012:
39).
The convergence of technical devices (Jucker and Dürscheid 2012: 37) seen
in recent years acts as a further complicating factor: As mobile phones turn into
hand-held computers continuously connected to the internet and medium-size
portable devices become commonplace, people can access the various platforms
whenever they want, receive notifications and respond immediately without hav-
ing to wait until they can access a stationary computer. Software providers react by
developing smartphone and tablet applications to make their web-based communi-
cation platforms easier to navigate from mobile devices. These technical advances
and accompanying shifts in practice can even lead to a conflation of tradition-
ally separate forms of communication (cf. Herring 2013b). This can be observed
in texting and instant messaging, for instance, whose main differentiating factor
was the kind of device used to engage in it (mobile phones with 12-key keypads
vs. stationary computers with a full keyboard) – in combination with resulting
restrictions such as character limits etc. Both are now continued in the form of
newer internet-based messenger services that work on both stationary and mobile
devices, the latter of which have seen major changes to their text input hard- and
software. In a similar vein, the once predominant deeply hierarchical structure of
asynchronous forums has gradually been reduced to a facultative display option
or even completely given up in favour of a purely chronological arrangement of
postings, which appears to be more mobile-friendly.
For all these reasons, synchronicity is best understood as a continuum (Baron
2008: 15) of synchronous, quasi-synchronous and asynchronous forms of com-
munication. By and large, however, bearing in mind that true synchronicity in the
sense of simultaneity is not facilitated by most text-based platforms, a rough cate-
gorization into synchronous and asynchronous forms of computer-mediated com-
munication still seems justifiable. Also, the functional range the software affords
is essentially geared towards its intended (quasi-)synchronous or asynchronous
use, and those platforms meant to facilitate asynchronous interaction can also host
quasi-synchronous exchanges.
Synchronicity has been shown to play a decisive role in how interactants design
and organize their contributions linguistically. As speed is particularly important
in multiparticipant synchronous environments, these contributions tend to be
shorter and (lexically and syntactically) less complex (e.g., Herring 2001: 617;
Beißwenger 2003: 225) – as compared to asynchronous exchanges, where individ-
ual contributions can be more complex and can accommodate various functional
and topical moves (Condon and Čech 2010; Black et al. 1983; Herring 1999). In
terms of conceptual differences manifest in the linguistic organization of individ-
ual contributions, the difference between (quasi-)synchronous and asynchronous
exchanges can be traced along the lines of the continuum of communicative imme-

Authenticated
286 Elisabeth Fritz
diacy according to Koch and Oesterreicher (1985) (Dürscheid and Brommer 2009:
7; Androutsopoulos 2007: 87; Jucker and Dürscheid 2012: 39).9
Concerning the interactional organization across messages, both synchronous
and asynchronous interactions are prone to exhibiting what Herring (1999) calls
“disrupted turn adjacency”, i.e. the tendency for interactionally and topically
unrelated turns to “interrupt” sequences of related messages. Tracking individual
co-occurring exchanges, then, is a “cognitively challenging” task (Herring 1999),
which becomes even more demanding as interactions fragment into various par-
allel threads on different topics, leading to persistent claims that “topics decay
quickly” (Herring 1999; also Lambiase 2010; Herring and Nix 1997).10 In both
synchronous and asynchronous communication, participants cannot fully rely on
the linear order of messages determined by the time of transmission in order to
trace topically and interactionally coherent sequences. In asynchronous comput-
er-mediated communication, the relative lack of time constraints in combination
with the persistence of messages additionally opens up the possibility for users to
contribute responses to temporally distant messages, thus creating webs of multi-
linear exchanges (Gruber 1998: 36, 2013: 65).
It thus becomes a particularly important strategy for creating coherence to
overtly signal links to previous contributions. Common verbal strategies, such as
addressing particular participants, paraphrasing or repeating (parts of) previous
contributions, are typically directed at people and/or the subject matter at hand
and are clearly adopted from established communicative practices of topic and
turn management in face-to-face interactions (cf. Schubert, Ch. 12, this volume).
Others are particular to computer-mediated communication and involve “semi-au-
tomatized” (Bublitz and Hoffmann 2011: 438) practices, most prominently the
quoting templates in asynchronous forms such as discussion forums and email
and the software-based positioning of messages in hierarchical tree structures in
some discussion forums or comment sections on social media platforms. Research
into the communicative use of such linking devices thus can constitute a first step
9
Dürscheid (2003: 47) rightly points out that, contrary to some conceptualizations, the
differentiation of conceptual orality and literacy cannot be attributed to the particular
forms of communication (such as email, chat, etc.) as such, but can only be applied to
particular genres emerging within the various platforms (cf. also Kilian 2001). In this
vein, a love email (or text message, etc.) will be more conceptually immediate, or oral,
than one sent in a business context (Dürscheid 2003: 47).
10
In contrast to this, Crystal (2006: 151) suggests that “the vast majority of messages […]
stay surprisingly on-topic”, whereas Stomer-Galley and Martinson (2009) arrive at a
more nuanced conclusion and claim that the pre-determined global topic, together with
the “seriousness” of the interaction determines whether or not participants tend to drift
from the original topic, and that technologically-conditioned factors are not “inherently
problematic” (Stromer-Galley and Martinson 2009: 195; see Section 4.4).

Authenticated
to approaching patterns of collaborative topic management in and across various

platforms (see Section 4.3).
3.2. Polylogal organization

Interaction on social media typically involves a large number of individuals. It
has been noted repeatedly that, with the “convergence between the public and the
private sphere” (Chovanec and Dynel 2015: 10) resulting from the technological-
ly-conditioned permanence and reproducibility of content, it is impossible to know
even the size of one’s audience. The blurring along the lines of audience and par-
ticipants necessitates modifications (Marcoccia 2004: 131–143; Arendholz 2013a:
44; Chovanec and Dynel 2015) of traditional frameworks of participation formats
established for face-to-face encounters in order to account, inter alia, for lurk-
ers, who inadvertently or intentionally read exchanges without actively participat-
ing. Once they start to contribute, however, they can easily turn into ratified and
addressed participants (Arendholz 2013a: 44). Conversational groups are dynamic
and “continuously under construction” (Marcoccia 2004: 140). The sheer amount
of people actively engaging in an interaction opens the door for intricate and mul-
tifarious patterns of interleaving and more or less separable exchanges. In doing
so, it poses considerable challenges both for the ongoing interactional organization
as well as an adequate linguistic description of these structures.
Fragmentation in itself is not unknown from the study of face-to-face commu-
nication. Polylogues are characterized by their increasing “flexibility, instability,
and unpredictability […] on all levels of their functioning” compared to dyadic
configurations (Kerbrat-Orecchioni 2004: 7). Placing fewer constraints on indi-
vidual participants in terms of the mechanics of turn taking and speaker selection,
they exhibit more frequent and varied instances of interruptions and simultaneous
talk (Kerbrat-Orecchioni 2004: 5–6). In interactions involving four or more partic-
ipants, these instances are prone to function as bifurcation points triggering what
Parker (1984) calls conversational grouping, i.e., the fragmentation into parallel
exchanges between smaller groups. A conversation can thus split up into two or
more simultaneously proceeding threads, which, in turn, can compete with each
other, quiet down and/or converge into new configurations, giving rise to a vast
variety of different dynamically evolving interactional patterns. In example (3),
Sara’s prompt in line 5 leads to a fragmentation into two parallel exchanges. The
competition is short-lived and ends with Elsa’s turn (lines 12–14), in which she
closes down the ongoing topic and refocuses on Sara’s original attempt at a topic
introduction. She thereby initiates a topic change towards a topic to which all par-
ticipants then contribute in a single shared conversational exchange:

Authenticated
288 Elisabeth Fritz
(3) 1 Elsa so we saw in any case that although topics could recur if you like
2 we saw differences [ uh in:::
3 Max [hm hm
4 (silence 3s)
5 Sara but uh I return to [the example of this morning\=
6 Max [yes that Léa =yes yes because it’s a good
7 is (inaud.) of the first example/
8 frame Sara yes
9 it will be a (inaud.) Léa (taking notes) emerging
10 Elsa [topical [simply but all the frame (silence) inside/
11 Max [yes [yes
12 Elsa rest has changed [
13 [yes
14 all the rest has changed yes
(2s) what did you say↑ this
morning/
15 Sara yes↑ this morning/ I arrived at half past 9\ and also uh::: uh came/
[ and at half past
16 Inès [yes
17 Sara 9 I came for a polylogue meeting/ [ and I came into the office
18 Elsa [yes↑
19 Sara here there was already an ongoing dialogue (.) between uh:
20 Léa and: – and Anne/ (Xs) then uh: well after a while – I stayed/ (.) after a
while
21 I went out because for me/ it was – we were – I was no longer in
22 the participation framework poly – uh polylogue meeting [
23 Elsa [yes↑
24 Inès [that’s it
(Traverso 2004: 71)
Such structures form more or less discrete organizational units at an intermedi-

ate, or “macro-local” (Traverso 2004: 56), level of discourse organization. This
level is located below the global level of the interactive encounter as a whole and
above that of local turn management. Structures at this macro-local level emerge
as “islands of organization and regularity” (Kerbrat-Orecchioni 2004: 7) and con-
stitute a fruitful middle way to identify and describe structures of coherence in
polylogal interaction. They have been described under the term of conversational
floors (Edelsky 1981; Shultz, Florio, and Erickson 1982; Hayashi 1991; Coates
1997; Traverso 2004; Jones and Thornborrow 2004; Simpson 2005, 2013), and
there are overlaps with other concepts from interactional discourse analysis such
as phases (and subphases), episodes, etc. (Korolija and Linell 1996; Linell and
Korolija 1997; Linell 1998: 203; Pavlidou 2014). The concept of floor is “notori-
ously slippery” (Simpson 2013: 518), and definitions vary. At their core, however,
floors are taken to be an interactional achievement of the participants, reflect-

Authenticated
ing their understanding of “what’s-going-on within a psychological time/space”

(Edelsky 1981: 405). As such, individual floors are characterized by their rela-
tively unified participation framework,11 as well as a shared communicative pur-
pose or activity and topic. Thus, they constitute a valuable springboard for the
analysis of collaboratively negotiated topics and topic management in polylogal,
interactionally complex, discourse.
As with topic-focused approaches, conversational floors can be studied from
structural and dynamic perspectives: The structural, or “cross-sectional” (Traverso
2004: 54), perspective yields descriptions of various, more or less abstract and
‘typical’ floor types: Edelsky (1981) identifies two such types, viz. one-at-a-time,
singly-developed type F1 and collaboratively-developed type F2. The former is
characterized by orderly turn-by-turn talk with little overlap, a smaller number of
speakers taking (comparatively) longer turns, a single topical focus and an over-
all more formal and hierarchical organization. Conversely, F2 types evolve in a
more egalitarian fashion, with a larger number of central participants, diverse top-
ics, shorter turns (and shorter overall duration) and a larger extent of overlapping
speech (cf. also Herring 2010: 3–4). Accounting for the possibility of simulta-
neously developing floors in polylogal interactions, Shultz, Florio and Erickson
(1982) incorporate the distinction of multiple and single floors into their frame-
work, which has been taken up in various taxonomic modification and extensions
(e.g., Hayashi 1991; Traverso 2004; Cherny 1999; cf. Simpson 2013 for a concise
overview). In the dynamic, or “longitudinal” (Traverso 2004: 55), approach, the
focus lies on describing the dynamics in which floors evolve, split off and con-
verge. It is closely connected to patterns identified for topic management such as
topic shifts, changes, digressions etc.
Floors are, essentially, topically bounded units; transition points between floors
typically exhibit the discontinuity markers identified by conversation analysis to
indicate topic boundaries (Traverso 2004: 61; see example (3)) – so that shifts in
participation statuses and/or ongoing communicative activity are likely to indicate
transitions in topic. Here again, structural and dynamic perspectives complement
each other.
Social media constitute ideal platforms to foster “massive polylogues” (Bou-
Franch and Garcés-Conejos Blitvich 2014: 19). Their tendency to fragment into
various threads or sub-threads has been widely noticed (e.g., Marcoccia 2004; Bou-
11
Participation framework, in this context, does not only refer to different participation
statuses in terms of Goffman’s (1981) framework (such as primary/secondary speaker,
addressed/unaddressed recipients etc.) but also extends to participant statuses based on
institutional and social roles vis-a-vis the overall interaction (such as moderator, teacher
vs. student etc.), and the ongoing communicative activity (e.g. asking for advice vs.
proffering advice), as well as those discursive roles related to local turn management
(e.g., giving the floor) (Kerbrat-Orecchioni 2004: 16).

Authenticated
290 Elisabeth Fritz
Franch, Lorenzo-Dus and Garcés-Conejos Blitvich 2012). As messages are per-

sistent and constantly retrievable on most – and especially asynchronous – CMC
platforms, fragmentation and multiple floors “are in fact easier to achieve than in
face-to-face conversations” (Cherny 1999: 180; cf. also Herring 1999), and it can
even be rather effortless for one participant to contribute extensively to two or more
simultaneously developed floors. Contrastively, in face-to-face settings, the ability
of interactants to monitor and actively engage in multiple parallel floors at the same
time is severely constrained by the ephemeral (spoken) nature of the participants’
contributions and the simultaneity of the exchange (Traverso 2004: 72).
Conversational floors have been used to conceptualize such mid-level units
in the organization of polylogal CMC interactions (Cherny 1999; Simpson 2005,
2013; Herring 2010), particulary with a view to the relative conversational domi-
nance of one or more individual participants. Other studies have brought forth tax-
onomies of similar types of patterns under different analytic foci. These include,
e.g., instances of disrupted vs. linear messages sequences, exchanges between
dyads or (smaller) sub-groups and larger groups, orientation toward different
communicative activities or patterns of linking back to specific prior contribu-
tions or participants vs. addressing and orienting towards a larger audience, which
may be accomplished with the help of software-afforded functions (e.g., Largier
2002; Schütte 2004; Holmer 2008; Bou-Franch, Lorenzo-Dus and Garcés-Conejos
Blitvich 2012; Bou-Franch and Garcés Conejos-Blitvich 2014; Haugh and Chang
2015). Discourse topics and their dynamics in interaction are rarely focused on
in any explicit manner, but seem to play an important role in the development of
such mid-level organizational structures.12 These structures can serve as a starting
point for more detailed investigations into local and global patterns of topic man-
agement in social media polylogues.
3.3. Monotopicality and off-topic behaviour

Perhaps most directly connected to matters of dynamic topic management is the
question of whether or not a communicative encounter is set out to cover a particu-
lar topic and/or serve a particular purpose. Depending on the presence or absence
of such an agenda, ensuing interactions can be classified as monotopical or polyto-
12
Regarding the formation of subgroups in synchronous MUDs (Multi-User Dungeons,
i.e., a text-based multiparticipant virtual world incorporating, inter alia, chat inter-
faces), Cherny (199: 176) claims that they “cannot be formed with physical proximity
or volume control, so if they do form, they must be purely topic driven”. In Herring’s
(2010) adaptation of floor analysis to academic mailing lists, aspects of topical organi-
zation feature in her category of thematic focus (i.e., measured as messages on main or
sub-themes) as well as, within her participation category, as “distribution of new topic
initiations across participants” (Herring 2010: 10).

Authenticated
pical respectively, with the terms referring to a projection of the acceptable topical
range rather than the actual outcome:
With the term ‘monotopical’ we intend not an ex post facto finding that a single topic
was talked about, especially in view of the complexity with which topic talk is done,
wherein each successive utterance can revise what the topic has been ‘all along’. We
have in mind, rather, conversations produced from their beginnings with an orientation
to their expectable monotopicality. (Schegloff and Sacks 1973: 307)
Lacking such an orientation, spontaneously evolving everyday conversations have

typically been attested a “relative lack of [topical] order, which cannot be foreseen
(or easily directed)” (Bublitz 1988: 40); topic changes, shifts and digressions are
common and do not in and of themselves constitute sanctionable behaviour. Topics
may develop by means of creative and unpredictable associations, and this can be
an obstacle to the analytical identification of topic boundaries.
In contrast, in monotopical interactions the shared focus on a global topic places
severe constraints on norm-conforming topical management. Although tangential
topics may de facto ensue, which temporarily shift the attention away from the main
topic, they are, in face-to-face environments, usually short-lived and marked as
digressions or as subordinate side floors (cf. Bublitz 1988: 94–129; Traverso 2004).
Although the distinction may certainly leave some leeway, most (asynchro-
nous) polylogal communication on social media, at the level of whole exchanges,
is intended to be monotopical. Synchronous chat in topically undefined chatrooms
– as far as the notion of a bounded ‘single exchange’ is applicable – seems to be
the most perspicuous exception.
Irrespective of their mediated or unmediated nature, the distinction of global
vs. local topics gains additional momentum in monotopical interactions. Gener-
ally, global and local topics may be defined by means of conversational scope
alone (e.g., Bublitz 1988: 35; Linell 1998: 183). In a monotopical setting, how-
ever, the global topic is pre-determined to span the entirety of the interaction;
it acts as a point of orientation for participants both in the construction of their
own contributions as well as for interpreting and judging the topical nature of
those made by other interactants. It is open for the duration of the interaction, and
in spoken conversations, its closure typically triggers “the initation of the clos-
ing section” (Schegloff and Sacks 1973: 308) terminating the whole interaction.
Locally-instantiated topics are as a rule expected to contribute in some way to the
global topic at hand. In this sense, the differentiation of local and global topics can
be extended by a notion of hierarchy in addition to their scope.
The differentiation of on-topic and off-topic behaviour is particularly promi-
nent in monotopically-oriented social media interactions. Platforms usually pro-
vide an explicit code of conduct, which normally includes the request for members
to stay on topic – a direct reminder of monotopicality. This clearly affects the
overall organization of topical management, as locally-instantiated topics are not

Authenticated
292 Elisabeth Fritz
only developed and interpreted with reference to their connection to sequentially

preceding topics, but are also continually assessed with respect to their relevance
for the global topic at hand.
However, from an analytic point of view, clear criteria and boundaries for on-
and off-topic exchanges are nearly impossible to establish. What falls within the
range of a particular global topic remains just as flexible and dependent on inter-
pretation and reinterpretation as the notion of discourse topic itself. Detailed mod-
els are rare and strongly dependent on how they operationalize discourse topics.
Simple semantic relations, as suggested, e.g., in Stromer-Galley and Martinson
(2009), however, may not do justice to the complexity of perceptions of relevance.
As seen before, topics and communicative actions are closely related in orient-
ing participants in an ongoing exchange (Linell 1998: 182). Monotopical interac-
tions are tied to a more or less clearly laid out purpose, which entails a particular
trajectory of relevant communicative actions and preferred responses (cf. Haugh
and Chang 2015). Most global topics such as politics, auto racing or cancer sup-
port (Stromer-Galley and Martinson’s (2009) examples for global topics), already
imply a purpose or range of relevant action patterns (e.g., discussion/argumenta-
tion, exchanging opinions or information, connecting to likeminded people, giving
and receiving support etc.). Especially in asynchronous exchanges, where mes-
sages can be longer, the initiating postings have been shown to regularly include
more or less explicit calls for particular actions (Arendholz 2013a: 148–156; see
Section 4.2). In this way, topics and communicative actions complement each
other, and messages may still be perceived as relevant to the overall exchange
when they are recognizable as being relevant to either the global topic or the com-
municative activity determined for the interaction.
It follows that the distinction between on-topic and off-topic behaviour should
best be conceived of as gradual. From the analyst’s point of view, the safest judge-
ments can be made in cases where participants display their perception that an
ongoing exchange has gone off-topic in their communicative behaviour. While
“silence”, or more accurately the absence of responses to particular contributions,
may not be intentional in this respect (Markman 2009: 157–158), explicit ver-
bal strategies can be less ambiguous. They include, most prominently, metacom-
municative utterances commenting on the lack of relevance to the global topic
(Lambiase 2010: 13). But participants may also display their discontent with the
relevance of the ongoing local topic less explicitly by engaging in recognizable
topic closing procedures adapted from face-to-face settings.
Whether or not an interaction is set out as monotopical places considerable
constraints on its development both in terms of discourse topics as well as commu-
nicative actions. Thereby, projected monotopicality impacts on the collaborative
organization of the exchanges as a whole. From a dynamic point of view, off-topic
exchanges are mostly the result of patterns of topic drift rather than abrupt changes,
which tend to occur more frequently at a point in time when the interaction has

Authenticated
been going on for a while, so that the global topic is exhausted and participants are
losing interest in continuing it (Lambiase 2010; see Section 4.4).
4. Topic management in multiparticipant social media interactions
This section will review linguistic research into topic management in multipartic-
ipant social media interaction, taking into account both approaches and findings
from synchronous (IRC and public web-based chatrooms, MUD systems, instant
messaging services) and asynchronous forms (mailing lists, web-based discussion
forums and their predecessors such as usenet newsgroups, and comment sections
on YouTube, news websites, etc.).
Setting off with an overview of predominant operationalizations of the notion
of discourse topic and topic progression, the focus will be narrowed onto recurrent
strategies of topic initiations across various platforms. With respect to stepwise
local topic transitions across individual contributions, the technological resources
afforded by the software to provide cross-message links play a prominent role in
shaping how users manage sequential and topical coherence. After exploring their
potential, the section will conclude with an overview of linguistic research on
local and global patterns of topic development in public multiparticipant comput-
er-mediated discourse.
4.1. Tendencies in researching discourse topics in social media

Only few studies focus specifically on matters of topic management in social
media. Of those, most concentrate on the phenomenon of topic drift. The clear
majority studies synchronous interaction (Herring and Nix 1997; Herring 2003;
Hoffmann 2004; Herring and Kurtz 2006; Stromer-Galley and Martinson 2009).
Topic investigations into asynchronous interactions are under-represented (e.g.,
Gruber 1997, 1998; Lambiase 2010).
These topic-focused approaches (with the exception of Hoffmann 2004) draw
on two topical frameworks in particular, viz. Ochs Keenan and Schieffelin (1976)
and Hobbs (1990). Both of them encompass relatively specific definitions of dis-
course topics as propositions as well as taxonomies for local topic transitions:
Hobbs’ model of topic drift describes various strategies of stepwise topic transi-
tions between utterances (termed parallelism, explanation and metatalk) on the
basis of how they relate to each other.13 Performed over a series of utterances,
13
Parallelism is based on similarity of argument or proposition, explanation refers to a
causal relation, and metatalk transitions have to do with a shift towards the goals of the
conversation (Hobbs 1990: 8–9).

Authenticated
294 Elisabeth Fritz
such transitions can steer the conversation far away from the original topic. Ochs
Keenan and Schieffelin (1976: 341–342) cover a broader spectrum and make a
four-fold distinction between the general categories of continuous and discontin-
uous discourse, which each cover two transition types based on the topical simi-
larity between utterances.14 These frameworks seem to lend themselves more eas-
ily to adaptation for computer-mediated environments than conversation-analytic
approaches. This is because they define discourse topics and their boundaries by
means of propositional content, whereas conversation analysis has always been
more cautions with content analysis and instead focuses on organizational pro-
cedures that strongly hinge on simultaneity. Viewing topics as a collaborative
achievement (Lambiase 2010: 5), the transferability at this local level also relies
on short message length. As a general tendency, this makes chat more accessible
than asynchronous interaction.
A major issue for topic analysis in computer-mediated environments is that
adjacent turns tend to get disrupted by (topically and/or interactionally) unrelated
turns (cf. Herring 1999). Topical and sequential coherence typically go hand in
hand; indeed, Schegloff (1990: 64) observes that “the sequential structure [can]
provide […] the basis for finding some topical linkages across what are, at the
surface, topically unrelated and noncoherent utterances.” As such, an analysis of
dynamic topic development crucially depends on tracing sequences of contribu-
tions, which are often non-adjacent in the chronological order of the interactions.
Synchronous and asynchronous platforms typically differ in both the complexity
of such sequences and the range of linking strategies that participants can draw
upon. While some of them are practices adopted from turn management strategies
in face-to-face conversations, others (especially on asynchronous platforms) rely
on functions afforded by the software. Once such sequences are established, the
models usually take referential ties as the basis for describing patterns of topic
development.
Sequences of topically related contributions are commonly referred to as
threads (Herring 1999; Crystal 2006: 142; Lambiase 2010: 5; Markman 2013:
545). Synchronous platforms usually do not provide any formal means to indi-
cate such sequences; they have to be signalled and interpreted by means of the
verbal content of the contributions alone, and the complex interplay of multiple
threads can become “confusing” (Cherny 1999: 182).15 Asychronous platforms,
14
Continuous discourse subsumes collaborating and incorporating discourse topics, with
collaborating topic matching that of the preceding utterance and an incorporating topic
sharing an element from the preceding topic. Discontinuous discourse, in contrast, is
characterized by introducing or re-introducing discourse topics, depending on whether
or not the topic has been active in more distant utterances of the ongoing discourse
(Ochs Keenan and Schieffelin 1976: 341–342).
15
There are a number of software projects concerned with visualizing interactional pat-

Authenticated
on the other hand, differ widely in terms of their specific functions, but they usu-
ally provide some technological resources to structure exchanges by arranging
messages into hierarchical sequences, which are intended to visualize local topic
development. They have the potential to make coherent exchanges more effortless,
in the sense that users can outsource a large part of their linking work. However,
they are also susceptible to human error (Gruber 1997: 109; Marcoccia 2004: 123–
126) and/or strategic misuse (Lambiase 2010: 10). As asynchronous interactions
are typically monotopically oriented towards a global topic, they are usually as
a whole referred to as threads (and can, in turn, develop into several local (sub-)
threads). In contrast to chatrooms, these threads are usually explicitly started by a
thread-initiating message, which lays out the global topic and the action trajectory
that all responding messages are supposed to relate to.
4.2. Topic initiations

Establishing a discourse topic is a collaborate achievement and thus necessar-
ily involves at least two messages by two different participants (e.g., Lambiase
2010: 5). In the potentially noisy environments of social media, users do not com-
pete so much for the right to speak as to make themselves heard (Cherny 1999:
174), i.e., to receive responses to their postings. Consequently, the first message in
topic initiations is particularly important. Users regularly self-select to start new
topics (Markman 2013: 549), and unilateral topic nominations (and profferings; cf.
Schegloff 2007: 169–180) generally seem to be much more promising a strategy
than topic elicitations, whereby recipients are invited to propose a topic (Schegloff
2007: 170; cf. Button and Casey 1984, 1985). Linguistic patterns are adapted from
face-to-face conversations and are similar across synchronous and asynchronous
platforms, but the placement of the postings and their complexity is heavily influ-
enced by the synchronicity and intended monotopicality of the exchange.
4.2.1. Topic initiations in synchronous chat

Local topic initiations can occur throughout the chat, even more so if it is not
intended as monotopical (cf. Herring and Nix 1997). Pre-arranged chat sessions
typically exhibit opening (and closing) phases similar to those of face-to-face or
telephone conversations. Markman (2009), in her investigation of informal virtual
terns, based on message links, participation and/or temporal structure etc., both as a tool
for analysis in retrospect (e.g., Herring and Nix 1997; Herring 2003; Herring and Kutz
2006; Donath, Karahalios and Viégas 1999; Donath 2002; Holmer 2008; Severinson
Eklundh and Rodriguez 2004 for an asynchronous platform) as well as bringing struc-
ture to synchronous exchanges while they are taking place (e.g., Donath, Karahalios
and Viégas 1999; Donath 2002).

Authenticated
296 Elisabeth Fritz
meetings, observes a “two-stage process” (Markman 2009: 155) by which partici-

pants habitually move into the topical stage of the meetings of a fixed group, con-
sisting of “(a) an opening move referencing prior communication from the team
and (b) an agenda-setting turn focusing the talk on a specific topic” (Markman
2009: 155). While the first stage is frequently introduced by the discourse marker
so (e.g., So, guess you found some neat people according to Dr Kimble’s notes …;
Markman 2009: 156; also 2013: 550), the second stage is typically “phrased as
a first-pair part (questions, assessments) that offer[s] a specific candidate for the
current meeting’s discussion” (Markman 2009: 156), e.g., what areas are we all
interested in researching? (Markman 2009: 160). It is thus this second stage that
functions as a move to initiate the first (and often global) topic.
This strategy of opening the main (topical) phase is characteristic of meetings
in general and not in itself specific to the chat setting. What seems to be a direct
consequence of the quasi-synchronous setting, however, is the ease with which
openings can be interrupted, so that it can take several attempts to complete the
two stages and open the meeting (Markman 2009: 161).
Such clearly delineated opening phases are naturally absent from open cha-
trooms, in which users continually log on and off (Schönfeldt 2001: 48). For those
who have recently logged in, topic initiations constitute an easily accessible entry
point for them to join in (cf. Hoffmann 2004: 107).
In terms of linguistic strategies, directives in the form of interrogative clauses
(i.e., Leech’s (2014) rogatives) are the primary means by which new topics are
initiated in chats, and they are typically (implicitly or explicitly) directed at every-
one (Hoffmann 2004: 109; Markman 2013: 547). As they establish conditional
relevance, they are a particularly effective means to trigger responses. Assertives
are clearly less effective in this respect, but may be successful, particularly when
participants use them to offer assessments which clearly invite contrastive assess-
ments (Hoffmann 2004: 111).
4.2.2. Thread initiations on asynchronous platforms

In contrast to chats, asynchronous multiparticipant interactions are as a rule ori-
ented towards a particular global topic and formally separated from other threads.
Consequently, individual thread-initiatiating messages can be clearly identified.
The asynchronous setting encourages longer messages,16 which means that post-
16
Nonetheless, messages tend to stay relatively short: In her corpus of 50 discussion
forum threads, e.g., Arendholz (2013b: 360) measures a mean of 171.82 words in the
thread-initiating messages, with longer specimens containing around 550 words. As
there is no obligation whatsoever for users to pick up on potential thread initiations,
overly long messages may be felt to overburden prospective respondents and are thus
avoided.

Authenticated
ings can be more complex and elaborate on the envisioned topical scope to make it
interesting and relevant for users to respond. How this is achieved is influenced by
a multitude of social factors, most prominently the communicative purpose of the
thread and the projected actions as well as the characteristics of the group as such,
e.g., user demographics, interests and expertise on particular subjects.
Similarly to chats, directive speech acts realized by interrogative or impera-
tive clauses are common strategies to elicit responses from the recipients (Gruber
1997: 114, 1998: 30; Arendholz 2013a: 150) and are at the core of topic initiations.
Arendholz’ (2103a) study of interaction in a web-based discussion forum indicates
that this is also reflected in the thread headings, where interrogative structures
make up more than half (29 of 50) of the threads under investigation (Arendholz
2013a: 139). Headings are particularly powerful means of attracting attention on
these platforms, as threads are listed by their headings and are only displayed in
full once they are clicked on. They serve as “condensed abstracts” (Arendholz
2013a: 128) of the thread-initiating postings, often indicating both intended topi-
cal range and communicative purpose.
The way in which the global topic is introduced and elaborated on is primarily
influenced by the communicative purpose the authors have in mind: Most prom-
inently, the global topic is built around a central (personal or otherwise) prob-
lem, about which authors seek advice or opinions, or simply wish to vent their
anger. Personal problem expositions are performed by means of narratives, which
are typically concluded by an evaluation if users mean to “blow […] off steam”
(Arendholz 2013a: 156). Advice, on the other hand, is elicited implicitly or explic-
itly through interrogative or imperative clauses, which can explicitly address the
community (Arendholz 2013a: 154). Thread initiations that aim for other people’s
opinions rather than advice equally employ interrogatives to elicit them. In con-
trast to the other categories, the topic need not be linked to the author’s personal
life. If it is not based on personal experience, authors frequently draw on thread- or
forum-external sources to establish its relevance and build up common ground for
discussion (Arendholz 2013a: 149–155; Gruber 1997: 116).
While a distinction between different communicative purposes is important, it
should be noted that this is often not as clear-cut as one might suppose. Example
(4) demonstrates that the trajectory of the directives includes both opinions and
narratives of personal experiences. In fact, it has been frequently observed that
communicative activities such as troubles telling, seeking advice and complaining
frequently conflate in online interactions (Smithson et al. 2011; cf. Haugh and
Chang 2015: 104–108 for a succinct overview). However, they can be differenti-
ated and exhibit different trajectories of preferred responses.
(4) Is the UK sleepwalking into a surveillance state? Do we need more adequate safe-
guards? Have you been snooped on?
(adapted from Arendholz 2013a: 155)

Authenticated
298 Elisabeth Fritz
Which topics exactly users find interesting to talk about is highly dependent on
interests and purposes shared by the community. Gruber (1997, 1998) observes that
the central points of interest in academic mailing list communities seem to centre
on theoretical and empirical issues. Either are usually common and accepted foci
for responding messages to pursue. For this reason, he argues that the global topic
laid out in thread-initiating postings can best be regarded as a set of related (more
specific) topics (Gruber 1997: 115). Arendholz’ finding that authors often ask mul-
tiple questions (e.g., example (4)) is consistent with this interpretation. Global
topics are relatively broad and can be interpreted as a range of more specific topics
for respondents to choose from.
Thread-initiating messages compete with each other, and not all messages turn
out to be successful. Marcoccia (2004: 121), e.g., notes that 50 % of thread-initi-
ating messages to a French newsgroup within a two-month period fail to trigger
responses; in contrast to this, Gruber (1998: 32) finds “only a few” such postings
on two academic mailing lists within four month. Both hypothesize that the lack
of responses may be (at least partly) attributed to the linguistic presentation of the
topic. Just as in chat interaction, responses may be inhibited by the failure to pro-
duce first pair parts (Marcoccia 2004: 122). Also, presenting the topic in a way that
does not make the problematic issue evident, e.g. by phrasing their exposition in
“unmodalized positive polarity” (Gruber 1998: 33), seems to have a similar effect,
as readers are not offered particular “possible points [or topics] of discussion”
(Gruber 1998: 33) and may feel reluctant to disagree with an account that has not
been marked as problematic. Other factors that may negatively influence users’
propensity to respond include extensive message length (Gruber 1998: 33) and the
possibility to get back to the author on other channels (i.e., through private email
etc.; Marcoccia 2004: 122).
4.3. Quoting and threading in the service of topic development

Asynchronous systems usually provide specific technological options for partic-
ipants to indicate sequential patterns of the ongoing exchange, viz., most prom-
inently quoting templates and a threading mechanism of some sort. They com-
plement the array of ‘manual’ linking strategies, such as paraphrasing, direct
references to authors etc., which users have at their disposal throughout synchro-
nous and asynchronous platforms (cf. Severinson Eklundh and Rodriguez 2004 for
an overview of such linking strategies; see also Section 4.4).
Threading and quoting can be considered semi-automatized strategies in the
sense that a considerable part of the linking work is taken over by the software.17
17
The distinction between non-, semi- and fully-automatized is originally introduced by
Bublitz and Hoffmann (2011) with reference to quoting in CMC. However, it can be

Authenticated
But they differ in their degree of optionality and granularity, as threading (if fea-
tured by the software) is usually obligatory, while the use of quoting is optional.
With threading, the software generates hierarchical thread structures based on the
participants’ choices in marking their messages as replies to specific prior con-
tributions to the thread. With quoting, the system will automatically generate a
quote of the selected message, i.e., adding a verbatim repetition of the message
body, formally separated from the new contribution (e.g. displayed as a box or
bubble) and complemented by a pointer to the original message. Participants can
usually manipulate this quote, e.g. by selecting parts of the quote and deleting
others, segmenting the original message into multiple quotes or selecting content
from other messages to include as additional quotes (by means of copy-and-past-
ing) (cf. Bublitz and Hoffmann 2011: 439; Bös and Kleinke 2015: 74). This high
degree of adaptability makes quoting a particularly versatile strategy,18 which can
be employed to serve an array of functions going beyond simple indication of
sequential message order.
On a broad scale, semi-automatized quoting practices can induce coherence as
well as carry out relational functions (Bös and Kleinke 2015; Arendholz 2015; Bub-
litz 2015). These main functions are, however, not always clearly separable in prac-
tice; Bös and Kleinke (2015: 78) suggest viewing this distinction as a continuum,
with quoting patterns “typically […] display[ing] a complex functional profile”.
The coherence-inducing function of quoting is probably most prominent and
has been most widely discussed in linguistic literature (e.g. Severinson Eklundh
and Rodriguez 1994; Herring 1999; Severinson Eklundh and Macdonald 2004;
Severinson Eklundh 2010; Bublitz and Hoffmann 2011). Aiming to disentangle
conceptual ambiguities, Bös and Kleinke (2015) list three coherence-related func-
tions, viz. the text-deictic, the adjacency-creating and the trigger/mental access
function. The text-deictic function (Bös and Kleinke 2015: 79) relates to basic
linking, tying a message to a particular preceding one, which is thus taken as the
departure point for the new contribution. This function can equally be performed
by means of threading. However, quoting is a particularly economic strategy, as by
providing a verbatim repetition of a prior message’s textual material, it invites the
extended to strategies of (thread-internal) linking in a more general sense with little

difficulty.
18
Note that pragmatically-oriented research on quoting in social media emphasizes the
fact that the communicative act of quoting invariably encompasses both the representa-
tion of a source text as well as its (evaluative) re-contextualization on the part of the
quoter (Bublitz and Hoffmann 2011; Arendholz 2015; Bublitz 2015; Bös and Kleinke
2015). This understanding is mirrored conceptually and terminologically in a differ-
entiation between quote (i.e. the inclusion of textual material from another source in a
particular message) and quotation or the act of quoting, i.e. “the combination of quote
and comment” (Bös and Kleinke 2015: 72; also Arendholz 2015: 55).

Authenticated
300 Elisabeth Fritz
use of cohesive ties in the new contribution much more effectively than a simple
link of messages in the thread structure would (cf. Severinson Eklundh 2010: 20).
From the perpective of interactional organization, quoting can contribute to
eliciting adjacency and lending the interaction a flavour of dia- or polyloguicity.
Within a single message, a quote not only creates a link to a prior message, but it
“creates the illusion of adjacency in that it incorporates and juxtaposes (portions
of) two turns, an initiation and a response, within a single message. When portions
of previous text are repeatedly quoted and responded to, the resulting message can
have the appearance of an extended conversational exchange” (Herring 1999).
Severinson Eklundh (2010: 10) reports that (newsgroup) users prefer to crop
the original message and select only a part of it to appear as a quote, i.e. to quote
selectively. Her observation that “there seems to be a tendency to quote elicitative
parts of a message [i.e., most typically a question] as the primary source” (Sever-
inson Eklundh 2010: 10) clearly highlights the adjacency-creating function. Poly-
loguicity (as opposed to dialoguicity) is particularly promoted by the inclusion of
multiple quotes from various messages, interspersed with new comments (Bös and
Kleinke 2015: 80) – a strategy facilitated by the asynchronous setting (cf. Black
et al. 1983). In terms of topic management, this can be a strategy to contribute to
multiple (local) topics (developed by different people) in one message in a trans-
parent and economic way.
The trigger/mental access function builds on this and focuses on the way in
which the quote is constructed by the quoter as an anchor for a reinterpretation and
shift in focus carried out in their comment. In this sense, quotes “initiate […] and
trigger […] processes of metonymic elaboration by providing a conceptual (and
contextual) anchorage point from which the quoter accesses the situation referred
to in the comment” (Bös and Kleinke 2015: 82). This is exemplified in example
(5), where the pope is taken as an anchor point to trigger the target domain the
church , for which G od and sinner are sub-domains:
(5) Bush is a man who killed thousands of people in iraq in the name of terrorism, what the
hell is the pope doing with a murderer. Regards
I thought God loves everyone, including sinners.
(adapted from Bös and Kleinke 2015: 83; the quote is marked in italics)
This last function is clearly an effective strategy in the service of topical work:
Explicitly repeating particular textual material from a preceding message not only
re-establishes this material in the participants’ common ground (cf. Arendholz
2015: 53). It also, particularly in the case of selective quoting, explicitly brings
to the foreground a specific part of the preceding talk and presents this as the
departure point for the following comment: “By selecting a part of the message
to which a response is provided, the sender simultaneously excludes a number of
other possible interpretations of the message and focuses explicitly on the quoted
parts as salient or noticeable” (Severinson Eklundh 2010: 19–20).

Authenticated
Narrowing the focus down in this way, quoters prepare recipients for an
impending shift. This shift can consist in Bös and Kleinke’s metonymic elabo-
ration patterns, in which the focus is transferred to referents inferable from, but
not explicitly mentioned in the quote. But similary, a shift in focus can be accom-
plished by means of what Bös and Kleinke (2015) and Arendholz (2015) describe
as primarily interpersonally-oriented quoting practices – i.e. in comments with
evaluative elements directed either at the proposition(s) and the stance conveyed
in the quote or directly at the author of the quoted material (ad res and ad hominem
in Arendholz’ (2015) terminology).19 The latter is exemplified in (6):
(6) sorry, but i dont see how your coming up with these theories when youve admitted that
youve never spoke to someone whos been home educated??
its called using your imagination fool
(adapted from Arendholz 2015: 63)
While quoting mostly exhibits both relational and coherence-related functional
components to varying degrees, it is the refocusing on an element drawn from
the quote that carries potential for more or less evident shifts in topic. Bearing in
mind that topics are negotiated interactionally, the shift in focus would need to
get picked up on (and thus ratified) by other participants in order to be identifi-
able as an interactively-achieved topic shift. With a view to the enacted partici-
pation structure, the adjacency-creating function is particularly highlighted when
the quoter comments by means of a first pair part directed at the originator of the
quoted material, thus shifting from a polylogal to a dyadic alignment:
(7) I wasn’t home-schooled, but provided I have the time when I’m older I’d like to home-
school my future children (of course allowing them to attend school if they want); not
because I expect them to be bullied or to be agoraphobic, but beause I’m strongly
against the current UK education system and don’t see it changing drastically any time
soon.
Why are you against the UK education system?
(adapted from Arendholz 2015: 65)
As a technology-facilitated practice, quoting stands out from non-automatized
ways of linking messages. It can be regarded as a particularly prominent device
that posters can use in order to signal a shift in focus and, potentially, topic.
Selective quoting further accentuates this by enabling posters to deliberately and
efficiently select the material they consider immediately relevant for their contri-
bution (cf. Severinson Eklundh 2010: 20). Incorporating multiple quotes into a
single message can be a valuable tool for handling multiple local topics in a single
message. Despite their potential, however, there is hardly any specific research
19
By extension, evaluative comments can also be concerned with intergroup relations
of various scopes rather than being directly targeted at individuals (Bös and Kleinke
2015).

Authenticated
302 Elisabeth Fritz
on the impact of these software-facilitated linking functions on interactive topic

management throughout a whole thread.20
In an endeavour to compare threading and quoting functions and find out
which of them provides a more accurate representation of a thread’s topical struc-
ture, Barcellini et al. (2005) look at an email-based discussion centred on software
design. Contrary to the assumed purpose of threading, they find that a representa-
tion of the thread structure based on (pervasively employed) quoting links pro-
vides a better resource “for the reconstruction of thematic coherence” (Barcellini
et al. 2005: 302) than links provided by the threading mechanism. The main reason
for this seems to lie in the greater versatility and specificity of quoting as a linking
strategy (Barcellini et al. 2005: 312).
Consequently, the role of threading for indicating coherence and topic devel-
opment in asynchronous exchanges seems to be smaller than it is often assumed.
This is also supported by observations that, even in cases where threading errors
are made, this does not necessarily affect coherence (Bou-Franch, Lorenzo-Dus
and Garcés-Conejos Blitvich 2012: 512; cf. also Berglund 2009; Marcoccia 2004).
Frequently occurring misplacement of messages seems to indicate that users tend
not to be particularly attentive to the threading structure and resort to quoting if
they feel that explicit linking is important.
Social media platforms seem to have reacted to this tendency by introducing
changes in their linking resources in recent years and now offer a condensed range
of semi-automatized linking strategies. Traditionally fully threaded platforms (i.e.,
most prominently discussion forums) increasingly switch off the threading func-
tion and offer quoting as the only technologically-aided linking strategy. On the
other hand, ‘newer’ social media platforms such as Facebook and YouTube, which
have never featured explicit linking functionality, now provide a reduced threading
structure, which distinguishes between comments to the main post and comments
to a first-level comment by means of simple indentation (cf. also Bou-Franch,
Lorenz-Dus and Garcés Conejos-Blitvich 2012: 509). This may be understood as
a move adapting the technological environment to user practices striving towards
a middle ground between communicative effort and benefit.
20
However, quantitative results on the frequency, position and form of quotes in various
forms of asynchronous interactions are provided in, e.g., Bös and Kleinke 2015; Arend-
holz 2015; Severinson Eklundh 2010 and Barcellini et al. 2005. Quoting tends to be
a widely adopted strategy, with the percentage of messages containing a quote often
exceeding 50 %. Particularly Severinson Eklundh (2010: 13) points out that frequency
and position of quotes are likely to be influenced by the software default option (cf. also
Bös and Kleinke 2015: 75).

Authenticated
4.4. Local and global topic development
As seen before, dynamic topic management in interaction crucially depends on

signalling and reconstructing (sub-)threads of related messages. Software for syn-
chronous interaction, as a rule, does not offer resources to visualize topical threads
and requires users to manually type in their entire contributions. In the absence of
quoting and threading templates, links to preceding messages are established by
means of addressivity (Werry 1996), metacommunicative elements, lexical (and
grammatical) cohesion as well as along the lines of adjacency pairs (e.g., Mark-
man 2013: 545; Holmer 2008: 2; Herring 1999). Much of topic-related research
on computer-mediated discourse has focused on local topic development in chat,
often with the purpose of describing larger patterns and their interdependence with
other factors.
Based on the initial observation that “topics decay quickly” (Herring 1999),
Herring and Nix (1997; also Herring 2003; Herring and Kurtz 2006; Stromer-Gal-
ley and Martinson 2009) adapt Hobbs’ (1990) categorization of local topic shift
patterns and develop a framework for analyzing and classifying moves of topic
development on a local basis. As suggested in Hobb’s model, the main idea is that
sequences of consecutive local transitions result in topic drift, i.e., by means of
“successive small modifications in the topic” (1990: 3), interactants can eventually
end up talking about a topic that bears no resemblance to the one they started out
on.
As this kind of semantic analysis of local topic dynamics relies on linear
sequences of related utterances, the first analytic step is to identify the threads of
related utterances in the chat interaction. For this, recognizable linking devices
(see above) provide a valuable reference point. Since messages in synchronous
exchanges are typically short and “[e]ach utterance [= message] is primarily on
one topic and directed to one person” (Cherny 1999: 160), local topics are typi-
cally operationalized at the “level of chat message” (Stromer-Galley and Martin-
son 2009: 201). With reference to Hobbs’ (1990) proposition-based framework,
each proposition (which is typically co-extensive with a message) is coded accord-
ing to its relation to the “previous proposition or message in the conversation to
which it appears to have been most directly intended to relate” (Herring 2003: 4).
Herring’s Dynamic Topic Analysis directly imports Hobbs’ coherence rela-
tions of parallelism, explanation and metatalk, and adds break and on-topic as
additional coding categories. On-topic relations are created by messages that
continue on the topic without adding new material, predominantly “simple agree-
ments, reactions, rephrasings and clarifications” (Herring 2003: 5). Breaks exhibit
no identifiable relation to preceding propositions and are – provided the mes-
sages receive responses – a strategy to initiate new topical threads and develop
stretches of discontinuous discourse in Ochs Keenan and Schieffelin’s (1976)
terms.

Authenticated
304 Elisabeth Fritz
In order to quantify the extent of topic drift, semantic distance is measured and
operationalized along a scale of 0–4, depending on the inferences that have to be
drawn in order to interpret the new message as topically coherent. This, however,
is highly interpretive (Herring 2003: 5–7; Stromer-Galley and Martinson 2009:
202), and a reworking of this coding scheme by Stromer-Galley and Martinson
(2009) replaces both sets by a two-fold coding scheme for old/new topic and old/
new referent(s).21 This then results in four different types of transitions on a scale
from on-topic/no drift-transitions, along small and greater drift, to full breaks with
a new topic and referent being established (Stromer-Galley 2009: 203).
Topics, in this framework, are loosely understood as “umbrellas for linked
knowledge structures” (Stromer-Galley and Martinson 2009: 211). Greater shifts
involving new topics with a continuity of referents can be found in metacommu-
nicative shifts or (as in Bill’s message in example (8)) “when an utterance shifts
from talking about a topic that is something in the world to discussing a person in
the interaction” (Stromer-Galley and Martinson 2009: 212):
(8) Sue George Bush is Great!
Teri I agree
Bill Sue, how can you say that? (or: Sue, are you a Republican?)
(adapted from Stromer-Galley and Martinson 2009: 212)
In contrast, example (9) is an instance of smaller topic drift, in which Groovy con-
tinues on the old topic (“the war in Afghanistan”; Stromer-Galley and Martionson
2009: 203) but introduces new referents:
(9) Christa Hey I just saw on CNN talking about finding “low load” nukes in the cave
Groovy Chris; … you sure you heard that right? This is just another example that Bush
was right to go into Afghanistan.
(Stromer-Galley and Martinson 2009: 203)
Both frameworks include an additional classification of individual local topics as
on- or off-topic. Herring and Nix (1997), closely adhering to Hobbs’ local perspec-
tive, define any message as off-topic that does not fit into the on-topic category;
the reference topic is always that of the directly preceding message in the sequence
(Herring and Nix 1997: 8). Stromer-Galley and Martinson (2009), on the other
hand, specifically focus on chatroom interactions with a pre-determined global
topic. In a separate analytic step, they classify each local topic as on-topic or off-
topic, depending on whether or not they consider it a relevant aspect of the global
21
Topical coding necessarily has to remain fundamentally interpretive (Herring 2003: 5;
Stromer-Galley and Martinson 2009: 204). Nonetheless, Stromer-Galley and Martinson
(2009: 210) speak of an “acceptable” intercoder agreement for the assessment of local
transitions (80 %) and a high degree (of 98 %) for the classification of local topics as
on-topic or off-topic.

Authenticated
topic. This allows them to not only determine degrees of topic drift over longer
stretches of interaction, but also the number of threads and/or individual messages
that focus on topics deviating from the global monotopical orientation.
Both Herring and Nix’s as well as Stromer-Galley and Martinson’s findings
indicate a tendency for serious and/or educational chat to exhibit a higher percent-
age of on-topic messages than purely recreational chats.22 With regard to the distri-
bution of local topic development patterns, Stromer-Galley and Martinson’s study
shows that entertainment and politics chats exhibit a relatively high frequency of
“messages that introduced a new interactional [local] topic while retaining a refer-
ential connection with prior messages” (2009: 206), whereas old topic, old referent
relations make up the largest part of transitions in all of the chats. In the chat on
politics, these shifts tended to result in a personal attack on another chatter; on the
entertainment chat, they usually introduced plays on words. Full breaks, on the
other hand, were particularly prominent in chats on auto racing and cancer support
(32 % and 24 %, respectively), while still accounting for 12 % of the messages
in the entertainment chat. However, Stromer-Galley and Martinson (2009: 206)
mainly attribute them to technical difficulties, which participants intermittently
focused on.
This shows again that the purpose of the interaction is a predominant factor in
global topical management. It is not the technical environment of computer-medi-
ated discourse itself that impedes topic-centred coherent interactions (Stromer-Gal-
ley and Martinson 2009: 209; Herring 2003: 3, 11). Instead, participants tend to
accommodate their topic management behaviour to whether the interaction is
framed as serious or non-serious, instructional/problem-solving or recreational.
The former will, as a tendency, lead to longer and fewer threads with a common
topical focus, while in the latter case, participants will feel less obliged to pur-
sue a narrow topical focus, so that multiple shorter (and overlapping) threads can
develop (Herring 2003: 3), which may often deal with relational rather than strictly
topic-focused matters. The relative frequency of messages loosely connected by
means of superficial lexical connections and word associations (i.e. greater shifts
in Stromer-Galley and Martinson’s terms) tends to be so prevalent that Herring
(2013a) speaks of loosened relevance as a communicative norm in recreational
synchronous computer-mediated discourse.
As for global patterns of topic management in asynchronous exchanges,
Lambiase’s (2010) study of a discussion group centred on the exchange of news
after a bombing in Oklahoma City in 1995 shows that attention to the original
global topic decreases “within a few days” (Lambiase 2010: 1): The focus on news
22
However, the results are not directly comparable as the two studies employ different
reference points (the preceding message vs. the global topic). Still, the figures point in
the same direction.

Authenticated
306 Elisabeth Fritz
exchange is gradually broadenened towards more argumentative discourse on mar-

ginal topics, which were developed from the global topic by means of stepwise
local shifts as well as full breaks (Lambiase 2010: 6). Most original topics and
purposes become exhausted and/or obsolete after some time, and whether or not a
thread is kept alive depends on whether or not the participants have succeeded in
deriving new (and potentially off-topic) topical strands that can hold their interest.
Global topic management also seems to be majorly influenced by how the inter-
action is set off. In contrast to many older forms of computer-mediated discourse
(especially chat), most Web 2.0 interactions are initiated by a particular prompt,
such as a video, news story etc. As Herring (2013b) points out, the presence of
such a prompt leads to a reconfiguration of topical interaction patterns, as partici-
pants tend to continuously orient their talk towards this prompt. The consequence
of this is “limited topical development” (Herring 2013b: 13), with most contri-
butions being oriented to the initial prompt rather than (exclusively) at preceding
comments (cf. also Bou-Franch, Lorenzo-Dus and Garcés-Conejos Blitvich 2012).
However, interactions on older platforms, such as discussion forums, are equally
clearly triggered by an identifiable prompt (viz. the initiating post), which is per-
sistently displayed on topic of the ensuing interaction. But while these prompts
are directly directed at users and typically feature direct requests for interaction,
prompts on Web 2.0 platforms can vary a lot in this respect (e.g., YouTube vlogs
vs. articles on news websites). In the light of the scarcity of research into topic
development on asynchronous platforms in general, it would be interesting for
future research to investigate which role the nature of this prompt plays for topic
development in the ensuing discussion, and whether the pervasiveness of Web 2.0
interaction patterns, in turn, can be shown to leave traces in topic management on
older discussion platforms.
5. Conclusion
Discourse topics, while being a major factor for construing coherent talk, have
always been an inherently fuzzy concept, touching upon conceptually distinct but
empirically interwoven matters of sequential discourse organization and commu-
nicative purpose or activity. Online interactions constitute a further challenge for
participants as well as analysts. As varied as computer-mediated communication
is, social media tend to be geared towards facilitating instances of multiparticipant
monotopical discourse.
There is, as of yet, little research specifically on topic management in com-
puter-mediated discourse, but those studies that do exist tend to pursue a narrow
topic definition and track patterns of topic development back to a measurement
of semantic similarity between individual messages. Their findings suggest that,
while patterns of topic drift as well as discontinuous breaks can be found regularly,

Authenticated
it is not the technological setting itself that necessarily impedes monotopical talk.
Instead, users accommodate the way in which they manage topics to other key fac-
tors characterizing the interaction, most prominently perhaps its communicative
purpose. From this perspective, users’ perception of acceptable (i.e., “on-topic”)
talk may have as much to do with maintaining a relevant topical focus as with
orienting to communicative actions conforming to the projected purpose of the
interaction.
Asynchronous platforms facilitate especially complex interaction structures,
with threads potentially stretching over extended periods of time and developing
multiple more or less distinct sub-threads, which can develop and interleave in
various ways. Individual messages have the potential to be linguistically complex
and extend on more than one locally-instantiated topic at the same time, thus being
able to give rise to complex multilinear structures. The practice of using semi-au-
tomatized linking devices provided by the system is adopted by users not only
to indicate sequential structures alone; it is also regularly exploited as a device
aiding local shifts of topic, which are often accompanied by modifications in the
participation structure. For a detailed description of the dynamics of topic man-
agement, it thus seems to be helpful to look at mid-level patterns of organization,
which should be characterized by relatively homogeneous constellations regard-
ing participation structure, (range of local) topics and communicative actions, but
also, e.g., temporal pace, message complexity, use of linking templates etc. These
structures can then provide the backbone from which strategies of interactional
development and management of distinct local topics at various degrees of speci-
ficity and with different degrees of attention to the global topic can be described.
References
Adato, Albert
1979 Unanticipated topic continuations. Human Studies 2(1): 171–186.
Adato, Albert
1980 “Occasionality” as a constituent feature of the known-in-common character of
topics. Human Studies 3(1): 47–64.
2007 Neue Medien – neue Schriftlichkeit? Mitteilungen des Deutschen Germanisten
verbandes 54(1): 72–97.
Arendholz, Jenny
2013a (In)Appropriate Online Behavior. A Pragmatic Analysis of Message Board
Arendholz, Jenny
2013b “How to stop strange people speaking to me” – A syntactic and interpersonal
perspective on offering advice online. In: Katrin Röder and Isle Wischer (eds.),
Anglistentag 2012 Potsdam. Proceedings, 357–370. Trier: Wissenschaftlicher
Verlag Trier.

Authenticated
308 Elisabeth Fritz
Arendholz, Jenny
2015 Quoting on online message boards: An interpersonal perspective. In: Jenny
ics of Quoting Now and Then, 53–69. Berlin/Boston: de Gruyter.
Baron, Naomi S.
2008 Always On. Language in an Online and Mobile World. Oxford: Oxford Uni-
versity Press.
Barcellini, Flore, Françoise Détienne, Jean-Marie Burkhardt and Warren Sack
2005 A study of online discussions in an open-source community: Reconstructing
thematic coherence and argumentation from quoting practices. In: Peter van
den Besselar, Giorgio De Michelis, Jenny Preece and Carla Simone (eds.),
Communities and Technologies 2005, 301–320. Proceedings of the Sec-
ond Communities and Technologies Conference, Milano 2005. Dordrecht:
Springer.
Baym, Nancy K.
2006 Language in computer-mediated communication. In: Keith Brown (ed.), Ency-
clopedia of Language & Linguistics. Second edition. Vol. 6, 523–529. Amster-
dam: Elsevier.
Baym, Nancy K.
2015 Personal Connections in the Digital Age. Second edition. Cambridge: Polity
Press.
2007 Sprachhandlungskoordination in der Chat-Kommunikation. Berlin/New York:
de Gruyter.
Berglund, Therese Örnberg
2009 Disrupted turn adjacency and coherence maintenance in instant messaging
conversations. Language@Internet 6, article 2: 1–25. http://www.languageat
internet.org/articles/2009/2106/Berglund.pdf.
Black, Steven D., James A. Levin, Hugh Mehan and Clark N. Quinn
1983 Real and non-real time interaction: Unravelling multiple threads of discourse.
Discourse Processes 6(1): 59–75.
de Gruyter.
2014 Conflict management in polylogues: A case study from YouTube. Journal of
2012 Social interaction in YouTube-based polylogues: A study of coherence. Jour-
nal of Computer-Mediated Communication 17(4): 501–521.
Brinker, Klaus, Hermann Cölfen and Steffen Pappert
2014 Linguistische Textanalyse. Eine Einführung in Grundbegriffe und Methoden.
8th edition. Berlin: Erich Schmidt.
Brinker, Klaus and Jörg Hagemann
2001 Themenstruktur und Themenentfaltung in Gesprächen. In: Klaus Brinker, Gerd
Antos, Wolfgang Heinemann and Sven F. Sager (eds.), Text- und Gesprächs-

Authenticated
linguistik. Ein internationales Handbuch zeitgenössischer Forschung. Vol-

ume 2, 1253–1263. Berlin/New York: de Gruyter.
Brown, Gillian and George Yule
1983 Discourse Analysis. Cambridge: Cambridge University Press.
Bublitz, Wolfram
1988 Supportive Fellow-Speakers and Cooperative Conversations. Discourse Top-
ics and Topical Actions, Participant Roles and ‘Recipient Action’ in a Particu-
lar Type of Everyday Conversation. Amsterdam/Philadelphia: Benjamins.
Bublitz, Wolfram
2006 It utterly boggles the mind: Knowledge, common ground and coherence. In:
Hanna Pishwa (ed.), Language and Memory. Aspects of Knowledge Rep-
resentations, 359–386. Berlin/New York: de Gruyter.
Bublitz, Wolfram
2015 Introducing quoting as a ubiquitous meta-communicative act. In: Jenny Arend-
holz, Wolfram Bublitz and Monika Kirner-Ludwig (eds.), The Pragmatics of
Quoting Now and Then, 1–26. Berlin/Boston: de Gruyter.
Bublitz, Wolfram
2017 Oral features in fiction. In: Miriam A. Locher and Andreas H. Jucker (eds.),
Pragmatics of Fiction, 235–263. Berlin/Boston: de Gruyter.
2011 “Three men using our toilet all day without flushing – This may be one of the
worst sentences I’ve ever read”: Quoting in CMC. In: Joachim Frenk and Lena
Steveker (eds.), Anglistentag Saarbrücken 2010. Proceedings, 433–447. Trier:
Button, Graham
1987 Moving out of closings. In: Graham Button and John R.E. Lee (eds.), Talk and
Social Organisation, 101–151. Clevendon/Philadelphia: Multilingual Matters.
Button, Graham and Neil Casey
1984 Generating topic: The use of topic initial elicitors. In: Maxwell Atkinson and
John Heritage (eds.), Structures of Social Action. Studies in Conversation
Analysis, 167–190. Cambridge: Cambridge University Press.
Button, Graham and Neil Casey
1985 Topic nomination and topic pursuit. Human Studies 8(1): 3–55.
Cherny, Lynn
1999 Conversation and Community. Chat in a Virtual World. Stanford, CA: CSLI
Publications.
2015 Researching interactional forms and participation structures in public and
social media. In: Marta Dynel and Jan Chovanec (eds.), Participation in Pub-
lic and Social Media Interactions, 1–23. Amsterdam/Philadelphia: Benjamins.
Coates, Jennifer
1997 The construction of a collaborative floor in women’s friendly talk. In: Talmy
Givón (ed.), Conversation. Cognitive, Communicative and Social Perspec-
tives, 55–89. Amsterdam/Philadelphia: Benjamins.
Condon, Sherri L. and Claude G. Čech
2010 Discourse management in three modalities. Language@Internet 7, article 6:
1–32. http://www.languageatinternet.org/articles/2010/2770/Condon_Cech.
pdf.

Authenticated
310 Elisabeth Fritz
Crow, Bryan K.
1983 Topic shifts in couples’ conversations. In: Robert T. Craig and Karen Tracy
(eds.), Conversational Coherence. Form, Structure, and Strategy, 136–156.
Beverly Hills, CA: Sage.
Crystal, David
Dascal, Marcelo and Tamar Katriel
1979 Digressions: A study in conversational coherence. PTL: A Journal for Descrip-
tive Poetics and Theory of Literature 4: 202–232.
Donath, Judith
2002 A semantic approach to visualizing online conversations. Communications of
the ACM 45(4): 45–49.
Donath, Judith, Karrie Karahalios and Fernanda Viégas
1999 Visualizing conversation. Journal of Computer-Mediated Communication
4(4). doi: 10.1111/j.1083-6101.1999.tb00107.x.
Drew, Paul and Elizabeth Holt
1998 Figures of speech: Figurative expressions and the management of topic transi-
tions in conversation. Language in Society 27(4): 495–522.
Dürscheid, Christa
2003 Medienkommunikation im Kontinuum von Mündlichkeit und Schriftlichkeit.
Theoretische und empirische Probleme. Zeitschrift für angewandte Linguistik
38: 37–56.
Dürscheid, Christa
2005 Medien, Kommunikationsformen, kommunikative Gattungen. Linguistik
Online 22. https://bop.unibe.ch/linguistik-online/article/view/752/1283.
Dürscheid, Christa and Sarah Brommer
2009 Getippte Dialoge in neuen Medien. Sprachkritische Aspekte und linguistische
Analysen. Linguistik online 37(1): 3–20.
Edelsky, Carole
1981 Who’s got the floor? Language in Society 10(3): 383–421.
Fritz, Gerd
1982 Kohärenz. Grundfragen der linguistischen Kommunikationsanalyse. Tübin-
gen: Gunter Narr.
Fritz, Gerd
1994 Grundlagen der Dialogorganisation. In: Gerd Fritz and Franz Hundsnurscher
(eds.), Handbuch der Dialoganalyse, 177–201. Tübingen: Max Niemeyer.
Gardner, Roderick
1987 The identification and role of topic in spoken interaction. Semiotica 65(1/2):
129–141.
Goffman, Erving
1981 Forms of Talk. Oxford: Blackwell.
Gruber, Helmut
1997 Themenentwicklung in wissenschaftlichen E-mail-Diskussionslisten. Ein
Vergleich zwischen einer moderierten und einer nicht-moderierten Liste. In:
Rüdiger Weingarten (ed.), Sprachwandel durch Computer, 105–128. Opladen:
Westdeutscher Verlag.

Authenticated
Gruber, Helmut
1998 Computer-mediated communication and scholarly discourse: Forms of topic-
initiation and thematic development. Pragmatics 8(1): 21–45.
Gruber, Helmut
2013 Mailing list communication. In: Susan C. Herring, Dieter Stein and Tuija Vir-
tanen (eds.), Pragmatics of Computer-Mediated Communication, 55–82. Ber-
phia: Benjamins.
Hayashi, Reiko
1991 Floor structure of English and Japanese conversation. Journal of Pragmatics
16(1): 1–30.
Herring, Susan C.
1996 Two variants of an electronic message schema. In: Susan C. Herring (ed.),
Computer Mediated Communication. Linguistic, Social and Cross-Cultural
Herring, Susan C.
1999 Interactional coherence in CMC. Journal of Computer-Mediated Communica-
tion 4(4). doi: 10.1111/j.1083-6101.1999.tb00106.x.
Herring, Susan C.
2003 Dynamic topic analysis of synchronous chat. New Research for New Media.
Innovative Research Methodologies Symposium Working Papers and Read-
ings, 1–18. Minneapolis, MN: University of Minnesoty School of Journalism
and Mass Communication.
http://ella.slis.indiana.edu/~herring/dta.2003.pdf.
Herring, Susan C.
Internet 4, article 1: 1–37.
Herring, Susan C.
2010 Who’s got the floor in computer-mediated conversation? Edelsky’s gender
patterns revisited. Language@Internet 7, article 8, 1–29. http://www.lan-
guageatinternet.org/articles/2010/2801/herring.intro.pdf.
Herring, Susan C.
2013a Relevance in computer-mediated conversation. In: Susan C. Herring, Dieter
Stein and Tuija Virtanen (eds.), Pragmatics of Computer-Mediated Communi-
cation, 245–268. Berlin/Boston: de Gruyter.
Herring, Susan C.
2013b Discourse in Web 2.0: Familiar, reconfigured, and emergent. In: Deborah Tan-
nen and Anna Marie Trester (eds.), Discourse 2.0. Language and New Media,
1–25. Washington, DC: Georgetown University Press.
Herring, Susan C. and Andrew J. Kurtz
2006 Visualizing dynamic topic analysis. Proceedings of CHI’06, 1–6. ACM Press.
http://ella.slis.indiana.edu/~herring/chi06.pdf.

Authenticated
312 Elisabeth Fritz
Herring, Susan C. and Carole G. Nix

1997 Is “serious chat” an oxymoron? Pedagogical vs. social uses of Internet Relay
Chat. Paper presented at the American Association of Applied Linguistics
annual conference, 1–19. Orlando, FL. http://ella.slis.indiana.edu/~herring/
aaal.1997.pdf.
Hobbs, Jerry R.
1990 Topic drift. In: Bruce Dorval (ed.), Conversational Organization and Develop-
ment. 3–22. Norwood, NJ: Ablex.
Hoffmann, Ludger
2004 Chat und Thema. In: Michael Beißwenger, Ludger Hoffmann and Angelika
Storrer (eds.), Internetbasierte Kommunikation, 103–122. (Osnarbrücker
Beiträge zur Sprachtheorie 68.) Oldenburg: Redaktion OBST.
Holmer, Torsten
2008 Discourse structure analysis of chat communication. Language@Internet 5, arti-
cle 10: 1–17. http://www.languageatinternet.org/articles/2008/1633/holmer.pdf.
Holt, Elizabeth and Paul Drew
2005 Figurative pivots: The use of figurative expressions in pivotal topic transi-
tions. Research on Language and Social Interaction 38(1): 25–61.
Jefferson, Gail
1983 Caveat speaker: Preliminary notes on recipient topic-shift implicature. Tilburg
Papers in Language and Literature 30: 1–25.
Jefferson, Gail
1984 On stepwise transition from talk about a trouble to inappropriately next-posi-
tioned matters. In: Maxwell J. Atkinson and John Heritage (eds.), Structures of
Social Action. Studies in Conversation Analysis, 191–222. Cambridge: Cam-
Jones, Rod and Joanna Thornborrow
2004 Floors, talk and the organization of classroom activities. Language in Society
33(3): 399–423.
Keller, Eric
1979 Gambits: Conversational strategy signals. Journal of Pragmatics 3(3–4): 219–
238.
2004 Introducing polylogue. Journal of Pragmatics 36(1): 1–24.
Kilian, Jörg
2001 T@astentöne. Geschriebene Umgangssprache in computervermittelter Kom-
munikation. Historisch-kritische Ergänzung zu einem neuen Feld der lin-
guistischen Forschung. In: Michael Beißwenger (ed.), Chat-Kommunikation.
Sprache, Interaktion, Sozialität & Identität in synchroner computervermit-
telten Kommunikation. Perspektiven auf ein neues Forschungsfeld, 55–78.
Stuttgart: ibidem.

Authenticated
Korolija, Natascha and Per Linell

1996 Episodes: Coding and analyzing coherence in multiparty conversation. Lin-
guistics 34(4): 799–831.
Lambiase, Jacqueline
2010 Hanging by a thread: Topic development and death in an online discussion of
breaking news. Language@Internet 7, article 9: 1–22. http://www.languageat
internet.org/articles/2010/2814/Lambiase.pdf.
Landert, Daniela
2014 Personalisation in Mass Media Communication. British Online News Between
Public and Private. Amsterdam/Philadelphia: Benjamins.
Largier, Céline
2002 Aspekte der Debatte in argumentationsorientierten Internet-Foren: die Abtrei-
bungsdebatte in Frankreich und Deutschland. Deutsche Sprache 30(4): 287–
306.
Lee, Benny P.H.
2001 Mutual knowledge, background knowledge and shared beliefs: Their roles in
establishing common ground. Journal of Pragmatics 33(1): 21–44.
Leech, Geoffrey
2014 The Pragmatics of Politeness. Oxford: Oxford University Press.
Lenk, Uta
1998 Marking Discourse Coherence. Functions of Discourse Markers in Spoken
English. Tübingen: Gunter Narr.
Linell, Per
1998 Approaching Dialogue. Talk, Interaction and Contexts in Dialogic Perspec-
tives. Amsterdam/Philadelphia: Benjamins.
Linell, Per and Natascha Korolija
1997 Coherence in multi-party conversation. Episodes and contexts in interaction.
In: Talmy Givón (ed.), Conversation. Cognitive, Communicative and Social
Marcoccia, Michel
internet newsgroups. Journal of Pragmatics 36(1): 115–145.
Markman, Kris M.
2013 Conversational coherence in small group chat. In: Susan C. Herring, Dieter
Stein and Tuija Virtanen (eds.), Pragmatics of Computer-Mediated Communi-
cation, 539–564. Berlin/Boston: de Gruyter.
Markman, Kris M.
2009 “So what shall we talk about”. Openings and closings in chat-based virtual
meetings. Journal of Business Communication 46(1): 150–170.
Maynard, Douglas W.
1980 Placement of topic changes in conversation. Semiotica 30(3/4): 263–290.
Ochs Keenan, Elinor and Bambi B. Schieffelin
1976 Topic as a discourse notion: A study of topic in the conversations of children
and adults. In: Charles N. Li (ed.), Subject and Topic, 335–384. New York:
Academic Press.

Authenticated
314 Elisabeth Fritz
Parker, Richard
1984 Conversational grouping and fragmentation: a preliminary investigation.
Semiotica 50(1–2): 43–68.
Pavlidou, Theodossia-Soula
2014 Phases in discourse. In: Klaus P. Schneider and Anne Barron (eds.), Pragmat-
ics of Discourse, 353–383. Berlin/Boston: de Gruyter.
Planalp, Sally and Karen Tracy
1980 Not to change the topic but …: A cognitive approach to the management of
conversation. In: Dan Nimmo (ed.), Communication Yearbook 4, 237–258.
New Brunswick, NJ: Transaction.
Pons Bordería, Salvador and Maria Estellés Arguedas
2009 Expressing digression linguistically: Do digressive markers exist? Journal of
Pragmatics 41(5): 921–936.
Reichman, Rachel
1990 Communication and mutual engagement. In: Bruce Dorval (ed.), Conversa-
tional Organization and its Development, 23–48. Norwood, NJ: Ablex.
Reichman, Rachel
1978 Conversational coherency. Cognitive Science 2(4): 283–327.
Sacks, Harvey
1995 Lectures on Conversation. Volume 1. Oxford: Blackwell.
Schank, Roger C.
1977 Rules and topics in conversation. Cognitive Science 1(4): 421–441.
Schegloff, Emanuel A.
1990 On the organization of sequences as a source of ‘coherence’ in talk-in-inter-
action. In: Bruce Dorval (ed.), Conversational Organization and its Develop-
ment, 51–77. Norwood, NJ: Ablex.
2007 Sequence Organization in Interaction. A Primer in Conversation Analysis
I. Cambridge: Cambridge University Press.
Schegloff, Emanuel A. and Harvey Sacks
1973 Opening up closings. Semiotica 8(4): 289–327.
Schönfeldt, Juliane
2001 Die Gesprächsorganisation in der Chat-Kommunikation. In: Michael Beiß
wenger (ed.), Chat Kommunikation. Sprache, Interaktion, Sozialität & Iden-
tität in synchroner computervermittelter Kommunikation. Perspektiven auf ein
interdisziplinäres Forschungsfeld, 25–53. Stuttgart: ibidem.
Schütte, Wilfried
2004 Diskursstrukturen in fachlichen Mailinglisten: Zwischen Einwegkommunika-
tion und Interaktion. In: Michael Beißwenger, Ludger Hoffmann and Ange-
lika Storrer (eds.), Internetbasierte Kommunikation, 55–75. (Osnarbrücker
Beiträge zur Sprachtheorie 68.) Oldenburg: Redaktion OBST.
Severinson Eklundh, Kerstin
2010 To quote or not to quote: Setting the context for computer-mediated dialogues.
Language@Internet 7, article 5: 1–30. http://www.languageatinternet.org/arti-
cles/2010/2665/Severinson_Eklundh.pdf.
Severinson Eklundh, Kerstin and Clare Macdonald
1994 The use of quoting to preserve context in electronic mail dialogues. IEEE
Transactions on Professional Communication 37(4): 197–202.

Authenticated
Severinson Eklundh, Kerstin and Henry Rodriguez

2004 Coherence and interactivity in text-based group discussions around web doc-
uments. Proceedings of the 37th Hawaii International Conference on System
Sciences: 1–10.
Shultz, Jeffrey J., Susan Florio and Frederick Erickson
1982 Where’s the floor? Aspects of the cultural organization of social relation-
ships in communication at home and in school. In: Perry Gilmore and Allan
A. Glatthorn (eds.), Children In and Out of School. Ethnography and Educa-
tion, 88–123. Washington, DC: Center for Applied Linguistics.
Simpson, James
2005 Conversational floors in synchronous text-based CMC discourse. Discourse
Studies 7(3): 337–361.
Simpson, James
2013 Conversational floor in computer-mediated discourse. In: Susan C. Herring,
Dieter Stein and Tuija Virtanen (eds.), Pragmatics of Computer-Mediated
Communication, 515–538. Berlin/Boston: de Gruyter.
Sirois, Patricia and Bruce Dorval
1988 The role of returns to a prior topic in the negotiation of topic change: A devel-
opmental investigation. Journal of Psycholinguistic Research 17(3): 185–210.
Smithson, Janet, Siobhan Sharkey, Elaine Hewis, Ray Jones, Tobit Emmens, Tamsin Ford
and Christabel Owens
2011 Problem presentation and responses on an online forum for young people who
self-harm. Discourse Studies 13(4): 487–501.
Stromer-Galley, Jennifer and Anna M. Martinson
2009 Coherence in political computer-mediated communication: Analyzing topic
relevance and drift in chat. Discourse & Communication 3(2): 195–216.
Thurlow, Crispin, Laura Lengel and Alice Tomic
2007 Computer Mediated Communication. Social Interaction and the Internet. Lon-
don/Thousand Oaks: Sage.
Tiittula, Liisa
1993 Metadiskurs. Explizite Strukturierungsmittel im mündlichen Diskurs. Ham-
burg: Helmut Buske.
Traverso, Véronique
2004 Interlocutive ‘crowding’ and ‘splitting’ in polylogues: The case of a research-
ers’ meeting. Journal of Pragmatics 36(1): 53–74.
van Dijk, Teun A.
1977 Text and Context. Explorations in the Semantics and Pragmatics of Discourse.
London/New York: Longman.
van Dijk, Teun A.
1980 Macrostructures. An Interdisciplinary Study of Global Structures in Discourse,
Interaction, and Cognition. Hillsdale, NJ: Lawrence Erlbaum.
Werry, Christopher C.
1996 Linguistic and interactional features of Internet Relay Chat. In: Susan C. Her-
ring (ed.), Computer-Mediated Communication. Linguistic, Social and Cross-
Cultural Perspectives, 47–63. Amsterdam/Philadelphia: Benjamins.

Authenticated
316 Elisabeth Fritz
West, Candace and Angela Garcia

1988 Conversational shift work: A study of topical transitions between women and
men. Social Problems 25(5): 551–575.

Authenticated
12. Discourse and cohesion
Christoph Schubert
Abstract: This paper investigates the occurrence of cohesion in the non-linear genre
of social media platforms, taking into account both grammatical and lexical ties at
the intra-, inter-, and extranodal levels. In addition to verbal cohesion, the focus is
on the functions of software-dependent nonverbal signs such as response buttons
and navigation tools, which often provide global orientation and contextualization
cues. The platforms under scrutiny are the media-sharing service YouTube, the
social network site Facebook, the microblogging service Twitter as well as personal
weblogs. The cohesive ties on these websites are examined with regard to types of
cross-modal relations, degrees of author control, and structural composition. By
contextualizing such ties in their social circumstances of interpersonal networking,
it is shown that hypercohesion in social media is the result of textual collabora-
tion between users in the form of mutually relevant posts, entries, and comments.
1. Introduction
When the notion of cohesion was originally established in the 1970s (cf. Halliday
and Hasan 1976; Gutwinski 1976), electronic discourse did not play a decisive role
in everyday social interaction. Since the genres discussed in these seminal stud-
ies belong to linear spoken and written discourse, the application of cohesion to
non-linear social media platforms poses new challenges. In their ground-breaking
monograph, Halliday and Hasan define cohesion as a “semantic relation” (1976: 6)
that manifests itself in grammatical and lexical ties across the sentences of a text.1
1
Only a few years later, the complementary term coherence was programmatically estab-
lished (cf. de Beaugrande and Dressler 1981: 84–112; Brown and Yule 1983: 223–271).
Originally conceptualized as a continuity of sense that is not explicitly realized in texts
but cognitively constructed by recipients, the term has proved controversial over the
past decades. According to Bublitz, coherence is both “an interpretive process” and a
“mental concept” (2006: 363), resulting from strategies of discourse comprehension.
In addition, coherence is intuitively expected by recipients as a “default assumption”
(Bublitz 2006: 364, emphasis original) and varies according to individual interpreters.
As regards the relationship between cohesion and coherence, it has been demonstrated
that cohesion is neither sufficient nor necessary for the construction of coherence, yet
cohesion strongly supports the process of establishing unexpressed meaning across sen-
tences and paragraphs (cf. Tanskanen 2006: 23–29).
DOI 10.1515/9783110431070-012
Authenticated
318 Christoph Schubert
While sentences are internally marked by “structure”, the patterns of cohesive ties
throughout a text are said to constitute its “texture” (Halliday and Hasan 1976:
2). Since the often idiosyncratic punctuation in social media is no valid parameter
for the determination of written sentence boundaries, cohesive ties will here be
examined not at the intersentential but at the interclausal level, which depends on
syntactic functions (cf. Hoffmann 2012: 73).
Cohesive ties can be subdivided into grammatical and lexical devices that
weave textual patterns mainly at the discursive micro-level (cf. Halliday and Mat-
thiessen 2004: 532–538; Schubert 2012: 31–64). Grammatical cohesion comprises
pro-forms, coordinators, syntactic constructions such as ellipsis and parallelism
as well as the grammatical categories of tense and aspect. Lexical cohesion is
realized chiefly by means of partial and total repetition, semantic relations, par-
aphrases as well as lexical fields. On social media platforms, cohesive ties are
additionally established by technical devices such as response buttons, navigation
tools, embedded hyperlinks and menus. From the perspective of computer-medi-
ated discourse analysis (CMDA), texture in social media is constituted primar-
ily at the level of “interaction management”, as realized by “turns, sequences,
exchanges, threads etc.” (Herring 2013: 5).
Recent research on cohesion has concentrated on various areas, such as lex-
ical cohesion from a statistical, corpus-based vantage point (e.g. Flowerdew and
Mahlberg 2009), the spoken-written divide in cohesion (e.g. Dontcheva-Navra-
tilova and Povolná 2009), cross-modal cohesion in film discourse (Tseng 2013) as
well as contrastive and variational issues of cohesion (e.g. Neumann 2014; Klein
2012; Neumann and Fest 2016). However, cohesion has been neglected by most
major studies dealing with digital discourse, Netspeak and computer-mediated
communication (CMC) (e.g. Crystal 2006; Frehner 2008; Yus 2011; Zappavigna
2012; Tannen and Trester 2013). A few notable exceptions to this practice are
specific articles by Fritz (1999), Storrer (1999), Bublitz (2005), Markman (2012)
and Adami (2015) as well as a monograph by Hoffmann (2012), which provides an
examination of cohesive profiling in the electronic genre of weblogs.
The present contribution intends to treat cohesion not as a purely text-linguistic
phenomenon in its own right but aims to recontextualize cohesive ties in their social
circumstances, as proposed already by Martin’s systemic-functional account of
cohesion and texture (cf. 2001: 47). Since the main function of social network sites
(SNS) is the creation of social “connections” (boyd and Ellison 2008: 211), it is to
be expected that these are reflected by linguistic linkage as well. On these grounds,
this article will investigate the internal texture of individual messages, posts and
comments in social media and the ways in which they are cohesively interwoven.
In doing so, the paper will first provide a taxonomy of cohesive devices in order
to lay the conceptual foundation and will then successively apply the single cat-
egories to local discourse units of the media-sharing service YouTube, the social
network site Facebook, the microblogging service Twitter and (we)blogs, so as to

Authenticated
Discourse and cohesion 319
cover the most prominent social media platforms. These services are all associated
with the concept of Web 2.0, which is generally marked by “participatory informa-
tion sharing; user-generated content; an ethic of collaboration; and use of the web
as a social platform” (Herring, Stein and Virtanen 2012: 12). On this basis, it will
be possible to discuss the pragmatic issue of participatory interaction with regard
to the intentions and expectations of users. Owing to the hypermedial and multi-
modal character of these electronic genres, the focus will be not only on verbal
cohesion but also on software-dependent nonverbal signs such as navigation tools,
buttons and embedded hyperlinks.
2. Cohesive ties on social media platforms
As for verbal cohesive devices, a terminological synopsis of Halliday and Hasan

(1976), de Beaugrande and Dressler (1981) as well as Halliday and Matthiessen
(2004) yields a taxonomy of grammatical and lexical items, as outlined in Table 1.
Table 1. A survey of verbal cohesive devices (adapted from Schubert 2012: 61–62)
grammatical cohesion lexical cohesion
(1) pro-forms reference (1) repetition total

substitution partial (morphemic)
(2) syntactic ellipsis (2) semantic synonymy
constructions parallelism relations antonymy
(3) conjunction coordinators/ hyponymy
subordinators meronymy
[adverbs] (3) paraphrase expansion
[prepositional condensation
phrases]
(4) grammatical tense
categories aspect
(4) collocation lexical fields
lexical sets
Since pro-forms in texts function as search instructions, they create cohesive ties
with a presupposed phrase or passage that constitutes the lexical antecedent. While
items of “reference”, such as personal and demonstrative pronouns, show coref-
erential relations, the members of “substitution” (e.g. one, do or so) replace their
antecedents only at the grammatical surface level (cf. Halliday and Hasan 1976:
89). Ellipsis creates cohesion due to the fact that grammatically incomplete struc-
tures presuppose a source in the surrounding context which provides the missing
lexical items. Syntactic parallelism has a cohesive effect because of its formal

Authenticated
similarity, which is often reflected by analogies in content as well. For instance, in

the weblog “1000 Awesome Things” (see Section 6), two comments discussing the
entry “#977 The smell of gasoline” show SPO structures, both of them express-
ing forms of (dis-)agreement: the post “I like your blog..” (15 September 2011)
is answered by the comment “u guys are crazy i hate the smell of gasoline” (22
September 2011).
Conjunctive items may be purely grammatical in the case of coordinators and
subordinators or can contain a lexical component if they are realized as linking
adverbs (e.g. moreover) or conjunctive prepositional phrases (e.g. in addition, as
a result). The interclausal logical connection expressed by them can be additive,
adversative, causal or temporal (cf. Halliday and Hasan 1976: 242–243). The cat-
egory of conjunction also comprises so-called “continuatives”, such as yes, no,
oh, of course, anyway, after all and well, which “signal a move in the discourse:
a response, in dialogue, or a new move to the next point if the same speaker is
continuing” (Halliday and Matthiessen 2004: 81). In their function of organizing
and structuring discursive interaction, they are similar to discourse markers such
as you know, you see and right, which can indicate “an interactive relationship
between speaker, hearer, and message” or “a transition in the evolving progress”
of communication (Biber et al. 1999: 1086). Since there are structural analogies
between social media discourse and spoken conversation, it may be expected that
such items occur online between users as well. For instance, related Twitter posts
constitute a written polylogue that is likewise frequently organized by “connecting
adverbs” such as well (cf. Crystal 2011: 48). Tense and aspect are two grammat-
ical categories that trigger cohesion mainly by means of temporal continuity or
sequence.
In the area of lexical cohesion, the repeated occurrence of complete words is
labelled “total” repetition, while the recurrence of morphemes in different combi-
nations constitutes “partial” repetition (cf. de Beaugrande and Dressler 1981: 49).
Lexical cohesion is also realized by various kinds of semantic relations, which
may form cohesive chains both within and across web pages, since here no uni-
directional pointing procedure is necessary. The category of “paraphrase” shows
“expansion” if the content of a shorter expression such as a noun phrase is speci-
fied by longer expressions in the ensuing text, while “condensation” refers to the
opposite sequence (cf. Schubert 2012: 51–52). In a text-linguistic framework, the
phenomenon of collocation is defined as “the association of lexical items that
regularly co-occur” (Halliday and Hasan 1976: 284). While lexical fields contain
items that belong to the same word class and have common semantic features,
the members of lexical sets are more loosely associated and do not necessarily
belong to the same part of speech (cf. Schubert 2012: 53–54). Since sections of
social media sites are commonly grouped around thematic foci, it is to be expected
that on these pages collocation looms large among the various types of cohesive
ties.

Authenticated
When the traditional systemic-functional concept of cohesion is applied to

social media, it needs to be adapted and modified with respect to both the hyper-
textual and the multimodal character of these websites. A hypertext may be
defined as “a non-linear text that consists of nodes, that is to say textual units, and
links between these nodes” (Jucker 2002: 29). Storrer (cf. 2000: 240) addition-
ally distinguishes between medial (non-)linearity, based on the respective medium
(e.g. a film reel vs. YouTube), and conceptual (non-)linearity, depending on the
arrangement of textual sections by the authors of hypertexts (cf. Eisenlauer 2013:
62). Although there may be occasional “pre-defined paths”, as in online shopping
platforms, users commonly need to proceed along a “self-selected path” through
the hypertext, establishing individual cohesion through browsing (cf. Fritz 1999:
222). Instead of paragraphs and chapters prearranged in a fixed sequence, users
find nodes or clusters of nodes that they can process according to their own choos-
ing (cf. Bublitz 2005: 322). This type of cohesion that occurs in non-linear and
multimodal hypertexts such as social media platforms will here be labelled hyper-
cohesion.
Herring (1999) already shows that discursive interaction in many genres of
computer-mediated communication is challenged by asynchronous feedback and
interrupted adjacency. For instance, this is characteristic of YouTube comments
whenever the “top comments” button is clicked, since in this case the chronology
of comments is neglected. As a result, sequence-based types of grammatical cohe-
sion are hardly existent across individual posts. While in linear texts anaphoric
and cataphoric relations can be clearly distinguished, these backward- and for-
ward-pointing processes may become ambiguous in non-linear hypertext. Since
ellipsis relies on empty spaces in the discursive information flow, its use is highly
problematic across hypertextual nodes as well (cf. Eisenlauer 2013: 70; Fritz
1999: 230). However, lexical cohesive devices frequently appear in the form of
“semantically filled links” (Jucker 2002: 41), which are often realized as repeated
content words such as hashtags on Twitter. By contrast, “semantically empty
links” can occur as general directives in the form of view all and more options on
Twitter or show more on YouTube, triggering transfer to a different or extended
node.
As regards cohesion across nodes, a tripartite distinction may be drawn, for
textual ties may operate “within one node (intranodal cohesion), within one hyper-
textual database (internodal cohesion) or across different hypertextual databases
(extranodal cohesion)” (Eisenlauer 2013: 71). For example, personal weblogs
basically consist of the bloggers’ entries and corresponding user comments, which
together form individual nodes. In turn, single entries may be internodally linked
to other entries within the same blog or they can be extranodally connected to other
websites which are not under the present blogger’s control. As a consequence,
intranodal texture typically displays a higher degree of cohesive density than inter-
and extranodal texture, since there is a stronger thematic continuity within a node

Authenticated
such as a specific blog entry than between different hypertexts. Thus, the concept
of intertextuality is highly pertinent to the make-up and use of social media sites.
The presence of cohesive devices also corresponds with different functional
types of hyperlink triggers (cf. Jucker 2005: 289–291). First, what-triggers repre-
sent the question “what can you tell me about x?”, so that they initiate a discursive
structure resembling a question-answer adjacency pair. For instance, if users click
on a specific hashtag presented by Twitter, they will be provided with topically
related tweets. Second, where-triggers express the enquiry “where am I?”, offering
users textual orientation, for example by guiding them to the beginning or end of
the respective node. Along these lines, the trigger “top” on Twitter refers users
back to the most recently posted entry within the present hashtag. Third, do-trig-
gers represent a hypertextual order or command, such as printing a document,
showing a video clip or downloading a file. In YouTube, for example, this typi-
cally happens by clicking on the title of a video or by using the “upload” button,
which enables users to post a video. In general, by clicking a trigger, users initially
anticipate specific hypertextual connections and are afterwards enabled to decode
the cohesive ties between the root and target nodes.
Taking into account the layout and structure of hypertexts on the screen, Stor-
rer (cf. 1999: 49–61) develops three types of navigation devices operating on the
discursive macrolevel (cf. also Jucker 2002: 44). These techniques closely coop-
erate with website-internal cohesion and can be adapted to various subgenres of
social media discourse. First, “global orientation devices” help users to gain an
overview of the thematic and functional macrostructure of a hypertext. A typical
device belonging in this category is the hypertextual frame, forming a subsection
on the screen which remains constant while the hypertext is browsed and which
contains essential links that may be clicked at any time. On Facebook, for instance,
users may switch between their profile and their home page, while at the top of the
screen the verbal links “home”, “find friends” and the user’s name stay unchanged.
Hence, through easily remembered logos, such frames also fulfil the commercial
functions of brand recognition and corporate identity construction.
Second, “contextualisation devices” (Jucker 2002: 44) support users by indi-
cating the hypertextual node in which they are currently situated. This can be
achieved by specific colours, by typographic means or by a sequential list of steps
users have taken from the website’s cover page. For instance, when users choose
one of the five sections “top”, “live”, “accounts”, “photos” and “videos” under a
Twitter hashtag, the respective section heading is underlined while browsing this
part. Third, “retrospective orientation devices” (Jucker 2002: 44) provide informa-
tion on previously visited nodes and websites, so that users know which links they
have already clicked. This is most commonly done by means of the back button,
the history option and the use of bookmarks. In addition, so-called “bread crumbs”
(Storrer 1999: 60; emphasis original) introduce colour to highlight all triggers that
have already been clicked within a given space of time before the present moment

Authenticated
of browsing. Moreover, on highly personalized social media sites such as Face-

book profiles, retrospective orientation dominantly manifests itself in links such as
“friends” or “favorites”, which provide a survey of preferred sites.
Like many pages on the internet, social media services are marked by a mul-
timodal architecture, so that it is indispensable for the study of hypercohesion to
include nonverbal signs as well. Along these lines, Adami argues that cohesive
devices may operate “trans-modally, i.e., what is spoken in a video can be shown
as an image in its response, thus establishing a multi- or intermodal cohesive tie
of repetition” (2015: 234). Van Leeuwen (2005: 179–267) discusses “multimodal
cohesion” from a social semiotic perspective, providing a typology of informa-
tion linking with regard to images and words. In principle, he argues that visual
linking as well as visual-verbal linking rely on either “elaboration” or “extension”
from image to text or vice versa (2005: 229–230). Along these lines, for instance,
YouTube videos typically provide specifying and illustrating “elaborations” of
the clips’ verbal titles given below the video screen (cf. Section 3). By contrast,
“extension” implies complementary information, as it may be provided by knowl-
edgeable user comments referring to YouTube clips.
In semiotic research, the typology of signs is a controversially disputed issue,
which is chiefly due to deviant terminologies and diverging classificatory criteria
(cf. Nöth 1995: 107). A well-known starting point which can be utilised for the
discussion of social media is the Peircean “index-icon-symbol trichotomy” (Nöth
1995: 108). Along these lines, verbal signs are “symbols” on grounds of their
arbitrary and conventional relationship between signifier and signified, whereas
hypermedial images are often “icons”, since in these cases the form imitates, or
at least resembles, the content (cf. Kress and van Leeuwen 2006: 8). For instance,
with the help of iconic pictograms, users are able to express opinions (e.g. the
“thumbs up” button) or to carry out requests. By virtue of their imitative character,
chains of iconic signs are particularly immediate in their representation of seman-
tic continuity.
The third category, the so-called “indexical” signs, are characterized by a
causal or physical relationship between signifier and signified. According to Chan-
dler (2007: 37), this type not only comprises natural phenomena such as thunder
and lightning but also includes “recordings (a photograph, a film, video or televi-
sion shot, an audio-recorded voice)”. Consequently, this category is pertinent to
hypermedial communication as well, for social media sites often contain docu-
mentations of extralinguistic events and situations. In view of these distinctions,
the semantic connections that constitute multimodal cohesion are realized at the
discursive surface by intertwined patterns of symbolic, iconic and indexical signs.
At the same time, it needs to be considered that one hypertextual node as a whole
typically constitutes a multilayered sign that may comprise embedded signs of
various types. For instance, a Facebook home page is a complex assemblage of
clickable keywords such as “edit profile”, usually in combination with indexical

Authenticated
video clips and icons such as a stylized pen. Cohesion may further be supported by
visual means of page design as well as layout features based on colours and font
types (cf. Boardman 2005: 18; Hoffmann 2012: 20–21).
Occasionally, hypercohesion may fail or be disrupted as a result of the transi-
tory and dynamic character of the web. If single pages are deleted or shut down
without notification, this process is called “link rot” and thus entails “dead links”
(Crystal 2006: 211). Consequently, instead of the expected target page with rele-
vant cohesive ties, the message “error 404 – page NOT found” will be displayed.
If there is a temporary problem with the internet connection, the message usu-
ally reads “the page cannot be displayed” (Boardman 2005: 77), which implies a
short-term or merely momentary failure of hypercohesion. Another phenomenon
that interrupts cohesion is pop-up windows, which obstruct the users’ direct path
to the desired target page and are often motivated by commercial objectives (cf.
Boardman 2005: 63). After the reading process has been discontinued in this way,
users will decide whether they intend to engage in this discursive side sequence
or whether they discard the pop-up window and proceed with the cohesive chain
originally intended.
3. Cohesion on YouTube
As indicated by the personal pronoun you and the noun tube, a colloquial American
English expression for ‘television’, YouTube is a media-sharing site that allows
users to post and watch video clips (cf. Sindoni 2013: 172). Since users may also
comment on the clips, YouTube represents a mixture of media that is labelled “con-
vergent media computer-mediated communication” (CMCMC) by Herring (2013:
4). In this type of “hyper-intertextuality” (Sindoni 2013: 177, emphasis original),
cohesion operates inter- and hypermedially between the different discursive units
on this platform. Since YouTube has a collaborative character, Androutsopoulos
counts it among the so-called “participatory spectacles”, which are “multiau-
thored, multimodal, multimedia, inherently dialogic, dynamically expanding, and
open ended” (2013: 50; see also Johansson, Ch. 7, this volume).
As far as structure is concerned, YouTube consists of three main segments,
which will be investigated with regard to their respective cohesive potentials and
interconnected patterns: (1) the video clips, (2) a section of user comments, which
can be viewed according to the two parameters “top comments” or “newest first”,
and (3) “a hosting environment that includes a list of related videos and other
peripheral elements” (Androutsopoulos 2013: 47–48). These other elements com-
prise a navigational frame with interactive buttons entitled “review”, “sign in” and
“upload”, all of which trigger various online activities. Moreover, YouTube allows
users to look for topics by typing search expressions in an input field and provides
icons that enable the audience to “like” and “dislike” as well as to “subscribe”,

Authenticated
“add to”, “share” and carry out “more” actions by means of do- and what-triggers.
While segment (1) constitutes the main part and prime level, segments (2) and (3)
are paratextual planes that form a frame below and to the right of the video screen.
In terms of multimodal cohesion based on spatial composition, the video clip in
the top left area of the screen represents the “given” element (van Leeuwen 2005:
201), the starting point of the page design, if the sequentiality of Western writing
from left to right and from top to bottom is transferred to the reading of images (cf.
Schubert 2009: 35–36). Accordingly, the recommended videos on the right-hand
side constitute the “new” section with a sequential structure in both spatial and
temporal terms, while the comments below the video form the concluding section
on the page.
As regards segment (1), uploaded video clips are multimodal texts with a linear
organization, similar to unalterable verbal e-texts such as pdf documents. As these
clips are integrated into a hypertextual website, users are often confronted with
pop-up video clip suggestions or advertisements, which cover a certain section
of the screen and can be accepted or discarded. Characteristically, advertisements
show a thematic connection to the subject of the current video clip, so that lex-
ical cohesion links the video text proper with the quasi-parasitical paratext. For
instance, while watching the clip “the best of David Cameron” (8:20 minutes),
which is a compilation of humorous and sarcastic utterances by the former prime
minister in the House of Commons, a pop-up window notes “suggested: the best of
Nigel Farage”, establishing cohesion through lexical repetition and collocation in
the form of British politicians’ personal names. Moreover, the website recommends
users to subscribe to the YouTube channel “UK political humour”, whose very title
contains superordinate terms of the respective lexical set. Furthermore, the title of
the channel is cross-medially connected to a visual logo consisting of a photo of the
Houses of Parliament combined with the Union Jack. Beneath the video frame, the
content of the clip is briefly paraphrased by the caption “David Cameron’s funniest
moments inside the Chamber!”. Hence, in this additional piece of information, the
adjective “best” is not only grammatically continued by the superlative “funniest”
but also semantically expanded. When the mouse cursor is moved over the icons
representing thumbs up or thumbs down, a small window displays the brief texts
“I like this” and “I dislike this”, respectively. In this way, the demonstrative pro-
noun “this” establishes “extended” (Halliday and Hasan 1976: 66) demonstrative
reference with regard to the complete multimodal video clip.
As for segment (2), the comments usually show cohesive connections with
the prompting video clip and are also internally linked by means of grammatical
and lexical cohesive ties. If the presentation mode “top comments” is chosen, the
comments are not in chronological order, since the comments with most “thumbs
up” clicks are displayed first. As a result, the cohesion between adjacent comments
is restricted, since the antecedent utterance of conjunctive items or pronominal
reference is difficult to determine. If, however, the mode “newest first” is selected,

Authenticated
comments appear in chronological order, so that cohesion is more tightly knit. This
is demonstrated by example (1), which contains the five most recent comments on
the video clip “the best of David Cameron”.
(1) Alberto Barbosa 1 day ago
He may be a cuckservative, but at least he’s got some banter.
Kaustubh Misra 2 days ago
That ‘Bill Somebody’ roast though … Poor Ed.
Dr Penguin 3 days ago
Ed Balls just sitting there getting roasted xD
Rusty Shackleford 4 days ago
That is some top shelf banter right there
erik lindström 4 days ago
This is the first time I’m seeing this. I don’t know much about this guy, since I’m from
Sweden, but he acts like the Wolf of Wall Street.
(YouTube, “The Best of David Cameron”, retrieved 1 February 2016,
<https://www.youtube.com/watch?v=2U5R6ka4nlc>)
Hypercohesion here manifests itself in lexical repetition (“banter”, “day[s] ago”,
“Ed”) as well as in morphological repetition (“roast” and “roasted”) across indi-
vidual comments, so that the multi-authored discourse forms a cohesive unity to
some extent. In addition, pro-forms also contribute to cross-medial cohesion, as
the demonstrative pronoun in “seeing this” refers to the complete video clip and
the determiner in “this guy” relates both to Cameron in the clip and to his name in
the caption. In addition, the three occurrences of the personal pronoun “he” form a
cohesive chain across different comments. While “right there” refers to the events
in the video clip, the pro-adverb in “sitting there” creates a cohesive link to the
clip and at the same time deictically points to Ed Milliband’s seat on the opposition
bench, as visible in the clip.
Concerning the participatory framework, YouTubers commonly regard them-
selves as members of one of many “digital communities of practice” (Sindoni
2013: 174, emphasis original), interactively negotiating and interpreting shared
media content. As example (2) demonstrates, this practice can lead to a dialogue
of adjacent utterances.
(2) iM3GTR 1 week ago
And this load of jeering school children are running the country. Why can’t nice people
ever be in charge for once.
Scott Mairs 4 days ago
I find Cameron very strong … especially against that pathetic excuse of a man who is
Corbyn. And +iM3GTR I’d rather someone with backbone and authority than someone
who’d throw British patriotism and principles down the drain.
iM3GTR 4 days ago
+Scott Mairs Fair enough.
(YouTube, “The Best of David Cameron”, retrieved 1 February 2016
<https://www.youtube.com/watch?v=2U5R6ka4nlc>)

Authenticated
In extract (2), the critical comment posted “1 week ago” by a tuber named iM3GTR
is the prompt for a response by tuber Scott Mairs, who speaks in defence of Cam-
eron. This adjacency pair structure is made visible by the plus sign “+” in colloca-
tion with the original tuber’s name. On the same day (“4 days ago”), iM3GTR again
replied to this post, tentatively agreeing with the phrase “fair enough”. Hence, this
leads to a three-turn conversational exchange, in which the final response can be
analysed as an instance of clausal ellipsis, since it presupposes a previous state-
ment. In terms of cohesion, such posts and counter-posts show structures of spo-
ken conversation transferred to digital discourse.
Bou-Franch, Lorenzo-Dus and Garcés-Conejos Blitvich (2012: 506–511)
exclusively consider the comment section on YouTube as a written polylogue,
analysing types of “cross-turn cohesion” between the individual posts. Drawing
from a corpus of 300 consecutive Spanish comments on two YouTube clips, they
demonstrate that lexical cohesion is quantitatively dominant (1,018 items), while
the number of pronominal references is significantly lower (645 items). The other
three types of cohesion according to Halliday and Hasan (1976) hardly play a role,
as corroborated by the numbers for substitution (49 items), ellipsis (56 items) and
conjunction (51 items) (Bou-Franch, Lorenzo-Dus and Garcés-Conejos Blitvich
2012: 510). Since this type of interaction is similar across languages, comparable
quantities may be expected for English comment sections. Extending the scope,
the authors also point out “YouTubers’ preference for managing turns through
turn-entry/exit devices and cross-turn addressivity signals instead of employing
such strategies as cross-turn quoting and backchannelling” (2012: 515).
Finally, segment (3) of YouTube pages offers thematically related videos under
the heading “up next”, in which the adverb “next” invariantly represents tempo-
ral conjunction. The videos suggested at the top of the list bear the titles “David
Cameron vs Gordon Brown – very entertaining!!” and “David Cameron – thug life
compilation”, so that they display lexical repetition of the personal name in collo-
cation with new thematic foci. Under the eye-catching still frames which represent
the individual clips, users find the current number of “views”, which means that
lexical repetition of this word collocating with the lexical field of ordinal numbers
leads to additional cohesion. In case the autoplay option is activated, the single
clips will be played back to back in a linear succession suggested by the You-
Tube software, so that cohesion is then constructed not only by users but also by
automatized electronic processes. Scrolling down to the end of the list of videos,
users reach a do-trigger entitled “show more”, in which the pronominal quantifier
“more” creates comparative reference between the clips currently on display and
the clips to come.
With regard to audience participation, Adami (2015) points out that besides
comments, users may also react asynchronously in the form of home-made video
responses, which are thus linked to previously posted video clips. In her analysis of
various videos responding to a popular prompting video, she postulates a “related-

Authenticated
ness continuum” (Adami 2015: 254, emphasis original). At the one end of the scale,
responses show very strong cohesion with the initiating video, while at the other
end, the videos may show no explicit hints of relatedness at all. In between the two
extreme poles, there are different degrees of (non-)cohesion, realized by various
verbal and nonverbal devices. In particular, responding video clips may show the
immediate emotional reactions of users watching the prompting video (e.g. through
facial expressions), so that they function as direct second pair parts of multimodal
adjacency pairs. Alternatively, video bloggers may verbally comment on the orig-
inal clip, using some degree of topic deviation. Finally, transformations can occur
in the form of “[r]emixes, parodies and recontextualizations” (Adami 2015: 246).
As for the general theoretical framework for such open and multimodal types of
cohesion, Adami refers to Gunther Kress’s social semiotic idea “that communica-
tion is always a response by one participant to a prompt by other participants in
social events” (Kress 2010: 35, emphasis original). Accordingly, video clips that
trigger attention and interest may well provide the starting point for corresponding
interpretive activities, which in turn give the original post the status of a prompt.
4. Cohesion on Facebook
Over the past ten years, Facebook has become one of the primary social network-
ing services (SNS) across the world, since it is not only widespread among private
individuals but also used for business and commercial purposes (cf. Dąbrowska
2013: 127–128; Eisenlauer 2013: 32; see also Eisenlauer, Ch. 9, this volume). As
the “Social Network Theory” (Yus 2011: 116, emphasis original) propounds, there
is a close relationship between social bonding and the use of SNS communication,
which will in turn have an effect on cohesion at the linguistic level. As regards their
central functional components, SNSs comprise a “profile”, “a list of other users”
that serve as contacts and the technical possibilities to browse other profiles (boyd
and Ellison 2008: 211). More precisely, typical SNS profiles contain photographs,
names, personal information, friends, actions on the profile, feeds of latest events
and public messages as well as other applications (cf. Yus 2011: 113). Furthermore,
SNSs are characterized by the four significant features of persistence, searcha-
bility, replicability and invisible audiences (cf. Yus 2011: 112). In comparison to
YouTube, Facebook offers more options of initiating, continuing and intensifying
social contacts. As regards audience design on Facebook, a user who posts a com-
ment or status update does not only communicate with a specific addressee but
also needs to take into account “[a]ctive Friends”, “[w]ider Friends” as well as
all internet users who may metaphorically eavesdrop on postings (cf. Tagg 2015:
156). As Lee (cf. 2011: 115–117) points out, status updates can fulfil a wide range
of discursive and potentially cohesive functions, such as responding to the Face-
book prompt, starting a discussion or quoting from other online sources.

Authenticated
Facebook account
home page personal profile
navigational devices profile data advertisements
info layer wall layer photo layer
Figure 1. Structural constituents of a Facebook account
In terms of macrostructural constituents, a typical Facebook account of an individ-

ual user consists of a home page and a personal profile (cf. Figure 1). While users
switch between these two nodes, the frame, which includes the semantically filled
hyperlinks “home” and “friends” as well as icons for friend requests , messages
and notifications , remains unchanged and thus functions as a global orientation
device. The personal home page serves as an interface for various social con-
tact-establishing and -maintaining activities, such as selecting favourites, using
applications or creating groups (cf. Locher 2014: 564). Centrally, the default news
feed initiates a dialogue by means of the prompt “what’s on your mind”, which can
be answered by typing in a text box or by clicking an icon that allows members to
add photos, tag people or inform others about current activities.
The personal profile page of individual users comprises three structural con-
stituents that can be labelled “navigational devices”, “advertising texts” and “pro-
file data” (Eisenlauer 2013: 33). The profile data comprise a variety of details that
other members can possibly relate to, so that the web of social contacts resem-
bles the woven texture of cohesive ties. Within the next level of “profile layers”
(Eisenlauer 2013: 34), the “info” section contains an additive sequence of social
features and individual interests, so that interpersonal connections can be based
on common hobbies or shared pop culture predilections. The “wall” layer gives an
account of the member’s recent Facebook activities, while the “photo” layer con-
tains personal pictures. The advertisements are semantically connected to items in
the user profile, since the advertisers are provided with information on consumer
preferences of their respective target groups. Finally, the navigational devices con-
sist of semantically filled links mainly based on what- and partially on do-trig-
gers, which allow users to find self-selected paths across other profiles. As can
be expected from a networking service, online activities by individual users have
significant repercussions on other nodes, for
the corresponding text actions are heavily interwoven: Changes to member details on
the Personal Profile are dispersed throughout a user’s network of friends via the News
Feed stream. Likewise, commenting on other member’s [sic] posts within the Home
Page environment is recorded on one’s own profile Wall and thus contributes to a mem-
ber’s self-presentation (Eisenlauer 2013: 36).

Authenticated
Consequently, cohesion is to some extent initiated and created by the Facebook

software itself, which automatically links nodes in a cross-medial manner. As dis-
cursive items are generated and distributed by the software, Facebook has been
labelled a “third author” (Eisenlauer 2013: 42) that structurally intervenes in the
communication between network members.
Cohesion on Facebook strongly relies on lexical sets and fields, since most of
the triggers on the site ultimately represent content-related slots. This will be illus-
trated by means of semantic areas realized as drop-down lists or vertical columns
on the home page. For instance, if the semantically filled hyperlink “About” on the
profile page is clicked, the subordinate hypertextual node provides a vertical col-
umn with additional links which subcategorize biographical details (cf. example 3).
(3) About
Overview
Work and Education Work + add a workplace
Professional skills + add a professional skill
College + add a college
High school + add a high school
Places You’ve Lived
Contact and Basic Info
Family and Relationships
Details About You
Life Events
(Facebook, retrieved 12 February 2016, <https://www.facebook.com/>)
The anthropomorphic pictogram of a close shot, located next to the preposition

“About”, indicates the focus on personal matters. The result is an onomasiologi-
cal and hierarchically structured display of this semantic field. The nouns which
are listed here serve as clickable what-triggers offering more specific data. For
instance, when “work and education” is selected, additional options are given at
the centre of the screen, as displayed in example (3). These are do-triggers which
lead to text boxes provided for the user to enter written information. Hence, inter-
nodal cohesion here results from the lexical set curriculum vitae, subsuming facts
about the user’s professional and private life. This is complemented by the sec-
ond-person pronoun you, which in example (3) occurs twice, pointing in exophoric
and endophoric directions, since it refers both to the extralinguistic user and to his
or her “cyberego” (Bublitz 2005), as represented on the screen by a proper name
in combination with a photograph.
Under the Facebook prompt “what’s on your mind?” on the personal home-
page, a laughing emoticon serves as a trigger for posts responding to the more spe-
cific stimulus “add what you’re doing or how you’re feeling”. By clicking on this
pictogram, an extensive drop-down list offers a variety of choices in the form of
present participles or verbal nouns in -ing. If any of them is selected, such as “feel-
ing” or “celebrating”, verbal complements appear, as demonstrated in example (4).

Authenticated
(4) Feeling Celebrating

happy your birthday
loved her birthday
excited his birthday
blessed your special day
sad this special day
thankful success
crazy life
fantastic love
(Facebook, retrieved 12 February 2016, <https://www.facebook.com/>)
Regarding the first domain, “feeling” can be considered as the superordinate noun
or verb, while the adjectives offered represent subordinate terms. Similarly, the
default smiley which represents “feeling” forms an iconic sign with fewer seman-
tic features than the more specific emoticons, so that a semantic hierarchy exists
also on the pictorial level. In the case of the verb form “celebrating”, the drop-
down list provides typical syntactic collocates, which may likewise be subsumed
under the text-linguistic notion of collocation, since their co-occurrence in textual
segments is highly probable. Although grammatical cohesion is rare at the inter-
nodal level, the present drop-down list contains the personal pro-forms “your”,
“his” and “her” as well as the demonstrative determiner “this”. The third-person
pronouns here presuppose social contacts, or, more precisely, Facebook Friends,
whose names have to be retrieved from other nodes of the hypertext, so that cohe-
sion manifests itself in the form of internodal reference. The demonstrative pro-
noun in “this special day” constitutes exophoric reference but also presupposes a
specifying item that needs to be endophorically recovered from a different node.
The ubiquitous “like” button, which expresses positive reactions to a specific
content, fulfils the structural function of a “clickable response device” (West and
Trester 2013: 138). Thus, it forms the second part of a potentially multimodal adja-
cency pair and presupposes a cohesively linked antecedent. The introduction of
further “emoji reactions” in addition to the “like” button was officially announced
by Facebook on 24 February 2016 (“love”, “haha”, “wow”, “sad” and “angry”)
(Facebook Newsroom). With these pictograms, users can express their emotional
responses to posts in more detailed ways than with a general “dislike” button,
whose introduction has been avoided by Facebook. These updates impressively
underline the statement that electronic discourse is a “moving target” (Locher
2014: 557) for any linguistic analysis. As Lee cogently points out, such alterations
“pose real challenges” to discourse analysis but at the same time are “a perfect
opportunity for tracing creative adaptations in people’s new media textual prac-
tices” (2011: 111), for commercial software applications are bound to react to the
communicative needs and preferences of their users.

Authenticated
5. Cohesion on Twitter
Although microblogging is a possibility also offered by Facebook, Twitter has

been the default platform for this digital communicative practice since 2006
(cf. Crystal 2011: 36–56; Yus 2011: 135–149; Zappavigna 2012, 2013, and also
Zappavigna, Ch. 8, this volume). Twitter is based on microposts (i.e. microblog
posts) or “tweets”, which are “time-stamped, chronologically ordered entries”
(Zappavigna 2012: 195), continuously published in microblogging feeds. From a
structural viewpoint, tweets contain three vertically arranged components: (a) the
user name, marked by “@” and accompanied by a photo and the date of the tweet,
(b) the message in the form of text and possibly images, followed by the topic,
which is indicated by the hash character “#”, and (c) options of reacting to the
tweet by means of clickable pictograms. In this way, users can reply to the author,
retweet the post or express approval for the tweet by clicking on an iconic heart
symbol. In particular, retweeting is the practice of forwarding a message to your
“followers”, as it enables users “to republish another user’s tweet within your own
tweet” (Zappavigna 2012: 35).
As opposed to Facebook, Twitter users can have asymmetrical or “non-re-
ciprocal relationships” (Draucker 2015: 52), since their followers do not need to
confirm mutual friendship. This has consequences on the direction of endophoric
reference, as the tweets by followers typically refer back to the prompting tweets.
Moreover, “[i]n contrast to Facebook and YouTube, the user-generated content
in Twitter lends itself well to re-transmission” (Chovanec and Dynel 2015: 9).
Accordingly, this activity of forwarding and spreading messages strongly sup-
ports the intricate web of cohesive ties. Hashtags indicate “the topic of tweets in
a process of ‘ambient affiliation’, whereby people sharing associated values bond
around these user-defined topics” (Zappavigna 2012: 14). This system involves
linguistic cohesion as well, since hashtags connect posts with similar lexical con-
tent and facilitate cross-referencing as well as mutual commenting. Thus, as Zap-
pavigna points out, Twitter is strongly based on the development of “searchable
talk”, in which “we mark our discourse so that it can be found by others, in effect
so that we can bond around particular values” (2012: 1). As a consequence, social
grouping is likely to correspond with linguistic linkage on Twitter in the form of
lexical cohesion. Terms like friends and followers imply the presence of a social
relation, which is reflected by the language use in the posts.
Since the prompt for the Twitter text box was changed from “what are you
doing” to “what’s happening” in 2009, “Twitter now has far fewer isolated post-
ings and far more semantic threads” (Crystal 2011: 11). These cohesive chains
are linked mainly by lexical cohesion, especially by repetition and collocation.
Thus, for an investigation of cohesion on Twitter, it is important to check cohesive
ties both (a) between tweets and their corresponding replies by other users and
(b) between tweets of the same semantic thread posted by one individual user.

Authenticated
Accordingly, cohesion will rely either on a quasi-conversational exchange or on

an additive sequence of contributions. Despite the predominance of lexical link-
age, Crystal (2011: 48) found various items of grammatical cohesion in tweets as
well, such as coordinators, conjunctive adverbs, anaphoric reference and repair
strategies (e.g. “I meant …, really?”). Example (5) is an excerpt from a conversa-
tional polylogue among Twitter users from 12 February 2016. It was retrieved by
clicking on a trigger labelled “view conversation” under the topic “Griffith Obser-
vatory”, a famous sight in Los Angeles, California.
(5) Jonah Green @Jonah Green ∙ Feb 12
Low light grain destroyed pic at the observatory last night actually turned out pretty
sick
Margarita @singingasian ∙ Feb 12
@JonahGreen where where where where ?! also take me!!!
Jonah Green @Jonah Green ∙ Feb 12
@melizabeth143 GRIFFITH OBSERVATORY!!
We all should go next week :)
Margarita @singingasian ∙ Feb 12
@JonahGreen ah! Yes, so down.
Jessica Danielle @ohnoitsjessicaa ∙ Feb 12
@melizabeth143 @JonahGreen take me too Marg or else
Emma:) @emmalemonhillis ∙ Feb 12
@JonahGreen @melizabeth143 and me
(Twitter, “Griffith Observatory”, retrieved 16 February 2016,
<https://twitter.com/search?q=griffith%20observatory&src=typd>)
This polylogue is initiated by tweeter Jonah Green, who posts a photo of him-
self sitting on the wall of the observatory during nighttime. Margarita responds
to his post with a highly elliptical tweet consisting of the repeated interrogative
adverb “where” and an emphatic imperative. Jonah again answers by means of
clausal ellipsis and introduces the inclusive pronoun we in collocation with the
quantifier all. This triggers responses not only by Margarita but also by two other
women, whose tweets are likewise connected mainly by verbal and clausal ellip-
sis. Margarita’s interjection “ah!” in her second tweet signals recognition, while
the response form “[y]es” realizes anaphoric clausal ellipsis. In Jessica Danielle’s
tweet the vocative “Marg” indicates that now the addressee has changed, while the
ellipted destination “GRIFFITH OBSERVATORY!!” remains unchanged. Finally,
the member Emma:) adds the highly elliptical “and me”, presupposing informa-
tion from various antecedent tweets. Thus, as Yus argues, such “sub-sentential
utterances” (2001: 146) are typical of the structurally reduced discourse of Twit-
ter, which calls for endophoric reference. Hence, example (5) corroborates the
fact that Twitter contains grammatical cohesion in polylogic units at the intran-
odal level. In addition, the pervasive repetition of users’ names and the calendar
dates constitute dominant chains of lexical cohesion. In a series of tweets by one

Authenticated
individual user over a period of time, the occurrence of cohesive ties is differ-
ent, as example (6) demonstrates, taken from the feed by American actor Kevin
Spacey.
(6) Kevin Spacey @KevinSpacey ∙ 2 Dec 2015
Here it is … Tardiness will not be tolerated! Learn more about my new project here:
masterclass.com/ks
Kevin Spacey @KevinSpacey ∙ 8h
I’m a man of the people. Even I like to take a selfie outside my house from time to
time … instagram.com/p/BCEjTHpClIX/
Kevin Spacey @KevinSpacey ∙ Feb 20
Two weeks left to apply for the @KS_Foundation #KSFArtistsofChoice £10K/$10K +
mentoring for your creative project
http//bit.ly/KSFArtistsofChoiceapplication
(Twitter, “#Kevin Spacey”, retrieved 22 February 2016,
<https://twitter.com/KevinSpacey>)
In contrast to the polylogue in example (5), these tweets are not addressed to spe-
cific Twitter members but appeal to the wide online public. The list starts with a
“pinned” tweet from 2 December 2015, which invariantly remains at the top of
the stream, since it fulfils an advertising function that is valid over a longer period
of time. By clicking the link in the tweet, the user reaches Spacey’s acting school
website, which is extranodally tied to the present trigger by the repetition of the
noun “masterclass” and by Spacey’s name. In addition, a teaser video serves as a
commercial for his acting lessons, providing multimodal cohesion as well, for the
URL and the video are cataphorically referred to in the topmost tweet by demon-
strative (“here”) as well as personal reference (“it”).
The next tweet in example (6) contains an extranodal link to Instagram, where
a selfie of Spacey in front of the White House can be viewed. The extranodal
cohesive item consists of an exact repetition of the tweeted text, followed by the
greeting “[w]elcome to my instagram” and accompanied by a reappearance of
the actor’s well-known countenance. The third tweet again contains a link to an
external site, here the Kevin Spacey Foundation, which is likewise extranodally
connected through lexical cohesion. This is achieved through morphological rep-
etition in the form of the verb “apply” in the root node and the compound noun
“application form” in the target node. Apart from first-person pronominal refer-
ence (“my”, “I”), the three posts are intranodally linked chiefly by the lexical
repetition of Spacey’s name, his Twitter address and calendrical details. Still, there
is no lack of overall cohesion, for the contextual frame provided by Twitter guar-
antees uninterrupted orientation and thus allows tweets to be perceived as coherent
even in situations of low cohesive density.
When Spacey’s tweets are answered in the form of numerous retweets, this
Twitter thread becomes polylogic, so that discourse markers or continuatives
occur as well. Investigating discourse markers such as right, well, great, okay

Authenticated
or anyway in text messages, Tagg argues that “their functions include opening
or closing interactions, sequencing interaction and marking topic boundaries and
shifts, focusing attention and monitoring shared knowledge” (2012: 105). Since
Twitter may principally be labelled “the SMS of the Internet” (Crystal 2011: 36),
similar uses of discourse markers appear in the retweets responding to Spacey’s
prompting tweet, such as “right then … time to enroll myself on Masterclass!!!”
(@Deborah J Ball ∙ 2 Dec 2015) or “Hi! Well, THIS is a touch of terrific!”
(@dianeborbes129 ∙ 2 Dec 2015). Here the discourse marker “right then” signals
attention and approval of Spacey’s post, while “well” indicates confirmation and
interest. Both of them clearly underline the similarity between Twitter discourse
and the register of informal conversation.
6. Cohesion in Weblogs
In comparison to YouTube, Facebook and Twitter, weblogs are somewhat less

polylogic and interactive, since usually a single author is in control of the site.
Nevertheless, blogs typically contain extranodal hyperlinks and include comment
sections, in which other users may post additional remarks (cf. Yus 2011: 95;
Puschmann 2009: 58–59, and also Heyd, Ch. 6, this volume). Like on other social
media sites, entries appear in reverse chronological order, so that this sequential
structure lays the foundation for possible cohesive patterns. As far as functional
subgenres are concerned, blogs may be related to the users’ private (diaries, friend-
ship weblogs) or professional lives (career, commercial/corporate blogs) (cf. Hoff-
mann 2012: 39). Regarding the fundamental constituents of blogs, it is necessary
to differentiate between two compositional layers (cf. Hoffmann 2012: 19). The
first level contains the primary sections of entries and comments, while the second
refers to peripheral segments that have a paratextual character. In particular, this
fringe on the screen can be divided into the upper, lower and side panels, which
commonly contain the blog title, navigational tools as well as a blog archive and
a blog roll (cf. Hoffmann 2012: 19–26). The presence of extranodal hyperlinks,
which is an essential feature of weblogs, may result in a complex “network of
inter-connected blogs” (Yus 2011: 105).
In terms of cohesion, personal blogs are generally closer to written monologi-
cal texts than to spoken dialogical ones, as demonstrated by Hoffmann (cf. 2012:
206–213). Accordingly, the frequencies of substitution and repetition are quite
low, while collocation and conjunction show relatively high quantities. Ellipsis is
scarce in blog entries but occurs more regularly in comments, whereas there is only
little cohesion between entries and comments. Moreover, blogs hardly contain any
cohesion between different comments, so that “[m]ost entries and comments are
individual, autonomous monologues which showed only a few intertextual bonds”
(Hoffmann 2012: 209). As for personal pronouns, first-person reference is particu-

Authenticated
larly salient, placing the bloggers and their opinions at the centre of interest (cf.
Puschmann 2009: 74).
Myers (cf. 2010: 28–47) discusses “creative linking” in blogs and emphasizes
their strong hyper-intertextual character, which is caused by high numbers of links.
He reaches the conclusion that “[t]here have been many studies of what bloggers
link to, but rather few of what they link with, the information contained in the
text about the link” (Myers 2010: 34). For instance, if longer quotations or textual
segments are used in the highlighted trigger of the root node, extranodal lexical
cohesion is often particularly strong, since it may be based on several literal repeti-
tions. Alternatively, if the trigger is formed by single words, names, titles or URLs,
there are usually fewer cohesive ties at work. Multimodal cohesion is established
by images as triggers, while deictic expressions functioning as semantically empty
links may give rise to grammatical cohesion. As blogs typically contain individual
opinions about the respective topics under discussion, stance markers are quite fre-
quent, especially in the comment sections. As Myers shows (cf. 2013: 269–270),
this practice of stance-taking includes items that can also contribute to cohesion,
such as adversative conjunction with but or conversational features in the form of
discourse markers or response forms.
For the purpose of illustration, extracts from the popular blog entitled “1000
Awesome Things” (<http://1000awesomethings.com/>) will be employed. This
site is operated and regularly updated by Canadian-born author and blogger Neil
Pasricha, who intended to post notes about the simple pleasures of everyday life
on a daily basis. The single entries range in length from only a few words to longer
stretches of text in combination with images. Since these posts internally behave
like linear texts in terms of cohesion, the focus will here be on the interplay of
entries, comments and hyperlinks. Example (7) contains entry #43 from 21 Febru-
ary 2012, followed by four brief comments by bloggers.
(7) #43 Finding out from a doctor or mechanic that nothing’s wrong
February 21, 2012 Neil Pasricha
Weird thump, strange lump, funny bump? It’s nothing. AWESOME!
48 thoughts on “#43 Finding out from a doctor or mechanic that nothing’s wrong”
ATUL10ER
February 21, 2012 at 12:11 am
Very True..
WENDY
Oh, I should probably see *that* doctor, or mechanic perhaps …hahaha:)
ATUL JAIN
lol
CASEY
Finding awesome in everyday things.. I love it.

Authenticated
(“1000 Awesome Things”, retrieved 23 February 2016, <http://1000awesomethings.

com/2012/02/21/43-finding-out-from-a-doctor-or-mechanic-that-nothings-wrong/>)
In contrast to personal weblogs, this site addresses a wide online audience and
has a commercial background, since it advertises topically related books by the
blogger. Although the comments quoted here are very short, they show cohesive
ties both with the original entry and among themselves. The prompting entry is
accompanied by a photograph of a doctor giving the viewer literally the thumbs
up, so that the image cohesively reinforces the verbal meaning of the post. The
comments are separated from and connected to the entry by the lexically cohesive
link “48 thoughts on […]”. The first comment shows clausal ellipsis with regard
to the entry, for the attribution of the adjective phrase “[v]ery [t]rue” presupposes
the complete antecedent post. The second entry starts with the interjection “[o]h”,
which functions as a discourse marker of mild surprise, while the demonstrative
determiner “that” cross-modally refers to the physician in the photo of the entry.
Since this comment ends with the humorous interjection “hahaha:)”, the ensuing
comment “lol” is directly tied to the immediately previous one. The interconnec-
tion of these comments is strongly supported by the exact posting time, since this
detail indicates the rapid succession of the messages within minutes. The final
comment quoted here is linked to the entry by the repetition of “awesome” and
by personal reference (“it”), which in this case establishes extended reference not
only to the present entry but to the blog in its entirety.
As for global orientation devices, the blog’s upper panel contains the naviga-
tional links “HOME”, “ABOUT”, “BOOK”, “BOOK 2”, “THE TOP 1000” and
“SUBMIT”. At the left-hand side, bloggers find an optional panel containing the
icons of the current social media sites Pinterest, Google+, Facebook and Twit-
ter. The right-hand panel comprises advertisements of the author’s bestselling and
upcoming publications, a search box, contact information, the top posts and a list
of newspapers and TV channels under the title “FEATURED IN”. Thus, the intran-
odal cohesion present in example (7) is complemented by internodal ties between
the 1,000 entries of the blog and by extranodal hyperlinks. However, a blogroll in
the narrow sense of the word is not offered here, which may be due to the self-con-
tained commercial framework of this blog.
7. Summary and conclusions
To sum up, both the theoretical considerations and the illustrative examples have
demonstrated that the conceptual outline of social media platforms, whose main
purpose is interpersonal networking, encompasses linguistic and multimodal cohe-
sion. Along these lines, the metaphors of follower or friend already presuppose an
interest in the semantic content of other users’ posts, which often results in cohe-

Authenticated
sive ties as well. Thus, Tanskanen’s (cf. 2006: 23–27) notion of textual collabora-
tion is particularly salient in the case of social media, in which chains of meaning
are collectively negotiated by users. For instance, whenever a topic is initiated by
a brief post and sequentially elaborated on by a number of users, this process can
be labelled collaborative expansion in the area of lexical cohesion. Moreover, this
concerted effort is to some extent supported and intensified by automatized con-
nections that are established by the social media software.
From the perspective of hypertextual nodes, intranodal texture comprises a
variety of both grammatical and lexical cohesive ties, while inter- and extranodal
texture is realized mainly by lexical cohesion. The most common ties are literal
repetition and collocation in the form of lexical sets, since these types of cohesion
do not imply a search instruction and thus do not demand a particular referential
direction. Between chronological posts, cohesive ties are only possible anaphor-
ically, for although users per se need to rely on forward-looking strategies (cf.
Fritz 1999: 230), cataphoric cohesion dependent on other users’ future posts is not
feasible. Whenever the cohesive tie of reference is present across individual posts,
the search instruction evoked by pronouns is more demanding than in linear texts.
The presupposed item may be more difficult to find, for it can be realized in mul-
timodal form and in order to retrieve it users might have to scroll up.
With regard to hypermedial design, the frequent use of images in social media
results in pervasive cross-modal cohesion, particularly at the personal level, when,
for instance, interpersonal pronouns or proper names refer to photographs of users.
Furthermore, iconic pictograms typically function as do-triggers, allowing users
to express their emotions or opening other nodes and platforms. In general, the
informationally dense and emotionally appealing character of most iconic images
may attract users’ attention and trigger semantic threads more easily than purely
symbolic signs.
Although the four social media platforms discussed in this contribution share
many cohesive features due to their similar communicative functions, a few signif-
icant differences can be determined as well. First, regarding cross-modal cohesion,
YouTube is predestined for word-image connections, since short video clips are at
the very centre of this platform, while on the other three sites, images are not the
starting point but rather fulfil supportive functions. Second, as for author control,
the original creators of weblogs can exert influence over their entries and Face-
book users can supervise their own profiles. By contrast, since the discourses of
YouTube and Twitter represent mosaics of individual contributions, lower degrees
of cohesive uniformity are present on these platforms. Thus, owing to the prac-
tice of microblogging, Twitter shows the highest degree of textual fragmentation,
whereas weblogs often contain more extended textual chunks that are internally
marked by greater cohesive density. Accordingly, the conversational register and
concomitant cohesive devices such as ellipsis are most salient on YouTube and
Twitter, for these sites are based almost exclusively on dialogical interaction.

Authenticated
Third, concerning structural composition, the predetermined autobiographical

slots on Facebook constitute a relatively fixed template that results in a given set
of collocations, whereas the other platforms do not offer such prearranged lexical
fields. Hence, in the conceptual layout of Facebook, individual self-presentation
through personal profiles serves as the point of departure for cohesion, while in
blog comments, YouTube posts or retweets, cohesive threads are typically initiated
by prompting posts or video clips.
Provided by the social media software, explicit markers of semantic threads,
such as the hashtag, help users to trace cohesive chains. Conclusively, the hyper-
cohesive “harmony” (Halliday and Hasan 1985: 94) between interlinked chains is
strongly supported by these digital devices. Whenever cohesive ties between indi-
vidual posts are scarce, global orientation and contextualization cues play a signif-
icant role. Despite these hypertextual tools, hypercohesion is not as text-centered
as linear cohesion, since the recipient not only has more freedom in establishing
hypertrail texture but also needs to expend more effort in tracking cohesive ties.
Having acquired the necessary New Media literacy, however, average users are
acquainted with the challenges of social media cohesion, so that the (re-)construc-
tion of hypertexture is usually unproblematic.
References
Adami, Elisabetta
2013 Participatory culture and metalinguistic discourse: Performing and negotiating
German dialects on YouTube. In: Deborah Tannen and Anna Marie Trester
(eds.), Discourse 2.0: Language and New Media, 47–71. Washington, DC:
Georgetown University Press.
Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad and Edward Finegan
1999 Longman Grammar of Spoken and Written English. London: Longman.
Boardman, Mark
2005 The Language of Websites. London: Routledge.
2012 Social interaction in YouTube text-based polylogues: A study of coherence.
2008 Social Network Sites: definition, history, and scholarship. Journal of Computer-
Brown, Gillian and George Yule
1983 Discourse Analysis. Cambridge: Cambridge University Press.

Authenticated
Bublitz, Wolfram
2005 The user as ‘cyberego’: Text, hypertext and coherence. In: Lilo Moessner and
Christa M. Schmidt (eds.), Anglistentag 2004 Aachen: Proceedings, 311–324.
Trier: Wissenschaftlicher Verlag Trier.
Bublitz, Wolfram
2006 It utterly boggles the mind: knowledge, common ground and coherence. In:
Hanna Pishwa (ed.), Language and Memory: Aspects of Knowledge Represen
tation, 359–386. Berlin/Boston: de Gruyter.
Chandler, Daniel
2007 Semiotics: The Basics. 2nd ed. London: Routledge.
Chovanec, Jan and Marty Dynel
media. In: Marty Dynel and Jan Chovanec (eds.), Participation in Public and
Crystal, David
Crystal, David
2011 Internet Linguistics: A Student Guide. London: Routledge.
Dąbrowska, Marta
2013 Variation in language: Faces of Facebook English. Frankfurt am Main: Peter
Lang.
De Beaugrande, Robert-Alain and Wolfgang Ulrich Dressler
1981 Introduction to Text Linguistics. London: Longman.
Dontcheva-Navratilova, Olga and Renata Povolná, eds.
2009 Coherence and Cohesion in Spoken and Written Discourse. Newcastle: Cam-
bridge Scholars.
Draucker, Fawn
2015 Participation structures in Twitter interaction: Arguing for the broadcaster
role. In: Marty Dynel and Jan Chovanec (eds.), Participation in Public and
Eisenlauer, Volker
2013 A Critical Hypertext Analysis of Social Media: The True Colours of Facebook.
Facebook Newsroom. Reactions Now Available Globally. 24 February 2016. http://news-
room.fb.com/news/2016/02/reactions-now-available-globally/ (retrieved 26
February 2016).
Flowerdew, John and Michaela Mahlberg (eds.)
2009 Lexical Cohesion and Corpus Linguistics. Amsterdam/Philadelphia: Benja-
mins.
Frehner, Carmen
2008 E-Mail – SMS – MMS: The Linguistic Creativity of Asynchronous Discourse in
the New Media Age. Bern: Peter Lang.
Fritz, Gerd
1999 Coherence in hypertext. In: Wolfram Bublitz, Uta Lenk and Eija Ventola
(eds.), Coherence in Spoken and Written Discourse: How to Create it and
How to Describe it, 221–232. Amsterdam/Philadelphia: Benjamins.

Authenticated
Gutwinski, Waldemar
1976 Cohesion in Literary Texts: A Study of Some Grammatical and Lexical Fea-
tures of English Discourse. The Hague: Mouton.
1976 Cohesion in English. London: Longman.
1985 Language, Context, and Text: Aspects of Language in a Social-Semiotic Per-
spective. Geelong, Victoria: Deakin University.
Halliday, Michael A.K. and Christian M.I.M. Matthiessen
2004 An Introduction to Functional Grammar. 3rd ed. London: Hodder Education.
Herring, Susan
1999 Interactional coherence in CMC. Journal of Computer-Mediated Communica-
tion 4. doi: 10.1111/j.1083-6101.1999.tb00106.x.
Herring, Susan C., Dieter Stein and Tuija Virtanen
2012 Introduction to the pragmatics of computer-mediated communication. In:
puter-mediated Communication, 3–32. Berlin/Boston: de Gruyter.
Herring, Susan C.
2013 Discourse in Web 2.0: familiar, reconfigured, and emergent. In: Deborah Tan-
nen and Anna Marie Trester (eds.), Discourse 2.0: Language and New Media,
2012 Cohesive Profiling: Meaning and Interaction in Personal Weblogs. Amster-
Jucker, Andreas
2002 Hypertextlinguistics: Textuality and typology of hypertext. In: Andreas Fischer,
Gunnel Tottie and Hans Martin Lehmann (eds.), Text Types and Corpora: Stud-
ies in Honour of Udo Fries, 29–51. Tübingen: Narr.
Jucker, Andreas H.
2005 Hypertext research: Some basic concepts. In: Lilo Moessner and Christa
M. Schmidt (eds.), Anglistentag 2004 Aachen: Proceedings, 285–295. Trier:
Klein, Yvonne
2012 Cohesion in English and German. In: Silvia Hansen-Schirra, Stella Neumann
and Erich Steiner (eds.), Cross-Linguistic Corpora for the Study of Transla-
tions: Insights from the Language Pair English-German, 161–172. Berlin/
Boston: de Gruyter.
Kress, Gunther
2010 Multimodality: A Social Semiotic Approach to Contemporary Communication.
London: Routledge.
Kress, Gunther and Theo van Leeuwen
2006 Reading Images: The Grammar of Visual Design. 2nd ed. London: Routledge.
Lee, Carmen K. M.
2011 Micro-blogging and status updates on Facebook: texts and practices. In:
Crispin Thurlow and Kristine Mroczek (eds.), Digital Discourse. Language in
the New Media, 110–128. Oxford: Oxford University Press.

Authenticated
Locher, Miriam A.
2014 Electronic discourse. In: Klaus P. Schneider and Anne Barron (eds.), Pragmat-
Markman, Kris
2012 Conversational coherence in small group chat. In: Susan Herring, Dieter Stein
and Tuija Virtanen (eds.), Pragmatics of Computer-Mediated Communication,
Martin, James R.
2001 Cohesion and Texture. In: Deborah Schiffrin, Deborah Tannen and Heidi
E. Hamilton (eds.), The Handbook of Discourse Analysis, 35–53. Oxford:
Blackwell.
Myers, Greg
2010 Discourse of Blogs and Wikis. London: Continuum.
Myers, Greg
2013 Stance in blogs. In: Ken Hyland (ed.), Discourse Studies Reader: Essential
Excerpts, 253–271. London: Bloomsbury.
Neumann, Stella
2014 Contrastive Register Variation: A Quantitative Approach to the Comparison of
English and German. Berlin/Boston: de Gruyter.
Neumann, Stella and Jennifer Fest
2016 Cohesive devices across registers and varieties: The role of medium in Eng-
lish. In: Christoph Schubert and Christina Sanchez-Stockhammer (eds.), Vari-
ational Text Linguistics: Revisiting Register in English, 195–220. Berlin/Bos-
ton: de Gruyter.
Nöth, Winfried
1995 Handbook of Semiotics. Bloomington: Indiana University Press.
2009 Lies at Wal-Mart: style and the subversion of genre in the Life at Wal-Mart
blog. In: Janet Giltrow and Dieter Stein (eds.), Genres in the Internet: Issues
in the Theory of Genre, 49–84. Amsterdam/Philadelphia: Benjamins.
Schubert, Christoph
2009 Raumkonstitution durch Sprache: Blickführung, Bildschemata und Kohäsion
in Deskriptionssequenzen englischer Texte. Tübingen: Niemeyer.
Schubert, Christoph
2012 Englische Textlinguistik: Eine Einführung. 2nd ed. Berlin: Erich Schmidt.
2013 Spoken and Written Discourse in Online Interactions: A Multimodal Approach.
New York, NY: Routledge.
Storrer, Angelika
1999 Kohärenz in Text und Hypertext. In: Henning Lobin (ed.), Text im digitalen
Medium: Linguistische Aspekte von Textdesign, Texttechnologie und Hyper-
text Engineering, 33–65. Opladen/Wiesbaden: Westdeutscher Verlag.
Storrer, Angelika
2000 Was ist ‘hyper’ am Hypertext? In: Werner Kallmeyer (ed.), Sprache und neue
Medien, 222–249. Berlin/Boston: de Gruyter.
Tagg, Caroline
2012 The Discourse of Text Messaging: Analysis of SMS Communication. London:
Continuum.

Authenticated
Tagg, Caroline
Tannen, Deborah and Anna Marie Trester (eds.)
2013 Discourse 2.0: Language and New Media. Washington, DC: Georgetown Uni-
versity Press.
Tanskanen, Sanna-Kaisa
2006 Collaborating towards Coherence: Lexical Cohesion in English Discourse.
Tseng, Chiao-I.
2013 Cohesion in Film: Tracking Film Elements. Houndmills: Palgrave Macmillan.
Van Leeuwen, Theo
2005 Introducing Social Semiotics. London: Routledge.
West, Laura and Anna Marie Trester
2013 Facework on Facebook: Conversations on social media. In: Deborah Tannen
and Anna Marie Trester (eds.), Discourse 2.0: Language and New Media,
Yus, Francisco
Zappavigna, Michele
2012 Discourse of Twitter and Social Media: How we Use Language to Create Affil-
Zappavigna, Michele
2013 The language of tweets. In: Ken Hyland (ed.), Discourse Studies Reader, 301–
327. London: Bloomsbury.

Authenticated
Authenticated
13. Discourse and cognition
Andreas Langlotz
Abstract: This chapter focusses on the relationship between discourse and cog-
nition by concentrating on one social medium, Twitter, and by analysing concrete
tweets as discursive examples. Starting from a working definition of the central
notions ‘discourse’ and ‘cognition’, it provides a brief historical overview of the
recent convergence between discourse analysis and cognitive science. In line with
this trend, the chapter adopts a socio-cognitive perspective of discourse manage-
ment. With reference to Herbert Clark’s theory of language use (Clark 1996), it
outlines the socio-cognitive parameters that play a decisive role in the human
management of discourse in general and the management of tweets in particular.
Finally, the socio-cognitive affordances and limitations of Twitter will be briefly
surveyed with regard to the following questions: How can users perform joint
actions online? What signalling strategies can they use for doing so? How do they
manage and creatively play with common-ground construction online?
1. Introduction
More than 20 years ago, in the early phases of research on computer-mediated

communication (CMC), Nancy Baym predicted the great social potential of new
media:
If anything, the ways in which people have appropriated the commercial and noncom-
mercial networks demonstrate that CMC not only lends itself to social uses but is, in
fact, a site for an unusual amount of social creativity. […] participants in CMC devel-
op forms of expression which enable them to communicate social information and to
create and codify group-specific meanings, socially negotiate group-specific identities,
form relationships […], and create norms which serve to organize interaction and to
maintain desirable social climates. (Baym 1995: 160)
Indeed, the massive spread and proliferation of social media platforms and social
networking sites (SNS) such as Twitter, Facebook, Instagram, flickr, Pinterest,
Linkedin, and many more prove that human beings enjoy communicating with each
other online. As psychologists and neuroscientists are showing, the great amount
of communication on and through social media also has an impact on human psy-
chology and cognition (Meshi, Tamir and Heekeren 2015). For instance, SNS such
as Facebook or Twitter are reported to increase working memory and spelling
abilities and to enhance the feeling of social connectedness (Alloway and Alloway
2012; Alloway et al. 2013). On the other hand, various studies have supported
DOI 10.1515/9783110431070-013
Authenticated
346 Andreas Langlotz
arguments that point to more negative cognitive and psychological consequences

for “homo zappiens” (Veen and Vrakking 2006), especially on the youth (Ahn
2011): multitasking and information overload on SNS reduces academic perfor-
mance (Kirschner and Karpinski 2010; Junco and Cotton 2012); social media
reduce the feeling of socioemotional well-being in young girls (Pea et al. 2012);
and they may increase feelings of social anxiety and depression (Becker, Alzahabi
and Hopwood 2013).
While the reported insights into the striking adaptive effects of new media on
the neuropsychological, psychological, and cognitive setup of their users are very
interesting and intriguing, even more fundamental questions have to be asked from
the perspective of linguistic pragmatics. Social media present human beings with
particular interactional settings or socio-communicative environments (Langlotz
2015a: 52). Apart from written verbal communication, these platforms also include
a number of additional communicative tools (icons, images, emoticons, videos,
etc.) that allow their users to organise their interactions with other online com-
municators. On the other hand, the media – excluding Skype or other audiovisual
platforms such as Snapchat, Youtube, videochats, and videoblogs – do not support
the immediate use of embodied communicative modes such as facial expressions
and body language that are naturally employed in face-to-face interaction (Bublitz
2013: 29). To become able to navigate through the complex and dynamic semiotic
worlds established by and through social media, communicators need to exploit
their cognitive, social, and emotional capacities in order to find orientation for
themselves and to create orientation for their communicative partners. Therefore,
from a pragmatic point of view the impact of social media can only be explained
if we better understand the cognitive tasks and challenges – closely linked to the
corresponding social-interactional and emotional challenges (Langlotz and Locher
2012) – that these new communicative environments present to their users in the
first place. Put differently, what socio-cognitive skills do online users have to bring
to the task that allow them to find social and communicative orientation in them?
Harré (2001: 695–696) proposes a threefold distinction between tool, task, and
skill to discuss the complex interrelationship between discourse, socio-commu-
nicative tasks, and the cognitive capacities that support them. By analogy with
this threefold distinction we can illustrate the interdependent relationship between
cognition, discourse and social media by Figure 1. I have used the term tripartite
reciprocal adaptation to denote the interdependent relationship between the cog-
nitive skills, the communicative tools, and the discursive tasks (Langlotz 2015a:
111–115) that communicators are dealing with in communicative activities. We
can employ this general idea in order to scrutinise the complex relationships
between discourse, cognition, and social media in the present chapter. With regard
to the model we can ask three general questions with each question addressing one
fundamental relationship in Figure 1:

Authenticated
Discourse and cognition 347
1. What cognitive skills are necessary for the management of discourse?

2. By means of which socio-cognitive processes is discourse managed in general
and when communicating through social media?
3. How are these cognitive skills and socio-cognitive processes affected when
cognisers adapt to the technological affordances and constraints provided by
these platforms?
Figure 1. Tripartite reciprocal adaptation between cognition, discourse, and social media
(adapted from Langlotz 2015a: 112)
To engage with the complex interrelations between cognition, discourse, and

social media along these interrogative lines, the chapter concentrates on one social
medium, namely Twitter (see Zappavigna, Ch. 8, this volume), only. This limi-
tation of scope to the micro-blog is taken due to reasons of space and coherence.
However, it is assumed that the general points addressed in this chapter can also
be applied to other social media. Engaging with concrete tweets as discursive
examples throughout, the chapter is divided into three main sections. After this
introduction, Section 2 discusses the relationship between cognition and discourse
and gives a brief historical overview of the more recent convergence between dis-
course analysis and cognitive science as a growing area of linguistic interest. In the
third Section, the socio-cognitive parameters that play a decisive role in the human
management of discourse will be specified with reference to Herbert Clark’s theory
of language use (Clark 1996). In Section 4, we will then zoom in on the particular
discursive environment offered by Twitter to deal with question 2 and 3, i.e. how
its particular features involve special socio-cognitive challenges and opportunities
for its users. In doing so, some concrete applications, for which investigating the

Authenticated
relationship between cognition and discourse on Twitter is particularly interesting,

will be presented: How can users perform joint actions online? (Section 4.1) What
signalling strategies can they use for doing so? (4.2) How do they manage and
creatively play with common-ground construction online? (4.3). The chapter then
ends with a short conclusion that reinterprets the insights from the above-men-
tioned psychological studies from the perspective of socio-cognitive linguistics.
2. Discourse and cognition – What cognitive skills are necessary to

manage discursive tasks both online and offline?
2.1. What is discourse? What can be achieved through it?

Following Schiffrin (2006: 169), I define discourse as “the use of language
above and beyond the sentence: how people use language in texts and contexts”.
While the preposition ‘above’ points to the fact that discourse analysts are inter-
ested in communicative phenomena that are more complex than single sentences
or utterances, the word ‘beyond’ highlights that discourse involves much more
than merely conveying and sharing information (see also Jaworski and Coupland
1999). One can refer to Halliday’s famous distinction between three “communica-
tive meta-functions” (Halliday and Matthiessen 2004: 29–31) to scrutinise what
tasks communicators can manage through discursive engagement.
(1) Ideational function:
a. represent the world of experience (including inner-psychic and fictional worlds of
experience)
b. convey ideas to communicative partners
(2) Textual function:
a. structure discursive activities into distinct communicative actions
b. make information accessible and mutually comprehensible
(3) Interpersonal function:
a. engage in interactions
b. establish relationships
c. show stances
d. convey identities
e. negotiate power
The following tweet by Barack Obama can be used to illustrate these dimensions.
Tweet 1. President Barack Obama’s first official tweet

Authenticated
(1) Ideational function:

a. With his tweet, the President represents the new state-of-affairs about him having
finally got his own presidential Twitter account
b. He also communicates his view that he got it rather late, after six years of being in
office.
(2) Textual function:
a. The President clearly structures his message into two parts: a greeting to the Twitter
community Hello, Twitter! It’s Barack. Really! and a critical statement about how
long it took for him to get his presidential Twitter account. The short message fea-
tures the brevity that can be expected from a tweet.
b. To render his message more emphatically and to hammer it in, the President employs
four speech acts: an address formula (Hello, Twitter!), a personal introduction (It’s
Barack.), an intensifier (Really!), and an explanatory statement (Six years in, they’re
finally giving me my own account). He also uses exclamation marks for emphasis.
(3) Interpersonal function:
a. The President connects to other people using Twitter.
b. The way in which he addresses the community indicates that he probably wishes to
establish a relationship to a group of young and modern people who are not com-
monly in contact with the most powerful man on the planet. This is best shown by
his use of his first name, Barack. In doing so, he symbolically puts himself on the
same hierarchical level as other Twitter users.
c. Obama jocularly expresses his critical stance against the circumstances of his late
joining of the community. It was they who gave him the account late. They probably
refers to the US governmental administration. But this does not become fully clear
from the message.
d. Various elements in the message are relevant to convey Obama’s personal and pro-
fessional identity. The image depicts Obama. Next to the image his official title and
last name are given: President Obama. This information is accompanied by the user-
name/handle @POTUS – an acronym for President Of The United States. Moreover,
Obama reveals his personal identity by using his first name.
e. The President clearly avoids a demonstration of his authority and political power.
On the contrary, he establishes the personalised image of a friendly, funny, self-
ironic and jovial Twitter-friend.
It seems uncontroversial that both producers and interpreters of such messages

need considerable cognitive skills to competently manage the highly complex
array of communicative tasks that are achieved through such a message. In the
next section, these cognitive skills are sketched.
2.2. What is cognition? What skills does it comprise?

The Merriam Webster Dictionary simply defines cognition as: “conscious men-
tal activities: the activities of thinking, understanding, learning, and remember-
ing” (http://www.merriam-webster.com/dictionary/cognition). While such a basic
definition is certainly not sufficient from a scientific perspective, it points to the
central capacities that a cogniser must be able to achieve when engaging with his

Authenticated
Figure 2. The dynamics of sense-making
or her world of experience. More technically speaking, cognition comprises the

intelligent ability to produce meaningful and creative behavioural responses by
applying mental representations and processes to the input of environmental stim-
uli (Bless, Fiedler and Strack 2004: 8). This view of the mind encapsulates two
central metaphors. On the one hand, the mind is conceived as a container which
stores mental representations, including ideas, thoughts, and beliefs. On the other
hand, the mind works like a machine that can manipulate this mental content in
working memory to produce a behavioural response, which allows the cogniser to
solve a problem or manage a task appropriately (see also Langlotz 2015a: 88–89).
The psychologist Gibson (1979) introduced the term affordance to point to the
fact that a cogniser can only engage with the world of experience – in our case
a social medium – to the extent that this is made possible by his or her cognitive
capacities and the range and variety of stimuli that emanate from the environment.
Along these lines, Figure 2 depicts the interaction between a cogniser (on the left)
and the social medium as his or her world of experience (on the right). The plat-
form contains tweets as stimuli to which a cogniser can potentially orient. This
pool of stimuli both affords and constrains the cogniser’s potential contact with
the world of experience. Contact is further afforded and constrained by the cognis-
er’s sensorimotor system as well as his or her interaction with the emitted stimuli.
Only those stimuli that come into the cogniser’s limited scope of attention can be
perceived by him or her. These factors conjointly determine the set of potential

Authenticated
conceptualisations that result from the interplay of the cognitive and environmen-
tal constraints and affordances (see also A. Clark 1997: 44). The activated con-
ceptual patterns can then be mapped back onto the stimuli to make sense of them.
While the messages in the social medium influence the activation of sense-making
processes bottom-up, the cognitive capacities determine sense-making top-down.
This turns sense-making into a highly dynamic and adaptive process.
When applying this general model to the task of understanding social media
discourse, one can sketchily describe the necessary cognitive skills as follows:
(A) Perception:
i. orient to and detect the message, i.e. bring it to one’s scope of attention;
ii. actively attend to the presented signals by which the tweet is made up.
(B) Categorisation and conceptualisation:
i. categorise the input relative to previous knowledge that is stored in long-term mem-
ory;
ii. activate frames as knowledge schemas (see Ziem 2014: 218–242) relative to which
the input can be understood in a broader cognitive context.
(C) Making inferences and judgements:
i. infer the speaker’s meaning by computing, integrating or blending the activated
conceptualisations into complex and meaningful situated conceptualisations (Bar-
salou 2005);
ii. scrutinise the message for its potential informational, social, and emotional impli-
cations;
iii. take an affective stance towards the underlying conceptualisations.
When applying these dimensions to the President’s tweet, we can describe the
underlying cognitive dimensions as follows:
(A) Perception:
i. Readers must detect the tweet on their computational device (computer, mobile
phone) and orient to it.
ii. They must attend to the multimodal semiotic environment presented by the tweet,
including its imagery, verbal components and layout. Paying attention to the mes-
sage corresponds to representing it in working memory. When posting the message,
Obama had to exploit the semiotic affordances (Kress 2011: 55) offered by the tech-
nological environment of Twitter in a way that fosters attention to and engagement
with his personal message.
(B) Categorisation and conceptualisation:
i. Readers must categorise the multimodal input in working memory. This involves:
– communicative categorisation (recognising the message as a tweet with its medi-
um-specific elements, e.g. the handle);
– linguistic categorisation (recognising the verbal signals);
– social categorisation (recognising Obama in the image and rightly classifying
this person as the President of the United States). When composing the message,
the President had to try and make sure that the signs used support and appropri-
ately enable such processes of categorisation.

Authenticated
ii. To support these categorisation processes the readers have to activate conceptual
structures or frames from long-term memory that provide the necessary knowledge
about:
– the communicative activity (What is Twitter? How does it work?) = discourse
frames;
– the types of speech-activities (Why does the President present the message in
this particular form?) = speech-activity frames about greeting, introducing one-
self, giving an explanation;
– linguistic knowledge (e.g. What does @POTUS stand for?) = semantic frames;
– social knowledge (Who is Barack Obama?) = social frames;
– world knowledge (How is American presidency organised?) = encyclopedic
frames
As a sender of his message, Obama must have assumed that readers can (partially)
retrieve these chunks of knowledge when engaging with his tweet.
(C) Making inferences and judgements:
i. The readers must integrate the activated chunks of knowledge about Obama’s cur-
rent presidency, about Twitter, about Obama, and about the form and meaning of the
message into a complex conceptual network in order to;
ii. appreciate the informational, social, and emotional value of the message, including
the relationship that it creates to the audience, the identity it creates for Obama, and
the affective stance that it expresses. When producing the message, Obama must
have intended to change the mental representations of his readers via the conceptual
knowledge activated in (B) in order to achieve several aims along the three commu-
nicative metafunctions outlined above.
iii. The message may also trigger emotional reactions in the audience. This affec-
tive stance may cause them to appreciate or reject Obama’s communicative move
(Langlotz and Locher 2012).
Although this overview of cognitive representations and processes that have to be

recruited for understanding even a very short message like Obama’s tweet is only
cursory and superficial, it can give us an inkling of the wide-ranging cognitive
skills that are involved in finding and structuring communicative orientation in
social media. Several cognitive-linguistic, pragmatic, and socio-cognitive frame-
works exist that model different dimensions of these cognitive skills and could
potentially offer alternative descriptive, explanatory, and methodological frame-
works to engage with them. As it is impossible to present these approaches in any
sufficient detail here, Table 1 is meant to provide readers with a rough map of how
they can be connected to the dimensions sketched above. Interested readers can
then explore these frameworks further via the references:

Authenticated
Table 1. Overview of cognitively-oriented frameworks modelling sense-making in and

through discourse
Theory Short description of Typical questions that References

cognitive dimensions could be asked about the
covered tweet
Relevance Post-Gricean theory that How can cognisers compute Sperber
theory focuses on the cognitive the utterance meaning of and Wilson
processes of meaning gen- Obama’s tweet on the basis (1995)
eration in context and that of the linguistic input in
models language-based combination with contex-
meaning inference tual knowledge?
Frame Analyses frames as complex What semantic frames Fillmore
theory knowledge schemas that are activated by Obama’s (1982);
are recruited for (linguistic) words? What broader Cienki
sense-making. Promotes an knowledge do they entail (2010)
encyclopedic view of mean- for making sense of the
ing. tweet?
Cognitive Semantic theory that focuses What conceptual categories Cruse and
semantics on how linguistic units are are activated by Obama? Croft (2004)
related to conceptual struc- How is conceptual meton-
tures (including metaphor ymy relevant for under-
and metonymy) to catego- standing the phase six years
rise the world of experience. in?
Cognitive Usage-based theory of How does the grammatical Langacker
grammar language that scrutinises form of Obama’s statement (2008)
the relationship between conceptualise the world of
language (use) and concep- experience?
tualisation.
Mental Cognitive theory of online How are mental spaces Fauconnier
space and processes involved in situ- activated and blended for and Turner
blending ated sense-making. Focusses the situated construction of (2002)
theory the conceptual creativity meaning?
involved in sense-making.
Socio- Diverse approaches that How are cognitive pro- Clark
cognitive combine insights from cesses of conceptualisation (1996);
linguistics discourse analysis with coupled with social pro- Van Dijk
cognitive heuristics in order cesses with interaction? (2010);
to model the interaction (see Section 3) Langlotz
of social and cognitive (2015);
components involved in Zima (2013)
sense-making (see Sec-
tion 3).

Authenticated
Theory Short description of Typical questions that References

cognitive dimensions could be asked about the
covered tweet
Discursive Theory of social psychology How does Obama discur- Edwards
psychology that employs conversa- sively construct his social (2005)
tion-analytical methods to identity as the President on
deal with cognitive concepts Twitter?
such as emotion, memory,
social categorisation or eval-
uation
Discourse Psycholinguistic models of What are the incremental Graesser et
processing how discourse is processed. steps involved in processing al. (2003)
the tweet?
Dialogic Model of how syntactic reso- How does formal repetition Du Bois
syntax nance in interaction supports of Obama’s phrasing reflect (2014)
cognitive processes of infer- intersubjectivity?
encing and stance-taking.
To the best of my knowledge, only very few of these frameworks have been sys-
tematically applied for the explicit purpose of analysing discourse in social media
from a cognitive angle. Wherever possible I will refer to such studies in the fol-
lowing parts of this chapter. The striking absence of cognitively inspired linguis-
tic research of social media discourse is probably due to historical, theoretical,
and methodological reasons. Indeed, a fruitful combination of discourse analy-
sis with cognitive science has only emerged recently. This convergence is shortly
addressed in the next section.
2.3. Bringing discourse and cognition together

From a historical perspective cognitive approaches to discourse had been sleep-
ing for a long time because cognitive science – due to the primacy of cognitivism
(Bechtel, Abrahamsen and Graham 1998: 38–77; Waskan 2006: 15–35) – adopted
a highly individualist and universalist perspective on language and cognitive pro-
cessing up to the 1990s. By comparison, discourse analysis had always cherished
a fundamentally social and functionalist view of language as a form of social and
cultural practice.1 These orientations had also prevented an inspired dialogue
between these two fields on the level of theory. While – following the metaphor
1
A systematic comparison of the two conceptions is presented in Langlotz (2015a:
83–107).

Authenticated
of the mind as a computer – a fairly mechanistic and internalised view of infor-

mation-processing used to be prominent in cognitivism, meaning-creation in dis-
course analysis was seen as an externalised social process that always transcends
the capacities of an individual. In grossly simplified terms, meaning is fully con-
trolled by the individual cogniser in traditional cognitive science, whereas meaning
in discourse analysis is conceived as a form of social convention and the result of
social-interaction that necessarily transcends individual achievement. Finally, the
marriage of discourse and cognition was also hindered by methodological differ-
ences. While cognitive science had primarily favoured the measurability offered
by experimental methods, discourse analysis had long been more closely allied
with qualitative sociolinguistic and ethnographic methods that place the analyti-
cal control in the hands of the participant-observer or culturally-informed interac-
tional analyst (Harré 2001: 693–695).
In recent years several developments in cognitive science and discourse analy-
sis have encouraged people from both camps to explore the fascinating interfaces
between discourse and cognition more closely. Most importantly, cognitive sci-
ence has changed its basic view of the cognitive processor away from the comput-
er-metaphor to the idea of the embodied cogniser (see, e.g., Pfeifer and Bongard
2007). In accordance with the new role-model in artificial intelligence – the robot –
an embodied cogniser is seen to solve everyday problems in very close physical
and social interaction with the environment. For human cognisers, environments
constitute culturally-constructed, historically-determined, and socially-mediated
environments (Langlotz 2015a: ch. 4.1). As a consequence, the embodied view of
cognition brings human cognitive activity much closer to analytical phenomena
that have been traditionally scrutinised by discourse analysts.
The paradigm shift towards the idea of embodiment (Varela, Thompson and
Rosch 1991; A. Clark 1997, 2008; Noë 2009) is also compatible with ideas from
social constructivism, which claims that the construction of mental representa-
tions and cognitive processing are socially-distributed activities (Vygotsky and
Cole 1978). Following this idea, human cognisers depend on processes of social
“scaffolding” to fully exploit their cognitive capacities when creating meaning
(A. Clark 1997: 45). The socially interdependent nature of human cognition has
also been stressed by representatives of mediated cognition (Hutchins 1995,
2006) and situated cognition (Suchman 1987). These researchers have analysed
the highly complex interactional and technological environments that pattern and
mediate mindful human practices. Instead of experimental methods, these situ-
ated, mediated, and socially-distributed approaches to cognition have embraced
discourse-analytical methods, most importantly conversation analysis and ethnog-
raphy (see Streeck and Mehus 2005). With regard to the pragmatic analysis of
social media discourse, these approaches are very interesting because social media
also constitute situated, mediated, and socially-distributed communicative envi-
ronments. They provide very complex arrangements of communicative tools and

Authenticated
multimodal semiotic resources (pin walls, chatrooms, postings, imagery, emoti-

cons, etc.), which structure socio-communicative processes beyond merely pro-
viding additions to verbal interaction (Eisenlauer 2013).
However, it is not only within the field of cognitive science that researchers
have moved more closely to discourse. In recent years, various linguists – primar-
ily having their academic roots in cognitive linguistics, critical discourse anal-
ysis, and interpersonal pragmatics – have also become more interested in com-
bining the cognitive underpinnings of language and discourse management with
their social-interactional foundations. Based on pioneering work in cognitive and
socio-cognitive linguistics (e.g., Clark 1996; Du Bois 2003; Enfield and Levinson
2006; Langacker 2001, 2008; Tomasello 1999, 2003; Van Dijk 1990, 2006a; Ver-
hagen 2006; Zlatev 2007; Zlatev and Sinha 2008), several researchers have ven-
tured to develop integrative models that combine concepts from cognitive science
and discourse analysis (e.g., Fischer 2010; Langlotz 2010, 2015a, 2015b; Van Dijk
2014; Zima 2013; Zima and Brône 2015). Trying to overcome the research-his-
torical divide between social practices and individual cognition, this socio-cogni-
tive research strand has stressed the inseparability of the social and the cognitive
dimensions of discourse by highlighting the joint and intersubjective nature of
discursive achievements as well as the dynamic, socially-distributed nature of con-
ceptualisation processes.
In the following section, I will specify how these socio-cognitive parameters
can play a decisive role in the management of social media discourse with refer-
ence to Herbert Clark’s theory of language use. The theory is taken as a baseline to
carve out how the discourse-analytical tasks listed in Section 2.1 interact with the
cognitive skills sketched in Section 2.2 in social media communication.
3. The socio-cognitive dimensions of managing social media discourse
The central mystery of communication – both online and offline – resides in the
fact that whenever we communicate with other people, we cannot convey or trans-
port any meaning. Although this “conduit metaphor”, as Reddy (1979: 186) terms
it, captures a common folk conception about how communication works, meaning
simply cannot be encoded into words for it to be unpacked by its recipients (Sper-
ber and Wilson 1995: 3–9). Instead, the task of managing meaning through dis-
course amounts to the communicators’ complex interactional task of coordinating
their mental states or mutual cognitive environments (Sperber and Wilson 1995:
42) in order to create an intersubjective orientation to social, transactional, or eval-
uative states-of-affairs. In his seminal work Using Language (1996) Herbert Clark
has proposed a socio-cognitive theory of communication. His central argument is
that the creation of meaning is achieved by joint actions in which the speaker and
the listener simultaneously play active roles. Therefore, acts of communication

Authenticated
cannot be analysed in simplified terms as mere acts encoding and decoding prede-
fined meanings.
3.1. Managing common ground

For Clark language use amounts to the complex management of joint commu-
nicative actions. These actions have the purpose of establishing and advancing
common ground between the communicative partners (Clark 1996: 92–121). Com-
mon ground is “the sum of [the interactors’] mutual, common, or joint knowledge,
beliefs, and suppositions” (Clark 1996: 93), i.e. the knowledge that both com-
municative partners are supposed to share. Some of this knowledge – communal
common ground – is presupposed by both communicative partners when they enter
their interaction. Other knowledge is built up through the interlocutors’ shared
history of joint engagement. This is termed personal common ground.
Communal common ground describes knowledge that is shared by the mem-
bers of a cultural group or community of practice (Wenger 1998). It is this mutu-
ally shared basis of expertise that defines the set of people as a community. Apart
from such insider knowledge, people may also have public access to outside infor-
mation about other cultural groups (Clark 1996: 101). When posting his message,
Obama obviously assumed a great deal of communal ground between his audience
and himself. For instance, he assumed that the recipients knew that he is the cur-
rent President of the United States (POTUS) in office and that they therefore refers
to the US government as an institution. Moreover, he took for granted that people
could immediately link his first name Barack to his last name Obama.
By contrast, personal common ground comprises interpersonal knowledge
that is derived from the history of shared experiences between individuals. When
tweeting the message, Obama obviously also wished to advance personal common
ground between himself and the readership, namely that he has just got a presiden-
tial Twitter account. This newly established common ground between himself and
his audience must be inferred from the multimodal input that is provided by his
tweet. Thus, cognitively speaking, the construction of common ground is bound to
cognition in the form of internal representations of knowledge. These representa-
tions are retrieved as cognitive frames from long-term memory, which comprise
stable knowledge about language, cultural facts and norms (encyclopedic knowl-
edge), typical procedures (how-to knowledge), or more personal experiences (see
Section 2.2).
In highly simplified terms, communication thus amounts to advancing shared
states of personal common ground by exploiting the presupposed knowledge that
is retrieved via communal common ground. By activating and constructing pieces
of common ground, users can create internal representations of what is going on
in discourse. But how can these processes of activating the frames be managed
communicatively?

Authenticated
3.2. Coordination devices
Since meaning cannot be grasped physically, cognitive knowledge states can only
be coordinated if communicators can rely on solid, perceptually-accessible evi-
dence that both (or more) interactors conceptualise the world of mutual experience
in a similar way. Such communicative evidence establishes an external representa-
tion of what is going on in the given interaction. Clark calls these externalised
pieces of evidence coordination devices: “What coordination devices do is give the
participants a rationale, a basis, for believing they and their partners will converge
on the same joint action” (Clark 1996: 65, italics in the original). Coordination
devices are semiotic tools to channel the participants’ mutual conceptualisation
of their immediate ‘reality’ for establishing common ground (see Langlotz 2015a:
160–166). To infer and construct common ground by means of such coordination
devices, cognisers must proceed incrementally (Clark 1996: 116–120). They must:
1. detect cues that can serve as evidence for common ground;

2. derive common ground by exploiting these semiotic resources;
3. establish conceptual connections by integrating old pieces of common ground
with new ones, i.e. elaborate and fine-tune the joint knowledge bases.
Let us use another Twitter message, Bill Clinton’s reaction to Obama’s tweet, to
illustrate these concepts.
Tweet 2. Bill Clinton’s reaction to Barack Obama’s first Twitter message
With reference to this tweet, one can point to the following coordination devices
with the corresponding pieces of common ground that they can potentially trigger
(Table 2):

Authenticated
Table 2. Coordination devices involved in Clinton’s reaction to Obama’s first tweet
coordination (presumed) shared inference of common type

device basis ground
message design knowledge of this this is a tweet communal
social medium
@billclinton knowledge of discur- these are usernames or communal
@Twitter sive conventions on handles
@POTUS Twitter
#askingforafriend knowledge of discur- this is a hashtag communal
sive conventions on
Twitter
photo encyclopedic knowl- this is Bill Clinton communal
edge about Clinton’s
physical appearance
Bill Clinton name denotes this Clinton is the sender of this communal
particular person message
Welcome to conventional welcom- sender wishes to establish communal
(@Twitter …) ing formula contact to receiver personal
one question: linguistic knowledge, Clinton wishes to ask communal

performative formula Obama a question personal
Does that user- lexical knowledge Clinton signals that he has communal
name [@POTUS] about username understood the meaning of
the username
stay with the encyclopedic knowl- Clinton alludes to the lim- communal
office edge about the maxi- ited duration of Obama’s
mum 8-year period in presidency
presidential office
Does that user- integration of pieces of cheeky allusion that Obama personal
name stay with common ground might lose the account
the office again soon
#askingforafriend linguistic knowledge jocular allusion to the fact personal
that this is a question for a
particular “friend” that both
Clinton and Obama only
know too well – Hillary
Clinton, of course.

Authenticated
Note that the conceptualisations associated with these pieces of common ground
implement all the dimensions of the ideational, interpersonal (including their
affective dimensions), and textual metafunctions of discourse (1) – (3). Clinton
reacts to Obama’s attempt at sharing his new experience with the world of Twitter
and does so in a strikingly jocular and informal manner in order to establish a
strong interpersonal sense of proximity to Obama. His affective stance can thus
presuppose to align him with the President on a hierarchical level of social equal-
ity and friendship.
3.3. Signalling methods employed in meaning coordination

As becomes obvious in Table 2, an array of multimodal coordination devices can
be employed by users of social media even within the scope of a very short tweet.
Clark’s notion of coordination device is compatible with this semiotic character-
istic of social media communication, because it replaces the classical model of
the linguistic sign with the more dynamic notion of signalling (Clark 1996: 155–
189). He proposes that instead of just employing linguistic symbols as vehicles
for encoding and decoding conventional semantic content, a “signal is really the
presentation of a sign by one person to mean something for another” (Clark 1996:
160, italics in the original).
Signalling in communication consists of a mixture of three methods of con-
structing and using coordination devices for joint communicative action: “describ-
ing-as, indicating and demonstrating” (Clark 1996: 188). The three methods of
signalling are related to Peirce’s (1977) tripartite model of the sign in terms of
icons, indexes, and symbols. Icons, such as pictograms, resemble their referents
perceptually. Indexes are physically connected with their referents. For instance,
smoke points to fire. Finally, symbols, such as linguistic symbols or traffic signs,
constitute purely artificial signs which are associated with their referents by con-
vention. Clark (1996: 160) proposes that communicators use the three methods of
signalling to create such signs for their recipients. This can be nicely illustrated
on the basis of the two tweets. When demonstrating a thing to the communicative
partner, the speaker relies on or creates an icon. Thus, Obama and Clinton used a
picture of themselves to demonstrate their identity. Obama also relied on the strat-
egy of indicating; when using the explanation marks in his posting, he employed
an index that directs the readers to his state of excitement. Finally, when formulat-
ing their messages, both interlocutors also relied on linguistic symbols to describe
the new state-of-affairs as well as the question, respectively.
In communicative activities, users combine the different signalling strategies
to produce complex composite signals: “In language use, indicating is usually
combined with describing or demonstrating” (Clark 1996: 168). Both tweets
also feature such complex composite signals. For instance, in Obama’s tweet the
image is combined with the name President Obama and the username or handle

Authenticated
@POTUS. These elements describe, demonstrate, and indicate social information

about their prominent user. The name verbally underlines the personal identity
that is demonstrated by the image, whereas the handle indicates the role of the
user in the form of an acronym for President Of The United States. In the text,
Obama further describes his personal identity by using his first name Barack. By
means of these different methods of signalling social information, he can channel
the construction of his self-presentation by evoking both communal and personal
common ground in his followers’ minds. This obviously worked for Bill Clinton
as is evidenced by his answer.
3.4. Grounding communicative behaviour through joint actions

As the different signalling methods indicate, the discursive management of com-
mon ground does not merely apply to the content of what is said but also to the
complex semiotic actions involved in how the speaker says it. As Brennan states:
“To communicate successfully, two people need to coordinate not only the content
of what they say, but also the process of saying it” (Brennan 1998: 201). When
producing communicative actions, speakers must therefore monitor whether their
communicative partners have established common ground and whether the pro-
cedure of using the specific coordination devices has been successful in doing
so. This socio-communicative process of managing common-ground construction
is called grounding: “people try to ground what they do together. To ground a
thing, […], is to establish it as part of common ground […]” (Clark 1996: 221,
italics in the original).
Grounding works through the interactors’ mutual engagement in joint actions:
“A joint action is one that is carried out by an ensemble of people acting in coordi-
nation with each other. As simple examples think of two people waltzing, paddling
a canoe, playing a piano duet, or making love” (Clark 1996: 3). Clark’s idea of
joint action is fully compatible with the situated and praxeological view of cogni-
tion developed in conversation analysis (Mondada 2006). From this perspective,
the sequential arrangement of interactions into turns as well as the construction of
these turns for the management of the interactors’ joint orientation to the ongoing
conversational activity is the basis for guaranteeing intersubjective understanding.
According to Clark (1996: 148–153), joint communicative actions feature a
very complex participatory structure organised along four interdependent levels
as depicted in Table 3.
Levels 1 and 2 involve the production of an utterance by producing composite
signals with the three different methods of signalling sketched above. The speaker
executes a behaviour to which the listener attends. The behaviour has the purpose
of presenting coordination devices to the listener, which they must try to identify
correctly. With reference to Obama’s tweet this involves the President’s writing
and posting of the message and the followers’ reception of it. These levels of

Authenticated
Table 3. Joint communicative actions
Level Speaker A’s actions Addressee B’s actions

4 A is proposing joint project w to B B is considering A’s proposal of w
3 A is signalling that p for B B is recognizing that p from A
2 A is presenting signal s to B B is identifying signal s from A
1 A is executing behaviour t for B B is attending to behaviour t from A
communicative work thus correspond to the dimensions (Ai), (Aii) and (Bi) in
our overview of cognitive skills and to the textual function (2) involved in dis-
course-management. Note, however, that Clark regards execution/attention and
presentation/identification as socially-distributed joint actions: speaker’s must
capture the listener’s attention when producing an utterance and listeners must
orient to the semiotic activities of the speaker and take on the perspective of the
speaker when interpreting an utterance (see also Langlotz 2015a: 168–169).
Levels 3 and 4 also constitute complex joint actions. Level 3 involves the
speaker’s informative intention and its recognition by the addressee. With his post-
ing Obama primarily wished to communicate that he is now a member of the
Twitter community. For communication this level is the most decisive one because
it involves the listener’s adaptation of his or her mental state for the construction
of common ground (Clark 1996: 139). With regard to the cognitive skills and dis-
cursive tasks, level 3 thus corresponds to the dimensions (Bii) and (Ci), and the
realisation of the ideational (1) and interpersonal functions (3).
Through this process of cognitive adaptation, the addressee is invited to con-
sider a joint communicative project, which is proposed by the speaker (Level 4).
The readers of Obama’s tweet are invited to engage with Obama’s representation of
the new state of affairs and engage with Obama’s message by reacting to it. Cogni-
tively speaking, they can further scrutinise the message (Cii) and take an affective
stance towards it (Ciii). Ideally, these cognitive and affective changes may lead to
taking up the proposed joint project and to the recipient’s writing a response.
But how could Barack Obama assume that his communicative move actually
succeeded in triggering the participatory actions and the corresponding cognitive
activities that are necessary for sharing common ground with his potential audi-
ence? How do we know that his communicative project was sufficiently grounded?
For grounding to become successful, joint actions must obey the principle of clo-
sure:
Principle of closure. Agents performing an action require evidence, sufficient for cur-
rent purposes, that they have succeeded in performing it. (Clark 1996: 222)
Two further complementary principles make it possible for the communicative

partners to obtain explicit evidence on the success of their joint convergence on

Authenticated
common ground: the principle of upward completion and the principle of down-
ward evidence.
Upward completion. In a ladder of actions, it is only possible to complete actions from
the bottom level up through any level in the ladder. (Clark 1996: 147)
Downward evidence. In a ladder of actions, evidence that one level is complete is also
evidence that all levels below it are complete. (Clark 1996: 148)
We have already described how Obama’s tweet incrementally completes the four
levels of speaker’s actions to communicate his informative intention: ‘I wish to
share my excitement about being on Twitter with you!’ However, the President
cannot be certain that his potential followers have actually understood this mean-
ing before he receives any direct positive evidence of their recognition of his inten-
tion and their willingness to take up the joint communicative project (Clark 1996:
222). Such an evidence is reflected in Bill Clinton’s reaction to Obama’s tweet.
With his own tweet, Clinton orients back to Obama’s message by welcom-
ing him to Twitter. By replying to Barack, the former POTUS takes up the joint
project proposed by the current president (Level 4). This provides Obama with
positive evidence that his attempt at establishing personal common ground with
this particular user of Twitter actually succeeded. An addressee’s uptake provides
the most powerful direct evidence of joint closure because it also entails evidence
that all the lower levels of the joint action-ladder have also been completed suc-
cessfully. More concretely, Clinton reacts to two pieces of common ground a)
Obama’s gesture of addressing the Twitter community as a new member, and b)
Obama’s new official username or handle @POTUS. Thus, by reacting to these
two pieces of information, Clinton indicates that he has understood Obama’s move
in a way that was probably intended by him. Following the principle of downward
evidence, Clinton also shows that he successfully attended to the corresponding
signals (Level 1), positively identified them (Level 2), and recognised Barack’s
purpose in producing them (Level 4). In line with the concept of grounding, Clin-
ton’s reply thus signals to Obama explicitly that he has grounded what Obama
wished to achieve.
The recipients themselves have to actively signal their understanding of what
is being said to them. Coordination devices thus play a double role. On the one
hand, they must serve as appropriate tools for triggering the speaker’s message:
I am using these signals to communicate something to you. On the other hand,
the cues must also serve as appropriate vehicles for signalling that the speaker’s
message has been successfully processed by the listener. This double function of
communicative signals creates what Clark (1996: 241) terms “two parallel tracks
of actions” as illustrated in Figure 3.
On track 1 the actual content of what is communicated has to be derived and
negotiated by the speaker and the hearer, whereas on track 2, the speaker and the
hearer have to adopt a meta-communicative stance to control the functionality of

Authenticated
the coordinated actions that are employed to guarantee the joint alignment of their
shared conceptualisations. In this light, it is interesting to note that Clinton does
not only react to Obama’s communicative move, but he also mockingly addresses
the signals that Barack uses for establishing common ground with his audience
on Twitter. His question: Does that username stay with the office? implies a very
cheeky and teasing meta-commentary on Obama’s tweet. On track 2, he reacts to
Obama’s central identificatory signal on Twitter: the handle @POTUS. Indirectly,
this question also makes fun of the President’s own comment (on track 1) on his
rather late joining of Twitter. Since the American presidency can only last a max-
imum period of eight years – a piece of communal common ground that can obvi-
ously be presupposed by Clinton –, it seems evident that Obama will only be able
to use the handle @POTUS for the rather short period of another one-and-a-half
years. Clinton’s seemingly innocent question about one of Obama’s most impor-
tant coordination devices thus also provides good evidence that Clinton actively
observed the signalling strategies that Obama employed for executing and present-
ing his message. Clinton’s question-as-meta-comment shifts the communicative
activity from welcoming to jocular banter between friends. Clinton underlines his
tease by using the hashtag #askingforafriend.
Figure 3. Two tracks for communicative actions (based on Clark 1996: 241)

Authenticated
4. Socio-cognitive affordances and limitations of social media
From the socio-cognitive perspective articulated above, one can now ask how
the specific communicative features of social media afford and constrain (see
Section 2.2.) social and cognitive processes of sense-making online. In several
approaches to discourse – most notably conversation analysis – spoken, face-to-
face talk-in-interaction is conceived as the most natural and primordial site for lin-
guistic and social activity (Drew 2005: 74; Schegloff 1996; see also Bublitz 2013:
29). A similar stance is also adopted by Clark, who developed his socio-cognitive
model of discourse primarily – but by no means exclusively – with a focus on con-
versational settings (Clark 1996: 8–9). With reference to an earlier study by Clark
and Brennan (1991), he defines the discursive parameters of “natural settings” as
follows and groups them into immediacy (features 1–4), medium (features 5–7),
and control (features 8–10) features:
Table 4. Communicative features of face-to-face conversation
1 Copresence The participants share the same physical environment.

2 Visibility The participants can see each other.
3 Audibility The participants can hear each other.
4 Instantaneity The participants perceive each other’s actions at no percep-
tible delay.
5 Evanescence The medium is evanescent – it fades quickly.
6 Recordlessness The participant’s actions leave no record or artifact.
7 Simultaneity The participants can produce and receive at once and simul-
taneously.
8 Extemporaneity The participants formulate and execute their actions extem-
poraneously, in real time.
9 Self-determination The participants determine for themselves what actions to
take when.
10 Self-expression The participants take actions themselves.
As Goodwin’s (2007: 56) interactional research in the field of situated and

embodied cognition shows, interactors actively construct joint attention through
“embodied participation frameworks.” In such frameworks, gestures and physical
behaviours such as bodily stance-taking, postures or pointing are intertwined with
joint sense-making processes. Moreover, these embodied actions are “environ-
mentally coupled” with material resources that are jointly used for problem-solv-
ing (Goodwin 2007: 63). Through these embodied practices interactors can liter-
ally demonstrate their instrumental, epistemic, cooperative, moral, and affective
stances (Goodwin 2007: 70). In other words, they can demonstrate which tools
they actively use for solving a problem, how they use these tools, what their cog-

Authenticated
nitive orientation to the problem at hand is, to what extent they actually cooperate,
as well as what their attitudes and emotions towards the issue at hand is. All of
these complex dimensions are accessible to the interlocutors’ direct observation.
As a result, facets of meaning can directly be seen in the interlocutors’ behaviours.
They do not have to be inferred through internalised and effortful mental processes
(see also A. Clark 1997: 180).
From this perspective of cognition-in-interaction, social media display various
features, which differ strikingly from the embodied arrangements of face-to-face
cooperation. It is well known that social media are often conceived as a “dimin-
ished form of face to face conversation” (Baym 2010: 51). While this negative
evaluation is obviously not shared by the many users of online communication
platforms, it is nevertheless the case that some of the parameters of situated and
embodied cognitive processing through joint action are radically affected by social
media. The parameters from Table 4 can be taken as a point of departure to high-
light the main features of Twitter in comparison (see Table 5):
Table 5. Communicative features of Twitter
1 Virtual presence The participants share the same electronic communication

platform as a communicative environment.
2 Non-visibility The participants cannot see each other.
3 Readability The participants can read each other’s contributions.
4 Non-instantaneity The participants perceive each other’s actions with a (some-
times considerable) temporal delay.
5 Permanence The medium is permanent.
6 Storage and repli- The participant’s actions leave a record – a permanent trace
cability – on the platform. These traces can be referred to or recycled
by other users.
7 Asynchronicity Simultaneity is considerably reduced on Twitter. The reac-
tion to and exchange of messages may occur with a consider-
able time lapse.
8 Distribution and The interlocutors’ actions become spatially distributed
fragmentation throughout the medium. However, actions by specific users
and particular topics can be followed via handles and hash-
tags.
9 Planning The participants can formulate their contributions with delib-
eration and consideration.
10 Partial self-deter- The participants determine for themselves what actions to
mination take when, but they cannot control when and how their com-
municative moves are taken up.
11 Guided self-ex- While the participants take actions themselves, their freedom
pression is constrained by the features of the communication plat-
form.

Authenticated
12 Extended reach The users can address many more users than the people
present in their physical context. But they can only partially
control their readership.
13 Mobility The users of Twitter can reach one another anytime and any-
where.
As can be seen from this sketch, Twitter has new communicative affordances, but
it also limits communicative interactions in striking ways. What consequences
may these differences have on the socio-cognitive dynamics of sense-making
sketched in the Section 3?
4.1. Affordances and constraints on joint actions in social media

The characteristics of Twitter have a direct effect on the management of joint
actions. Due to its non-instantaneous and non-simultaneous nature, the joint char-
acter of the coordinated actions along the four levels of Clark’s action-ladder-
model is considerably reduced. While the communicative partners can mutually
control and monitor the performance of each other’s participatory actions at each
singular level in face-to-face conversations (Richardson, Dale and Kirkham 2007),
they can only orient to the products of their communicative actions online. For
instance, readers of Obama’s Twitter message can only participate in his com-
municative move once they have actually detected and read the message. But
they cannot immediately react to the ways the writer is executing or presenting
it. Moreover, the producer of a message does not have access to any evidence of
whether the recipient has understood his communicative intention and considered
the proposed communicative project before he/she has received a reaction by the
addressee in the form of a communicative response (a reply in the form of another
tweet). Along these lines, the diverse uptake of Obama’s tweet is only evidenced
by the many reactions that the tweet received on the Twitter platform (https://
twitter.com/potus). However, these reactions do not all appear right after Oba-
ma’s move. Rather they are spatially distributed as a long strand of reactions that
also include reactions to these reactions. This fragmentation of the communicative
take-up of a tweet disrupts the sequential arrangement of communicative moves
that is typical of face-to-face interaction (Drew 2005: 89–94). To capture joint
communicative actions on Twitter, Clark’s model must therefore be adapted by
separating the producer’s and the addressee’s participatory actions (see Table 6).
Moreover, due to the extended reach of social media, it is not always clear
to whom a message is actually addressed. As a consequence the dyadic struc-
ture of a face-to-face conversation is broken up (Bublitz 2013: 30–34) and the
participant roles are decomposed (Hoffmann 2012: 63). Obama’s message was
addressed to the entire Twitter-community, which massively increased the range of
its potential readership. (@POTUS currently has more than 6 million followers).

Authenticated
Table 6. Joint communicative actions in Twitter
Level Producer A’s actions Addressee X’s actions

4 A is proposing joint project w X is considering A’s proposal
separation of “joint”
to X of w
Spatio-temporal
3 A is signalling that p for X X is recognizing that p from A
2 A is planning and writing a X is detecting and reading the
message m for X message m from A
actions
1 A is planning a message m for X X is detecting and reading the
message m from A
In the action-ladder model the addressee must therefore be labelled as a potentially

indefinite X (see Table 6).
This different setup of joint communicative actions on Twitter has cognitive
consequences for the producers and receivers of messages. In the absence of any
direct feedback, the producer has to carefully design the tweet to create intersub-
jectivity with the desired audience through the carefully planned use of coordi-
nation devices. This demands considerable cognitive effort. Moreover, due to the
record left on the platform, the producer cannot easily self-repair what he or she
has said. This becomes apparent in Obama’s careful formulation of his tweet as
discussed above.
When anticipating the shared bases of presupposed knowledge in the absence
of an interlocutor, the social-psychological principle of valuing ‘me and mine’
over ‘them and theirs’ works as a very strong driving force (Smith and Mackie
2000: 16–17). Following Clark (1996: 111), communicators derive a feeling of
what others know (feeling of others’ knowing) from what they know themselves
(feeling of knowing). This may lead to a false consensus effect, i.e. an overinter-
pretation of common ground shared by a ‘me’ and a given ‘other’. This effect is
reflected in some reactions to Obama’s post:
Tweets 3 and 4. Unclear reactions to Obama’s post
While these replies indicate an uptake of the President’s tweet, they obviously do
not attempt to provide any evidence to the President that his communicative inten-
tion has been recognised. This lack of control over the recipients’ communicative

Authenticated
cooperation constitutes one of the central socio-cognitive limitations of managing

joint communicative projects via this social medium. While the split-up of joint
socio-cognitive coordination in the action-ladder liberates the recipient and gives
him/her more personal control over his/her engagement with particular messages,
the complementarity of participant-roles in natural conversation (Langlotz 2015a:
169) is potentially – but not necessarily – eroded. This is clearly reflected in the
following post by DOPE Magazine, which deliberately reinterprets the handle
@POTUS as a reference to marihuana:
Tweet 5. Pot for us!
Van Dijk (2006b) stresses the cognitive dimension that underlies the construction
of such shifts of communicative context: “[…] contexts are […] subjective par-
ticipant interpretations, constructions or definitions of […] aspects of the social
environment” (Van Dijk 2006b: 163; italics in the original). The cogniser works
as the mediating link between situational and social norms and their actual effect
on the production of discursive behaviour. The interpretative act of perceiving and
mentally constructing a given situation as being of type X makes it possible for us
to relate to contextual parameters in a purposeful way. Accordingly, context mod-
els are the interlocutors’ mental models of a communicative situation. The catego-
risation of what the situation is about provides a cognitive frame relative to which
discursive choices can be planned: “context models ongoingly control discourse
production and understanding. That is, they are not fixed, but flexible and dynamic
and ongoingly adapt themselves to the situation, to what has been said before, to
changes of plan, and so on” (Van Dijk 2006b: 171).
Van Djik highlights cognitive conceptualisation processes as the actual medium
for controlling “ways of speaking” (Van Dijk 2006b: 171). In DOPE Magazine’s
uptake of Obama’s handle @POTUS we can see a radical shift in the contextual
meaning of the coordination device. The meaning of @POTUS shifts from the
concept of ‘presidency’ to the idea of ‘drug consumption’. This recontextualisation
also channels a shift of speech activity from ‘presidential greeting’ to ‘Obama
campaigning for marihuana’. Due to the highly distributed and fragmented nature
of Twitter, producers of messages can hardly repair such adversive recontextual-
isations of their communicative intentions. This shows how the construction of
social-psychological categories such as the ‘image of the President of the United
States’ become socially-distributed achievements for which the producer’s dis-
cursive control over the entire sense-making process is strikingly reduced. From

Authenticated
a socio-cognitive perspective, this may explain the feelings of social anxiety and
the experiences socioemotional stress that are reported in psychological studies on
social media (see Section 1).
4.2. Constraints and affordances on signalling through Twitter

In his critical discussion of the limiting factors of the social networking site Face-
book (Fb), Eisenlauer (2013: 99) shows how: “the specific functional properties of
the Fb platform condition the semiotic choices of a Fb participant”. While afford-
ing the upload of images, videos, and other iconic tools to demonstrate – following
Clark’s terminology – personal experiences and opinions, the technical environ-
ment further causes people to report a feeling of loss of control “over the textual
self-presentation” (Eisenlauer 2013: 201). As Bublitz (2013: 40–41) argues with
reference to Eisenlauer, the technological templates offered by social media pre-
mediate the creation of composite signals and therefore strongly intervene with the
users’ freedom of self-expression as if the media acted as ‘third authors’.
This critical stance towards constrained signalling is aligned with the imper-
sonal view of online communication that was already developed in the early
phases of CMC. Due to the absence of important social signalling methods like
nonverbal gestures, facial expressions, voice and accent, the semiotic affordances
for generating a sense of an authentic and personal social presence are rather poor
in cyberspace. This cues-filtered-out-approach was introduced in the 1980s (Cul-
nan and Markus 1987) with reference to the social psychological social presence
theory (Short et al. 1976) and the social context cues theory (Sproull and Kiesler
1976). From a socio-cognitive perspective, the cues-filtered-out-approach implies
that recipients cannot find enough bases for coordinating social information with
their virtual interlocutors. Hence, the limited number of cues prevents them from
activating frames for social and normative knowledge that would allow them to
sufficiently conceptualise their social presence as well as the social context of the
given communicative activity. As Wood and Smith (2005: 82) state “[…] the short
supply of social context cues can create perceptions of impersonal replies and
impersonal interpretations of messages”.
The same ideas can also be applied to Twitter. A tweet consists of a limit of
140 characters and contains options for uploading photos or embedding videos.
Features like handles and hashtags are also characteristic of this type of electronic
message. With regard to signalling, tweeters can thus potentially exploit all the
signalling strategies outlined in Section 3.3. However, their freedom of semiotic
activity only operates within the systemic limits. This has caused some users to
give expert advice on how tweets should be optimally constructed as small-text
types (http://www.socialmediatoday.com/content/structure-perfect-tweet).
The active search for optimal signalling strategies on Twitter is compatible
with the social identification/deindividuation (SIDE) model developed by Postmes,

Authenticated
Spears and Lea (1998). This model suggests that people may overcome the restric-
tions of computer-mediated signalling. To compensate for the lack of social con-
text cues users may create a new sense of community and norm by developing
novel, socially-accepted behavioural patterns. Interestingly, Twitter is also termed
Twitterverse by its users (see: https://twitter.com/hashtag/twitterverse). The very
existence of this name points to the communal common ground shared by Twitter
users. For them this platform constitutes a discursive environment with its very
own semiotic characteristics and discursive norms. These patterns are based on
conventionalised, but platform-specific interactional styles (Wood and Smith
2005: 82). From a socio-cognitive perspective, the SIDE-model therefore implies
that Twitter users must develop specialised frames for semiotically managing the
activity type (Levinson 1992: 69) called ‘tweeting’. Hence, interpersonal engage-
ment is of great importance in the SIDE model because the communication of
common ground is essential to sustain relational ties. For people not sharing these
group-specific communicative norms this may have negative social and cognitive
consequences. On the one hand, they may be excluded from the community, on
the other hand, the community may actively prevent them from developing enough
communal common ground to become an accepted member of the platform. Such
norm-imposing, excluding moves can also be seen in some adversive reactions to
Obama’s tweet:
Tweets 6 and 7. Rejection of Obama’s tweets due to group-internal norms
Here Obama’s act of signalling his presence on Twitter is explicitly contested.

Obama’s suggested concept of ‘US president as Twitter friend’ may clearly run
counter to the social identification of some users with the Twitter community.
Thus, instead of incorporating the President through a platform-specific process of
deindividuation, the commentators in tweets 6 and 7 clearly individuate the Pres-
ident’s posting to harshly demarcate it as running counter to the signalling norms
of the community.

Authenticated
4.3. Social and cognitive consequences of grounding via Twitter
Due to the extended reach of social media, their non-instantaneous nature as well as
the readability, permanence and replicability of the recorded communicative traces,
the producer of an online message cannot fully control the construction of common
ground. But the restrictions for grounding on Twitter and other social media may
also lead to more positive socio-cognitive effects. Such effects are addressed by the
concept of hyperpersonal communication as introduced by Walther (1996; see also
Walther 2007). Running counter to the cues-filtered-out theory, this view claims
that due to the better control and the arrangement of signals online, users may over-
come the social pressures of co-presence and feel more liberated when it comes to
presenting themselves. This effect also becomes visible in Obama’s tweet which
is considerably less formal than other addresses issued by the President. It can be
assumed that the discursive environment of Twitter with its flat social hierarchies
gave Obama the liberty of formulating his message rather informally. This more
controlled and deliberate self-presentation may further cause recipients to exagger-
ate the positive qualities of their communicative partners and thus trigger hyper-
positive feedback. Again, the same effect can be seen in reactions to Obama whose
informal welcome to the Twitter community led to equally informal responses that
would seem rather unlikely in a face-to-face encounter with the President.
Tweet 8. Love eyes
The love-eyed smiley in this tweet clearly goes beyond the level of intimacy that
may have been desired by Obama in his self-presentation to the community. As
this and the other examples discussed suggest, hyperpersonal attempts at estab-
lishing common ground online have direct consequences on interpersonal concep-
tualisation processes.
The construction of common ground on Twitter has cognitive as well as social
repercussions. Since producers have to presuppose communal common ground
between themselves and their audience when writing a message, they naturally
imply a certain degree of familiarity or distance to their addressees through how
they formulate their message – “style is the relationship” as Sperber and Wilson
(1995: 217) put it. Tweeters expect that the used coordination devices are sufficient
for the recipient to activate the necessary knowledge chunks from long-term mem-
ory and to recognise the content and purpose of the message. According to Enfield
(2006: 422): “the practices by which we manage and exploit common ground in
interaction demonstrate a particular commitment to particular relationships and
particular communities”. This is apparent in Obama’s tweet, which expects a con-

Authenticated
siderable amount of knowledge about him as a person and thus suggests interper-
sonal familiarity. Thus, the more communal and personal common ground two
interactors share, the more likely their practices are to converge around the con-
struction of common ground. This has become most obvious in Bill Clinton’s reac-
tion to Obama’s tweet. He advanced personal common ground between himself
and Obama by means of his cheeky tease. In teasing Obama, he demonstrated a
very high degree of personal common ground and intimacy.
Barack Obama himself replied to Clinton in the following interesting manner,
which can be regarded as a form of creative hyperpersonal identity construction
(Langlotz 2015b):
Tweet 9. Obama’s reply to Clinton
Obama creatively exploits the semiotic affordances and peculiarities of the

medium to strike back against Clinton’s tease. He ambiguates the word handle to
mean both ‘username’ and ‘door handle’. So when Obama leaves the White House,
he will also have to abandon the Twitter handle. But there is another Twitter han-
dle: @FLOTUS, which is used by the First Lady Of The United States. It is well
known that Bill Clinton’s wife, Hillary, had already been campaigning for the US
presidency when Obama and Clinton posted their tweets. Obama’s question about
@FLOTUS is thus directly, but implicitly, addressed at Bill Clinton, who could
indeed become the future First ‘Lady’. Obama’s allusion is funny and mean at the
same time. He makes fun of Clinton by engaging in a little transgender mockery.
He symbolically emasculates the former president and thus trumps (Veale et al.
2006: 312) Clinton’s previous tease. Moreover, he may make fun of the fact that
the US constituency does not have a name for the male partner of a female pres-
ident because Hillary would indeed be the first female leader of America. Thus,
Obama communicates both on Track 1 and on Track 2 of communicative signal-
ling. His allusion to @FLOTUS is both a meta-commentary on the specific Twit-
ter-handle and a cheeky remark to Clinton becoming the First Lady. Hence, this
example demonstrates that close convergence and social creativity are possible
even within the restricted semiotic space offered by a tweet. This construction,

Authenticated
however, demands the activation of a great deal of presupposed common ground

and the clever exploitation of coordination devices that are explicitly designed for
the take-up by a specific addressee. Moreover, the sequentiality of the tweets must
be mutually traced by the interactors and controlled by both of them.
5. Conclusion
Social media, discourse and cognition co-determine one another through tripartite
reciprocal adaptation (see Figure 1). Users must be able to adapt their socio-cog-
nitive skills and processes to the affordances and limitations of the social medium.
The medium is itself designed for managing specific discursive tasks and for dif-
ferent types of information. This design is then exploited to the extent that this is
afforded by the users’ socio-cognitive skills and dispositions. Most importantly,
they must recruit chunks of conceptual knowledge from their long-term memo-
ries for the construction of common ground with their communicative partners.
This knowledge involves both the content of messages as well as the processes of
jointly managing the sense-making process.
The chapter has surveyed some of the socio-cognitive opportunities and chal-
lenges that emerge for Twitter-users. While Twitter certainly enables the joint
construction of shared situated conceptualisations, the brevity of tweets and the
absence of bodily signalling methods, the spatio-temporal separation as well as the
fragmentation of the sequentiality of joint actions undermine the embodied nature
of face-to-face interaction. This communicative disembodiment is also accom-
panied with a cognitive disembodiment. Instead of orienting to their interlocu-
tors’ subtle actions of jointly coordinating common ground, Twitter-users must
press out meaning from the semiotic space that is afforded by the medium. The
reduction of socio-cognitive coordination puts more load on internalised and indi-
vidual sense-making processes because a great deal of communal and personal
common ground has to be activated and mentally constructed to make sense of
tweets. Internalised processing is likely to be increased because the control over
the joint-meaning construction is lowered both for the producers and the recipients
of tweets. This higher degree of socio-cognitive disembodiment may explain the
positive and the negative cognitive and socio-emotional effects reported in psy-
chological studies: the increase of working memory and spelling abilities, infor-
mation overload, and a reduced feeling of socio-emotional well-being that may be
accompanied by social anxiety.
This said, very little is actually known about the socio-cognitive dynamics
of social-media-communication. This is primarily due to the fact that embodied,
socio-cognitive studies of discourse processing have only emerged recently. In the
future, the potential of socio-cognitive linguistics has yet to be fully developed
both with regard to offline and online communication.

Authenticated
References
Ahn, June
2011 The effect of social network sites on adolescents’ academic and social devel-
opment: Current theories and controversies. Journal of the American Society
for Information Science and Technology 62(8): 1435–1445.
Alloway, Tracy P. and Ross G. Alloway
2012 The impact of engagement with social networking sites (SNSs) on cognitive
skills. Computers and Human Behavior 28: 1748–1754.
Alloway, Tracy P., John Horton, Ross G. Alloway and Clare Dawson
2013 The impact of technology and social networking on working memory. Com-
puters and Education 63: 10–16.
Barsalou, Lawrence W.
2005 Situated conceptualization. In: Henri Cohen and Claire Lefebvre (eds.), Hand-
book of Categorization in Cognitive Science, 619–650. Amsterdam: Elsevier.
Baym, Nancy K.
1995 The emergence of community in computer-mediated communication. In: Ste-
ven Jones (ed.), CyberSociety, 138–163. Newbury Park, CA: Sage.
Bechtel, William, Adele Abrahamsen and George Graham
1998 The life of cognitive science. In: William Bechtel and George Graham (eds.),
A Companion to Cognitive Science, 1–104. Oxford: Blackwell.
Becker, Mark W., Reem Alzahabi and Christopher J. Hopwood
2013 Media multitasking is associated with symptoms of depression and social anx-
iety. Cyberpsychology, Behavior, and Social Networking 16(2): 132–135.
Bless, Herbert, Klaus Fiedler and Fritz Strack
2004 Social Cognition: How Individuals Construct Social Reality. New York: Psy-
chology Press.
Brennan, Susan E.
1998 The grounding problem in conversations with and through computers. In:
Susan R. Fussell and Roger J. Kreuz (eds.), Social and Cognitive Psycho-
logical Approaches to Interpersonal Communication, 201–225. Hillsdale, NJ:
Lawrence Erlbaum.
Bublitz, Wolfram
2013 Der duale Internetnutzer: Ansätze einer dissoziativen Kommunikation. In:
Konstanze Marx and Monika Schwarz-Friesel (eds.), Sprache und Kommunika
tion im technischen Zeitalter: Wieviel Internet (v)erträgt unsere Gesellschaft?,
Cienki, Alan
2010 Frames, idealized cognitive models, and domains. In: Dirk Geeraerts and Her-
bert Cuyckens (eds.), The Oxford Handbook of Cognitive Linguistics, 170–
187. Oxford: Oxford University Press.
Clark, Andy
1997 Being There: Putting Brain, Body, and World Together Again. Cambridge,
MA: MIT Press.
Clark, Andy
2008 Supersizing the Mind: Embodiment, Action, and Cognitive Extension, Oxford:
Oxford University Press.

Authenticated
Clark, Herbert H.
Clark, Herbert H. and Susan E. Brennan
1991 Grounding in communication. In: Lauren B. Resnick, John M. Levine and
Stephanie D. Teasley (eds.), Perspectives on Socially Shared Cognition, 127–
149. Washington: American Psychological Association.
Croft, William and David A. Cruse
2004 Cognitive Linguistics. Cambridge: Cambridge University Press.
Culnan, Mary J. and M. Lynne Markus
1987 Information technologies. In: Frederic M. Jablin, Linda L. Putnam, Karlene
H. Roberts and Lyman W. Porter (eds.), Handbook of Organizational Com-
munication: An Interdisciplinary Perspective, 420–443. Thousand Oaks, CA:
Sage Publications.
Drew, Paul
2005 Conversation analysis. In: Kristine L. Fitch and Robert E. Sanders (eds.),
Handbook of Language and Social Interaction, 71–102. Mahwah, NJ: Law-
rence Erlbaum.
Du Bois, John W.
2003 Discourse and grammar. In: Michael Tomasello (ed.), The New Psychology
of Language: Cognitive and Functional Approaches to Language Structure,
Vol. 2, 47–87. London: Erlbaum.
Du Bois, John W.
2014 Towards a dialogic syntax. Cognitive Linguistics 25 (3): 359–410.
Edwards, Derek
2005 Discursive psychology. In: Kristine L. Fitch and Robert E. Sanders (eds.),
Handbook of Language and Social Interaction, 257–273. Mahwah, NJ: Law-
rence Erlbaum.
Eisenlauer, Volker
2013 A Critical Hypertext Analysis of Social Media: The True Colours of Facebook.
Enfield, Nick J.
2006 Social consequences of common ground. In: Nick J. Enfield and Stephen
C. Levinson (eds.), Roots of Human Sociality: Culture, Cognition and Interac-
tion, 399–430. Oxford: Berg.
Fauconnier, Gilles and Mark Turner
2002 The Way We Think. Conceptual Blending and the Mind’s Hidden Complexities.
New York: Basic Books.
Fillmore, Charles J.
2006 [1982] Frame semantics. In: Dirk Geeraerts (ed.), Cognitive Linguistics: Basic
Readings, 373–400. Berlin/Boston: de Gruyter.
Fischer, Kerstin
2010 Beyond the sentence: Constructions, frames and spoken interaction. Construc-
tions and Frames 2(2): 185–207.
Gibson, James J.
1979 The Ecological Approach to Visual Perception. Boston: Houghton-Mifflin.
Goodwin, Charles
2007 Participation, stance, and affect in the organization of activities. Discourse and
Society 18(1): 53–73.

Authenticated
Graesser, Arthur C. and Carol Forsyth

2013 Discourse comprehension. In: Daniel Reisberg (ed.), Oxford Handbook of
Cognitive Psychology, 475–491. Oxford: Oxford University Press
Halliday, Michael A.K. and Christian Matthiessen
2004 Introduction to Functional Grammar, 3rd ed. London: Edward Arnold.
Harré, Rom
2001 The discursive turn in social psychology. In: Deborah Schiffrin, Deborah
Tannen and Heidi E. Hamilton (eds.), The Handbook of Discourse Analysis,
688–707. Oxford: Blackwell.
Hutchins, Edwin
1995 Cognition in the Wild. Cambridge, MA: MIT Press.
Hutchins, Edwin
2006 The distributed cognition perspective on human interaction. In: Nick J. Enfield
and Stephen C. Levinson (eds.), Roots of Human Sociality: Culture, Cognition
and Interaction, 375–398. Oxford: Berg.
Jaworski, Adam and Nikolas Coupland
1999 Introduction: perspectives on discourse analysis. In: Adam Jaworski and Niko-
las Coupland (eds.), The Discourse Reader, 1–44. London: Routledge.
Junco, Reynol and Sheila R. Cotten
2012 No A 4 U: The relationship between multitasking and academic performance.
Computers and Education 59(2): 505–514.
Kirschner, Paul A. and Aryn C. Karpinski
2010 Facebook® and academic performance. Journal of Computer in Human
Behavior 26 (6): 1237–1245.
Kress, Gunther
2011 What is mode? In: Carey Jewitt (ed.), Routledge Handbook of Multimodal
Analysis, 54–67. London/New York: Routledge.
Langacker, Ronald W.
2001 Discourse in cognitive grammar. Cognitive Linguistics 12(2): 143–188.
Langacker, Ronald W.
2008 Cognitive Grammar. A Basic Introduction. Oxford: Oxford University Press.
Langlotz, Andreas
2010 Social cognition. In: Miriam A. Locher and Sage L. Graham (eds.), Interper-
sonal Pragmatics. Handbook of Pragmatics Vol. 6, 167–204. Berlin: de Gruyter.
Langlotz, Andreas
2013 Yo, who be the main gansta in our phat gang? – Linguistic creativity and
the construction of hyperpersonal identity. In: Tony Veale, Kurt Feyaerts and
Charles Forceville (eds.), Creativity and the Agile Mind: A Multi-Disciplinary
Study of a Multi-Faceted Phenomenon, 159–179. Berlin/Boston: de Gruyter.
Langlotz, Andreas
2015a Creating Social Orientation Through Language: A Socio-Cognitive Theory of
Situated Social Meaning (Converging Evidence in Language and Communica-
tion Research 17). Amsterdam/Philadelphia: Benjamins.

Authenticated
Langlotz, Andreas
2015b Local meaning-negotiation, activity types, and the current-discourse-space
model. Language and Cognition 7: 515–545.
Langlotz, Andreas and Miriam A. Locher
2012 Ways of communicating emotional stance in online disagreements. Journal of
Pragmatics 44(12): 1591–1606.
Levinson, Stephen C.
1992 Activity types and language. In: Paul Drew and John Heritage (eds.), Talk
at Work. Interaction in Institutional Settings, 66–100. Cambridge: Cambridge
University Press.
Merriam Webster Dictionary
2015 http://www.merriam-webster.com/dictionary/, last accessed: 11th April 2016.
Meshi, Dar, Diana I. Tamir and Hauke R. Heekeren
2015 The emerging neuroscience of social media. Trends in Cognitive Sciences 19
(12): 771–782.
Mondada, Lorenza
2006 Participants’ online analysis and multi-modal practices: projecting the end of
the turn and the closing of the sequence. Discourse Studies 8(1): 117–129.
Noë, Alva
2009 Out of Our Heads: Why You Are Not Your Brain, and Other Lessons from the
Biology of Consciousness. New York: Hill and Wang.
Roy Pea, Clifford Nass, Lyn Meheula, Marcus Rance, Aman Kumar, Holden Bamford,
Matthew Nass, Aneesh Simha, Benjamin Stillerman, Steven Yang and Michael
Zhou
2012 Media use, face-to-face communication, media multitasking, and social well-be-
ing among 8- to 12-year-old girls. Developmental Psychology 48 (2): 327–336.
Peirce, Charles S.
1977 Semiotics and Significs. Ed. by Charles Hardwick. Bloomington IN: Indiana
University Press.
Pfeifer, Rolf and Josh Bongard
2007 How the Body Shapes the Mind: A New View of Intelligence. Cambridge, MA:
MIT Press.
Postmes, Tom, Russell Spears and Martin Lea
1998 Breaching or building social boundaries? Side-effects of CMC. Communica-
tion Research 25: 689–716.
Reddy, Michael J.
1979 The conduit metaphor: A case of frame conflict in our language about lan-
guage. In: Andrew Ortony (ed.), Metaphor and Thought, 284–310. Cambridge:
Cambridge University Press.
Richardson Daniel C., Rick Dale and Natasha Z. Kirkham
2007 The art of conversation is coordination: Common ground and the coupling of
eye movements during dialogue. Psychological Science 18: 407–413.
Schegloff, Emanuel A
1996 Issues of relevance for discourse analysis: Contingency in action, interaction,
and co-participant context. In: Eduard H. Hovy and Donia R. Scott (eds.),
Computational and Conversational Discourse: Burning Issues: An Interdisci-
plinary Account, 3–35. New York: Springer.

Authenticated
Schiffrin, Deborah
2006 Discourse. In: Jeffrey Connor-Linton and Ralph Fasold (eds.), Introduction to
Linguistics, 169–203. Cambridge: Cambridge University Press.
Short, John, Ederyn Williams and Bruce Christie
1976 The Social Psychology of Telecommunications. London: John Wiley and Sons.
Sperber, Dan and Deirdre Wilson
1995 Relevance: Communication and Cognition. Oxford: Blackwell Publishers.
Sproull, Lee and Sara Kiesler
1986 Reducing social context cues: Electronic mail in organizational communica-
tion. Management Science 32(11): 1492–1512.
Streeck, Jürgen and Siri Mehus
2005 Microethnography: The study of practices. In: Kristine L. Fitch and Robert
E. Sanders (eds.), Handbook of Language and Social Interaction, 381–404.
Mahwah: Lawrence Erlbaum.
Suchman, Lucy
1987 Plans and Situated Action. Cambridge: Cambridge University Press.
Tomasello, Michael
1999 The Cultural Origins of Human Cognition. Cambridge, MA: Harvard Univer-
sity Press.
Tomasello, Michael
2003 Constructing a Language: A Usage-Based Theory of Language Acquisition.
Cambridge, MA: Harvard University Press.
Van Dijk, Teun A.
1990 Social cognition and discourse. In: Howard Giles and W. Peter Robinson
(eds.), Handbook of Language and Social Psychology, 163–183. Oxford: John
Wiley and Sons.
Van Dijk, Teun A.
2006a Discourse, Interaction and Cognition. Special issue of Discourse Studies
8(1).
Van Dijk, Teun A.
2006b Discourse, context, and cognition. Discourse Studies 8(1): 159–177.
Van Dijk, Teun A.
2009 Society and Discourse. Cambridge: Cambridge University Press.
Van Dijk, Teun A.
2014 Discourse and Knowledge: A Sociocognitive Approach. Cambridge: Cam-
Varela, Francisco J., Evan Thompson and Eleonor Rosch
1991 The Embodied Mind: Cognitive Science and Human Experience. Cambridge,
MA: MIT Press.
Veale, Tony, Kurt Feyaerts and Geert Brône
2006 The cognitive mechanisms of adversarial humor. Humor 19(3): 305–338.
Veen, Wim and Ben Vrakking
2006 Homo Zappiens: Growing Up in a Digital Age. London: Continuum.
Verhagen, Arie
2005 Constructions of Intersubjectivity: Discourse, Syntax, and Cognition. Oxford:
Vygotsky, Lev and Michael Cole
1978 Mind in Society. Cambridge, MA: Harvard University Press.

Authenticated
Walther, Joseph B.
1996 CMC: Impersonal, interpersonal, and hyperpersonal interaction. Communica-
tion Research 19: 52–90.
Walther, Joseph B.
2007 Selective self-presentation in computer-mediated communication: Hyper-
personal dimensions of technology, language, and cognition. Computers in
Human Behavior 23(5): 2538–2557.
Waskan, Jonathan A.
2006 Models and Cognition: Prediction and Explanation in Everyday Life and in
Science. Cambridge, MA: MIT Press.
Wenger, Etiene
Wood, Andrew F. and Matthew J. Smith
2005 Online Communication: Linking Technology, Identity, and Culture. Mahwah,
NJ: Lawrence Erlbaum.
Ziem, Alexander
2014 Frames of Understanding in Text and Discourse: Theoretical Foundations and
Descriptive Applications. Amsterdam/Philadelphia: Benjamins.
Zima, Elisabeth
2013 Kognition in der Interaktion: eine kognitiv-linguistische und gesprächsanaly-
tische Studie dialogischer Resonanz in österreichischen Parlamentsdebatten.
Heidelberg: Universitätsverlag Winter.
Zima, Elisabeth and Geert Brône
2015 Cognitive linguistics and interactional discourse. Time to enter into dialogue.
Language and Cognition 7(4): 485–498.
Zlatev, Jordan
2007 Intersubjectivity, mimetic schemas and the emergence of language. Intellec-
tica: 123–152.
Zlatev, Jordan and Chris Sinha (eds.)
2008 The Shared Mind: Perspectives on Intersubjectivity. Amsterdam/Philadelphia:
Benjamins.

Authenticated
14. Discourse and ideology
Stephen Pihlaja and Andreas Musolff
Abstract: This chapter focuses on the manifestation of ideology in social media

interaction, both in explicit contestation and in implicit frameworks. Following
a discussion of the key concepts of ideology and discourse, several exemplary
studies are presented that show how ideology can be investigated in different ways
and on different scales, from looking at blogs and forums, where ideologies are the
explicit topics of research, to trolling on Twitter, where ideology is enacted in the
way users respond to specific users and topics. Finally, we discuss key methodo-
logical issues and controversies in the study of ideology in social media.
1. Introduction
Social media is diverse. What you present about yourself differs drastically
depending on the site or platform. You’re unlikely to use YouTube in the same way
that you use Facebook because the different sites serve different purposes for you,
and both have specific elements which are suited for different kinds of interaction.
When we speak of ideology in social media, we might think of political memes
shared on Facebook walls or offensive, racist tweets. There are, of course, impor-
tant examples of such ideologies being displayed online, but there are also more
subtle instances of ideologies that impact on social media discourse. If you want
only your friends to see a video of your child learning their first words, you might
put that video on Facebook and protect the post so that only your friends can see it.
Alternatively, if you think the video has the chance of going viral, thus providing
you with minor Internet celebrity status, you might post the video on YouTube. In
either of the choices you make, there are ideological issues to be examined. The
choices you make will likely say something about the way you see the world and
how you think it should be.
Social media and online contexts also provide the chance for people to meet
and engage with like-minded individuals that might be absent in their close phys-
ical proximity, such as atheists in America who might not know any other ‘out’
or self-described atheists in their day-to-day life (Cimino and Smith 2014). This
ability to meet and connect with others can lead to movements that challenge dom-
inant ideological positions, as users interact in spaces that allow for more diversity.
At the same time, as spaces for dissent grow and become influential, voices from
dominant ideologies (both in the positions of individual users and corporations)
can begin to assert influence and create conflict.
DOI 10.1515/9783110431070-014
Authenticated
382 Stephen Pihlaja and Andreas Musolff
This chapter focuses on how issues of ideology are present in social media
interaction, both when they are explicitly contested and in the implicit ideolog-
ical frameworks that the platforms are built on. The aim is to highlight several
central issues in ideology and social media. We will look at the role of complex
performance of ideologies, how ideologies are voiced online, how users come to
consensus, the role of fluid identities in online places, and the effects of constant
interaction on the development of ideologies. The chapter will offer some impor-
tant framing questions and guidance for how to think about the role of ideology
in the construction and use of social media platforms. The first section introduces
the key concepts of ideology and discourse, followed by a sketch of the historical
background to research on social media and language. We then describe several
exemplary studies of ideology online, showing how ideology can be investigated
in different ways and on different scales, from looking at blogs and forums, where
ideologies are the explicit topics of research, to trolling on Twitter, where ideology
is enacted in the way users respond to specific users and topics. Finally, we intro-
duce some key methodological issues in the study of ideology in social media and
discuss some key critical issues and controversies.
2. Ideology and discourse
The concept of ideology remains contested, oscillating between pre- and post-Marx-
ist definitions, which usually focus on unequal power-relationships, and reformu-
lations such as the basis of any kind of social “representations shared among a
group” (van Dijk 1998: 8; see also Eagleton 2007; Fairclough 1989; Laclau and
Mouffe 1989) and the sub-concept of ideologies of language use, i.e., norms and
presuppositions about how to communicate ‘properly’ (Schieffelin, Woolard and
Kroskrity 1998; Verschueren 2012). The latter may not even be recognised by
their users as ‘ideological’ but reveal their socio-political force in conflict, when
they are explicitly contested and invested by their users with normatively loaded
interpretations. Ideology in this perspective can be defined as a set of evaluative
beliefs and attitudes regarding socially relevant topics, including language itself,
which is constructed in the process of discourse as the social production of mean-
ing (Halliday 1978). The emphasis is then on the discursive production of ideology
rather than on its cognitive (or “socio-cognitive”) aspect as highlighted by van
Dijk (1998, 2016), which seems to presuppose pre-established shared concepts
such as “capitalism” and proceeds to analyze and categorize their discursive mani-
festations, for example, “polarization” and “positive self-descriptions and negative
other-descriptions” (van Dijk 2016: 67–77). The seeming objectivity of such an
approach is undercut by the pre-conceived identification of “dominant” and “dom-
inated groups”, which reinstates by the backdoor a vaguely “emancipatory” bias,
in which polemical discourse strategies such as “othering” or de-legitimization of

Authenticated
Discourse and ideology 383
an out-group are criticized if practiced by “dominant” groups and legitimized as

“resistant” when practiced by “dominated” ones (van Dijk 2016: 77–82). Instead
we take the view that power-relationships implicit in ideologies are in fact estab-
lished and renegotiated in and through discourse and that all discourse aspects that
serve to convey interactional, socio-politically and ethically evaluative messages
can be characterized as ideological. The ‘sharing’ of socially evaluative beliefs and
attitudes is itself the product, not the pre-condition of ideology in discourse.
In social media, ideologically loaded discourse can most easily be found in
explicitly politicised text types, such as discussion forums and blogs, but also
manifests itself in the meta-communicative commenting of each other’s contribu-
tions in the community of users. One oft-commented example is the correction of
other participants’ spelling mistakes, which establishes interactional relationships
of power and authority among each other (see Crystal 2011: 61–63; Harris and
Hiltunen 2014; Markman 2010). Comments can also express stances on politeness
rules (‘netiquette’), rules of addressing and quoting, dialect vs. standard varieties,
multilingualism and many more issues. Insofar as such language-focused norms
are highlighted and debated as conceptual-evaluative systems in their own right,
they constitute ‘language ideologies’ that can also impact strongly on debates
about other ideological contents (Verschueren 2012: 21–50). The public debate on
immigration conducted on the BBC’s popular Have Your Say website, for instance,
provides a forum not just for vigorous argumentation about the effect of mass
immigration into Britain but also gives an insight into one of the rarely explicitly
spelt-out negative assumptions about multilingualism, such as the contention that
one’s home culture is threatened by any evidence of other languages being used ‘in
our streets’ or by learning ‘more than one language’ or the public purse ‘wasting
money’ on translation/interpreting services (BBC 2010a-c). The analysis of such
computer-mediated communication (CMC) data helps to uncover folk-linguistic
attitudes and trends that are hidden from or even officially denied in the ‘main-
stream’ public discussion of contentious topics such as migration because they are
deemed taboo or at least unfashionable.
Computer-based social media has changed fundamentally the way the estab-
lished print media and electronic media function. It has become inconceivable for
any press organ or broadcaster not to have their own Facebook and Twitter accounts
as well as forums for interactive user-participation, which can also be cross-posted
on other forums or blogs (Berker et al. 2005). Some forms of highly topical and
contentious news reporting, e.g. crisis news, have been directly affected by the
need to include emerging and ongoing social media coverage as part of the news
event itself and consumers’ reactions to it (Belair-Gagnon 2015). This ‘feedback
loop’ between news reporting commenting and re-commenting has forced but also
re-empowered powerful media institutions such as the BBC to reformulate their
ideological stance as the ‘normal’ consensus of the public. In a sense, the so-called
‘traditional’ media and their ideologies have become partly ‘socialised’ themselves.

Authenticated
3. Social Media
Among other things, social media is about producing and consuming content and
therefore has consequences for the concept of ideology as continuous, dynamic,
and variable emergence in discourse patterns. Whether on Facebook, Twitter or
YouTube, users engage videos, status updates, photos in a way that promotes ide-
ologies they align with and those with which they disagree.1 The obvious examples
come to mind quickly: the obnoxious politician spewing racist rhetoric in 140
characters on Twitter or an ‘inspirational’ message on Facebook that assumes a
particular religious ideology. These examples are certainly interesting in terms of
the ideologies they embody, but we should also think beyond the explicit displays
of ideology, and consider how all interaction is influenced by and evidences users’
ideological positions.
Before investigating ideology as it relates to interaction on social media, we
need to discuss how the media platforms themselves are created from particular
ideological positions and have, embedded in them, certain ideological assump-
tions. Marwick (2013) shows how radical ideologies present in early Web 2.0 cul-
ture became in many ways subsumed in the rise of social media. While originally
driven by principles of community and sharing, social media has become more
about the commodification of social interaction online, where ‘sharing’ becomes a
means for apps and platforms to benefit from the expression of their users.
Mark Zuckerberg, the founder of Facebook, describes this commodification of
interaction in an interview with Wired magazine:
Think about what people are doing on Facebook today. They’re keeping up with their
friends and family, but they’re also building an image and identity for themselves,
which in a sense is their brand. They’re connecting with the audience that they want to
connect to. It’s almost a disadvantage if you’re not on it now. (Vogelstein 2009)
Zuckerberg’s use of the language of commodity is instructive as we think about
how social technology platforms are produced and the ideology that underpins
these platforms. If users are seen as ‘building identities’ around self-promotion
and image, how that interaction is curated by the site or platform will reflect that
ideology. The statement suggests the company does not only see the site as a com-
munity of users, but rather as individuals looking to get ahead socially. The activ-
ities of ‘liking’, ‘commenting’ and ‘sharing’ of Facebook entries and to pick up
supposedly ‘trending’ posts and videos are understood as self-advertising or even
competitive moves in a virtual market.
For the average user of social media, however, commodification is not explic-
itly present beyond ads appearing on the sites they visit. Instead, interaction with
1
For Facebook see also Eisenlauer, Ch. 9, this volume, for Twitter Zappavigna, Ch. 8,
this volume, and for YouTube Johannson, Ch. 7, this volume.

Authenticated
others is dictated by a complex set of assumptions about what is and isn’t appro-
priate in certain online environments. These assumptions and practices lead to
a diverse range of interactions on different sites. Within these different sites for
engagement, there are then different ways in which ideology is present. Ideology is
enacted in different ways on different sites, depending on the affordances the tech-
nology provides and how users interact with it. The study of ideology on social
media is then not just a question of analysing the discourse of interactants on the
site, but also of how the site facilitates interaction.
Online interaction has long been known for aggressive, antagonistic interac-
tion. It has been one of the main historical interests about online interaction, going
back to the early eighties (Kiesler, Siegel and McGuire 1984). In online forums,
for example, particularly those dedicated to controversial issues that are likely to
draw opposing sides, contestation is the norm. The nature of a social media site and
how users interact on it will impact on what norms emerge. Users are competing to
be heard, and that competition can emerge as antagonism when given the right set
of factors. There are, of course, places where consensus is the norm and users are
not prone to attack one another. On Facebook, for example, when a young father
posts a picture of his new-born daughter with the message ‘#blessed’, another set
of complex factors will lead to an outpouring of positivity, both in the ‘likes’ the
photo will receive from friends and family and in positive text comments on the
photo. It is very unlikely that anyone will say something negative or criticise the
father for posting the picture.
To understand why there is such a difference in the responses, we must take
into account the different factors affecting interaction: the topic, audience, text
producers and site for interaction (see also Fritz, Ch. 11, this volume). Changing
any one element of the scenarios we have given as examples means a possible
change to how the audience might react and the extent to which the user will be
met with praise, animosity or indifference. The ideologies of users and enacting of
the ideology in interaction will affect how people respond to the actions of others.
Uncovering what ideologies are at work, and how they impact on any given inter-
action leads to important methodological issues for researchers, such as what data
to collect and when to collect it.
4. Methodological issues
So far, this chapter has focused on ideology in the practices of users in their inter-
action on social media platforms. However, ideology is not only embedded in the
production of social media data. It is also present in the ways in which data is
managed and made publicly available. For researchers considering which data to
analyse and whether permission is needed to access the date, ethical considera-
tions are particularly important.

Authenticated
While historically interaction online could roughly be considered either ‘pub-

lic’ or ‘private’, social media blurs these distinctions, something Lange’s work
(2007) identified early in the life of YouTube. For example, a Facebook user with
10 or 15 friends may post personal information about themselves that is publicly
available, but doesn’t appear to be oriented towards a public audience. Users’
physical location may be shared on their page, and they may assume that no one
would be looking at their page except their friends. A researcher might say that
this material is publicly available and therefore requires no consent to analyse.
Although this may be the case in legal terms, particularly if the posts are of a
personal or controversial nature, the researcher must weigh carefully the effect of
recontextualising social media texts in other environments (like research papers or
conference presentations).
Social media interaction, like Internet interaction more generally, also prob-
lematises the notion of multiple identities (boyd 2011: 40–41), and the extent to
which one can have a ‘real’ identity online in contrast to an offline identity. The
extent to which these opinions represent what a user really thinks about a par-
ticular topic is difficult to judge. Platforms like YouTube, for example, encourage
users to behave in more antagonistic ways than they might otherwise, in attempts
to generate interest around their videos (see Pihlaja 2014a for an analysis of drama
videos on YouTube). Although careful discussion of important issues is certainly
present on social media, more often than not, extreme and antagonistic positions
rise to the top as users are attracted to the drama surrounding these positions and
their responses. In these cases, it can be difficult to extract what a user’s ‘real’
position is, rather than what they are presenting to garner attention. This can be
seen in atheists attacking Christians on YouTube (and vice-versa) claiming that
their opponents are acting like Nazis. One user was repeatedly called ‘Hitler’ for
suggesting that non-Christians would be burned in hell (Pihlaja 2014a: 74). The
extent to which they genuinely hold that position, or are simply using the language
of antagonism for rhetorical effect, is not always simple to deduce.
How to analyse texts on social media is also fraught with difficulty, taking
into account all the complexities of interaction. In terms of linguistic pragmat-
ics, there are numerous ways to investigate the ways ideology is present in social
media interaction. We’ve touched on identity construction, discussing how users
might present more extremist views in online settings than they might otherwise.
In terms of the actual texts and utterances produced, ideology can be considered
by looking at a variety of different elements in communication. For example, the
use of quotations in presenting one’s own position can be very important in pre-
senting a particular ideology. This can include quoting authoritative texts, as when
Christians quote the Bible to show that their position is the same as Jesus’ (Pihlaja
2013). The quotation of other users can vary in a chatting ‘thread’ and can serve
a variety of purposes, ranging from consensus creation over dialogic uptake and
critique to the polemical dismissal of (presumptive) common ground presupposi-

Authenticated
tions (Arendholz 2013, 2015; Bös and Kleinke 2015). Metaphor is also a rich topic
of analysis, particularly when investigating how ideology is implicit in talk about
controversial issues. Metaphor allows users to state more controversial positions
in a veiled way, like calling someone a ‘wolf’ rather than explicitly saying that
they are lying. By using the metaphor of ‘wolf’, users can avoid making a direct
accusation, while still implying that a person is dangerous or unreliable (Pihlaja
2016).
Although linguistic expression is an important part of expressing ideology in
social media, ideologies are also implicit in the non-linguistic, multimodal ele-
ments of interaction. In text-based computer-mediated interaction, these issues
have been investigated in the use of emoticons in text and instant messaging
online. As technology has developed, interaction around images, gifs, and videos
has increased, and with these technologies, analysis of language is not sufficient
to provide complete descriptions of how interaction develops. For example, ‘inter-
net memes’ (in the sense of iconic images matched with formulaic language) are
effective not simply because of their linguistic content, but of the interaction of
the language with the image and the audiences’ shared knowledge of the formula
of the meme. An example can be seen in the sharing of the “Pepe the Frog” meme
during the 2016 Presidential election. The meme had become a symbol of White
Nationalism in the USA and was assertively used in the campaign for presiden-
tial candidate Trump (BBC 2016; Clinton 2016), but recognising this required
contextual, shared knowledge as the frog cartoon isn’t inherently a racist icon.
Here again, there is both the potential for ideologies to be explicitly present in the
semantic value of the meme, but also implicitly in its shared use over time and the
meanings of particular images, formulaic language, and video styles garner in use,
as with Internet memes (Gal, Shifman and Kampf 2015). Unpacking the meanings
of the memes requires not only the analysis of the text and image, but also some
input from the community of users producing and consuming the images, and
their histories, which can be gained from observing the community over time and/
or asking them directly, as Androutsopoulos (2008) uses in Discourse-Centred
Online Ethnography.
With any research of social media, there is also an important issue of scale.
Researchers can look very closely at a small number of interactions, or can look at
huge datasets with hundreds of millions of words. How data is gathered, and the
kind of data that is analyzed will always be subject to the research question one
is interested in investigating. To exemplify how some of these methodical issues
play out in research both in scale and in the analytic approach, we will now look
briefly at examples of research, taken from our own work and that of colleagues.
The examples show how these methods for the analysis of ideology can differ
and elicit the ways in which researchers can approach questions of ideology both
implicitly and explicitly.

Authenticated
5. Examples of ideology on social media
Internet discussion forums and blogs can be ideology-laden and polemical. This is
hardly surprising given that many of them are maintained and managed by polit-
ical institutions, parties, the media and individual ‘public’ personalities. Forums,
as publicly accessible discussion platforms, invite and foster debate and polemical
dispute, with varying degrees of editorial ‘gatekeeping’ but usually open to all
visitors who register or simply post under an assumed online identity (Suler 2004;
Wiszniewski and Coyne 2002). The main gatekeeping sanction available to the
forum website’s editor is the removal of a posting, which is resented by the ‘vic-
tims’ as a form of censorship but on the other hand helps to exclude hate-speech
and openly insulting behaviour (Li 2013; Ihlebæk and Krumsvik 2015), so that the
flow of communication can be maintained and a (precarious) community-specific
ethos is established. The interventions of forum managers thus concern mainly the
interactional dimension of user communication, rather than a monitoring of taboo
areas for ideational content (with the exception of criminal activities). Of course,
this does not mean that taboo topics such as sex, death, bodily functions etc. are
plainly spoken about in forums; as in other forms of discourse they are elaborately
treated by way of lexical and textual means of euphemistic (and dysphemistic)
reference and representation (Crespo-Fernández 2015).
Forum participants have a chance to interact with each other emphatically and
in real time by quoting and commenting both on press and online articles (and
reacting to those genres’ own characteristic quotation practices) and on each oth-
er’s postings, thus continually constructing and negotiating individual and com-
munal ‘self’-, ‘other’- and mutual face-relationships (Angouri and Tseliga 2010;
Bös and Kleinke 2015; Korenmann and Wyatt 1996; Landert 2015; Landert and
Jucker 2011; Neurauter-Kessels 2011). In the comment sections of online media,
the respective original articles and/or other ‘precedent’ texts serve as primary ‘ref-
erence points’ for follow-up quotations and comments but they are also instru-
mentalized to build up the commentators’ own communicative identities through
evaluation, stance-taking, repositioning, selective or ironic meta-representation
(Johansson 2015; Reich 2011). Thus, in an online debate on the popular history (or
perhaps better: mythology) of German(ic) national identity, participants competed
for authority status by arguing over the right to quote their most famous source, i.e.
the writings of the ancient historian Tacitus, e.g. by correcting each other’s trans-
lations of key-passages, accusing each other of misquotation and some adopting
the ‘online identity’ of Tacitus for themselves (Musolff 2015a: 131–140). Forum
participants also quote each other cumulatively, as it were, to escalate an argument
as in this debate on the BBC’s “Have Your Say” website about Pope Benedict’s
public condemnation of paedophile priests: “‘So one elderly German in a badly fit-
ting white frock says sorry in English and that makes everything OK does it?’ [par-
ticipant 1] quoted by participant 2: ‘No the offenders will be judged, Judged that

Authenticated
is, when time comes’; both quoted by participant 3: ‘It also helps that the church
is compensating many of the victims with money’” (Bös and Kleinke 2015: 79).
Whilst Internet forums are oriented towards providing a platform for wide pos-
sible audiences and open debates, blogs, on account of their being maintained by
individually identifiable persons, groups or institutions, often cultivate an ‘insider’
ethos (Bolander 2012, 2013; Bruns and Jacobs 2006; Herring et al. 2005; Hoff-
mann 2012; Miller and Shepherd 2004; see also Heyd, Ch. 6, this volume). Politi-
cal and communicative ideologies of blogs are strongly intertwined, although this
does not completely exclude the possibility of disagreement and robust discussion;
however, they take place within the blog’s in-group that defines itself through
their shared interest in maintaining the blog. With the establishment of blogs as a
“stabilized genre” (Yus 2011) they provide ample material for pragmatic research.
This has focused so far, for example, on the authors’ “intentions when uploading
information”, their adherence and/or deviation to the genre-specific standards, the
readers’ interpretations and follow-up reactions, and their success in communi-
ty-building and maintenance (Yus 2011: 95; see also Kopytowska 2013; Singer
2006).
The difference between blogs and related but distinct CMC genres such as
online press publishing and internet forums has become a special focus of research.
Musolff (2015b) has investigated a mixed corpus of texts in all three genres that
discussed UK immigration policies, with a focus on dehumanizing metaphors such
as parasites for ‘immigrants’, as in this defamatory characterisation: “If they hav-
en’t been detected for ten years then they are either living via the proceeds of
crime or tax dodging. And that makes them parasites and criminals” (BBC 2010b).
In terms of their overall ideological stance, all genres were characterised by a
majority of cautious to hostile anti-immigration attitudes and judgements. How-
ever, only the Blogosphere exhibited a consistent xenophobic and polemic bias
insofar as the parasite metaphor was regularly used together with further ‘disgust-
ing/dangerous organism’ terminology to stigmatize immigrants. In online forums,
such as the BBC’s Have Your Say website, and to an even greater extent in the
press, the use of the immigrants-as-parasites metaphor was explicitly criticized
as xenophobic and racist and attributed to the ‘extremist’ section of the political
spectrum. Beyond this extremist fringe there is the grey zone of terrorist groups’
websites, whose success in disseminating ideologies and acquiring followers is
still hotly debated (Benson 2014; Brunst 2010; Ducol 2012).
YouTube, since its foundation in 2005, has also been a site for contentious
arguments, and served, inter alia, the conflict-rich interaction of religious com-
munities (Pihlaja 2011, 2014). It has, for instance, become an important catalyst
for the emergence of so-called ‘New Atheism’, a position that gained prominence
on the site (Pihlaja 2014). While the basic starting point of the New Atheist move-
ment was simply a lack of belief in God or gods, it has become attached to other
positions, such as a belief in the supremacy of science and the scientific method,

Authenticated
a rejection of institutionalised religion, and a focus on a critical approach to any

and all other beliefs.
As atheist voices grew prominent on YouTube, Evangelical Christians appeared
on YouTube as well, often championing voices which opposed the scientists and
humanists that atheist users supported and cited. The authority of prominent athe-
ists like Richard Dawkins and Sam Harris was met with Creationist Ken Hamm
and Christian apologist William Lane Craig. In instances where users argued,
prominent thinkers from both sides were often present either in the arguments,
or in video clips or lectures posted on YouTube. In this way, ideological positions
of particular individuals were revoiced and recontextualized in the social media
space, where they could be debated and reworked to meet users’ own contextual
arguments. The result was often a hardening of positions, with less focus on dis-
covering the truth, and more on winning an argument and shaming an opponent.
Within the disagreements among Christians and atheists on YouTube, a key
topic of argument has been Creationism. For atheist users, Christians’ lack of
belief in evolution is incoherent and illogical, evidence that Christian belief is, at
its core, unreasonable. Christians in general and Creationists in particular should
be the object of scorn and ridicule, with atheist users making parodies of promi-
nent Christians and mocking their positions (Pihlaja forthcoming). For Christians
(often Evangelical Christians, based in the USA) on YouTube, the belief in the
Bible and a literal reading of the Christian creation myth is essential to their belief
about the authority of the Bible and their particular reading of the text. The argu-
ment about Creationism was part of a larger argument about the authority of the
Bible and how Christians interacted with it.
In these arguments, metaphor and us-vs.-them categorisation have been par-
ticularly important for users attempting to construct their own group identities
through attacking and defending other users. Metaphor analysis (Cameron and
Maslen 2010) showed, for example, that the use of words and phrases from the
Bible could be used to support a user’s particular position by embedding it in an
authoritative discourse that sounded vaguely biblical. A user could mix language
taken from the Bible and mix in their own interpretation to suggest that some peo-
ple were ‘wolves’ while others were ‘sheep’, using the language of the Bible to
categorise people in their own social context (Pihlaja 2016). User-categorisation
is of key-importance in these discussions (Housley and Fitzgerald 2002). Users
employ categories as a way of affiliating with friends and then attacking enemies,
e.g. by using categories like ‘good Christians’ and ‘bad Christians’ to differentiate
among other users, and alienate and judge others with whom they disagreed (Pih-
laja 2014b). Finally, impoliteness and causing offense was an important way for
users to draw attention to their own videos and the message they were attempting
to spread. By employing language that others judged as offensive such as calling
another person ‘human garbage’, users could draw attention to themselves and
build their popularity both among those who agreed with them and those who,

Authenticated
even if they were offended, engaged with the discourse to express their disagree-
ment.
This ongoing argument between atheists and Christians on YouTube serves as
a good example of how ideology can be identified and analysed in social media
context. While Creationism is simply one argument of many, by focusing dis-
course analysis on how users from both sides argue, what sources of authority
they appeal to, and how they react to challenges, ideological positions become
clear, beyond their own stated positions. Pihlaja (2014a: 151) showed that when
given the chance to sit down together in a shared physical space, two of the most
antagonistic users from both sides of the argument could find common ground and
shared experience in conversation. The physical meeting of the users which they
recorded and uploaded to YouTube suggested that the positions taken by both on
their own channels, and particularly their antagonism towards one another, was
likely heightened by the distance both in time and physical location that YouTube
video facilitated, and that when sat together over lunch, their own ideological posi-
tions could at least accommodate the other’s position.
Other scholars have sought to offer broader descriptions of online behaviour
that have implications for the study of ideology on social media. Hardaker and
McGlashan’s (2016) work on Twitter looks at rape threats made on feminist cam-
paigner Caroline Criado-Perez. Their research investigates a corpus of tweets with
the goal of understanding the language surrounding sexual aggression on the site
and shows disturbing, if perhaps unsurprising, language directed at women on Twit-
ter, and in particular, towards Caroline Criado-Perez. Hardaker and McGlashan
(2016) use a mixture of discourse analysis and corpus linguistics methods, start-
ing with a focus on frequency, collocation, and keywords. This method allowed
them to identify places to investigate how sexual aggression is enacted and what
the language might say about different communities of users. The use of corpus
linguistic tools then allows the researchers to provide a quantitative assessment of
how a particular pattern of action occurs. These patterns could then have important
implications for the ideologies that different user communities hold.
The findings also show how ideologies spread on social media. Hardaker and
McGlashan note, that
perhaps the most crucial issue here was how cleanly and neatly different “communities”
or “groups” can be identified, especially when dealing with a highly fluid, fast-moving
environment like Twitter populated by users who may coalesce around a topic or user
and engage in transient interactions for a mere matter of seconds before moving on. In-
deed, terms like “community” or “group” seem far too strong for a collection of people
who may have no further connection to each other than to have tweeted the same target
with either support or abuse. (Hardaker and McGlashan 2016: 91–92)
This quotation highlights an important factor in researching ideological statements
online: Similar expressions of beliefs or positions are not necessarily co-ordinated,
and, more interestingly, might not be the independent position of a user. Instead,

Authenticated
users may be ‘joining in’ without considering what they’re saying or why they’re
saying it. What the research shows is thus not necessarily a community of users
with shared beliefs working together (either explicitly or implicitly) to spread a
particular set of beliefs. Instead, in the moment of the social media face attack,
latent ideologies (in this case, anti-feminist ideology) can become apparent. The
users who threatened and attacked Caroline Criado-Perez were not clearly affili-
ated with one another, but behaved in similar ways, using similar language. The
‘coalescing’ around a particular event and the different reactions that come out
in response are then sites for ideological expression (Hardaker and McGlashlan
2016). While ‘community’ remains a recurring theme in research on online inter-
action (Jones 1995; Herring 2004; Seargeant and Tagg 2014), the term is problem-
atic given its variety of uses and its comparison to a traditional notion of commu-
nity which simply can’t be applied to online spaces (Yuan 2013) where users may
or may not share the same ideologies even within the same ‘community’.
This is an important point to consider when researching ideology in interaction
online. While our other examples showed people who were committed to specific
positions and beliefs over time, the users making threats on Twitter may not be
operating from a specific set of explicit beliefs, but react to something they have
seen and then move on immediately to something else. This does not negate the
importance of the ideology which is implicit in the action, but it does need to be
considered in a different light than the sustained positions held by users in Pihla-
ja’s (2013, 2014a, 2016) studies of Christians and atheists, which represent a more
coordinated, systematic presentation of ideology.
These different case studies highlight the different ways that ideology can be
investigated on social media, from the explicit object of study as in Musolff’s
(2015a) work, to research which uncovers underlying ideologies in particular social
media events. The different studies also highlight different approaches to social
media data. Researchers have looked at particular pages or websites as a location of
interaction, seeing, for example, how racist ideology is discussed and spread, often
by way of explicitly disclaiming racism, in blogs and Youtube forums (Goodman
and Row 2013; Kettrey and Laster 2014). Experiments have furthermore indicated
that that frequency of Facebook use related positively to message acceptance, par-
ticularly messages with overt racist content (Rauch and Schanz 2013: 613).
We will now move on to some key issues and controversies in the study of
ideology on social media and suggest ways in which difficulties might be met.
6. Critical issues and controversies
Historically, the issue of online anonymity has been a key issue for the analysis of
CMC in general and has important implications for the study of ideology online. If
users are able to communicate without the repercussions of immediate social feed-

Authenticated
back as in face-to-face communication, they may be more willing to take extreme

positions, such as posting racist memes. Most recently, this concern can be seen
in fears about radicalisation online and fears that young people could be drawn to
more extremist views. Anonymity has been seen as contributing to users taking
more radical positions online and recruiting others to their positions (Neo et al.
2016). Whereas physical communities tend to discourage users from speaking in
overtly radical ways whether it be espousing racist opinion or encouraging hate
against specific individuals, in online contexts without these social restraints and
feedback, users may take more radical positions.
While these concerns persist in public discourses about social media and the
Internet more generally, there is a role for researchers to play in understanding if
such concerns can be substantiated and to investigate also the ideological posi-
tions of the corporations that own the means of production. Thus, it is not only
important to analyse how users employ Twitter to spread extremist positions, but
the response by Twitter and the effect of responses to extremism are of equal if not
higher significance. In the case of Hardaker and McGlashan (2016) on trolling, for
example, it is not only important to understand why users post the threats that they
do, but why Twitter allows for the threats to be posted on the site.
The answer appears to relate to the role and necessity of ‘free speech’ on social
media, a critical issue for users of social media (Swigger 2013). Concerns about
free speech are often contrasted with the need to curtail ‘hate speech’ online, which
encourages violence or discrimination towards particular people or groups (Brown
2015). Social media companies like Facebook and Twitter operate in multi-na-
tional and -cultural contexts, so their decisions about what should and should not
be allowed has an important impact not only on the content of their own sites, but
how other sites are modelled after them. The concept of free speech and its limits
also has important implications for influencing how users interact with their own
governments, particularly when Twitter or Facebook are censored by national gov-
ernments, which happened for example during an event called “Everybody Draw
Muhammad Day” in 2010. Users on social media sites encouraged others to draw
pictures of Muhammad and post them on social media, leading to government
agencies in Pakistan to censor the sites and limit access (Greene 2010). As the cor-
porations play an important role in deciding what is acceptable, there is important
work still to be done on critiquing the institutions and their processes, and how
national governments interact with these decisions and support or oppose them, or
attempt to influence them.
In addition to fears about extremism, there is also difficulty with online dis-
course and identity in deciding to what extent ideological positions displayed
online, particularly in anonymous settings, represent a user’s actual ideology. In
research on Twitter like the work above (Hardaker and McGlashan 2016) and stud-
ies of trolling (Buckels, Trapnell and Paulhus 2014; Hardaker 2010; Herring et
al. 2002), the extent to which individual users stand by their words and actions in

Authenticated
online environments is a complicated issue. These issues have been evident even
from early narrative research about YouTube, where users expressed complex feel-
ings about flaming and trolling on the site and their own engagement in it (Moor,
Heuvelman and Verleur 2010). Questions of multiple identities not only compar-
ing offline and online interaction, but also different sites of interaction online,
remain an important issue to investigate.
The issue of multiple identities is not, of course, bound to online spaces, as
people perform different identities in different settings. Social media, however,
can highlight disparities among positions, particularly on social networking sites,
where users perform multiple identities for a mixed audience that may include
family, friends, colleagues, and the public. In work done by Marwick and boyd
(2011), the complexity of these issues has come up regularly in the use of social
media. Whereas in offline contexts, users can mitigate (in some ways) different
audiences with different physical spaces, the pressure to ‘friend’ colleagues and
family members on places like Facebook may mean that having multiple authen-
tic selves in the context of that site is limited when audiences are ‘flattened’ into
one (Marwick and boyd 2011: 122). Again, the response of corporations in the
development of the platforms is a key to how users manage these issues, with
different ‘circles’ in Google’s social media site, Google+, or in different “lists” on
Facebook. These technologies are continually changing and the extent to which
the platforms adapt and innovate their platforms will continue to dictate how users
present themselves and their beliefs on the sites.
The effect of ‘context collapse’ on ideology is also worth considering, par-
ticularly in how ideological positions are displayed in social media contexts. If
users must perform different identities, it is also then plausible to consider the
different ideologies they may perform, given a different audience. Muslim users
of social media, for example, can focus on the compatibility of Christianity and
Islam, when speaking to a predominantly Christian audience (Pihlaja forthcom-
ing). In this case, ideology might not then be a static position that a user holds
and promotes and defends online, but rather a malleable set of beliefs that comes
out in different ways in different contexts, depending on the perceived audience
and the affordances of the technology. Increasingly, however, there has been more
pressure for users to produce ‘authentic’ versions of themselves on social media.
With the rise of social networking sites and the integration of mobile technology
and the Internet, research now also focuses on this so-called “context collapse”
(Marwick and boyd 2011: 122) online, where users struggle to maintain authentic-
ity in online spaces where they interact with family, friends, and co-workers, along
with Internet users whom they may have never met in offline contexts. This is, in
many ways, directly tied to the ideology of the Facebook platform, which values
authenticity in a way that previous platforms have avoided. Users are encouraged,
if not required, to use real names and connect with people they know and have
met through offline interactions. This coupled with the proliferation of mobile

Authenticated
technologies that allow users to check in at physical locations and integrate their
interaction on social media with their physical movements, may mean a future
where maintaining different personas and selves online could become increasingly
difficult.
The extent to which the integration of online mobile technologies with offline
contexts will further this ‘collapse’ and require users to present more consistent
positions across platforms is yet to be seen. With these shifting positions and con-
texts, researchers looking at language must be particularly in tune with why users
are interacting in specific social media spaces at specific times, and take care to
look carefully at the various factors surrounding expressions of beliefs and posi-
tionings, including the site of interaction, who is interacting, and the history of
interaction in a particular context. Understanding that expressions of ideology are
plastic and highly contextual in these environments is of key importance to under-
standing why certain expressions are occurring.
7. Conclusion
In this chapter, we have focused on the ways in which ideologies are present both
in the explicit practices of social media interaction and the implicit ways in which
the platforms themselves promote particular ideologies. Depending on the focus
of discussion and research, individuals will encounter different theoretical, prac-
tical, and ethical issues in their investigation and use of social media platforms.
Social media presents a specific challenge in the study of ideology because the
interaction is mediated by the different platforms, changing how users express
themselves and what positions they take. Researchers must look carefully at the
contexts of interaction over time to provide nuanced, careful descriptions of how
ideology is shaped by the technology used. From there, researchers can then begin
to analyse discourse and offer insight into how beliefs are spread and adapt in
social media contexts.
This analysis must take into account the different factors that influence the pro-
duction of ideology on different sites. First, the platform itself must be analysed,
and the ideological positions of the platform producers. The ideology of Face-
book plays an important role in how users interact on the site, encouraging certain
behaviours, while discouraging others. The analyst must be aware of the features
specific to certain platforms and how these impact on the use of the technology.
No platform is without an ideological position, and awareness of the position is
necessary for understanding how users interact on a site.
Second, researchers must be aware of the ways in which historical positions
held in the so-call ‘real world’ are at work in new media. While social media may
give an increased access to broad audiences in a way that hasn’t been possible
historically, this does not mean that interaction on social media is necessarily new

Authenticated
or novel. In many cases, the Internet and social media specifically have become
simply new sites for old arguments. As in the example of Fundamentalists argu-
ing for Creationism, the argument itself is not new, but rooted in a long standing
disagreement and presentation of a specific ideological position that is quite old.
The historical context is then essential for doing analysis of social media ideology.
Third, there must be an awareness that arguments online are often not about
clarifying positions or convincing others, but ways of presenting ideological posi-
tions to build communities and serve social purposes like insulting or demean-
ing others. Moreover, as Verschueren states, we must remember that “[t]he idea
that an issue – any issue – could be settled once and for all is an illusion” (Ver-
schueren 2012: xi). The analyst’s role is then to understand both the ideology that
is being represented and the way in which language is being used as social action
in online communities. Linguistic analysis provides the tools for describing how
users accomplish this work. Analysis of, for example, linguistic impoliteness can
provide insight into the positions that users hold based on what they do or do not
find offensive and how other users respond to (and are offended by) the claims that
they make and/or how they act online (Angouri and Tseliga 2010).
Fourth, there is a need to understand that the access social media provides
allows users to interact with individuals outside of their physical social connec-
tions who might share minority ideological positions. As in the case of atheism on
YouTube, users can develop movements and communities around ideologies that
have been historically seen as suspect and discouraged. When given the chance to
interact with users on social media, users can come to build online communities
around these ideas and impact offline social change, using the Internet as the main
means of publicising their message. Alternatively, given the need to interact with
broad audiences coming from different ideological positions, social media interac-
tion can also be influenced by ‘context collapse’, where users have to maintain a
kind of authenticity among a diverse group of users, including those that disagree
with them and those whom they might wish to influence.
While historically anonymity has been of central concern to researchers look-
ing at online interaction, social media and pressure to maintain an ‘authentic’ posi-
tion online change the dynamics of how users see interaction online. While anon-
ymous interaction is still possible and does still lead to antagonism and hatred, the
interaction on places like Facebook, with user names attached to real names and
photos, rather than avatars and screennames, the interaction between offline and
online lives will seemingly continue.
Social media will remain an interesting and engaging site for research as long
as users of diverse backgrounds interact with and challenge one another. As users
are constantly challenging the notion of what counts in a particular community
culture as common sense, ideologies must be constantly re-asserted, revised, and
defended. With another user with a differing opinion only a click away, there is an
almost constant back-and-forth among users over their ideological positions. This

Authenticated
contestation can lead to users rethinking their ideology, but can also lead to users
becoming more entrenched in their positions and building communities around
their particular ideology. Social media then provides a useful arena for seeing both
how antagonism and argument between ideologies are played out, and how com-
munities emerge around ideological positions.
References
2008 Potentials and limitations of Discourse-Centred Online Ethnography. Language
@Internet 5. http://www.languageatinternet.org/articles/2008/1610 (accessed
September 26, 2016).
2010 “You have no idea what you are talking about!” From e-disagreement to e-im-
politeness in two online fora. Journal of Politeness Research: Language,
Behaviour, Culture 6(1): 57–82.
Arendholz, Jenny
2013 (In)Appropriate Online Behavior: A Pragmatic Analysis of Message Board
Arendholz, Jenny
2015 Quoting in Online Message Boards: An interpersonal perspective. In: Jenny
ics of Quoting Now and Then. 53–69. Berlin/Boston: de Gruyter.
BBC
2010a Have your Say: Should politicians be talking about immigration? http://www.
bbc.co.uk/blogs/haveyoursay/2010/04/should_politicians_be_talking.html
(accessed December 15, 2013).
BBC
2010b Have your Say: How should immigration be tackled? http://www.bbc.co.uk/
blogs/haveyoursay/2010/04/how_should_immigration_be_tack.html.
(accessed December 15, 2013).
BBC
2010c Have your Say: Are immigration rules fair? http://www.bbc.co.uk/blogs/
haveyoursay/2010/06/are_immigration_rules_fair.html (accessed December
15, 2013).
BBC
2016 Pepe the frog branded a hate symbol. 28 September 2016. http://www.bbc.com/
news/world-us-canada-37493165 (accessed September 29, 2016).
Belair-Gagnon, Valerie
2015 Social Media at BBC News: The Re-making of Crisis Reporting. London/New
York: Routledge.
Benson, David C.
2014 Why the Internet is not increasing terrorism. Security Studies 23(2): 293–328.
Berker, Thomas, Maren Hartmann, Yves Punie and Katie Ward (eds.)
2005 Domestication of Media and Technology. Maidenhead, UK: Open University
Press.

Authenticated

de Gruyter.
Bolander, Brook
Bolander, Brook
boyd, danah
2011 Social Network Sites as Networked Publics: Affordances, dynamics, and
implications. In: Zizi Papacharissi (ed.), A Networked Self. Identity, Commu-
nity, and Culture on Social Network Sites, 39–58. London: Routledge.
Brown, Alexander
2015 Hate Speech Law. A Philosophical Examination. London: Routledge.
Bruns, Axel and Joanne Jacobs (eds.)
2006 Uses of Blogs. New York: Peter Lang.
Brunst, Philip W.
2010 Terrorism and the Internet: New threats posed by cyberterrorism and terrorist
use of the Internet. In: Marianne Wade and Almir Maljevic´ (eds.), A War on
Terror?: The European Stance on a New Threat, Changing Laws and Human
Rights Implications, 51–78. New York: Springer.
Buckels, Erin E., Paul D. Trapnell and Delroy L. Paulhus
2014 Trolls just want to have fun. Personality and Individual Differences 67:
97–102.
Cameron, Lynne and Rob Maslen (eds.)
2010 Metaphor Analysis: Research Practice in Applied Linguistics, Social Sciences
and the Humanities. London: Equinox.
Cimino, Richard and Christopher Smith
2014 Atheist Awakening: Secular Activism and Community in America. Oxford:
Clinton, Hillary
2016 https://www.hillaryclinton.com/feed/donald-trump-pepe-the-frog-and-white-
supremacists-an-explainer/ (accessed September 29, 2016).
Crespo-Fernández, Eliecer
2015 Sex in Language: Euphemistic and Dysphemistic Metaphors in Internet
Forums. London: Bloomsbury.
Crystal, David
2011 Internet Linguistics. London/New York: Routledge.
Ducol, Benjamin
2012 Uncovering the French-speaking jihadisphere: An exploratory analysis. Media,
War and Conflicts 5(1): 51–70.
Eagleton, Terry
2007 Ideology: An Introduction. London: Verso.

Authenticated
Fairclough, Norman
1989 Language and Power. London: Longman.
Gal, Noam, Limor Shifman and Zohar Kampf
2015 “It Gets Better”: Internet memes and the construction of collective identity.
New Media and Society 18(8): 1698–1714.
Goodman, Simon and Lottie Rowe
2013 Maybe it is prejudice … but it is NOT racism. Negotiating racism in discus-
sion forums about Gypsies. Discourse and Society 5 (1): 32–46.
Greene, Richard
2010 Pakistan shuts down Facebook over “Draw Mohammed Day”. CNN. http://
http://edition.cnn.com/2010/TECH/05/19/pakistani.facebook.shutdown/
(accessed September 27, 2016).
Halliday, Michael A.K.
1978 Language as Social Semiotic. The Social Interpretation of Language and
Meaning. London: E. Arnold.
Hardaker, Claire
2010 Trolling in asynchronous computer-mediated communication: From user dis-
cussions to academic definitions. Journal of Politeness Research. Language,
Behaviour, Culture 6(2): 215–242.
2016 “Real men don’t hate women”: Twitter rape threats and group identity. Journal
Harris, Kathleen and Turo Hiltunen
2014 “It’s you’re not your”: Exploring misspelled words in YouTube comments.
Studies in Variation, Contacts and Change in English 15. http://www.helsinki.
fi/varieng/series/volumes/15/nm_fifteen/harris_hiltunen/.
Herring, Susan
behavior. In: Sasha Barab, Rob Kling and James Gray (eds.), Designing for
Virtual Communities in the Service of Learning, 338–376. Cambridge: Cam-
Herring, Susan, Kirk Job-Sluder, Rebecca Scheckler and Sasha Barab
2002 Searching for safety online: Managing “trolling” in a feminist forum. The
Information Society, 18(5): 371–384.
Herring, Susan, Inna Kouper, John C. Paolillo, Lois Ann Scheidt, Michael Tyworth, Peter
Welsch, Elijah Wright and Ning Yu
2005 Conversations in the blogosphere: An analysis “from the bottom up”. Pro-
ceedings of the Thirty-Eighth Hawai’i International Conference on System
Sciences (HICSS-38), IEEE. //http://ella.slis.indiana.edu/~herring/blogconv.
pdf/.
Housley, William and Richard Fitzgerald.
2002 The reconsidered model of membership categorization analysis. Qualitative
Research 2(1): 59–83.

Authenticated
Ihlebæk, Karoline Andrea and Arne H. Krumsvik

2015 Editorial power and public participation in online newspapers. Journalism
16(4): 470–487
Johansson, Marjut
2015 Bravo for this editorial! Users’ comments in discussion forums. In: Elda
Weizman and Anita Fetzer (eds.), Follow-ups in Political Discourse, 83–107.
Jones, Steven
1995 Understanding community in the information age. In: Steven Jones (ed.),
Cybersociety: Computer-mediated Communication and Community, 10–35.
Thousand Oaks, CA: Sage.
Kettrey, Heather Hensman and Whiteney Nicole Laster
2014 Staking territory in the “world White web”: An exploration of the roles of
overt and color-blind racism in maintaining racial boundaries on a popular
web site. Social Currents 1(3): 257–274.
Kiesler, Sara, Jane Siegel and Timothy W. McGuire
1984 Social psychological aspects of computer-mediated communication. American
Psychologist 39(10): 1123–1134.
Kopytowska, Monika
2013 Blogging as the mediatization of politics and a new form of social inter-
action. A case study of ‘proximization dynamics’ in Polish and British political
blogs. In: Piotr Cap and Urszula Okulska (eds.), Analyzing Genres in Political
Communication, 379–421. Amsterdam/Philadelphia: Benjamins.
Korenman, Joan and Nancy Wyatt
1996 Group dynamics on an e-mail forum. In: Susan C. Herring (ed.), Comput-
er-mediated Communication: Linguistic, Social and Cross-cultural Perspec-
tives, 225–242. Amsterdam/Philadelphia: Benjamins.
Laclau, Ernesto and Chantal Mouffe
1985 Hegemony and Socialist Strategy: Towards a Radical DemocraticPolitics.
London: Verso.
Landert, Daniela
2015 Reportable facts and a personal touch: The function of direct quotes in online
news. In: Jenny Arendholz, Wolfram Bublitz and Monika Kirner-Ludwig
(eds.), The Pragmatics of Quoting Now and Then, 29–52. Berlin/Boston: de
Gruyter.
2011 Private and public in mass media communication. From letters to the editor to
online commentaries. Journal of Pragmatics 43: 1422–1434.
Lange, Patricia
of Computer-Mediated Communication 13(1): 361–380.
Li, Xigen
2013 Internet Newspapers: The Making of a Mainstream Medium. London: Rout-
ledge.
Markman, Kris M.
2010 Learning to work virtually. Conversational repair as a resource for norm devel-
opment in computer-mediated team meetings. In: Jun-Ran Park and Eileen

Authenticated
G. Abels (eds.), Interpersonal Relations and Social Patterns in Communica-

tion Technologies: Discourse Norms, Language Structures and Cultural Vari-
ables, 220–236. Hershey/New York: Information Science Reference.
Marwick, Alice
2013 Status Update: Celebrity, Publicity, and Branding in the Social Media Age.
New Haven, CT: Yale University Press.
imagined audience. New Media & Society 13(1): 114–133.
Miller, Carolyn R. and Dawn Shepherd
2004 Blogging as social action: A genre analysis of the weblog. In: Laura Gurak,
Smiljana Antonijevic, Laurie A. Johnson, Clancy Ratliff and Jessica Reyman
(eds.), Into the Blogosphere: Rhetoric, Community, and Culture of Weblogs.
https://conservancy.umn.edu/bitstream/handle/11299/172818/Miller_
Blogging%20as%20Social%20Action.pdf?sequence=1&isAllowed=y/
(accessed September 27, 2016).
Moor, Peter, Ard Heuvelman and Ria Verleur
2010 Flaming on YouTube. Computers in Human Behavior, 26(6): 1536–1546.
Musolff, Andreas
2015a Quotation and online identity: The voice of Tacitus in German newspapers
and Internet discussions. In: Jenny Arendholz, Wolfram Bublitz and Monika
Musolff, Andreas
2015b Dehumanizing metaphors in UK immigrant debates in press and online media.
Journal of Language Aggression and Conflict 3(1): 41–56.
Neo, Loo Seng, Leevia Dillon, Priscilla Shi, Jethro Tan, Yingmin Wang and Danielle Gomes
2016 Understanding the psychology of persuasive violent extremist online plat-
forms. In: Majeed Khader, Loo Seng Neo, Gabriel Ong, Eunice Tan Mingyi,
and Jeffery Chin (eds.), Combating Violent Extremism and Radicalization in
the Digital Era, 1–15. Hershey, PA: IGI Global.
Neurauter-Kessels, Manuela
2011 Im/polite reader responses on British online news sites. Journal of Politeness
Research 7(2): 187–214.
Pihlaja, Stephen
2011 Cops, popes, kings, and garbage collectors: Metaphor and antagonism in an
atheist/Christian YouTube video thread. Language@Internet, 8(1). Retrieved
from Language@Internet: http://www.languageatinternet.org/articles/2011/
Pihlaja/ (accessed September 26, 2016).
Pihlaja, Stephen
2013 “It’s all red ink”: the interpretation of Biblical metaphor among Evangelical
Christian YouTube users. Language and Literature, 22(2): 103–117.
Pihlaja, Stephen
2014a Antagonism on YouTube: Metaphor in Online Discourse. London: Bloomsbury.
Pihlaja, Stephen
2014b “Christians” and “bad Christians”: Categorization in atheist user talk on You-
Tube. Text and Talk, 34(5): 623–639.

Authenticated
Pihlaja, Stephen
2016 “What about the wolves?”: The use of scripture in YouTube arguments. Lan-
guage & Literature, 25(3): 226‒238.
Pihlaja, Stephen.
forthcoming Religious Talk Online: The Evangelical Discourse of Muslims, Christians,
and Atheists. Cambridge: Cambridge University Press.
Rauch, Shannon M. and Kimberley Schanz
2013 Advancing racism with Facebook: Frequency and purpose of Facebook use
and the acceptance of prejudiced and egalitarian messages. Computers in
Human Behavior, 29(3): 610‒615.
Reich, Zvi
2011 User comments: The transformation of participatory space. In: Jane B. Singer,
Alfred Hermida, David Domingo, Ari Heinonen, Steve Paulussen, Thorsten
Quandt, Zvi Reich and Marina Vujnovic (eds.), Participatory Journalism.
Guarding Open Gates at Online Newspapers, 96‒117. Oxford: Wiley-Black-
well.
Schieffelin, Bambi, Kathryn A. Woolard, and Paul V. Kroskrity (eds.)
1998 Language Ideologies: Practice and Theory. New York: Oxford University
Press.
Seargeant, Philip and Caroline Tagg
2014 The Language of Social Media: Identity and Community on the Internet. Bas-
ingstoke, UK: Palgrave Macmillan.
Singer, Jane B.
2006 Journalists and news bloggers: Complements, contradictions, and challenges.
In: Axel Bruns and Joanna Jacobs (eds.), The Uses of Blogs, 23‒32. New York:
Peter Lang.
Suler, John
2004 The online disinhibition effect. CyberPsychology & Behavior 7(3): 321–326.
Swigger, Nathaniel
2013 The online citizen: Is social media changing citizens’ beliefs about democratic
values? Political Behavior, 35(3): 589‒603.
van Dijk, Teun A.
1998. Ideology: A Multidisciplinary Approach. London: Sage.
van Dijk, Teun A.
2016. Critical discourse studies: a sociocognitive approach. In: Ruth Wodak and
Michael Meyer (eds.), Methods of Critical Discourse Studies, 62‒85. London:
Sage.
Verschueren, Jef
2012 Ideology in Language Use: Pragmatic Guidelines for Empirical Research.
Vogelstein, Fred
2009 The Wired Interview: Facebook’s Mark Zuckerberg. Wired.com. http://www.
wired.com/2009/06/mark-zuckerberg-speaks/ (accessed September 26, 2016).
Wiszniewski, Dorian and Richard Coyne
2002 Mask and identity: The hermeneutics of self-construction in the Information
Age. In: Ann Renninger and Wesley Shumar (eds.), Building Virtual Commu-

Authenticated
nities: Learning and Change in Cyberspace, 191–214. Cambridge: Cambridge

University Press.
Yuan, Elain
2013 A culturalist critique of ‘online community’ in new media studies. New Media
& Society, 15(5): 665–679.
Yus, Francisco

Authenticated
Authenticated
15. Facework and identity
Miriam A. Locher and Brook Bolander
Abstract: This chapter reviews studies which focus on Internet users’ attempts to
change (challenge, reinforce, negotiate) current or past, stereotypical, individual
and/or group identities in interactions. It thereby acknowledges that the literature
regularly draws on various theoretical conceptions of identity. Perspectives on
identity range from sociolinguistic understandings of the impact of social variables
on linguistic variation to constructivist and discursive negotiations of identity as
employed in conversation analysis, discourse analysis and discursive psychology.
These latter approaches share a particular view of identity as a complex, emergent,
context-sensitive, social, ephemeral/changing and negotiable concept. Methodol-
ogies vary according to theoretical orientation so that we find a rich mixture of
quantitative and qualitative studies. In addition to addressing these different con-
ceptualisations of and approaches to identity, the chapter also reviews a number of
key themes which emerge in the literature review: the importance of (im)polite-
ness; the impact and negotiation of gender; the construction of expertise, authen-
ticity and trust; the surfacing of emotions; the creation of in- and out-groups and
community building; and the intertwining of offline and online acts of positioning.
1. Introduction
It has been long established that language contains a relationship component and
that any act of communication thus implicitly and often also explicitly says some-
thing about the relationship of the interactants involved (e.g. Watzlawick, Bea-
vin and Jackson 1967). The same pertains to computer-meditated communication
(CMC), such that research on CMC also lends itself to studies on identity and
relationship construction. A famed cartoon by Peter Steiner (1993) in The New
Yorker depicts two dogs in front of a computer, one of which had been typing
and then informs the other that “on the Internet, nobody knows you’re a dog”.
Ever since this cartoon, it is not just scholars but also lay people who have been
aware of the fact that the keyboard can be used to write oneself into being (Sundén
2003). While the quotation highlights the anonymity that online interaction can
grant its users, more recent social media practices show that users also reveal their
name, post pictures and videos of themselves, and write texts (for example sta-
tus updates) to show their worldviews. They thus share a considerable amount of
information about themselves. This is not to deny that anonymous and/or creative
identity construction can still take place, but with the advent of web 2.0 platforms,
DOI 10.1515/9783110431070-015
Authenticated
408 Miriam A. Locher and Brook Bolander
the spectrum of different acts of positioning is even wider than before. Complex
acts of positioning which create affiliations and disaffiliations result in creative
practices of online identity construction.
This chapter first touches on a number of linguistic theories that discuss iden-
tity construction and its link to facework (Section 2). It then offers a succinct
review of recurrent themes in identity construction research dealing with online
data (Section 3). The chapter then moves to a presentation of a number of studies
in recent social media practices, such as Facebook and Twitter (Section 4), before
concluding in Section 5.
2. Theories of identity construction
Within linguistics, the topic of identity construction has a long tradition and it is
defined and approached in many different ways. The study of how interactants use
language to shape their persona or image has been of concern within many subfields,
among them pragmatics, sociolinguistics, discourse analysis, conversation analy-
sis, anthropological linguistics and literary pragmatics. Linguistics has benefitted
much from insights from other disciplines, among them rhetorics, gender studies,
sociology, anthropology, literary studies and discursive psychology, which focus
on language as well as other forms of communication. Depending on how widely or
how narrowly one defines identity, the above-mentioned disciplines and fields can
be drawn on for our respective research questions and methodological decisions.
Rhetorics, for example, has shown us that politicians who want to persuade
others in order to be elected or to convince people of their ideas, try to appear
knowledgeable but not scholarly, friendly and approachable but not chummy, etc.,
depending on their target audience and cultural context (see, e.g., Burke 1969 on
the use of pathos, ethos and logos). The study of how persuasion is performed and
power exercised is also of concern for linguists working on CMC, such as public
health sites or political campaign sites (see Section 3.3 below).
From sociolinguistics in the Labovian (1972) tradition, we learn that the ways
in which speakers utter words and choose lexemes reveal something about their
class, ethnicity, age and/or gender. Individuals may also aspire to sound like a
member of a higher social class (see, e.g., the phenomena of hypercorrection and
overt prestige) or adhere to their sociolect in order to distance themselves from
other groups and create ingroup solidarity (covert prestige; see, e.g., Labov 1972;
Meyerhoff 2006). Identity in these studies is functionalized through the inclusion
of social variables into the research design. As a consequence, identity is depicted
with a broad brush that considers the values of gender, class, age and ethnicity as
methodologically given.
Studies that use insights from linguistic anthropology, ethnology, and conver-
sation analysis provide further important findings with respect to the indexical

Authenticated
Facework and identity 409
power of language. These lines of investigation target a more fine-grained pic-

ture of identity since the focus is on the construction of identity in interaction.
For example, Irvine (2001) studies in minute detail how language indexes social
belonging in a Wolof village in Senegal and reports on Gal’s (1992) work on the
Hungarian village Bóly where cultural and ethnic belonging is indexed through
linguistic style. And Ochs (1993: 288) demonstrates how “speakers attempt to
establish the social identities of themselves and others through verbally perform-
ing certain social acts and verbally displaying certain stances” (emphasis in origi-
nal). The concept of membership categorization developed in conversation analy
sis (see, e.g., Sacks 1992; Antaki and Widdicombe 1998; Hester and Eglin 1997;
Fitzgerald 2015) helps to show the practices through which interactants them-
selves draw on lexemes for self-identification (such as mother, sister, daughter)
and style in order to highlight those aspects of their identities that they wish to
make salient in interaction.
The constructivist stance is also key in Eckert (1989) and Mendoza-Denton
(2008), who demonstrate how students use linguistic and other practices to index
ingroup and outgroup belonging. The latter work creatively combines quantitative
aspects from variationist sociolinguistics with qualitative observations about the
indexicality of linguistic markers. Furthermore, the study of identity construction
is expanded beyond a focus on language use to include other practices like after-
school activities, hairstyle, and clothing.
Research on stylistics and the pragmatics of fiction highlights that language is
a crucial means to create characters in fictional worlds. The authors’ and actors’
choices for register, accent and syntax tap into the readers’ and viewers’ world
knowledge in that they evoke stereotypes, which act as recognizable cues for the
readers/audience.1
From discursive psychology, we get the concept of positioning (Davies and
Harré 1990; Bamberg 1997; Deppermann 2013), which contributes to a better
understanding of how people create storylines for themselves in face-to-face inter-
actions, therapy and storytelling. This tradition complements the sociologist Goff-
man’s (1955, 1959, 1967, 1974) metaphor of the stage2 and the idea that people
take on and enact (dynamic) different roles in particular situations. Positioning
theory has also influenced Bucholtz and Hall’s work (2005, 2008, 2010; Hall and
Bucholtz 2013), which further demonstrates the interdisciplinarity of this research
field. They explore “the social positioning of self and other” (Bucholtz and Hall
1
Here, the reader is referred to the handbook Pragmatics of Fiction (Locher and Jucker
2017), which reviews crucial linguistic strategies for characterization, among them
regional, social and ethnic linguistic cues, stance markers and code-switching.
2
In Goffman’s “dramaturgical approach” (Willems 2001: 6298) the metaphors of theatre
and the stage are used to point to “the manipulative and the moral aspects of social life”
(Léon 2006: 98).

Authenticated
2005: 586), and posit that there are five principles which show the complexity
of identity construction. These principles highlight that identity is a relational
phenomenon (relationality principle), which emerges in interaction (emergence
principle), and which is indexed (indexicality principle) through processes of posi-
tioning through which interactants attempt to position self and other (positionality
principle). By studying such acts we only ever get a partial glimpse of identity
construction (partialness principle), which can be complemented (contradicted,
challenged, reconfirmed) in previous and following interactions and which is
influenced by competing (cultural) discourses (e.g. on gender and roles).
The concept of face is also important to (im)politeness research (and inter-
personal pragmatics more generally; see Locher and Graham 2010). This field of
research is concerned with relationship creation through interaction and it thus stud-
ies how face is negotiated and how facework patterns in different situations and cul-
tural contexts. Here Goffman (1955, 1967) is again important, since his concept of
face and facework are crucially interlaced with (im)politeness studies. To be judged
a polite or well-mannered person might be a motivation for choosing between dif-
ferent possibilities of performing a speech act. The ways we use language thus say
something about our understanding of the social situation we find ourselves in and
how we judge our relationship towards others (see also Section 3.1).
While research on facework overlaps with politeness research (and might be
considered synonymous by some researchers), there are scholars who work with
the concept of face without necessarily studying politeness norms. For exam-
ple, the concepts of rapport management (Spencer-Oatey 2007), relational work
(Locher and Watts 2005) and face-constituting theory (Arundale 2015) all draw
on the notion of face3 but do not necessarily or exclusively study (im)politeness
concerns. These concepts are thus more encompassing than (im)politeness. Yet
they can all be tied to identity management and relationship creation (see Locher
2008 for this link).
The field of gender studies, itself interdisciplinary in nature, also illustrates the
different approaches to the study of identity. There are numerous overview articles
(e.g. Mullany 2012; Sunderland and Litosseliti 2002; Swann 2002) that trace the
development from a more essentialist to a more constructivist approach and that
highlight that a gender identity may be one of many identity traits that might be
made salient by interactants in situ. Studying the impact of gender on CMC will
be taken up in Section 3.2.
3
The metaphor of face itself has been discussed extensively and its definitions vary
from assigning it the status of universal psychological wants (e.g. Brown and Levinson
1987) to more negotiable understandings (e.g. Goffman 1955). It is beyond the scope of
this chapter to give a history of this concept (but cf. Section 3.1). For an overview, the
reader is referred to Bargiela-Chiappini (2006).

Authenticated
This chapter cannot do justice to all of these different traditions of studying

identity construction, nor to all of the concepts that are associated with this process
(see, e.g., style, stance, indexicality, stereotype, (im)politeness). For overviews
on research on identity, see, among others, De Fina (2010), Joseph (2004, 2010),
Kiesling (2006), Locher (2008) and Mendoza-Denton (2002). What is central here
is recognising the diversity of approaches, since this same diversity is mirrored in
studies that explore identity construction in computer-mediated contexts and social
media in particular. For example, we find qualitative case studies conducted in an
ethnomethodological, conversation analytic or discourse analytic tradition, as well
as large-scale quantitative studies of linguistic expression where data is separated
according to social variables. Much of the research we review in Sections 3 and 4
also alludes to (im)politeness theories as identity construction goes hand in hand
with (im)politeness considerations (for observations on this interface, see Locher
2008; Hall and Bucholtz 2013).
3. Themes in identity construction research on online data
Reviewing the literature on identity construction in online data, we first need

to reiterate that any form of language use contains elements of recipient/audi-
ence design and thus represents a choice of a particular interactant as to available
linguistic variants. This means that any type of CMC can be studied from what
Locher and Graham (2010: 1) call an interpersonal pragmatics perspective, i.e.
with respect to its relational component. Relational work, i.e. the communicative
choices interactants make and how these are interpreted in light of relationship
creation (e.g. address terms, register, levels of mitigation, Locher and Watts 2005),
then leads to identity construction. Social network sites such as Facebook might
be immediately associated with relationship maintenance and creation since this is
also the declared aim of the platform. Yet we wish to stress that it is not only such
social network sites that lend themselves to the study of identity construction. Nor
does interactivity have to be built into the technical make-up of the practice as in
discussion boards or newspaper comments sections. Also less obviously social
or interactive online practices such as information websites can be explored with
respect to identity construction via a focus on concepts such as stance, style or
positioning.
In the following we introduce a series of pertinent themes for the study of iden-
tity construction online: the importance of (im)politeness; the impact and negoti-
ation of gender; the construction of expertise, authenticity and trust; the surfacing
of emotions; the creation of in- and out-groups and community building; and the
intertwining of offline and online acts of positioning. In light of our conviction that
identity construction can be performed in various modes, these themes do not stem
from studies on just one type of online interaction but from a mixture of sources.

Authenticated
In turn, insights from the studies may feed into more than one theme, such that
individual studies may be relevant with respect to a series of the identified themes.
3.1. (Im)politeness: From face-enhancing to face-aggravating behavior

Studying concerns about politeness and impoliteness is an important theme in lin-
guistic research on online communication (see also Graham, Ch. 17, this volume).
The link to identity construction hinges around the concept of face and facework
(see Goffman 1955, 1967; Bediijs, Held and Maaß 2014b). Face is defined by
Goffman (1967: 5), as “the positive social value a person effectively claims for
himself [sic] by the line others assume he has taken during a particular contact”.
Considering someone’s face in interaction means that speakers engage in facework
and project identities that might challenge, confirm or negotiate roles that surface
as relevant in a particular interaction. Since communication online is not faceless
in its metaphorical sense, we can assume that facework is equally important as it is
in face-to-face communication (see also the discussion in Bediijs, Held and Maaß
2014b and Held 2014).
In many ways what happens in this research field mirrors general trends in
(im)politeness research. These include discussions on methodological and theo-
retical concerns such as the role of universality within a given framework or the
distinction between emic and etic approaches; broadening the scope of data and
interest, which now also includes impoliteness and aggression; the role of emo-
tions; the emergence and negotiation of norms; processes of judging and historical
developments of modes of conduct in given societies (for overviews on develop-
ments in this field, see Locher 2013a, 2014, 2015). Given this wide field of interest
covered by (im)politeness research nowadays, it comes as no surprise that we find
many studies on CMC and (im)politeness, but no unified theoretical or methodo-
logical approach.
The literature review shows that there are several reasons why (im)politeness
scholars have been particularly motivated to research computer-mediated lan-
guage. These reasons are often intertwined and thus not mutually exclusive:
– To further theory building: The last two decades have seen an upsurge in theo-
retical discussions about how to approach the study of (im)politeness phenom-
ena. The new data available from computer-mediated contexts was readily uti-
lised as testing grounds for established and new ideas. Brown and Levinson’s
(1987) key concepts of face (adopted and developed from Goffman 1955) as
well as face-threatening act (FTA), and their taxonomy of mitigating the force
of FTAs (from bold on record, with mitigation, to off record) is also applied to
online data, at times in adapted forms (e.g. Bou-Franch and Garcés-Conejos
Blitvich 2014; Kleinke and Bös 2015; Herring 1994; Neurauter-Kessels 2011,
2013; Yus 2011). At the same time, scholars also employ more recent discur-

Authenticated
sive understandings of politeness and particularly focus on local and situated

emergent forms of (im)politeness (e.g. Angouri and Tseliga 2010; Graham
2008; Haugh 2010; Locher 2008; Planchenault 2010).
– To study emic negotiations of norms of conduct in and about computer-medi-

ated communication: As interactants adopt new forms of CMC, new norms of
conduct are explicitly commented on and addressed. As a result, we find many
practices where meta-discussions on appropriate or inadequate behavior are
negotiated. This is the case with respect to forms of Netiquette that were pub-
lished and discussed (e.g. for email conduct and behavior on message boards).
In addition, there are many threads in discussion boards or online articles
where interactants comment on offline codes of conduct. This data is thus rich
for scholars who are interested in the negotiation of norms on personal, local
or societal levels and who focus on meta-pragmatic comments. Representative
examples of such studies are Arendholz (2013, 2014), Graham (2007, 2008),
Haugh (2010), Haugh, Chang and Kádár (2015) and Locher (2008).
– To explore face-aggravating behavior: Scholars interested in developing polite-

ness theories which can also deal with impoliteness and aggressive face behavior
found rich data of conflictual and face-aggravating behavior in computer-medi-
ated contexts. While recording aggression and conflict in face-to-face contexts
might often be a matter of chance, online data seems to give easy access to such
data (for ethics and other research challenges, see Bolander and Locher 2014).
Numerous studies focus especially on face-aggravating data and explore it with
(im)politeness theories (e.g. Arendholz 2013, 2014; Kunkel 2014; Neuraut-
er-Kessels 2011, 2013; Langlotz and Locher 2012). In early research on online
communication, the possibility for anonymous posting was seen as an unin-
hibiting factor that resulted in less mitigation, directness and bluntness (Baym
1996; Reid 1999). In addition, practices such as trolling, flaming and sham-
ing and the ensuing discussion about appropriate behavior received attention
(e.g. Arendholz 2013; Hardaker 2010; Helfrich 2014; Kluge 2014; Perelmutter
2013, 2015; Rentel 2014; Turnage 2007). While not every study on face-ag-
gravating behavior in a computer-mediated context necessarily draws on (im)
politeness theories, the general link to identity construction and facework is
usually a given. For a review of studies on consensual and conflictual disagree-
ment in online contexts, see Bolander and Locher (Ch. 22, this volume).
The overview of (im)politeness studies on computer-mediated data draws a rich

picture of interpersonal strategies that can be used for face-maintenance, face-en-
hancing and face-aggravation. Screening the texts with respect to epistemology
and methodology, it becomes apparent that we find studies adopting a construc-
tivist understanding of politeness and identity; and a leaning toward qualitative

Authenticated
explorations, although there are also quantitative and mixed-method approaches.

Special journal issues and edited collections (Bedijs, Held and Maaß 2014a; Locher
2010; Locher, Bolander and Höhn 2015) particularly tackle the interface of iden-
tity construction, relational work and (im)politeness studies in CMC. A number of
monographs highlight aspects of relational work in different computer-mediated
contexts (e.g. Arendholz 2013; Dayter 2016; Locher 2006; Yus 2011) and recent
papers and handbooks offer overviews of (im)politeness research and CMC (e.g.
Dynel 2015; Hardaker and Graham 2017; Graham, Ch. 17, this volume).
The contexts studied vary, and include:
– blogs (e.g. Bolander 2012, 2013; Luzón 2013; Perelmutter 2013),
– chats (e.g. Linnemann, Brummernhenrich and Jucks 2014; Vandergriff 2013;
van Compernolle, Williams and McCourt 2011),
– discussion boards and fora (e.g. Angouri and Tseliga 2010; Arendholz 2014;
Ehrhardt 2014; Eisenchlas 2012; Fröhlich 2014; Haugh 2010; Haugh, Chang
and Kádár 2015; Held 2014; Herring 1994; Kleinke and Bös 2015; Kreß 2014;
Kunkel 2014; Maaß 2014; Nishimura 2010; Perelmutter 2015; Planchenault
2010; Placencia 2012; Schrader-Kniffki 2014; Shum and Lee 2013; Thaler
2014),
– email interaction (e.g. Darics 2010; Hössjer 2013),
– mailing lists (e.g. Graham 2007, 2008),
– newspaper comments (e.g. Upadhyay 2010; Langlotz and Locher 2012; Neu
rauter-Kessels 2011, 2013),
– Youtube videos and comments and other polylogues (e.g. Bedijs 2014; Bou-
Franch and Garcés-Conejos Blitvich 2014; Dayter and Rüdiger 2014; Lange
2014; Placencia 2012),
– social network sites (see Section 4 in this chapter, and Eisenlauer, Ch. 9, this
volume),
– and text messaging (Spilioti 2011; Georgakopoulou 2013).
3.2. Gender
As mentioned in Section 2, studies on gender and identity construction have a
long tradition, which has been continued with data from many different comput-
er-mediated contexts (for overviews, see, e.g., Herring 2003; Herring and Stoerger
2014). In their recent review article, Herring and Stoerger (2014: 567) explore
the impact of anonymity on CMC in light of gender and the popular claim that
CMC is “inherently democratic, leveling traditional distinctions”. They report that
online communication has erroneously been perceived as a context where gender
can be neutralized or power differences made to disappear (Herring and Stoerger
2014: 578). The studies they review demonstrate that gender still becomes salient
both with respect to access and use, and the distribution of textual or multimodal
features. The picture, however, is diverse and context-sensitive. Thus, while men

Authenticated
are typically seen as adopting a more aggressive communicative style compared

with women, in the case of asynchronous discussion lists and forums, “Herring
(1996 […]) observed a majority-gender effect: women tend to be more aggressive
in male-dominated groups than among other women, and men tend to be more
aligned in female-dominated groups than in groups dominated by men” (Herring
and Stoerger 2014: 570).
The results are also mixed with respect to language use, e.g., vocabulary, style
and speech acts (Herring and Stoerger 2014: 572). From a discursive perspective
this is not surprising, since gender ideologies are not expected to disappear in com-
puter-mediated contexts, but are instead made salient in varying ways by interact-
ants. As Page’s (2012: 91) study of storytelling in various online, polylogue con-
texts confirms, “gender persists as a meaningful category in computer-mediated
contexts and […] offline values do not disappear in online interaction”. And in
a review on studies of texting, Thurlow and Poff (2013: 166) report that the fact
“[t]hat gender differences emerge in young people’s preferred communication
styles is hardly surprising (Thurlow 2001); these findings do, however, reiterate
the variability that exists between texters and the messages they send”.
With respect to research methodologies employed in studies on gender in com-
puter-mediated contexts, we find the same broad distinctions that also characterize
the research on facework and identity construction in general. There are studies
that are interested in exploring whether and how men and women use language
differently by considering gender a social variable for exploring different comput-
er-mediated contexts, often with a quantified study design (e.g. Chen and Abedin
2014; Herring and Paolillo 2006; Knupsky and Nagy-Nell 2011; Panyametheekul
and Herring 2003). On the other end of the spectrum, we also find qualitative
research on how individuals or groups of individuals negotiate gender. For exam-
ple, Herring’s (1999) two case studies from an Internet Relay Chat channel and a
listserv discussion group show how men and women are treated differently when
disagreeing. This results in what Herring (1999: 151) calls a “rhetoric of har-
assment” against women, which surfaces differently in the two online contexts:
“Whereas female participants on IRC are kicked off the channel, in the discussion
group harassers must rely exclusively on language to intimidate and silence”. A
further example is Lee’s (2011) study of one woman who used status updates while
giving birth (see also Section 4), or Perelmutter’s (2015) work on how Russian
women negotiate appropriate and inappropriate behavior on a forum dedicated
to marital infidelity. By employing acts of shaming addressed to mistresses and
wives, the interactants in Perelmutter’s (2015) research make gender roles salient:
Since the overarching societal norms and expectations of family mores and gendered
behavior in the post-Soviet society are often unclear, these shaming practices help Rus-
sian-speaking women construct and negotiate their identities within a group of peers.
These negotiations integrate individual, group, and societal face concerns. (Perelmutter
2015: 149)

Authenticated
There are also studies that combine quantitative with qualitative analysis and
insights derived from different datasets (e.g. Eisenchlas 2012; Hampel 2015; Page
2012). Planchenault’s (2010, 2015) study on how transvestites construct gender
roles to position themselves in their online communities is a case in point. She
reports how roles and experiences with roles are discussed on a meta-level and
how gender markers in French are used in an intricate combination of acts of
positioning: grammatical female endings visible in orthography in adjectives
or past participles but not always pronounced are combined with the male and
female form of nouns. This allows the posters to create ambiguity and to play with
their identities on a lexical, grammatical level (Planchenault 2010: 93). Another
example is Zhang and Kramarae’s (2014: 66) work on online meta-discussions of
gender norms in the context of Chinese online debates about “Shanghai Metro’s
official ‘dress-code’ warning to women passengers”. The authors use a corpus of
comments as reactions to the initial message posted by Shanghai Metro and com-
plement this with further online sources and the response by Shanghai metro. They
then explore the discursive strategies that the different players employ. This results
in a very heterogeneous, often contradictory and at times ambiguous picture about
how gender and feminism are negotiated online in present-day China.
3.3. Expertise, authenticity and trust: Experts and laypeople

in e-health interaction
The negotiation of facework and the construction of identity is particularly salient
in the field of professional and lay interaction, where expertise, authenticity and
trust are at stake. This can be a concern for information outlets such as newspaper
articles, governmental information websites, websites of political parties, e-cam-
paigns and e-presence of political candidates and e-health communication. In this
review, we focus on e-health practices.
In Locher and Schnurr’s (2017) literature review on health and (im)polite-
ness research, studying e-health practices is considered one of the up-and-coming
themes. Within both interpersonal pragmatics in general, as well as for e-health in
particular, four key concerns emerge:
– The face-threatening potential of many interactions in health contexts;
– The negotiation of roles pertaining to health interaction in dynamic encounters;
– The creation and maintaining of trust and expertise;
– The importance of advice giving, information giving, counselling, etc.
(Locher and Schnurr 2017: 698)
We thus find, for example, a number of studies which explore how interactants
attempt to negotiate roles and identities in order to come across as credible experts
(patients and health practitioners alike) in different online contexts. This includes
research from both a linguistic and psychological perspective (among them, Harri-

Authenticated
son and Barlow 2009; Harvey and Koteyko 2013; Hunt and Koteyko 2015; Koteyko
and Hunt 2016; Locher 2006, 2013b; Locher and Hoffmann 2006; Rudolf von
Rohr 2015; Sillence 2010, 2013; Thurnherr, Rudolf von Rohr and Locher 2016).
One example of this research conducted from a linguistic perspective is Loch-
er’s (2006; cf. also Locher and Hoffmann 2006) study of advice online in the
American Internet advice column “Lucy Answers”. Here, the professional advisor
persona Lucy, who is created by a team of health professionals and presented as an
agony aunt, distinguishes herself by, for example, providing up-to-date researched
factual information on health concerns (e.g. by quoting statistics or referring to
books and other sources), embedding her advice within other discursive moves to
lower its face-threatening character, displaying a sense of humour, not being afraid
to criticize and appearing approachable through her choice of vocabulary (while
not being informal, there is a clear avoidance of medical jargon) (see Locher and
Hoffmann 2006). The importance of embedding advice and thus mitigating it to a
certain extent has also been reported for other online advice practices (e.g. Mor-
row 2006, 2012 and Ch. 24, this volume; Placencia 2012 and references in the next
paragraph).
With respect to peer-to-peer interactions, we can refer to studies by Harri-
son and Barlow (2009), Harvey and Koteyko (2013), Hunt and Koteyko (2015),
Kouper (2010), Koteyko and Hunt (2016), Rudolf von Rohr (2015), Sillence (2010,
2013) and Thurnherr, Rudolf von Rohr and Locher (2016), which all convincingly
demonstrate that peers also engage in the creation of expertise and trust. For exam-
ple, in a study of how peers help each other in their journey to quit smoking in an
online forum, Rudolf von Rohr (2015: 264) observes that “[p]articipants who seek
help have to convince other forum members of the authenticity of their claim, while
helpers need to establish their expertise to give advice or emotional support”. In
other words, it is not just the self-selected advice-givers who engage in warranting
strategies to demonstrate their expertise and thus legitimize their role, but also the
advice-seekers, who need to establish their authentic identity as experts of their
personal quitting situation, so that they are “considered help-worthy” (Harvey and
Koteyko 2013: 165). To give another example of the intricate negotiation of roles
in e-health practices, both Harrison and Barlow’s (2009) research on a diabetes
peer-to-peer forum and Thurnherr, Rudolf von Rohr and Locher’s (2016) analysis
of various online e-health sources show how sharing narratives can function as a
resource for giving advice and creating a credible identity as a resource for help.
Research on e-health practices is still developing and to approach it with an
interpersonal pragmatics lens, which is interested in identity construction and
facework, proves promising in light of exploring the themes listed at the begin-
ning of this section. The interest guiding this research can be both applied (e.g.
health practitioners might benefit from insights on how expertise can be indexed
linguistically) and theoretical (e.g. how do peers differ from experts in their use of
strategies for imparting advice and sharing expertise, etc.).

Authenticated
3.4. Further themes
A number of further themes will be mentioned in passing only. The first has to do
with in-group and out-group creation and community building in general. This
is connected to the negotiation of face and the construction of identity since acts
of positioning employed to affiliate and disaffiliate oneself from others influence
emerging identities. For example, many studies on conflict also discuss online
community building and involve meta-discussions of group norms (e.g. Baker
2001; Baym 1996, 1998; Dayter 2016; DuVal Smith 1999; Reid 1999; references
mentioned in Section 3.1 on face-aggravating behavior).
A further strand of research zooms in on the expression of emotions and how
emotional stance cues add to identity construction, the negotiation of interper-
sonal relations and community building (e.g. Maíz-Arévalo and García-Gómez
2013; Langlotz and Locher 2012; Skovholt, Gronning and Kankaanranta 2014).
For example, for their data of Taiwanese parenting fora, Haugh and Chang (2015:
99) report that emotional support can be expressed via both “affiliative responses”
(including “mutual encouraging, mutual bemoaning, and empathic suggesting”)
and “dissafiliative responses” (including “accusing and advising”). These diverse
possibilities of expression highlight the importance of “soliciting emotional sup-
port” as a relational practice in these discussion boards.
To give another example, in their research on emoticons in workplace emails,
Skovholt, Gronning and Kankaanranta (2014: 780) underscore several functions
of emoticons in workplace e-mails, which are not primarily used to index the writ-
ers’ emotions, as one might at first glance expect. Instead, the emoticons func-
tion as “contextualization cues”, which “provide information about how an utter-
ance is supposed to be interpreted” (Skovholt, Gronning and Kankaanranta 2014:
780). More specifically, these include marking “a positive attitude” when they
occur after signatures; functioning as markers of jokes or irony when they follow
utterances intended as humorous; and serving as “strengtheners” after expressive
speech acts and “softeners” after directives (Skovholt, Gronning and Kankaan-
ranta 2014: 780). From a perspective that takes facework and identity construction
into account, exploring emotional stance clearly promises more insights for the
study of relational work.
As a final theme, we would like to mention studies that focus on the intri-
cate interplay between online and offline activities of interactants with respect to
identity construction (e.g. Georgakopoulou 2013; Lee 2011; Mak and Hin Leung
Chui 2014). There are many studies on CMC where no background information is
available on the participants or where studying interactants’ offline communica-
tive behavior would not be possible (e.g. polylogue chatrooms, fora, online games,
etc.). There are, however, studies that focus on the individual and that thus follow
him/her in order to understand better how offline and online acts of communica-
tion intertwine. Georgakopoulou (2013), for example, accompanies young people

Authenticated
during their school days and observes how acts of CMC and information shared in
such acts (e.g. texting) also shape their face-to-face negotiation of identities and
become part of the small stories that they share. And Lee (2011) shows how a Chi-
nese women uses status updates to keep her friends informed about the progress of
her birthing experience (see below). These studies are by nature qualitative.
4. Examples from Facebook and Twitter
To illustrate some of the findings previously discussed, this section reviews a

selection of studies on identity construction on the social networking site (SNS)
Facebook and the microblogging platform Twitter. SNS are “web-based services
that allow individuals to (1) construct a public or semi-public profile within a
bounded system, (2) articulate a list of other users [‘friends’] with whom they
share a connection, and (3) view and traverse their list of connections and those
made by others within the system” (boyd and Ellison 2007). boyd and Ellison
(2007) established the following sequence of appearance of SNS: MySpace in
2003; Flickr and Facebook (the Harvard-only version) in 2004; YouTube and Face-
book (for high-school networks) in 2005; and Twitter and Facebook (for everyone)
in 2006. SNS combine previously separate online ways of communication within
one platform (e.g., posting status updates, sharing photographs and videos, repost-
ing/sharing other users’ material, chatting, commenting on each other’s activities,
playing games and participating in surveys; see, e.g. Herring 2013; Jucker and
Dürscheid 2012; Locher 2014; Thurlow and Mroczek 2011). Characteristically,
the technological affordances of these providers are constantly developing (see,
e.g., Bolander 2017; Eisenlauer 2013 and Ch. 9, this volume; Locher 2014: 558).
As a consequence, the literature on social media interaction always needs to
state clearly how a particular practice worked at the time during which the data
was recorded. This ensures transparency on whether the data is comparable, and
prompts for reflection on the ways it might be more or less comparable given
the change/s in question. It also encourages increased research on the diachronic
development of Facebook practices over time. While such work is scant, Lee
(2011) and Page (2012), for example, include analysis of the shift in 2009 in the
status update prompt from “What are you doing right now?” to “what is on your
mind”; and Koteyko and Hunt (2016) explore health identities on Facebook via a
longitudinal (four-month long) observation of 20 Facebook profiles.
In our own research on identity construction in Facebook (inspired by Zhao,
Grasmuck and Martin 2008), we focused on what people reveal in their profile
pages and in status updates. Our choice to prioritise status updates over other activ-
ities performed by our interlocutors was based on them being the primary activity
in our data collected from two groups of ten students (living in the UK and Swit-
zerland) between December 2008 and January 2009 (Bolander and Locher 2010,

Authenticated
2015; Locher and Bolander 2014, 2015; Locher 2014). Drawing on a previous
study on away messages in Instant Messaging (Nastri, Peña and Hancock 2006),
we analysed what the students write about on a speech act level and what acts of
positioning are evoked within the status updates. We found that the two groups
of students put different emphasis on identities related to personality, pastime
endeavours, sense of humour, work and relationships, and that individuals draw
on and position themselves with respect to these categories to differing degrees.
In the reactions to status updates, we found a predominance of moves that confirm
acts of positioning by the status updates writers and only a handful of challenges.
However, many status updates received no comments at all. We also found that
the ten individuals living in Switzerland draw on more language varieties when
writing their status updates than the UK-based focus group, with multilingualism
thus serving as a resource for identity construction.
While our choice of data (status updates) emerged from the nature of the Face-
book platform at the time of analysis, working with current data warrants a more
multi-modal approach that includes photographs, videos and memes and which
captures interaction (reactions to interactional moves within the platform and also
interaction between offline and online practices) (e.g. Bolander and Locher 2015;
Dayter and Rüdiger 2014; Lee 2011; Locher and Bolander 2014, 2015; Maíz-
Arévalo 2013; Maíz-Arévalo and García-Gómez 2013; Mak and Hin Leung Chui
2014; Page 2012, 2014a; Peña and Brody 2014; Theodoropoulou 2015). In this
vein, Lee (2011) focuses on a female user and author of status updates, who doc-
uments the process of giving birth and thus her transition into motherhood. This
example highlights that the distinction between offline and online life is not ten-
able. In documenting the act of giving birth via Facebook in front of witnesses
(Facebook “friends”), it also highlights a blurring of private and public (see also
Jucker and Landert 2011; Landert 2014).
Our second example of a computer-mediated context is from the microblog-
ging platform Twitter, which started in 2006 and provides its users the opportunity
to post messages that are restricted to 140 characters. As Zappavinga (2012: 3)
succinctly describes, these posts are addressed either to the “general internet” or
a specific set of “followers” who have subscribed to an individual’s “message
stream”; the posts are subsequently displayed in reverse chronological sequence
“as an unfolding stream of content”. Research on identity construction on Twitter
has been growing rapidly since the turn of the decade, with Zappavigna (2012,
2014a, 2014b), Page (2012, 2014a, 2014b), Dayter (2014, 2016) and Rentel
(2014), for example, exploring tweets as instances of microblogging that – in their
cumulative effect over time – add up to acts of identity construction.
Zappavigna (2012: 14) maintains that “people use Twitter and other microblog-
ging services to share their experiences and enact relationships rather than to
simply narrate the mundane details of their activities, as has been claimed in the
popular press”. This enacting of relationships through tweets makes “microblog-

Authenticated
ging […] an ongoing performance of identity” (Zappavigna 2012: 38). For Zap-
pavigna (2012: 38), these processes of identity construction are connected with the
broader aim of connecting with others, or our “human desire for affiliation”. Since
we “exist within communities of other voices with which we wish to connect” our
tweeting practices become bound to, and relative to the community and its prac-
tices of “meaning-making” (Zappavigna 2012: 38).
It is important to stress that Twitter is not a practice that functions in isolation.
As mentioned in the quote above, tweets connect to different actions and activities
and tap into other meaningful repertoires. In her work, Zappavigna (2012) utilises
a corpus of over 7 million tweets (and sub-corpora thereof). Her methodology
draws on systemic functional analysis and corpus linguistic expertise. Throughout
her monograph she identifies and discusses practices that are part of identity per-
formance. For example, she explores the rallying function of hashtags, how evalu-
ative stance is embedded, or how the use of humor, memes and slang serve to fore-
ground certain identities. In Zappavigna (2014a: 140) on tweets about coffee, she
differentiates between “affiliation (personae aligning into communities of value)
and identity (personae enacting particular evaluative dispositions)”. She further
shows how the hashtag or affiliative stance markers in processes of coupling (from
a systemic functional perspective “[ideation: coffee/evaluation: positive appreci-
ation: positive reaction]”, 2014a: 148) are used to “align personae around shared
values” (Zappavigna 2014a: 156). As a consequence, a community of coffee lovers
who use coffee as treat and reward emerges (Zappavigna 2014a: 155). Her study
thus shows how “linguistic strategies available to personae [are employed] in elec-
tronic discourse for construing community” (Zappavigna 2014a: 156).
In her research monograph on microblogging, Dayter (2016: 215) follows a
group of eleven non-professional, yet dedicated and passionate ballet students.
Using a corpus of 1000 tweets, she explores their acts of self-disclosure, achieved
through “self-praise, implicit positive disclosure, third party complaints and per-
sonal narratives”. In studying the pragmatic force of the tweets, Dayter (2016)
reports that there are contradicting repertoires, an “ego repertoire” and a “member
of community repertoire”. Whereas the former serves to confirm the speaker’s
face and contribute to the formation of the “ballet hero image”, the latter serves
to confirm the “others’ face”, such that it signals the desire to be accepted “into
the community” and the willingness to “award social capital to others” (Dayter
2016: 217). Both repertoires are drawn on by individual tweeters. Next to the fact
that the members of her focus group mix strategies individually, another finding
is that self-praise was a frequent and accepted practice (i.e. there was no criticism
of this practice by other tweeters; Dayter 2016: 215, 217), which contributed to
the image of a ballet hero. For example, references to “bloody bunioned feet”
demonstrate in-group knowledge and thus serve to distinguish between “proper”
dancers as opposed to fans; mentioning physical ailments thus translates into proof
of a dedication to ballet (Dayter 216: 134). (For further discussion of Facebook

Authenticated
and Twitter, please consult Zappavigna, Ch. 8 and Eisenlauer, Ch. 9, this hand-
book).
5. Conclusions
There is neither a unified theory nor singular methodological approach to the study
of identity construction in computer-mediated contexts. As shown in the literature
review, the field is vibrant and insights from studies cross-fertilize each other. As a
consequence, we find a rich interface of different approaches, with scholars draw-
ing on previous research conducted within linguistics, but also other neighbouring
disciplines.
In this review, we focused on a number of themes which emerge in research
on identity construction in various modes of CMC. Much attention was paid to
studies of (im)politeness that have adopted online data as their site for research
with scholars approaching identity construction by drawing most explicitly on the
notions of face and facework (or relational work/rapport management, etc.). Three
particular orientations have been shown to be pursued: (a) to further theory build-
ing; (b) to study emic negotiations of norms of conduct in and about CMC; and (c)
to explore face-aggravating behavior in addition to face-maintaining and face-en-
hancing behavior. These also mirror general trends within (im)politeness research.
Studies inspired by research on gender have been shown to constitute an equally
dominant trend, where scholars have underscored, for example, both the pertinence
of gender ideologies about men and women, while also drawing attention to the
ways factors beyond sex influence the ways these men and women interact online.
To illustrate the importance of the construction of expertise, authenticity and trust,
we chose the context of e-health communication, where we addressed some of the
textual and multimodal ways interlocutors position self and other when engaging
in e-health discourse. Finally, the chapter briefly explored the surfacing of emo-
tions, the creation of in- and out-groups and community building, as well as the
intertwining of offline and online acts of positioning.
More research on the study of the construction of identity within different
computer-mediated contexts is clearly called for. We also predict that there will
be increased attention paid to the relationship between online and offline contexts,
and the implications of the blurring of these contexts for online identity practices;
as well as to multimodal acts of positioning.
Acknowledgments
Our heartfelt thanks goes to Aline Bieri, who was instrumental in compiling the
literature reading list and chasing up the texts for our perusal. For feedback on this

Authenticated
chapter, we thank Wolfram Bublitz and Christian Hoffmann. All remaining faults
and ommissions are our own.
References

2010 ‘you HAVE NO IDEA WHAT YOU ARE TALKING ABOUT!’ From e-dis
6(1): 57–82.
Antaki, Charles and Sue Widdicombe (eds.)
1998 Identities in Talk. London: Sage.
Arendholz, Jenny
Relations Amsterdam/Philadelphia: Benjamins.
Arendholz, Jenny
2014 “You sound very talented” – Negotiating face in online message boards. In:
Kristina Bedijs, Gudrun Held and Christiane Maaß (eds.), Face Work and
Social Media, 213–234. Münster: Lit-Verlag.
Arundale, Robert B.
2015 Face constituting theory. In: Karen Tracy (ed.), The International Encyclope-
dia of Language and Social Interaction, 1–9. Hoboken, NJ: John Wiley and
Sons. doi: 10.1002/9781118611463.wbielsi094.
Baker, Paul
2001 Moral panic and alternative identity construction in Usenet. Journal of Compu
ter-Mediated Communication 7(1). doi: 10.1111/j.1083-6101.2001.tb00136.x.
Bamberg, Michael G. W.
1997 Positioning between structure and performance. Journal of Narrative and Life
History 7(1–4): 335–342.
Bargiela-Chiappini, Francesca
2006 Face. In: Keith Brown (ed.), Encyclopedia of Language and Linguistics, 421–
423. Oxford: Elsevier.
Baym, Nancy
1996 Agreements and disagreements in a computer-mediated discussion. Research
on Language and Social Interaction 29(4): 315–345.
Baym, Nancy
1998 The emergence of on-line community. In: Steven G. Jones (ed.), Cybersociety
2.0: Revisiting Computer-Mediated Communication and Community, 35–68.
Thousand Oaks, CA: Sage.
Bedijs, Kristina
2014 Shared face and face-enhancing behaviour in social media: Commenting on
the Spanish goalkeeper’s tears on YouTube. In: Kristina Bedijs, Gudrun Held
and Christiane Maaß (eds.), Face Work and Social Media, 136–156. Münster:
Lit-Verlag.
Bedijs, Kristina, Gudrun Held and Christiane Maaß (eds.)
2014a Face Work and Social Media. Münster: Lit-Verlag.

Authenticated
Bedijs, Kristina, Gudrun Held and Christiane Maaß

2014b Introduction: Face work and social media. In: Kristina Bedijs, Gudrun Held
Lit-Verlag.
Bolander, Brook
Bolander, Brook
Bolander, Brook
2017 Language and identity on Facebook. In: Steven Thorne and Stephen May
(eds.), Encylopedia of Language and Education: Language and Technology,
1–13. New York: Springer.
Bolander, Brook and Miriam A. Locher
2010 Constructing identity on Facebook: Report on a pilot study. In: Karen Junod
and Didier Maillat (eds.), Performing the Self, 165–185. Tübingen: Narr
Francke.
methodological issues. Discourse, Context and Media 3: 14–26.
2015 “Peter is a dumb nut”: Status updates and reactions to them as ‘acts of posi-
tioning’ in Facebook. Pragmatics 25(1): 99–122.
2007 Social network sites: Definition, history, and scholarship. Journal of Comput-
1987 Politeness: Some Universals in Language Usage. Cambridge, UK: Cambridge
University Press.
Bucholtz, Mary and Kira Hall
2005 Identity and interaction: A sociocultural linguistic approach. Discourse Stud-
ies 7(4–5): 585–614.
2008 Finding identity: Theory and data. Multilingua 27(1–2): 151–163.
2010 Locating identity in language. In: Carmen Llamas and Dominic Watt (eds.),
Language and Identities, 18–28. Edinburgh: Edinburgh University Press.
Burke, Kenneth
1969 A Rhetoric of Motives. Berkeley, CA: University of California Press.
Chen, Gina Masullo and Zainul Abedin
2014 Exploring differences in how men and women respond to threats to positive
face on social media. Computers in Human Behavior 38: 118–126.

Authenticated
Darics, Erika
2010 Politeness in computer-mediated discourse of a virtual team. Journal of Polite-
ness Research 6(1): 129–150.
1990 Positioning: The discursive production of selves. Journal for the Theory of
Social Behaviour 20(1): 43–63.
Dayter, Daria
2014 Self-praise in microblogging. Journal of Pragmatics 61: 91–102.
Dayter, Daria
2016 Discursive Self in Microblogging: Speech Acts, Stories and Self-Praise.
Amsterdam/Philadephia: Benjamins.
Dayter, Daria and Sofia Rüdiger
2014 Speak your mind, but watch your mouth: Complaints in CouchSurfing ref-
erences. In: Kristina Bedijs, Gudrun Held and Christiane Maaß (eds.), Face
Work and Social Media, 193–212. Münster: Lit-Verlag.
De Fina, Anna
2010 The negotiation of identities. In: Miriam A. Locher and Sage L. Graham (eds.),
Interpersonal Pragmatics, 205–224. Berlin/Boston: de Gruyter.
Deppermann, Arnulf
2013 How to get a grip on identities-in-interaction: (What) does ’positioning’ offer
more than ‘membership categorization’? Evidence from a mock story. Narra-
tive Inquiry 23(1): 62–88.
DuVal Smith, Anna
1999 Problems of conflict management in virtual communities. In: Marc Smith and
Peter Kollock (eds.), Communities in Cyberspace, 134–163. New York: Rout-
ledge.
Dynel, Marta
2015 The landscape of impoliteness research. Journal of Politeness Research 11(2):
329–354.
Eckert, Penelope
1989 Jocks and Burnouts: Social Categories and Identity in the High School. New
York: Teachers College Press.
Ehrhardt, Claus
2014 Politeness and face work in German forum communication. In: Kristina
Bedijs, Gudrun Held and Christiane Maaß (eds.), Face Work and Social
Media, 84–107. Münster: Lit-Verlag.
Eisenchlas, Susana A.
2012 Gendered discursive practices on-line. Journal of Pragmatics 44(4): 335–345.
Eisenlauer, Volker
Fitzgerald, Richard
2015 Membership categorization analysis. In: Karen Tracy (ed.), The International
Encyclopedia of Language and Social Interaction, 1–11. Hoboken, NJ: John
Wiley and Sons. doi: 10.1002/9781118611463.wbielsi018.
Fröhlich, Uta
2014 Reflections on the psychological terms self and identity in relation to the
concept of face for the analysis of online forum communication. In: Kris-

Authenticated
tina Bedijs, Gudrun Held and Christiane Maaß (eds.), Face Work and Social
Gal, Susan
1992 Dialect variation and language ideology. Paper presented at the 91st Annual
Meeting of the American Anthropology Association, San Francisco, CA.
Georgakopoulou, Alexandra
2013 Small stories and identities analysis as a framework for the study of im/polite-
ness-in-interaction. Journal of Politeness Research 9(1): 55–74.
Goffman, Erving
1955 On face work: An analysis of ritual elements in social interaction. Psychiatry
18(3): 213–231.
Goffman, Erving
1959 The Presentation of Self in Everyday Life. New York: Doubleday Anchor Books.
1967 Interaction Ritual: Essays on Face-to-Face Behavior. Garden City, NY:
Anchor Books.
Goffman, Erving
1974 Frame Analysis: An Essay on the Organization of Experience. Cambridge,
Graham, Sage Lambert
2007 Disagreeing to agree: Conflict, (im)politeness and identity in a computer-me-
diated community. Journal of Pragmatics 39(4): 742–759.
2008 A manual for (im)politeness?: The impact of the FAQ in electronic communi-
ties of practice. In: Derek Bousfield and Miriam A. Locher (eds.), Impoliteness
in Language: Studies on Its Interplay with Power in Theory and Practice,
Hall, Kira and Mary Bucholtz
2013 Epilogue: Facing identity. Journal of Politeness Research 9(1): 123–132.
Hampel, Elisabeth
2015 “Mama Zimbi, pls help me!” – Gender differences in (im)politeness in Ghana-
ian English advice-giving on Facebook. Journal of Politeness Research 11(1):
99–130.
Hardaker, Claire
2010 Trolling in asynchronous computer-mediated communication: From user dis
cussions to academic definitions. Journal of Politeness Research 6(2): 215–242.
Hardaker, Claire and Sage L. Graham
2017 Impoliteness and computer-mediated communication. In: Jonathan Culpeper,
Michael Haugh and Dániel Kádár (eds.), Handbook of Linguistic (Im)Polite-
ness, t.b.a. London: Palgrave.
Harrison, Sandra and Julie Barlow
2009 Politeness strategies and advice-giving in an online arthritis workshop. Jour-
nal of Politeness Research 5(1): 93–111.
Harvey, Kevin and Nelya Koteyko
2013 Exploring Health Communication: Language in Action. London: Routledge.
Haugh, Michael
2010 When is an email really offensive?: Argumentativity and variability in evalua-
tions of impoliteness. Journal of Politeness Research 6(1): 7–31.

Authenticated

pation in Public and Social Media Interactions, 99–133. Amsterdam/Philade-
phia: Benjamins.
Haugh, Michael, Wei-Lin Melody Chang and Dániel Z. Kádár
2015 “Doing deference”: Identities and relational practices in Chinese online dis-
cussion boards. Pragmatics 25(1): 73–97.
Held, Gudrun
2014 Figura … or face? Reflections on two sociopragmatic key concepts in the light
of a recent media conflict between Italians and Germans and its negotiation in
Italian internet forums. In: Kristina Bedijs, Gudrun Held and Christiane Maaß
(eds.), Face Work and Social Media, 32–81. Münster: Lit-Verlag.
Helfrich, Uta
2014 Face work and flaming in social media. In: Kristina Bedijs, Gudrun Held and
Christiane Maaß (eds.), Face Work and Social Media, 297–318. Münster:
Lit-Verlag.
Herring, Susan C.
1994 Politeness in computer culture: Why women thank and men flame. In: Mary
Bucholtz, A.C. Liang, Laurel A. Sutton and Caitlin Hines (eds.), Cultural Per-
formances: Proceedings of the Third Berkeley Women and Language Confer-
ence, 278–294. Berkely, CA: Berkeley Women and Language Group.
Herring, Susan C.
1996 Two variants of an electronic message schema. In: Susan C. Herring (ed.),
Computer-Mediated Communication: Linguistic, Social, and Cross-Cultural
Perspectives, 81–106. Amsterdam/Philadephia: Benjamins.
Herring, Susan C.
1999 The rhetorical dynamics of gender harassment on-line. The Information Soci-
ety 15(3): 151–167.
Herring, Susan C.
2003 Gender and power in on-line communication. In: Janet Holmes and Miriam
Meyerhoff (eds.), The Handbook of Language and Gender, 202–228. Malden,
MA: Blackwell.
Herring, Susan C.
2013 Discourse in Web 2.0: Familiar, reconfigured, and emergent. In: Deborah Tan-
nen and Anna M. Trester (eds.), Georgetown University Round Table on Lan-
guages and Linguistics 2011: Discourse 2.0: Language and New Media, 1–25.
Washington, DC: Georgetown University Press.
Herring, Susan C. and John C. Paolillo
2006 Gender and genre variation in weblogs. Journal of Sociolinguistics 10(4):
439–459.
Herring, Susan C. and Sharon Stoerger
2014 Gender and (a)nonymity in computer-mediated communication. In: Susan
Ehrlich, Miriam Meyerhoff and Janet Holmes (eds.), The Handbook of Lan-
guage, Gender and Sexuality, 567–586. Chichester, UK: Wiley-Blackwell.
Hester, Stephen and Peter Eglin
1997 Culture in Action: Studies in Membership Categorisation Analysis. Washing-
ton, DC: University Press of America.

Authenticated
Hössjer, Amelie
2013 Small talk, politeness, and email communication in the workplace. In: Susan
C. Herring, Dieter Stein and Tuija Virtanen (eds.), Pragmatics of Computer-
Mediated Communication, 613–638. Berlin/Boston: de Gruyter.
Hunt, Daniel and Nelya Koteyko
2015 “What was your blood sugar reading this morning?”: Representing diabetes
self-management on Facebook. Discourse and Society 26(4): 445–463.
Irvine, Judith T.
2001 ‘Style’ as distinctiveness: The culture and ideology of linguistic differentia-
tion. In: Penelope Eckert and John R. Rickford (eds.), Style and Sociolinguis-
tic Variation, 21–43. Cambridge, UK: Cambridge University Press.
Joseph, John Earl
2004 Language and Identity: National, Ethnic, Religious. Basingstoke, UK: Pal-
grave Macmillan.
Joseph, John Earl
2010 Identity. In: Carmen Llamas and Dominic Watt (eds.), Language and Identi-
ties, 9–17. Edinburgh: Edingburgh University Press.
2012 The linguistics of keyboard-to-screen communication: A new terminological
Jucker, Andreas H. and Daniela Landert
Kiesling, Scott
2006 Identity in sociocultural anthropology and language. In: Keith Brown (ed.),
Encyclopedia of Language and Linguistics, 2nd ed., 495–502. Oxford: Else-
vier.
Kluge, Bettina
2014 The collaborative construction of an outsider as a troll in the blogosphere of
Latin American immigrants to Quebec, Canada. In: Kristina Bedijs, Gudrun
Held and Christiane Maaß (eds.), Face Work and Social Media, 323–348.
Münster: Lit-Verlag.
Knupsky, Aimee C. and Natalie M. Nagy-Bell
2011 Dear professor: The influence of recipient sex and status on personalization
and politeness in e-mail. Journal of Language and Social Psychology 30(1):
103–113.
Koteyko, Nelya and Daniel Hunt
2016 Performing health identities on social media: An online observation of Face-
book profiles. Discourse, Context and Media 12: 59–67.
Kouper, Inna
2010 The pragmatics of peer advice in a LiveJournal community. Language@Inter-
net 7. http://www.languageatinternet.org/articles/2010/2464.
Kreß, Beatrix
2014 “Hofnarr” and “Bürgerschreck” vs. “kamir-batir” and “barakobamas”: Face
work strategies and stylising in Russian and German online discussion forums.

Authenticated
In: Kristina Bedijs, Gudrun Held and Christiane Maaß (eds.), Face Work and
Kunkel, Melanie
2014 Impoliteness in the negotiation of expert status: Folk linguistic debates in a
French online forum. In: Kristina Bedijs, Gudrun Held and Christiane Maaß
(eds.), Face Work and Social Media, 403–421. Münster: Lit-Verlag.
Labov, William
1972 Sociolinguistic Patterns. Philadelphia, PA: University of Pennsylvania Press.
Landert, Daniela
2014 Personalisation in Mass Media Communication: British Online News between
Public and Private. Amsterdam/Philadephia: Benjamins.
Lange, Patricia G.
2014 Commenting on YouTube rants: Perceptions of inappropriateness or civic
engagement? Journal of Pragmatics 73: 53–65.
Pragmatics 44(12): 1591–1606.
Lee, Carmen
2011 Micro-blogging and status updates on Facebook: Texts and practices. In:
Crispin Thurlow and Kristine Mroczek (eds.), Digital Discourse: Language in
the New Media, 110–128. Oxford, UK: Oxford University Press.
Léon, Jacqueline
2006 Goffman, Erving (1922–1982). In: Keith Brown (ed.), Encyclopedia of Lan-
guage and Linguistics, 97–99. Oxford: Elsevier. http://dx.doi.org/10.1016/
B0-08-044854-2/02587-6.
Linnemann, Gesa, Benjamin Brummernhenrich and Regina Jucks
2014 A matter of politeness? On the role of face-threatening acts in online tutoring.
Locher, Miriam A.
2006 Advice Online: Advice-Giving in an American Internet Health Column.
Locher, Miriam A.
2008 Relational work, politeness and identity construction. In: Gerd Antos and Eija
Ventola (eds.), Handbook of Interpersonal Communication, 509–540. Berlin/
Boston: de Gruyter.
Locher, Miriam A. (ed.)
2010 Politeness and Impoliteness in Computer-Mediated Communication. Special
issue, Journal of Politeness Research 6(1).
Locher, Miriam A.
2013a Politeness. In: Carol E. Chapelle (ed.), The Encyclopedia of Applied Linguis-
tics. Oxford: Wiley-Blackwell. doi: 10.1002/9781405198431.wbeal0916.
Locher, Miriam A.
2013b Internet advice. In: Susan Herring, Dieter Stein and Tuija Virtanen (eds.),
de Gruyter.

Authenticated
Locher, Miriam A.
2014 Electronic discourse. In: Klaus P. Schneider and Anne Barron (eds.), Pragmat-
Locher, Miriam A. and Brook Bolander
2014 Relational work and the display of multilingualism in two Facebook groups.
2015 Humour in microblogging: Exploiting linguistic humour strategies for iden-
tity construction in two Facebook groups. In: Marta Dynel and Jan Chovanec
(eds.), Participation in Public and Social Media Interactions, 135–155.
Locher, Miriam A., Brook Bolander and Nicole Höhn (eds.)
2015 Relational Work in Facebook and Discussion Boards/Fora. Special issue,
Pragmatics 25(1).
Locher, Miriam A. and Sage L. Graham
2010 Introduction to interpersonal pragmatics. In: Miriam A. Locher and Sage
Lambert Graham (eds.), Interpersonal Pragmatics, 1–13. Berlin/Boston: de
Gruyter.
Locher, Miriam A. and Sebastian Hoffmann
2006 The emergence of the identity of a fictional expert advice-giver in an Amer-
ican Internet advice column. Text and Talk 26(1): 69–106. doi: 10.1515/
TEXT.2006.004.
Locher, Miriam A. and Andreas H. Jucker (eds.)
2017 Pragmatics of Fiction. Berlin/Boston: de Gruyter.
Locher, Miriam A. and Stephanie Schnurr
2017 (Im)politeness in health settings. In: Jonathan Culpeper, Michael Haugh and
Dániel Kádár (eds.), Palgrave Handbook of Linguistic (Im)Politeness, 689–
711. London: Palgrave.
9–33.
2013 “This is an erroneous argument”: Conflict in academic blog discussions. Dis-
course, Context and Media 2(2): 111–119. doi: 10.1016/j.dcm.2013.04.005.
Maaß, Christiane
2014 On the markedness of communication on online message boards as part of
a perception-oriented politeness approach. In: Kristina Bedijs, Gudrun Held
Lit-Verlag.
Maíz-Arévalo, Carmen
2013 “Just click ‘Like”’: Computer-mediated responses to Spanish compliments.
Journal of Pragmatics 51: 47–67. doi: 10.1016/j.pragma.2013.03.003.
2013 ‘You look terrific!’ Social evaluation and relationships in online compliments.
Discourse Studies 15(6): 735–760. doi: 10.1177/1461445613490011.
Mak, Bernie Chun Nam and Hin Leung Chui
2014 Impoliteness in Facebook status updates: Strategic talk among colleagues ‘out-

Authenticated
side’ the workplace. Text and Talk 34(2): 165–185. doi: 10.1515/text-2013-
0042.
Mendoza-Denton, Norma
2002 Language and identity. In: J. K. Chambers, Peter Trudgill and Natalie Schilling-
Estes (eds.), Handbook of Language Variation and Change, 475–499. Oxford,
UK: Blackwell.
Mendoza-Denton, Norma
2008 Homegirls: Language and Cultural Practice among Latina Youth Gangs.
Malden, MA: Blackwell.
Meyerhoff, Miriam
2006 Prestige, overt and covert. In: Keith Brown (ed.), Encyclopedia of Language
and Linguistics, 2nd ed., 77–79. Oxford, UK: Elsevier.
Morrow, Phillip R.
2006 Telling about problems and giving advice in an Internet discussion forum:
Some discourse features. Discourse Studies 8(4): 531–548.
Morrow, Phillip R.
Mullany, Louise
2012 Discourse, gender and professional communication. In: James Paul Gee and
Michael Handford (eds.), The Routledge Handbook of Discourse Analysis,
509–522. London: Routledge.
Nastri, Jacqueline, Jorge Peña and Jeffrey Hancock
2006 The construction of away messages: A speech act analysis. Journal of Compu
ter-Mediated Communication 11(4): 1025–1045.
Neurauter-Kessles, Manuela
2011 Im/polite reader responses on British online news sites. Journal of Politeness
Research 7(2): 187–214.
Neurauter-Kessles, Manuela
2013 Impoliteness in cyberspace: Personally abusive reader responses in online news
media. Ph.D. dissertation, University of Zurich. http://opac.nebis.ch/ediss/
20131752.pdf.
Nishimura, Yukiko
2010 Impoliteness in Japanese BBS interactions: Observations from message ex-
changes in two online communities. Journal of Politeness Research 6(1):
33–55.
Ochs, Elinor
1993 Constructing social identity: A language socialization perspective. Research
Panyametheekul, Siriporn and Susan C. Herring
2003 Gender and turn allocation in a Thai chat room. Journal of Computer-Medi-
ated Communication 9(1). doi: 10.1111/j.1083-6101.2003.tb00362.x.
Page, Ruth
2012 Stories and Social Media: Identities and Interaction. London: Routledge.
Page, Ruth
2014a Hoaxes, hacking and humour: Analysing impersonated identity on social net-
work sites. In: Philip Seargeant and Caroline Tagg (eds.), The Language of

Authenticated
Social Media: Identity and Community on the Internet, 46–64. Basingstoke,

UK: Palgrave Macmillan.
Page, Ruth
2014b Saying ‘sorry’: Corporate apologies posted on Twitter. Journal of Pragmatics
62: 30–45. doi: 10.1016/j.pragma.2013.12.003.
Peña, Jorge and Nicholas Brody
2014 Intentions to hide and unfriend Facebook connections based on perceptions of
sender attractiveness and status updates. Computers in Human Behavior 31:
143–150. doi: 10.1016/j.chb.2013.10.004.
Perelmutter, Renee
2013 Klassika zhanra: The flamewar as a genre in the Russian blogosphere. Journal
of Pragmatics 45(1): 74–89. doi: 10.1016/j.pragma.2012.10.006.
Perelmutter, Renee
2015 Shaming, group face, and identity construction in a Russian virtual community
for women. In: Marina Terkourafi and Jonathan Culpeper (eds.), Interdiscipli-
nary Perspectives on Im/politeness, 149–180. Amsterdam/Philadephia: Benja-
mins.
Placencia, María Elena
2012 Online peer-to-peer advice in Spanish Yahoo!Respuestas. In: Holger Limberg
and Miriam A. Locher (eds.), Advice in Discourse, 281–305. Amsterdam/Phil-
Planchenault, Gaëlle
2010 Virtual community and politeness: The use of female markers of identity and
solidarity in a transvestites’ website. Journal of Politeness Research 6(1):
83–103.
2015 Voices in the Media: Performing French Linguistic Otherness. London:
Bloomsbury Academic.
Reid, Elizabeth
1999 Hierarchy and power: Social control in cyberspace. In: Marc A. Smith and
Peter Kollock (eds.), Communities in Cyberspace, 107–133. London: Rout-
ledge.
Rentel, Nadine
2014 Linguistic strategies for the realisation of face work in Italian tweets. In: Kris-
tina Bedijs, Gudrun Held and Christiane Maaß (eds.), Face Work and Social
Rudolf von Rohr, Marie-Thérèse
2015 ‘You will be glad you hung onto this quit’: Sharing information and giving
support when stopping smoking online. In: Catherine Arnott-Smith and Alla
Keselman (eds.), Meeting Health Information Needs Outside of Healthcare:
Opportunities and Challenges, 263–290. Waltham, MA: Chandos/Elsevier.
Sacks, Harvey
1992 Lectures on Conversation, Vol. II. Ed. by Gail Jefferson. Oxford: Blackwell.
Schrader-Kniffki, Martina
2014 Subject emergence, self-presentation, and epistemic struggle in French lan-
guage forums. In: Kristina Bedijs, Gudrun Held and Christiane Maaß (eds.),
Face Work and Social Media, 376–401. Münster: Lit-Verlag.

Authenticated
Shum, Winnie and Cynthia Lee

2013 (Im)politeness and disagreement in two Hong Kong Internet discussion forums.
Journal of Pragmatics 50(1): 52–83. doi: 10.1016/j.pragma.2013.01.010.
Sillence, Elizabeth
2010 Seeking out very like minded others; exploring trust and advice issues in an
online health support group. International Journal of Web Based Communities
6(4): 376–394.
Sillence, Elizabeth
2013 Giving and receiving peer advice in an online breast cancer support group.
Cyberpsychology, Behavior, and Social Networking 16(6): 480–485. doi: 10.1089/
cyber.2013.1512.
Skovholt, Karianne, Anette Gronning and Anne Kankaanranta
2014 The communicative functions of emoticons in workplace e-mails: :-). Journal of
Computer-Mediated Communication 19(4): 780–797. doi: 10.1111/jcc4.12063.
Spencer-Oatey, Helen
2007 Theories of identity and the analysis of face. Journal of Pragmatics 39(4):
639–656.
Spilioti, Tereza
2011 Beyond genre: Closings and relational work in text-messaging. In: Crispin
New Media, 67–85. Oxford: Oxford University Press.
Steiner, Peter
1993, July 5 On the Internet, nobody knows you’re a dog. The New Yorker [Cartoon].
Sundén, Jenny
2003 Material Virtualities: Approaching Online Textual Embodiment. New York:
Peter Lang.
Sunderland, Jane and Lia Litosseliti
2002 Gender identity and discourse analysis: Theoretical and empirical considera-
tions. In: Lia Litosseliti and Jane Sunderland (eds.), Gender Identity and Dis-
course Analysis, 1–39. Amsterdam/Philadephia: Benjamins.
Swann, Joan
2002 Yes, but is it gender? In: Lia Litosseliti and Jane Sunderland (eds.), Gender
Identity and Discourse Analysis, 43–67. Amsterdam/Philadelphia: Benjamins.
Thaler, Verena
2014 Negative evaluation and face work in French and Italian online comments.
2015 Politeness on Facebook: The case of Greek birthday wishes. Pragmatics 25(1):
23–45.
Thurlow, Crispin
2001 Talkin’ ’bout my communication: Communication awareness in early adoles-
cence. Language Awareness 10(2–3): 213–231.
Thurlow, Crispin and Kristine Mroczek
2011 Introduction: Fresh perpespectives on new media sociolinguistics. In: Crispin
New Media, xix-xliv. Oxford: Oxford University Press.

Authenticated
Thurlow, Crispin and Michele Poff

2013 Text messaging. In: Susan Herring, Dieter Stein and Tuija Virtanen (eds.),
de Gruyter.
Thurnherr, Franziska, Marie-Thérèse Rudolf von Rohr and Miriam A. Locher
2016 The functions of narrative passages in three written online health contexts.
In: Daria Dayter and Susanne Mühleisen (eds.), Personal Narrative Online.
Special issue, Open Linguistics 2(2).
Turnage, Anna
2007 Email flaming behaviours and organizational conflict. Journal of Comput-
Upadhyay, Shiv R.
2010 Identity and impoliteness in computer-mediated reader responses. Journal of
Politeness Research 6(1): 105–127.
Vandergriff, Ilona
2013 Emotive communication online: A contextual analysis of computer-mediated
communication (CMC) cues. Journal of Pragmatics 51: 1–12. doi: 10.1016/j.
pragma.2013.02.008.
van Compernolle, Rémi A., Lawrence Williams and Claire McCourt
2011 A corpus-driven study of second-person pronoun variation in L2 French syn-
chronous computer-mediated communication. Intercultural Pragmatics 8(1):
67–91. doi: 10.1515/iprg.2011.003.
Watzlawick, Paul, Janet Helmick Beavin and Don D. Jackson
1967 Pragmatics of Human Communication: A Study of Interactional Patterns,
Pathologies and Paradoxes. New York: Norton.
Willems, Herbert
2001 Goffman, Ervin (1921–82). In: Paul B. Baltes (ed.), International Encyclope-
dia of the Social and Behavioral Sciences, 6297–6301. Oxford: Pergamon.
Yus, Francisco
2011 Cyberpragmatics: Internet-Mediated Communication in Context. Amsterdam/
Zappavigna, Michele
iation on the Web. London: Continuum.
Zappavigna, Michele
2014a CoffeeTweets: Bonding around the bean of Twitter. In: Philip Seargeant and
Caroline Tagg (eds.), The Language of Social Media: Identity and Community
on the Internet, 139–160. Basingstoke: Palgrave Macmillan.
Zappavigna, Michele
2014b Enacting identity in microblogging through ambient affiliation. Discourse and
Communication 8(2): 209–228. doi: 10.1177/1750481313510816.
Zhang, Wei and Cheris Kramarae
2014 “SlutWalk” on connected screens: Multiple framings of a social media discus-
sion. Journal of Pragmatics 73: 66–81. doi: 10.1016/j.pragma.2014.07.008.

Authenticated
16. Evaluation
Michele Zappavigna
Abstract: This chapter explores linguistic research into evaluation in social media
texts. After a survey of some of the dominant linguistic theories of stance and eval-
uation, introducing the Appraisal Framework as a model that has been used across
a range of social media discourses, it explores work on evaluation in social tag-
ging, focusing on the evaluation functions of hashtags. The chapter concludes by
considering how values are negotiated in social media texts and how alignments
are forged through ambient affiliation.
1. Introduction
Social media texts rarely present bald facts or narrate activities and events without
adopting some kind of evaluative stance, since sharing and contesting opinion and
sentiment is central to social media discourse. Even the most ideationally focused
post implies some form of assessment, having been selected by a social media
user as worthy of being presented to the social network, whether for criticism or
praise. The discursive domain at stake in such sharing is interpersonal meaning, a
region of meaning concerned with adopting evaluative positions in discourse and
enacting identities and social affiliations (Martin and White 2005). This chapter
explores how attitudinal meaning is construed in social media texts. The focus
is on the various ways in which stances are enacted and values negotiated. After
a brief overview of the theoretical space of linguistic work on evaluation, I will
consider research focused specifically on social media discourse.
Social media research outside of linguistics has been preoccupied with the role
that ‘emotion’ (in a broad sense) plays in social media communication across a
range of groups of people and contexts. For example, research, typically relying
on forms of content analysis, has explored how social media networks cluster in
terms of the emotions expressed by users and the tendency of users to engage in
“valence-based homophily” when interacting with people who share their opinions
(Himelboim, McCreery and Smith 2013: 2). Other studies have also considered
how social groups differ in terms of emotional expression (Ritter, Preston and
Hernandez 2014). In media studies, there is currently a concentration of work on
‘affect’, leading to concepts such as “affective publics” (Papacharissi 2015) and
“affective networks” (Shaw 2013: 221) centered on how feelings of belonging in
relation to new media. Within health research there has been work on how neg-
ative emotion talk might impact on dimensions of health such as heart disease
DOI 10.1515/9783110431070-016
Authenticated
risk (Eichstaedt et al. 2015: 164), support during cancer (Myrick et al. 2016), and
mental health (De Choudhury et al. 2013). The expression of public opinion has
been seen as relevant to crisis communication (Shaw et al. 2013), and political dis-
course (Bruns et al. 2015). However, while all of this research clearly presupposes
some conceptualisation of evaluative meaning as a dimension of communication,
specific linguistic parameters remain undefined or underspecified due to the disci-
plinary orientation of this work.
The most widely used type of quantitative study of evaluative discourse and
social media is computational work known as “sentiment analysis” (Pang and Lee
2008). It draws on linguistic theory to varying extent, often favouring a ‘bag of
words’ approach that, by focusing on lexis alone, factors out a range of interper-
sonal meanings at clause level and above. Sentiment analysis aims to improve the
automatic detection of evaluative features in corpora by applying various forms
of machine learning to social media texts. It has been preoccupied with how to
use these approaches to ‘mine’ social media as a datasource for exploring public
‘mood’, particularly in relation to predicting change in areas in which shifts in
public opinion are critical, such as voting (Ceron et al. 2014; Tumasjan et al. 2010)
and stock market fluctuation (Bollen, Mao and Zeng 2011).
This chapter begins by providing a brief overview of linguistic theories of
evaluation, focusing on interpersonal pragmatics, the Appraisal Framework (Mar-
tin and White 2005), and coupling theory (a theory of how evaluation and expe-
rience come together to form values). It then surveys pragmatic approaches to
social media that are focused on the role of evaluative language. Because many
studies which have an evaluative focus have been in the area of social tagging
(e.g. hashtag use), three subsections are devoted to this domain. Finally, because
forging social relationships through facilitating interaction and building commu-
nities is the main aim of social media, the chapter concludes by introducing work
on “ambient affiliation” (Zappavigna 2011, 2012, 2014a) that aims to understand
how language works to enact relationships.
2. Linguistic theories of evaluation
Research into evaluative language has become an important area of inquiry within
linguistics despite the extent to which the affective functions of grammatical and
discursive structures have tended to be underestimated by linguists for many
decades (Martin 1992). Historically, interpersonal meaning has been marginal-
ised in linguistics (Poynton 1990). The reasons are both practical and political.
Interpersonal meaning is difficult to study because it is mostly prosodic in nature
and thus not reducible to constituent components (Martin 1996; Halliday 1979).
Since defining a constituent structure is a prerequisite for the kinds of counting of
features needed for statistical analysis, prosodic meanings have tended to evade

Authenticated
Evaluation 437
quantification. The Chomskian hegemony entrenched in the 1960s also resulted in

a structural hierarchy in linguistics which tended to privilege “the cognitive and
individual over the interpersonal and social” (Poynton 1990: 8).
Nevertheless, linguistic research into the nature of evaluative language is cur-
rently a burgeoning area. Most studies now recognize that “beyond the function
of communicating referential information, languages are responsive to the fun-
damental need of speakers to convey and assess feeling, moods, dispositions and
attitudes” (Ochs and Schieffelin 1989: 9). Attitudinal meanings have variously
been viewed as dimensions of “stance” (Biber and Finegan 1989; Hunston and
Thompson 2000; Hunston 2007), “appraisal” (Martin and White 2005; White
2002) and “evaluation” (Bednarek 2008). Most approaches attempt to taxonomise
or systematise evaluative meaning using strategies such as defining “evaluative
parameters” (Bednarek 2008) or mapping relevant discursive resources (Martin
and White 2005) which will be explored in the following section. In particular,
there has been a concentration of corpus-based work on evaluation (Conrad and
Biber 2000; Bednarek 2006, 2008; Hunston and Thompson 2000; Hunston 2011).
2.1. Interpersonal pragmatics

Work within pragmatics on evaluation has largely arisen out of the focus on the
interpersonal dimensions of communication originating in Leech’s (1983) study of
interpersonal rhetoric as well as Brown and Levinson’s studies (1987) of polite-
ness in interactions. Haugh, Kádár and Mills (2013) provide a useful overview
of this history in terms of the issues and debates arising out of related work into
pragmatic perspectives on interpersonal communication. They note Locher and
Graham’s (2010: 1) definition of interpersonal pragmatics as being about investi-
gating how “social actors use language to shape and form relationships in situ”. In
their overview, they also point out that such concern with exploring relationality
in communication and interaction is seen in other disciplines (e.g. anthropology)
and methodologies (e.g. ethnography).
Locher (2014) has surveyed pragmatic approaches to computer-mediated com-
munication, of which social media communication might be seen as a sub-regis-
ter. Her overview notes the difficulties that have been faced in finding a term for
describing language use mediated by computer technologies that is not too broad
or too specific:
[…] terms such as “digital discourse”, “electronic discourse”, “e-communication”,
“digitally mediated communication”, etc. are too broad “since it would also include
mass media communication via TV and radio, which is not in the focus of researchers
analyzing language use in the new media” (Jucker and Dürscheid 2012: 40). For this
reason, Dürscheid and Jucker (2012: 40) prefer the term keyboard-to-screen commu-
nication (KSC), which, for example, allows mobile phones and text messages to be
included in the analysis in a more transparent way […]. (Locher 2014: 556)

Authenticated
Herring, Stein and Virtanen’s (2013) edited collection brings together pragmatic
work on computer-mediated communication, focusing largely on email, instant
messaging and blogging. Much of the work surveyed in their 2013 handbook was
focused on “interpersonal metacommunicative” communication such as the use
of emoticons and various kinds of abbreviation (Gruber 2013: 61). As we will see
later in this chapter, such interest has been reignited in relation to social tagging
(e.g. hashtag use) in social media communication.
2.2. The Appraisal Framework

The most comprehensively theorized model of evaluative language that has been
used across a wide range of contexts, and by researchers from both qualitative and
corpus-based traditions, is the Appraisal Framework (Martin and White 2005).
This framework was developed within Systemic Functional Linguistics (SFL) to
account for how evaluative language functions in situational and cultural contexts.
Appraisal has been used in a number of social media studies, including studies of
affiliation (Zappavigna 2015, 2014a, 2014b, 2014c, 2012, 2011), solidarity-build-
ing (Drasovean and Tagg 2015), identity (Vásquez 2014), narrative (Page 2012),
computational sentiment analysis (Vásquez 2014), and social tagging (Chiluwa
and Ifukor 2015; Zappavigna 2015).
As noted earlier, evaluation is a domain of interpersonal meaning where lan-
guage is used to express attitudes and to adopt stances regarding other texts and
voices. Martin and White’s (2005) model, based on Painter’s (1984) work on lan-
guage development, has suggested that the kinds of emotional reaction to the world
expressed through infant protolanguage develops into more complex systems of
meaning-making as we are socialised into a culture and into institutions. Feeling
becomes institutionalized as ethics and morality, forming the judgement 1 system,
with which we construe rules and regulations regarding behaviour (top example,
Figure 1). Feeling is also institutionalized as aesthetics and value, forming the
appreciation system, with which we generate assessments based on our reactions
to phenomena (bottom example, Figure 1).
Before considering each of these discursive systems of evaluation, it is neces-
sary to explain what ‘system’ means within the theory and the particular modelling
strategy that it uses to represent systems, namely the system network. SFL system-
atizes meaning as choice. This systemic orientation arose out of the Firthian tradi-
tion in linguistics (Firth 1957). Firth asserted that a distinction needed to be made
between structure and system, that is, between syntagmatic and paradigmatic rela-
tions in language. Firth’s ideas about how the notion of a system might be used
1
Small caps are used here to differentiate the technical use of these terms from their
everyday counterparts.

Authenticated
Evaluation 439
Figure 1. The institutionalisation of affect, adapted from Martin and White (2005: 45)
to model language were taken up in Halliday’s development of system networks.

System networks are networks of interrelated options that are organised paradig-
matically, in terms of “what could go instead of what”, rather than syntagmatically
in terms of structure (Halliday and Matthiessen 2004: 22). They are an alternative
to modelling language as a catalogue of structures. Figure 2 provides examples of
each region of appraisal (shown in bold) as a system network theorised by Martin
and White (2005). The network adopts the convention whereby capitalised labels
affect Absolutely adore the people I’ve met here so far
ATTITUDE judgment Stupid people amaze me. They are fun. #Idiots
appreciation The linguine looks delicious #mkr
monogloss This is the end of the road.
APPRAISAL ENGAGEMENT
@User Hey you! Saw your Mom the other day, she said you
heterogloss were very happy! Congrats!!
force @User This is an extremely epic tweet!

GRADUATION
focus @User you are truly my hero
Figure 2. The appraisal system, adapted from Martin and White (2005: 38)

Authenticated
in front of the arrows indicate different systems of meaning, and the lowercase
labels at the end of each path mark features within systems. A square bracket rep-
resents a choice between two options in a system (an ‘or’ relation), while a brace
represents simultaneous choices (an ‘and’ relation).
The example tweets appended to this network in the speech bubbles each illus-
trate a type of appraisal . The network may be further specified to greater levels
of delicacy depending on the kind of analysis for which it is being used. Evalua-
tive meaning is clearly a complex semantic domain, and to a certain extent every
utterance we make is tinged with evaluation. The framework attempts to account
for this complexity by distinguishing attitude which is ‘inscribed’, that is, real-
ised via explicit evaluative choices, from attitude which is ‘invoked’, that is,
suggested via ideational choices that imply evaluation.
The discourse semantic region of meaning concerned with adopting stances
is attitude . Appraisal specifies three domains of attitude : affect (expressing
emotion), judgement (assessing behaviour) and appreciation (estimating value)
(Martin and White 2005; see Figure 3). These domains are not intended to simply
act as devices for categorising particular lexical items, but instead for systema-
tising the different semantic regions of evaluation. Depending on the target and
source of the evaluation, lexis such as ‘good’ can construe any of the attitude sys-
tems. Consider the difference between judgement in the following tweet assessing
the morality of a human actor:
Your beliefs don’t make you a good person, your attitude and behavior does.
and affect , expressing an emotion:
Now I know why I don’t feel good.
or appreciation , estimating the aesthetic value of something:
Heading to Covent Garden, where is the best place to get good coffee?
affect Absolutely adore the people I’ve met here so far

ATTITUDE TYPE
judgement Stupid people amaze me. They are fun. #Idiots
appreciation The linguine looks delicious #mkr
Figure 3. The attitude system, based on Martin and White (2005: 38)
The rest of this section will focus on affect , the region of meaning concerned
with “registering positive and negative feelings: do we feel happy or sad, confident
or anxious, interested or bored?” (Martin and White 2005: 42). In microblogging,

Authenticated
Evaluation 441
for instance, it is not uncommon for users to devote an entire post to detailing their
current emotional state. For example, searching for I feel via the Twitter search
interface results in tweets involving emotion talk such as the following (affect
shown in bold):
I feel like I can’t breathe.
Why do I feel so awkward when I talk to you?
I feel Good Today …
I’m wayyyyy up I feel blessed!
I feel really really shitty
Table 1 gives examples of the types of affect Martin and White (2005) specify,
along with illustrative tweets (instances marked in bold).
Table 1. Types of affect , adapted from Martin and White (2005: 51)
Affect Positive Negative

Dis/inclination @User seriously – I long to @User I’m wary to say. Don’t
come back and lounge poolside want you laughing at me.
for five or six days …
Un/happiness @user and I adore you. Can’t shake this melancholy.
Early back to work from the
doctor’s office. I hope work
feels good.
In/security @User has emily been getting @User I have not thought much
more confident with her skat- about anteaters before, and I am
ing? now truly freaked out.
Dis/satisfaction @User Congrats Darius! So fckn sick of takin these nasty
chuffed you made the final! ass pills! ughhhh :(
Good luck for next week!
Zappavigna (2012) has employed this model of evaluation to consider patterns of

attitude in a 100 million word corpus of tweets. In this corpus ‘love’ was the most
common instance of affect , and I love you was the only inscribed affect in the
top twenty 3-grams.2 This pattern was frequently followed by the heart emoti-
con <3 and coupled with a direct @ address. An example is the following, where
@user is the trigger for the positive affect :
Going to bed. Goodnight dear @User ! I love you <3
2
Within the field of corpus linguistics, an n-gram is a set of words that co-occur within
a particular window, n, in the case above a window of 3 words (a trigram).

Authenticated
Further examples are given in the ensuing concordance lines:

@User whooo! have a great i love you !!!:D
show everyone:) kelly
@User I love you lotss and lotsss like jelly tots!! ;-)
xxxxxxxxxx
@User I love you so much girly!! Your the best and
everything will be okay! Your awesome
B!
@User HEHEHEHEHHE! I love you too and think ’bout you all the time.
I can’t wait to see you again! x Night.
@User thankyou for i love you! Xx
following me :D
Other common 3-grams, involving love, are shown in Table 2. A very frequent
target of the second most common 3-gram listed, I love the, was the domain of
music. This pattern was associated with the most frequent hashtag in HERMES,
#nowplaying, used to bond around values of music appreciation (see Section 5 of
this chapter for more on social bonding). Similarly, I love that was usually used to
share a common positive assessment by referencing something in another tweet:
@User I love that show!
Table 2. The top 10 3-grams for ‘love*’ in HERMES
N Cluster Freq. Example

1 I LOVE YOU 14,102 @User thanks sweetie for #FF i love you too :)
2 I LOVE THE 9,786 Is it weird that I love the smell of #windex
3 I LOVE THAT 6,553 @User wow, I LOVE that movie!
4 I LOVE IT 6,134 SOOOOON as UberTwitter stops acting up
I will show yall how crazy this snow is lol I
LOVE IT
5 I LOVE THIS 5,991 #nowplaying We gotta liiiiiiiiiiiiiiive like we’re
dying *.* oh gosh, i love this song!
6 LOVE LOVE LOVE 5,366 @User Omg – the acoustic in that song makes
me weeeeeak! I love love love it too!
While the Appraisal Framework provides a comprehensive model for analysing

the various forms of evaluative language that are possible, evaluation does not of
course occur without reference to experience. Thus in order to account for how
values are formed in discourse we need to consider how appraisal enters into rela-
tionships with particular forms of ideation, that is particular topics or themes. This
is the focus of the next section.

Authenticated
Evaluation 443
2.3. Coupling: Construing evaluations about experience
Attitudinal stances are always associated with particular targets in order to con-
strue value positions in discourse, in other words, when we make evaluations about
dimensions of the experiential world we construe evaluations about things. Thus
the Appraisal Framework has been extended in order to consider how evaluations
‘couple’ with ideation as we share values and form communities (Knight 2013,
2010b). According to Knight (2010a), social bonds are negotiated via couplings
of ideational and interpersonal meaning in discourse. She comments on these cou-
plings as follows: “Specifically, couplings realize bonds of value with experience
linguistically, as bonds are on a higher order of abstraction in the socio-cultural
context. We discursively negotiate our communal identities through bonds that we
can share, and these bonds make up the value sets of our communities and culture,
but they are not stable and fixed” (Knight 2010a: 43).
Knight’s model of affiliation is based on the idea that social alignments are
realized in discourse as patterns of ‘couplings’ of ideation and experience. For
instance Apple pie tastes amazing involves a coupling of ‘apple pie’ and positive
appreciation that we might notate as:
[evaluation: positive appreciation/ideation: apple pie]
The / symbol is used to suggest that evaluation and ideation are ‘fused’ together
to create a value that can be negotiated by interlocutors. For example, Knight has
noted the role of unfolding prosodies of attitude produced in dialogic interaction in
the communication of “three conversational participants [who] commune around a
bond realised by a coupling of intensified positive appreciation for a pie party that
they regularly participate in together” (Knight 2010b: 219):
Table 3. Positive evaluation of a pie party in a causal conversation

(adapted from Knight 2010b: 219)
Speaker Talk Coupling

A Party’s over. For me;
B No more pie party ==
C == This was an awesome pie party [evaluation: positive appreciation/
guys ideation: pie party]
B I love pie parties [evaluation: positive t-appreciation/
ideation: pie party]
C I can’t wait to have another one [evaluation: positive t-appreciation/
ideation: pie party]

Authenticated
According to Knight’s model, ideation-evaluation couplings when shared, con-

tested, or ‘laughed off’ establish a ‘bond’ between interlocutors (Knight 2013).
3. Pragmatic approaches to social media communication
There has recently been a concentration of work adopting a pragmatic perspective

on the language of social media. Some of this work is surveyed in a 2014 spe-
cial issue of the Journal of Pragmatics ‘The Pragmatics of Textual Participation
in the Social Media’ (for the introduction to this special issue see Bou-Franch
and Garcés-Conejos Blitvich 2014a). Most of this work does not position itself
as being specifically focused on evaluative language, though many of the studies
do deal with dimensions of interpersonal meaning such as social media ‘sharing’
practices (Androutsopoulos 2014: 4), the management of conflict (Bou-Franch and
Garcés-Conejos Blitvich 2014b; also Bolander and Locher, Ch. 22, this volume),
the interactive dimensions of YouTube participation (Dynel 2014; also Johansson,
Ch. 7, this volume) and ‘ranting’ as a social media genre (Lange 2014: 55; also
Hardaker, Ch. 18, this volume). Most of these studies develop issues arising out of
the body of work associated with Brown and Levinson’s study (1987) of the nego-
tiation of face in interactions (cf. also Graham, Ch. 17, this volume). For example,
Zhang and Kramarae (2014) use the concept of face to interpret gendered dis-
courses about women’s dress in microblogging in China, suggesting that the rise of
capitalism in this country has resulted in the reinvention for partriarchal ideology
and the redeployment of these ideologies on social media platforms. Some studies
extend this work into investigation of metadiscourse about politeness itself. For
example, Sifianou (2015) has investigated how the concept of politeness is dis-
cussed on Twitter (cf. also Zappavigna, Ch. 8, this volume).
The area of social media research with a pragmatic orientation that has most
directly focused on evaluative language has been work on how interpersonal
meanings are made through social tagging. Thus this is the focus of the subsec-
tions that follow.
3.1. Social tagging and evaluation

Evaluation can be realized in social media texts directly in the ‘body’ of a post
via the systems of attitude explored above. It can also be realized through semi-
otic resources that are more specific to social media channels, for example, forms
of ‘re-posting’ such as retweeting (Puschmann 2012), and social tagging prac-
tices, such as the use of hashtags. Hashtags have developed important interper-
sonal functions associated with the expression of attitudinal stance (Zappavigna
2012, 2015). These functions include contributing to “conversational style” (Scott
2015), with emotive and emphatic use linked to variations in illocutionary force of

Authenticated
Evaluation 445
the kind achieved through paralanguage in offline conversation (Wikström 2014).

They may also function as meta-pragmatic markers, for instance, playing a role
in negotiating levels of self-praise in practices of visual self-representation (Mat-
ley forthcoming). In addition, hashtags have been viewed as brief “spontaneous”
interjections with increased emotive denotative power compared with the body of
a post (Cislaru 2015: 466). In summary, hashtags, while a noted resource for indi-
cating the topic of a post, have been seen as significant to construing opinion and
sentiment in social media discourse.
Some work has suggested that hashtags may provide clues for retrieving mean-
ings that would perhaps have been realized paralinguistically in offline discourse
through communicative modes such as body language. This kind of perspective,
that likens social media texts to conversational interactions, suggests that eval-
uative hashtags often act as forms of interjections (e.g. #ahah, #haha). How-
ever, according to Cislaru (2015: 466), most “emotion hashtags are intertextually
acquired and reproduced as a topic, or as a marker of group affiliation” rather than
as a spontaneous expression of emotion.
A large body of work in this area is computational, exploring how to leverage
evaluative hashtags to improve the performance of automated detection and clas-
sification of opinion and emotion in tweets (Mohammad and Kiritchenko 2015;
Davidov, Tsur and Rappoport 2010). For instance, Simeon and Hilderman (2015:
169) attempt to automatically classify ‘sentiment and non-sentiment bearing’
hashtags so that the former may be used in computational processing. Examples
provided by the authors of what they refer to as hashtags ‘sentiment information’
include the following: “#goodluck”, “#enjoy”, “#wellplayed”, and “#worldcup-
fever”. There has been considerable interest in this area in the semantic domain
of sarcasm and irony, in particular in computational linguistics where automati-
cally detecting sarcasm is a difficult computational problem. Some studies have
explored how hashtags can help distinguish sarcastic microblogs. For instance,
Kunneman et al. (2015) collected a large corpus of tweets featuring the hashtags
#sarcasm and synonymous tags. This study found that posts containing these kinds
of tags tended to include fewer additional communicative features for signaling
sarcasm such as exclamations and intensifiers, and suggested that hashtags may
function in a “digital extralinguistic equivalent” to paralinguistic expressions seen
in face to face interaction for conveying sarcasm (Kunneman et al. 2015: 500).
3.2. Hashtags as interpersonal punctuation

Hashtags are an important resource for foregrounding metadiscourse, with the #
symbol functioning as a form of punctuation marking the point in the text where
the metadiscourse begins, for example:
Don’t be mad at me, I tried, which is a hell of a lot more than you did. #stupid

Authenticated
While the exact target of the evaluation in the hashtag above is not directly retriev-
able (does it refer to the whole situation or to the person addressed in the post?),
we might liken it to a muttered sottovoce comment in spoken discourse, express-
ing the speaker’s negative opinion of the interlocutor. The tag concurrently seems
to operate as a form of interpersonal emphasis akin to the kinds of intensification
that might be achieved typographically via capital letters or punctuation such as
exclamation marks. This effect is most obvious when the hashtag itself is an inten-
sifier upscaling the attitudinal meaning realised in the body of the post as in the
following:
These are the reasons I dislike Wachusett #very strongly.
Hashtags are very flexible in terms of structural position. They can either occur as
an adjunct to the group, clause, or clause complex constituting the main content
of a post, or, alternatively, integrate themselves seamlessly into that content. In
terms of positioning within syntactic structure, the # symbol may mark tags at any
point within a post, though it is most commonly found at either the beginning or
end of a post (Tsur and Rappoport 2012: 2). Tsur and Rappoport (2012) term these
three possible locations (middle, beginning, end) infix, prefix and suffix positions.
The hash symbol may also mark multiple hashtags in a sequence at each of these
locations, without any intervening content between the tags. In suffix position tags
are typically not integrated into the clause but are appended at the end of the post.
These are most often tags marking the ‘Theme’ of a post, but may also be interper-
sonal tags realizing ‘New’ information as in the following example:3
Off to start my third job in this lovely rain! #nothappy
In this type of instance the tag has a prosodic resonance (Hood 2010: 141), in other
words, the scope of the meta-evaluation enacted in the hashtag extends across the
whole text rather than a single constituent unit. In this particular example the target
of the negative emotion (nothappy), is the entire clause that precedes it rather than
a particular unit within the clause.
3.3. Hashtags as evaluative metacommentary

In analogy with spoken discourse, we may also think of the interpersonal mean-
ing realised by evaluative metadiscourse in social media discourse as the user’s
“ongoing intrusion into the speech situation … his [or her] perspective on the
exchange” (Halliday 1979: 66–67). Hashtags are an important resource for con-
3
‘Theme’ relates typically to information at the beginning of a clause, and ‘New’ relates
to information about what is contained in the ‘Theme’. For an explanation of this tex-
tual pattern cf. Halliday and Matthiessen 2014: 88–133.

Authenticated
Evaluation 447
struing what might be thought of as a kind of metadiscursive ‘conversational

aside’, for instance a meta-complaint:
Ordered some new heels and they don’t even look like what they did on the website!
#nothappy
literally catching every red light right now #itsoneofthosedays #itsokimjamming
Just found out you have to run the beep test in gym this year #youhavetobekiddingme
my dad calls for a workbee every fucking day #wtfistheretoclean #itsalreadyclean
The metacomments realized in the hashtags shown above operate at a different

order of experience to the main content of the post. For example, the hashtag in
the first example (#nothappy) is a meta-appraisal of the entire situation described
in the main body of the post, namely the experience of purchasing a product online
that does not match its description.
In terms of the different relationships that can be instantiated between an eval-
uative hashtag and the main body of a post, there appear to be three main choices:
instances where a hashtag supplies attitude that is not directly presented in the
body of a post, supports, through elaboration or extension, existent attitude co-in-
scribed in the body, or challenges the stance presented; cf. Figure 4:
Figure 4. Relationships between an evaluative hashtag and the main body of a post
An example of a hashtag supplying attitudinal meaning that could not be retrieved

by inspecting the content of the main body of a post itself is the following:
Looking at your pics …#amazing
The body of this post does not give any clue to the user’s orientation toward the
pictures referenced, and instead the opinion about these images is inscribed in the
hashtag. Alternatively, a tag may support evaluation already present in the post:
If I get mad at someone can easily shut that person out of my life without any second
thoughts..#ihateitbutithappens

Authenticated
In this example the negative affect (mad) in the body of the post is echoed in the
hashtag (hate). A hashtag may also challenge the evaluation presented, typically
to invoke humour or sarcasm:
So love it when my twitter gets hacked #Not #NoBueno
These options are presented in Figure 3 with further examples. In terms of textual
meaning, that is, how interpersonal meaning is organised by the textual metafunc-
tion, these hashtags operate in the New position in the clause, adding additional
information to the Theme realised in the main body of the post.
4. Enacting community: aligning around values
Up until this point we have considered patterns of evaluation in social media texts.
However, evaluation is also an important resource for establishing and maintain-
ing the alignments that are central to community membership. In terms of general
internet research on communication, the basic notion of online community was
popularised in Rheingold’s (1993) work on ‘virtual community’ which asserts that
communities emerge when discourse is produced “with sufficient human feeling,
to form webs of personal relationships in cyberspace” (Rheingold 1993: 5). Since
the emergence of this definition, often criticised for its vagueness, there has been
a debate surrounding which criteria establish the bounds of an online community,
the structure of such community, and how communities are built or emerge (see for
example Wellman 2001; Hagel and Armstrong 1997; Jones 1997; Burnett 2000;
Herring 1996, 2004, 2008). No stable definition of online community has prevailed.
In addition, whether analysis of linguistic function and structure of discourse
can serve as evidence for defining communities is an emergent area of inquiry.
Approaches to affiliation within sociolinguistics and discourse analysis have used
concepts such as speech communities (Labov 1972; Bolinger 1975), discourse
communities (Swales 1990; Nystrand 1982) and communities of practice (Wenger
1998; Meyerhoff 2008) to explore linguistic patterning and social structure. These
concepts have resounded with the interest of computer-mediated communication
(CMC) studies in online community formation (for an overview see Androut-
sopoulos 2006, who has summarised some of the main tendencies in CMC in this
respect).
Using linguistics as a lens on community means that we are using semiotic evi-
dence to group instances of meaning-making, rather than simply using contextual
speaker variables such as age or geography. In other words, this kind of perspec-
tive considers “the degree of connection, communality, or rapport [that is] indi-
cated through the interlocutors’ linguistic choices” (White 2008: 567). Evidence
of this kind is important to ensuring that the classification is more than a folk
categorisation. As Herring (2004) notes much research into CMC tends towards

Authenticated
Evaluation 449
generalisation rather than applying a particular strategy for empirically analysing

the evidence about the online behaviour studied:
Internet research often suffers from a premature impulse to label online phenomena in
broad terms, e.g., all groups of people interacting online are “communities”; the language
of the Internet is a single style or “genre”. Notions such as “community” and “genre” are
familiar and evocative, yet notoriously slippery, and unhelpful (or worse) if applied in-
discriminately. An important challenge facing Internet researchers is thus how to identify
and describe online phenomena in culturally meaningful terms, while at the same time
grounding their distinctions in empirically observable behavior. (Herring 2004: 338)
Some language-based research into social media communities has begun by com-
putational linguists in the area of sentiment analysis, mentioned at the beginning
of this chapter. This work attempts to automate the detection of language patterns
for categorising groups of users into social networks (Haythornthwaite and Gruzd
2007; Gruzd 2009). Nguyen et al. (2010: 23) propose “grouping patterns of com-
munities purely from their sentiment”. This study used both mood tags, a form
of metadata applied to blog posts by their authors, and “emotion-bearing words”
(Nguyen et al. 2010: 26), extracted from the content of posts, to classify sentiment.
When attempting to model how communities form in discourse, it is worth
noting, however, that perspectives developed from studies of dialogic commu-
nication (e.g. face to face casual conversation) cannot be directly applied, since
“[d]igitally mediated communication is identical to neither speech nor writing,
but selectively and adaptively displays properties of both […] it does things that
neither of these other mediums do” (Crystal 2010: 235). In addition, social media
discourse displays features of monologue, dialogue and broadcast communication,
despite often simply being termed a form of public ‘conversation’. Most defini-
tions of conversation presuppose some version of adjacency pairing usually based
on Sacks, Schegloff and Jefferson (1974), occurring in dialogic interaction, the-
orised as either sequences of ‘turns’ or as more elaborated exchange structures
(Martin 2000; Eggins and Slade 2005). However, it has long been noted that dig-
ital communication can produce turn-taking that is very different to face to face
discourse, since how turns unfold or overlap “is dictated by the software, and not
by the participants” (Crystal 2010: 238). This is one of the many reasons that the
notion of conversational dialogue cannot be directly ported from studies of offline
interaction to social media communication. Just as dimensions of social media
communication, such as social tagging, afford meaning potential not possible in
dialogue, conversational situations such as chat at the dinner table draw on mean-
ings afforded by body language and the material context in ways not possible
online. However, it should also be noted that although software may predefine the
unfolding and overlap possibilities of turns in digital communication (to a certain
degree, at least), the participants still have a bearing on how they choose to pro-
duce and unfold their turns and even some sort of overlaps can be staged on certain
forms of digital communication.

Authenticated
From the point of view of functional linguistics and semiotics, the key thing to
look at in order to explore how values are negotiated in social media is interper-
sonal meaning. However, in social media communication, users do not necessarily
know who will interact with the material they post to the social stream. This issue
has been flagged within media studies as the problem of an ‘imagined audience’
(Marwick and boyd 2011; Litt and Hargittai 2016). This has been described as the
potential audience that a social media user envisages they are addressing when the
actual or material audience is not visible/traceable: “Given the various ways peo-
ple can consume and spread tweets, it is virtually impossible for Twitter users to
account for their potential audience, let alone actual readers […] Without knowing
the audience, participants imagine it” (Marwick and boyd 2011: 117).
The need to ‘imagine’ the audience is a result of what this research terms “con-
text collapse” (Wesch 2009). Context collapse, according to this perspective, refers
to the impossibility of exhaustively assessing and attuning to the various param-
eters of a given context because the audience is largely unknown. For instance,
a vlogger recording a YouTube video is under considerable semiotic pressure to
regulate themselves impersonally despite the indeterminacy of their interaction
with their potential audience:
The problem is not lack of context. It is context collapse: an infinite number of contexts
collapsing upon one another into that single moment of recording. The images, actions,
and words captured by the lens at any moment can be transported to anywhere on the
planet and preserved (the performer must assume) for all time. The little glass lens [of
the webcam] becomes the gateway to a black hole sucking all of time and space—vir-
tually all possible contexts—in on itself. (Wesch 2009: 23)
Beyond asking users who they think they are talking to, with all the attendant prob-
lems of ‘extracting’ tacit knowledge that such an approach incorporates, social
media research has made little progress in terms of understanding how language
is operating interpersonally to support this user-audience dynamic. Part of this
problem may be the very idea that context is stable in other forms of communica-
tion, rather than dynamic and emerging. Any public or mass media communication
integrates a range of meanings from different contexts to which any particular
audience may not necessarily have unmediated access. It is nevertheless possible
for a media text to target specific groups, audiences, reader-/viewerships through
modulating particular communicative choices (e.g. see Clark and Brennan’s 1991
work on the negotiation of common ground between discourse participants).
Rather than framing the question of the imagined audience in such literal terms
we suggest that it may in fact be a question of how social media texts themselves
construe a putative addressee through their linguistic patterning and how they
‘convoke’ (Zappavigna forthcoming-a) an audience to align around particular cou-
plings. In other words, we need to consider how these texts position potential
respondents, in the sense of both potential individual interlocutors and broader

Authenticated
Evaluation 451
communities of interest, through the linguistic resources they deploy. This involves
exploring affiliation that is “ambient” (Zappavigna 2011), rather than necessarily
dialogic, involving communing around rather than necessarily directly negotiat-
ing particular couplings in a clearly defined conversational exchange (Zappavigna
forthcoming-b). Work in this area is ongoing within social semiotic, pragmatics,
sociolinguistics and elsewhere.4
5. Conclusion
This chapter has surveyed the emerging area of research into evaluative language
and social media. It has detailed the Appraisal Framework (Martin and White
2005), a significant model of evaluation that has been adopted in social media
research into stance across a range of contexts. Beyond taxonomising evaluative
meanings, I have attempted to highlight the issue of how evaluation is associated
with ideational targets in discourse when forming particular values (referred to as
“couplings” by Knight 2010a: 35). Beyond this issue is the additional question
of how values are positioned as negotiable or ‘bondable’ in social media texts, in
other words, how they are presented as open to the assessment of the social stream.
This is a question of the discursive processes involved in “ambient affiliation”
(Zappavigna 2011, 2012).
What is at stake when attempting to theorise and analyse affiliation in rela-
tion to social media is how users draw on evaluation as an interpersonal resource.
Social media affiliation is often ambient in the sense that it occurs across platforms
where users may share values without necessarily interacting directly in the dia-
logic sense we as linguists are much more familiar with when analysing offline
dialogue. The challenge is to explain how social media texts ‘convoke’ or ‘call
together’ communities by both construing particular values and proposing particu-
lar orientations to attitudinal stances.
References
2006 Introduction: Sociolinguistics and computer–mediated communication. Jour-
nal of Sociolinguistics 10(4): 419–438.
4
See, for example, forthcoming work on social tagging in an upcoming special issue of
Discourse, Context and Media.

Authenticated
Bednarek, Monika
2006 Evaluation in Media Discourse: Analysis of a Newspaper Corpus. London:
Continuum.
Bednarek, Monika
2008 Emotion Talk Across Corpora. London: Palgrave Macmillan.
Biber, Douglas and Edward Finegan
1989 Styles of stance in English: Lexical and grammatical marking of evidentiality
and affect. Text – Interdisciplinary Journal for the Study of Discourse 9(1):
93–124.
Bolinger, Dwight
1975 Aspects of Language. 2nd ed. New York: Harcourt Brace Jovanovich.
Bollen, Johan, Huina Mao and Xiaojun Zeng
2011 Twitter mood predicts the stock market. Journal of Computational Science 2
(1): 1–8.
2014a The pragmatics of textual participation in the social media. Journal of Prag-
matics 73: 1–3.
2014b Conflict management in massive polylogues: A case study from YouTube.
versity Press.
Bruns, Axel, Gunn Enli, Eli Skogerbø, Anders Olof Larsson and Christian Christensen
2015 The Routledge Companion to Social Media and Politics. New York/London:
Routledge.
Burnett, Gary
2000 Information exchange in virtual communities: A typology. Information
Research 15(4). http://www.informationr.net/ir/5-4/paper82.html.
Ceron, Andrea, Luigi Curini, Stefano M. Iacus and Giuseppe Porro
2014 Every tweet counts? How sentiment analysis of social media can improve our
knowledge of citizens’ political preferences with an application to Italy and
France. New Media and Society 16(2): 340–358.
Chiluwa, Innocent and Ifukor Presley
2015 ‘War against our Children’: Stance and evaluation in #BringBackOurGirls
campaign discourse on Twitter and Facebook. Discourse and Society 26: 267–
296. doi: 10.1177/0957926514564735.
Cislaru, Georgeta
2015 Emotions in tweets: From instantaneity to preconstruction. Social Science
Information 54(4): 455–469.
Clark, Herbert and Susan Brennan
1991 Grounding in communication. In: Lauren B. Resnick, John M. Levine and
Stephanie D. Teasley (eds.), Perspectives in Socially Shared Cognition, 127–
149. Washington, DC: American Psychological Association.
Conrad, Susan and Douglas Biber
2000 Adverbial marking of stance in speech and writing. In: Susan Hunston and
Geoff Thompson (eds.), Evaluation in Text: Authorial Stance and the Con-
struction of Discourse, 56–73. Oxford: Oxford University Press.

Authenticated
Evaluation 453
Crystal, David
2010 The changing nature of text: a linguistic perspective. In: Wido van Peursen,
Ernst Thoutenhoofd and Adriaan Weel (eds.), Text Comparison and Digital
Creativity, 229–251. Leiden: Brill.
Davidov, Dmitry, Oren Tsur and Ari Rappoport
2010 Enhanced sentiment learning using twitter hashtags and smileys. Proceedings
of the 23rd International Conference on Computational Linguistics, Beijing,
August 2010, 241–249. Association for Computational Linguistics. http://
dl.acm.org/citation.cfm?id=1944594.
De Choudhury, Munmun, Michael Gamon, Scott Counts and Eric Horvitz
2013 Predicting depression via Social Media. Proceedings of the 2013 AAAI Confer-
ence on Weblogs and Social Media, Boston 2013, 128–137. Association for the
Advancement of Artificial Intelligence. http://course.duruofei.com/wp-con-
tent/uploads/2015/05/Choudhury_Predicting-Depression-via-Social-Media_
ICWSM13.pdf.
Drasovean, Anda and Caroline Tagg
2015 Evaluative language and its solidarity-building role on TED.com: An appraisal
and corpus analysis. Language @ Internet 12. http://oro.open.ac.uk/id/eprint/
44604.
Dynel, Marta
ics 73: 37–52.
Eggins, Suzanne and Diana Slade
2005 Analysing Casual Conversation. London: Equinox.
Eichstaedt, Johannes, Hansen Andrew Schwartz, Margaret L. Kern, Gregory Park, Darwin
R. Labarthe, Raina M. Merchant, Sneha Jha, Megha Agrawal, Lukasz A. Dzi-
urzynski and Maarten Sap
2015 Psychological language on Twitter predicts county-level heart disease mortal-
ity. Psychological Science 26(2): 159–169.
Firth, John Rupert
1957 Papers in Linguistics 1934–1951. London: Oxford University Press.
Gruber, Helmut
2013 Mailing list communication. In: Susan Herring, Dieter Stein and Tuija Virta-
nen (eds.), Pragmatics of Computer-mediated Communication, 55–82. Berlin/
Boston: de Guyter.
Gruzd, Anatoliy
2009 Automated discovery of social networks in text-based online communities.
Proceedings of the ACM 2009 International Conference on Supporting Group
Work, Sanibel Island, Florida, 379–380. New York: Association for Comput-
ing Machinery.
Hagel, John and Arthur Armstrong
1997 Net Gain: Expanding Markets through Virtual Communities. Boston, MA:
Harvard Business School Press.
Halliday, Michael
1979 Modes of meaning and modes of expression: types of grammatical structure,
and their determination by different semantic functions. In: David J. Allerton,
Edward Carney and David Holdcroft (eds.), Function and Context in Linguis-

Authenticated
tic Analysis: A Festschrift for William Haas, 57–79. Cambridge: Cambridge

University Press.
Halliday, Michael and Christian Matthiessen
2004 An Introduction to Functional Grammar. 3rd ed. London: Arnold.
Haythornthwaite, Caroline and Anatoliy Gruzd
2007 A noun phrase analysis tool for mining online community conversations. In:
Charles Steinfield, Brian T. Pentland, Mark Ackerman and Noshir Contractor
(eds.), Communities and Technologies 2007, 67–86. London: Springer.
Haugh, Michael, Daniel Z. Kádár and Sara Mills
2013 Interpersonal pragmatics: Issues and debates. Journal of Pragmatics 58: 1–11.
Herring, Susan
1996 Computer-Mediated Communication. Linguistic, Social and Cross-Cultural
Herring, Susan
communities. In: Sasha A. Barab, Rob Kling and James H. Gray (eds.), Design-
ing for Virtual Communities in the Service of Learning, 338–376. Cambridge:
Herring, Susan
2008 Virtual community. In: Lisa M. Given (ed.), Encyclopedia of Qualitative
Research Methods, 920–921. Los Angeles: Sage.
2013 Pragmatics of Computer-Mediated Communication. Handbooks of Pragmatics
9. Berlin/Boston: de Gruyter.
Himelboim, Itai, Stephen McCreery and Marc Smith
2013 Birds of a feather tweet together: Integrating network and content analyses to
examine cross-ideology exposure on Twitter. Journal of Computer-Mediated
Hood, Susan
2010 Appraising Research: Evaluation in Academic Writing. London: Palgrave
Macmillan.
Hunston, Susan
2007 Using a corpus to investigate stance quantitatively and qualitatively. In: Rob-
ert Englebretson (ed.), Stancetaking in Discourse, 27–48. Amsterdam/Phila-
delphia: Benjamins.
Hunston, Susan
2011 Corpus Approaches to Evaluation: Phraseology and Evaluative Language.
London: Routledge.
Hunston, Susan and Geoffrey Thompson
2000 Evaluation in Text: Authorial Stance and the Construction of Discourse.
Oxford: Oxford University Press.
Jones, Quentin
1997 Virtual-communities, virtual settlements and cyber-archaeology: A theoretical
outline. Journal of Computer-Mediated Communication 3(3). http://jcmc.indi-
ana.edu/vol3/issue3/jones.html.
framework. Linguistik Online 56: 35–60.

Authenticated
Evaluation 455
Knight, Naomi
2010a Wrinkling complexity: Concepts of identity and affiliation in humour. In:
Knight, Naomi
2010b Laughing our bonds off: Conversational humour in relation to affiliation. Uni-
versity of Sydney: Department of Linguistics. http://hdl.handle.net/2123/6656.
Knight, Naomi
2013 Evaluating experience in funny ways: how friends bond through conversa-
tional hum. Text and Talk 33(4–5): 553–574.
Kunneman, Florian, Christine Liebrecht, Margot van Mulken and Antal van den Bosch
2015 Signaling sarcasm: From hyperbole to hashtag. Information Processing and
Management 51(4): 500–509.
Labov, William
1972 Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press.
Lange, Patricia
2014 Commenting on YouTube rants: Perceptions of inappropriateness or civic
engagement? Journal of Pragmatics 73: 53–65.
Leech, Geoffrey
1983 Principles of Pragmatics. London/New York: Longman.
Litt, Eden and Eszter Hargittai
2016 The imagined audience on social network sites. Social Media and Society 2 (1).
doi: 10.1177/2056305116633482.
Locher, Miriam
2014 Electronic discourse. In: Klaus P. Schneider and Naomi Barron (eds.), Prag-
matics of Discourse, 555–581. Berlin/Boston: de Gruyter.
Locher Miriam and Sage Graham (eds.)
2010 Interpersonal Pragmatics. Handbooks of Pragmatics 6. Berlin/Boston: de
Gruyter.
Martin, James R.
1992 English Text: System and Structure. Amsterdam/Philadelphia: Benjamins.
Martin, James R.
1996 Types of structure: deconstructing notions of constituency in clause and text.
In: Eduard H. Hovy and Donia R. Scott (eds.), Computational and Conversa-
tional Discourse: Burning Issues – an Interdisciplinary Account, 39–66. Ber-
lin: Springer Verlag.
Martin, James R.
2000 Factoring out exchange: Types of structure. In: Malcolm Coulthard, Janet Cot-
terill and Frances Rock (eds.), Working with Dialogue, 19–40. Tübingen: Nie-
meyer.
Martin, James R. and Peter White
2005 The Language of Evaluation: Appraisal in English. New York: Palgrave Mac-
millan.
imagined audience. New Media and Society 13(1): 114–133.

Authenticated
Mately, David
forthcoming This is NOT a #humblebrag, this is just a #brag”: The pragmatics of self-
praise, hashtags and politeness in Instagram posts. Discourse, Context and
Media, Special Issue on the Discourse of Social Tagging.
Meyerhoff, Miriam
2008 Communities of practice. In: J. K. Chambers and Natalie Schilling (eds.),
The Handbook of Language Variation and Change, 526–548. Hoboken, NJ:
Wiley-Blackwell.
Mohammad, Saif and Svetlana Kiritchenko
2015 Using hashtags to capture fine motion categories from Tweets. Computational
Intelligence 31(2): 301–326.
Myrick, Jessica Gall, Avery E. Holton, Itai Himelboim and Brad Love
2016 #Stupidcancer: Exploring a typology of social support and the role of emo-
tional expression in a social media community. Health Communication 31(5):
596–605.
Nguyen, Thin, Dinh Phung, Brett Adams, Truyen Tran and Svetha Venkatesh
2010 Hyper-community detection in the blogosphere. Proceedings of Second ACM
SIGMM Workshop on Social Media, Firenze, Italy. doi:10.1145/1878151.1878159.
Nystrand, Martin
1982 What Writers Know: The Language, Process, and Structure of Written Dis-
course. New York: Academic Press.
Ochs, Elinor and Bambi B. Schieffelin
1989 Language has a heart. Text – Interdisciplinary Journal for the Study of Dis-
course 9(1): 7–26.
Painter, Clare
1984 Into the Mother Tongue: A Case Study in Early Language Development. Lon-
don: Frances Pinter.
Painter, Clare, James Martin and Len Unsworth
2013 Reading Visual Narratives: Image Analysis of Children’s Picture Books. Lon-
don: Equinox.
Pang, Bo and Lillian Lee
2008 Opinion mining and sentiment analysis. Foundations and Trends in Informa-
tion Retrieval 2(1–2): 1–135.
Papacharissi, Zizi
2015 Affective Publics: Sentiment, Technology, and Politics. London: Oxford Uni-
versity Press.
Poynton, Cate
1990 Address and the Semiotics of Social Relations: A Systemic-Functional Account
of Address Forms and Practices in Australian English. PhD thesis. Sydney:
University, Department of Linguistics. http://hdl.handle.net/2123/2297.
2012 The Pragmatics of Retweeting. A Case Study of Academic Uses of Twitter.
http://de.slideshare.net/coffee001/the-pragmatics-of-retweeting.
Rheingold, Howard
1993 The Virtual Community: Homesteading on the Electronic Frontier. Reading,
MA: Addison-Wesley.

Authenticated
Evaluation 457
Ritter, Ryan S., Jesse Lee Preston and Ivan Hernandez

2014 Happy tweets: Christians are happier, more socially connected, and less ana-
lytical than atheists on Twitter. Social Psychological and Personality Science
5(2): 243–249.
Sacks, Harvey, Emanuel A. Schegloff and Gail Jefferson
1974 A simplest systematics for the organization of turn-taking for conversation.
Language 50(4): 696–735.
Scott, Kate
2015 The pragmatics of hashtags: Inference and conversational style on Twitter.
Shaw, Frances
2013 Emotional investments: Australian feminist blogging and affective networks.
In: Tova Benski and Eran Fisher (eds.), Internet and Emotions, 211–224. Lon-
don: Routledge.
Shaw, Frances, Jean Burgess, Kate Crawford and Axel Bruns
2013 Sharing news, making sense, saying thanks. Australian Journal of Communi-
cation 40(1): 23–39.
Sifianou, Maria
Simeon, Credell and Robert Hilderman
2015 Using combined lexical resources to identify hashtag types. 6th Workshop on
Computational Approaches to Subjectivity, Sentiment and Social Media Anal-
ysis, Lisboa, Portugal, 17 September, 2015, 169–174. Association for Com-
putational Linguistics. http://www.aclweb.org/anthology/W/W15/W15-29.
pdf#page=181.
Swales, John
1990 Genre Analysis: English in Academic and Research Settings. Cambridge:
Tsur, Oren and Ari Rappoport
2012 What’s in a hashtag?: Content based prediction of the spread of ideas in
microblogging communities. Proceedings of the Fifth ACM International
Conference on Web Search and Data Mining. doi:10.1145/2124295.2124320.
Tumasjan, Andranik, Timm O. Sprenger, Philipp G. Sandner and Isabell M. Welpe
2010 Election forecasts with Twitter: How 140 characters reflect the political land-
scape. Social Science Computer Review 29(4): 402–418.
Vásquez, Camilla
2014 ‘Usually not one to complain but …’: Constructing identities in user-generated
online reviews. In: Philip Seargeant and Caroline Tagg (eds.), The Language
of Social Media: Identity and Community on the Internet, 65–90. London:
Palgrave Macmillan.
Wellman, Barry
2001 Physical place and CyberPlace: The rise of personalized networking. Interna-
tional Journal of Urban and Regional Research 25: 227–252.
Wenger, Etienne

Authenticated
Wesch, Michael
2009 Youtube and you: Experiences of self-awareness in the context collapse of the
recording webcam. Explorations in Media Ecology 8(2): 19–34.
White, Peter
2002 Appraisal. In: Jef Verschueren, Jan-Ola Östman and Jan Blommaert (eds.),
Handbook of Pragmatics, 1–27. Amsterdam/Philadelphia: Benjamins.
White, Peter
2008 Praising and blaming, applauding and disparaging – solidarity, audience posi-
tioning and the linguistic of evaluation disposition. In: Gerd Antos and Eija
Ventola (eds.), Handbook of Interpersonal Communication, 567–594. Berlin/
Boston: de Gruyter.
Wikström, Peter
2014 #srynotfunny: Communicative functions of hashtags on Twitter. SKY Journal
of Linguistics 27: 127–152.
Zappavigna, Michele
ety 13(5): 788–806.
Zappavigna, Michele
2012 Discourse of Twitter and Social Media. (Continuum Discourse Series.) Lon-
don: Continuum.
Zappavigna, Michele
2014a Ambient affiliation in Microblogging: Bonding around the quotidian. Media
Zappavigna, Michele
2014b Coffeetweets: Bonding around the bean on Twitter. In: Philip Seargeant and
Caroline Tagg (eds.), The Language of Social Media: Communication and
Community on the Internet, 139 –160. London: Palgrave.
Zappavigna, Michele
2014c Enacting identity in Microblogging. Discourse and Communication 8(2):
209–228.
Zappavigna, Michele
2015 Searchable talk: the linguistic functions of hashtags. Social Semiotics 25(3):
274–291.
Zappavigna, Michele
forthcoming–a #Convocation and ambient affiliation: Social tagging as a resource for
aligning around values in social media. Discourse, Context and Media. Special
Issue on the Discourse of Social Tagging.
Zappavigna, Michele
forthcoming–b ‘had enough of experts’: Intersubjectivity and quotation in social
media. In: Eric Friginal (ed.), Studies in Corpus-Based Sociolinguistics. Lon-
don: Routledge.
sion. Journal of Pragmatics 73: 66–81.

Authenticated
Sage Lambert Graham
Abstract: This chapter will explore the concepts of politeness and impoliteness as
they are enacted in various forms of social media. Using an ‘interpersonal prag-
matic’ approach (Locher and Graham 2010), the focus is on interactive mediated
settings such as vlogs, email, discussion boards and virtual worlds. The chapter
examines the parameters of how we define (im)politeness, the terminology we
employ, and the ways that the digital setting influences the way (im)politeness is
enacted. This is followed by an illustrative case study on online gaming that exam-
ines how impoliteness is identified and challenged by participants while moder-
ators simultaneously attempt to define and control for this type of problematic
behaviour. The chapter concludes with a brief overview of directions for future
research.
1. Introduction
Notions of (im)politeness are present in every interaction; we draw on them as

we communicate to accomplish tasks, establish social connections and construct
our identities in a wide array of settings. We are taught from the youngest ages
how to behave ‘politely’, and throughout our lives these notions evolve and we
evaluate our own and others’ actions and make our communicative choices based
on our ideas about what is or is not (im)polite. And while as children much of this
instruction tends to focus on face-to-face contexts, the role of (im)politeness is no
less present or important in digital interactions. Locher (2010) notes that online
communication tends to develop its own set of communicative norms and prac-
tices due to factors such as the merging of public/private audiences and the multi-
modal capabilities of digital media, and as digitally–mediated communication has
increased and evolved, the ways that we navigate the expectations of (im)polite-
ness and interpret others’ behavior in digital contexts has increasing importance in
our more-and-more-digital world.
It is the goal of this chapter to explore the ways that the context of social media
influences and reflects the ways that (im)politeness as a pragmatic device is con-
ceptualized, negotiated and enacted in digital interaction. That said, it is important
to note that the title of this chapter, Politeness and Impoliteness (in Social Media),
might be used to refer to any number of digital platforms and media channels –
far too many to be addressed here. This chapter will therefore restrict the focus to
platforms that lend themselves to interpersonal (and interactive) pragmatic dis-
DOI 10.1515/9783110431070-017
Authenticated
460 Sage Lambert Graham
course. After addressing aspects of digital communication generally, it will move

to an overview of some important concepts in (im)politeness research, examine
where (im)politeness and digitally-mediated discourse research come together,
and explore the manifestation of this overlap in a case study of one particular
digital arena: online gaming. The chapter concludes with a discussion of future
developments and directions for further research.
2. Important issues
In any exploration of (im)politeness in digital communication (DC),1 there are

a number of influential factors that come into play: the nature of DC itself, our
current understanding of politeness and impoliteness and the roles they play in our
interactions, and the ways that they intersect and interact when they come together.
The next section will therefore address each of these in turn.
2.1. The nature of digital communication

The array of digital platforms available for communication today is staggering,
and is evolving daily, both in terms of capabilities as well as popularity. While
any explanation of ‘current’ digital media is, therefore out-of-date before it can
ever reach the printer, there are some common factors shared by various digital
platforms that affect the nature of communication in digital contexts and how (im)
politeness is created and interpreted. In addition to the fluctuations in the popular-
ity and creation of different media, moreover, the array of digital platforms goes
far beyond the platforms that were common in early CMC research, which focused
largely on email, discussion groups and online role-playing communities such as
MUDs and MOOs.2 For reasons of space, I will focus here on communication that
1
Terms used to refer to communication in digital contexts have evolved along with
technologies, and at this point early terms such as computer-mediated communication
(CMC) are no longer accurate since much digital interaction takes place without a com-
puter. While alternative terms such as computer-mediated discourse analysis (CMDA)
have been proposed, for the purposes of this chapter I will use the term digital commu-
nication (DC) except when referring to previous research that uses other terms or when
referring to contexts that are limited to only interaction via computer.
2
Although not as commonly used as they once were, MUDs (Multi-User Dungeons or
Multi-User Domains) and MOOs (which are related to MUDs but are Object-Oriented
virtual realities) are text-based formats that allow multiple users to interact concurrently
via a common server. These platforms were the basis for many current Role-Playing
games (RPGs) and other multi-player formats where players (who do not necessarily
know one another) come together and play/create through a shared server.

Authenticated
Politeness and impoliteness 461
is mediated via computer within social platforms rather than attempting to address
video, television, news media or a whole range of other mediated communicative
settings; it is important to remember, however, that this only scratches the surface
of the array of communicative acts now taking place via digital channels.
2.1.1. Digital media versus face-to-face interaction

One question that has often been asked is whether, or how much, digital commu-
nication is different from face-to-face interaction (f-t-f) (e.g. Benwell and Stokoe
2006). Locher, Bolander and Höhn (2015) rightfully argue that DC and f-t-f inter-
action are not even truly separable at this point because in many cases “we cannot
clearly split the norms interactants bring to the keyboard with those that emerge in
online interactions” (2015: 9). While it is certainly the case that digital interaction
and f-t-f interaction have become more-and-more intertwined in recent years, there
are still some features that distinguish digital interaction from f-t-f encounters.
One place where this distinction can be seen is in the wide array of tools in dig-
ital interaction that encourage users to evaluate others’ behavior as (in)appropriate
(or (non-)politic to use Watts’ 2005 and Locher and Watts’ 2005 terms, discussed
further below). These tools include buttons that are prominently placed on the
screen that ask us to declare our assessments of others’ behavior via ‘blocking’,
using spam filters, using ‘report abuse’ buttons, and, on the positive end, ‘Like-
ing’, ‘Share-ing’ and marking as ‘Favorite’, through which we are continually
asked to rate others’ communicative acts. If we were to argue that Like-ing indi-
cated a positive (politic) assessment and Block-ing indicated a negative (non-pol-
itic) assessment of another’s behavior, then this would mean that assessments of
(im)politeness (on the (im)politeness ‹—› (non-)politic spectrum) are encoded into
the media themselves via these tools, and there is an implicit notion that these
evaluations are something that can (or even should) be declared openly and con-
tinuously. This is in contrast to face-to-face interactions, where, although we may
be constantly evaluating others’ behavior, we are not asked in every interaction
to ‘click buttons’ to openly declare our (ongoing) assessments. So, although we
may question how different the pragmatic strategies are in communicating via
computer or other digital channels versus face-to-face, most would agree that they
are different at least in some ways (just as having a conversation via telephone is
different than having the same conversation f-t-f).
2.1.2. Paralinguistic tools

In digitally-mediated contexts, we also do not have access to the same paralinguis-
tic cues like intonation, stress, and facial expressions that we have in face-to-face
interactions, and these tools are a critical component in conveying the pragmatic
intent of a given message. When DC first became popular among the general pop-

Authenticated
ulation, there was early adaptation of keyboard symbols meant to approximate

facial expressions in written form (also known as ‘smilies’ or emoticons). These
early adaptations, however, were in large part limited to characters and symbols
available on a QWERTY keyboard, since most devices and platforms were limited
to keyboard-based communication rather than touch-screens with graphic capa-
bilities or real-time audio. There was therefore a limit to the scope of information
they could convey.
In more recent years, there has been an explosion in the development of new
digital platforms and modalities, most of which try to provide tools to compensate
for the lack of paralinguistic cues (Vandergriff 2013). Smartphones, for example,
have new ‘emojis’ added with almost every software update and include not just
facial expressions as the first emoticons did, but also pictures that abbreviate mes-
sages that are too tangential to type or that the users feel should be abbreviated
(examples include country flags, waving hands, zodiac symbols, pictures of rain-
bows and dozens more). These emojis are certainly expressive – they allow users
to convey a great deal of information without having to type long messages, and
take a step toward indicating whether someone intends an utterance as humorous
or serious just as the early emoticons did. The increased number of available emo-
jis, however, has not necessarily simplified our ability to clarify communicative
intent. While users may choose emojis to save time or simply avoid keying text,
there is a price in that there are still ambiguities that cannot necessarily be captured
in a picture. Sending an IM or SMS message with an emoji of a spiderweb (one
of the emojis available on the iPhone 6), for example, would allow a user to avoid
typing, but also could be interpreted in many ways – “I’m tangled up so I’ll be
late”, “I’m in a sticky situation”, “I’m afraid of arachnids”, etc.
It is generally agreed that emojis as substitutes for facial expressions and para-
linguistic cues is imperfect – particularly when we take into account the ambig-
uous use as either emotive (e.g. a winking face to express a humorous intent) or
as a neutral abbreviating device (e.g. a picture of a ‘thumbs-up’ in lieu of typ-
ing ‘okay’), but they are nevertheless an important part of digital interaction and
one that is critical to the interpretation of (im)politeness because, no matter how
imperfect, they help interactants clarify their intentions.
2.1.3. A/synchronicity
In digital communication, speed of response is connected to assessments of (im)
politeness more and more frequently. As transmission times have become faster
through fiber-optic cable and extensive internet availability, a common expectation
is that people will both receive and respond to messages more and more quickly.
To fail to respond within this ever-quickening timeframe is often considered impo-
lite. One factor that affects response time is the relative a/synchronicity of the
communicative medium. Chat platforms, for example, are quite synchronous and

Authenticated
approximate real-time ongoing conversation much more closely than more asyn-
chronous platforms like email or Facebook, where messages are sent or posted and
will be read and replied to at some (undetermined) point in the future. And even
when response-time on these relatively asynchronous channels is comparatively
quick, there is not an expectation that it will be quick (or even instantaneous) every
time for every participant. In terms of impoliteness in fact, asynchronous conflicts
often have a tendency to grow a life of their own and may spiral out of control
before they can be recognized and then offset/controlled (Graham 2007b), whereas
in a synchronous chat, a conflict could be resolved quickly in real time, allowing
everyone to move on and resume the interaction. The longer the conflict extends,
however, the longer it will take to resolve, and this is why a/synchronicity plays a
role in online impoliteness.
2.1.4. Anonymity
Yet another aspect of DC that is frequently discussed is anonymity. It is often
posited that people are more impolite online than would ever be f-t-f, and that this
willingness to behave impolitely/inappropriately (terms that are further discussed
in Section 2.2.1 below) stems from the fact that they can act anonymously and are
therefore free from any negative ramifications that their actions might have (see
Lea, Spears and de Groot 1991 and Vinagre 2008 for further discussion of both the
positive and negative results of anonymity online; also Bös and Kleinke, Ch. 4,
this volume; boyd and Hargiattai 2010; Herring and Stoerger 2014; and Marwick
and boyd 2014 for further discussion of privacy and context in social media).
When using anonymity as a lens through which to examine (im)politeness, it is
important to remember that anonymity online (and its counterpart privacy) is not
universal – it operates on a spectrum, with some platforms affording a greater
degree than others. Nevertheless, it is an important part of how people create and
navigate their behavior (either polite or impolite) in digital settings.
2.1.5. Longevity
In addition to anonymity, DC is also affected by the longevity of any given digi-
tal platform. Users might justifiably ask themselves “Will my previous behavior,
polite or impolite, benign or malicious, discreet or indiscreet, come back to haunt
me?” With scandals such as the Ashley Madison case in 2015 where a website
for people having extramarital affairs was hacked and private information leaked,
users have become more aware of the longevity of what they communicate and
how they communicate it via digital channels. The fear that impolite or inappro-
priate behavior will be resurrected and possibly used against them in the future has
led to the development of social media sites like Snapchat, where digital content is
automatically deleted after a specified time period. The knowledge that our actions

Authenticated
are recorded and may expose our prior behavior must certainly affect our polite or
impolite language choices.
2.1.6. Other motivational factors and goals of digital interaction

In addition to the factors listed above, there are other considerations as well; dig-
ital communication is also influenced by 1) whether the audience is public or pri-
vate, and 2) what the goals of the interaction are (e.g. whether participants want to
establish relationships or accomplish some task).
Public interaction makes notions of (im)politeness potentially less certain and
predictable than private discourse, since the audience may be unknown. Private
communication, since the audience is likely known, may be easier to predict/con-
trol – for interactants may have a greater awareness of the norms and expectations
of their audiences and might therefore be more conscious of meeting those expec-
tations since they will likely be held accountable for the choices they make.
The goals of interaction also have an impact on both the norms and execution
of (im)politeness. In a case where an individual wants to declare a stance on a
particular topic (e.g. “I support my country’s Olympic team!”), the statement is
intended to declare a stance. Discussion may ensue if others respond, but that isn’t
a certainty. In other cases, the goal of interaction might be to establish a connection
with others with similar interests and/or join a community (for further discussion
of community formation online see Schwämmlein and Wodzicki 2012 and Baym
2015). In this case, we might expect communication to attend more closely to the
norms of politeness that exist within the group since the goal might not be accom-
plished if these rules are not obeyed.
2.2. The nature of (im)politeness and (non-)politic action/(in)appropriateness

(Im)politeness is not only a fascinating lens through which to examine human
behavior, it is also a hotly debated topic in pragmatic research, much of which
builds on seminal works such as Brown and Levinson (1978/1987), Leech (1983)
and Lakoff (1973). Some of the major themes and questions that have emerged
from this research involve 1) what the primary goals and focus of (im)politeness
research should be, 2) how (im)politeness should be defined, 3) what terms we
should use to define and explain (im)politeness, 4) what role intent plays in estab-
lishing/identifying (im)politeness, and 5) how (im)politeness intersects with vari-
ables such as language, ethnicity, geography/nationality and gender.
It is far beyond the scope of this chapter to give an overview of several decades
of politeness research, but a large portion of research on the pragmatics of (im)
politeness involves the balance people strike between satisfying their own need
to be autonomous (i.e. not imposed upon) with catering either explicitly to the
wants and needs of others or to societal expectations for what constitutes appro-

Authenticated
priate or desirable (and their counterparts inappropriate and undesirable) behav-

ior (for comprehensive overviews of theoretical approaches see Arendholz 2013;
Garcés-Conejos Blitvich 2010; Kádár and Haugh 2013; Locher 2012).
In addition to the themes listed above, an additional question that has arisen is
whether the classification of utterances as (im)polite should be the primary focus
of exploration or whether the interactional impact should be the focus of attention.
The later point of view posits that research should examine how notions of (im)
politeness affect communication through emergent interaction, how people inter-
pret behaviors as (im)polite and how this affects their communicative practices and
relationships. Watts (2003), Locher (2006) and Locher and Watts (2005) take this
stance in their framework of Relational Work, where interactions are categorized
as (non)-politic, or more or less appropriate for a given situation and how interact-
ants use their assessments of these factors in establishing relationships. According
to this framework, behavior must go above-and-beyond what is simply appropriate
(and therefore unmarked) in order to be classified as polite or impolite. A driver
switching lanes to avoid hitting a cement block that had fallen in the roadway, for
example, is taking a neutral (appropriate/politic) action. The same driver stopping
and moving the block to the side of the road is going above-and-beyond because
s/he considers how hitting the block might negatively affect other drivers and
(being polite) takes extra action to prevent those negative consequences for others.
2.2.1. Terminology and definitions

One of the stickier elements of the ongoing exploration of (im)politeness relates to
the terms we use – these include polite and impolite, rude, appropriate and inap-
propriate and (non-)politic. Many researchers have attempted to tease these terms
apart and identify the characteristics that differentiate them (e.g. Bousfield 2010;
Culpeper 2008; Terkourafi 2008), and much of this research hangs on what Bous-
field calls “the thorny concept of ‘intentionality’” (2010: 111). In other words, a
hearer’s assessment of whether someone means to be offensive/impolite/rude has
an effect on how s/he evaluates and classifies the speaker’s behavior. Building
on Culpeper (2008), Bousfield (2010), for example, makes a distinction between
impoliteness and rudeness by defining impoliteness as intentional and rudeness as
unintentional (2010: 114). Others (e.g. Graham 2008), however, argue that making
a distinction between terms such as impolite and rude is not useful, since we can
never be certain of another’s intent and should therefore focus on the impact of an
utterance (which may however be based in part on the hearer’s interpretation of
the speaker’s intent) in the subsequent interaction as it unfolds.
In addition to the terminological debate about how/whether to distinguish
impoliteness from rudeness, etc., another less-explored consideration is where we
draw lines between behaviors that are (im)polite and ones that index other pos-
itive or negative behaviors. Even if we could agree on definitions for the terms

Authenticated
impolite, inappropriate, offensive and rude, for example, there are also terms that
seem related, but have not quite been examined in the way that (im)politeness and
rudeness have, such as aggressive, violent and harassing (although see Culpeper
2011 for an in-depth discussion of aggression and its relationship to impoliteness).
This is consistent with the spectrum of behavior proposed by Locher and Watts
(mentioned above), in which the labels we apply to specific behaviors are not
clearly differentiated. As we will see in the case study below, in practice, these
distinctions are not always easily made, and, just as (im)politeness is emergent
in interaction, part of that emergence can be the evolution of an interaction from
an exchange that is polite to one that is impolite to one that is angry to one that
is aggressive. If we are to attempt to differentiate these terms, it is critical to ask
ourselves how we intend to draw the lines between what is impolite and what is
aggressive, or when something crosses over from being simply inappropriate to
being outright offensive (and for that matter, does offensive equal rude?). To what
extent can they overlap or co-exist? Where does impoliteness stop and aggres-
sion/anger/violence begin? And perhaps most importantly, does differentiating the
terms matter?
Finally, there is the question of how localized terms of (im)politeness are. Dif-
ferent Communities of Practice (Lave and Wenger 1991) may each have their own
terms that identify behavior as (un)problematic (e.g. toxic, or salty, to choose a
few from online gaming). There may also be different culturally-influenced under-
standings of politeness phenomena. There has been a great deal of research in
recent years, for example, that has argued convincingly that the concept of ‘face’
as it was used by Brown and Levinson (1987) in their seminal work on politeness,
cannot be adequately applied to interactions in Asian contexts where the notion of
face is distinct from the notion of politeness (e.g. papers in the Journal of Polite-
ness Research special issue on face and politeness 8(1)), and any examination of
(im)politeness terminology must take these factors into account.
2.2.2. Predictive models of (im)politeness – FAQs and Codes of Conduct

One of the most hotly debated topics in current research on (im)politeness relates
to first-order versus second-order approaches, which are differentiated largely
through the idea of a predictive model of (im)politeness. A first-order approach
refers to (im)politeness as it is experienced and understood by the actors within an
interaction, while a second-order approach refers to a theoretical scientific exam-
ination that depends on an outsider/objective perspective to outline and predict
(im)politeness. While a comprehensive discussion of the two approaches is well
beyond the scope of this chapter, they are foundational to any examination of (im)
politeness, and (as discussed in Section 4 below) are especially striking in a digital
context where the idea of a predictive model is pivotal to how behavioral regula-
tions are enforced.

Authenticated
Researchers have long debated the values and shortcomings of the first-order
versus second-order perspectives (see Kádár and Haugh 2013 for a comprehensive
overview). Proponents of the predictive model (P2) contend that lay users have
their own understandings of the terms (im)polite which are loose and often too
arbitrary to be meaningful (e.g. Terkourafi 2008; Leech 2014), while researchers
have a more specified and ‘sharpened’ use of the terms which illuminates “statis-
tically observable politeness norms” (Leech 2014: 47). I would argue that taking
a solely second-order perspective, however, questions the abilities of lay persons
to assess (im)politeness in a systematic way and runs counter to practice in digital
settings.
In digital contexts, lay users create predictive tools all the time in the form of
FAQs and Codes of Conduct. Such documents address behaviors that the writers
of the guidelines (who in many cases are likely to be lay users) predict will occur
and outline rules for behavior that is deemed (in)appropriate within the community
in question. It seems a stretch to think that all FAQs, employee manuals, Codes of
Conduct etc. have been written by researchers; I would argue, however, that those
written by lay users are no less predictive.3 Any person (research or lay user) can
attempt to predict behavior and write guidelines to control it, but since norms of
interaction are constantly in flux, any guidelines we write might or might not be
reflective of practice that changes to meet the needs of the interactants. In a digital
environment this is particularly true because change occurs so quickly in modali-
ties and subsequent strategies for interaction.
One response to this line of argument is that documents such as FAQs do
not address (im)politeness explicitly, but instead address a general guideline for
appropriateness. While these manuals may not use the word (im)politeness specif-
ically, however, we can once again draw on Watts’ (2003) perspective that appro-
priate behavior is linked to polite behavior on a spectrum and that these guidelines,
which describe (in)appropriate behavior, therefore implicitly suggest that behavior
that falls outside these norms is (im)polite. From an alternative perspective, I also
would say that these behavioral manuals address conduct that falls in the middle of
Leech’s (2014) (im)politeness spectrum, where there is a neutral center and behav-
ior (positive or negative) moves away from this point as it becomes more polite
in one direction and more impolite in the other. Using either of these frameworks,
FAQs and Codes of Conduct ask that readers strive to maintain this middle point
of neutral behavior while implying that behavior that progresses away from it is
more- or less polite or impolite.
3
It is also important to note here, that ‘predictive’ is not necessarily synonymous with
‘accurate’ (see Graham 2008 for how this notion materializes in FAQs).

Authenticated
2.3. The nature of (im)politeness and (in)appropriateness in digital

communication – where they come together
If we agree with the early observation that digital (im)politeness has at least some
characteristics that differentiate it from f-t-f interaction, then it is worthwhile to
examine where the norms of digital interaction and the pragmatics of (im)polite-
ness come together, and this will be the focus of the following sections (for addi-
tional discussion of how the parameters of digital contexts intersect with (im)-
politeness see Graham and Hardaker 2017).
2.3.1. Guidelines/rules/FAQs
With regard to (im)politeness in digital contexts, there are baseline rules that are in
large part agreed upon by people communicating in this environment. These draw
on principles of Netiquette such as those outlined by Virginia Shea (1994), whose
original guidelines said:
1) Remember the human

2) Adhere to the same standards of behavior online that you follow in real life
3) Know where you are in cyberspace
4) Respect other people’s time and bandwidth
5) Make yourself look good online
6) Share expert knowledge
7) Help keep flame wars under control
8) Respect other people’s privacy
9) Don’t abuse your power
10) Be forgiving of other people’s mistakes
While most guidelines for digital communication address at least some of these
expectations, they are still in flux and are violated by some members of any given
online community while others are restrained by them (Graham 2008). Neverthe-
less, even decades later, these foundational rules are still in play in most digital
encounters.
2.3.2. Digital (im)politeness terminology

Just as noted above in Section 2.2.1, terminology is a focus of much discussion in
research on (im)politeness, and adding a digital environment further expands the
terms used to describe (im)polite digital interactions. While there is digital termi-
nology to describe both polite and impolite behaviors, terms that describe negative
or impolite behaviors receive a great deal of discussion (e.g. trolling, spamming,
flaming and doxing). Hardaker, for example, defines a troller as “someone who

Authenticated
constructs the identity of sincerely wishing to be part of the group in question […]
but whose real intention(s) is/are to cause disruption and to trigger or exacerbate
conflict for the purposes of their own amusement” (2010: 237). Flaming involves
an overt attack on an individual or group. Spamming occurs when individuals
distribute empty or meaningless context to a widespread audience with the goal
of being disruptive. Finally, doxing occurs when a user distributes someone’s
personal contact information or details without that person’s knowledge, thereby
exposing that person to risk of f-t-f harm or harassment (see also Hardaker, Ch.
18, this volume).
These terms vary widely in both meaning and importance across digital con-
texts and communities. While in a book club, a troll might be a serious threat and
quite disruptive, while in an online gaming community spammers and trolls are an
expected and accepted (and in some ways encouraged and sanctioned) part of the
interaction. This shifts the import of what trolling actually means – whether the act
of trolling is defined by the impact it has (i.e. being disruptive) or by its packaging
(e.g. whether it contains racist content, which might or might not be offensive and
disruptive within the community in which it is said).
3. Important trends in digital (im)politeness research
Many of the other chapters in this volume will provide an in-depth exploration of
the pragmatics of interaction in separate modalities. The discussion in this chapter,
however, focuses on providing an overview of (im)politeness and related concepts
in these modalities (rather than an exploration of general pragmatic practice). A
key concept in this exploration is that of interpersonal pragmatics, which Locher
and Graham (2010) note should focus on “the relational aspect of interactions
between people that both affect and are affected by their understandings of culture,
society, and their own and others’ interpretations” (2010: 2, emphasis added). We
could argue that social media has its own culture and that examining it with digital
cultural factors (particularly as they are manifested in different digital contexts) in
mind is a worthy enterprise.
As noted above, the array of digital interactional settings and modalities cur-
rently operating in our society is far too vast to explore every one. For reasons
of space, here I will focus on (im)politeness that is both digitally-mediated and
interactive (and therefore emergent) in some way – and is therefore best suited
to illuminate the relational, interpersonal aspect of pragmatics described above.
Under this umbrella, for example, a YouTube video on cooking would not be an
appropriate focus since it is not in-and-of-itself interactive, while the discussion
about the video via the ‘comments’ section (which would fall under the category
of a bulletin or discussion board) would.

Authenticated
3.1. Email and discussion boards
Email and discussion board interaction is perhaps the largest area of research on
(im)politeness in digital contexts (for further discussion of message boards see
also Arendholz, Ch. 5, this volume). In these types of digital interactions (particu-
larly email), the audience is more likely to be known than in blogs/micro-blogs
and therefore the navigation of (im)politeness may be different than in other con-
texts. With regard to email, recent examinations of (im)politeness in teacher/stu-
dent and workplace interactions include Bolken and Holmgren (2012), Chejnová
(2014) and Merrison et al. (2012), all of which examine email interactions where
the audience is known and relatively small. Other research on email impoliteness
focuses on emails sent to large groups (e.g. Graham 2005, 2007a and Ho 2011)
which are more public and frequently focus more on establishing relationships
and communities rather than accomplishing specific tasks. Haugh (2010) and
O’Driscoll (2013) bridge the public and private arenas in examining private email
exchanges that are then interpreted on a public stage. The variety of influences on
(im)politeness and relational work in these studies illuminates the variability of
this platform and further research is warranted as the email context continues to
evolve in the ways it is used to perform relational work, particularly with regard to
public/private and task/relational intersections.
Discussion fora, like blogs (which will be discussed in greater depth below),
encompass a broad spectrum of media platforms and are often embedded within
other multimodal settings. We would place discussion comments about a YouTube
video, for example, in the category of discussion board posts rather than blogs
because, while the YouTube videos themselves could be classified as vlogs (vid-
eo-blogs), the comments that they generate more closely follow the format of dis-
cussion board postings – discussion about a blog (e.g. Garcés-Conejos Blitvich’s
2013 study on responses to a YouTube broadcast).
While the first linguistic research on discussion boards tended to focus on
community building and interactional practice, a sharp increase in recent research
(far too much to be listed exhaustively here) has incorporated both politeness and
impoliteness as specific research foci. Nishimura (2010) and Planchenault (2010),
for example, draw on politeness theory to examine Discussion Board interactions,
while other research has begun to focus more on impoliteness (e.g. Angouri and
Tseliga 2010; Dynel 2012; Hardaker 2010, 2015; Langlotz 2010; Lorenzo-Dus,
Garcés-Conejos Blitvich and Bou-Franch 2011; Richet 2013; Upadhyay 2010) and
flaming (e.g. Perelmutter 2013) to name a few.
Despite the large number of studies on these types of DC, however, there is
little focus on how the constraints of the medium itself may affect norms of (im)
politeness. Despite (or perhaps because of) the breadth of these formats, continued
research is needed that explores the technology itself and the ways that the con-
straints of the media influence how (im)politeness is accomplished.

Authenticated
3.2. Facebook, Instagram, and other social media
Research on social network sites (hereafter SNSs) such as Facebook and Insta-
gram has grown as these modalities have evolved and increased in popularity (see
also Eisenlauer, Ch. 9, this volume). Bolander and Locher (2015), Mak and Chui
(2014) and Theodoropoulou (2015), for example, explore the ways that identities
are negotiated through status updates and posted messages such as birthday wishes
(see also Locher and Bolander, Ch. 15, this volume). With regard to the ongoing
presence of (im)politeness evaluators as discussed in Section 1, Maíz-Arévalo’s
(2013) study of ‘liking’ is of particular interest since it examines the effects of a/
synchronicity, disembodiment and their relationships with (im)politeness (see also
Maíz-Arévalo, Ch. 21, this volume). These recent studies begin to address how
SNSs operate as a context for relational work, but it is important to remember that,
as is the case with all media, these platforms are constantly changing. Each time
there are changes to the privacy policies on Facebook, for example, the potential
audience for any given posting changes and this, in turn, alters users’ strategies for
navigating (im)politeness norms and performing relational work.
3.3. Short-form messages: Blogs/micro-blogs and SMS/IM

Yet another genre of digital interactions is that of blogs and micro-blogs such as
Instagram, Snapchat and Twitter (see also Heyd, Ch. 6, this volume, and Zappav-
igna, Ch. 8, this volume) and SMS/IM messages. In blogs and micro-blogs, users
post content in the form of pictures, videos, and/or text-based messages without
any clear confirmation of who their audience is, when their messages will be read,
or even if they will receive any response. This uncertainty leaves posters unsure
of how they might anticipate or conform to the (im)politeness expectations of
their audience. As Myers notes, “the vast majority of bloggers get along with just
a few people looking at their work. […] but the readers can be complete strangers
anywhere in the world” (2010: 77). He goes on to say that “Bloggers are often sur-
prisingly careful in how they present their opinions to an audience, using a range
of markers that suggest how they are making this statement” (2010: 95). Research
in this area is still growing as new media platforms continue to develop. In spite
of its prominence, for example, there is relatively little research on abuse and (im)
politeness on Twitter (see, however, Page 2014; Sifianou 2015). Since audience
awareness is critical to the examination of (im)politeness, this is a key factor in
any assessment of how (im)politeness works in this environment.
Yet another form of digital communication is SMS (Short Message Services)
and IM (Instant Messages). Like microblogging systems such as Twitter these plat-
forms produce short-form messages; unlike microblogging sites, however, they
are more likely to target known audiences and relatively small groups rather than
the hundreds or even thousands of unknown individuals who may receive a tweet.

Authenticated
As with other digital platforms already discussed, the ways that people use
SMS/IM have changed and expectations about what is normal (e.g. writing full
words v. using txt-speak) have changed as well. Current researchers, building on
previous research examining the oral/written blend of SMS/IM, have continued
in that trend as the capabilities of the digital platforms have evolved, but research
that specifically focuses on (im)politeness as a component of relational work in the
SMS/IM context is still somewhat lacking.
3.4. Virtual Realities and Virtual Worlds

While originally primarily text-based (e.g. MUDs), Virtual Worlds have evolved
as technology has advanced to include visual components as well. Platforms such
as Minecraft allow users to create computer-generated worlds (e.g. by designing
buildings, adding landscaping, etc. (alone or in collaboration with others) and cre-
ate characters/avatars that can operate within the world(s) they have created. An
offshoot of this type of platform – online gaming – allows users to create avatars
which perform in-game tasks and interact with others in attempting to achieve
them. Chat is a critical feature of interaction in virtual worlds, which are highly
synchronous, though they may be interaction- or goal-oriented.
While use of Virtual Worlds (especially online gaming platforms) is increas-
ing, there is not a great deal of research that examines (im)politeness and relational
work in these contexts. Pojanapunya and Jaroenkitboworn’s (2011) examination
of face in Second Life and Arendholz’s (2014) exploration of the online world The
Student Room are exceptions, but investigations that examine (im)politeness in
greater depth in these contexts would be fruitful.
Similarly, there is increasing interest across disciplines in research into online
gaming (which historically draws on virtual world platforms), but linguistic
research in this area almost all focuses on gaming as a socialization tool for L2
learners. There is very little linguistic research that addresses how players negoti-
ate interaction in highly synchronous and frequently multimodal settings (excep-
tions include Rusaw 2011 and Ensslin 2012), and this would be another fruitful
area for additional exploration.
4. Case study – FAQs and guidelines for polite/impolite/politic behavior
Gaming is an ideal venue for examining (im)politeness in a digital context for

multiple reasons. Gaming is highly prevalent in first-world countries and reaches
across multiple demographic categories. As it is put by Maher (2016): “By the
age of 21, the average young gamer will have logged thousands of hours playing
time. That fact alone makes dichotomies such as ‘digital world’ and ‘real world’
ring false – for many, game-playing is the real world” (emphasis added). In addi-

Authenticated
tion, video games are often associated with aggressive behavior. As Maher (2016)
notes:
Online gamers have a reputation for hostility. In a largely consequence-free environ-
ment inhabited mostly by anonymous and competitive young men, the antics can be
downright nasty. Players harass one another for not performing well and can cheat,
sabotage games and do any number of things to intentionally ruin the experience for
others — a practice that gamers refer to as griefing. Racist, sexist and homophobic
language is rampant; aggressors often threaten violence or urge a player to commit
suicide; and from time to time, the vitriol spills beyond the confines of the game. In the
notorious ‘gamergate’ controversy that erupted in late 2014, several women involved in
the gaming industry were subjected to a campaign of harassment, including invasions
of privacy and threats of death and rape. (http://www.nature.com/news/can-a-video-
game-company-tame-toxic-behaviour-1.19647)
In addition to the fact that, as a digitally-mediated genre, gaming is a setting where

(as noted above) assessments/declarations of (im)politeness are an ever-present
feature of interaction, in online gaming, (im)politeness is also encoded to some
degree into the games themselves – even the most benign-seeming game like ‘bub-
ble pop’ is based on the notion of hunting down bubbles and destroying them (one
assumes that most of us would classify destruction of someone/something as an
act that is at least impolite).
If we accept the premise that online gaming is a productive environment in
which to examine norms of (im)polite behavior, then there are several questions
we might ask: 1) what are the rules of (im)polite behavior in this context?, 2) who
determines what counts as (im)polite (i.e. who defines the rules for appropriate
behaviour), 3) once identified via these rules, how are impolite acts sanctioned,
and 4) how are polite acts rewarded? After a brief description of the data set and
context from which it was collected, I will address each of these questions in turn
below.
4.1. Twitch.tv and the genre of online gaming

Twitch.tv is an internet social video platform that is devoted to the broadcasting
of digital content by individuals worldwide and provides a forum for gamers to
establish communities. According to its website, as of May 2016, Twitch.tv had
over 100+ million unique viewers per month, 1.7+ million unique broadcasters per
month, and 106 minutes watched per person per day. Twitch streams are publically
available and can be watched by anyone, although viewers are invited to ‘sub-
scribe’ (by donating a small fee) to streamers that they particularly enjoy watch-
ing. According to Twitch, people broadcast over their channel(s) “to connect with
friends and fans over a shared love of games” and people watch others on Twitch
because “it is live social video that relies on audio and chat to enable broadcasters
and their audiences to interact about everything from pop culture to life in general

Authenticated
as they game” (https://www.twitch.tv/p/about). The content of streams includes

User Generated Content (e.g. live streams of gamers playing video games), edi-
torial programs, events and conventions with panels and press conferences, and
eSports competitions.4
Streamers use cameras and microphones to broadcast their content, which
either appears ‘Live’ or in some cases is recorded and available for replay. Game
streamers can have anywhere from zero viewers to hundreds of thousands of
viewers at any given time. Moreover, Twitch.tv manages game streams from hun-
dreds of games of all categories –1st person ‘shooter’ games such as Call of Duty,
MMORPGs (Massive-multiplayer online role playing games) such as World of
Warcraft, MOBAs (Multiplayer Online Battle Arenas) such as League of Leg-
ends, Virtual World games such as Minecraft, and even ‘non-digital’ games such
as poker.
For the purposes of this study, I will focus on the MOBA (Multiplayer Online
Battle Arena) game League of Legends. At the time of this writing, League of
Legends was the most-popular game on the market, having surpassed World of
Warcraft, with competitively ranked players on Asian, European and North Amer-
ican servers and millions of players and viewers worldwide.
4.2. League of Legends: Predictive rules and guidelines
4.2.1. The Summoner’s Code

The developer and publisher of League of Legends, Riot Games, makes the game
available for free download and maintains control of all game content. Users are
provided with a base set of characters and resources and can build upon these
resources with points and experience earned within the game. The learning curve
for League of Legends is high – it takes time to build expertise and proficiency
within the game and there are large amounts of resources available to help guide
beginners as they attempt to gain proficiency. In addition to Riot’s official instruc-
tions for getting started, they also provide the following “Summoner’s Code”,
which outlines expectations for behavior while playing the game:
1) Support your team
2) Drive constructive feedback
3) Facilitate civil discussion
4
eSports is a genre of competition that is structured just as traditional sports competitions
are, with video of the game itself, close-up television shots of the competitors’ faces
throughout the progression of the gameplay, and commentators offering their assess-
ments throughout the progression of the games regarding the strategies players choose,
the backgrounds and worldwide rankings and experience of individual players, etc.

Authenticated
4) Enjoy yourself, but not at anyone else’s expense

5) Build relationships
6) Show humility in victory, and grace in defeat
7) Be resolute, not indignant
8) Leave no newbie behind!
9) Lead by example
(emphasis added) http://gameinfo.na.leagueoflegends.com/en/game-info/get-started/sum-
moners-code/
The guide ends with the following dictum: “So remember! Stay positive, remain
calm, and keep to the code!”
Each of the guidelines listed above is followed by a paragraph explaining it in
greater detail, including an illustrative quote by a famous figure. The explanation
for rule #6 – Show humility in victory and grace in defeat – for example, begins
with a quote by Benjamin Franklin: “To be humble to superiors is duty, to equals
is courtesy, to inferiors is nobility”. Rule #5 – Build Relationships – includes the
observation that “it would behoove you to adopt a cordial demeanor and attempt
to make friends”. This (sometimes overly flowery) language seems to attempt to
create a tone of refinement that presupposes an attendance to the ‘wants and needs
of others’ that Brown and Levinson (1978/1987) identify as one of the key com-
ponents of politeness.
A quick overview of these guidelines also gives the impression of this game
and the game environment as overwhelmingly positive and polite. The guidelines
are filled with positive lexicon (support your team, build relationships, facilitate
civil discussion) and language that encourages empathy and support for other
players (show humility in victory and grace in defeat, leave no newbie behind),
and behave in a polite way (engage in civil discussion, don’t enjoy yourself at any-
one else’s expense, don’t be indignant). The fact that half of these guidelines use
positive (rather than negative) phrasing, I would argue, immediately swings the
tone to the ‘polite’ end of the polite/impolite spectrum (if we want to adopt Leech’s
2014 framework). By focusing on all of the positive things that players should do,
the game designers propose a code of conduct that is implicitly polite by being
respectful of other players within the confines of the gaming world.
4.2.2. Twitch.tv Code of Conduct

A second layer of behavioral guidelines is provided by Twitch.tv, the broadcaster
for the gaming streams. Twitch views its guidelines as a ‘living document’, provid-
ing frequent updates. The core content of each of their updated versions, however,
is the same – the key differences involve the tone (earlier editions tended to be
more playful and sarcastic/flippant, while later versions tend to be more brusque
and legalistic) and the level of detail provided. Twitch states explicitly that the

Authenticated
behaviors that could lead to sanctions “fall under a common sense philosophy
and apply to all video content, chat messages, channel content, profile content,
usernames, and any other content on our services” (https://www.twitch.tv/p/rules-
of-conduct).
In contrast to the Summoner’s Code described above, however, the Twitch.tv
guidelines adopt an approach that outlines behaviors that are NOT allowed –
instead of telling players what they should do (as in the Summoner’s Code above),
the Twitch rules tell participants what they should not do. Their guidelines are
more vague than the League of Legends code, and leave more interpretive license
to the user to determine what behaviors are (in)appropriate. In addition to legal
rules related to copyrights, rebroadcasting without permission, nudity, and other
similar legal violations, the Twitch behavioral rules (as of August 1, 2016) also
include rules against:
– Targeted harassment, threats, and violence against others, and
– Hate speech or other harassment
The more specific guidelines published in May 2015 described this this way:
We’re not going to tell you to watch the potty mouth, that’s between you and your
mother who doesn’t approve anyway. What we will tell you, however, is that if you
choose to use language or produce content that is racist, sexist, homophobic or falls
under generally accepted guidelines for hate speech you will disappear from Twitch.
(https://www.twitch.tv/p/rules-of-conduct)
The guidelines as amended in November 2015 simply say this: “Any content that
promotes or encourages discrimination, harassment, or violence based on race,
ethnicity, gender identity, sexual orientation, age, religion, or nationality is pro-
hibited”. The guidelines also prohibit “Spam, scams and other malicious content”.
Unlike The Summoner’s Code, the guidelines for behavior on Twitch focus on
what NOT to do, (rather than what to do) and tend to refer to actions with legal
ramifications such as harassment, threats and violence (although as noted in Sec-
tions 2.2.1 and 2.3.2, it is difficult to draw lines as to where something ceases to
be impolite and begins to be hateful or harassing).
The shift from the more casual description of (in)appropriate behavior from
May 2015 to the more legalistic guidelines in November 2015 is also noteworthy,
since it represents a shift from guidelines (suggestions?) to enforceable rules that
will have legal consequences. Those in charge have gone from “we’re not going
to tell you …” (implying that you should know better and should act appropriately
on your own) to “we’re telling you”.

Authenticated
4.2.3. Streamers’ Rules
A third layer of guidelines for appropriate behavior consists of rules created by the
broadcasters themselves. Twitch allows, indeed encourages, streamers to create
their own individual rules for their streams according to their comfort levels and
their assessments of their audiences. While not all streamers post detailed rules
for their viewers on their Twitch home pages, Twitch suggests that they at least
communicate their expectations to their moderators so that the mods know what
should be deleted/prohibited and what should be allowed.
One League of Legends streamer with a strong following on Twitch.tv posts
the following guidelines on his page:
Da Rulez
– No backseat gaming of any kind
– Don’t be a dick
– Don’t advertise anything. This includes other streams or your own personal business
– No racism
– No spreading false rumors
– No flaming
– Don’t one man spam
– Don’t fight with other viewers
– No ascii pictures (Especially the D)
– Don’t tell me how to run my stream. If you want to give constructive feedback do so
on my social medias
– Negative remarks may be timed out or banned at the mod’s discretion
Xanbot will automatically time out for the following:
– Messages that contain too many characters, don’t know the amount
– All caps
– Spamming twitch emotes (I think more than 5 on the same line)
– Colored messages
– Hyperlinks
(http://www.twitch.tv/wingsofdeath)
These rules are in alignment with the Twitch.tv rules, in that they forbid disruptive,
illegal, and hateful behaviors (i.e. ‘remember the human’). Advertising, posting
ascii pictures, and the specific types of posts that the Xanbot (a robot moderator)
is programmed to delete (all caps, messages that contain too many characters, too
many Twitch emotes, colored messages and hyperlinks) fall under the category of
standard spam in that, when posted over-and-over in quick succession they disrupt
the flow of the chat and disrupt the flow of the conversation. Other behaviors listed
above such as the rule against racism align with Twitch’s policy. The no flaming,
no spreading false rumors, don’t fight, no negative remarks, and the less-specific,
but much more descriptive “don’t be a dick” rules ban negative behaviors that
disturb the cohesiveness of the community by creating divisiveness.
Another streamer takes a more succinct approach, providing the following:

Authenticated
General Guidelines:
– No sexism, racism, homophobia of any kind
– No promotion of other channels, no slandering of other channels
– No spoilers5 of any kind (yes, this includes haha just trollin bro it was a fake spoiler)
– Respect the moderators. That means why are you arguing with the mods about your
rights to post porn in twitchchat? Hint: No, Michael is not going to unmod them
because you got banned. Yes, your ban length may be increased.
– Moobot is in charge of auto banphrase moderation. Moobot is a bot. He only has
powers to timeout. Nothing personal … kid.
(https://www.twitch.tv/imaqtpie)
Whereas the Summoner’s Code highlights polite/positive behaviors that focus on

protecting the game experience of the players, and the Twitch guidelines focus on
protecting the legal responsibilities of the organization, both of these streamers
bring these two together and provide guidelines that are designed to meet the legal
demands of Twitch (hence references to no advertising and spoilers) while also
creating a positive environment for their viewers in order to enhance their expe-
rience and thereby ensure the success of the stream. Despite the fact that specific
definitions vary, these guidelines highlight two types of problematic behaviors:
1) offensive content (manifested in racism, sexism, etc.) and 2) spam – both of
which are harmful to the community.
4.3. ‘Policing’ of impoliteness – Breaking ‘Da Rulez’

In any situation where there are published guidelines for behavior, there is an
expectation that these ‘rules’ will be enforced in some way. In Twitch, this process
is handled by moderators (mods), both human and robot (also known as botmods
or bots). Mods are chosen by the streamers themselves, and Twitch offers the fol-
lowing suggestions for finding viewers who are suitable mod candidates:
What should you look for in a moderator?
When choosing a moderator, you should have a general idea of what you consider
offensive and what you consider spam.6 In some channels, large capitalized text blocks
are considered spam, as they make others’ text very hard to distinguish. Other chan-
nels, particularly channels with slow moving chat, occasional text blocks are less intru-
sive. This is only one example of spam, but there are many others that you should
consider:
5
In some cases, gamers are allowed early access to games by the game developers before
they are released to the public. In such cases, those with early access are expected to
refrain from giving spoilers – details about the game (such as the storyline) – so that
people can experience and discover the game for themselves when it is released to the
public.
6
Emphasis added.

Authenticated
– Copy pasted text, often large

– Links
– Geometric block images
– Repeated text (texttexttexttexttexttexttext, lollollollollollollollollollollollol)
– Excessive use of emoticons
– Excessive symbol usage (are you serious?!?!?!?!?? Wow!!!!!!!!!!!!!!!!!!!!!)
When choosing moderators, make sure you communicate what you want removed, and
what is fine.
(http://help.twitch.tv/customer/protal/articles/1360598-moderation-guide)
These guidelines highlight two of the most prominent themes: no ‘offensive’ con-
tent and no spam. The question remains, however, how these problematic behav-
iors are identified and controlled, particularly in light of the fact that streamers are
encouraged to make their own definitions.
4.3.1. BotMods v. human mods

Streamers have two systems in place for regulating behavior in their streams:
human mods and botmods. Botmods by nature must rely on a predictive framework
for impoliteness – they must be programmed ahead of time to recognize posts that
break the rules of politeness and delete them. For a botmod to work, someone will
need to write code that instructs it on how to recognize impolite/offensive/prob-
lematic messages. This requires a definition, however, that is specific enough to be
enforceable by a robot that is not capable of assessing the nuances of impoliteness
in context. A botmod can be programmed to recognize and delete messages that
have more than a certain number of emotes or symbols (as noted above in both
the Twitch Rules and Da Rulez), for example, and can reliably identify and delete
hyperlinks as well as large blocks of ascii text (also frequently known as ‘copypas-
tas’). They can also (in theory) be programmed to recognize and delete swearing.
You could, for example, compile a list of swear words and program the botmod to
delete all messages that include them. This predictive model is flawed, however,
since users can manipulate language in ways that will sidestep the coding while
still making the meaning clear (e.g. substituting fcuk for the word fuck or s*it for
the word shit) and it would be impossible to anticipate and program a botmod to
delete all of the possible combinations.
While botmods can provide some amount of regulation and enforcement
(e.g. in identifying hyperlinks), streamers must also rely on human mods to recog-
nize content that could be classified as ‘offensive’ according to the definitions they
(the streamers) have identified. The following example illustrates how messages
that do not contain objectively identifiable markers (the kind that botmods could
recognize) that would mark a given post as a violation, can still contain offensive
content.

Authenticated
Example 1
1 igivenofuks: poe7 is for sihtkids get this trash off twitch tv you fukn loser […]
2 r3cp3ct: ban this dude […]
3 igivenofuks: read my name see how i feel about ban in this siht chat
4 igivenofuks: fukboy […]
5 r3cp3ct: o im a fukboi […]
6 igivenofuks: this game is for retarded pussies who seek no challenge at all […]
7 igivenofuks: and have zero gaming skills
8 Dormantg: someones on their period
9 Jekke888: @igivenofuks, I could beat the *** out of you with no challenge at all. […]
10 r3cp3ct: jekke888 don’t bother, he is trolling
11 Jekke888: I’m aware […]
12 igivenofuks: oh my god guys look at him hes proving valid points and don’t agree with
him he is a troll wahh wahh
13 igivenofuks: fk urself […]
14 Dormantg: you must be 12 […]
15 igivenofuks: age joke great comeback […]
16 MisterMycelium: TT.TT @igivenofuks next time ask daddy to be a little more gentle
when he gets his jolly’s off
17 Jekke888: I don’t give a *** if you’re a troll or not if you are not a troll you are stupid
and if you are a troll you are stupid congratualtions you are stupid either way […]
18 igivenofuks: anyone who plays this game should be shot […]
19 igivenofuks: end of story
20 <igivenofuks has been banned from talking>
21 jekke888: end of story
In this exchange, igivenofks’ posts are clear examples of offensive messages that
are interpreted as trolling, an attempt to provoke a reaction and get attention (see
Hardaker, Ch. 18, this volume for further discussion of trolling). The messages,
however, do not contain predictable features that a botmod might use to identify
them. The terms used for swearing have letters that are inverted or deleted (e.g.
siht for shit and fuk for fuck), and it would be impossible to anticipate all of the
potential spelling or other ways that trollers and spammers could manipulate the
texts so that you could program a botmod to identify and delete them. Similarly,
igivenofuks’ statement in line 6, “this game is for retarded pussies”, is not only
offensive but also highly context-dependent. While a streamer might attempt to
program a botmod to delete words or phrases that s/he finds particularly offensive,
it would be impossible to anticipate every potentially problematic or impolite term.
Given this, streamers must rely on human mods to determine, based on the con-
text, which utterances are disruptive, offensive, impolite or constitute hate speech.
In the following exchange, posted during the open chat of a female streamer, the
use of the term ‘pussy’ has multiple meanings:
7
poe = Path of Exile – an online role playing game (RPG).

Authenticated
Example 2
1 Wozleee:[to the female streamer]: can I see your pussy
2 EliteShogun: @Wozlee: can I see yours? […]
3 Lolnignog: you can see my pussy
4 Lolnignog: it’s in my backyard
In this exchange, Wozleee directs an offensive message to the female streamer.

The responses come from other chat participants who challenge Wozleee’s state-
ments by appropriating his language. Of particular interest here is the fact that
Wozleee and Lolnignog use the same term. Wozlee, however, is banned from the
chat temporarily, while Lolnignog is not. In this case, Lolnignog’s use of the term
pussy challenges Wozleee by highjacking the word and redefining it with a more
socially acceptable meaning (i.e. by shifting the reference from a female body part
to a more benign and less personally offensive reference to a pet cat). This type of
exchange would be impossible for a botmod to interpret. Even if a streamer were
to compile a list of offensive words and program the botmods to delete any mes-
sage that contained them, in cases like this a botmod would be unable to discern
the difference between the offensive use and the benign use of the term – for that,
streamers must rely on human moderators who can judge (im)politeness in context
and determine what, in fact, is problematic and what is not.
Similarly, since spam is prohibited because it is disruptive, mods must judge
what counts as a disruption and how disruptive a given behavior is before they
decide whether to discipline the participant. Posting a copypasta message once, for
example, is unlikely to be interpreted as highly disruptive, but posting the same
message multiple times in quick succession makes it difficult for other participants
to continue the conversation because the spam is ‘clogging up’ all of the available
space. The following exchange took place after a participant called another partic-
ipant a black gay fat man.
Example 3
1 Aquafresht: Stop being racist guys its so wrong please, people actually get upset and
sad when you say racist things !! Show some respect! Show some humanity! God damn
internet kids thnks eberything is funny and that they can make fun of everything. [sic]
[15 intervening messages]
2 Luquitasc: @Aquafresht is that a copy pasta?
3 Aquafresht: no that is 100 % me
4 Luquitasc: Stop being racist guys its so wrong please, people actually get upset and
internet kids thnks eberything is funny and that they can make fun of everything. 6 [sic]
[4 messages]

Authenticated
internet kids thnks eberything is funny and that they can make fun of everything. pls no
copy pasterino
[4 messages]
internet kids thnks eberything is funny and that they can make fun of everything. 6
internet kids thnks eberything is funny and that they can make fun of everything.
[…]
Over the course of 2 minutes, this message (with only minor variations tagged
onto the end) is reposted a total of 10 times by three different users. None of the
three were timed out or otherwise punished, despite the fact that repost of this
(rather long) message could certainly be described as spam since it doesn’t convey
any new information and therefore just ‘takes up space’ in the chat stream. The
fact that there are intervening messages between the (re)posts of the spam seems to
indicate that the flow of the chat is not threatened too much, however – the conver-
sation continues around these copied and reposted messages. Just as in Examples 1
and 2 above, then, the mods must make a judgment about whether these messages
count as spam (i.e. they must identify impolite behavior) and decide whether pun-
ishment is needed for the good of the community. In this way, human mods must
make a judgment call as to whether a particular user should be banned by exam-
ining the content of the message, weighing its effects within the unfolding chat
stream and determine if it is something that could or should be banned based on
the streamer’s definitions of what is problematic (which as Twitch notes, should be
conveyed to the mods by the streamer prior to the chat). In an ideal case, these two
regulatory groups – botmods and human mods – should be able to work together
to limit impolite/counterproductive behavior. As we will see below, however, there
are still flaws in implementation of this system.
4.3.2. Flaws in the system

While one might expect the combination of botmods and human mods to cre-
ate a symbiotic balance in which the chat would flow smoothly and problematic
users could be controlled before becoming disruptive, the system, in fact, is inade-
quate to accomplish this goal for two reasons: 1) the fact that decisions about what
counts as impolite must be balanced against the ‘needs’ of the stream in order to
establish community and build/maintain viewership, and 2) community member-
ship is often marked by impoliteness and violations of the rules (just as has been
shown in many other on- and off-line communities; e.g. Graham 2008).
In the following exchange, a streamer is asked for information about a key-
board command to employ during a particular game to make something blow up.

Authenticated
He responds by telling the viewers (via audio mic rather than typing in chat) to
type QWEQEQ. Fifteen of the next 34 messages contain some variation of this
sequence of letters. Approximately 45 seconds after the messages begin, one of
the mods begins to timeout users (presumably because he interprets their postings
as spam – repetitive and devoid of meaningful content). Twenty-five seconds later,
the streamer realizes that users have been banned and responds with the following:
Example 4
1 Wingsofdeath [streamer]: who the ***
2 Wingsofdeath: Is timing them out
[1 message]
3 Wingsofdeath: holy ***
[5 messages]
4 Wingsofdeath: 3 mods
5 Wingsofdeath: one of you
[1 message]
6 Wingsofdeath: speak up now
[1 message]
7 eosmeep: sorry that was me
8 Wingsofdeath: or youre all gone
9 Wingsofdeath: ok
10 eosmeep: won’t happen again
[2 messages]
11 Wingsofdeath: WAT IS WRONG WITH U
[1 message]
12 Wingsofdeath: seriously
[4 messages]
13 Wingsofdeath: Were u completely absent
[2 messages]
14 Wingsofdeath: for the time I was talking about modding
[84 messages]
15 eosmeep: not sure how to answer you wings when chat is going too fast and you
probably wont read what i type
[5 messages]
16 eosmeep: I hit timeouts instead of bans on bttv
[43 messages]
17 <jtv removes channel operator (i.e. mod) status from eosmeep> (5 minutes after origi-
nal bans by eosmeep)
In this case, the problematic messages looked like spam – they showed a sequence
of letters without any identifiable meaning (i.e. they looked like random characters
strung together) and many were typed in all caps (one of the markers of spam as
noted above). While timing out a handful of users may seem like a minor action, in
this case the streamer, Wingsofdeath, was extremely upset, because if participants
are timed out (especially unjustly) then they won’t continue watching the stream.
In this case, the messages that had markers of spam were misinterpreted by one

Authenticated
of the mods and his response created a situation that the streamer felt might dis-
courage viewers from continuing to watch. This balance between allowing users
to post what they want to in order to keep them watching versus the need to con-
trol ‘empty’ content that disrupts the flow of the chat was derailed. In such cases,
rather than making controlling spam the priority, the streamer may decide to allow
a certain amount of spam so that his/her viewers will keep watching.
In addition to streamers allowing spam up to a certain point in order to keep
viewership, there are also instances where streamers take the idea a step further. In
the case of imaqtpie whose rules are posted above, there is an automated system
that tracks participation and not only allows problematic behavior like spam, it even
sanctions it and uses it as a reward for watching. For every minute users stay in the
chat room, they are awarded points (subscribers get more points than non-subscrib-
ers). They can then spend these points (which are tracked by a botmod) to buy the
right to post spam. For a certain number of points, users can post ascii pictures like
the one below as well as long text paragraphs where the content isn’t relevant to the
ongoing discussion in the chat stream as in examples 5 and 6 below:
Example 5
░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
░░░░░░░░▄▄▄▀▀▀▄▄███▄
░░░░░▄▀▀░░░░░░░▐░▀██▌
░░░▄▀░░░░▄▄███░▌▀▀░▀█
░░▄█░░▄▀▀▒▒▒▒▒▄▐░░░░█▌
░▐█▀▄▀▄▄▄▄▀▀▀▀▌░░░░░▐█▄
░▌▄▄▀▀░░░░░░░░▌░░░░▄███████▄
░░░░░░░░░░░░░▐░░░░▐██████████▄
░░░░░le░░░░░░░▐░░░░▐██████████▄
░░░░toucan░░░░░░▀▄░░░▐█████████▄
░░░░░░has░░░░░░░░▀▄▄██████████▄
░░░░░arrived░░░░░░░░░░░░█▀██████
Example 6
13:37 wowsobot: “Hello Miguel Santander, this is your cat, Mellow Cat, I heard that you
were complaining about my excesive meowing. I dont complain about having to hear you
talk to yourself for 8 hours while playing that stupid game so I expect the same. Come here
and feed me scrub”
13:37 wowsobot: qtpie, this is your dad here. i just wanted to say how proud of you i am
son, always hiding in your room playing your video games, talking to your internet friends.
your mom and i think that’s swell and we just wanted to let you know we love you. pls no
copy dadderino
Imaqtpie’s reward points allow for users (and subscribers have an advantage) to
break the rules in a sanctioned way. Chat participants with enough points can
‘buy’ access to spam that includes copy pastas – including blocky ascii images and
repeated text as well as links (which violate the rules set forth by Twitch.tv and

Authenticated
the streamer himself). The streamer himself, for that matter, breaks his own rules.
While both his rules and the Twitch.tv rules prohibit sexism, for example, every
time someone subscribes to his channel, he reads the subscriber’s name aloud and
says “Welcome to the big dick club”. Despite the fact that some would argue that
this statement is inherently sexist, the streamer in this case is not only allowing this
violation of the rules but is participating in the violations himself.
4.4. Remaining questions

So where does this leave us? In DC there are attempts to develop a predictive
model of impoliteness which can be used to control for some aspects of impolite
behavior. Even in non-gaming settings, there are spam filters (which most would
agree are imperfect) and an increasing number of tools (like the Clutter feature in
Outlook email that takes messages that it perceives as non-urgent and puts them
into a separate folder) that are designed to help identify and filter out unwanted
digital information. But these are doomed to failure and must be checked by
humans – everyone has found an email that ‘slipped through’ the spam filter or
identified a spammer who was able to slip past a mod by adjusting the spelling or
presentation of a problematic phrase. The difficulty with this is first defining (so
that effective codes and filters can be developed) inappropriate remarks and then
identifying impolite behavior in the moment quickly enough so that you can delete
messages before they become disruptive – all without discouraging viewership.
The tension between attempts at enforcement and the goals of the streams
leads to a situation where (once again) (im)politeness is different in practice than
in theory and cannot be theorized to be used in a predictive way. This brings us
back to relational work. As noted in Graham (2008 and 2016), impoliteness and
breaking the rules for behavior is used as a means of marking community and
in-group membership. In this case, the rule-breaking is even encouraged; as view-
ers show greater commitment to the group they are allowed (even encouraged) to
break the rules.
5. Future directions
It is clear that digital communication has become an ever-present (and increas-

ingly complex) part of our daily interactions and is therefore a setting where we
will continue to negotiate and revise our expectations of (im)polite behavior. As
the norms of interaction morph and change as new technologies and platforms
develop, it would be a worthy area of examination to explore the ways that early
concepts of Netiquette (which as discussed are still in play today) have changed
to meet the demands of new media and correlated new norms and expectations of
(im)politeness.

Authenticated
Another potential area for further investigation is the way that computer-gen-
erated controls are used in conjunction with humans (as with bots and human
moderators) to construct and regulate the ways that behavior is assessed and con-
trolled. As automated systems become more complex, we might ask ourselves
whether it will ever be possible for a bot to have the ability to recognize (im)polite
behaviors in digital contexts and how their shortcomings in this area relate to the
feasibility of a second-order model of (im)politeness.
The largest potential area for further research, however, is in the area of mul-
timodal interaction. Digital communication becomes more multimodal daily – we
can carry on a conversation via video link on Skype while simultaneously typing
messages and sending hyperlinks to the same user. In this context, we are employ-
ing audio capabilities in conjunction with video and text-based communication,
each of which draws on different communicative strategies and expectations
(including assessments of identity and (im)politeness). In the case study above,
for example (Example 4), a human moderator timed participants out for posting
messages that looked like spam. Those (written) messages, however, were posted
in response to an audio comment by the streamer. In this case, looking only at the
chat stream, one might misinterpret the messages as spam since they look like
this problematic behavior when they are taken out of the context of the stream-
er’s comments (which were transmitted via audio rather than writing). Linguistic
research on multimodality is limited (although see Naper 2011), but understanding
the ways that multiple modes of communication are intertwined in cases such as
this can illuminate factors of contextuality that can influence the identification
and interpretation of (im)politeness. The ability of multimodal platforms to bring
features of DC together such as a/synchronicity, relative anonymity, and public/
private communication in this way cannot help but alter our strategies in satisfying
the demands and expectations of (im)polite behavior, and is therefore, I would
argue, the most promising area for future research on (im)politeness.
References

2010 “you HAVE NO IDEA WHAT YOU ARE TALKING ABOUT!”: From
e-disagreement to e-impoliteness in two online fora. Journal of Politeness
Research 6: 57–82.
Arendholz, Jenny
Baym, Nancy K.
2015 Personal Connections in the Digital Age. 2nd ed. Malden, MA: Polity Press.

Authenticated

2015 “Peter is a dumb nut”: Status updates and reactions to them as ‘Acts of Posi-
boyd, danah and Eszter Hargiatti
2010 Facebook privacy settings: Who cares?. First Monday 15(8). doi: 10.5210/
fm.v15i8.3086.
Bousfield, Derek
2010 Researching impoliteness and rudeness: Issues and definitions. In: Miriam
Locher and Sage Graham (eds.), Interpersonal Pragmatics, 101–134. Berlin/
Boston: de Gruyter.
Brown, Penelope and Stephen Levinson
1978/1987 Politeness: Some Universals in Language Usage. Cambridge: Cambridge
University Press.
Chejnová, Pavla
2014 Expressing politeness in the institutional e-mail communications of university
students in the Czech Republic. Journal of Pragmatics 60: 175–192.
Culpeper, Jonathan
2008 Reflections on impoliteness, relational work, and power. In: Derek Bousfield
and Miriam Locher (eds.), Impoliteness in Language: Studies on Its Interplay
and Power in Theory and Practice, 17–44. Berlin/Boston: de Gruyter.
Culpeper, Jonathan
2011 Impoliteness: Using Language to Cause Offence. Cambridge: Cambridge Uni-
versity Press.
Dynel, Marta
2012 Swearing methodologically: The (im)politeness of expletives in anonymous
commentaries on YouTube. Journal of English Studies 10: 25–50.
Ensslin, Astrid
2012 The Language of Online Gaming. Basingstoke, UK: Palgrave.
2010 Introduction: The status quo and quo vadis of impoliteness research. Intercul-
tural Pragmatics 7(4): 535–559.
2013 Relational work in anonymous asynchronous communication: A study of (dis)
affiliation in YouTube. In: Istvan Kecskes and Jesus Romero-Trillo (eds.),
Research Trends in Intercultural Pragmatics, 343–366. Berlin/Boston: de
Gruyter.
Graham, Sage L.
2005 Cyberparish: Gendered identity construction in an online religious commu-
nity. In: Jule Allyson (ed.), Gender and the Language of Religion, 133–150.
Basingstoke, UK: Palgrave Macmillan.
Graham, Sage L.
2007a ‘Do Unto Others’: Gender and the construction of a ‘good christian’ identity
in an e-community. In: Jule Allyson (ed.), Language and Religious Identity:
Women in Discourse, 73–103. Basingstoke, UK: Palgrave Macmillan.
2007b Disagreeing to agree: Conflict, (im)politeness and identity in a computer-me-

Authenticated
Graham, Sage L.
2008 A manual for (im)politeness?: The impact of the FAQ in an electronic commu-
nity of practice. In: Derek Bousfield and Miriam Locher (eds.), Impoliteness
in Language: Studies on its Interplay with Power in Theory and Practice,
Graham, Sage L.
2016 Relationality, friendship and identity in digital communication. In: Alexandra
Georgakopoulou and Tereza Spilioti (eds.), The Routledge Handbook of Lan-
guage and Digital Communication, 305–320. New York: Routledge.
Graham, Sage L. and Claire Hardaker
2017 (Im)politeness in digital communication. In: Jonathan Culpeper, Michael
Haugh and Daniel Kadar (eds.), Handbook of Linguistic Impoliteness, 785–
814. Basingstoke, UK: Palgrave Macmillan.
Hardaker, Claire
cussions to academic definitions. Journal of Politeness Research 6: 215–242.
Hardaker, Claire
2015 “I refuse to respond to this obvious troll.”: An overview of (perceived) troll-
ing. Corpora 10 (2): 201–229.
Haugh, Michael
tions of impoliteness. Journal of Politeness Research 6: 7–31.
Erlich, Miriam Meyerhoff and Janet Holmes (eds.), Handbook of Language, Gen-
der and Sexuality, 567–586. Hoboken, NJ: John Wiley and Sons. doi: 10.1002/
9781118584248,ch29.
Ho, Victor
2011 A discourse-based study of three communities of practice: How members
maintain a harmonious relationship while threatening each other’s face
via email. Discourse Studies, 13(3): 299–326. http://search.proquest.com/
docview/925725317?accountid=14657.
Hymes, Dell
1974 Foundations in Sociolinguistics: An Ethnographic Approach. Philadelphia:
University of Pennsylvania Press.
Kádár, Daniel and Michael Haugh
2013 Understanding Politeness. Cambridge: Cambridge University Press.
Lakoff, Robin
1973 The logic of politeness, or minding your p’s and q’s. Chicago Linguistics Soci-
ety 9: 292–305.
Langlotz, Andreas
2010 Social cognition. In: Miriam Locher and Sage L. Graham (eds.), Interpersonal
Pragmatics, 167–202. Berlin/Boston: de Gruyter.

Authenticated

Lea, Martin, Russell Spears and Daphne de Groot
1991 Knowing me, knowing you: Anonymity effects on social identity processes
within groups. Personality and Social Psychology Bulletin 27: 526–537.
Leech, Geoffrey
Leech, Geoffrey
Locher, Miriam A.
2006 Polite behavior within relational work: The discursive approach to politeness.
Multilingua 25(3): 249–267.
Locher, Miriam A.
2010 Introduction: Politeness and impoliteness in computer mediated communica-
tion. Journal of Politeness Research 6(1): 1–5.
Locher, Miriam A.
2012 Politeness. In: Carol E. Chapelle (ed.), The Encyclopedia of Applied Linguis-
2005 Politeness theory and relational work. Journal of Politeness Research 1: 9–33.
Locher, Miriam, Brook Bolander and Nicole Höhn
2015 Introducing relational work in Facebook and discussion boards. Pragmatics,
25(1): 1–21.
2011 Online-polylogs and impoliteness: The case of postings sent in response to
the Obama Reggaeton YouTube video. Journal of Pragmatics, 43(10): 2578–
2593.
Maher, Brenden
2016 Can a video game company tame toxic behaviour? Nature. http://www.nature.
com/news/can-a-video-game-company-tame-toxic-behaviour-1.19647 doi:
10.1038/531568a.
2013 ‘Just Click “Like”’: Computer-mediated responses to Spanish compliments.
2014 Impoliteness in Facebook status updates: Strategic talk among colleagues
‘outside’ the workplace. Text and Talk 34(2): 165–185.
2014 Networked privacy: how teenagers negotiate context in social media. New
Media and Society. doi: 10.1177/1461444814543995.
Merrison, Andrew John, Jack J. Wilson, Bethan L. Davies and Michael Haugh
2012 Getting stuff done: Comparing e-mail requests from students in higher educa-
tion in Britain and Australia. Journal of Pragmatics 44(9): 1077–1098.
Moderation Guide – Broadcaster
2013 http://help.twitch.tv/customer/portal/articles/1360598-moderation-guide---
broadcaster [29 August 2015].

Authenticated
Myers, Greg
2010 The Discourse of Blogs and Wikis. New York: Continuum Press.
Naper, Ida
2011 Conversation in a multimodal 3D virtual environment. Language@Internet 8:
1–27.
Nishimura, Yukiko
2010 Impoliteness in Japanese BBS interactions: Observations from message
exchanges in two online communities. Journal of Politeness Research 6:
33–55.
O’Driscoll, Jim
2013 Situational transformations: The offensive-izing of an email message and the
publicization of offensiveness. Pragmatics and Society 4(3): 369–387.
Page, Ruth
62: 30–45.
Perelmutter, Renee
2013 Klassika Zhanra: The flamewar as a genre in the Russian blogosphere. Journal
of Pragmatics 45(1): 74–89.
2010 Virtual community and politeness: The use of female markers of identity
and solidarity in a transvestites’ website. Journal of Politeness Research 6:
83–103.
Pojanapunya, Punjaporn and Kandaporn Jaroenkitboworn
2011 How to say ‘goodbye’ in Second Life. Journal of Pragmatics 43(14): 3591–
3602.
Richet, Bertrand
2013 Fanning the flames?: A study of insult forums on the internet. In: Denis Jamet
and Manuel Jobert (eds.), Aspects of Linguistic Impoliteness, 223–231. New-
castle: Cambridge Scholars.
Rusaw, Erin
2011 Language and social interaction in the vitual space of World of Warcraft. Stud-
ies in the Linguistic Sciences 36: 66–88.
Schwämmlein, Eva and Katrin Wodzicki
2012 What to tell about me?: Self-presentation in online communities. Journal of
Shea, Virginia
1994 Netiquette. http://www.albion.com/netiquette/corerules.html.
Sifianou, Maria
Terkourafi, Marina
2008 Toward a unified theory of politeness, impoliteness, and rudeness. In: Derek
Bousfield and Miriam A. Locher (eds.), Impoliteness in Language: Studies
on its Interplay with Power and Theory in Practice, 45–74. Berlin/Boston: de
Gruyter.
23–45.

Authenticated
Twitch.tv Rules of Conduct. https://www.twitch.tv/p/rules-of-conduct.

Upadhyay, Shiv
Politeness Research 6: 105–127.
Vandergriff, Ilona
communication (CMC) cues. Journal of Pragmatics 51: 1–12.
Vinagre, Margarita
2008 Politeness strategies in collaborative email exchanges. Computers and Educa-
tion 50(3): 1022–1036.
Watts, Richard J.
2003 Politeness. Cambridge: Cambridge University Press.
Watts, Richard J.
2005 Linguistic politeness and politic verbal behavior: Reconsidering claims for
universality. In: Richard Watts, Sachiko Ide and Konrad Ehlich (eds.), Polite-
ness in Language, 21–42. 2nd ed. Berlin/Boston: de Gruyter.

Authenticated
Authenticated
Sage Lambert Graham
Abstract: This chapter will explore the concepts of politeness and impoliteness as
they are enacted in various forms of social media. Using an ‘interpersonal prag-
matic’ approach (Locher and Graham 2010), the focus is on interactive mediated
settings such as vlogs, email, discussion boards and virtual worlds. The chapter
examines the parameters of how we define (im)politeness, the terminology we
employ, and the ways that the digital setting influences the way (im)politeness is
enacted. This is followed by an illustrative case study on online gaming that exam-
ines how impoliteness is identified and challenged by participants while moder-
ators simultaneously attempt to define and control for this type of problematic
behaviour. The chapter concludes with a brief overview of directions for future
research.
1. Introduction
Notions of (im)politeness are present in every interaction; we draw on them as

we communicate to accomplish tasks, establish social connections and construct
our identities in a wide array of settings. We are taught from the youngest ages
how to behave ‘politely’, and throughout our lives these notions evolve and we
evaluate our own and others’ actions and make our communicative choices based
on our ideas about what is or is not (im)polite. And while as children much of this
instruction tends to focus on face-to-face contexts, the role of (im)politeness is no
less present or important in digital interactions. Locher (2010) notes that online
communication tends to develop its own set of communicative norms and prac-
tices due to factors such as the merging of public/private audiences and the multi-
modal capabilities of digital media, and as digitally–mediated communication has
increased and evolved, the ways that we navigate the expectations of (im)polite-
ness and interpret others’ behavior in digital contexts has increasing importance in
our more-and-more-digital world.
It is the goal of this chapter to explore the ways that the context of social media
influences and reflects the ways that (im)politeness as a pragmatic device is con-
ceptualized, negotiated and enacted in digital interaction. That said, it is important
to note that the title of this chapter, Politeness and Impoliteness (in Social Media),
might be used to refer to any number of digital platforms and media channels –
far too many to be addressed here. This chapter will therefore restrict the focus to
platforms that lend themselves to interpersonal (and interactive) pragmatic dis-
DOI 10.1515/9783110431070-017
De Gruyter Mouton. Brought to you by | University of Cambridge
Authenticated
Download Date | 11/3/17 9:18 AM
course. After addressing aspects of digital communication generally, it will move

to an overview of some important concepts in (im)politeness research, examine
where (im)politeness and digitally-mediated discourse research come together,
and explore the manifestation of this overlap in a case study of one particular
digital arena: online gaming. The chapter concludes with a discussion of future
developments and directions for further research.
2. Important issues
In any exploration of (im)politeness in digital communication (DC),1 there are

a number of influential factors that come into play: the nature of DC itself, our
current understanding of politeness and impoliteness and the roles they play in our
interactions, and the ways that they intersect and interact when they come together.
The next section will therefore address each of these in turn.
2.1. The nature of digital communication

The array of digital platforms available for communication today is staggering,
and is evolving daily, both in terms of capabilities as well as popularity. While
any explanation of ‘current’ digital media is, therefore out-of-date before it can
ever reach the printer, there are some common factors shared by various digital
platforms that affect the nature of communication in digital contexts and how (im)
politeness is created and interpreted. In addition to the fluctuations in the popular-
ity and creation of different media, moreover, the array of digital platforms goes
far beyond the platforms that were common in early CMC research, which focused
largely on email, discussion groups and online role-playing communities such as
MUDs and MOOs.2 For reasons of space, I will focus here on communication that
1
Terms used to refer to communication in digital contexts have evolved along with
technologies, and at this point early terms such as computer-mediated communication
(CMC) are no longer accurate since much digital interaction takes place without a com-
puter. While alternative terms such as computer-mediated discourse analysis (CMDA)
have been proposed, for the purposes of this chapter I will use the term digital commu-
nication (DC) except when referring to previous research that uses other terms or when
referring to contexts that are limited to only interaction via computer.
2
Although not as commonly used as they once were, MUDs (Multi-User Dungeons or
Multi-User Domains) and MOOs (which are related to MUDs but are Object-Oriented
virtual realities) are text-based formats that allow multiple users to interact concurrently
via a common server. These platforms were the basis for many current Role-Playing
games (RPGs) and other multi-player formats where players (who do not necessarily
know one another) come together and play/create through a shared server.
Brought to you by | University of Cambridge

Authenticated
is mediated via computer within social platforms rather than attempting to address
video, television, news media or a whole range of other mediated communicative
settings; it is important to remember, however, that this only scratches the surface
of the array of communicative acts now taking place via digital channels.
2.1.1. Digital media versus face-to-face interaction

One question that has often been asked is whether, or how much, digital commu-
nication is different from face-to-face interaction (f-t-f) (e.g. Benwell and Stokoe
2006). Locher, Bolander and Höhn (2015) rightfully argue that DC and f-t-f inter-
action are not even truly separable at this point because in many cases “we cannot
clearly split the norms interactants bring to the keyboard with those that emerge in
online interactions” (2015: 9). While it is certainly the case that digital interaction
and f-t-f interaction have become more-and-more intertwined in recent years, there
are still some features that distinguish digital interaction from f-t-f encounters.
One place where this distinction can be seen is in the wide array of tools in dig-
ital interaction that encourage users to evaluate others’ behavior as (in)appropriate
(or (non-)politic to use Watts’ 2005 and Locher and Watts’ 2005 terms, discussed
further below). These tools include buttons that are prominently placed on the
screen that ask us to declare our assessments of others’ behavior via ‘blocking’,
using spam filters, using ‘report abuse’ buttons, and, on the positive end, ‘Like-
ing’, ‘Share-ing’ and marking as ‘Favorite’, through which we are continually
asked to rate others’ communicative acts. If we were to argue that Like-ing indi-
cated a positive (politic) assessment and Block-ing indicated a negative (non-pol-
itic) assessment of another’s behavior, then this would mean that assessments of
(im)politeness (on the (im)politeness ‹—› (non-)politic spectrum) are encoded into
the media themselves via these tools, and there is an implicit notion that these
evaluations are something that can (or even should) be declared openly and con-
tinuously. This is in contrast to face-to-face interactions, where, although we may
be constantly evaluating others’ behavior, we are not asked in every interaction
to ‘click buttons’ to openly declare our (ongoing) assessments. So, although we
may question how different the pragmatic strategies are in communicating via
computer or other digital channels versus face-to-face, most would agree that they
are different at least in some ways (just as having a conversation via telephone is
different than having the same conversation f-t-f).
2.1.2. Paralinguistic tools

In digitally-mediated contexts, we also do not have access to the same paralinguis-
tic cues like intonation, stress, and facial expressions that we have in face-to-face
interactions, and these tools are a critical component in conveying the pragmatic
intent of a given message. When DC first became popular among the general pop-

Authenticated
ulation, there was early adaptation of keyboard symbols meant to approximate

facial expressions in written form (also known as ‘smilies’ or emoticons). These
early adaptations, however, were in large part limited to characters and symbols
available on a QWERTY keyboard, since most devices and platforms were limited
to keyboard-based communication rather than touch-screens with graphic capa-
bilities or real-time audio. There was therefore a limit to the scope of information
they could convey.
In more recent years, there has been an explosion in the development of new
digital platforms and modalities, most of which try to provide tools to compensate
for the lack of paralinguistic cues (Vandergriff 2013). Smartphones, for example,
have new ‘emojis’ added with almost every software update and include not just
facial expressions as the first emoticons did, but also pictures that abbreviate mes-
sages that are too tangential to type or that the users feel should be abbreviated
(examples include country flags, waving hands, zodiac symbols, pictures of rain-
bows and dozens more). These emojis are certainly expressive – they allow users
to convey a great deal of information without having to type long messages, and
take a step toward indicating whether someone intends an utterance as humorous
or serious just as the early emoticons did. The increased number of available emo-
jis, however, has not necessarily simplified our ability to clarify communicative
intent. While users may choose emojis to save time or simply avoid keying text,
there is a price in that there are still ambiguities that cannot necessarily be captured
in a picture. Sending an IM or SMS message with an emoji of a spiderweb (one
of the emojis available on the iPhone 6), for example, would allow a user to avoid
typing, but also could be interpreted in many ways – “I’m tangled up so I’ll be
late”, “I’m in a sticky situation”, “I’m afraid of arachnids”, etc.
It is generally agreed that emojis as substitutes for facial expressions and para-
linguistic cues is imperfect – particularly when we take into account the ambig-
uous use as either emotive (e.g. a winking face to express a humorous intent) or
as a neutral abbreviating device (e.g. a picture of a ‘thumbs-up’ in lieu of typ-
ing ‘okay’), but they are nevertheless an important part of digital interaction and
one that is critical to the interpretation of (im)politeness because, no matter how
imperfect, they help interactants clarify their intentions.
2.1.3. A/synchronicity
In digital communication, speed of response is connected to assessments of (im)
politeness more and more frequently. As transmission times have become faster
through fiber-optic cable and extensive internet availability, a common expectation
is that people will both receive and respond to messages more and more quickly.
To fail to respond within this ever-quickening timeframe is often considered impo-
lite. One factor that affects response time is the relative a/synchronicity of the
communicative medium. Chat platforms, for example, are quite synchronous and

Authenticated
approximate real-time ongoing conversation much more closely than more asyn-
chronous platforms like email or Facebook, where messages are sent or posted and
will be read and replied to at some (undetermined) point in the future. And even
when response-time on these relatively asynchronous channels is comparatively
quick, there is not an expectation that it will be quick (or even instantaneous) every
time for every participant. In terms of impoliteness in fact, asynchronous conflicts
often have a tendency to grow a life of their own and may spiral out of control
before they can be recognized and then offset/controlled (Graham 2007b), whereas
in a synchronous chat, a conflict could be resolved quickly in real time, allowing
everyone to move on and resume the interaction. The longer the conflict extends,
however, the longer it will take to resolve, and this is why a/synchronicity plays a
role in online impoliteness.
2.1.4. Anonymity
Yet another aspect of DC that is frequently discussed is anonymity. It is often
posited that people are more impolite online than would ever be f-t-f, and that this
willingness to behave impolitely/inappropriately (terms that are further discussed
in Section 2.2.1 below) stems from the fact that they can act anonymously and are
therefore free from any negative ramifications that their actions might have (see
Lea, Spears and de Groot 1991 and Vinagre 2008 for further discussion of both the
positive and negative results of anonymity online; also Bös and Kleinke, Ch. 4,
this volume; boyd and Hargiattai 2010; Herring and Stoerger 2014; and Marwick
and boyd 2014 for further discussion of privacy and context in social media).
When using anonymity as a lens through which to examine (im)politeness, it is
important to remember that anonymity online (and its counterpart privacy) is not
universal – it operates on a spectrum, with some platforms affording a greater
degree than others. Nevertheless, it is an important part of how people create and
navigate their behavior (either polite or impolite) in digital settings.
2.1.5. Longevity
In addition to anonymity, DC is also affected by the longevity of any given digi-
tal platform. Users might justifiably ask themselves “Will my previous behavior,
polite or impolite, benign or malicious, discreet or indiscreet, come back to haunt
me?” With scandals such as the Ashley Madison case in 2015 where a website
for people having extramarital affairs was hacked and private information leaked,
users have become more aware of the longevity of what they communicate and
how they communicate it via digital channels. The fear that impolite or inappro-
priate behavior will be resurrected and possibly used against them in the future has
led to the development of social media sites like Snapchat, where digital content is
automatically deleted after a specified time period. The knowledge that our actions

Authenticated
are recorded and may expose our prior behavior must certainly affect our polite or
impolite language choices.
2.1.6. Other motivational factors and goals of digital interaction

In addition to the factors listed above, there are other considerations as well; dig-
ital communication is also influenced by 1) whether the audience is public or pri-
vate, and 2) what the goals of the interaction are (e.g. whether participants want to
establish relationships or accomplish some task).
Public interaction makes notions of (im)politeness potentially less certain and
predictable than private discourse, since the audience may be unknown. Private
communication, since the audience is likely known, may be easier to predict/con-
trol – for interactants may have a greater awareness of the norms and expectations
of their audiences and might therefore be more conscious of meeting those expec-
tations since they will likely be held accountable for the choices they make.
The goals of interaction also have an impact on both the norms and execution
of (im)politeness. In a case where an individual wants to declare a stance on a
particular topic (e.g. “I support my country’s Olympic team!”), the statement is
intended to declare a stance. Discussion may ensue if others respond, but that isn’t
a certainty. In other cases, the goal of interaction might be to establish a connection
with others with similar interests and/or join a community (for further discussion
of community formation online see Schwämmlein and Wodzicki 2012 and Baym
2015). In this case, we might expect communication to attend more closely to the
norms of politeness that exist within the group since the goal might not be accom-
plished if these rules are not obeyed.
2.2. The nature of (im)politeness and (non-)politic action/(in)appropriateness

(Im)politeness is not only a fascinating lens through which to examine human
behavior, it is also a hotly debated topic in pragmatic research, much of which
builds on seminal works such as Brown and Levinson (1978/1987), Leech (1983)
and Lakoff (1973). Some of the major themes and questions that have emerged
from this research involve 1) what the primary goals and focus of (im)politeness
research should be, 2) how (im)politeness should be defined, 3) what terms we
should use to define and explain (im)politeness, 4) what role intent plays in estab-
lishing/identifying (im)politeness, and 5) how (im)politeness intersects with vari-
ables such as language, ethnicity, geography/nationality and gender.
It is far beyond the scope of this chapter to give an overview of several decades
of politeness research, but a large portion of research on the pragmatics of (im)
politeness involves the balance people strike between satisfying their own need
to be autonomous (i.e. not imposed upon) with catering either explicitly to the
wants and needs of others or to societal expectations for what constitutes appro-

Authenticated
priate or desirable (and their counterparts inappropriate and undesirable) behav-

ior (for comprehensive overviews of theoretical approaches see Arendholz 2013;
Garcés-Conejos Blitvich 2010; Kádár and Haugh 2013; Locher 2012).
In addition to the themes listed above, an additional question that has arisen is
whether the classification of utterances as (im)polite should be the primary focus
of exploration or whether the interactional impact should be the focus of attention.
The later point of view posits that research should examine how notions of (im)
politeness affect communication through emergent interaction, how people inter-
pret behaviors as (im)polite and how this affects their communicative practices and
relationships. Watts (2003), Locher (2006) and Locher and Watts (2005) take this
stance in their framework of Relational Work, where interactions are categorized
as (non)-politic, or more or less appropriate for a given situation and how interact-
ants use their assessments of these factors in establishing relationships. According
to this framework, behavior must go above-and-beyond what is simply appropriate
(and therefore unmarked) in order to be classified as polite or impolite. A driver
switching lanes to avoid hitting a cement block that had fallen in the roadway, for
example, is taking a neutral (appropriate/politic) action. The same driver stopping
and moving the block to the side of the road is going above-and-beyond because
s/he considers how hitting the block might negatively affect other drivers and
(being polite) takes extra action to prevent those negative consequences for others.
2.2.1. Terminology and definitions

One of the stickier elements of the ongoing exploration of (im)politeness relates to
the terms we use – these include polite and impolite, rude, appropriate and inap-
propriate and (non-)politic. Many researchers have attempted to tease these terms
apart and identify the characteristics that differentiate them (e.g. Bousfield 2010;
Culpeper 2008; Terkourafi 2008), and much of this research hangs on what Bous-
field calls “the thorny concept of ‘intentionality’” (2010: 111). In other words, a
hearer’s assessment of whether someone means to be offensive/impolite/rude has
an effect on how s/he evaluates and classifies the speaker’s behavior. Building
on Culpeper (2008), Bousfield (2010), for example, makes a distinction between
impoliteness and rudeness by defining impoliteness as intentional and rudeness as
unintentional (2010: 114). Others (e.g. Graham 2008), however, argue that making
a distinction between terms such as impolite and rude is not useful, since we can
never be certain of another’s intent and should therefore focus on the impact of an
utterance (which may however be based in part on the hearer’s interpretation of
the speaker’s intent) in the subsequent interaction as it unfolds.
In addition to the terminological debate about how/whether to distinguish
impoliteness from rudeness, etc., another less-explored consideration is where we
draw lines between behaviors that are (im)polite and ones that index other pos-
itive or negative behaviors. Even if we could agree on definitions for the terms

Authenticated
impolite, inappropriate, offensive and rude, for example, there are also terms that
seem related, but have not quite been examined in the way that (im)politeness and
rudeness have, such as aggressive, violent and harassing (although see Culpeper
2011 for an in-depth discussion of aggression and its relationship to impoliteness).
This is consistent with the spectrum of behavior proposed by Locher and Watts
(mentioned above), in which the labels we apply to specific behaviors are not
clearly differentiated. As we will see in the case study below, in practice, these
distinctions are not always easily made, and, just as (im)politeness is emergent
in interaction, part of that emergence can be the evolution of an interaction from
an exchange that is polite to one that is impolite to one that is angry to one that
is aggressive. If we are to attempt to differentiate these terms, it is critical to ask
ourselves how we intend to draw the lines between what is impolite and what is
aggressive, or when something crosses over from being simply inappropriate to
being outright offensive (and for that matter, does offensive equal rude?). To what
extent can they overlap or co-exist? Where does impoliteness stop and aggres-
sion/anger/violence begin? And perhaps most importantly, does differentiating the
terms matter?
Finally, there is the question of how localized terms of (im)politeness are. Dif-
ferent Communities of Practice (Lave and Wenger 1991) may each have their own
terms that identify behavior as (un)problematic (e.g. toxic, or salty, to choose a
few from online gaming). There may also be different culturally-influenced under-
standings of politeness phenomena. There has been a great deal of research in
recent years, for example, that has argued convincingly that the concept of ‘face’
as it was used by Brown and Levinson (1987) in their seminal work on politeness,
cannot be adequately applied to interactions in Asian contexts where the notion of
face is distinct from the notion of politeness (e.g. papers in the Journal of Polite-
ness Research special issue on face and politeness 8(1)), and any examination of
(im)politeness terminology must take these factors into account.
2.2.2. Predictive models of (im)politeness – FAQs and Codes of Conduct

One of the most hotly debated topics in current research on (im)politeness relates
to first-order versus second-order approaches, which are differentiated largely
through the idea of a predictive model of (im)politeness. A first-order approach
refers to (im)politeness as it is experienced and understood by the actors within an
interaction, while a second-order approach refers to a theoretical scientific exam-
ination that depends on an outsider/objective perspective to outline and predict
(im)politeness. While a comprehensive discussion of the two approaches is well
beyond the scope of this chapter, they are foundational to any examination of (im)
politeness, and (as discussed in Section 4 below) are especially striking in a digital
context where the idea of a predictive model is pivotal to how behavioral regula-
tions are enforced.

Authenticated
Researchers have long debated the values and shortcomings of the first-order
versus second-order perspectives (see Kádár and Haugh 2013 for a comprehensive
overview). Proponents of the predictive model (P2) contend that lay users have
their own understandings of the terms (im)polite which are loose and often too
arbitrary to be meaningful (e.g. Terkourafi 2008; Leech 2014), while researchers
have a more specified and ‘sharpened’ use of the terms which illuminates “statis-
tically observable politeness norms” (Leech 2014: 47). I would argue that taking
a solely second-order perspective, however, questions the abilities of lay persons
to assess (im)politeness in a systematic way and runs counter to practice in digital
settings.
In digital contexts, lay users create predictive tools all the time in the form of
FAQs and Codes of Conduct. Such documents address behaviors that the writers
of the guidelines (who in many cases are likely to be lay users) predict will occur
and outline rules for behavior that is deemed (in)appropriate within the community
in question. It seems a stretch to think that all FAQs, employee manuals, Codes of
Conduct etc. have been written by researchers; I would argue, however, that those
written by lay users are no less predictive.3 Any person (research or lay user) can
attempt to predict behavior and write guidelines to control it, but since norms of
interaction are constantly in flux, any guidelines we write might or might not be
reflective of practice that changes to meet the needs of the interactants. In a digital
environment this is particularly true because change occurs so quickly in modali-
ties and subsequent strategies for interaction.
One response to this line of argument is that documents such as FAQs do
not address (im)politeness explicitly, but instead address a general guideline for
appropriateness. While these manuals may not use the word (im)politeness specif-
ically, however, we can once again draw on Watts’ (2003) perspective that appro-
priate behavior is linked to polite behavior on a spectrum and that these guidelines,
which describe (in)appropriate behavior, therefore implicitly suggest that behavior
that falls outside these norms is (im)polite. From an alternative perspective, I also
would say that these behavioral manuals address conduct that falls in the middle of
Leech’s (2014) (im)politeness spectrum, where there is a neutral center and behav-
ior (positive or negative) moves away from this point as it becomes more polite
in one direction and more impolite in the other. Using either of these frameworks,
FAQs and Codes of Conduct ask that readers strive to maintain this middle point
of neutral behavior while implying that behavior that progresses away from it is
more- or less polite or impolite.
3
It is also important to note here, that ‘predictive’ is not necessarily synonymous with
‘accurate’ (see Graham 2008 for how this notion materializes in FAQs).

Authenticated
2.3. The nature of (im)politeness and (in)appropriateness in digital

communication – where they come together
If we agree with the early observation that digital (im)politeness has at least some
characteristics that differentiate it from f-t-f interaction, then it is worthwhile to
examine where the norms of digital interaction and the pragmatics of (im)polite-
ness come together, and this will be the focus of the following sections (for addi-
tional discussion of how the parameters of digital contexts intersect with (im)-
politeness see Graham and Hardaker 2017).
2.3.1. Guidelines/rules/FAQs
With regard to (im)politeness in digital contexts, there are baseline rules that are in
large part agreed upon by people communicating in this environment. These draw
on principles of Netiquette such as those outlined by Virginia Shea (1994), whose
original guidelines said:
1) Remember the human

2) Adhere to the same standards of behavior online that you follow in real life
3) Know where you are in cyberspace
4) Respect other people’s time and bandwidth
5) Make yourself look good online
6) Share expert knowledge
7) Help keep flame wars under control
8) Respect other people’s privacy
9) Don’t abuse your power
10) Be forgiving of other people’s mistakes
While most guidelines for digital communication address at least some of these
expectations, they are still in flux and are violated by some members of any given
online community while others are restrained by them (Graham 2008). Neverthe-
less, even decades later, these foundational rules are still in play in most digital
encounters.
2.3.2. Digital (im)politeness terminology

Just as noted above in Section 2.2.1, terminology is a focus of much discussion in
research on (im)politeness, and adding a digital environment further expands the
terms used to describe (im)polite digital interactions. While there is digital termi-
nology to describe both polite and impolite behaviors, terms that describe negative
or impolite behaviors receive a great deal of discussion (e.g. trolling, spamming,
flaming and doxing). Hardaker, for example, defines a troller as “someone who

Authenticated
constructs the identity of sincerely wishing to be part of the group in question […]
but whose real intention(s) is/are to cause disruption and to trigger or exacerbate
conflict for the purposes of their own amusement” (2010: 237). Flaming involves
an overt attack on an individual or group. Spamming occurs when individuals
distribute empty or meaningless context to a widespread audience with the goal
of being disruptive. Finally, doxing occurs when a user distributes someone’s
personal contact information or details without that person’s knowledge, thereby
exposing that person to risk of f-t-f harm or harassment (see also Hardaker, Ch.
18, this volume).
These terms vary widely in both meaning and importance across digital con-
texts and communities. While in a book club, a troll might be a serious threat and
quite disruptive, while in an online gaming community spammers and trolls are an
expected and accepted (and in some ways encouraged and sanctioned) part of the
interaction. This shifts the import of what trolling actually means – whether the act
of trolling is defined by the impact it has (i.e. being disruptive) or by its packaging
(e.g. whether it contains racist content, which might or might not be offensive and
disruptive within the community in which it is said).
3. Important trends in digital (im)politeness research
Many of the other chapters in this volume will provide an in-depth exploration of
the pragmatics of interaction in separate modalities. The discussion in this chapter,
however, focuses on providing an overview of (im)politeness and related concepts
in these modalities (rather than an exploration of general pragmatic practice). A
key concept in this exploration is that of interpersonal pragmatics, which Locher
and Graham (2010) note should focus on “the relational aspect of interactions
between people that both affect and are affected by their understandings of culture,
society, and their own and others’ interpretations” (2010: 2, emphasis added). We
could argue that social media has its own culture and that examining it with digital
cultural factors (particularly as they are manifested in different digital contexts) in
mind is a worthy enterprise.
As noted above, the array of digital interactional settings and modalities cur-
rently operating in our society is far too vast to explore every one. For reasons
of space, here I will focus on (im)politeness that is both digitally-mediated and
interactive (and therefore emergent) in some way – and is therefore best suited
to illuminate the relational, interpersonal aspect of pragmatics described above.
Under this umbrella, for example, a YouTube video on cooking would not be an
appropriate focus since it is not in-and-of-itself interactive, while the discussion
about the video via the ‘comments’ section (which would fall under the category
of a bulletin or discussion board) would.

Authenticated
3.1. Email and discussion boards
Email and discussion board interaction is perhaps the largest area of research on
(im)politeness in digital contexts (for further discussion of message boards see
also Arendholz, Ch. 5, this volume). In these types of digital interactions (particu-
larly email), the audience is more likely to be known than in blogs/micro-blogs
and therefore the navigation of (im)politeness may be different than in other con-
texts. With regard to email, recent examinations of (im)politeness in teacher/stu-
dent and workplace interactions include Bolken and Holmgren (2012), Chejnová
(2014) and Merrison et al. (2012), all of which examine email interactions where
the audience is known and relatively small. Other research on email impoliteness
focuses on emails sent to large groups (e.g. Graham 2005, 2007a and Ho 2011)
which are more public and frequently focus more on establishing relationships
and communities rather than accomplishing specific tasks. Haugh (2010) and
O’Driscoll (2013) bridge the public and private arenas in examining private email
exchanges that are then interpreted on a public stage. The variety of influences on
(im)politeness and relational work in these studies illuminates the variability of
this platform and further research is warranted as the email context continues to
evolve in the ways it is used to perform relational work, particularly with regard to
public/private and task/relational intersections.
Discussion fora, like blogs (which will be discussed in greater depth below),
encompass a broad spectrum of media platforms and are often embedded within
other multimodal settings. We would place discussion comments about a YouTube
video, for example, in the category of discussion board posts rather than blogs
because, while the YouTube videos themselves could be classified as vlogs (vid-
eo-blogs), the comments that they generate more closely follow the format of dis-
cussion board postings – discussion about a blog (e.g. Garcés-Conejos Blitvich’s
2013 study on responses to a YouTube broadcast).
While the first linguistic research on discussion boards tended to focus on
community building and interactional practice, a sharp increase in recent research
(far too much to be listed exhaustively here) has incorporated both politeness and
impoliteness as specific research foci. Nishimura (2010) and Planchenault (2010),
for example, draw on politeness theory to examine Discussion Board interactions,
while other research has begun to focus more on impoliteness (e.g. Angouri and
Tseliga 2010; Dynel 2012; Hardaker 2010, 2015; Langlotz 2010; Lorenzo-Dus,
Garcés-Conejos Blitvich and Bou-Franch 2011; Richet 2013; Upadhyay 2010) and
flaming (e.g. Perelmutter 2013) to name a few.
Despite the large number of studies on these types of DC, however, there is
little focus on how the constraints of the medium itself may affect norms of (im)
politeness. Despite (or perhaps because of) the breadth of these formats, continued
research is needed that explores the technology itself and the ways that the con-
straints of the media influence how (im)politeness is accomplished.

Authenticated
3.2. Facebook, Instagram, and other social media
Research on social network sites (hereafter SNSs) such as Facebook and Insta-
gram has grown as these modalities have evolved and increased in popularity (see
also Eisenlauer, Ch. 9, this volume). Bolander and Locher (2015), Mak and Chui
(2014) and Theodoropoulou (2015), for example, explore the ways that identities
are negotiated through status updates and posted messages such as birthday wishes
(see also Locher and Bolander, Ch. 15, this volume). With regard to the ongoing
presence of (im)politeness evaluators as discussed in Section 1, Maíz-Arévalo’s
(2013) study of ‘liking’ is of particular interest since it examines the effects of a/
synchronicity, disembodiment and their relationships with (im)politeness (see also
Maíz-Arévalo, Ch. 21, this volume). These recent studies begin to address how
SNSs operate as a context for relational work, but it is important to remember that,
as is the case with all media, these platforms are constantly changing. Each time
there are changes to the privacy policies on Facebook, for example, the potential
audience for any given posting changes and this, in turn, alters users’ strategies for
navigating (im)politeness norms and performing relational work.
3.3. Short-form messages: Blogs/micro-blogs and SMS/IM

Yet another genre of digital interactions is that of blogs and micro-blogs such as
Instagram, Snapchat and Twitter (see also Heyd, Ch. 6, this volume, and Zappav-
igna, Ch. 8, this volume) and SMS/IM messages. In blogs and micro-blogs, users
post content in the form of pictures, videos, and/or text-based messages without
any clear confirmation of who their audience is, when their messages will be read,
or even if they will receive any response. This uncertainty leaves posters unsure
of how they might anticipate or conform to the (im)politeness expectations of
their audience. As Myers notes, “the vast majority of bloggers get along with just
a few people looking at their work. […] but the readers can be complete strangers
anywhere in the world” (2010: 77). He goes on to say that “Bloggers are often sur-
prisingly careful in how they present their opinions to an audience, using a range
of markers that suggest how they are making this statement” (2010: 95). Research
in this area is still growing as new media platforms continue to develop. In spite
of its prominence, for example, there is relatively little research on abuse and (im)
politeness on Twitter (see, however, Page 2014; Sifianou 2015). Since audience
awareness is critical to the examination of (im)politeness, this is a key factor in
any assessment of how (im)politeness works in this environment.
Yet another form of digital communication is SMS (Short Message Services)
and IM (Instant Messages). Like microblogging systems such as Twitter these plat-
forms produce short-form messages; unlike microblogging sites, however, they
are more likely to target known audiences and relatively small groups rather than
the hundreds or even thousands of unknown individuals who may receive a tweet.

Authenticated
As with other digital platforms already discussed, the ways that people use
SMS/IM have changed and expectations about what is normal (e.g. writing full
words v. using txt-speak) have changed as well. Current researchers, building on
previous research examining the oral/written blend of SMS/IM, have continued
in that trend as the capabilities of the digital platforms have evolved, but research
that specifically focuses on (im)politeness as a component of relational work in the
SMS/IM context is still somewhat lacking.
3.4. Virtual Realities and Virtual Worlds

While originally primarily text-based (e.g. MUDs), Virtual Worlds have evolved
as technology has advanced to include visual components as well. Platforms such
as Minecraft allow users to create computer-generated worlds (e.g. by designing
buildings, adding landscaping, etc. (alone or in collaboration with others) and cre-
ate characters/avatars that can operate within the world(s) they have created. An
offshoot of this type of platform – online gaming – allows users to create avatars
which perform in-game tasks and interact with others in attempting to achieve
them. Chat is a critical feature of interaction in virtual worlds, which are highly
synchronous, though they may be interaction- or goal-oriented.
While use of Virtual Worlds (especially online gaming platforms) is increas-
ing, there is not a great deal of research that examines (im)politeness and relational
work in these contexts. Pojanapunya and Jaroenkitboworn’s (2011) examination
of face in Second Life and Arendholz’s (2014) exploration of the online world The
Student Room are exceptions, but investigations that examine (im)politeness in
greater depth in these contexts would be fruitful.
Similarly, there is increasing interest across disciplines in research into online
gaming (which historically draws on virtual world platforms), but linguistic
research in this area almost all focuses on gaming as a socialization tool for L2
learners. There is very little linguistic research that addresses how players negoti-
ate interaction in highly synchronous and frequently multimodal settings (excep-
tions include Rusaw 2011 and Ensslin 2012), and this would be another fruitful
area for additional exploration.
4. Case study – FAQs and guidelines for polite/impolite/politic behavior
Gaming is an ideal venue for examining (im)politeness in a digital context for

multiple reasons. Gaming is highly prevalent in first-world countries and reaches
across multiple demographic categories. As it is put by Maher (2016): “By the
age of 21, the average young gamer will have logged thousands of hours playing
time. That fact alone makes dichotomies such as ‘digital world’ and ‘real world’
ring false – for many, game-playing is the real world” (emphasis added). In addi-

Authenticated
tion, video games are often associated with aggressive behavior. As Maher (2016)
notes:
Online gamers have a reputation for hostility. In a largely consequence-free environ-
ment inhabited mostly by anonymous and competitive young men, the antics can be
downright nasty. Players harass one another for not performing well and can cheat,
sabotage games and do any number of things to intentionally ruin the experience for
others — a practice that gamers refer to as griefing. Racist, sexist and homophobic
language is rampant; aggressors often threaten violence or urge a player to commit
suicide; and from time to time, the vitriol spills beyond the confines of the game. In the
notorious ‘gamergate’ controversy that erupted in late 2014, several women involved in
the gaming industry were subjected to a campaign of harassment, including invasions
of privacy and threats of death and rape. (http://www.nature.com/news/can-a-video-
game-company-tame-toxic-behaviour-1.19647)
In addition to the fact that, as a digitally-mediated genre, gaming is a setting where

(as noted above) assessments/declarations of (im)politeness are an ever-present
feature of interaction, in online gaming, (im)politeness is also encoded to some
degree into the games themselves – even the most benign-seeming game like ‘bub-
ble pop’ is based on the notion of hunting down bubbles and destroying them (one
assumes that most of us would classify destruction of someone/something as an
act that is at least impolite).
If we accept the premise that online gaming is a productive environment in
which to examine norms of (im)polite behavior, then there are several questions
we might ask: 1) what are the rules of (im)polite behavior in this context?, 2) who
determines what counts as (im)polite (i.e. who defines the rules for appropriate
behaviour), 3) once identified via these rules, how are impolite acts sanctioned,
and 4) how are polite acts rewarded? After a brief description of the data set and
context from which it was collected, I will address each of these questions in turn
below.
4.1. Twitch.tv and the genre of online gaming

Twitch.tv is an internet social video platform that is devoted to the broadcasting
of digital content by individuals worldwide and provides a forum for gamers to
establish communities. According to its website, as of May 2016, Twitch.tv had
over 100+ million unique viewers per month, 1.7+ million unique broadcasters per
month, and 106 minutes watched per person per day. Twitch streams are publically
available and can be watched by anyone, although viewers are invited to ‘sub-
scribe’ (by donating a small fee) to streamers that they particularly enjoy watch-
ing. According to Twitch, people broadcast over their channel(s) “to connect with
friends and fans over a shared love of games” and people watch others on Twitch
because “it is live social video that relies on audio and chat to enable broadcasters
and their audiences to interact about everything from pop culture to life in general

Authenticated
as they game” (https://www.twitch.tv/p/about). The content of streams includes

User Generated Content (e.g. live streams of gamers playing video games), edi-
torial programs, events and conventions with panels and press conferences, and
eSports competitions.4
Streamers use cameras and microphones to broadcast their content, which
either appears ‘Live’ or in some cases is recorded and available for replay. Game
streamers can have anywhere from zero viewers to hundreds of thousands of
viewers at any given time. Moreover, Twitch.tv manages game streams from hun-
dreds of games of all categories –1st person ‘shooter’ games such as Call of Duty,
MMORPGs (Massive-multiplayer online role playing games) such as World of
Warcraft, MOBAs (Multiplayer Online Battle Arenas) such as League of Leg-
ends, Virtual World games such as Minecraft, and even ‘non-digital’ games such
as poker.
For the purposes of this study, I will focus on the MOBA (Multiplayer Online
Battle Arena) game League of Legends. At the time of this writing, League of
Legends was the most-popular game on the market, having surpassed World of
Warcraft, with competitively ranked players on Asian, European and North Amer-
ican servers and millions of players and viewers worldwide.
4.2. League of Legends: Predictive rules and guidelines
4.2.1. The Summoner’s Code

The developer and publisher of League of Legends, Riot Games, makes the game
available for free download and maintains control of all game content. Users are
provided with a base set of characters and resources and can build upon these
resources with points and experience earned within the game. The learning curve
for League of Legends is high – it takes time to build expertise and proficiency
within the game and there are large amounts of resources available to help guide
beginners as they attempt to gain proficiency. In addition to Riot’s official instruc-
tions for getting started, they also provide the following “Summoner’s Code”,
which outlines expectations for behavior while playing the game:
1) Support your team
2) Drive constructive feedback
3) Facilitate civil discussion
4
eSports is a genre of competition that is structured just as traditional sports competitions
are, with video of the game itself, close-up television shots of the competitors’ faces
throughout the progression of the gameplay, and commentators offering their assess-
ments throughout the progression of the games regarding the strategies players choose,
the backgrounds and worldwide rankings and experience of individual players, etc.

Authenticated
4) Enjoy yourself, but not at anyone else’s expense

5) Build relationships
6) Show humility in victory, and grace in defeat
7) Be resolute, not indignant
8) Leave no newbie behind!
9) Lead by example
(emphasis added) http://gameinfo.na.leagueoflegends.com/en/game-info/get-started/sum-
moners-code/
The guide ends with the following dictum: “So remember! Stay positive, remain
calm, and keep to the code!”
Each of the guidelines listed above is followed by a paragraph explaining it in
greater detail, including an illustrative quote by a famous figure. The explanation
for rule #6 – Show humility in victory and grace in defeat – for example, begins
with a quote by Benjamin Franklin: “To be humble to superiors is duty, to equals
is courtesy, to inferiors is nobility”. Rule #5 – Build Relationships – includes the
observation that “it would behoove you to adopt a cordial demeanor and attempt
to make friends”. This (sometimes overly flowery) language seems to attempt to
create a tone of refinement that presupposes an attendance to the ‘wants and needs
of others’ that Brown and Levinson (1978/1987) identify as one of the key com-
ponents of politeness.
A quick overview of these guidelines also gives the impression of this game
and the game environment as overwhelmingly positive and polite. The guidelines
are filled with positive lexicon (support your team, build relationships, facilitate
civil discussion) and language that encourages empathy and support for other
players (show humility in victory and grace in defeat, leave no newbie behind),
and behave in a polite way (engage in civil discussion, don’t enjoy yourself at any-
one else’s expense, don’t be indignant). The fact that half of these guidelines use
positive (rather than negative) phrasing, I would argue, immediately swings the
tone to the ‘polite’ end of the polite/impolite spectrum (if we want to adopt Leech’s
2014 framework). By focusing on all of the positive things that players should do,
the game designers propose a code of conduct that is implicitly polite by being
respectful of other players within the confines of the gaming world.
4.2.2. Twitch.tv Code of Conduct

A second layer of behavioral guidelines is provided by Twitch.tv, the broadcaster
for the gaming streams. Twitch views its guidelines as a ‘living document’, provid-
ing frequent updates. The core content of each of their updated versions, however,
is the same – the key differences involve the tone (earlier editions tended to be
more playful and sarcastic/flippant, while later versions tend to be more brusque
and legalistic) and the level of detail provided. Twitch states explicitly that the

Authenticated
behaviors that could lead to sanctions “fall under a common sense philosophy
and apply to all video content, chat messages, channel content, profile content,
usernames, and any other content on our services” (https://www.twitch.tv/p/rules-
of-conduct).
In contrast to the Summoner’s Code described above, however, the Twitch.tv
guidelines adopt an approach that outlines behaviors that are NOT allowed –
instead of telling players what they should do (as in the Summoner’s Code above),
the Twitch rules tell participants what they should not do. Their guidelines are
more vague than the League of Legends code, and leave more interpretive license
to the user to determine what behaviors are (in)appropriate. In addition to legal
rules related to copyrights, rebroadcasting without permission, nudity, and other
similar legal violations, the Twitch behavioral rules (as of August 1, 2016) also
include rules against:
– Targeted harassment, threats, and violence against others, and
– Hate speech or other harassment
The more specific guidelines published in May 2015 described this this way:
We’re not going to tell you to watch the potty mouth, that’s between you and your
mother who doesn’t approve anyway. What we will tell you, however, is that if you
choose to use language or produce content that is racist, sexist, homophobic or falls
under generally accepted guidelines for hate speech you will disappear from Twitch.
(https://www.twitch.tv/p/rules-of-conduct)
The guidelines as amended in November 2015 simply say this: “Any content that
promotes or encourages discrimination, harassment, or violence based on race,
ethnicity, gender identity, sexual orientation, age, religion, or nationality is pro-
hibited”. The guidelines also prohibit “Spam, scams and other malicious content”.
Unlike The Summoner’s Code, the guidelines for behavior on Twitch focus on
what NOT to do, (rather than what to do) and tend to refer to actions with legal
ramifications such as harassment, threats and violence (although as noted in Sec-
tions 2.2.1 and 2.3.2, it is difficult to draw lines as to where something ceases to
be impolite and begins to be hateful or harassing).
The shift from the more casual description of (in)appropriate behavior from
May 2015 to the more legalistic guidelines in November 2015 is also noteworthy,
since it represents a shift from guidelines (suggestions?) to enforceable rules that
will have legal consequences. Those in charge have gone from “we’re not going
to tell you …” (implying that you should know better and should act appropriately
on your own) to “we’re telling you”.

Authenticated
4.2.3. Streamers’ Rules
A third layer of guidelines for appropriate behavior consists of rules created by the
broadcasters themselves. Twitch allows, indeed encourages, streamers to create
their own individual rules for their streams according to their comfort levels and
their assessments of their audiences. While not all streamers post detailed rules
for their viewers on their Twitch home pages, Twitch suggests that they at least
communicate their expectations to their moderators so that the mods know what
should be deleted/prohibited and what should be allowed.
One League of Legends streamer with a strong following on Twitch.tv posts
the following guidelines on his page:
Da Rulez
– No backseat gaming of any kind
– Don’t be a dick
– Don’t advertise anything. This includes other streams or your own personal business
– No racism
– No spreading false rumors
– No flaming
– Don’t one man spam
– Don’t fight with other viewers
– No ascii pictures (Especially the D)
– Don’t tell me how to run my stream. If you want to give constructive feedback do so
on my social medias
– Negative remarks may be timed out or banned at the mod’s discretion
Xanbot will automatically time out for the following:
– Messages that contain too many characters, don’t know the amount
– All caps
– Spamming twitch emotes (I think more than 5 on the same line)
– Colored messages
– Hyperlinks
(http://www.twitch.tv/wingsofdeath)
These rules are in alignment with the Twitch.tv rules, in that they forbid disruptive,
illegal, and hateful behaviors (i.e. ‘remember the human’). Advertising, posting
ascii pictures, and the specific types of posts that the Xanbot (a robot moderator)
is programmed to delete (all caps, messages that contain too many characters, too
many Twitch emotes, colored messages and hyperlinks) fall under the category of
standard spam in that, when posted over-and-over in quick succession they disrupt
the flow of the chat and disrupt the flow of the conversation. Other behaviors listed
above such as the rule against racism align with Twitch’s policy. The no flaming,
no spreading false rumors, don’t fight, no negative remarks, and the less-specific,
but much more descriptive “don’t be a dick” rules ban negative behaviors that
disturb the cohesiveness of the community by creating divisiveness.
Another streamer takes a more succinct approach, providing the following:

Authenticated
General Guidelines:
– No sexism, racism, homophobia of any kind
– No promotion of other channels, no slandering of other channels
– No spoilers5 of any kind (yes, this includes haha just trollin bro it was a fake spoiler)
– Respect the moderators. That means why are you arguing with the mods about your
rights to post porn in twitchchat? Hint: No, Michael is not going to unmod them
because you got banned. Yes, your ban length may be increased.
– Moobot is in charge of auto banphrase moderation. Moobot is a bot. He only has
powers to timeout. Nothing personal … kid.
(https://www.twitch.tv/imaqtpie)
Whereas the Summoner’s Code highlights polite/positive behaviors that focus on

protecting the game experience of the players, and the Twitch guidelines focus on
protecting the legal responsibilities of the organization, both of these streamers
bring these two together and provide guidelines that are designed to meet the legal
demands of Twitch (hence references to no advertising and spoilers) while also
creating a positive environment for their viewers in order to enhance their expe-
rience and thereby ensure the success of the stream. Despite the fact that specific
definitions vary, these guidelines highlight two types of problematic behaviors:
1) offensive content (manifested in racism, sexism, etc.) and 2) spam – both of
which are harmful to the community.
4.3. ‘Policing’ of impoliteness – Breaking ‘Da Rulez’

In any situation where there are published guidelines for behavior, there is an
expectation that these ‘rules’ will be enforced in some way. In Twitch, this process
is handled by moderators (mods), both human and robot (also known as botmods
or bots). Mods are chosen by the streamers themselves, and Twitch offers the fol-
lowing suggestions for finding viewers who are suitable mod candidates:
What should you look for in a moderator?
When choosing a moderator, you should have a general idea of what you consider
offensive and what you consider spam.6 In some channels, large capitalized text blocks
are considered spam, as they make others’ text very hard to distinguish. Other chan-
nels, particularly channels with slow moving chat, occasional text blocks are less intru-
sive. This is only one example of spam, but there are many others that you should
consider:
5
In some cases, gamers are allowed early access to games by the game developers before
they are released to the public. In such cases, those with early access are expected to
refrain from giving spoilers – details about the game (such as the storyline) – so that
people can experience and discover the game for themselves when it is released to the
public.
6
Emphasis added.

Authenticated
– Copy pasted text, often large

– Links
– Geometric block images
– Repeated text (texttexttexttexttexttexttext, lollollollollollollollollollollollol)
– Excessive use of emoticons
– Excessive symbol usage (are you serious?!?!?!?!?? Wow!!!!!!!!!!!!!!!!!!!!!)
When choosing moderators, make sure you communicate what you want removed, and
what is fine.
(http://help.twitch.tv/customer/protal/articles/1360598-moderation-guide)
These guidelines highlight two of the most prominent themes: no ‘offensive’ con-
tent and no spam. The question remains, however, how these problematic behav-
iors are identified and controlled, particularly in light of the fact that streamers are
encouraged to make their own definitions.
4.3.1. BotMods v. human mods

Streamers have two systems in place for regulating behavior in their streams:
human mods and botmods. Botmods by nature must rely on a predictive framework
for impoliteness – they must be programmed ahead of time to recognize posts that
break the rules of politeness and delete them. For a botmod to work, someone will
need to write code that instructs it on how to recognize impolite/offensive/prob-
lematic messages. This requires a definition, however, that is specific enough to be
enforceable by a robot that is not capable of assessing the nuances of impoliteness
in context. A botmod can be programmed to recognize and delete messages that
have more than a certain number of emotes or symbols (as noted above in both
the Twitch Rules and Da Rulez), for example, and can reliably identify and delete
hyperlinks as well as large blocks of ascii text (also frequently known as ‘copypas-
tas’). They can also (in theory) be programmed to recognize and delete swearing.
You could, for example, compile a list of swear words and program the botmod to
delete all messages that include them. This predictive model is flawed, however,
since users can manipulate language in ways that will sidestep the coding while
still making the meaning clear (e.g. substituting fcuk for the word fuck or s*it for
the word shit) and it would be impossible to anticipate and program a botmod to
delete all of the possible combinations.
While botmods can provide some amount of regulation and enforcement
(e.g. in identifying hyperlinks), streamers must also rely on human mods to recog-
nize content that could be classified as ‘offensive’ according to the definitions they
(the streamers) have identified. The following example illustrates how messages
that do not contain objectively identifiable markers (the kind that botmods could
recognize) that would mark a given post as a violation, can still contain offensive
content.

Authenticated
Example 1
1 igivenofuks: poe7 is for sihtkids get this trash off twitch tv you fukn loser […]
2 r3cp3ct: ban this dude […]
3 igivenofuks: read my name see how i feel about ban in this siht chat
4 igivenofuks: fukboy […]
5 r3cp3ct: o im a fukboi […]
6 igivenofuks: this game is for retarded pussies who seek no challenge at all […]
7 igivenofuks: and have zero gaming skills
8 Dormantg: someones on their period
9 Jekke888: @igivenofuks, I could beat the *** out of you with no challenge at all. […]
10 r3cp3ct: jekke888 don’t bother, he is trolling
11 Jekke888: I’m aware […]
12 igivenofuks: oh my god guys look at him hes proving valid points and don’t agree with
him he is a troll wahh wahh
13 igivenofuks: fk urself […]
14 Dormantg: you must be 12 […]
15 igivenofuks: age joke great comeback […]
16 MisterMycelium: TT.TT @igivenofuks next time ask daddy to be a little more gentle
when he gets his jolly’s off
17 Jekke888: I don’t give a *** if you’re a troll or not if you are not a troll you are stupid
and if you are a troll you are stupid congratualtions you are stupid either way […]
18 igivenofuks: anyone who plays this game should be shot […]
19 igivenofuks: end of story
20 <igivenofuks has been banned from talking>
21 jekke888: end of story
In this exchange, igivenofks’ posts are clear examples of offensive messages that
are interpreted as trolling, an attempt to provoke a reaction and get attention (see
Hardaker, Ch. 18, this volume for further discussion of trolling). The messages,
however, do not contain predictable features that a botmod might use to identify
them. The terms used for swearing have letters that are inverted or deleted (e.g.
siht for shit and fuk for fuck), and it would be impossible to anticipate all of the
potential spelling or other ways that trollers and spammers could manipulate the
texts so that you could program a botmod to identify and delete them. Similarly,
igivenofuks’ statement in line 6, “this game is for retarded pussies”, is not only
offensive but also highly context-dependent. While a streamer might attempt to
program a botmod to delete words or phrases that s/he finds particularly offensive,
it would be impossible to anticipate every potentially problematic or impolite term.
Given this, streamers must rely on human mods to determine, based on the con-
text, which utterances are disruptive, offensive, impolite or constitute hate speech.
In the following exchange, posted during the open chat of a female streamer, the
use of the term ‘pussy’ has multiple meanings:
7
poe = Path of Exile – an online role playing game (RPG).

Authenticated
Example 2
1 Wozleee:[to the female streamer]: can I see your pussy
2 EliteShogun: @Wozlee: can I see yours? […]
3 Lolnignog: you can see my pussy
4 Lolnignog: it’s in my backyard
In this exchange, Wozleee directs an offensive message to the female streamer.

The responses come from other chat participants who challenge Wozleee’s state-
ments by appropriating his language. Of particular interest here is the fact that
Wozleee and Lolnignog use the same term. Wozlee, however, is banned from the
chat temporarily, while Lolnignog is not. In this case, Lolnignog’s use of the term
pussy challenges Wozleee by highjacking the word and redefining it with a more
socially acceptable meaning (i.e. by shifting the reference from a female body part
to a more benign and less personally offensive reference to a pet cat). This type of
exchange would be impossible for a botmod to interpret. Even if a streamer were
to compile a list of offensive words and program the botmods to delete any mes-
sage that contained them, in cases like this a botmod would be unable to discern
the difference between the offensive use and the benign use of the term – for that,
streamers must rely on human moderators who can judge (im)politeness in context
and determine what, in fact, is problematic and what is not.
Similarly, since spam is prohibited because it is disruptive, mods must judge
what counts as a disruption and how disruptive a given behavior is before they
decide whether to discipline the participant. Posting a copypasta message once, for
example, is unlikely to be interpreted as highly disruptive, but posting the same
message multiple times in quick succession makes it difficult for other participants
to continue the conversation because the spam is ‘clogging up’ all of the available
space. The following exchange took place after a participant called another partic-
ipant a black gay fat man.
Example 3
internet kids thnks eberything is funny and that they can make fun of everything. [sic]
2 Luquitasc: @Aquafresht is that a copy pasta?
3 Aquafresht: no that is 100 % me
internet kids thnks eberything is funny and that they can make fun of everything. 6 [sic]
[4 messages]

Authenticated
internet kids thnks eberything is funny and that they can make fun of everything. pls no
copy pasterino
[4 messages]
internet kids thnks eberything is funny and that they can make fun of everything. 6
internet kids thnks eberything is funny and that they can make fun of everything.
[…]
Over the course of 2 minutes, this message (with only minor variations tagged
onto the end) is reposted a total of 10 times by three different users. None of the
three were timed out or otherwise punished, despite the fact that repost of this
(rather long) message could certainly be described as spam since it doesn’t convey
any new information and therefore just ‘takes up space’ in the chat stream. The
fact that there are intervening messages between the (re)posts of the spam seems to
indicate that the flow of the chat is not threatened too much, however – the conver-
sation continues around these copied and reposted messages. Just as in Examples 1
and 2 above, then, the mods must make a judgment about whether these messages
count as spam (i.e. they must identify impolite behavior) and decide whether pun-
ishment is needed for the good of the community. In this way, human mods must
make a judgment call as to whether a particular user should be banned by exam-
ining the content of the message, weighing its effects within the unfolding chat
stream and determine if it is something that could or should be banned based on
the streamer’s definitions of what is problematic (which as Twitch notes, should be
conveyed to the mods by the streamer prior to the chat). In an ideal case, these two
regulatory groups – botmods and human mods – should be able to work together
to limit impolite/counterproductive behavior. As we will see below, however, there
are still flaws in implementation of this system.
4.3.2. Flaws in the system

While one might expect the combination of botmods and human mods to cre-
ate a symbiotic balance in which the chat would flow smoothly and problematic
users could be controlled before becoming disruptive, the system, in fact, is inade-
quate to accomplish this goal for two reasons: 1) the fact that decisions about what
counts as impolite must be balanced against the ‘needs’ of the stream in order to
establish community and build/maintain viewership, and 2) community member-
ship is often marked by impoliteness and violations of the rules (just as has been
shown in many other on- and off-line communities; e.g. Graham 2008).
In the following exchange, a streamer is asked for information about a key-
board command to employ during a particular game to make something blow up.

Authenticated
He responds by telling the viewers (via audio mic rather than typing in chat) to
type QWEQEQ. Fifteen of the next 34 messages contain some variation of this
sequence of letters. Approximately 45 seconds after the messages begin, one of
the mods begins to timeout users (presumably because he interprets their postings
as spam – repetitive and devoid of meaningful content). Twenty-five seconds later,
the streamer realizes that users have been banned and responds with the following:
Example 4
1 Wingsofdeath [streamer]: who the ***
2 Wingsofdeath: Is timing them out
[1 message]
3 Wingsofdeath: holy ***
[5 messages]
4 Wingsofdeath: 3 mods
5 Wingsofdeath: one of you
[1 message]
6 Wingsofdeath: speak up now
[1 message]
7 eosmeep: sorry that was me
8 Wingsofdeath: or youre all gone
9 Wingsofdeath: ok
10 eosmeep: won’t happen again
[2 messages]
11 Wingsofdeath: WAT IS WRONG WITH U
[1 message]
12 Wingsofdeath: seriously
[4 messages]
13 Wingsofdeath: Were u completely absent
[2 messages]
14 Wingsofdeath: for the time I was talking about modding
[84 messages]
15 eosmeep: not sure how to answer you wings when chat is going too fast and you
probably wont read what i type
[5 messages]
16 eosmeep: I hit timeouts instead of bans on bttv
[43 messages]
17 <jtv removes channel operator (i.e. mod) status from eosmeep> (5 minutes after origi-
nal bans by eosmeep)
In this case, the problematic messages looked like spam – they showed a sequence
of letters without any identifiable meaning (i.e. they looked like random characters
strung together) and many were typed in all caps (one of the markers of spam as
noted above). While timing out a handful of users may seem like a minor action, in
this case the streamer, Wingsofdeath, was extremely upset, because if participants
are timed out (especially unjustly) then they won’t continue watching the stream.
In this case, the messages that had markers of spam were misinterpreted by one

Authenticated
of the mods and his response created a situation that the streamer felt might dis-
courage viewers from continuing to watch. This balance between allowing users
to post what they want to in order to keep them watching versus the need to con-
trol ‘empty’ content that disrupts the flow of the chat was derailed. In such cases,
rather than making controlling spam the priority, the streamer may decide to allow
a certain amount of spam so that his/her viewers will keep watching.
In addition to streamers allowing spam up to a certain point in order to keep
viewership, there are also instances where streamers take the idea a step further. In
the case of imaqtpie whose rules are posted above, there is an automated system
that tracks participation and not only allows problematic behavior like spam, it even
sanctions it and uses it as a reward for watching. For every minute users stay in the
chat room, they are awarded points (subscribers get more points than non-subscrib-
ers). They can then spend these points (which are tracked by a botmod) to buy the
right to post spam. For a certain number of points, users can post ascii pictures like
the one below as well as long text paragraphs where the content isn’t relevant to the
ongoing discussion in the chat stream as in examples 5 and 6 below:
Example 5
░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
░░░░░░░░▄▄▄▀▀▀▄▄███▄
░░░░░▄▀▀░░░░░░░▐░▀██▌
░░░▄▀░░░░▄▄███░▌▀▀░▀█
░░▄█░░▄▀▀▒▒▒▒▒▄▐░░░░█▌
░▐█▀▄▀▄▄▄▄▀▀▀▀▌░░░░░▐█▄
░▌▄▄▀▀░░░░░░░░▌░░░░▄███████▄
░░░░░░░░░░░░░▐░░░░▐██████████▄
░░░░░le░░░░░░░▐░░░░▐██████████▄
░░░░toucan░░░░░░▀▄░░░▐█████████▄
░░░░░░has░░░░░░░░▀▄▄██████████▄
░░░░░arrived░░░░░░░░░░░░█▀██████
Example 6
13:37 wowsobot: “Hello Miguel Santander, this is your cat, Mellow Cat, I heard that you
were complaining about my excesive meowing. I dont complain about having to hear you
talk to yourself for 8 hours while playing that stupid game so I expect the same. Come here
and feed me scrub”
13:37 wowsobot: qtpie, this is your dad here. i just wanted to say how proud of you i am
son, always hiding in your room playing your video games, talking to your internet friends.
your mom and i think that’s swell and we just wanted to let you know we love you. pls no
copy dadderino
Imaqtpie’s reward points allow for users (and subscribers have an advantage) to
break the rules in a sanctioned way. Chat participants with enough points can
‘buy’ access to spam that includes copy pastas – including blocky ascii images and
repeated text as well as links (which violate the rules set forth by Twitch.tv and

Authenticated
the streamer himself). The streamer himself, for that matter, breaks his own rules.
While both his rules and the Twitch.tv rules prohibit sexism, for example, every
time someone subscribes to his channel, he reads the subscriber’s name aloud and
says “Welcome to the big dick club”. Despite the fact that some would argue that
this statement is inherently sexist, the streamer in this case is not only allowing this
violation of the rules but is participating in the violations himself.
4.4. Remaining questions

So where does this leave us? In DC there are attempts to develop a predictive
model of impoliteness which can be used to control for some aspects of impolite
behavior. Even in non-gaming settings, there are spam filters (which most would
agree are imperfect) and an increasing number of tools (like the Clutter feature in
Outlook email that takes messages that it perceives as non-urgent and puts them
into a separate folder) that are designed to help identify and filter out unwanted
digital information. But these are doomed to failure and must be checked by
humans – everyone has found an email that ‘slipped through’ the spam filter or
identified a spammer who was able to slip past a mod by adjusting the spelling or
presentation of a problematic phrase. The difficulty with this is first defining (so
that effective codes and filters can be developed) inappropriate remarks and then
identifying impolite behavior in the moment quickly enough so that you can delete
messages before they become disruptive – all without discouraging viewership.
The tension between attempts at enforcement and the goals of the streams
leads to a situation where (once again) (im)politeness is different in practice than
in theory and cannot be theorized to be used in a predictive way. This brings us
back to relational work. As noted in Graham (2008 and 2016), impoliteness and
breaking the rules for behavior is used as a means of marking community and
in-group membership. In this case, the rule-breaking is even encouraged; as view-
ers show greater commitment to the group they are allowed (even encouraged) to
break the rules.
It is clear that digital communication has become an ever-present (and increas-

ingly complex) part of our daily interactions and is therefore a setting where we
will continue to negotiate and revise our expectations of (im)polite behavior. As
the norms of interaction morph and change as new technologies and platforms
develop, it would be a worthy area of examination to explore the ways that early
concepts of Netiquette (which as discussed are still in play today) have changed
to meet the demands of new media and correlated new norms and expectations of
(im)politeness.

Authenticated
Another potential area for further investigation is the way that computer-gen-
erated controls are used in conjunction with humans (as with bots and human
moderators) to construct and regulate the ways that behavior is assessed and con-
trolled. As automated systems become more complex, we might ask ourselves
whether it will ever be possible for a bot to have the ability to recognize (im)polite
behaviors in digital contexts and how their shortcomings in this area relate to the
feasibility of a second-order model of (im)politeness.
The largest potential area for further research, however, is in the area of mul-
timodal interaction. Digital communication becomes more multimodal daily – we
can carry on a conversation via video link on Skype while simultaneously typing
messages and sending hyperlinks to the same user. In this context, we are employ-
ing audio capabilities in conjunction with video and text-based communication,
each of which draws on different communicative strategies and expectations
(including assessments of identity and (im)politeness). In the case study above,
for example (Example 4), a human moderator timed participants out for posting
messages that looked like spam. Those (written) messages, however, were posted
in response to an audio comment by the streamer. In this case, looking only at the
chat stream, one might misinterpret the messages as spam since they look like
this problematic behavior when they are taken out of the context of the stream-
er’s comments (which were transmitted via audio rather than writing). Linguistic
research on multimodality is limited (although see Naper 2011), but understanding
the ways that multiple modes of communication are intertwined in cases such as
this can illuminate factors of contextuality that can influence the identification
and interpretation of (im)politeness. The ability of multimodal platforms to bring
features of DC together such as a/synchronicity, relative anonymity, and public/
private communication in this way cannot help but alter our strategies in satisfying
the demands and expectations of (im)polite behavior, and is therefore, I would
argue, the most promising area for future research on (im)politeness.
References

2010 “you HAVE NO IDEA WHAT YOU ARE TALKING ABOUT!”: From
e-disagreement to e-impoliteness in two online fora. Journal of Politeness
Research 6: 57–82.
Arendholz, Jenny
Baym, Nancy K.
2015 Personal Connections in the Digital Age. 2nd ed. Malden, MA: Polity Press.

Authenticated

2015 “Peter is a dumb nut”: Status updates and reactions to them as ‘Acts of Posi-
boyd, danah and Eszter Hargiatti
2010 Facebook privacy settings: Who cares?. First Monday 15(8). doi: 10.5210/
fm.v15i8.3086.
Bousfield, Derek
2010 Researching impoliteness and rudeness: Issues and definitions. In: Miriam
Locher and Sage Graham (eds.), Interpersonal Pragmatics, 101–134. Berlin/
Boston: de Gruyter.
Brown, Penelope and Stephen Levinson
1978/1987 Politeness: Some Universals in Language Usage. Cambridge: Cambridge
University Press.
Chejnová, Pavla
Culpeper, Jonathan
2008 Reflections on impoliteness, relational work, and power. In: Derek Bousfield
and Miriam Locher (eds.), Impoliteness in Language: Studies on Its Interplay
and Power in Theory and Practice, 17–44. Berlin/Boston: de Gruyter.
Culpeper, Jonathan
2011 Impoliteness: Using Language to Cause Offence. Cambridge: Cambridge Uni-
versity Press.
Dynel, Marta
2012 Swearing methodologically: The (im)politeness of expletives in anonymous
commentaries on YouTube. Journal of English Studies 10: 25–50.
Ensslin, Astrid
2012 The Language of Online Gaming. Basingstoke, UK: Palgrave.
2010 Introduction: The status quo and quo vadis of impoliteness research. Intercul-
tural Pragmatics 7(4): 535–559.
2013 Relational work in anonymous asynchronous communication: A study of (dis)
affiliation in YouTube. In: Istvan Kecskes and Jesus Romero-Trillo (eds.),
Research Trends in Intercultural Pragmatics, 343–366. Berlin/Boston: de
Gruyter.
Graham, Sage L.
2005 Cyberparish: Gendered identity construction in an online religious commu-
nity. In: Jule Allyson (ed.), Gender and the Language of Religion, 133–150.
Basingstoke, UK: Palgrave Macmillan.
Graham, Sage L.
2007a ‘Do Unto Others’: Gender and the construction of a ‘good christian’ identity
in an e-community. In: Jule Allyson (ed.), Language and Religious Identity:
Women in Discourse, 73–103. Basingstoke, UK: Palgrave Macmillan.
2007b Disagreeing to agree: Conflict, (im)politeness and identity in a computer-me-

Authenticated
Graham, Sage L.
nity of practice. In: Derek Bousfield and Miriam Locher (eds.), Impoliteness
in Language: Studies on its Interplay with Power in Theory and Practice,
Graham, Sage L.
2016 Relationality, friendship and identity in digital communication. In: Alexandra
Georgakopoulou and Tereza Spilioti (eds.), The Routledge Handbook of Lan-
guage and Digital Communication, 305–320. New York: Routledge.
Graham, Sage L. and Claire Hardaker
2017 (Im)politeness in digital communication. In: Jonathan Culpeper, Michael
Haugh and Daniel Kadar (eds.), Handbook of Linguistic Impoliteness, 785–
814. Basingstoke, UK: Palgrave Macmillan.
Hardaker, Claire
Hardaker, Claire
2015 “I refuse to respond to this obvious troll.”: An overview of (perceived) troll-
ing. Corpora 10 (2): 201–229.
Haugh, Michael
Erlich, Miriam Meyerhoff and Janet Holmes (eds.), Handbook of Language, Gen-
der and Sexuality, 567–586. Hoboken, NJ: John Wiley and Sons. doi: 10.1002/
9781118584248,ch29.
Ho, Victor
2011 A discourse-based study of three communities of practice: How members
maintain a harmonious relationship while threatening each other’s face
via email. Discourse Studies, 13(3): 299–326. http://search.proquest.com/
docview/925725317?accountid=14657.
Hymes, Dell
Kádár, Daniel and Michael Haugh
2013 Understanding Politeness. Cambridge: Cambridge University Press.
Lakoff, Robin
1973 The logic of politeness, or minding your p’s and q’s. Chicago Linguistics Soci-
ety 9: 292–305.
Langlotz, Andreas
2010 Social cognition. In: Miriam Locher and Sage L. Graham (eds.), Interpersonal

Authenticated

Lea, Martin, Russell Spears and Daphne de Groot
1991 Knowing me, knowing you: Anonymity effects on social identity processes
within groups. Personality and Social Psychology Bulletin 27: 526–537.
Leech, Geoffrey
Leech, Geoffrey
Locher, Miriam A.
2006 Polite behavior within relational work: The discursive approach to politeness.
Multilingua 25(3): 249–267.
Locher, Miriam A.
2010 Introduction: Politeness and impoliteness in computer mediated communica-
tion. Journal of Politeness Research 6(1): 1–5.
Locher, Miriam A.
2012 Politeness. In: Carol E. Chapelle (ed.), The Encyclopedia of Applied Linguis-
2005 Politeness theory and relational work. Journal of Politeness Research 1: 9–33.
Locher, Miriam, Brook Bolander and Nicole Höhn
2015 Introducing relational work in Facebook and discussion boards. Pragmatics,
25(1): 1–21.
2011 Online-polylogs and impoliteness: The case of postings sent in response to
the Obama Reggaeton YouTube video. Journal of Pragmatics, 43(10): 2578–
2593.
Maher, Brenden
2016 Can a video game company tame toxic behaviour? Nature. http://www.nature.
com/news/can-a-video-game-company-tame-toxic-behaviour-1.19647 doi:
10.1038/531568a.
2013 ‘Just Click “Like”’: Computer-mediated responses to Spanish compliments.
2014 Impoliteness in Facebook status updates: Strategic talk among colleagues
‘outside’ the workplace. Text and Talk 34(2): 165–185.
2014 Networked privacy: how teenagers negotiate context in social media. New
Media and Society. doi: 10.1177/1461444814543995.
tion in Britain and Australia. Journal of Pragmatics 44(9): 1077–1098.
Moderation Guide – Broadcaster
2013 http://help.twitch.tv/customer/portal/articles/1360598-moderation-guide---
broadcaster [29 August 2015].

Authenticated
Myers, Greg
2010 The Discourse of Blogs and Wikis. New York: Continuum Press.
Naper, Ida
2011 Conversation in a multimodal 3D virtual environment. Language@Internet 8:
1–27.
Nishimura, Yukiko
2010 Impoliteness in Japanese BBS interactions: Observations from message
exchanges in two online communities. Journal of Politeness Research 6:
33–55.
O’Driscoll, Jim
2013 Situational transformations: The offensive-izing of an email message and the
publicization of offensiveness. Pragmatics and Society 4(3): 369–387.
Page, Ruth
62: 30–45.
Perelmutter, Renee
2013 Klassika Zhanra: The flamewar as a genre in the Russian blogosphere. Journal
of Pragmatics 45(1): 74–89.
2010 Virtual community and politeness: The use of female markers of identity
and solidarity in a transvestites’ website. Journal of Politeness Research 6:
83–103.
Pojanapunya, Punjaporn and Kandaporn Jaroenkitboworn
2011 How to say ‘goodbye’ in Second Life. Journal of Pragmatics 43(14): 3591–
3602.
Richet, Bertrand
2013 Fanning the flames?: A study of insult forums on the internet. In: Denis Jamet
and Manuel Jobert (eds.), Aspects of Linguistic Impoliteness, 223–231. New-
castle: Cambridge Scholars.
Rusaw, Erin
2011 Language and social interaction in the vitual space of World of Warcraft. Stud-
ies in the Linguistic Sciences 36: 66–88.
Schwämmlein, Eva and Katrin Wodzicki
2012 What to tell about me?: Self-presentation in online communities. Journal of
Shea, Virginia
1994 Netiquette. http://www.albion.com/netiquette/corerules.html.
Sifianou, Maria
Terkourafi, Marina
2008 Toward a unified theory of politeness, impoliteness, and rudeness. In: Derek
Bousfield and Miriam A. Locher (eds.), Impoliteness in Language: Studies
on its Interplay with Power and Theory in Practice, 45–74. Berlin/Boston: de
Gruyter.
23–45.

Authenticated
Twitch.tv Rules of Conduct. https://www.twitch.tv/p/rules-of-conduct.

Upadhyay, Shiv
Politeness Research 6: 105–127.
Vandergriff, Ilona
communication (CMC) cues. Journal of Pragmatics 51: 1–12.
Vinagre, Margarita
2008 Politeness strategies in collaborative email exchanges. Computers and Educa-
tion 50(3): 1022–1036.
Watts, Richard J.
Watts, Richard J.
2005 Linguistic politeness and politic verbal behavior: Reconsidering claims for
universality. In: Richard Watts, Sachiko Ide and Konrad Ehlich (eds.), Polite-
ness in Language, 21–42. 2nd ed. Berlin/Boston: de Gruyter.

Authenticated
Authenticated
18. Flaming and trolling
Claire Hardaker
Abstract: In the last two decades, pragmatic explorations of flaming and troll-
ing in computer-mediated communication have gained momentum. However,
although there is little doubt that “attacks, assaults and contemptuous remarks”
(Jucker and Taavitsainen 2000: 73) have always been commonplace, some
researchers (cf. Moor, Heuvelman and Verleur 2010; Nitin, Bansal and Kanzashi
2011) have recently argued that certain types of conflict, such as flaming and troll-
ing are particularly native to social media. To this end, this chapter aims to provide
new insights into current pragmatic research into flaming and trolling, including
how these terms are defined and deployed, case studies that illuminate how these
behaviors are accounted for by existing literature, and the current challenges that
face these fields and their development into the future.
1. Introduction
In recent years, the media has pounced on what I have, in previous research, loosely
termed negatively marked online behaviors (NMOB) (Hardaker 2010, 2013, 2015;
Hardaker and McGlashan 2016). NMOB is behavior that, for whatever reason,
others consider to fall outside of the bounds of social decorum for that particular
online interaction. It could involve something as mild as raising a topic that falls
outside of the scope of a given forum, or more seriously, posting crude jokes on a
tribute page set up to honor a deceased person. At the criminal end of the scale, it
might consist of repeatedly sending threats via emails or text messages, or sharing
someone’s personal, private details on public platforms.
The media’s interest in NMOB is unsurprising. The sheer breadth, depth, and
range of unpleasant online behaviors provide an endless source of attention-grab-
bing headlines and exposés. Likewise, the people behind that behavior, whether
characterized as trolls, cyberstalkers, or online predators, make for simplistic,
almost cartoon-like villains that the vast majority of media-consumers can readily
position themselves against.
This fervor of interest in online abuse, however, though so pervasive now, is a
relatively new phenomenon. The internet could be described as well-established
across numerous countries by the 1990s, yet as late as 2010, antisocial online
behavior was virtually ignored by the news, academia, and even to an extent, the
online platforms that were home to those behaviors themselves. Even at the time
of writing, there is still surprisingly little linguistic research on NMOB, and what
DOI 10.1515/9783110431070-018
Authenticated
494 Claire Hardaker
interest we can find tends to be distributed unevenly. We can readily find interest
in topics such as spam (e.g. Barron 2006; Stivale 1997), cyberbullying (e.g. Strom
and Strom 2005; Topçu, Erdur-Baker and Çapa-Aydin 2008), cyberstalking (e.g.
Bocij 2004; Whitty 2004), aggressive video games (e.g. Dill and Dill 1998; Scott
1995; van Schie and Wiegman 1997) and computer-related depression (e.g. Kraut
et al. 1998). Likewise, flaming had arguably arrived by the turn of the millennium,
with interest from fields as diverse as psychology (Collins 1992; Lea et al. 1992),
linguistics (Arendholz 2013; Herring 1994; Jucker and Taavitsainen 2000), media
and cultural studies (Millard 1997; O’Sullivan and Flanagin 2003), sociology (Lee
2005) and information science (Kayany 1998).
Trolling, however, has struggled to gain validity as field of interest until as late
as 2010. For instance, by the end of the 1990s, there had been a handful of articles
on the politics (Tepper 1997) and deceit (Donath 1999) found in online practices.
By 2005 new research was being published about online deception (Placks 2003;
Utz 2005; Zhou et al. 2004), but very little directly about trolling (see, however
Herring et al. 2002). Finally, by 2010, research dealing specifically with troll-
ing (e.g. Hardaker 2010; Shachaf and Hara 2010; Shin 2008) started to appear
in greater numbers. And by 2015, the growth had become exponential, spanning
fields as diverse as Artificial Intelligence (Dlala et al. 2014), computing (Cheng,
Danescu-Niculescu-Mizil and Leskovec 2015; Fichman and Sanfilippo 2015),
communication studies (Brabazon 2012), cultural studies (Phillips 2015), media
(Binns 2011; McCosker 2014), politics (Virkar 2014), psychology (Buckels, Trap-
nell and Paulhus 2014; Maltby et al. 2015), public relations (Weckerle 2013) and
of course, linguistics (Hardaker 2013, 2015; Hardaker and McGlashan 2016; Hop-
kinson 2013).
Why such reluctance to confront what might be viewed as an online epidemic?
One problem may be a lingering sense that computer-mediated communication
(CMC) is somehow not as worthy a subject of investigation as offline interaction,
and some research has compounded this devaluation by dichotomising virtual and
real-life activity (e.g. Baym 1996: 342; Chiou 2006: 547; Strom and Strom 2005:
41). A secondary problem from the academic perspective may be with
[…] the variability in the perceptions of norms and expectations underlying evaluations
of behaviour as polite, impolite, over-polite and so on, and thus inevitably discursive
dispute or argumentativity in relation to evaluations of im/politeness in interaction. Yet
with the exception of work by Locher (2006) and Graham (2007, 2008), there has been
little research on im/politeness in various forms of computer-mediated communication
from this perspective. (Haugh 2010: 8)
These problems are not limited to the analyst. CMC, by its very nature, is both rel-
atively new and changing extremely quickly. Analysts and users alike may strug-
gle to evaluate the appropriacy of online utterances in light of their own norms and
expectations. With this in mind, this chapter moves next into a brief overview of

Authenticated
Flaming and trolling 495
CMC and the peculiarities that this context affords. The current state of the art of
academic research on flaming and trolling in CMC in general and social media in
particular is discussed in Sections 3 and 4 respectively. Finally, Section 5 looks
towards promising future directions.
2. Computer-mediated communication
The overarching term computer-mediated communication is not necessarily clear-

cut, so it is worth briefly defining. Firstly, computer may intuitively seem to index
the prototypical desktop machine, but it can also refer to smartphones, smart-
watches, tablets, games consoles, and more besides. Our interest is also in com-
munication via computers, rather than with or from the computer, as covered by
fields such as human-computer interaction (HCI) and artificial intelligence (AI).
In other words, for the purposes of this chapter, we will define CMC as commu-
nication occurring via any mediating, networked technology (December 1997: 5;
Ferris 1997; Herring 1996, 2003: 612) regardless of where that communication
falls on the scale from almost instantaneous (synchronous CMC) to delayed (asyn-
chronous CMC).
In early research, CMC was generally viewed from a technologically deter-
ministic (Bolter and Grusin 1998) perspective. Strong technological determinism
positions technology as in control, particularly of its place and development in
culture and society. As a very small example of this, the following headline, How
Facebook ruined my holiday: The Internet and mobile technology make it increas-
ingly difficult to switch off (Kirsh 2012) casts Facebook, the internet, and mobile
technology in the role of autonomously self-aware, and even malignant actor,
whilst the user is merely the passive victim.
Researchers such as Kraut et al. (1992: 375) argued that CMC not only
impeded the formation of new offline social ties, but even damaged those already
in existence. In later research, they suggested that, “greater use of the internet
is associated with declines in participants’ communication with family members
in the household, declines in the size of their social circle, and increases in their
depression and loneliness” (Kraut et al. 1998: 1017). Turkle (1990) describes the
phenomenon of the subjective computer in which we project onto a new tech-
nology our own fears and desires. Tyrannical dictatorships are more likely to see
the threat of revolution and insurgency in CMC, whereas a democratic and open
leadership will find in it an opportunity for more voices to be heard, and for the
free communication of ideas and knowledge. Latterly, CMC research shifted its
attention from the technology itself to the user in context (Androutsopoulos 2006:
420). This notion of self-determinism posits that we use, or allow ourselves to be
affected by technology. For example, the sentence, I ruined my life (via CMC)
casts the user as primary actor responsible for her own actions (Chandler 1995).

Authenticated
496 Claire Hardaker
In short, as we will see in the coming sections, the current focus of CMC
research is on exploring the diversity, creativity, and meaning of linguistic choices,
the construction of online identities, and the interplay between these two.
Not all the diversity, creativity, and constructions are positive, however. In
fact, almost from its first days, CMC became a fertile ground for misbehavior and
crime, whether in the form of sharing illegal images, finding potential victims, or
just being extremely unpleasant to others from the safety of a distant keyboard. As
both Donath (1999) and Sternberg (2000) observed, as a direct result of this, many
CMC platforms are built and managed with conflict at the forefront of almost
every aspect. The evidence for this can be found in the technology, in the form of
block, ban, or ignore options features. They are usually also documented in the
site’s own literature: “An extensive description of killfile techniques in a group’s
FAQ is a kind of virtual scar-tissue, an indication that they have had previous trou-
ble with trolls or flame-wars” (Donath 1999: 48).
In the space of a short chapter, it is impossible to do justice to all the factors
that play into abusive online behavior, so only three major aspects – anonymity,
meaning, and format – are covered below.
2.1. Anonymity, pseudonymity, and disclosure

Probably the earliest avenue of investigation into online abuse involved analyzing
the way in which anonymity affected behavioral outcomes. In simplistic terms,
depending on the platform in question, CMC can offer users a choice between
complete anonymity (one changes one’s username very frequently or interacts
without signing in) through to pseudonymity (one maintains a long-term username
that does not identify that person offline, but that accrues its own history) through
to full disclosure (one uses an identifier that would allow another to find them
offline) (Bernstein et al. 2011).
Research into online identity has revealed that almost regardless of a site’s
imposed level of disclosure, users will themselves choose just how much per-
sonal information they wish to convey (Chester and Bretherton 2007). Anonymity
and pseudonymity both facilitate the exploration of other identities that may be
deemed inconsistent with their offline selves (Bucholtz 1999; Bucholtz and Hall
2005), and as a consequence, Vinagre (2008: 321) observed that this can encour-
age users to feel that they are beyond the reach of consequences:
Sometimes people share very personal things about themselves. They reveal secret
emotions, fears, wishes. They show unusual acts of kindness and generosity, sometimes
going out of their way to help others. We may call this benign disinhibition. However,
the disinhibition is not always so salutary. We witness rude language, harsh criticisms,
anger, hatred, even threats. Or people visit the dark underworld of the Internet—places
of pornography, crime, and violence—territory they would never explore in the real
world. We may call this toxic disinhibition. (Vinagre 2008: 321)

Authenticated
As long ago as 380BC, Plato (2007: 2.359c-2.360d) used the story of the Ring of
Gyges as an insight into the ways in which visibility, invisibility, and anonymity
could encourage a shepherd to murder a king. More recently, Lea and Spears (1991)
observed how anonymity could influence group decision-making processes, whilst
Reicher, Levine and Gordijn (1998) considered the ways that it affected power
relations within groups.
Particularly on very large social networks, little prevents users from carry-
ing out online abuse with fabricated, cloned, or even their own accounts if social
stigma does not concern them (Chester and O’Hara 2009; Phillips 2002; Zarsky
2004). Siegel et al. (1986) elaborate on the ways that anonymity can foster a sense
of impunity, a loss of self-awareness, and a likelihood of acting upon normally
inhibited impulses – an effect known as deindividuation (1986: 161). Psycholog-
ically, interactants may give less consideration to the recipient’s feelings. This,
according to Douglas and McGarty, is manifested in NMOB like flaming and troll-
ing (2001: 399). When we add to this, “the potential for reaching a diverse global
audience, consisting of hundreds of cultures, it is unsurprising that conflict is a
common phenomenon in Usenet” (Baker 2001).
This disjuncture between offline and online communication, however, is not
the only anvil on which conflict is wrought. Even the communication of straight-
forward meanings can be more complex than first appears.
2.2. Richness, assumptions, and meanings

Herring (2003) describes text-based CMC as “free from competing influences from
other channels of communication and from physical context” (2003: 612). Whilst
this claim is perhaps a little strong, it does highlight a supposed advantage of
CMC – that others cannot leap to conclusions about us before they have read what
we have to say. However, this apparent advantage can become a serious drawback.
Zdenek (1999) suggests that CMC is “clearly at a disadvantage, because [users] do
not have access to the wide array of social cues that FTF [face-to-face] speakers
do” (1999: 390). In different words, users who are heavily accustomed to face-to-
face interaction may find themselves communicating in an environment that does
not readily convey all the supporting metadata about the words, such as the tone of
voice, gestures, or facial expressions. It is easy to forget this, or to make assump-
tions that one’s intention will be far more obvious than it really is, and the result is
a greatly increased potential for incorrect inferences.
In this precise vein, much early research into CMC implicitly held face-to-
face interaction as the gold-standard from which all other communicative forms,
including CMC, deviated. For instance, Kraut et al. (1992) argued that face-to-face
was expressively and interactively ‘richer’ than CMC. Baker develops this further
when he argues:

Authenticated
498 Claire Hardaker
To a “newbie,” posting to Usenet can be a daunting experience. It is impossible to know

who will read the message, or to determine how they will react. Also, unlike face-to-
face interaction, gestural, facial or prosodic cues are absent, and so many posters rely
on smilies (Elmer-Dewitt 1994, Sanderson 1993) to function as indicators of emotion.
(Baker 2001)
As above, a typical focus in CMC research is on what CMC lacks in comparison

to face-to-face interaction (e.g. instant audiovisual feedback). Little, if anything,
is said about the online features that are impossible offline, such as the ability to
share multimedia, to overtly and anonymously rate user interaction, and to fully
filter out undesired content.
Perhaps because of our heavy reliance on the face-to-face schema, however,
current research highlights that even experienced users can over-estimate their
online communicative abilities, and this over-estimation increases the potential for
miscommunication (Herring 2003: 612; Kruger et al. 2005; Zdenek 1999: 390).
A complicating factor, already alluded to above, is that CMC is dynamic, and is
continually developing new genres and styles (Abbasi and Chen 2008: 812), some
of which cannot exist offline (e.g. vlogs; fan pages; synchronous, worldwide social
networks).
CMC has also developed a rich body of cues, including symbols, imagery, style,
formatting, greetings, signatures, memes, and animations (also known as GIFs, i.e.
Graphics Interchange Formats). As a result, we can find research into the ways
that these and many besides can express complex and subtle relational, social, and
pragmatic information, such as affect (Subasic and Huettner 2001), rank (Hara,
Bonk and Angeli 2000), opinion (Nigam and Hurst 2004), genre knowledge (Yates
and Orlikowski 2002), ingroup norms, power relations and positioning (Henri
1992; Panteli 2002).
One glaring gap in the literature relates to innovation and memes. In their
efforts to enrich the communicative experience and convey precise meanings,
users will often circumvent interactional limitations by creating their own lin-
guistic and multimodal varieties (cf. the lolcat, doge, and leetspeak varieties, as
well as emoticons known as emoji), or imbuing existing features with new mean-
ings (Vaughan and Gawne 2011; Wilson and Peterson 2002). These can become
standardised, valued markers which form powerful micro-cultures. Knowledge of
how to appropriately use features can be a key index of ingroup membership, and
successful innovations can even become mainstream commercial ventures. For
instance, emojis can now be bought as a range of merchandise. Despite this, it may
be that the very informality, popularity and transience of these features militates
against them being perceived as ‘serious’ topics of study.

Authenticated
2.3. Interactivity, format, and transience
A significant proportion of our time online is dedicated not simply to consuming

static content, but to interacting with others, including via video chats, private
messages, emails, public posts, blog comments and far more besides. Inevitably,
all such communication entails a potential for miscommunication. From this, con-
flict and disharmony can escalate into offline consequences, such as legal action,
violence, and even murder. It might then seem easier to simply abolish online inter-
action altogether, but there are several potential benefits to offering interactivity.
Sites with busy communities can generate enormous revenue from advertising,
commission, and merchandise. Cothrel (2000) investigates how users can be an
excellent source of free content for media outlets, determine the development of a
brand (e.g. Walkers’ ‘Do us a flavour!’ campaign) and give invaluable feedback.
Users may also provide free trouble-shooting and technical support to other users,
thereby reducing the load (and expense) on the retailer.
However, there is relatively little research, certainly within linguistics, that
investigates the ways that the transience of some content (e.g. platforms where
interactions are automatically deleted after a short period of time) can shape anti-
social online behavior. Similarly, there is little to be found on the ways that the
format can impact the spread of online conflict. For instance, by its very nature,
the prototypical bulletin-board format (e.g. the comments sections of many news
outlets, blogs and sites such as 4chan, Yahoo! Answers, Mumsnet, YouTube and
Github) promotes a form on interaction in which one user posts an item to many
people (one-to-many), and many are invited to respond to that one (many-to-one).
By contrast, threaded or nested environments (e.g. Slashdot, Reddit and Usenet)
promote interactions where many people post messages for the many to read
(many-to-many). In short, on a threaded forum, each comment may have a greater
number of readers, meaning that NMOBs may gain a wider audience and therefore,
greater uptake. Empirical research into such matters, though, is in short supply.
3. Flaming
Before reviewing the literature, it is interesting to look back into the history of the
word flaming, because, though it may sound or feel new, this concept has been
with us for some time.
3.1. Etymology
A quick glance at any reasonably thorough dictionary shows that flame (v.) was
used as early as the 1500s to describe a violent, passionate outburst. As might
be expected of an ancient word that describes one of the earliest elemental con-

Authenticated
500 Claire Hardaker
cepts, the metaphorical extension of the word flame sits within a constellation of
other metaphorical extensions, all dedicated to conveying the nuances of anger and
provocation. For instance, a fiery, hot-tempered, explosive person might fan the
flames by making heated, inflammatory or incendiary comments, and we might
simmer over an insult, burn with indignation, boil with rage, or even blow our top
like an erupting volcano. In short, high temper has long been associated with high
temperature, and when we turn to the modern era, CMC simply appears to have
co-opted this ancient fire-as-rage metaphor and applied it to the new phenomenon
of hasty, vitriolic and excessive online outbursts.
When we turn back the pages of academic research, however, we find a differ-
ent story. As already mentioned above, unlike trolling, flaming has received more
academic attention (see, for instance, Avgerinakou 2008; Chester 1996; Herring
1994; Kayany 1998; Lea et al. 1992; Millard 1997), and greater interest typically
results in a proliferation of definitions, rather than a coalescing agreement on one
accepted understanding. Despite this greater interest and the long history that this
metaphor has beyond the realms of CMC, though, as recently as fifteen years ago,
the concept was still regarded in a somewhat fuzzy manner. For instance, under the
heading, 20th century adolescents: Sounding and flaming, Jucker and Taavistainen
(2000) describe flaming thus:
The other institutionalized form of insults is the practice of flaming on the internet. It
appears to be particularly common in news groups, where a large number of partici-
pants can submit email postings under the cover of anonymity. In this context, flaming
is considered to be bad style and is rejected by the code of behavior on the internet, the
so-called netiquette.1 (Jucker and Taavitsainen 2000: 90)
This description really doesn’t tell us what flaming is at all, beyond the fact that
it is considered “bad style”, and the example that Jucker and Taavistainen (2000)
subsequently provide comes from alt.flame, a newsgroup dedicated specifically
to flaming. Jucker and Taavistainen (2000) acknowledge that by the very nature
of such a group, the flamewars carried out there are probably “of an entirely ludic
nature” (Jucker and Taavitsainen 2000: 91). In short, this tells us very little about
what we might think of as genuine flaming. Only a year later, however, Baker
(2001) gives us a more thorough definition:
Antagonistic postings are known as flames (Siegel et al. 1986) and prolonged, escalat-
ing conflicts are often referred to as flame wars. In flame wars flames can give rise to
other flames, involving more and more posters, some who may be angry that the flame
war is taking over the newsgroup. The tone of flames is intentionally aggressive and
numerous methods of attack are used, ranging from intellectualized debate, through
biting sarcasm to scatological abuse. (Baker 2001)
1
This also rather worryingly seems to suggest the existence of only one universal neti-
quette!

Authenticated
And only a few years later, Johnson, Cooper and Chin (2008) describe flaming as
“the antinormative hostile communication of emotions […] that includes the use
of profanity, insults, and other offensive or hurtful statements” (Johnson, Cooper
and Chin 2008: 419).
3.2. Case studies: Chaplin and Chambers

Beyond definitions, it is illuminating to look at two brief examples of flaming,
both from the UK, to see how the literature applies to this behavior in practice.
Our first case involves Gary Chaplin, an executive for recruitment business Stark
Brooks. In November 2011, Chaplin, along with approximately 4,000 other peo-
ple, received the following email from a client named Emmanouil Katsampoukas:
(1) Email from Katsampoukas to Chaplin

Dear Sirs,
My name is Manos Katsampoukas and I am interested in finding a job in the banking/
marketing sector in the UK. Please find attached my CV. Further information available
upon request. Looking forward to hearing from you.
Kind Regards
Manos Katsampoukas.
For Chaplin, Katsampoukas’ email appears to have been provocation enough, and
using the pseudonym Richard Vickers, he reacted as follows:
(2) Email from Chaplin to Katsampoukas

Emmanouil – I think I speak for all 4000 people you have emailed when I say, ‘Thanks
for your CV’ – it’s nice to know you are taking this seriously and taking the time to
make us all feel special and unique.
If you are not bright enough to learn how to ‘bcc’ and thus encourage cock-jockey
retards to then spam everyone on the list (yes Dan McCarthy from One Search I’m
talking about you – you opportunistic thundercunt) then please fuck off …you are too
stupid to get a job, even in banking.
I get enough retarded spam from idiots – I don’t need the Dan ‘fucktard’ McCunthy’s
of the world thinking they are being smart and original by spamming back to your 4000
best friends. (PS – is ‘One Search’ what you’ve successfully completed on this year?).
Yours hitting the delete button. Have a nice day!
Intentionally or otherwise, Chaplin appears to have clicked “Reply All” and sent
this response not only to Katsampoukas, but to all the other 4,000 recipients of the
original email as well. Chaplin’s identity was eventually traced through his ISP,
and as a consequence, Stark Brooks asked him to resign (Atkinson 2011).

Authenticated
502 Claire Hardaker
Whilst this is only one example, we find, as the literature suggests, language in
Chaplin’s email that could be characterized as “intentionally aggressive” (Baker
2001) due to the “profanity, insults, and other offensive or hurtful statements”
(Johnson, Cooper and Chin 2008: 419).
In another case, an irritated reaction that found its outlet in sarcastic humor
resulted not only in the producer of the content being fired, but also in an extended
legal battle and the involvement of multiple celebrities.
In January 2010, as snow and bad weather closed in around Doncaster, Robin
Hood Airport began to issue alerts about possible delays and closures. One
would-be passenger was trainee accountant Paul Chambers (then 26), who was
planning to fly to Belfast to finally meet face-to-face with Sarah Tonner (@crazy-
colours). As the weather worsened, Chambers tweeted several times:
(3) Paul Chambers @pauljchambers 06 Jan 2010
@Crazycolours: I was thinking that if it does then I had decided to resort to terrorism
@Crazycolours: That’s the plan! I am sure the pilots will be expecting me to demand a
more exotic location than NI
We no longer have Sarah’s tweets, but we might infer that she had advocated
taking control of the aircraft. Having already mentioned a penchant for terrorism,
Chambers then tweeted his six hundred or so followers with the following:
Crap! Robin Hood airport is closed. You’ve got a week and a bit to get your shit together
otherwise I’m blowing the airport sky high!!
Despite the airport manager who found it, the senior airport official who was told
about it, and even a police officer investigating it all considering the tweet a joke
rather than a credible threat, a week later, four South Yorkshire police officers
arrested Chambers at his workplace on suspicion of making a hoax bomb threat.
As a result, Chambers lost his job, and though he defended the tweet as a sarcastic
joke borne of frustration, he was ultimately convicted of “sending a public elec-
tronic message that was grossly offensive or of an indecent, obscene or menacing
character contrary to the Communications Act 2003”, leaving him with a fine and
a lifelong criminal record.
Over the next two and a half years Chambers lodged multiple appeals that
were increasingly vocally backed by celebrities including Stephen Fry. Finally, in
the high court, before the country’s most senior judge, Chambers’ conviction was
overturned – a very late acknowledgement, perhaps, that his ‘threat’ to bomb an
airport was really nothing more than a careless moment of irritated online venting.
If we are to draw out a common theme from both the available data and litera-
ture, then it is that flaming is an over-reaction to some sort of provocation, whether
that is an expletive-laden rant in reply to an unsolicited email or a sarcastic bomb-

Authenticated
threat in response to a delayed flight. More simplistically, unlike trolling, which

we turn to shortly, flaming is reactive, rather than proactive.
3.3. Current challenges

As noted throughout this chapter, the research into flaming is further ahead than that
on trolling, though this does not necessarily mean that such research is without its
problems or challenges, and the most prominent issue, which recurs with trolling,
is that of disambiguating a behavior from other possible behaviors in the first place.
Within the remit of NMOB, we can include flaming, trolling, cyberbullying,
cyberstalking, and more besides, and it is no mean feat to consistently distinguish
them from each other. To exemplify this, for a moment we will step outside of
flaming and consider Bocij’s (2004) definition of cyberstalking:
[Cyberstalking is] a group of behaviors in which an individual, group of individuals
or organization, uses information and communication technologies to harass another
individual, group of individuals, or organization. Such behaviors may include, but are
not limited to, the transmission of threats and false accusations, identity theft, damage
to data or equipment, computer monitoring, solicitation of minors for sexual purposes,
and any form of aggression. Harassment is defined as a course of action that a reasona-
ble person, in possession of the same information, would think causes another reasona-
ble person to suffer emotional distress. (Bocij 2004: 14, emphasis mine)
It is difficult to see how this definition could be used in such a way that it would not
also capture flaming, trolling, cyberbullying, cyberharassment, and sex offenders
grooming children online, as well as any CMC contribution perceived to be objec-
tionable or distressing for other reasons (e.g. posting a video of animal abuse). The
problem is little better when we consider legislation and policy guidance. From
the UK, for instance, the House of Lords’ Communications Committee published
its first report on social media and criminal offences. In this, trolling is described
as the “intentional disruption of an online forum, by causing offence or starting an
argument” (2014: ch2 § 9c). However, this extremely simplistic definition doesn’t
differentiate between someone disrupting a forum because they have (or think they
have) a genuine grievance and someone doing so simply for the sake of amusement.
Similarly, in guidelines on prosecuting cases involving communications sent
via social media, the Crown Prosecution Service alludes to flaming thus:
Examples of cyberstalking may include:
– Threatening or obscene emails or text messages.
– Spamming (where the offender sends the victim multiple junk emails).
– Live chat harassment or ‘flaming’ (a form of online verbal abuse).
– Leaving improper messages on online forums or message boards.
– Sending electronic viruses.
– Sending unsolicited email.
– Cyber identity theft. (CPS 2016, emphasis mine)

Authenticated
504 Claire Hardaker
To describe flaming as a form of online verbal abuse confuses the boundaries

between this and most other kinds of NMOB. How is flaming different, then, from
trolling, cyberbullying, and so forth? We find similar issues in other jurisdictions.
In the US, states that do have statutes for NMOB differ widely in how clearly that
behavior is captured. For example, § 2 of North Carolina’s (2000) cyberstalking
provision states that it is unlawful to:
(2) Electronically mail or electronically communicate to another repeatedly, whether or
not conversation ensues, for the purpose of abusing, annoying, threatening, terrifying,
harassing, or embarrassing any person.
(3) Electronically mail or electronically communicate to another and to knowingly make
any false statement concerning death, injury, illness, disfigurement, indecent conduct,
or criminal conduct of the person electronically mailed or of any member of the per-
son’s family or household with the intent to abuse, annoy, threaten, terrify, harass, or
embarrass. (2000)
This legislation could be used to cover NMOB from repeatedly annoying a recip-
ient (i.e. this could capture spamming, flaming, or trolling) through to terrifying
them by lying about their loved ones (i.e. cyberbullying or cyberstalking).
To summarize, throughout academia, policy, legislation, and the media, there
is little consensus about these terms and their scope. This point is regrettably made
far more clearly in the following section.
4. Trolling
Just as we began the section on flaming with a brief glance into the history of
the word, so it is insightful to do the same with trolling. We find, however, a less
clear-cut answer.
4.1. Etymology
The current definition of trolling may have derived from one of two distinct routes.
The first captures troll as a noun, and finds its roots in the late fourteenth century.
In Old Norse and Scandinavian mythology, a tröll was a large, strong, nasty crea-
ture that possessed supernatural powers, but that would also turn to stone in the
sunlight (Jakobsson 2006: 1; MacCulloch 1930: 285–286). These ancient myths
persist in folklore tales such as the Norwegian fairytale The Three Billy Goats
Gruff. The second possibility captures troll as a verb, and dates back at least as far
as the 1600s. In this respect, trolling derives from fishing, and involves drawing
baited fishing lines through the water.2 This variant also has longstanding meta-
2
This is not to be confused with trawl-fishing, which involves dragging nets.

Authenticated
phorical extensions that include luring others along with some form of bait, and as
exhaustively searching for something.
Currently, users are as happy to invoke terms relating to mythology for the
person (e.g. get back under your bridge, don’t feed the troll) as to fishing for
the act (e.g. biting, baiting, netting, and hooking) and the reality is that we will
probably never know which history was being drawn on when the first person
uttered troll in some semblance of its modern, online sense. Certainly, there is
little enough evidence to give us much hope of deducing it from the earliest online
postings. One matter that is clear, however, is that this term has been far more
heavily influenced by the media than flaming. This is important to note, since
scholars are not impervious to being influenced by widespread, mainstream narra-
tives.
When we look to the media, we find that interest was initially slow, and prior
to 2010, there are only scattered reports on trolling (e.g. Black 2006; Cox 2006;
Moulitsas 2008; Thompson 2009). Where it was discussed, typical definitions
described trolling as the posting of incendiary comments designed to provoke con-
flict: “Hiding behind the pseudonymity of a Web alias, trolls disrupt useful dis-
cussions with ludicrous rants, inane threadjackings, personal insults, and abusive
language” (Naraine 2007: 146). Brandel (2007) adds that “[a] troll is a person who
posts with the intent to insult and provoke others. […] The goal is to disrupt the
normal traffic of a discussion group beyond repair” (Brandel 2007: 32).
Heffernan (2008) highlights the unprovoked nature of trolling, along with
another goal – amusement at another’s expense:
Consider this question from David Hume: “Would any man, who is walking alone, tread
as willingly on another’s gouty toes, whom he has no quarrel with, as on the hard flint
and pavement?” […] Internet trolls regularly tread on gouty toes. They trick vulnerable
people with whom they have no quarrel; they upset those people; they humiliate them;
they break their hearts; they mess with them. They do it for something Hume didn’t
perfectly name: the lulz—the spiteful high. (Heffernan 2008)
The earliest academic attention to the subject matter also largely created defi-
nitions from intuition, the media, and online ephemera like The Troller’s FAQ
(1996). The result is that the term has been, and continues to be used as an all-en-
capsulating term. For example, Herring et al. (2002: 372) and Turner et al. (2005)
describe trolling as luring others into frustratingly useless, circular discussion that
is not necessarily overtly argumentative. Donath (1999: 45) and Utz (2005: 50)
suggest that trollers can intentionally disseminate poor advice, thereby provok-
ing corrections from others. Tepper (1997: 41) explains how trolling can define
ingroup/outgroup membership: those who ‘bite’ signal novice, outgroup status,
whilst ingroup members will identify the troller, will not be baited, and may even
mock those who are. Donath (1999) and Dahlberg (2001) suggest that trolling is
a one-sided game of deception played on unwitting others: “The troll attempts to

Authenticated
506 Claire Hardaker
pass as a legitimate participant, sharing the group’s common interests and con-
cerns” (Donath 1999: 45). Then, once the troll has developed its false identity and
been accepted into the group, they will set about disrupting the forum whilst trying
to conceal their true intent (Dahlberg 2001).
The main point here is that whilst some of these definitions overlap to an
extent, all seem to be describing a collection of symptoms, rather than one coher-
ent behavior or motive that is responsible for those choices of actions. In an effort
to create an empirically informed definition, I analysed 3,727 user discussions of
trolling drawn from an eighty-six million word Usenet corpus and concluded that
trolling is “the deliberate (perceived) use of impoliteness/aggression, deception,
and/or manipulation in CMC to create a context conducive to triggering or antago-
nizing conflict, typically for amusement’s sake” (Hardaker 2013: 79).
In short, one distinction to be drawn between flaming and trolling, as hinted at
in the previous section, is that whilst flaming may be an over-reaction to a provo-
cation, trolling proactively strives to be a provocation in its own right.
4.2. Case studies: Brenda Leyland

Since 2010, there have been a string of prosecutions related to cases that the media
has typically described as trolling. For instance, on 29th October 2010, 36-year-old
Colm Coss pleaded guilty at Manchester City Magistrates’ Court after leaving
offensive messages and jokes on memorial pages set up to commemorate deceased
loved ones, including children. He was sentenced to eighteen weeks in prison.
Similarly, in September 2011, Sean Duffy, then aged 26, also pleaded guilty to the
same charge of improper use of a public electronic communications network at
Reading Magistrates’ Court after posting an offensive image to the tribute page of
a teenager who had been accidentally shot.
Deliberately targeting memorial and tribute sites with mockery is not the only
type of behavior that is picked up by the media as trolling, however. Other, more
sinister behavior also earns this name. In June 2013, 60-year-old Frank Zimmer-
man was given a suspended 26-week sentence after emailing abuse to a string of
celebrities, including Lord Alan Sugar, Louise Mensch MP, and journalist Terence
Blacker. In January 2014, John Nimmo and Isabelle Sorley were jailed for eight
and twelve weeks respectively for sending journalist and feminist activist Caroline
Criado-Perez Twitter abuse after she successfully petitioned the Bank of England
to ensure that a notable historic female figure would be represented on British
banknotes. And in September 2014, Peter Nunn was given an eighteen week custo-
dial sentence for sending abuse to Stella Creasy MP after she had supported Cria-
do-Perez’s campaign. Some of the abuse in this latter case involved rape, death,
and bomb-threats.
It is crucial to note, however, as has been alluded to above, that not all cases
of so-called trolling are this clear cut, but to understand the complexity we need

Authenticated
a far greater amount of context and background. For this, we turn to the case of
Brenda Leyland.
In 2007, three-year-old Madeleine McCann disappeared from the holiday resort
of Praia da Luz. Her parents, medical doctors Gerry and Kate McCann were dining
at a nearby restaurant at the time, whilst Madeleine and her younger siblings slept
in their hotel bedroom. At around 10pm, Kate McCann went to check on her chil-
dren and discovered that Madeleine was gone. Despite extensive searches, she has
never been found. During the early parts of the inquiry, the McCanns were ques-
tioned by police about her disappearance, and media outlets were quick to voice
their suspicions and theories about how the McCanns might have been involved.
Meanwhile, though the platform was only a year old at the time, users of the new
social media site, Twitter, also took an interest in the story. Whilst some invested
hours trading guesses about the truth behind Madeleine’s disappearance, others
spent their time sending threats of violence, murder and abduction of their other
children at the McCanns.
Responding to the media was relatively straightforward: the McCanns took
legal action and were ultimately awarded substantial damages, along with front-
page apologies. However, social media is a different and much more difficult plat-
form to regulate, and almost a decade later, havens of like-minded people still
interested in the Madeleine McCann case thrive. This brings us to 2014, and the
sleepy civil parish of Burton Overy in Leicestershire.
Brenda Leyland was educated at a convent school, went to church, enjoyed gar-
dening and photography, and was involved in the annual village scarecrow compe-
tition. Not all was perfect, however. A sixty-three year old mother to two sons, she
appears to have been estranged from the eldest, and her marriage had ended in 2001.
According to expert evidence at the inquest, Leyland had a history of attempted
suicide and was receiving both therapy and medication to help her deal with bouts
of severe depression and anxiety. A consultant psychiatrist who had treated Leyland
in the past also stated that she had lifelong unstable emotional personality traits.
On Twitter, Leyland had another identity: @sweepyface. The bio for sweepy-
face’s Twitter account simply read “Researcher”, and after a few dormant years, it
suddenly became active in November 2013. From that point to September 2014,
sweepyface sent roughly 4,600 tweets, and throughout 2014 alone, the number of
people following the account doubled from 93 to 183. The content of sweepyface’s
tweets skewed in a very particular direction: 87 % of them contained the word or
hashtag McCann. Meanwhile, references to the parents by their first names or ini-
tials occurred an average of once every ten tweets.
Within the online community of those campaigning about the McCanns,
sweepyface fell into a group that might be crudely titled the anti-McCanns – that
is, users who believe the McCanns to be responsible, to whatever degree, for their
daughter’s disappearance. By contrast, the pro-McCanns, as might be expected,
believe the parents innocent of most or all wrongdoing. Both groups employ the

Authenticated
508 Claire Hardaker
#McCann hashtag as a method of following discussion on this topic, and through

it, sweepyface and others could instantly jump to the latest tweets worldwide on
the subject.
How did sweepyface behave online? She was certainly not complimentary
towards the McCanns and on multiple occasions, she cast highly critical, personal
aspersions about them (see example 10, for instance). However, these tweets talk
about the McCanns, rather than at them, and there is also a lot of evidence to sup-
port the argument that sweepyface genuinely believed that she was pursuing justice:
(6) Sweepyface, 11:05:45, Fri 29 Nov 2013
#crimewatch I think you and the Met know as millions of us do that the McCanns may
be complicit in Madeleines fate, don’t let us down !!
(7) Sweepyface, 18:13:19, Mon 23 Dec 2013
I have never heard a pro say ” poor child, left alone, scared, not feeling well that day”
all I hear is poor #mccann s how they r suffering
(8) Sweepyface, 18:30:45, Mon 23 Dec 2013
#mccann So many people think the children were sedated, because the twins never
stirred for almost 12 hours, lights on, shouting crying?
(9) Sweepyface, 09:47:14, Tue 24 Dec 2013
#mccann It is and has always been, about Madeleine . A victim on so many levels, we
will help fight for as long as it takes 4 her
(10) Sweepyface, 15:21:34, Mon 08 Sep 2014
I think Kate #mccann sees herself as a modern day Eva Peron beautiful, suffering,
instead of a booze filled nymphomaniac
To briefly summarise some 4,600 tweets, sweepyface frequently described the
McCanns as neglectful parents, strongly objected to their ongoing media appear-
ances, and complained that they were profiting from their daughter’s disappear-
ance. She also tweeted police forces, crime investigation programmes, and media
outlets, campaigning for them to reopen the investigation. When pro-McCann
users and others challenged her, she called them unpleasant names, disputed their
evidence, and occasionally blocked them. In short, sweepyface appears to have
fixated on both the topic and on some controversial theories surrounding it, and
her conduct was annoying to some and offensive to others. However, as best as
can be inferred from the data, the intention to cause conflict for the sake of amuse-
ment – as we find in many definitions of trolling – appears to be missing.
Finally, in September 2014, sweepyface’s online world collided with Brenda
Leyland’s offline world. It began when sweepyface noticed that Sky reporter Mar-
tin Brunt had started to follow her. Sweepyface even tweeted him directly:
#mccann Just noticed @skymartinbrunt is following me, why Martin won’t you
investigate some of these facts and show neutrality ?
#mccann @skymartinbrunt Martin, any thoughts on the connection between Amy
Tierney, the witness and Basil ? how does that work ?

Authenticated
Rather than responding publically, Brunt contacted Leyland in private. Three days
later, on Thursday 02nd October, Sky News launched an exposé on “the McCann
trolls”. As well as radio reports and online articles, the investigation consisted of
an eleven-minute video report posted on YouTube.3 A shortened version of this
report was repeated on the main Sky News channel cycle throughout the day. In
that excerpt, Brenda Leyland is doorstepped by a camera crew led by Martin Brunt:
(13) Sky News, 18:12, Thu 02nd Oct 2014: ‘Evil’ Trolls In Hate Campaign Against McCanns:
video (04m 23s)4. Excerpt: 00s 00m to 01m 11s.
Brunt, voiceover: [Shaky footage as the camera person heads towards a car that Ley-
land is exiting.] This woman uses Twitter to attack the parents of
Madeleine McCann. On the internet, she’s anonymous. Not any-
more. [Leyland has been walking their way. As they meet, she looks
somewhat bewildered.]
Brunt: [Some speech obscured by voiceover.] …I’m Martin Brunt from
Sky News.
Leyland: Well I’m just about to go out.
Brunt: Well we’ve caught you. Can we talk to you about your Twitter?
Leyland: No.
Brunt: And your attacks on the McCanns.
Leyland: No.
Brunt: Erm. Why are you attacking them so regularly?
Leyland: Look I’m just going out with a friend. Okay?
Brunt: But why are you using your Twitter account …
Leyland: Excuse me. [Turns and walks back towards her car.]
Brunt: … to attack the McCanns?
Leyland: [Stops and turns back.] I’m entitled to do that Martin.
Brunt: You know you’ve been reported to the police, to Scotland Yard.
[Camera gets in front of Leyland.] They’re considering, er, a whole
file of Twitter accounts [camera gets closer to Leyland] and that …
Leyland: That’s fair enough.
Brunt: … what supporters say is a campaign of abuse against the McCanns.
Leyland: Okay well I’m going out. [Leyland turns and walks back towards
her car. She looks back briefly as if responding to something fur-
ther but the audio has cut to the voiceover at this point.]
Brunt, voiceover: On Twitter, she uses the name sweepyface with a profile picture of
a pet. She tweets many times a day, and mostly about the McCanns.
In one message she spread rumours about the couple’s marriage.
In another, she hoped Madeleine’s parents would suffer forever.
But sweepyface is not the worst. Almost from the day Madeleine
vanished seven years ago, her family have been targeted with vile
messages on the internet.
3
https://www.youtube.com/watch?v=qkAzz8Pwdvc
4
http://news.sky.com/story/1345871/evil-trolls-in-hate-campaign-against-mccanns

Authenticated
510 Claire Hardaker
In the report below this video, Brunt writes:

The Metropolitan Police is investigating a catalogue of vile internet abuse targeting the
family of Madeleine McCann including death threats, Sky News can reveal. Officers
are in talks with the Crown Prosecution Service after being handed a dossier of more
than 80 pages of Tweets, Facebook posts and messages on online forums aimed at Kate
and Gerry McCann. […] One troll – who uses the Twitter identity “Sweepyface” and
has posted dozens of anti-McCann messages using the #mccann hashtag – was con-
fronted by Sky News. When asked about her use of social media to attack the couple,
she replied: “I’m entitled to.”
Though Sky New does not give her offline name, Leyland’s face and voice is given
the prime slot at the start of the video, and she features heavily in Sky’s reports on
“the McCann trolls” throughout the day. The following tweet from her sweepyface
account is also used and widely requoted:
(14) Sweepyface, 12:30:20, Fri 29 Nov 2013
#mccann Q ” how long must the Mccanns suffer” answer ” for the rest of their miser-
able lives”
This tweet appears in amongst a wide range of other content from various sources,
some containing threats and incitement to violence towards the McCanns. The
identities of those other users are kept anonymous, however, leaving Leyland as
the sole face and identity of “the McCann trolls”. Within hours, other media out-
lets make the connection between sweepyface and Brenda Leyland. Her identity,
age and village are all published, and by the next day, across most of the media,
she is being vilified as a troll.
On Saturday 04th October, two days after the exposé, Leyland is found dead in
a room at the Marriott Hotel, Leicester. An inquest immediately opens, and some
five months later, on 20th March 2015, Coroner Catherine Mason records a verdict
of suicide. As part of the inquest, Detective Sergeant Steven Hutchings of Leices-
tershire Police verifies that none of Leyland’s tweets had amounted to a criminal
offence.
This case exemplifies, to an extreme degree, the problem raised repeatedly
throughout the chapter: namely, that the definitions of terms relating to negatively
marked online behaviors are unclear, inconsistent, and because of this, they can be
badly misapplied. To take the term troll specifically, the societal stigma that goes
with being branded such is serious enough to make an otherwise unknown individ-
ual front page news. However, this very term is also being semantically stretched,
particularly by the media, to such an extent that it is catching cases in which the
accused individual is arguably not a troll at all. In short, there are serious, real-
world consequences to the ways in which we define and apply terms like troll and
these need much more investigation.

Authenticated
4.3. Current challenges
The issue of terminology is not the only challenge faced by those researching
trolling. Given that trolling, as a field of enquiry, is virtually brand new, it would
be possible to list almost any aspect as a future challenge. However, there is one
defining element of trolling, when contrasted with flaming, that we find across the
literature: deception.
Literature in the field of im/politeness is currently grappling hard with the
place and scope of intentions, in our understandings and productions of socially
(un)acceptable behavior (see, for instance, Arundale 2008; Haugh 2008; Spencer-
Oatey and Xing 2006). But how does this relate to trolling or deception? Brown and
Levinson (1987) noted very early on that someone might express themselves such
that “there is more than one unambiguously attributable intention so that the actor
cannot be held to have committed [them]self to one particular intent” (Brown and
Levinson 1987: 69).
Whilst Brown and Levinson are talking about politeness, this can equally apply
to any form of interaction, from courtroom examination to political interviews to
online trolling. As in the definitions proposed in § 4.1 above, a troll may “attempt
to pass as a legitimate participant” (Donath 1999: 45) and then, “after developing
their false identity and becoming accepted within a group, the troll sets about dis-
rupting proceedings while trying to maintain his or her cover” (Dahlberg 2001).
If indeed a troll chooses a covert strategy, should they be challenged, they may
excuse their behavior as accidental, unwitting, incidental, and so on, whilst deny-
ing any intention to offend. Similarly, just as a troll might disguise their intentions
to deliberately offend, a target might be disingenuous about their interpretations of
that behavior, and pretend to be amused rather than offended as a way to save face.
In summary, the debate involving the role of intentions has received some con-
siderable attention both within and even outside of pragmatics5 but virtually none
considers deception as part of that process.
Nefarious deception (that is, deception motivated by personal gain or harming
others, as opposed to that motivated by kindness or politeness) reaches beyond
intentions, however. It can also encapsulate the formation of an identity that is
inconsistent with one’s offline self, if done with malicious purposes in mind – an
aspect that is heavily facilitated by the anonymity that CMC can offer. This may
even stretch as far as creating multiple accounts deliberately made to look like
separate people, so that the user can cause greater disruption by the appearance of
5
See, for instance, Bach (1987), Davis (1998, 2007, 2008), Gibbs (1999, 2001), Green
(2007, 2008), Jaszczolt (2005, 2006), Keysar (2007), Recanati (1986), Saul (2001),
Searle (1983, 1990), Thompson (2008).

Authenticated
512 Claire Hardaker
strength in numbers (Chester and O’Hara 2009; Phillips 2002; Zarsky 2004). Little
has been done in this area from linguistics.
Finally, deception also captures straightforward lies, such as giving advice that
the troll knows to be dangerous or incorrect in the hopes of causing hurt or harm,
or promoting themselves as trustworthy, to gain greater traction within the group.
Whilst there is very little work within linguistics on what we might crudely call
‘narrative deception’ (as opposed to intention deception, or identity deception)
fortunately there is a wealth of excellent work to be found in psychology6 which
may help move this area forward.
From almost every angle, research into flaming and trolling is very new. At the
time of writing, the field that these topics sit within – CMC – has been established
for barely thirty-five years (see, for instance, Kaye 1985; Kerr and Hiltz 1982;
Kiesler, Siegel and McGuire 1984). Flaming made its debut some ten years after
that (Lea et al. 1992), whilst trolling took another decade to arrive (Herring et al.
2002). The result, then, is that almost every possible avenue qualifies as a future
direction, but there are some that are worth repeating, or raising afresh.
Perhaps the most pressing issue, as discussed throughout this chapter, is clarity
on what terms like flaming and trolling mean. It is unsurprising for new fields to
go through a period of turmoil as concepts become established and certain con-
sensuses are reached. However, in this particular area of research, a lack of clarity
can have extremely serious, real-world impacts, as in the case of Brenda Leyland.
Clear definitions also become paramount when issuing legal guidance and even
legislation itself, however. Determining that a behavior is illegal relies heavily
on being able to identify that behavior in the first place, and here, linguists have
much to offer, particularly from the fields of forensic linguistics, corpus linguistics
(e.g. through observation of largescale patterns), and pragmatics.
Further, despite the now firmly-established field of impoliteness, which includes
seminal work by the likes of Culpeper (1996), Spencer-Oatey (2005), and Bousfield
(2008), there is surprisingly little research into either flaming or trolling that seems
to draw upon this (see however, Graham 2007, 2008; Herring 1994; Locher 2006).
In short, aligning impoliteness research and research into negatively marked
online behaviors – particularly flaming – could benefit both since they have marked
similarities, such as the variability in perceptions of what constitutes impoliteness,
trolling, or flaming in the first place, and the evaluations of degrees of hostility
6
See, for instance, Akehurst et al. (1996), Ekman (1996), Memon, Vrij and Bull (2003),
Rubin (2010), Vrij (2000), Vrij et al. (2000).

Authenticated
made by the participants (Graham 2007: 743). Similarly, research into flaming
and trolling may be substantially supported by work in psychology, criminology,
sociology, and computing.
In short, for those interested in researching negatively marked online behavior,
a wide vista of possibilities is open for exploration. If there remains any lingering
sense that CMC is somehow not as worthy a subject of investigation as offline
interaction, hopefully this chapter has shown that it is not only a serious area, but
also one that should be ventured into carefully, and with a great deal of sensitivity.
Acknowledgments
This work was supported by the Economic and Social Research Council through
two grants: the Twitter Rape Threats and the Discourse of Online Misogyny pro-
ject (grant reference ES/L008874/1) and the ESRC Centre for Corpus Approaches
to Social Science (grant reference ES/K002155/1).
References
Abbasi, Ahmed and Hsinchun Chen

2008 Cybergate: A design framework and system for text analysis of computer-me-
diated communication. MIS Quarterly 32(4): 811–837.
Akehurst, Lucy, Günter Köhnken, Aldert Vrij and Ray Bull
1996 Lay persons’ and police officers’ beliefs regarding deceptive behavior. Applied
Cognitive Psychology 10: 461–473.
2006 Introduction: Sociolinguistics and computer-mediated communication. Jour-
nal of Sociolinguistics 10(4): 419–438.
Arendholz, Jenny
Arundale, Robert B.
2008 Against (Gricean) Intentions at the heart of human interaction. Intercultural
Pragmatics 5(2): 229–258.
Atkinson, Jane
2011 £200,000 Exec Axed after Telling Jobseeker: **** Off. The Sun December
10th. http://www.thesun.co.uk/sol/homepage/news/3990329/200000–exec–
axed–after–telling–jobseeker–off.html.
Avgerinakou, Anthi
2008 Contextual Factors of Flaming in Computer-Mediated Communication. Edin-
burgh: Heriot-Watt University.
Bach, Kent
1987 On communicative intentions: A reply to Recanati. Mind and Language 2:
141–154.

Authenticated
514 Claire Hardaker
Baker, Paul
2001 Moral panic and alternative identity construction in Usenet. Journal of Compu
ter-Mediated Communication 7(1). http://jcmc.indiana.edu/vol7/issue1/baker.
html 08/12/09.
Barron, Anne
2006 Understanding spam: A macro-textual analysis. Journal of Pragmatics 38:
880–904.
Baym, Nancy
Bernstein, Michael S., Andrés Monroy-Hernández, Drew Harry, Paul André, Katrina Pano-
vich and Greg Vargas
2011 4chan and /B/: An analysis of anonymity and ephemerality in a large online
community. Association for the Advancement of Artificial Intelligence: 1–8.
Binns, Amy
2011 Don’t feed the trolls: Managing troublemakers in magazines’ online communi-
ties. Mapping the Magazine 3. http://www.people.vcu.edu/~dgolumbia/classes/
1314.2.spr2014/engl391/resources/Trolls.pdf.
Black, Lisa
2006 It’s a troll’s ‘life’ for some: Online games raise addiction concerns. Chicago
Tribune November 30th: 1.
Bocij, Paul
2004 Cyberstalking: Harassment in the Internet Age and How to Protect Your Fam-
ily. Westport: Praeger.
Bolter, Jay David and Richard Grusin
1998 Remediation: Understanding New Media. Cambridge, MA: The MIT Press.
Bousfield, Derek
2008 Impoliteness in Interaction. Amsterdam/Philadelphia: Benjamins.
Brabazon, Tara
2012 Digital Dialogues and Community 2.0: After Avatars, Trolls and Puppets.
Oxford: Chandos Publishing.
Brandel, Mary
2007 Blog trolls and cyberstalkers: How to beat them. Computerworld May 28: 32.
1987 Politeness: Some Universals in Language Use. Cambridge: Cambridge Uni-
versity Press. (Original edition 1978. Reprint 1987.)
Bucholtz, Mary
1999 "Why be normal?”: Language and identity practices in a community of nerd
girls. Language in Society 28(2): 203–223.
2005 Identity and interaction: A sociocultural linguistic approach. Discourse Stud-
ies 7(4–5): 585–614.
Buckels, Erin E., Paul D. Trapnell and Delroy L. Paulhus
2014 Trolls just want to have fun. Personality and Individual Differences 35(67):
97–102.
Chandler, Daniel
1995 Technological or Media Determinism. http://www.aber.ac.uk/media/
Documents/tecdet/tecdet.html.

Authenticated
Cheng, Justin, Cristian Danescu-Niculescu-Mizil and Jure Leskovec

2015 Antisocial behavior in online discussion communities. Proceedings of ICWSM
2015. http://arxiv.org/abs/1504.00680.
Chester, Andrea
1996 Braving the flames: Aggression and the Internet. Paper presented at the Media
and Ethnic Conflict Conference, November 1996, Melbourne, Australia.
Chester, Andrea and Di Bretherton
2007 Impression management and identity online. In: Adam Joinson, Katelyn
Y. McKenna, Tom Postmes and Ulf-Dietrich Reips (eds.), Oxford Handbook
of Internet Psychology, 223–236. Oxford: Oxford University Press.
Chester, Andrea and Agi O’Hara
2009 Image, identity, and pseudonymity in online discussions. International Jour-
nal of Learning 13(12): 193–204.
Chiou, Wen-Bin
2006 Adolescents’ sexual self-disclosure on the Internet: Deindividuation and
impression management. Adolescence 41(163): 547–561.
Collins, Mauri P.
1992 Flaming: The relationship between social context cues and uninhibited verbal
behaviour in computer-mediated communication. http://www.mediensprache.
net/archiv/pubs/2842.htm.
Cothrel, Joseph P.
2000 Measuring the success of an online community. Strategy and Leadership 20
(2): 17–21.
Cox, Ana Marie
2006 Making mischief on the Web. Time December 16th. http://www.time.com/time/
magazine/article/0,9171,1570701,00.html 03/02/10.
CPS
2016 Guidelines on Prosecuting Cases Involving Communications Sent Via Social
Media. Prosecution Policy and Guidance: Legal Guidance March 3rd. http://
www.cps.gov.uk/legal/a_to_c/communications_sent_via_social_media/.
Culpeper, Jonathan
1996 Towards an anatomy of impoliteness. Journal of Pragmatics 25: 349–367.
Cyberstalking
2000 North Carolina: § 14–196.3. http://www.ncga.state.nc.us/EnactedLegislation/
Statutes/HTML/BySection/Chapter_14/GS_14-196.3.html.
Dahlberg, Lincoln
2001 Computer-mediated communication and the public sphere: A critical analysis.
Journal of Computer-Mediated Communication 7 (1). http://jcmc.indiana.edu/
vol7/issue1/dahlberg.html 22/08/10.
Davis, Wayne
1998 Implicature. Intention, Convention, and Principle in the Failure of Gricean
Theory. Cambridge: Cambridge University Press.
Davis, Wayne
2007 How normative is implicature? Journal of Pragmatics 39: 1655–1672.
Davis, Wayne
2008 Replies to Green, Szabo, Jeshion, and Siebel. Philosophical Studies 137: 427–
445.

Authenticated
516 Claire Hardaker
December, John
1997 Notes on defining computer-mediated communication. CMC Magazine Janu-
ary. http://www.december.com/cmc/mag/1997/jan/december.html 19/07/08.
Dill, Karen E. and Jody C. Dill
1998 Video game violence: A review of the empirical literature. Aggression and
Violent Behavior: A Review Journal 3: 407–428.
Dlala, Imen Ouled, Dorra Attiaoui, Arnaud Martin and Boutheina Ben Yaghlane
2014 Trolls identification within an uncertain framework. Proceedings of the 2014
IEEE 26th International Conference on Tools with Artificial Intelligence.
http://arxiv.org/abs/1501.05272: 1011–1015.
Donath, Judith S.
1999 Identity and deception in the virtual community. In: Marc A. Smith and Peter
Kollock (eds.), Communities in Cyberspace, 29–59. London: Routledge.
Douglas, Karen M. and Craig McGarty
2001 Identifiability and self-presentation: Computer-mediated communication and
intergroup interaction. British Journal of Social Psychology 40(3): 399–416.
Ekman, Paul
1996 Why don’t we catch liars? Social Research 63(3): 801–817.
Elmer-Dewitt, Philip
1994 Bards of the Internet. Time July 4th: 66–67.
Ferris, Pixy
1997 What Is Cmc? An overview of scholarly definitions. CMC Magazine January.
http://www.december.com/cmc/mag/1997/jan/ferris.html.
Fichman, Pnina and Madelyn Rose Sanfilippo
2015 The bad boys and girls of Cyberspace: How gender and context impact per-
ception of and reaction to trolling. Social Science Computer Review 33(2):
163–180.
Gibbs, Raymond W.
1999 Intentions in the Experience of Meaning. Cambridge: Cambridge University
Press.
Gibbs, Raymond W.
2001 Intentions as emergent products of social interactions. In: Bertram Malle,
Louis Moses, and Dare Baldwin (eds.), Intentions and Intentionality, 105–
122. Cambridge, MA: MIT Press.
diated community. Journal of Pragmatics 39: 742–759.
nity of practice. In: Derek Bousfield and Miriam A. Locher (eds.), Impolite-
ness in Language: Studies on Its Interplay with Power in Theory and Practice,
324–352. Berlin/New York: de Gruyter.
Green, Mitchell
2007 Self-Expression. Oxford: Oxford University Press.
Green, Mitchell
2008 Expression, indication, and showing what’s within. Philosophical Studies 137:
389–398.

Authenticated
Hara, Noriko, Curtis J. Bonk and Charoula Angeli

2000 Content analysis of online discussion in an applied educational psychology
course. Instructional Science 28: 115–152.
Hardaker, Claire
cussions to academic definitions. Journal of Politeness Research. Language,
Behaviour, Culture 6(2): 215–242. doi: 10.1515/jplr.2010.011.
Hardaker, Claire
2013 “Uh.....Not to Be Nitpicky,,,,,But …The Past Tense of Drag Is Dragged, Not
Drug.”: An overview of trolling strategies. Journal of Language Aggression
and Conflict 1(1): 57–85.
Hardaker, Claire
2015 “I Refuse to Respond to This Obvious Troll.”: An overview of responses to
(perceived) trolling. Corpora 10(2): 201–229.
2016 “Real Men Don’t Hate Women”: Twitter rape threats and group identity. Jour-
nal of Pragmatics 91: 80–93.
Haugh, Michael
2008 Intention in pragmatics. Intercultural Pragmatics 5(2): 99–110.
Haugh, Michael
Heffernan, Virginia
2008 Trolling for ethics: Mattathias Schwartz’s awesome piece on Internet Polter-
geists. New York Times July 31st.
Henri, France
1992 Computer conferencing and content analysis. In: Anthony R. Kaye (ed.), Col-
laborative Learning through Computer Conferencing: The Najaden Papers,
115–136. Berkeley, CA: Springer.
Herring, Susan C.
1994 Politeness in computer culture: Why women thank and men flame. Cultural
Performances: Proceedings of the Third Berkeley Women and Language Con-
ference: 278–294. Berkeley Women and Language Group. http://ella.slis.indi-
ana.edu/~herring/politeness.1994.pdf.
Herring, Susan C. (ed.)
1996 Computer-Mediated Communication: Linguistic, Social and Cross-Cultural
Herring, Susan C.
2003 Computer-mediated discourse. In: Deborah Schiffrin, Deborah Tannen and
Heidi E. Hamilton (eds.), The Handbook of Discourse Analysis, 612–634.
Oxford: Blackwell.
Herring, Susan C., Kirk Job-Sluder, Rebecca Scheckler and Sasha Barab
2002 Searching for safety online: Managing “Trolling” in a feminist forum. The
Information Society 18: 371–384.
Hopkinson, Christopher
2013 Trolling in online discussions: From provocation to community-building.
Brno Studies in English 39(1): 5–25.

Authenticated
518 Claire Hardaker
Jakobsson, Ármann
2006 The Good, the Bad and the Ugly: Bárðar Saga and Its Giants. Paper read at The
13th International Saga Conference: The Fantastic in Old Norse/Icelandic Liter-
ature, 06th–12th August 2006, at Durham and York. http://opac.regesta-imperii.
de/lang_en/anzeige.php?sammelwerk=The+Fantastic+in+Old+Norse+Icelan-
dic+Literature.+Preprint+Papers.
Jaszczolt, Katarzyna M.
2005 Default Semantics. Foundations of a Compositional Theory of Acts of Commu-
nication. Oxford: Oxford University Press.
Jaszczolt, Katarzyna M.
2006 Meaning merger: Pragmatic inference, defaults, and compositionality. Inter-
cultural Pragmatics 3(2): 195–212.
Johnson, Norman, Randolph Cooper and Wynne Chin
2008 The effect of flaming on computer-mediated negotiations. European Journal
of Information Systems 17(4): 417–434.
Jucker, Andreas H. and Irma Taavitsainen
2000 Diachronic speech acts: Insults from flyting to flaming. Journal of Historical
Kayany, Joseph M.
1998 Contexts of uninhibited online behavior: Flaming in social newsgroups on
Usenet. Journal of the American Society for Information Science 49(12):
1135–1141.
Kaye, T.
1985 Computer-Mediated Communication Systems for Distance Education: Report
on a Study Visit to North America September/October 1985. Milton Keynes:
University of Open Institute of Educational Technology.
Kerr, Elaine B. and Starr Roxanne Hiltz
1982 Computer-Mediated Communication Systems: Status and Evaluation. New York/
London: Academic Press.
Keysar, Boaz
2007 Communication and miscommunication: The role of egocentric processes.
Intercultural Pragmatics 4(1): 71–84.
Psychologist 39: 1123–1134.
Kirsh, Elana
2012 Untangling the Web: How Facebook ruined my holiday. Jerusalem Post May
21st: http://www.jpost.com/Opinion/Columnists/Article.aspx?id=270839.
Kraut, Robert, Jolene Galegher, Robert Fish and Barbara Chalfonte
1992 Task requirements and media choice in collaborative writing. Human-Com-
puter Interaction 7: 375–407.
Kraut, Robert, Michael Patterson, Vicki Lundmark, Sara Kiesler, Tridas Mukopadhaya and
William Scherlis
1998 Internet paradox: A social technology that reduces social involvement and psy-
chological well-being. American Psychologist 53(9): 1101–1137.
Kruger, Justin, Nicholas Epley, Jason Parker and Zhi-Wen Ng
2005 Egocentrism over e-mail: Can we communicate as well as we think?’ Journal
of Personality and Social Psychology 89(6): 925–936.

Authenticated
Lea, Martin, T. O’Shea, P. Fung and Russell Spears

1992 ‘Flaming’ in computer-mediated communication. Observations, explanations,
implications. In: Martin Lea (ed.), Contexts of Computer-Mediated Communi-
cation, 89–112. New York: Harvester Wheatsheaf.
Lea, Martin and Russell Spears
1991 Computer-mediated communication, deindividuation and group decision-mak-
ing. International Journal of Man-Machine Studies 34: 283–301.
Lee, Hangwoo
2005 Behavioral strategies for dealing with flaming in an online forum. The Socio-
logical Quarterly 46(2): 385–403.
Locher, Miriam A.
2006 Advice Online: Advice-Giving in an American Internet Health Column.
MacCulloch, John Arnott
1930 Eddic Mythology: The Mythology of All Races in Thirteen Volumes. Vol. II.
Boston: Archaeological Institute of America, Marshall Jones Company.
Maltby, John, Liz Day, Ruth M. Hatcher, Sarah Tazzyman, Heather D.Flowe, Emma
J. Palmer, Caren A. Frosch, Michelle O’Reilly, Ceri Jones, Chloe Buckley,
Melanie Knieps and Katie Cutts
2015 Implicit theories of online trolling: Evidence that attention-seeking concep-
tions are associated with increased psychological resilience. The British Psy-
chological Society 1(1): 1–19.
McCosker, Anthony
2014 Trolling as provocation: YouTube’s agonistic publics. Convergence 20(2):
201–217.
Memon, Amina, Aldert Vrij and Ray Bull
2003 Psychology and Law: Truthfulness, Accuracy, and Credibility. 2nd ed. Chiches-
ter: John Wiley.
Millard, William B.
1997 I flamed Freud: A case study in teletextual incendiarism. In: David Porter
(ed.), Internet Culture, 145–160. New York: Hampton Press.
Moor, Peter J., Ard Heuvelman and Ria Verleur
2010 Flaming on Youtube. Computers in Human Behavior 26: 1536–1546.
Moulitsas, Markos
2008 Ignore ‘Concern Trolls’. The Hill January 09th. http://thehill.com/markos–
moulitsas/dems–ignore–concern–trolls–2008–01–09.html 06/06/09.
Naraine, Ryan
2007 The 10 biggest Web annoyances. www.pcworld.com December: 141–148.
Nigam, Kamal and Matthew Hurst
2004 Towards a robust metric of opinion. Proceedings of the AAAI Spring Sym-
posium on Exploring Attitude and Affect in Text: Theories and Applications,
598–603. Menlo Park, CA: The AAAI Press.
Nitin, Chanderwal, Ankush Bansal, and Deepak Kanzashi
2011 Understanding perceived flaming tendencies on social networking sites: An
exploratory study. Issues in Information Systems 12: 425–435.
O’Sullivan, Patrick B. and Andrew J. Flanagin
2003 Reconceptualization ‘Flaming’ and other problematic communication. New
Media and Society 5(1): 67–93.

Authenticated
520 Claire Hardaker
Panteli, Niki
2002 Richness, power cues and email text. Information and Management 40(2):
75–86.
Phillips, David J.
2002 Negotiating the digital closet: Online pseudonymity and the politics of sexual
identity. Information, Communication and Society 5(3): 406–424.
Phillips, Whitney
2015 This Is Why We Can’t Have Nice Things: Mapping the Relationship between
Online Trolling and Mainstream Culture. Cambridge, MA: The MIT Press.
Placks, Simon James
2003 Interpersonal Deceit and Lie-Detection Using Computer-Mediated Communi-
cation. Durham: University of Durham.
Plato
2007 The Republic. 3rd ed. London: Penguin Classic.
Recanati, Francois
1986 On defining communicative intentions. Mind and Language 1: 213–242.
Reicher, Steve, R. Mark Levine and Ernestine H. Gordijn
1998 More on deindividuation, power relations between groups and the rxpression
of social identity: Three studies on the effects of visibility to the in-group.
British Journal of Social Psychology 37: 15–40.
Rubin, Victoria L.
2010 On deception and deception detection: Content analysis of computer-medi-
ated stated beliefs. ASIST 2010 October 22nd–27th, 1–10. Pittsburgh, PA. http://
onlinelibrary.wiley.com/doi/10.1002/meet.14504701124/pdf.
Sanderson, David W.
1993 Smileys: Express Yourself Sideways. Sebastopol, CA: O’Reilly and Associates.
Saul, Jennifer
2001 Critical Studies: Wayne A. Davis, Conversational implicature: Intention and
convention in the failure of Gricean theory. Nous 35: 630–641.
Scott, Derek
1995 The effect of video games on feelings of aggression. The Journal of Psychol-
ogy 129: 121–132.
Searle, John R.
1983 Intentionality. Cambridge: Cambridge University Press.
Searle, John R.
1990 Collective intentions and actions. In: Philip R. Cohen, Jerry L. Morgan and
Martha E. Pollack (eds.), Intentions in Communication, 401–415. Cambridge,
MA: Bradford Books.
Shachaf, Pnina and Noriko Hara
2010 Beyond vandalism: Wikipedia trolls. Journal of Information Science 36(3):
357–370.
Shin, Jiwon
2008 Morality and Internet behavior: A study of the Internet troll and its relation
with morality on the Internet. In: Karen McFerrin, Roberta Weber, Roger
Carlsen and Dee Anna Willis (eds.), Proceedings of Society for Information
Technology and Teacher Education International Conference 2008, 2834–
2840. Chesapeake, VA: AACE.

Authenticated
Siegel, Jane, Vitaly J. Dubrovsky, Sara Kiesler and Timothy W. McGuire

1986 Group processes in computer-mediated communication. Organizational
Behaviour and Human Decision Processes 37(2): 157–187.
Social Media and Criminal Offences
2014 Communications Committee – First Report. House of Lords Select Committee.
London.
Spencer-Oatey, Helen D. M.
2005 (Im)Politeness, face and perceptions of rapport: Unpackaging their bases and
interrelationships. Journal of Politeness Research: Language, Behaviour, Cul-
ture 1(1): 95–119.
Spencer-Oatey, Helen D. M. and Jianyu Xing
2006 Impoliteness in intercultural interaction: Conventions, intentionality and the
role of the interpreter. Linguistic Impoliteness and Rudeness: Confrontation
and Conflict in Discourse, 03–04 July. University of Huddersfield.
Sternberg, Janet
2000 Virtual misbehavior: Breaking rules of conduct in online environments. Pro-
ceedings of the Media Ecology Association 1: 53–60.
Stivale, Charles J.
1997 Spam: Heteroglossia and harassment in Cyberspace. In: David Porter (ed.),
Internet Culture, 133–144. New York: Routledge.
Strom, Paris S. and Robert D. Strom
2005 When teens turn cyberbullies. The Education Digest, December: 35–41.
Subasic, Pero and Alison Huettner
2001 Affect analysis of text using fuzzy semantic typing. IEEE Transactions on
Fuzzy Systems 9(4): 483–496.
Tepper, Michele
1997 Usenet communities and the cultural politics of information. In: David Porter
(ed.), Internet Culture, 39–54. New York: Routledge.
Thompson, Clive
2009 Clive Thompson on the taming of comment trolls. Wired Magazine, March
23rd. http://www.wired.com/techbiz/people/magazine/17–04/st_thompson
30/03/09.
Thompson, Robert J.
2008 Grades of meaning. Synthese 161: 283–308.
Topçu, Çigdem, Özgür Erdur-Baker and Yesim Çapa-Aydin
2008 Examination of cyberbullying experiences among Turkish students from dif-
ferent school types. CyberPsychology and Behavior 11(6): 643–648.
Turkle, Sherry
1990 The psychology of personal computers. In: Tom Forester (ed.), The Informa-
tion Technology Revolution, 182–201. Oxford: Blackwell.
Turner, Tammara Combs, Marc A. Smith, Danyel Fisher and Howard T. Welser
2005 Picturing Usenet: Mapping computer-mediated collective action. Journal of
Computer-Mediated Communication 10(4). http://jcmc.indiana.edu/vol10/
issue4/turner.html 05/01/11.
Unknown
1996 The Troller’s Faq. http://www.altairiv.demon.co.uk/afaq/posts/trollfaq.html
10/01/01.

Authenticated
522 Claire Hardaker
Utz, Sonja
2005 Types of deception and underlying motivation: What people think. Social Sci-
ence Computer Review 23(1): 49–56.
van Schie, Emil G. M. and Oene Wiegman
1997 Children and video games: Leisure activities, aggression, social integration,
and school performance. Journal of Applied Social Psychology 27: 1175–1194.
Vaughan, Jill and Lauren Gawne
2011 I can has language play: Construction of language and identity in Lolspeak.
Australian Linguistics Society Annual Conference December 7th. http://vimeo.
com/33318759.
Vinagre, Margarita
2008 Politeness strategies in collaborative e-mail exchanges. Computers and Edu-
cation 50(3): 1022–1036.
Virkar, Shefali
2014 Trolls just want to have fun: Electronic aggression within the context of e-par-
ticipation and other online political behaviour in the United Kingdom. Inter-
national Journal of E-Politics 5(4): 21–51.
Vrij, Aldert
2000 Detecting Lies and Deceit. The Psychology of Lying and the Implications for
Professional Practice. Chichester: Wiley.
Vrij, Aldert, Katherine Edward, Kim P. Roberts and Ray Bull
2000 Detecting deceit via analysis of verbal and nonverbal behavior. Journal of
Nonverbal Behavior 24(4): 239–263.
Weckerle, Andrea
2013 Civility in the Digital Age: How Companies and People Can Triumph over
Haters, Trolls, Bullies, and Other Jerks. London: Pearson.
Whitty, Monica T.
2004 Cyberstalking. NSW Crime Division: Criminology Research Council.
Wilson, Samuel M. and Leighton C. Peterson
2002 The anthropology of online communities. Annual Review of Anthropology
31(449–467): 449.
Yates, JoAnne and Wanda J. Orlikowski
2002 Genre systems: Structuring interaction through communicative norms. Jour-
nal of Business Communication 39(1): 13–35.
Zarsky, Tal. Z.
2004 Thinking Outside the Box: Considering Transparency, Anonymity, and Pseu-
donymity as Overall Solutions to the Problems of Information Privacy in the
Internet Society. Unpublished PhD Thesis. New York: Columbia Law School.
Zdenek, Sean
1999 Rising up from the mud: Inscribing gender in software design. Discourse and
Society 10(3): 379–409.
Zhou, Lina, Judee K. Burgoon, Jay F. Nunamaker and Doug Twitchell
2004 Automating linguistics-based cues for detecting deception in text-based asyn-
chronous computer-mediated communications. Group Decision and Negotia-
tion 13(1): 81–106.

Authenticated
19. Narration
Ruth Page
Abstract: The analysis of narration in social media contexts exemplifies trends in

discourse-analytic approaches to narrative research, which have shifted towards
a more flexible view of narrative that takes into account the contexts in which
narration takes place. This contextualised view of narration is important, for it
takes into account the medium factors (Herring 2007) that shape the various ways
in which stories are told in social media sites and platforms. In particular, the
multimodal resources present new opportunities and challenges to reconsider how
narratives are produced, consumed and recontextualised by a range of tellers and
for different purposes.
1. Introduction
In this chapter, I set out an overview of the ways in which narration – the practices
of telling a story – have evolved in social media contexts. The various ways in
which narration takes place in social media contexts means that there is no sin-
gle narrative framework which analysts have used as they have sought to make
sense of these examples. Indeed, the stories told in social media genres test earlier
definitions of what might count as a narrative. The structure of this chapter thus
begins with a historical overview of the key definitions taken up within pragmatic
approaches to narrative and documents the shift towards a more flexible model of
narrative dimensions (Section 1). In Section 2, these narrative dimensions are con-
sidered in relation to the affordances of social media contexts, which include the
multimodal range of resources and mediated contexts that are used in the online
examples narration (Section 3). In Section 4, I move on to consider the identity
work that narration in social media contexts might achieve for particular tellers
and how these operate in relation to particular interactions (as discourse identi-
ties) and broader, socio-cultural contexts in which social media are situated (as
transportable identities). Whilst these ongoing concerns with the resources used
in narration and the identity work that narration can achieve bridge the study of
narration in offline and social media contexts, I close in Section 5 with a summary
of the most recent developments in computer-mediated discourse analysis that
employ a narrative approach.
DOI 10.1515/9783110431070-019
Authenticated
524 Ruth Page
2. Definitions and dimensions of narrative
The development of social media genres such as blogs, social network sites and
wikis has allowed people with relatively little technological expertise to publish
stories about their own experiences and the world around them. These examples
of story-telling provided important case studies against which scholars have tested
and refined existing definitions of and frameworks for analysing narrative across
a number of disciplines. Within pragmatics, narrative analysis has been influenced
by a number of research traditions which pre-date the emergence of social media,
such as the ethnopoetic work of Hymes (1996, 1998), studies in discursive psy-
chology, such as Bruner’s (1986, 1991) discussion of narrative and identity, and
the sociolinguistic work of Labov (1972, 2013). These studies scrutinised narra-
tion as it took place in various face-to-face contexts (such as performances and
interviews), but conceptualised the object of study in somewhat different ways.
For example, in Labov’s seminal work, narrative is defined as a particular text type
that can be observed as a discrete verbal artefact. In line with this, he proposed the
minimal definition of narrative in linguistic terms as “one method of recapitulat-
ing past experience by matching a verbal sequence of clauses to the sequence of
events which (it is inferred) actually occurred” (Labov 1972: 359–60). In contrast,
Bruner’s (1991) description of a “canonical narrative” relied less on the qualities
of a verbal artefact (the narrative product) and instead evoked cognitive models
in the form of scripts to position narratives as macro-social patterns, understand-
ing narrative as “an instrument of the mind” (1991: 6). These theoretical starting
points are important, for they shaped the parameters against which the norms for
narration could be evaluated. Following Labov’s structural description of a “fully
formed narrative”, narratives are expected to contain a core component, the com-
plicating action, which is resolved in the result or resolution. In addition, Labov
suggested that in order for the story to be told successfully, it must also be tella-
ble, where tellability was realised linguistically in a range of lexico-grammatical
resources brought together under the umbrella heading of evaluation. In Bruner’s
response to Labov’s work, he argued that tellability implied both normativity and
breach as central components for narrative. Canonical narratives thus typically
report and revolve around some kind of “trouble” (Bruner 1991: 16), in ways that
are culturally shaped and may bring to light the ways narratives operate as a form
of cultural legitimacy.
In the decades following the work of Labov and Bruner, narrative research in
pragmatics has considered a complex range and variety of storytelling practices
across different contexts. In moving beyond the narrative examples elicited by
Labov, scholars recognised increasingly that the formal properties of a narrative
could not be thought of in universal, abstract terms and instead a more flexible
approach to analysing narration as a process was required. This move has been
described as a shift towards narrativity, which following a cognitive approach

Authenticated
Narration 525
focuses on the interpretation of a text as more or less like a narrative (Ryan 2007).
Interpreting the narrativity of a text repositions the Labovian framework as one
option amongst many possibilities for narrative structure, albeit an option that is
(from a cognitive) perspective regarded as a proto-typical example of the category
against which other storytelling examples might vary. In their work on story gen-
res, Martin and Plum (1997) suggest that other story genres can include anecdotes
and recounts. The narrativity of these different genres varies, where some stories
might foreground affective response whilst others are structured as an unevaluated
event line. These genre-based differences help explain the patterns of storytelling
in a number of contexts, including those in social media. For example, in my anal-
ysis of blogs where the authors documented their experiences of being diagnosed
and treated for cancer, the Labovian pattern of a “fully formed” narrative rarely
occurred but anecdotes and recounts were plentiful (Page 2012).
Building on a more flexible, contextualised approach to narration, Ochs and
Capps (2001) argue that the structural qualities of the narrative text are but one
of five dimensions that may characterise any given storytelling instance. The five
dimensions include the structural features of narrative (linearity) but place a clear
emphasis on contextual elements, such as the persons involved in the telling (tell-
ership), and the sites in which the telling takes place (embeddedness), which may
inform the interpretation of moral stance and the assessment of selected narrative
content as more or less tellable. Ochs and Capps (2001) point out that each of these
dimensions can be realised in different ways, so that the kinds of narratives that
might be scrutinised can be thought of in more flexible terms. Whilst the stories
analysed by Labov were typically single teller accounts of highly tellable events,
which were told in relatively detached performances, other stories can be told by
many tellers, about mundane topics and be embedded in complex interactional
contexts. This is particularly the case in social media narration, which can incor-
porate the contribution of many tellers in various kinds of collaborations, be told
in episodic, post-by-post forms and be embedded in online templates quite unlike
the face-to-face interview, conversational or performative contexts considered in
earlier research. For example, Eisenlauer and Hoffmann (2010) explain how dif-
ferent weblog genres are characterised by multi-linearity in unfolding posts and
posts with appended comments that allow multiple reading paths, quite unlike the
“default narrative” suggested by Labov. The interaction with these mediated forms
of narration also differs in its participatory potential, so that co-tellers can contrib-
ute with different degrees of interactivity. This can include the kinds of interaction
found in print or conversational narrative (such as cognition), but also includes
options for clicking on hyperlinks or adding an additional comment (Eisenlauer
and Hoffmann 2010: 102–105).
Further interest in non-canonical examples of narration has gathered under the
umbrella of what has begun to be known as “small stories research”. Small story
research positions itself as a strategic shift; an “antidote to canonical research”

Authenticated
526 Ruth Page
(Bamberg and Georgakopoulou 2008: 377) in which the analytic focus is turned
towards “a gamut of underrepresented narrative activities” (2008: 381) that fall
beyond “fully formed” stories. Small stories include present tense accounts of
events as forms of breaking news, hypothetical stories and projections of future
events (Georgakopoulou 2007), refusals to tell (Bamberg and Georgakopoulou
2008), narrative stance-taking as an orientation towards telling a story, rather than
launching into the full account (Georgakopoulou 2013) and conversational shared
stories (Georgakopoulou 2005, 2007). These small stories are highly embedded in
their interactional contexts, often falling on the periphery or margins of the main
topic of the talk. In line with the emphasis on their emergent, interactional con-
texts, small stories are often co-constructed rather than the accomplishment of a
single narrator. Thus in her discussion of family stories, Georgakopoulou (2005)
notes how different members of the audience add evaluation to enhance, elaborate
or contest the versions of a story that are retold at social gatherings as kinds of
shared stories. The stories are typically fleeting rather than fully developed and
often report events which are mundane and every day in nature rather than land-
mark examples of tellable topics.
Like the sociolinguistic traditions of narrative research, the methods used in
small story research are empirically oriented and include close attention to the
micro-analysis of the narration, usually in verbal form. However, drawing on posi-
tioning theory as developed by Davies and Harré (1990), small story research
attempts to situate the micro-analysis of the narrative text in relation to the mac-
ro-social narratives that are invoked by the tellers. Crucially, in small stories
research this interpretive move takes place by analysing the narration in relation to
its interactional context, such as the turn-by-turn conversational co-text in which
the narrators are positioned. In this way, small stories research moves beyond the
limitations of structuralist models such as Labov’s that treat narrative primarily as
a product, rather than a process. At the same time, it also provides a method that
allows researchers to move between the empirical analysis of narrative practice
and the interpretation of what those practices might mean for the broader, mac-
ro-social meanings and ways of thinking that shape tellers’ experiences. The use
of positioning in small stories research has developed for the most part by tracing
how the identities of speakers emerge in conversational storytelling (see Depper-
mann 2013). However, more recently, examples of the ways in which small stories
have been analysed in new media contexts include Georgakopoulou’s (2008) study
of a teenaged girl’s use of breaking news in MSN messages as a way of maintain-
ing her social identity, in particular by indexing her gender within a heteronor-
mative framework. Other examples include the updates of breaking news told in
Facebook updates (Page 2012) which are co-constructed on semi-private accounts
between the Facebook members and their Friend lists, for example by one member
who might post an update and a ‘Friend’ who might post an evaluative comment
or a request for further information to extend the story.

Authenticated
Narration 527
3. Social Media contexts for narration: media affordances and

multimodality
In Ochs and Capps’ dimensional approach and in the small stories paradigm, the
narrative practices are understood in relation to the localised, co-texts of interac-
tion and broader, socio-cultural contexts in which those interactions take place.
The focus on a meso-level of analysis is a crucial turning point for theorising narra-
tion as a practice rather than understanding narrative as textual product. However,
many of the examples of storytelling considered by Ochs and Capps (2001) and
within the small stories paradigm were part of spoken interactions that took place
in offline, face-to-face contexts. Even in these offline situations for telling, con-
text is a multifaceted notion. For some researchers, the turn-by-turn co-texts of a
conversation may be the most important aspect of the communication, as in a con-
versation-analytic approach (see Antaki and Widdicombe 1998). Other researchers
may be interested in wider, cultural contexts. However, the intractable nature of
these broader contexts makes them difficult to identify. From a socio-cognitive
perspective, Van Dijk (2009) argues that social contexts are intersubjectively con-
structed as a set of features that are made relevant in any given communicative
context. He points out that some aspects of the context may be referred to, such
as the political affiliation of a particular speaker within parliament, whilst other
aspects of the context, such as the speaker’s dress code are not relevant. When we
turn to social media sites and platforms, the contextual complexity of narration
increases further still. As Jones (2005) pointed out, this can include the online con-
texts of the content created on a screen and uploaded into the particular formats of
a platform or communicative tool, and also the offline contexts in which the inter-
action takes place, such as accessing or creating the content in a public or private
space and through a particular kind of device, such as a smart phone. On the one
hand, this opens up new avenues for inquiry, for generally, unlike the ephemeral
nature of face-to-face narration, many of the formats found in many social media
sites and platforms preserve the interactional context in tractable forms that allow
the researcher to trace the processes of narrative production and reception. For
example, the quasi-conversational exchanges between participants can be archived
in posts and responses, and further contextual information such as the time and
place of the interaction may also be observed through meta-data appended to the
posts, comments and other kinds of interaction. In Page, Harper and Frobenius
(2013) we observed how the options within the social networking site Facebook
allowed the participants to indicate who they were with (through tagging other site
members) or show where they had been (through adding a “check in” that used
GPS data to show where a particular post had been published). On the other hand,
we might wish to be cautious about taking a media-blind approach and simply
transposing concepts of narration from face-to-face interactions to those found
online. Whilst there are clear points of continuity between the stories narrated in

Authenticated
528 Ruth Page
face-to-face contexts and those found in social media sites and platforms, there
are also important differences that relate to the mediated nature of the interaction.
Within computer-mediated discourse analysis, these differences have been
discussed in relation to media affordances. Gibson (1977) coined the term ‘affor-
dances’ to explain the ways in which the characteristics of a particular environ-
ment both constrained and enabled particular kinds of activity to take place. How-
ever, rather than assume that the technology directly determines what kinds of
narration take place in social media, the communicative functions of a particular
site or platform can be taken up, adapted and appropriated more or less creatively
by the participants involved in the narrative interaction. These media affordances
are thus part of the broader communicative situation in which narration takes
place, and inter-relate to other, non-technological characteristics. Herring’s (2007)
multi-faceted classification of computer-mediated discourse developed Hymes’
(1974) ethnography of communication model and sets out a useful set of medium
and situation factors that can be used to compare and contrast the different social
media contexts. In her model, the situation factors include those which span many
kinds of technology, such as the participants, the norms and purpose of the com-
munication and the linguistic code. An example of social media storytelling based
on the situation factors of the interaction is found in Page (2014), where I com-
pared how the editors of the English language Wikipedia and the Italian Wikipe-
dia co-constructed different narratives that reported the death of British student
Meredith Kercher. These articles varied in the selection of their content, and in
their language variety (English or Italian), and in their purpose (to represent the
suspects as more or less blameworthy in relation to the crime). Alongside situation
factors, Herring (2007) sets out a list of medium factors which may function as
affordances that open up or constrain particular narrative dimensions. The medium
factors include whether the communication is synchronous or asynchronous, pri-
vate or public, the choices for identity representation (such as anonymity), size of
the message, the persistence of the transcript, message format and channels for
communication. For example, the restrictions on a Twitter post which currently
prohibit messages of more than 140 characters mean that the stories told in this
context are briefer than those told in a Wikipedia page, which does not impose
any such restriction on length. These medium factors can also influence the tem-
poral orientation of a story, where the fast-paced nature of interacting via a site
like Twitter is reflected in the common use of breaking news stories (Page 2012),
whilst other sites where the asynchronicity and archiving allows its members a
longer timespan with which to interact with the text (such as Trip Advisor, Vásquez
2015), to foreground a more retrospective temporal orientation to the reports they
post as part of an online review.
The last of Herring’s medium factors focuses on the multimodal range of the
particular sites, which may include options to use images, sound, and audio-visual
content as part of the narration. The multimodal nature of social media narration

Authenticated
Narration 529
poses particular challenges to the models of narrative that were designed to ana-
lyse stories conveyed primarily in spoken words. However, within literary-critical
narratology, there has been an increased interest in the ways in which narratives
can be told in different media (Ryan 2004). Within the project of transmedial
narratology, one of the key questions that scholars have tried to answer is the
extent to which non-verbal semiotic resources can tell a story. Definitions of nar-
rative which foreground verbal factors such as past tense verbs and sequences of
clauses are difficult to apply to non-verbal phenomena such as pure music or still
images. Nonetheless, a more cognitive approach suggests that in certain condi-
tions, visual content can prompt the interpretation of narrativity (Ranta 2013).
For example, if the content of the image includes human participants who are
arranged in compositions which imply the transfer of actions from one participant
to another (Van Leeuwen and Kress 1996, Van Leeuwen 2008), a sequence of
events may be inferred by the viewer. In a similar vein, comics, graphic novels
and even films typically follow conventional spatio-temporal arrangements and
presentations to induce cognitive scripts which tie in with this notion of narrativity
(see McCloud 1993), Thus in some media (such as images), narratives are more
evoked than (explicitly) expressed while in others (like verbal art) it is the other
way round. Likewise, the potential for musical sequences to mirror a plot-like
progression from complication to resolution have been interpreted as narrative
resources (Rabinowitz 2004), especially where these are combined with verbal
content such as song lyrics and genres that carry particular narrative qualities, such
as opera (Hutcheon and Hutcheon 2010) or musical theatre (Krogh Hansen 2010).
More importantly for social media examples of narration, multimodality draws
attention to the ways in which semiotic resources are combined in ways that
contribute to the narrative dimensions of tellability, tellership and linearity. For
example, in some social media contexts, there can be a clear division between the
content that is conveyed in each of the multimodal elements or contributed by dif-
ferent tellers. In the video-sharing site, YouTube, the stories narrated in videos can
incorporate still and moving images, sound, words, gesture and dance. In contrast,
the comments are verbal and multimodal content such as image, emojis or links
to other sites are for the most part absent. In turn, this shapes the ways that the
video-maker and the commenters can contribute to the narration: commenters can
only write a response and cannot alter the audio-visual content. For example, in
the live-streamed trial of South African athlete, Oscar Pistorius, for the murder of
Reeva Steenkamp, the videos posted to YouTube include video recordings of the
trial, television news broadcasts, and parody videos featuring images and music.
However, the comments to these videos could include only written responses.
In contrast, social network sites like Facebook allow comments to also include
images or embedded video and are wider in the semiotic range that they afford
the site members. In other cases, where stories are retold in mashups, different
semiotic resources can be recontextualised and combined with new content. For

Authenticated
530 Ruth Page
example, in Georgakopoulou’s (2015) study of videos responding to a Greek polit-

ical incident, she showed how the verbal content and actions were transferred from
one video to another, but the orientation shown in the video changed to reframe the
political incident as a parody.
The multimodal combinations that are possible in social media narration also
challenge early text-based approaches that understand narrative as a product that
can be decontextualized from its interactional context. In social media narration
it remains important to distinguish between the reported events (the narrative
discourse) and the context of narration. However, in other cases, the boundaries
between the narrative discourse and certain aspects of the narrational context are
less clear cut. A case in point is the quasi-real time reporting of events in social
media where the time of reporting and the time of the reported events are assumed
to be similar. In these instances, such as the updates posted to sites such as Twitter,
Facebook and Instagram, the timestamp may contextualise both the narration and
the reported events. The possibility of embedding content relevant to the reported
events in the template around the space where the reported events are published has
increased over time. The medium factors of the international social network site,
Facebook, illustrate this process (cf. Eisenlauer, Ch. 9, this volume). In its early
days, the Facebook template provided a space for its members to post updates. The
update template was a text box, in which the member could complete the default
statement “[username] is …” The updates published on personal Facebook accounts
to the site member’s ‘wall’ could contain many kinds of text, including small stories
such as breaking news and projections (Page 2010, 2012). As the design of status
update evolved, the template changed to incorporate interactions from other tellers,
in the form of comments and response buttons such as the ‘like’ and most recently,
‘reaction’ icons. The modality of the updates also changed as it became possible to
upload images and video directly into the status update (rather than archiving these
separately in albums). In addition, the meta-data published in the status update
template also changed to include the possibility of representing places (through
check in buttons) and participants (by tagging other site members). Given that this
multimodal content can contribute to the narrative, for example, by providing ori-
entation or evaluation (Page, Harper and Frobenius 2013) of the events reported in
the update, it becomes difficult to treat the narrative as an isolated verbal product
alone. Instead, there has been a shift towards seeing the constellation of the mul-
timodal content, such as the written update along with images, video, comments
and responses as a composite “wall event” (Androutsopoulos 2014: 8) in which the
narration takes place and can be shared by multiple tellers, such as the updater, the
commenters and possibly other participants if the image and video in question were
produced in another context and then reproduced in the Facebook update.
The episodic nature of many social media sites, where stories are told in mul-
tiple posts over time (such as blog or forum posts), or through the incremental
evolution of documents such as wiki pages, stretches the boundaries of social

Authenticated
Narration 531
media narration further still. In line with research that traces iterative storytell-
ing as a form of identity construction over time, such as Wortham and Rhodes’
(2015) work on chains of narratives across speech events, the narration that can
be observed in social media contexts is highly open-ended and can evolve across
multiple, intertextually networked segments. Examples of mediated chains of sto-
rytelling can include updates posted over time to an individual’s profile or per-
sonal account on a social media site. The narrativised chains can be interpreted
relative to the architecture of the site’s archive. For example, Facebook exploits
this longitudinal unfolding of a site member’s narration over time by archiving
these in a timeline which appears searchable so that earlier updates and content
can be retrieved by the member or Friends viewing their profile. At the time of
writing, it is also possible to use semi-automated features, such as Facebook’s ‘On
this Day’ service to allow the site to reinsert content from an earlier period into
the current timeline of its members. However, the ways in which people make use
of multiple social media contexts for distributing and redistributing their stories
can make tracing these narrative chains a complex methodological process. For
example, Adami (2014) reports on the ways in which food bloggers cross posted
their texts (including stories) across interconnected platforms such as Twitter,
Facebook, Pinterest, Instagram and YouTube. Chains of stories do not just unfold
with open-ended linearity over time, but also across multiple sites of interaction,
with varying media factors and audiences. In her work, Adami argues for an eth-
nographically informed model of analysis that traces the cross-posting practices of
the blogger over time, taking into account the multimodal analysis of how the nar-
rative was remediated according to its new context. Adami’s analysis shows how
the process of social media narration can operate as process of assemblage, where
the written report of events on the blog post was created alongside other narrative
artefacts (such as video and images) that appeared on other sites (YouTube and
Instagram). The choice to assemble, remix and remediate the narrative content
across different sites reflected not only the medium factors of each site in question
(such as the restrictions on message length in Twitter, or the audio-visual empha-
sis of video and photo sharing sites), but also the different audiences with whom
the blogger engaged. For example, the multimodal composition of the words and
images projected closer social distance between the blogger and her audience on
Instagram, than did her use of the video and advice-giving tips on the noticeboard
of Pinterest (Adami 2014: 239). This example illustrates neatly how the analysis of
social media storytelling must stretch beyond the earlier, mono-modal work in dis-
course-analytic narrative research but retains similar concerns found in this earlier
work that considers narratives as a resource for negotiating the tellers’ identities.

Authenticated
532 Ruth Page
4. Tellers and narrative identity
The contextualising impetus in narrative inquiry exemplified in the work of Ochs

and Capps (2001) and extended within the small stories paradigm (Georgakopou-
lou 2007) moved the focus away from the narrative text as the sole object of scru-
tiny and towards broader concerns regarding how the narration might achieve
particular interpersonal outcomes for the participants. The increasing interest in
the tellership as a narrative dimension and the sites of telling (Georgakopoulou
2011) in which the narration was embedded went hand-in-hand with the broader
research that began to analyse identity as constructed discursively through interac-
tion (Antaki and Widdicombe 1998). The contexts of social media might seem to
offer the environment par excellence where this trend of narrative research might
further develop. First, the text-based forms of online interaction suggested that
online identity could be treated as a ‘mask’, separate from other, offline identity
(Turkle 1995) and so is in line with the view of identity as a fluid, discursive
project, rather than as a stable, essentialist attribute of tellers. Second, the identi-
ties-in-interaction approach benefited from the potential of social media contexts
to open up mediated forms of collaboration to enable innovative opportunities
for the co-construction of narration between multiple tellers. Narrative tellership
can be approached in relation to these inter-related aspects of identity. Zimmer-
man (1998) distinguishes between these aspects by differentiating a participant’s
transportable identity (the characteristics and attributes that are carried from one
interactional context to another) from the situated identities (the generic or insti-
tutional roles that a participant can adopt within a specific speech event, such as
a radio show host – interviewer, doctor – patient and so on) and discourse identi-
ties (the localised turn-by-turn roles determined by the interaction) that any given
participant may take up. The affordances of different social media sites shape the
transportable, situated and discourse identities for tellers, each of which underpin
the storytelling rights (Shuman 1986, 2005, 2015) for a teller to contribute to the
narration in the first place.
The medium factors that shape transportable identity include the options for
participants to represent their identity in ways that are more or less anonymous, and
more or less connected with aspects of the identities that are constructed in other
contexts outside social media. Typically, this kind of information about the tellers
is located in the profiles that a participant can create as they become members of a
particular site. The representation of the teller’s transportable identities is related
to notions of authenticity, and the assumption that the right to tell a story is asso-
ciated with personal experience of the reported events. As Schiffrin (2006) points
out, this assumption is rooted in Labov’s research, where he found that stories of
vicarious experience were seldom as fully evaluated as those where the narrator
had participated in events first hand. In social media contexts, despite early claims
that online environments would open up possibilities for identity play, expecta-

Authenticated
Narration 533
tions of authenticity persist and may be embedded in the tellership rights that are
formulated in site specific terms and conditions. Some sites encourage their mem-
bers to use their real names. For example, Facebook’s terms state that ‘We require
people to provide the name they use in real life; that way you always know who
you’re connecting with’ (Facebook 2016). Similarly, Twitter’s terms and condi-
tions (2016) prohibit impersonation and require that parody, commentary and fan
accounts are demarcated as such in the account name and profile information. In
contrast, other sites do not make such restrictions and may discourage their mem-
bers from disclosing information that identifies aspects of their offline identity.
For example, whilst Wikipedia allows its members to create usernames that can
be a person’s ‘real name’, in their Guidelines for new Users (2016) this is framed
as a ‘risk’ and outing a Wikipedia Editor’s offline identity is considered a serious
offence. These choices are related to the generic purpose of the site in question,
where the tellership rights do not rest upon the narrators’ first-hand experience of
events. For example, as an online encyclopedia, Wikipedia relies on the neutral
synthesis of existing, published knowledge rather than the sharing of personal
experience or opinion to shape the content of the articles that are co-constructed
by the editors. The rights for a teller to contribute to a particular Wikipedia article
are thus shaped by the core content policies, no original research and verifiability,
which propose that the contribution to articles be supported by ‘reliable sources’.
The first-hand, individual experience of events is replaced by the authority of
peer-reviewed, published source material as a criterion for the tellership.
Of course, participants do not necessarily respond to the policies, terms and
conditions of any given site in a uniform way. Wikipedia editors may choose to
use their real names in their accounts, and frequently debate what kinds of source
material can be regarded as reliable. It is possible for Facebook members to breach
the norms for authentic tellership, for example by posting updates on another
member’s account (Page 2012, 2015). The motivation for and relational outcomes
of these choices are often embedded in offline contexts of interactions that suggest
that a sharp distinction between the offline and online aspects of narration is dif-
ficult to uphold. For example, the Facebook participants who engaged in playful
impersonation were part of an offline community of practice, where the relational
work that emerged from these breaches of authenticity were for some members of
the audience a rapport-enhancing form of teasing, whilst for other members of the
audience the breaches in authenticity could result in loss of face, embarrassment
and ridicule. In part, the differences in the relational outcome reflect the distinctive
nature of the contexts in which tellership takes place in some social network sites.
Some social media sites are characterised by what has been called context collapse
(Marwick and boyd 2010), which describes the ways in which a networked audi-
ence (such as a Friend list on Facebook, or a Follower list on Twitter or Instagram)
can bring together groups of participants who would not typically be united in the
same interactional context. This may pose particular challenges for tellers who, on

Authenticated
534 Ruth Page
the one hand, may opt for authentic self-disclosure on a site like Facebook, yet not
wish to disclose exactly the same ‘breaking news’ to all members of their Friend
list via the one-to-many broadcasts of a status update. Prior to the introduction
of privacy settings in Facebook as affordances which allow members to filter the
intended audience for updates, some tellers negotiated this pragmatic dilemma by
narrating stories that were intentionally mundane in their tellability (Page 2012),
or sharing stories that were circulating in the mainstream media (Page, Harper and
Frobenius 2013). More generally, revealing too much in social media contexts
where authenticity is expected can also be interpreted as ‘over-sharing’ personal
information, and echo wider moral panics about the potential for social media sites
to reconfigure the boundaries between the private and public sphere. In yet other
cases, options for tellers to retain anonymity can result in other kinds of norms for
tellability being established. For example, Suler (2004: 13) argued that anonymity
might lead to disinhibition and “toxic outcomes” such as “harsh language, criti-
cism, even threats”. At least some evidence suggests that disinhibition does result
in community norms at the upper end of the mundane-troublesome spectrum in
sites like 4Chan, which require their members to use anonymous representation,
and whose content is deemed inappropriate for consumption in some contexts, as
indicated through filters that categorise these stories as ‘not safe for work’.
At a micro-level, the affordances of social media also provide novel oppor-
tunities for discourse identities to emerge. Social media sites provide different
technological formats for tellers to co-construct the narration via their interactions
with each other. These technological resources can be contrasted along a number
of parameters, including:
– whether the tellers contribute to separate or composite textual units (such as
post by post, or to a single wiki page),
– whether the interaction is controlled by one or all tellers (such as moderated or
unmoderated forums),
– the types of interaction that are possible (adding, altering or removing con-
tent),
– the design of the interactional context (how replies and responses are con-
nected to the originating post).
These factors can shape the discourse identities and interactional contexts in which
narration is embedded. In some cases, this may result in formats that resemble
face-to-face involvement in the narration. Examples include discussion forums,
where tellers may respond to an original post with second stories, evaluation,
requests for clarification or further breaking news in subsequent posts. In other
cases, the asynchronous, collapsed contexts of social network sites mean that the
interactional coherence associated with face-to-face co-narration is reconfigured.
For example, social media resources such as hashtags can be used in some sites
(such as Twitter) to aggregate all content containing that search term in a sin-

Authenticated
Narration 535
gle thread. This offers particular, mediated opportunities for co-narration between
multiple tellers who contribute content and evaluation to an emerging story around
a particular topic (such as the reports of a particular event, or commentary on a
television show). Unlike the interactional coherence of face-to-face conversation
that enables the co-construction of a narrative to be negotiated turn-by-turn, in
hashtag threads, the aggregated content is sequenced according to the time that the
post was published to the site and thus the tellers’ contributions need not be organ-
ised as adjacent responses within a single, linear structure. In media studies, the
more diffuse, fragmented streams of content are sometimes described as “ambi-
ent” (Zappavigna 2011), where participants scan, sift and select content rather than
reading each individual post as connected within a singular story (see also Zap-
pavigna, Ch. 16, this volume). In terms of co-tellership, this ambient co-tellership
no longer relies on tellers talking to each other, but rather talking together about a
topic with often little acknowledgement of each other (as indicated through replies
or direct address to other tellers).
The technological formats of other sites, such as those that use wiki pages,
provide options for discourse identities that draw on written, rather than spoken
traditions of narration. Wiki technology allows tellers to add to, alter and remove
the content contributed by another person. The unfolding narrative artefact of the
cumulatively revised document can thus decrease rather than increase in compo-
sition, with open-ended linearity as the text is updated over time. The revisions to
Wikipedia articles exemplify this kind of narration (see Page 2013, 2014), as do
similar technologies such as google docs. Within sites like Wikipedia, deleting the
text written by another editor can result in further, subsequent reactions from other
editors contributing to the article. For example, if one editor deletes the work of
another, the other editor may wish to reinstate the deleted text, in an action that
Wikipedia calls a ‘revert’. The guidance that the Wikipedia community provide for
carrying out a revert suggest that co-tellership may not always be harmonious and
that some kinds of co-narration, such as reversions, can be particularly contentious.
One measure to manage the risks and benefits of collective co-narration involve
assigning tellers particular rights within a particular telling situation. This can take
the form of site-specific situated identities, such as administrators (in Wikipedia),
or being a page or channel owner (in Facebook and YouTube respectively), or
moderating a forum or chat thread. The potential for discourse identities and situ-
ated identities to operate within particular hierarchies, where one teller can control
the contributions of another (for example, by blocking them from a particular site
and hence constraining their tellership rights) suggests that at a local level, co-nar-
ration need not be equal. This raises further questions about how asymmetries in
co-tellership might be embedded in broader, macro-social hierarchies where some
tellers hold greater power than others.
In the early rhetoric of what is sometimes called “Web 2.0” (O’Reilly 2005),
the collaborative potential of social media was optimistically framed as participa-

Authenticated
536 Ruth Page
tory culture (Jenkins 2006), that blurred the distinctions between producers and
consumers in forms of “produsage” (Bruns 2008) such as fanfiction and micro-ce-
lebrity practices (Page 2012). To some extent, the democratising possibilities for
co-tellership remain important political opportunities. Yet as social media have
developed through their early decades, other scholars have critiqued how equal
the opportunities for participation might be, both in terms of types of interaction
(Van Dijck 2013) and more widely in terms of global inequalities that persist in
access to technology and digital literacies (Deumert 2014). At a more localised
level, the co-tellers’ interactions can also be differentiated in terms of their par-
ticipatory potential. Some kinds of interactions contribute more to the creation of
content, whilst others curate that content, or redistribute it through actions such
as ‘sharing’ or reposting it within and across sites. The distribution of these types
of actions can be unequal, where fewer participants contribute to the creation of
content than those who consume it. This is true in many social media sites, such
as Instagram and Twitter, where large numbers of fans redistribute the content
created by the celebrity that they ‘follow’ or in YouTube, where videos can attract
large numbers of viewers that outnumber the accounts that upload video content
(Page forthcoming).
More generally, pragmatic approaches to social media narration might question
the ends to which the co-narration is put. Rather than doing away with economic
and institutional hierarchies, these continue to be played out in the macro-social
positioning of co-tellers within wider narratives. In some cases, companies might
wish to personalise their organization and build their reputation through employee
testimonials (Maagaard 2014). In other cases, companies can respond to the stories
that customers tell in their reviews of products and services (Zhang and Vásquez
2014). In turn, consumers can use narratives as the basis to build their own online
reputation as reviewers (Vásquez 2014). In other cases, celebrity practitioners
use stories that appear to give greater access to their backstage performances, but
are still characterised by synthetic personalisation that treats the audience as an
aggregated fan base (Page 2012). Alongside this, in other socio-economic con-
texts, social media is opening up spaces for stories to be told outside mainstream
or institutionalised media context. For example, in the health sector, narratives of
illness can be shared by advice-seekers in online forums (Locher 2008) and social
network sites (Koteyko and Hunt 2015) as a means of sharing the tellers’ self-man-
agement of medical conditions and thereby contrasting their personal expertise
with the rhetoric of the medical professionals. The development of citizen jour-
nalism practices of curating, commenting on and remixing the news similarly
allows stance-taking towards of first-hand reporting of events that contrast with
that reported through mainstream channels. Where censorship of online interac-
tion remains in place, participants can develop multimodal strategies to reference
political situations. Deumert (2014: 79) describes how in 2010, Chinese bloggers
who commented on the absence from Liu Xiaobo from the Nobel Prize ceremony

Authenticated
Narration 537
in Oslo, used images as an attempt to evade the linguistic restrictions that were
placed on political discussion of this situation. Clearly, the macro-social outcomes
of co-narration for particular tellers in different social media and socio-cultural
contexts are highly varied, but indicate the many ways in which narration can sup-
port, challenge or attempt to negotiate master narratives that position some tellers
or versions of events as dominant whilst marginalising others.
5. Current trends and future directions
The development of devices such as smart phones and tablets that allow par-
ticipants to access, produce and consume social media interactions means that
the contextualising trends in narrative research have begun to take an increasing
interest in the narrative dimension of embeddedness. In Ochs and Capps (2001),
embeddedness was framed in relation to the localised, interactional co-text that
framed the verbal narration. In the case of social media narration, there are other
elements of the context in which these co-texts are embedded. As Jones (2005) has
pointed out, people do not interact with technology in a vacuum but rather these
interactions take place in multi-layered contexts. The contexts for social media
narration include elements such as physical spaces, such as particular locations in
which a person might be situated (such as rooms, buildings, or landscapes). They
may tell stories via social media at particular times or during events (such as post-
ing an update whilst waiting for a bus, or live-tweeting whilst watching television).
The contexts also relate to the devices through which digital sites or platforms are
accessed (such as mobile technologies). The development of GPS (global position-
ing service) technologies within these devices has been further integrated into par-
ticular social media sites and platforms so that the division between physical and
digital contexts are increasingly difficult to maintain. Within the analysis of narra-
tion, this has led scholars to explore the rich and complex ways in which space and
place can be used as narrative resources. Some studies have explored the potential
for mobile devices to provide access to historical stories about particular places.
For example, Greenspan’s (2011) development of storytelling technologies such
as Storytrek make it possible for participants to co-construct historical narratives
by accessing archived material and displaying this on digital maps as they move
through particular spaces within particular location, such as the Rideau canal area
in Ottawa, Canada. In the heritage sector, the use of audio-tours precedes social
media uses of mobile storytelling but has developed more recently to include the
co-construction of oral history projects about particular places such as the Mur-
mur project. The Murmur project (Ruston 2010, Page 2012) allowed participants
to record stories about their lived experience of places, onsite, via cell phones
and then made these recordings available for others to listen to later, either onsite
or remotely via a web archive. From a linguistic perspective, these stories are of

Authenticated
538 Ruth Page
interest for the ways in which the narrators use place and person deictics to co-cre-
ate an imagined, shared interactional space and to re-imagine present locations in
the light of past, personal experience narratives.
The importance of space and place as narrative resources has also been doc-
umented in other day-to-day uses of social media sites via mobile devices. For
example, Cohen (2015) shows how particular kinds of breaking news stories
are part of the narrative practices that young Ethiopian men use when interact-
ing around a site like Tinder. Tinder is a location based service that allows its
members to connect selectively with other people in their immediate geographical
location, and is linked to information drawn from the person’s Facebook or Insta-
gram profile. In his ethnographic study of four, male Ethiopian teenagers, Cohen
showed how these participants engaged in a particular type of breaking news,
world attending sequences, to manage their polyfocal social activities of going
to ‘pick up’ events. Cohen’s study usefully points to the methodology required to
appreciate the complex, situated ways in which mobile storytelling takes place as
narrative practice, including observation of the participants in physical spaces (as
they gather together and hand around mobile devices) and the interactional co-text
of conversations in which the world attending sequences emerged.
The importance of space and place as narrative resources is also important
at a macro-social level. In computer-mediated discourse analysis, there has been
increasing recognition of the multilingual nature of online communication (Danet
and Herring 2007). At the same time, the complex patterns of migration and het-
erogenous urban societies were theorised in relation to the concept of superdiver-
sity. In sociolinguistics, this has been taken up by scholars such as Blommaert
and Rampton (2011) who have pointed to the complex ways in which individuals
might use linguistic repertoires to position their identity within these superdiverse
flows. The place of social media within these global patterns of mobility has drawn
particular attention (Androutsopoulos and Juffermans 2014), for digital contexts
for interaction allow individuals to remain in contact and involved in communities
where they may no longer be geographically resident. The analysis of narratives
of personal experience has begun to be included in this field of inquiry (Heyd
2014). In offline contexts, Myers (2006) examined how speakers might narrate a
single narrative of their place identity. In contrast, Heyd’s (2014) study of posts
made to a Nigerian web forum explores the ethnolinguistic repertoires that the
forum members used when they told ‘narratives of belonging’ which blended their
experiences of living and working in multiple locations. This narrative work builds
on a rich tradition of exploring the role of space and spatialization, particularly in
immigrant narratives (De Fina 2003). However, the methods used by Heyd con-
trast with those used to explore earlier, offline examples of narration. The relative
permanence of some forms of social media interactions mean that it is possible to
compile relatively large scale corpora of narratives and then use concordancing
tools to carry out corpus-assisted analysis of these materials. In Heyd’s (2014)

Authenticated
Narration 539
study, she used word lists to identify the most frequently used lexical labels that
the tellers used to project ethnic and racial distance and proximity between them-
selves and others. She then used qualitative analysis to explore how these fre-
quently occurring labels were used in a range of biographical narratives.
The complexity of the narrative embeddedness in social media narration is also
manifest in the convergence of mainstream and social media. The replicability
of social media which enables the recontextualisation of narrative content means
that the stories told in response to events reported in the mainstream news can be
readily incorporated into the news reports themselves. Likewise in other kinds of
mainstream narrative production, such as reality television series, the audience’s
use of social media as commentary can be co-opted by television corporations as
a means of boosting their online visibility. Typically, this kind of recontextualis-
ation relies on the affordances of hashtags in sites like Twitter to aggregate the
content created on social media before sifting and selecting particular posts to
be reproduced, either within social media (for example, as retweets), or within
the screened programme itself. However, there is no guarantee that the responses
to news events will be characterised by sentiments shared by all the participants.
Instead, the ambient nature and open ended linearity of many social media for-
mats means that there are opportunities for tellers to affiliate and disaffiliate from
the hashtag or from each other in forms of narrative stance-taking. For example,
Giaxoglou (2015) describes how the hashtag #jesuischarlie was used by different
members of the social media site, Twitter, to express affiliation or disaffiliation
towards the attacks on the satirical magazine, Charlie Hebdo in Paris, during 2015.
As these hashtags were translated into different languages (especially English and
French), adapted and recontextualised, the hashtag became an indexical reference
to larger, mediated stories of protest and resistance. Other kinds of multimodal
practices illustrate the variety of narrative stance-taking that are available. For
example, in November 2015, Facebook enabled its members to incorporate an
overlay of the French flag with their profile picture as a ‘sign of solidarity’, though
the extent to which such online actions result in actual, offline action remains open
to debate. Studies like Giaxoglou’s (2015) show how at the micro-level, small
stories can intersect with large scale political events. At the same time, the scale
with which this stance-taking can be taken up also poses challenges for the more
qualitative methods that are usually used within small stories research. In some
cases, it can be useful to use methods more usually associated with ‘big data’ to
contextualise particular narrative interactions (Page 2015), such as social network
analysis. Given the diversity of social media narration, there is room for narra-
tive research that trains the analytical foci toward the micro and the macro level,
though in order to do so, it may be that scholars in pragmatics will need to broaden
their interdisciplinary collaborations so that the affordances of social media sites
can be harnessed fully in the years to come.

Authenticated
540 Ruth Page
References
Adami, Elisabetta
2014 Retwitting, reposting, repinning; reshaping identities online: Towards a social
semiotic analysis of digital remediation. LEA – Lingue e Letterature d’Oriente
e d’Occidente 3: 233–243.
Androutsopoulos, Jannis and Kasper Juffermans (eds.)
2014 Digital Language Practices in Superdiversity. Special issue of Discourse,
Context and Media 4–5.
Antaki, Charles and Sue Widdicombe
1998 Identity as an achievement and as a tool. In: Charles Antaki and Sue Widdi-
combe (eds.), Identities in Talk, 1–14. London: Sage.
Bamberg, Michael and Alexandra Georgakopoulou
2008 Small stories as a new perspective in narrative and identity. Text and Talk 28:
377–396.
Blommaert, Jan and Ben Rampton
2011 Language and superdiversity: A position article. Working Papers in Urban
Language and Literacies 70. http://www.kcl.ac.uk/ldc (accessed: 13 January
2014).
Bruner, Jerome
1986 Two modes of thought. In: Actual Minds, Possible Worlds, 11–43. Cambridge,
Bruner, Jerome
1991 The narrative construction of reality. Critical Inquiry 18: 1–21.
Bruns, Axel
2008 Blogs, Wikipedia, Second Life, and Beyond. From Production to Produsage.
Berlin: Peter Lang.
Cohen, Leor
2015 World attending in interaction: Multitasking, spatializing, narrativizing with
mobile devices and Tinder. Discourse, Context and Media 9: 46–54.
Danet, Brenda and Susan Herring (eds.)
2007 The Multilingual Internet: Language, Culture and Communication Online.
New York: Oxford University Press.
1990 Positioning: the discursive construction of selves. Journal for the Theory of
Social Behaviour 20: 43–63.
Deppermann, Arnulf
2013 Editorial: Positioning in narrative analysis. Narrative Inquiry 23(1): 1–15.
De Fina, Anna
2003 Identity in Narrative: A Study in Immigrant Discourse. Amsterdam/Philadel-
phia: Benjamins.
Deumert, Ana
2014 Sociolinguistics and Mobile Communication. Edinburgh: Edinburgh Univer-
sity Press.

Authenticated
Narration 541

2010 Once upon a blog …Storytelling in weblogs. In: Christian R. Hoffmann (ed.),
Telling a Story in the Age of New Media, 79–108. Amsterdam/Philadelphia:
Benjamins.
Facebook helpdesk
2016 https://www.facebook.com/help/112146705538576.
2005 ‘Same old story?’ On the interactional dynamics of shared narratives. In:
Ulrike Quasthoff and Tabea Becker (eds.), Narrative Interaction, 223–241.
2007 Small Stories, Interaction and Identities. Amsterdam/Philadelphia: Benja-
mins.
2008 ‘On MSN with buff boys’ self- and other- identity claims in the context of
small stories. Journal of Sociolinguistics 15: 597–626.
2013 Small stories research and social media: The role of narrative stance-taking
in the circulation of a Greek news story. Working Papers in Urban Language
and Literacies, Paper 100. http://sins.au.dk/fileadmin/sins.au.dk/Texts/Georg-
akopoulou__2013_.pdf (accessed 27 November 2016).
2014 Small stories, transposition and social media: A micro-perspective on the
Greek crisis. Discourse and Society 25(4): 519–539.
2015 Sharing as rescripting: Place manipulations on YouTube, between narrative
and social media affordances. Discourse, Context and Media 9: 64–72.
Giaxoglou, Korina
2015 #JesuisCharlie: Narrative stance-taking in shared stories of global mourning.
Paper presented at The 6th International Language in the Media Conference,
Hamburg, 7–9 September 2015.
Gibson, James
1977 The theory of affordances. In: Robert Shaw and John Bransford (eds.), Per-
ceiving, Acting, and Knowing, 67–82. Hillsdale, NJ: Lawrence Erlbaum.
Greenspan, Brian
2011 Songlines in the streets: Story mapping with itinerant hypernarrative. In: Ruth
Page and Thomas Bronwen (eds.), New Narratives: Stories and Storytelling in
the Digital Age, 153–169. Lincoln: University of Nebraska Press.
Herring, Susan
Internet, 4, article 1. http://www.languageatinternet.org/articles/2007/761
(accessed 29 June 2016).
Heyd, Theresa
2014 Doing race and ethnicity in a digital community: Lexical labels and narratives
of belonging in a Nigerian Web Forum. Discourse, Context and Media 4–5:
37–48.

Authenticated
542 Ruth Page
Hutcheon, Michael and Linda Hutcheon

2010 Opera: Always and forever multimodal. In: Ruth Page (ed.), New Perspectives
on Narrative and Multimodality, 65–77. London/New York: Routledge.
Hymes, Dell
Hymes, Dell
1996 Ethnography, Linguistics, Narrative Inequality: Toward an Understanding of
Voice. London: Taylor and Francis.
Hymes, Dell
1998 When is oral narrative poetry? Generative form and its pragmatic conditions.
Jenkins, Henry
2006 Convergence Culture: Where Old and New Media Collide. New York: New
Jones, Rodney H.
2005 Sites of engagement as sites of attention: Time, space and culture in electronic
discourse. In: Sigrid Norris and Rodney H. Jones (eds.), Discourse in Action:
Mediated Discourse Analysis, 141–154. New York: Routledge.
Koteyko, Nelya and Daniel Hunt
2015 What was your blood sugar reading this morning? Representing diabetes
self-management on Facebook. Discourse and Society, 26(4): 445–463.
Krogh Hansen, Per
2010 All singing! All dancing! Prolegomena: On film musicals and narratives. In:
Marina Grishakova and Marie-Laure Ryan (eds.), Intermediality and Storytell-
ing, 147–164. Berlin/Boston: de Gruyter.
Labov, William
1972 Language in the Inner City. Philadelphia: University of Pennsylvania Press.
Labov, William
2013 The Language of Life and Death. Cambridge: Cambridge University Press.
Locher, Miriam
2008 Advice Online: Advice-Giving in an Internet Health Column. Amsterdam/Phil-
Maagaard, Cindie
2014 Employee testimonials: Animating corporate messages through employee sto-
ries. Discourse, Context and Media, 6: 22–32.
Martin, James and Gunther Plum
1997 Construing experience: some story genres. Journal of Narrative and Life His-
tory 7(1–4): 299–308.
2010 I tweet honestly, I tweet passionately. Twitter users, context collapse and the
McCloud, Scott
1993 Understanding Comics. New York: Harper Perennial.
Myers, Greg
2006 Where are you from? Identifying place in talk. Journal of Sociolinguistics 10
(3): 320–343.

Authenticated
Narration 543
O’Reilly Tim
2005 What is Web 2.0. http://oreilly.com/web2/archive/what-is-web-20.html
(accessed April 5, 2011).
Ochs, Lisa and Elinor Capps
2001 Living Narrative. Cambridge: Cambridge University Press.
Page, Ruth
2010 Re-examining narrativity: Small stories in status updates. Text and Talk 30(4):
423–444.
Page, Ruth
2012 Stories and Social Media: Identities and Interaction. London/New York:
Routledge.
Page, Ruth
2014 Counter narratives and controversial crimes in Wikipedia. Language and Lit-
erature 23(1): 61–76.
Page, Ruth
2015 Between the big and the small: Analysing identity and interaction in digital
contexts. In: Alexandra Georgakopoulou and Tereza Spilioti (eds.), The Hand-
book of Digital Communication, 403–407. London: Routledge.
Page, Ruth
forthcoming Narrative Online: Shared Stories and Social Media Controversy. Cam-
Page, Ruth, Richard Harper and Maximiliane Frobenius
2013 From small stories to networked narrative. The evolution of personal narra-
tives in Facebook status updates. Narrative Inquiry 23(1): 192–213.
Rabinowitz, Peter
2004 Music, genre and narrative theory. In: Marie-Laure Ryan (ed.), Narrative
across Media, 305–328. Lincoln/London: University of Nebraska Press.
Ranta, Michael
2013 (Re)creating order: Narrativity and world view in pictures. Storyworlds 5:
1–30.
Ruston, Scott
2010 Storyworlds on the move: Mobile media and their implications for narrative.
Storyworlds 2: 101–120.
Ryan, Marie-Laure (ed.)
2004 Narrative across Media. Lincoln/London: University of Nebraska Press.
Ryan, Marie-Laure
2007 Toward a definition of narrative. In: David Herman (ed.), The Cambridge
Companion to Narrative, 22–35. Cambridge: Cambridge University Press.
Schiffrin, Deborah
2006 In Other Words: Variation in Reference and Narrative. Cambridge: Cambridge
University Press.
Shuman, Amy
1986 Storytelling Rights. Cambridge: Cambridge University Press.
Shuman, Amy
2005 Other People’s Stories: Entitlement Claims and the Critique of Empathy.
Urbana/Chicago: University of Illinois Press.

Authenticated
544 Ruth Page
Shuman, Amy
2015 Story ownership and entitlement. In: Anna De Fina and Alexandra Georga
kopoulou (eds.), The Handbook of Narrative Analysis, 38–56. Oxford:
Wiley-Blackwell.
Suler, John
2004 The online disinhibition effect. CyberPsychology and Behavior 7(3): 321–326.
Turkle, Sherry
1995 Life on the Screen: Identity in the Age of the Internet. New York: Simon Schus-
ter.
Twitter terms and conditions (https://support.twitter.com/articles/106373)
Van Dijck, José
2013 The Culture of Connectivity: A Critical History of Social Media. Oxford:
Van Dijk, Teun
2009 Society and Discourse: How Social Contexts Influence Text and Talk. Cam-
Van Leeuwen, Theo
2008 Discourse and Practice: New Tools for Critical Discourse Analysis. Oxford:
Van Leeuwen, Theo and Gunther Kress
1996 Reading Images. London/New York: Routledge.
Vásquez, Camilla
2014 The Discourse of Online Consumer Reviews. London/New York: Bloomsbury.
Vásquez, Camilla
2015 Right now versus back then: Recency and remoteness as discursive resources
in online reviews. Discourse, Context and Media 9: 5–13.
Wikipedia editors
2016 Wikipedia guidelines for new users. https://en.wikipedia.org/wiki/Wikipe-
dia:Username_policy#Guidance_for_new_users.
Wortham, Stanton and Catherine C. Rhodes
2015 Narratives across speech events. In: Anna de Fina and Alexandra Georga
kopoulou (eds.) The Handbook of Narrative Analysis, 160–177. Oxford:
Wiley-Blackwell.
Zappavigna, Michele
2011 Ambient affilitation: A linguistic perspective on Twitter. New Media and Soci-
ety, 13(5): 788–806.
Zhang, Yi and Camilla Vásquez
2014 Hotels’ responses to online reviews: Managing customer dissatisfaction. Dis-
course, Context and Media 6: 56–64.
Zimmerman, Don H.
1998 Identity, context and interaction. Identities in talk. In Charles Antaki and Sue
Widdicombe (eds.) Identities in Talk, 87–106. London: Sage.

Authenticated
20. Fandom
Monika Bednarek
Abstract: This chapter focuses on the linguistic practices of fans and other media
consumers as evident in their contributions on social media platforms. The digital
environment has had significant impact on such practices, as summarized by Jen-
kins (2006a: 141–142): “It has increased the speed with which fans can communi-
cate with each other and the development of new fan communities. It has allowed
fans to access a much wider range of perspectives online and to practice more
effective activism. The international reach of the internet has made the community
larger and transnational. It has moved fandom towards the mainstream.” Never-
theless, fandom – and audience engagement more generally – has mainly been
considered in non-linguistic disciplines, with only a few exceptions. This chapter
discusses relevant linguistic practices through which audience members respond
to narrative mass media texts on social media platforms. It does so with reference
to fictional TV series and focuses primarily on the linguistic studies that have been
undertaken in this field so far.
1. Introduction
This chapter provides an introduction to the different types of linguistic actions, or

practices, that are undertaken by fans and other audience members on social media
platforms as they respond to narrative mass media texts (in particular, TV series).
Although there is no singular or unified academic definition of the concept of fan
(Costello and Moore 2007: 126), fans are often defined by their behaviour. In such
definitions, a fan engages in practices of consumption that go beyond viewing the
media text and that reflect a heightened emotional engagement – practices that fall
“outside the common expectations for a member of the audience” (Costello and
Moore 2007: 126).
However, in reality, practices of consumption by audience members, including
fans, vary quite considerably. For instance, there are different types and levels of
online fan activity, such as creating a fan site, liking and commenting on such a
site (or on other content), writing fan fiction, producing fan art, or tweeting as
a character (Costello and Moore 2007; Zhivov, Scheepers and Stockdale 2011;
Wood and Baughman 2012). Costello and Moore (2007: 130) therefore posit a
“continuum of passive/active audiences”.
In addition, it is difficult to clearly identify audience practices that go beyond
ordinary expectations or norms, in particular as such norms may also be shifting.
DOI 10.1515/9783110431070-020
Authenticated
546 Monika Bednarek
For example, are audience members who tweet about a particular TV program
engaging in a practice that is beyond the level of engagement now commonly
expected of members of the audience? Harrington (2014: 240–241) reports that
hashtags or keywords that relate to television programs frequently appear as
“trending topics”, indicating that live-tweeting about a TV program might be a
growing phenomenon, most often in relation to live sports and other events, but
also in relation to prime-time comedy/drama.
Building on such findings, in this chapter I assume that audience practices can
be situated on a scale in terms of their adherence to norm-based expectations, and
will thus not attempt to strictly differentiate fans from other audience members.
Rather, I will discuss linguistic research on how audience members engage with
narrative mass media texts (in particular, TV series) on social media platforms,
including but not limited to traditionally understood fandom.
Although the concept of “fan” is hence a fuzzy category with no clear-cut
boundaries, the study of fandom arguably represents an important new area for lin-
guistic research. Fan audiences are a significant aspect of contemporary consumer
culture in “mediatised” (Lundby 2009; Androutsopoulos 2014) societies, where
new digital means have moved fan activities towards the mainstream (Jenkins
2006a: 142). While the identity of the fan may have negative connotations (the ste-
reotype of the geeky or nerdy obsessed loner-fan), fans can also be conceptualised
more positively, for example as active and creative audiences, as producers who
build their own social identity and culture through selectively “poaching” from
media texts (Jenkins 1992). Through fan activities, people may acquire considera-
ble language, writing or technical skills, and work through important issues includ-
ing social justice, equality and diversity (Jenkins, Ito and boyd 2016: 93, 155).
Additionally, fans were early adopters of social media platforms such as Facebook
or Twitter (Jenkins, Ito and boyd 2016: 135) and may thus play an important part
in the diffusion of linguistic innovations, in particular when the form, function
and meaning of the respective innovation encourage such a diffusion from media
consumers to other speakers (Squires 2014). In disciplines outside linguistics, fans
have been an object of systematic and serious academic study for almost 25 years
now, starting with Jenkins’s seminal book in the early 1990s (Jenkins 1992). Hills
(2002: 183) states that there is no longer a need to fight for a place for fandom in
cultural studies research. However, most of this research does not consider linguis-
tic practices in depth, as it does not focus on language use per se.1 It is thus high
time for a linguistic “fanalysis”.
1
For example, Androutsopoulos and Weidenhöffer (2015: 23) note that “[m]ost existing
research on the use of Twitter by broadcast audiences originates in communication
studies and pays little attention to language use”.

Authenticated
Fandom 547
This chapter is not intended to replicate equivalent chapters on fandom, media

engagement, and social media in handbooks of Media and Communications Stud-
ies, and will hence focus squarely on linguistic research, although I will occasion-
ally draw on non-linguistic literature.
Linguistic media engagement studies have examined different ways in which
audiences deal with media language, especially in relation to young people, face-
to-face interaction, and transnational or minority communities (Androutsopoulos
2014: 9, 18–25). Some studies have drawn on the notion of appropriation (Ayaß
and Gerhardt 2012), while there has been recent interest in the concept of mediati-
sation (Androutsopoulos 2014). Such media engagement studies examine a range
of different media products, including narrative mass media texts such as films
and TV series (Coupland 2007: 171–176; contributions to Androutsopoulos 2012)
but also political speeches (Cole and Pellicer 2012) or talk shows (Squires 2014).
Overall, these studies are located in interactional sociolinguistics and linguistic
anthropology rather than in pragmatics, and studies on social media have only
emerged fairly recently (e.g. Squires 2014 on Twitter).
In this chapter, I limit my discussion firstly to research on social media plat-
forms and secondly to research on audience responses to contemporary fictional
TV series. By this I mean scripted series such as House of Cards or Mad Men,
which viewers may consume via a range of different platforms (television, com-
puter, tablet, mobile phone).
This focus is primarily motivated by the specific characteristics of audience
engagement with TV series (summarized from Bednarek 2012): Audiences engage
with the characters and narratives as well as the language of the fictional worlds
represented in TV series, including in conversations with others. Simply put, many
audience members are very interested in the lives and experiences of TV charac-
ters and in the unfolding of televisual narratives. Roman (2005: 130) and Cree-
ber (2004: 4) argue that TV series create audience involvement in unique ways
and more than other media texts. More specifically, audiences spend a lot of time
watching television series in general and often follow a particular TV series over
many seasons – this creates considerable depth of engagement. There is psycho-
logical involvement with television characters and with different ways in which
audience members form interpersonal connections with these media figures,
including parasocial interaction, (wishful) identification, or affinity/liking (Cohen
1999; Giles 2002). Viewers may thus build up a particularly close and intimate
relationship with characters in TV fiction (Wickham 2007: 91, 93).
Viewers of television series may also be very active in their online engage-
ment, and such engagement happens both within and across countries, and within
and across languages. For example, British audiences engage with British and
with American TV series, and both with TV series in their own and in other lan-
guages (for instance, Scandinavian crime series with English subtitles). In a study
of prime-time television series and daytime serials, Costello and Moore (2007)

Authenticated
548 Monika Bednarek
identified hundreds of fan sites that were related to these programs. The official
(i.e., network-managed) Facebook sites of the US television series The Big Bang
Theory, Game of Thrones and Castle currently have over 32, 16 and 4 million
members (‘likes’) respectively (26 Feb. 2016).
Although this chapter focuses on social media, it is worth noting that fandom
also exists outside social media, for instance at face-to-face events such as fan
conventions and has existed long before the digital age (Jenkins, Ford and Green
2013: 30; Jenkins, Ito and boyd 2016: 17). However, the digital media environ-
ment has expanded participation in fandom (Jenkins, Ito and boyd 2016: 17), and
the combination of television and the internet has reshaped fan cultures and prac-
tices (Wood and Baughman 2012: 333). The social media have created “a ‘virtual
loungeroom’: an online space where an audience can commune and centrally share
the television experience” (Harrington 2014: 241). The next section provides a
brief overview of some changes in the audiovisual landscape that are relevant to
the discussion in this chapter. This includes discussion of the new digital environ-
ment more generally and the social media in particular.
2. Background
Since the late 1970s (Sinclair, Jacka and Cunningham 1996: 1) or mid-1980s
(Barker 1997: 3) there has been a move towards a transnational television land-
scape, facilitated by technical developments and changes in media ownership.
There has since been extensive importing and localization of TV programmes,
genres and genre conventions beyond national boundaries (Sinclair, Jacka and
Cunningham 1996: 13). While these processes are multi-directional, they initially
favoured the West and in particular the US (Barker 1997: 5; Berry 2001: 267).2
Another important development lies in technological convergence and transmedia
franchises (Harrington 2014: 237; Jenkins 2006b: 2; Wood and Baughman 2012:
331): Media content is no longer limited to particular devices but can be accessed
using a variety of technologies (technological convergence), which allows view-
ers to engage with TV series independent of their broadcast schedule. In addi-
tion, transmedia franchises “cross both media and genres” (Lemke 2009: 292),
with multiple kinds of texts associated with a fictional world (apps, books, films,
games, music, merchandise, advertisements, theme park attractions, etc.).
2
Recent developments include the export of European TV drama into many countries,
for example Scandinavian series such as The Killling and The Bridge. UK series are
also widely exported, with Downtown Abbey a recent success in the US and elsewhere.
Mexico, Brazil, Hong Kong, Taiwan, Egypt and India export widely in particular geo-
linguistic regions (Berry 2001: 267). Asian popular culture (e.g. Japanese animation)
has also been exported successfully into the Western market (Jenkins 2006b: 160–165).

Authenticated
Fandom 549
Furthermore, new technologies and the rise of social media platforms have
lowered the barrier for engagement with TV series, since communicative activities
such as tweeting and blogging are less prescriptive than traditional forms of writ-
ing (Richardson 2010: 86, 89). They have also extended communication about TV
watching beyond close circles and geographical boundaries (Buschow, Schneider
and Ueberheide 2014: 130). In addition, new technologies have offered audience
members new means of linguistic interaction, starting with early tools such as
bulletin boards, mailing lists, discussion lists, web rings and chat rooms (Jenkins
2006a: 139). More recent developments include platforms such as YouTube, Twit-
ter, Tumblr, Facebook and Fanpop or other fansites which integrate multiple tools
like walls, videos, forums (with discussion threads), polls and so on; there are
websites that host fan transcripts, fan fiction, and fan art, etc. Audience members
can provide customer reviews of TV series on commercial transaction sites or
submit comments on dedicated TV/movie sites. Networks themselves may create
websites with official message boards for their own shows where viewers can
comment. There are also apps where fans can chat about TV shows, vote on polls,
answer trivia questions, watch videos, etc. Figure 1 shows a collage of selected
platforms for the TV series Flight of the Conchords.
Jenkins, Ford and Green (2013: 30) suggest that such platforms are “sites where
multiple existing forms of participatory culture – each with its own historical tra-
jectory, some over a century old – come together”. Hills (2002: 172), on the other
hand, argues that online fandoms alter fan practices and are not simply a version
of offline fandoms. Whatever the case, there is agreement that the digital envi-
ronment has had significant impact, as summarized by Jenkins (2006a: 141–142):
– It has increased the speed with which fans can communicate with each other and the
development of new fan communities.
– It has allowed fans to access a much wider range of perspectives online and more
effective activism.
– The international reach of the internet has made the community larger and transna-
tional.
– It has moved fandom towards the mainstream.
One of the ways in which the digital environment allows audience members (fans
and others) to connect is through enabling searches, allowing them to find par-
ticular topics or communities – whether via web searches or hashtag searches on
Facebook or Twitter. Twitter, for instance, has been called searchable talk, ena-
bling us to bond with others around particular values (Zappavigna 2012: 1, see
also Zappavigna, Ch. 8, this volume). This also makes it easier for linguists to
investigate audience discourses that were not produced under research conditions
(Richardson 2010: 89). To illustrate, Figure 2 shows selected Tweets about the US
TV sitcom The Big Bang Theory (BBT) on one day (13 November 2012), which
was found using the service “tweetarchivist” and the search terms “Sheldon” The
Big Bang Theory.

Authenticated
550 Monika Bednarek
Figure 1. Platforms (collage created by author)

Authenticated
Fandom 551
Figure 2. Tweets from http://www.tweetarchivist.com/
This “searchability” of the digital environment along with the rise of social media
platforms which facilitate sharing among users mean that audience responses to
TV series are no longer private and transient, but public, shareable, searchable, and
traceable – in other words, hyper mobile (Bednarek 2013). As Tagg and Seargeant
put it, we are dealing with “a wider internet discourse characterized by perma-
nence, replicability, searchability and scalability” (Tagg and Seargeant 2014: 179).
3. TV series and social media fandom
In the remainder of this chapter I will review linguistic research on audience

responses to fictional television series, especially those that consider the different
types of linguistic actions that occur in the digital environment and on social media
platforms.3 Although the reviewed studies do not draw on speech act theory, their
focus on various kinds of actions makes them relevant to pragmatics in a broad
sense (with a focus on linguistic action, interaction and function). There is also
emerging research on how particular linguistic actions are tied to stances or iden-
tities, which is of interest in pragmatic research (see Locher and Bolander, Ch. 15,
this volume).
3
All translations in this section are my own.

Authenticated
552 Monika Bednarek
3.1. Types of linguistic actions
When considering relevant research that offers categorisations of audience engage-

ment with TV series, two broad groups can be distinguished: The first group
focuses primarily on a particular social media platform such as Twitter or Tumblr,
and provides a classification of actions found on that platform, while the second
group proposes classification schemes that are more general and not tied to a spe-
cific platform. Three examples of platform-specific research include Gregoriou’s
(2012) critical linguistic analysis of an official message board, Petersen’s (2014)
study of fan talk on the micro-blogging site Tumblr, and Androutsopoulos and Wei-
denhöffer’s (2015) study of audience responses to a TV crime narrative on Twitter:
Gregoriou (2012) analyses the official message board for the US TV series Dex-
ter provided by its network (Showtime), but her study focuses specifically on how
viewers respond to the programme’s ideologies rather than categorizing audience
responses into different types. However, it is worth mentioning that some of the lin-
guistic actions that occur in these message board contributions include references
to the real world and intertextual references, as well as speculation about character
reactions and particular dialogue lines. Gregoriou’s (2012) study is a very targeted
analysis with the aim of answering questions about ideology. Thus, only some of
the message board threads are examined, which means that the linguistic actions
that are mentioned throughout the study may not be all that occur on the board. As
such, this study does not focus on media engagement or fandom more generally.
This is in contrast to Petersen’s (2014) study of Tumblr fan talk by fans of the
British TV series Sherlock (2010–).4 However, this study is located in Media Stud-
ies rather than Linguistics. I nevertheless mention it here, since it draws on con-
versation analysis and includes a relevant categorization of three types of fan talk:
talk through appropriation,5 talk through interpretation and talk through imitation.
The first category (appropriation) relates to cases where fans converse about
the televisual text but also to those where “conversations […] become new textual
objects, memes or popular fan-made narratives” (Petersen 2014: 93). On Tumblr,
examples of talk through appropriation include reaction gifs6 which respond to a
blog post/comment (for example showing two Sherlock characters hugging), or
sets of gifs that contain edited material from different Sherlock movies/TV series,
with rewritten dialogue.
4
Note that Petersen (2014) uses talk in a broad sense, encompassing both images and
verbal text.
5
According to Petersen (2014: 93) this term derives from Jenkins (2006b) who applies
the term cultural appropriation to various practices by fans (fanfiction, fanart, tropes,
memes …).
6
A gif (graphics interchange format) is usually a short video clip or a small animation on
a loop (Petersen 2014: 90).

Authenticated
Fandom 553
The second category (interpretation) concerns Tumblr posts that discuss dif-
ferent reactions and interpretations of the televisual narrative (Petersen 2014: 93),
when fans investigate, analyse, dissect and critique the TV series, for instance
interpreting storylines or character personalities, proposing scenarios or comment-
ing on specific scenes. Conversations between fans may proceed via gifs with
added text that are repeatedly re-used and hence promote “specific lingo and say-
ing” (Petersen 2014: 102), such as Oh my god. Why would you post that?
The third category (imitation) occurs when fans take on the identity of a char-
acter, including role playing (for instance using character names as the name of
Tumblr blogs and posting as those characters) and costume playing (conversing
via pictures or gifs where fans are dressed up as characters). Such talk “is an aspect
of talk as performance” (Petersen 2014: 102, italics in original) where fans appro-
priate some of the characteristics of televisual characters.
Petersen (2014: 96, 93) argues that ‘talk’ (in her definition) can be wordless and
expressed through visuals, that there is overlap between these categories on Tum-
blr, and that posts can include several types of talk at the same time. Although she
mentions that these types of talk can also occur offline, for example at fan meet-
ups, she proposes that Tumblr “enforces a particular manner of communication
that differs from other mediated and non-mediated contexts” (Petersen 2014: 90).
This study presents a useful contribution to research on fandom which is val-
uable for linguistic research because there is an attempt at categorizing types of
talk and some linguistic practices are also considered. It is also helpful in its focus
on a particular social media platform. At the same time, there is little systematic
or comprehensive micro-linguistic analysis, because of the disciplinary origin of
the study.
The most comprehensive linguistic study of media engagement on a particu-
lar social media platform is arguably that by Androutsopoulos and Weidenhöffer
(2015), who analyse tweets published during the broadcast of the German TV
crime narrative Tatort.7 The researchers’ categorization of tweets draws on and
adapts schemes by Klemm (2001) and Wood (2007). Thus, #tatort tweets are cate-
gorised into four general categories including seven types of actions (Handlungs
typen/Handlungsgruppen/Handlungskategorien) depending on their orientation to
the televisual narrative (Table 1).
7
Tatort is a highly popular German televisual crime narrative, which differs from typical
television series in that it includes different instalments with police teams from cities all
over Germany. In each 90-minute episode a police team solves a crime in a particular
city. For example, there are episodes that feature the Munich police team, or episodes
that feature the Berlin team, but there may be long gaps between the broadcast of epi-
sodes from the same city, rather than distinct seasons. Tatort therefore differs from
US crime franchises with spin-off series (e.g. NCIS; NCIS: Los Angeles; NCIS: New
Orleans; CSI; CSI: Miami; CSI: NY).

Authenticated
554 Monika Bednarek
Table 1. Seven types of actions
General category Type of action Relation to diegesis Interaction type

Thematising the Framing the Outside film Monologic/dialogic
reception reception
Commenting and Commenting locally Within film Monologic
evaluating the
Commenting Within film Monologic
televisual narrative
globally
Anticipating Within and outside Monologic/dialogic
film
Connections to Transferring and Within and outside Monologic/dialogic
own experience or comparing film
knowledge
Emphasis on inter- Asking questions Within and outside Dialogic
action with others film
Parasocial Within and outside Dialogic
interaction film
In addition, each of the seven types of action can be performed in different ways,
and Androutsopoulos and Weidenhöffer (2015) also identify specific linguistic
realisations, with multiple examples of tweets. Table 2 gives a flavour of how the
types of action are undertaken by German-speaking users of Twitter, according to
the researchers.
Table 2. Examples for performing different types of actions
Types of Main subcategories Examples Translations

action
Framing the naming the TV narra- #tatort Time ‘tatort time’
reception tive and its beginning
announcing participa- #Tatort leider ohne ‘Tatort unfortunately
tion or lack thereof mich without me’
with optional emo- #So, jetzt #dortmunder ‘Ok, now to watch
tional expression #tatort gucken. Freu #dortmund #tatort.
mich drauf. Looking forward
to it’
with optional infor- jetzt #Tatort :D Schön ‘now #tatort :D Nice
mation about channel mit Papa aufm Sofa sitting on the sofa with
or context of situation sitzen dad’

Authenticated
Fandom 555

action
Framing the thematising the los geht’s ihr Twerds :3 ‘here we go you
reception Twitter medium #tatort twerds’
asking for silence/ pscht #tatort ‘shhh #tatort’

attention
good-byes Sie hören zusammen ‘They are listening

Oper. In diesem Sinne to the opera together.
wars das für heute. So that’s it for today.
Danke fürs mitlesen. Thank you for reading.
Gute Nacht. #Tatort # Good night’
Münster
Comment- precise time reference 21.32 Chinesisches ‘9.32 pm Chinese New

ing locally Neujahrsfest. #tatort Year’
evaluation via inter- Awww. <3 #Tatort ‘awww’

jection/emoticon
evaluation via nomi- Guter Klingelton !!!! ‘Good ringtone’

nal/verbal group #tatort
direct quote … Sie haben mich ‘You have googled me.

gegoogelt … Reicht That should be enough’
doch #Tatort
Comment- list and evaluate a toller Anfang, span- ‘fantastic beginning,

ing globally range of aspects or nende Handlungen, gripping plot, very
plot elements sehr schöne Bilder beautiful images’
#tatort
comment on particu- Die erste Hälfte war ‘The first half was
lar sections of the irgendwie besser … somehow better’;
televisual narrative #tatort; bis jetzt ‘until now rather under-
leicht unterwältigend. whelming’
#dortmund #tatort
comment on a charac- Diesen Parkamann ‘I don’t like this Parka-

ter/actor #Tatort Schauspieler dude actor’
mag ich nicht
provide a concluding Fazit: 2–3 mal ge- ‘Conclusion: smiled

evaluative comment schmunzelt und die 2–3 times and rather
restliche Zeit eher bored for the rest of the
gelangweilt #tatort time’

Authenticated
556 Monika Bednarek

action
Anti Vermutungen Hoffentlich heute mal ‘hopefully today with
cipating (‘guesses’) mit mehr Story und we- more plot and less sil-
niger Klamauk #Tatort liness’;
eindeutig: der psy- ‘definitely: the psy-
cho-kommissar ist der cho-detective is the
mörder! #tatort murderer’
Vorwegnahmen #Tatort: Die chinesis- ‘The Chinese Princess
(‘revealers’; knowl- che Prinzessin – Kritik – critique’;
edge-based predic- [link] #spoiler; ‘BTW, she’s also going
tions) Die fährt übrigens to drive like that. But
auch so Auto! Aber erst only later.’
später. #tatort
Transfer- comparison with Wann sah mein Schlaf ‘When was the last
ring and user’s experience anzug mal so akkurat time my PJs looked so
comparing aus? #tatort slick?’
comparison with other Hui, der #Tatort ‘wow, the #Tatort is
media texts beginnt fast wie ein starting almost like a
James Bond Film. James Bond film’
Asking includes rhetorical Wer findet den Fehler ‘Who’ll find the error
questions and tag questions im heutigen #Tatort?; in today’s #Tatort?’;
Und wie fandet ihr ihn? ‘So how did you like
#Tatort; it?’;
Jemanden mit einem ‘To beat someone to
Monitor erschla- death with a monitor –
gen – das gilt in der does that also count as
Polizei-Statistik auch IT-crime in the crime
als IT-Kriminalität? stats?’;
#tatort; ‘Do you always have
Habt Ihr bei Euch cans of #ravioli some-
im Büro auch immer where in your office,
#Ravioli Dosen rumste- too?’
hen? #Tatort
Parasocial Addressing charac- Warum nimmst du nicht ‘Why don’t you just
interaction ters/actors, producers, einfach das Feuerzeug, take the lighter, you
screenwriters, etc du Depp! #tatort; idiot!’;
Gute Arbeit, Herr Wer- ‘Good job, Mr Werner
ner. #tatort #drehbuch #tatort #screenplay’
Ventriloquising the Lasst uns einen #Tatort ‘Let’s start a #Tatort
voices of characters/ mit Sex and Mord with sex and murder’
actors, producers, beginnen.
screenwriters, etc.

Authenticated
Fandom 557
The first category (“Framing the reception”) consists of monologic or dialogic

tweets that frame the reception, typically posted at the beginning or end of the
broadcast. This includes naming the TV narrative, announcing (lack of) partici-
pation, thematising the medium, asking for silence/attention and good-byes. The
second group (“Commenting locally”) covers all tweets that comment on particu-
lar scenes or actions of the televisual narrative. Such tweets may contain precise
time references and include minimal reactions consisting of evaluations using
interjections, emoticons or nominal/verbal groups. These tweets often reference
TV dialogue in various ways, including via direct and indirect ways of speech
representation. The third category (“Commenting globally”) provides comments
on larger chunks of the narrative rather than on particular scenes/actions, including
comments on the episode as a whole.
The fourth category (“Anticipating”) is distinguished from the previous two
categories by its temporal frame. It includes all tweets that anticipate future aspects
of the broadcast televisual narrative or future Tatort instalments, often with the use
of modal expressions (e.g. the adverb vielleicht/‘perhaps’ or the modal verb sol-
len/‘be supposed to’). Two sub-categories are distinguished by Androutsopoulos
and Weidenhöffer (2015): Vermutungen are ‘guesses’ about the televisual narra-
tive, whereas Vorwegnahmen (‘revealers’) are knowledge-based predictions which
may thus be marked with a spoiler alert.
The category “transferring and comparing” encompasses all those tweets in
which users compare and connect the televisual narrative to their own knowledge
and experience or to other media texts. These are distinguished as two sub-groups.
Also included here are #servicetweets which provide further information to other
users such as informing them about the brand of a pyjama worn by a character.
The sixth category is called “asking questions” and is fairly self-explanatory,
as it groups together those tweets that pose questions about the TV narrative or
other aspects. The category also includes rhetorical and tag questions. Question-
ing tweets may ask others to find errors or to provide an overall assessment, but
may also function to comment locally or to compare and transfer. This category
is thus primarily based on form, rather than on the type of media engagement or
the illocutionary force of utterances (e.g. request), as seen in the various examples
included in Table 2.
Androutsopoulos and Weidenhöffer (2015)’s final category is that of paraso-
cial interaction and concerns those tweets that either address the characters/actors,
producers, screenwriters, etc. or, more rarely, those tweets that ventriloquise
their voices. Parasocial tweets typically use the second person singular/plural or
imperatives, and may include named addressees. Requests (at the beginning of the
broadcast) and thanks (at the end) also occur (e.g. Danke für großartige Unter
haltung #tatort [‘Thank you for the great entertainment’]).
Comparing these categories of linguistic action with face-to-face communi-
cation, Androutsopoulos and Weidenhöffer argue that live tweeting in reaction to

Authenticated
558 Monika Bednarek
media events is a new communicative practice which borrows conventions from

face-to-face audience engagement, with both similarities and differences (2015:
27–29): There are Twitter-specific equivalents to particular types of face-to-face
audience engagement (e.g. evaluations) which make use of integral Twitter ele-
ments such as the hashtag (#). At the same time, Tweets are based on digitally-
mediated written language, dislocated, public/anonymous, less interactive, and
constrained by the character limit. They may also be circulated and disseminated
by other media platforms.
In sum, Androutsopoulos and Weidenhöffer’s (2015) study focuses on tweets
that occur during broadcast of a particular televisual narrative (and five minutes
before/after) and that are marked with a particular hashtag. It does not incorporate
analysis of non-verbal elements, such as images. The analysis seems to capture
tweets by fans, by other viewers, and even by non-viewers – for example, one
tweet that is provided simply states Während ihr #Tatort kuckt schaue ich das
#NCIS Staffelfinale [‘While you are watching #Tatort, I’m watching the #NCIS
season finale’]. While the focus on live-tweeted posts may be appropriate in the
context of this particular German media product (Tatort), with other TV narra-
tives such an approach might be less successful. As mentioned earlier, many audi-
ence members now engage with TV series without close temporal connection to
a live televisual broadcast, through services such as streaming or download. For
example, several German-speaking users of Twitter had clearly seen episodes
from the 2016 The X-files season before its German broadcast, as evident from
tweets that were posted before the day of broadcast (8 Feb. 2016) or that state as
much:
(1) Gerade die 3. Ep. der neuen Akte X Staffel gesehen. Eine fantastische Folge #AkteX
[‘Just watched the 3rd ep. of the new The X-files season. A fantastic episode #TheX-
files’], posted 5/2/2016
(2) Hab mich noch nicht entschieden, ob ich neue #AkteX Staffel cool oder schrecklich
finden soll. Wie war es bei euch? [‘Haven’t yet decided whether I should find new
#TheX-files season cool or terrible. How was it for you?’], posted 5/2/2016
(3) Der Zahn der Zeit hat auch an Scully und Mulder genagt. Alles andere wäre ja über-
natürlich. Hehe. Neue Staffel #Akte X [‘Scully and Mulder have also been subject to the
ravages of time. Everything else would of course be supernatural. Ha ha. New season
#TheX-files’], posted 5/2/2016
(4) @ProSieben ich konnte einfach nicht mehr warten. Ich hab 3 Episoden #AkteX auf
Englisch angeschaut [smiley face emoticon] [‘@Channelname I simply couldn’t wait
any more. I’ve watched 3 episodes of #TheX-files in English’]
Further, Androutsopoulos and Weidenhöffer’s (2015) data collection is hashtag-

based, which may not be suitable for analysis of posts by fans that tweet as tele-
visual characters, as such users do not necessarily mark their tweets with hashtags
(Kalviknes Bore and Hickman 2013: 2). Finally, Androutsopoulos and Weiden-

Authenticated
Fandom 559
höffer use a combination of functional, content-based and formal criteria to distin-

guish their categories, which sometimes makes a dual or alternative classification
possible, as they acknowledge themselves (2015: 50). Their most problematic cat-
egory (in my view) is that of “asking questions”, which is primarily form-based.
Rather than seeing it as a distinct category, this could instead be recognized as a
particular linguistic practice through which several categories of media engage-
ment can be realized.
As noted above, not all linguists take a platform-specific approach to the anal-
ysis of linguistic actions in audience responses to fictional TV series. Thus, Rich-
ardson (2010) and Bednarek (2017) make more general claims about audience
responses that are not tied to a specific social media platform.
Richardson (2010: 89–90) lists different “genres” for audience responses to TV
drama, namely fan forum thread, blog entry, review column, customer review, fan
fiction and Twitter stream, and mentions a range of different responses through-
out her chapter: criticism, character sketches, general and specific comments on
a televisual episode (including on the acting performance, character personality
or plot), appropriating particular memorable lines or catchphrases, and creating
fan fiction. She notes that online audience responses may either focus on meaning
and performance or focus on form (dialogue lines), but argues that even “when
audiences respond to meaning (in their interpretations of particular characters
for instance), dialogue – in other words, form – helps create those interpreta-
tions, as it is intended to do” (Richardson 2010: 85). When dialogue is singled
out by audience members, it “can be cited in evidence of a particular interpreta-
tion, […] criticized as unsatisfactory in a range of different ways, quoted as tribute
to the pleasure [it] has provided, and remembered, as catchphrases” (Richardson
2010: 104). While Richardson (2010) is not concerned with a typology of audi-
ence activities but rather focuses primarily on writing personae (see Section 3.2
below), this chapter has served as a starting point for my own classification
scheme:
In Bednarek (2017), I distinguish four ways of responding to TV series – four
key processes of (re-)circulation, which are not limited to online contexts and
which include multimodal practices. The label (re-)circulation “is not meant to
imply mere repetition, re-play or re-broadcasting of mass media content and mass
media language; rather, the label serves as a reminder that these practices can ulti-
mately be traced back (in different ways) to a fictional TV series” (2017: 122). The
four categories are: performative (re-)circulation, (re-)circulation of the narrative,
of the production or of the dialogue. These processes are summarised in Table 3
but will be illustrated with examples from different social media platforms below.
Since this is my own research I will provide new examples for each category here
rather than re-using examples from Bednarek (2017). Though I provide examples
from specific social media platforms, the classification scheme is meant to apply
more generally to media engagement with TV series.

Authenticated
560 Monika Bednarek
Table 3. Four key processes of (re-)circulation
Type of process Definition

performative (re-)circulation impersonation of television characters
(re-)circulation of the narrative references to the narrative world of the television
series, its characters and plots without any citing of
dialogue
(re-)circulation of the production references to the production or broadcast rather than
the narrative world
(re-)circulation of the dialogue partial or complete quotations of dialogue lines
Performative (re-)circulation concerns the impersonation of television characters.8

The category label was chosen in analogy to Georgakopoulou’s (2014: 228) use of
“performative enactment”. Accounts that are created for or named after television
characters are a good example from Twitter. For instance, the “Peggy Olson” Twit-
ter account for a Mad Men character has over 24,000 followers (Figure 3).
Figure 3. @Peggy Olson Twitter account, 17 December 2014
8
Compare Petersen’s (2014) category of imitation for Tumblr fan talk. Impersonation
also existed in the pre-digital area and still exists in face-to-face environments, whether
at fan conventions or in performative enactments in everyday classroom peer-talk,
where pupils may imitate TV characters (Georgakopoulou 2014: 221).

Authenticated
Fandom 561
What do tweets by such TV characters look like? What kind of linguistic actions
can be found? In their discussion of Mad Men characters on Twitter, Jenkins, Ford
and Green (2013: 31) note that during broadcast of the series these characters’
posts were often playful, self-referential, fit within the existing storyline and deep-
ened engagement with plots. It would be interesting to see how these linguistic
practices have developed over time, now that the series has come to an end (but
can still be consumed).
Wood and Baughman’s (2012) analysis of Glee character accounts identi-
fies two key practices: first, augmentation of existing storylines through first- or
third-person tweets about specific events or invention of new narrative events, and
second “identity control” through posts that reinforce the respective character’s
personality as it is constructed in the series (Wood and Baughman 2012: 338).
In their analysis of the practice of tweeting as a character from The West Wing,
Kalviknes Bore and Hickman (2013) point to “its emphasis on improvisation, its
combination of creative role playing and political participation, and the emphasis
on dialogue between participants and between participants and external individu-
als.” Note that the emphasis on political engagement here reproduces the thematic
concerns of the TV series (Kalviknes Bore and Hickman 2013), and the findings
may not be generalizable to other in-character tweeting. This means that findings
that are based on only one media text (i.e. one TV series) may not be fully gener-
alizable to other media texts, and that the nature of a TV series may have an effect
on the kinds of activities that occur on social media.
The next category, “(re-)circulation of the narrative”, encompasses any refer-
ences to the narrative world of the television series, its characters and plots which
does not include any partial or complete citing of dialogue. An example is pro-
vided below:
(5) Ich finde das tattoo schick [smiley face emoticon]

[‘I find the tattoo stylish’]
Above an image of Malcolm in the Middle character Dewey with naked upper body
with the tattooed words “LOIS IS MY MOTHER”)
(posted on a German Malcolm in the Middle Facebook fan site; image not reproduced
here for copyright reasons)
In this example, the user provides an evaluative and affective comment on the
posted image/scene through an evaluative adjective and emoticon, ‘citing’ the
image (and represented tattoo), but not citing any dialogue lines. Other social
media posts in this category may feature evaluative comments on characters, the
episode or the plot, or interpret and speculate on those aspects of the narrative.
The next category, “(re-)circulation of the production”, refers to references
to the production or broadcast rather than the narrative world itself. Such refer-
ences may be about the performance, actors, directors, channel, network, broad-
cast schedule, etc. and include comments about post-production aspects such as

Authenticated
562 Monika Bednarek
subtitling and dubbing. For example, when The X-files was revived in 2016, the
German version (Akte X) broadcast on the TV channel Pro Sieben featured a new
dubbing actor. This was much commented on via Twitter, with a selection of tweets
presented below:
(6) @ProSieben Ich freu mich zwar auf die neuen Folgen #AkteX aber mit der neuen
Syncrhonstimme [sic] für #Mulder wird es nie wieder das selbe sein. [‘I’m look-
ing forward to the new episodes of #TheX-files but with the new dubbing voice for
#Mulder it will never be the same again.’]
(7) Tragisch, wie die eigentlich coolen #AkteX-Teaser von @ProSieben wegen der voll
kommen unbekannten Mulder-Stimme einfach null funktionieren. [‘Tragic how the
actually cool #TheX-files teasers by @ProSieben don’t work at all because of the
completely unknown Mulder voice’]
(8) Wer heute Abend #AkteX schaut oder schon hat und ihm die #akteXsynchro nicht
gefiel, dann bitte hier unterschreiben [LINK] [‘Anyone watching #TheX-files tonight
or who has already and didn’t like #TheX-filesdubbing, then please sign here’]
(9) Da freust du dich erst das es neue #AkteX Folgen gibt, bis du erfahren hast das @Pro
Sieben das mit der Synchro mal so richtig versaut hat [thumb down emoticon] [‘There
you are at first happy about the new #TheX-files episodes until you find out that
@ProSieben completely fucked up the dubbing’]
(10) Wie kann man so eine unpassende Syncronstimme [sic] nur nehmen? @ProSieben
#Akte X #Mulder [‘How can anyone use such an unsuitable dubing voice?’]
Finally, as linguist I am most interested in the different ways in which dialogue

is re-circulated through partial or complete quotations. From Richardson (2010:
104), we know that this includes dialogue that is “quoted as tribute to the pleasure
[it] has provided”. I call this type of linguistic action tribute quoting. Tribute quot-
ing occurs in tweets, on Facebook, in fan forum discussions and on YouTube, with
several examples provided in Bednarek (2017). I will just offer one new example
here: A typical instance of tribute quoting on YouTube can be seen in the posting
of video compilations of sampled “best of” quotes. Figure 4 is a screenshot from
a YouTube video that is a collection of dialogue from The Wire – here the user’s
commentary is also verbalized through the title “100 greatest quotes”.
Another key way in which TV dialogue is cited is through appropriation and
play with memorable lines or catchphrases. In her discussion of a fan website and a
blog, Richardson (2010) notes that catchphrases in the form of templates enhance
audience creativity, but that more fixed catchphrases are also “played around with”
(Richardson 2010: 102). With respect to social media, Zappavigna discusses viral
memes involving phrasal templates with slots that can be modified, “allowing the
meme to ‘mutate’ as users add their own elements to the slots” (Zappavigna 2012:
106). Such memes include templates based on TV dialogue lines like the I for one
welcome our [ ] overlords from The Simpsons (Zappavigna 2012: 116) or those
based on the phrasal template Brace yourselves, [ ] is coming from the fantasy TV
drama Game of Thrones which often co-occur with visuals (Zappavigna, personal

Authenticated
Fandom 563
Figure 4.
The Wire quotes (http://www.youtube.com/watch?v=-Sgj78QG9Bg,
23 August 2013)
communication).9 Dialogue lines can also be appropriated as hashtags on Twitter

or Facebook (e.g. #D’oh from The Simpsons, #bazinga from The Big Bang Theory,
#thetruthisoutthere from The X-files).
While Richardson (2010: 101) proposes that the fragmentation of audiences
means that “[t]he catchphrases of the future are less likely to achieve the distri-
bution that those of the past were able to do”, social media may in fact help to
distribute them more widely than in the past. The searchable digital environment
will allow researchers to approach such questions confidently, drawing on large-
scale empirical analysis.
While tribute quoting and the appropriation of catchphrases and memorable
dialogue lines are two key ways of citing TV dialogue on social media, they are
not the only ones. Thus, Androutsopoulos and Weidenhöffer (2015: 48) mention
that quotes can be recontextualised in tweets that are posted at/after the end of the
broadcast so that they function communicatively as “essence” or epigram/motto
of the televisual narrative. Dialogue can also be cited to support an interpretation
or be criticized for quality/appropriateness (Richardson 2010: 104). In addition,
Leyda (in press) mentions “supercuts” (e.g. montages of each instance of a char-
acter uttering Yo or Bitch in Breaking Bad) and “vids” where clips are reorgan-
ized into new narratives. Such remixing of dialogue also exists for other series on
YouTube, e.g. supercuts of Sheldon’s bazingas in The Big Bang Theory or Clay
Davis’s shiiiiits in The Wire. Generally, the recontextualisation of TV dialogue
offers fertile ground for future linguistic research into different social media plat-
forms.
9
In both examples, [ ] stands for the slot that is filled with a variable element, e.g. Brace
yourselves, school is coming.

Authenticated
564 Monika Bednarek
It must be emphasized that the different categories may co-occur in one social
media post, as in the following tweet, which allows for a three-fold categorization:
No nominations for Big Bang’s Jim Parsons? What a farce. Is this one of Sheldon’s
(11) a.
classic jokes? Bazinga? #Emmys #BigBangTheory
b. No nominations for Big Bang’s Jim Parsons? What a farce. [(re-)circulation of the
production: comment on actor]
Is this one of Sheldon’s classic jokes? [(re-)circulation of the narrative: comment
on character]
Bazinga? [(re-)circulation of dialogue: appropriation of dialogue]
In this tweet, a comment on the actor Jim Parsons’s performance is combined with
a comment on the character he plays (Sheldon), and the user further appropriates
Sheldon’s catchphrase bazinga. (12) is an example from a different social media
platform, a post from a German Game of Thrones Facebook fan site:
(12) “Jaime and I are more than brother and sister. We are one person in two bodies”, sagte
Cersei einst.
Hat sich das jetzt geändert? Bereits im Finale von Staffel 3 zeigte sich, dass sie zurűck-
haltend auf ihren veränderten Bruder (mit nur einer Hand) reagierte. Und die erste Fol-
ger der 4. Staffel scheint das zu bestätigen. Was denkt ihr, wie es zwischen den beiden
weitergehen wird? Hier geht’s zu einem kleinen Artikel mit Mini-Spoiler: [link]
[“…” Cersei once said.
‘Has this now changed? Already in the finale of season 3 it became apparent that she
reacted with reserve towards her changed brother (with only one hand). And the first
episode of the fourth season seems to confirm this. What do you think, how are things
going to develop between the two? Here’s a brief article with mini-spoiler.’]
Here, the recirculation of dialogue occurs in the context of the user’s plot and
character speculation, i.e. is embedded in a broader recirculation of the narrative,
with direct appeal to other audience members to speculate with the writer and
establishing a community.
In relation to social media practices, it must be noted that Bednarek (2017)
only includes a brief application of this new classification scheme to social media
and excludes a range of fan practices such as fan fiction and fan art. A wider quan-
titative basis is missing and the scheme is perhaps too general as it encompasses
both offline and online (re-)circulation processes without attention to the particu-
larities of specific media platforms. However, as noted by the platform-specific
studies reviewed above, different social media platforms may indeed encourage
particular types of communication and particular realisations (linguistic and mul-
timodal). Even if all four key processes of (re-)circulation can be identified on all
platforms as well as in offline environments, there might be differences in how
linguistic actions are performed and the frequency with which they occur. More
empirical research is thus needed to identify both commonalities and differences
in the types of linguistic actions that occur in different contexts and media.

Authenticated
Fandom 565
3.2. Stance/identity
As mentioned earlier, there is also an emerging interest in how particular linguistic

actions are tied to stances or identities. Richardson (2010: 89–90) distinguishes
five different types of audiences: “wannabe writers, journalistic reviewers, schol-
arly critics, fans, and casual viewers”, who are seen as “ideal types” with over-
lapping interests (Richardson 2010: 88). Her particular interest is in the writing
persona or authorial voice, where she distinguishes between “The Professional
Reviewer,” “The Fan,” and “The ‘Ordinary’ Viewer” (Richardson 2010: 94–97).
For her:
It is in its online voices that the writing audience identifies itself— that members of the
viewing public constitute themselves as ordinary viewers, fans, or critics. Stances are
variously offered by writers to their readers. At one end of the spectrum viewers provide
unarguable expressions of personal taste at the level of “I never liked ER [Emergency
Room] as much after George Clooney left”, whereas at the other end there are lengthy
and highly articulate accounts of strengths and weaknesses. (Richardson 2010: 92, ital-
ics in original)
Richardson (2010) illustrates this by exploring audience activity in relation to
one episode of the British BBC series Doctor Who. For example, she argues that
reviews in the voice of the professional reviewer typically focus on performance
over dialogue and that fans construct themselves as “serious” viewers through
their detailed critical engagement and familiarity with characters and plots. Her
analysis of the “ordinary” viewer focuses on comments on a journalistic review
provided on a Guardian blog, with these viewers distancing themselves from other
viewers and providing general and specific comments as well as expressions of
disagreement and debate.
As Richardson herself notes, her approach is selective (Richardson 2010: 93).
The brief discussion of authorial voices thus remains rather vague on where the
line would be drawn between the different personae/identities (e.g. fan vs. ordi-
nary viewer), and Richardson does not analyse contemporary social media plat-
forms such as Facebook, Tumblr or Twitter. Further, the extent of variation within
the three personae is noted by Richardson – for instance, “the different flavors
of fandom” (Richardson 2010: 95), but not explored further, and transnational
engagement is not considered.
However, Bednarek (2017) emphasises that fandom of TV series is a transna-
tional phenomenon (cf. Section 2 above), using various examples of engagement
with US TV series by European fans. To provide some new transnational “fan-
dom” data here, Table 4 shows selected “unofficial” German/French Facebook
community fansites for US TV series which had more than 2000 “likes” in the
period between 29 January and 18 February 2016.
To be clear, not all English-language TV series have a German/French Face-
book fan site, and those that do vary in terms of their number of “likes”, active sta-

Authenticated
566 Monika Bednarek
Table 4. German/French Facebook fan sites for US TV series
Name of Number of “likes” on German (G) or French (F)

US TV series community Facebook site (between 29/1/2016–18/2/2016)
Breaking Bad 2069 (G)
Malcolm in the Middle 2488 (G)
Castle 2849 (G)
Glee 3252 (F)
Two Broke Girls 3842 (G)
Supernatural 4918 (G)
Teen Wolf 5949 (F)
The Vampire Diaries 5963 (G)
Grey’s Anatomy 8246 (F)
Two and a half men 9151 (G)
tus, and (inter)activity. Fans may also communicate on other Facebook sites (for
instance on “official” sites created by TV channels/networks) or on other social
media platforms. Thus, the official German Facebook site for Game of Thrones
(Warner Home Video) has over 54,000 members (“likes”), while the unofficial
German Facebook fan site A Game of Thrones (deutsche Fanseite) has less than
3000 members. This exemplifies the importance of corporate activity in fandom.10
Further, not all users who “like” a Facebook site are likely to post to the site (Zhi-
vov, Scheepers and Stockdale 2011: 8), since the “most common role in most,
if not all, online communities is that of ‘lurker,’ the person who reads but never
posts” (Baym 2015: 97). It is possible that smaller fan-managed Facebook sites
show a higher level of interaction than bigger official sites (Zhivov, Scheepers and
Stockdale 2011: 9). In any case, Table 4 indicates that there is a significant amount
of social media data that can be explored for linguistic research into identity con-
struction, transnational fandom and associated language practices.
10
Several researchers have noted that there is a high level of corporate intrusion and com-
modification, since corporations try to encourage fan activities and also (re-)circulate
televisual content themselves, including on social media (e.g. Jenkins 2006a: 147; Zhi-
vov, Scheepers and Stockdale 2011; Jenkins, Ito and boyd 2016: 14, Bednarek 2017).
Fan cultures, according to Hills (2002: 182) both intensify and challenge commodifica-
tion. Wood and Baughman (2012) point out that fan activities themselves may be seen
as uncompensated labour that benefits the entertainment industry, and ask: “Do these
practices benefit the fan, or the producer, or both?” (Wood and Baughman 2012: 330)

Authenticated
Fandom 567
Bednarek (2017: 135) suggests that both social media platforms and trans
national contexts allow for different “styles of fandom”. For example, “expert”
status – being a serious viewer – may be constructed through curation of selected
media content (such as “best of” quote compilations on YouTube) or through the
quoting of TV dialogue in the original, rather than dubbed version. Viewers may
also explicitly refer to their consumption practices, showing that they are ahead of
the game, as it were, and that they engage with (English) original versions rather
than dubbed TV broadcasts:
(13) @ProSieben #akte X habt ihr euch selbst versaut Pro7. Und den XPhiles schön in den
Hintern getreten damit. Ich schaue nur Original! [‘@TVChannel You have managed
to ruin #TheX-files all by yourself. And with that have given the XPhiles (X-files
fans) a good kick in the butt. I only watch the original version!’]
4. Conclusion
In conclusion, this chapter has shown that there is much room for linguistic
research into media engagement on social media platforms. Given the wealth of
research from other disciplines but the lack of relevant linguistic research, it is
high time for a linguistics of fandom. From a pragmatic perspective, a first task
might be to consider the different types of linguistic practices that can be identi-
fied. While the above schemes all have their limitations (in particular in relation to
the fuzziness/overlap between categories and the criteria adopted to differentiate
them) and further refinements are possible, a number of points have become clear:
Androutsopoulos and Weidenhöffer’s (2015) and Bednarek’s (2017) research
suggests that there are similarities between offline and online audience engage-
ment, and between engagement on different social media platforms. This also
becomes apparent when we compare findings from Petersen (2014) on Tumblr
and Gregoriou (2012) on message boards with findings from research on other
platforms where some of the same activities occur.
At the same time, there are also differences between offline and online audi-
ence engagement (Androutsopoulos and Weidenhöffer 2015), and between lin-
guistic practices on different social media platforms. Petersen (2014) makes a
persuasive argument for taking into account the specifics of communication on
a particular social media platform. Other non-linguistic research seems to con-
firm that there are similarities as well as differences across platforms (Wood and
Baughman 2012: 339, 341).
Androutsopoulos and Weidenhöffer’s (2015) study demonstrates that a com-
prehensive qualitative analysis can provide useful insights into typical linguistic
realisations of audience engagement practices, while Petersen (2014) highlights
the importance of visuals and multimodal combinations (e.g. gifs and verbal text).
Finally, both Richardson (2010) and Bednarek (2017) illustrate how linguistic

Authenticated
568 Monika Bednarek
practices can be tied to stances/identities/styles, while the latter also points to the
need to consider the transnational media landscape.
With respect to the pragmatics of social media, the social media platforms
that are mentioned in these studies include a wide variety, from fan forum dis-
cussion, blog entry, fan fiction, to Facebook and YouTube, but only Tumblr and
Twitter have been analysed systematically in terms of types of linguistic actions.
For example, the classification scheme that I propose in Bednarek (2017) distin-
guishes four ways of responding to TV series regardless of social media platform,
but such a claim needs empirical testing. In addition, the reviewed studies can
only be classified as pragmatic in a broad sense, sharing an interest in linguistic
action, interaction and functions, including the construction of identity. However,
analytical/theoretical frameworks are more often taken from sociolinguistics. Not
only is there scope for a linguistics of social media fandom, there is also the need
for a pragmatics of social media fandom.
It is also important to emphasise that the focus on TV series is not meant to
imply that other cultural products do not result in fandom or media engagement.
For example, there is fandom around anime and manga, video games, contempo-
rary and classic literature, science fiction, music, sports, etc. I hypothesise that
findings from the studies I reviewed in this chapter (on media engagement with
TV series) may be applicable to audience engagement with similar types of media
products, for instance other narrative texts (e.g. films, novels), but not necessarily
to other kinds of fandom (e.g. sports).
I want to conclude this chapter with some recommendations for future linguis-
tic research into media engagement and fandom on social media platforms. Such
research would arguably be well advised to:
– adopt clear criteria for categorizing media engagement activities in general,
and for sub-categorising what people do with dialogue in particular;
– consider the variety of such audience practices and tie them to different stances/
identities/styles, practices of community building/affiliation, constructions of
group identity (Zappavigna 2012; Baym 2015: 98–99), and “norm-based” par-
ticipation (Jenkins, Ito and boyd 2016: 66);
– consider media engagement activities with respect to a wide range of different
media texts;
– provide a uniquely linguistic perspective on the data, identifying the different
forms in which categories of talk can be and typically are realised;
– take into account non-linguistic elements and adopt a multimodal approach to
the data where relevant;
– explore both similarities and differences between audience practices in various
digital and non-digital contexts, for example on different social media plat-
forms;
– look beyond national boundaries and monolingual contexts.

Authenticated
Fandom 569
Last but not least, research into audience engagement on social media platforms
must continue to wrestle with ethical concerns (Kalviknes Bore and Hickman
2013; Page et al. 2014: 58–79) and explore more deeply the role of corporate
intrusion and commodification (Bednarek 2017). As Jenkins, Ito and boyd put
it, “any understanding of participatory culture today has to factor in the wave of
commercialization that has impacted contemporary digital culture” (2016: 13).
Acknowledgments
The research leading to these results has received funding from the People Pro-
gramme (Marie Curie Actions) of the European Union’s Seventh Framework Pro-
gramme (FP7/2007–2013) under REA grant agreement no [609305]. I am grateful
to the Freiburg Institute for Advanced Studies (FRIAS), University of Freiburg,
Germany, for awarding me an FCFP External Senior Fellowship. Thanks are also
due to Roland Muntschick for undertaking the survey on German/French Face-
book fan sites. My gratitude also goes to the editors for their constructive com-
ments on a previous version of this chapter.
References
2014 Mediatization and sociolinguistic change. Key concepts, research traditions,
open issues. In: Jannis Androutsopoulos (ed.), Mediatization and Sociolin-
guistic Change, 3–48. Berlin/Boston: de Gruyter.
Androutsopoulos, Jannis and Jessica Weidenhöffer
2015 Zuschauer-Engagement auf Twitter: Handlungskategorien der rezeptions
begleitenden Kommunikation am Beispiel von #tatort. Zeitschrift für ange-
wandte Linguistik 62(1): 23–59.
Ayaß, Ruth and Cornelia Gerhardt (eds.)
2012 The Appropriation of Media in Everyday Life. Amsterdam/Philadelphia: Ben-
jamins.
Barker, Chris
1997 Global Television: An Introduction. Oxford: Blackwell.
Baym, Nancy K.
2015 Personal Connections in the Digital Age, 2nd ed. Cambridge/Malden: Polity.
Bednarek, Monika
2012 Construing ‘nerdiness’: characterisation in The Big Bang Theory. Multilingua
31: 199–229.
Bednarek, Monika
2013 Mobile narratives. Paper presented at the Poetics and Linguistics Associa-
tion Annual Conference (PALA 2013), University of Heidelberg, Germany,
31 July – 4 August 2013.

Authenticated
570 Monika Bednarek
Bednarek, Monika
2017 (Re-)circulating popular television: Audience engagement and corporate prac-
tices. In: Janus Mortensen, Nikolas Coupland and Jacob Thøgersen (eds.),
Style, Mediation and Change. Sociolinguistic Perspectives on Talking Media,
115–140. Oxford: Oxford University Press.
Berry, Sarah
2001 Marketing and promotion. In: Roberta E. Pearson and Philip Simpson (eds.),
Critical Dictionary of Film and Television Theory, 264–267. London/New
York: Routledge.
Buschow, Christopher, Beate Schneider and Simon Ueberheide
2014 Tweeting television: Exploring communication activities on Twitter while
watching TV. Communications 39(2): 129–149.
Cohen, Jonathan
1999 Favorite characters of teenage viewers of Israeli serials. Journal of Broadcast-
ing and Electronic Media 43(3): 327–345.
Cole, Debbie and Régine Pellicer
2012 Uptake (un)limited: The mediatization of register shifting in US public dis-
course. Language in Society 41: 449–470.
Costello, Victor and Barbara Moore
2007 Cultural outlaws: An examination of audience activity and online television
fandom. Television and New Media 8(2): 124–143.
Coupland, Nik
2007 Style. Language Variation and Identity. Cambridge: Cambrige University
Press.
Creeber, Glen
2004 Serial Television: Big Drama on the Small Screen. London: BfI Publishing.
2014 ‘Girlpower or girl (in) trouble?’ Identities and discourses in the (new) media
engagements of adolescents’ school-based interaction. In: Jannis Androut-
sopoulos (ed.), Mediatization and Sociolinguistic Change, 217–244. Berlin/
Boston: de Gruyter.
Giles, David C.
2002 Parasocial interaction: A review of the literature and a model for future
research. Media Psychology 4(3): 279–304.
Gregoriou, Christiana
2012 ‘Times like these, I wish there was a real Dexter’: Unpacking serial murder
ideologies and metaphors from TV’s Dexter internet forum. Language and
Literature 21: 274–285.
Harrington, Stephen
2014 Tweeting about the telly: live TV, audiences, and social media. In: Katrin
Weller, Axel Bruns, Jean Burgess, Merja Mahrt and Cornelius Puschmann
(eds.), Twitter and Society, 237–247. New York: Peter Lang.
Hills, Matt
2002 Fan Cultures. New York/London: Routledge.
Jenkins, Henry
1992 Textual Poachers: Television Fans and Participatory Culture. New York/Lon-
don: Routledge.

Authenticated
Fandom 571
Jenkins, Henry
2006a Fans, Bloggers, and Gamers: Exploring Participatory Culture. New York:
New York University Press.
Jenkins, Henry
2006b Convergence Culture: Where Old and New Media Collide. New York: New
Jenkins, Henry, Sam Ford and Joshua Green
2013 Spreadable Media: Creating Value and Meaning in a Networked Culture. New
York: New York University Press.
Jenkins, Henry, Mizuko Ito and danah boyd
2016 Participatory Culture in a Networked Era. Cambridge/Malden: Polity.
Kalviknes Bore, Inger-Lise and Jonathan Hickman
2013 Studying fan activities on Twitter: Reflections on methodological issues
emerging from a case study on The West Wing fandom. First Monday l: http://
firstmonday.org/ojs/index.php/fm/article/view/4268 (accessed 26 February
2016).
Klemm, Michael
2001 Sprechhandlungsmuster. In: Werner Holly, Ulrich Püschel and Jörg R. Berg-
mann (eds.), Der sprechende Zuschauer: Wie wir uns Fernsehen kommunika-
tiv aneignen, 83–114. Wiesbaden: Westdeutscher Verlag.
Lemke, Jay
2009 Multimodal genres and transmedia traversals: Social semiotics and the politi-
cal economy of the sign. Semiotica 173-1/4: 283–297.
Leyda, Julia
in press Piracy and authorship in Breaking Bad fan remix videos. In: Ralph Poole and
Saskia Fürst (eds.), Auteur TV. Heidelberg: Winter Verlag.
Lundby, Knut (ed.)
2009 Mediatization: Concept, Changes, Consequences. New York: Peter Lang.
Page, Ruth, David Barton, Johann W. Unger and Michele Zappavigna
2014 Researching Language and Social Media: A Student Guide. Oxford/New York:
Routledge.
Petersen, Line Nybro
2014 Sherlock fan talk: Mediatized talk on tumblr. Northern Lights 12(1): 87–104.
Richardson, Kay
2010 Television Dramatic Dialogue: A Sociolinguistic Study. Oxford: Oxford Uni-
versity Press.
Roman, James
2005 From Daytime to Primetime: The History of American Television Programs.
Westport, Connecticut/London: Greenwood Press.
Sinclair, John, Elizabeth Jacka and Stuart Cunningham
1996 Peripheral vision. In: John Sinclair, Elizabeth Jacka and Stuart Cunningham
(eds.), New Patterns in Global Television: Peripheral Vision, 1–32. Oxford:
Squires, Lauren
2014 From TV personality to fans and beyond: Indexical bleaching and the diffusion
of a media innovation. Journal of Linguistic Anthropology 24(1): 42–62.

Authenticated
572 Monika Bednarek
Tagg, Caroline and Philip Seargeant

2014 Audience design and language choice in the construction and maintenance of
translocal communities on social network sites. In: Philip Seargeant and Car-
oline Tagg (eds.), The Language of Social Media: Identity and Community on
the Internet, 161–185. Basingstoke/New York: Palgrave Macmillan.
Wickham, Phil
2007 Understanding Television Texts. London: British Film Institiute.
Wood, Helen
2007 The mediated conversational floor: An interactive approach to audience recep-
tion analysis. Media, Culture and Society 29(1): 75–103.
Wood, Megan M. and Linda Baughman
2012 Glee fandom and Twitter: Something new, or more of the same old thing?
Communication Studies 63(3): 328–344.
Zappavigna, Michele
iation on the Web. London/New York: Continuum (now Bloomsbury).
Zhivov, Jacob, Helana Scheepers and Rosemary Stockdale
2011 Facebook – The final frontier for TV fandom: A lurker’s perspective. ACIS
2011 Proceedings, Paper 57. http://aisel.aisnet.org/acis2011/57 (accessed 26
February 2016).

Authenticated
21. Getting “liked”
Carmen Maíz-Arévalo
Abstract: This chapter explores the complex notion of “getting liked” in comput-
er-mediated communication. More specifically, it delves into the most widespread
techniques, strategies and multimodal features which social media users apply to
build up their positive face as well as defend and maintain their self-designed
online identities. It also aims to identify research which recognises how users rat-
ify, challenge, discuss or reject such online presentations and how other users
respond to these reactions. To do so, this chapter focuses on the interplay of four
main macro-strategies: self-presentation (also known as self-disclosure), social
expressivity, implicitness and humour.
1. Introduction
There seems to be a multidisciplinary consensus that human beings’ need to be

liked is universal (Baumeister and Leary 1995; Leary and Baumeister 2000; Elle-
mers, Doosje and Spears 2004; Srivastava and Beer 2005, among many others).
This necessity appears to stem from the evolutionary need to be part of a social
group in order to survive (cf. Weir 2012) but also from the actual physical reac-
tions our brain experiences when being liked. Thus, being liked has proven to acti-
vate primary (and pleasant) reward in parts of the brain such as the nucleus accum-
bens, the ventral tegmentum or the amygdala (Goleman 1995/2006; Punset 2007;
Davey et al. 2010), which explains the pleasure we all get when, for example, our
joke gets wholeheartedly laughed at by our fellow interlocutors. Getting liked has
also proven to play a crucial role in individuals’ psychological development and
welfare. Thus, being liked by others leads to a higher self-esteem (Srivastava and
Beer 2005) and positively influences individual self-development and self-clarity
(Yang and Brown 2016), as well as boosting the levels of self-satisfaction and
social trust (Valenzuela, Park and Kee 2009). Unfortunately, it is beyond the scope
of this chapter to zero in on such complex neurological and psychological aspects.
Suffice it to say that getting liked constitutes a core aspect of our own human
nature, and is hence worth studying.
This chapter aims to deal with the complex notion of “getting liked” from
a discourse pragmatic perspective by specifically focusing upon computer-me-
diated communication (CMC henceforth) and its effects of what “getting liked”
might entail. In fact, in the last few years, the vertiginous growth of Web 2.0
has multiplied users’ chances of getting liked by an incalculable number of other
DOI 10.1515/9783110431070-021
Authenticated
576 Carmen Maíz-Arévalo
users, which can ever render astonishing economic gains. For example, YouTubers
like the Swedish PewDiePie make billions of dollars by ‘simply’ getting other
YouTube users to like whatever they do online (e.g. playing video games, offer-
ing beauty advice, talking about religion, etc.).1 More specifically, this chapter
intends to seek out the most widespread techniques, strategies and multimodal
features which social media users apply recursively to build up their positive face
as well as defend and maintain their self-designed online identities. It also intends
to identify research which recognises how users ratify, challenge, discuss or reject
such online presentations and how other users respond to these reactions.
Multidisciplinary research has revealed different strategies individuals seem
to employ when trying to get liked by others. In my view, these ‘techniques’ can
be summarized into four main macro-strategies: self-presentation (aka self-disclo-
sure), social expressivity, implicitness and humour. For the sake of methodological
clarity, the rest of this chapter will focus on each of them individually even though
they are closely intertwined and often operate simultaneously.
2. Self-presentation (or self-disclosure)
Self-presentation can be defined as “the process through which individuals com-

municate an image of themselves to others” (Yang and Brown 2016: 404). Indi-
viduals’ choice of how to present themselves can be very limited, e.g. just their
name or even a nickname, or very extensive, e.g. by adding an explicit descrip-
tion of the self and a profile photograph (Bronstein 2014: 2). Much studied in
face-to-face communication (Baumeister 1982; Baumeister and Tice 1984; Kow-
alski and Leary 1990 among many others), self-presentation is a core element to
being socially accepted by a group. Self-presentation is often employed in lieu of
self-disclosure (see below) since the first is considered “a specific and more stra-
tegic form of self-disclosure” (Yang and Brown 2015: 406). In this chapter, they
will be used interchangeably.
Online self-presentation has been analysed in different Internet spaces like
social network sites (Livingstone 2005; boyd 2007; Zhao, Grasmuck and Martin
2008), blogs (Reed 2005; Qian and Scott 2007; Hodkinson and Lincoln 2008;
Bronstein 2012; García-Gómez 2010), chat rooms and Bulletin Boards (Rhein-
gold 1995; Turkle 1995; Surratt 1998), dating sites (Yurchisin, Watchravesringkan
and McCabe 2005; Ellison, Heino and Gibbs 2006; Toma, Hancock and Ellison
2008), personal homepages (Papacharissi 2002; Schau and Gilly 2003), Twitter
(Marwick and boyd 2011; Papacharissi 2012; Sifianou 2015), LinkedIn (van Dijck
2013), etc. Previous research has revealed that online self-presentation relies on
1
http://www.businessinsider.com/richest-youtube-stars-2014-3

Authenticated
Getting “liked” 577
a multimodal array of elements, e.g. personal photos, explicit self-descriptions,

adopting micro-celebrity tactics like sharing quotes, revealing positive features
of the self, etc. (Manago et al. 2008; Zhao, Grasmuck and Martin 2008). In their
study of Facebook, Zhao, Grasmuck and Martin (2008) put forward the notion of
a cline from implicit self-presentation (by visual means like photographs) to more
explicit, textual self-presentation, where users can produce “identity statements”
(Walker 2000: 99). In other words, users may choose to visually ‘show’ them-
selves implicitly as social actors or ‘tell’ about themselves explicitly in a textual
self-description.
Previous research has also revealed that anonymity seems to play a key role
in self-presentation routines. Many studies have investigated self-presentation in
settings like chats or forums, which allow users for a high degree of anonymity
(Rheingold 1995; Turkle 1995; Papacharissi 2002; Suler 2002; Rosenmann and
Safir 2006; Qian and Scott 2007, among others). These authors point out that, in
such anonymous contexts, users may feel freer to express themselves and even
behave in ways not always acceptable or permissible in their real offline lives
(Papacharissi 2002). Other users may exploit the anonymity afforded by these
settings to reinvent themselves (McKenna and Bargh 2000; Bargh, McKenna and
Fitzsimons 2002; Bronstein 2014). As summarised by Zhao, Grasmuck and Martin
(2008: 1819):
In a fully nonymous offline world where deviance from established social norms will
be punished or ridiculed, the masks people wear in everyday life become their “real” or
known identities (Goffman 1959) and a person’s “true” self often gets suppressed and
becomes hidden (Bargh, McKenna and Fitzsimons 2002). In contrast, in a fully anon-
ymous online world where accountability is lacking, the masks people wear offline are
often thrown away and their “true” selves come out of hiding, along with the tabooed
and other suppressed identities.
It may be argued, thus, that anonymous and nonymous are two extremes in the
same continuum. Anonymous settings can be defined as those where it may be
hard for users to trace back to their interlocutors offline.2 Nonymous settings, on
the other hand, reflect self-presentations easily traceable to offline contexts. For
example, in the following Twitter self-presentations, these users opt for presenting
themselves by using a nickname and adding an impersonal photo, which contrib-
utes to anonymizing their identity for other users:
2
It is beyond the scope of this chapter to enter the debate whether the online-offline
divide still holds. See Eklund (2015) for a detailed discussion on this issue.

Authenticated
(1) Alberto Caperas3

Just for fun
(2) Flannaghan McPhee4:
Marica oficial de Twitter y #TuitDiva por aclamación popular. Agente Secreto de Molar.
Y muy cuqui.
[Official Twitter fagot and #TuitDiva by popular acclaim. Secret agent of coolness. And
very cutie.]
In (2) the user employs a Spanish pun to create their fake Twitter name while
accompanying it with a clear statement of their intentions to use Twitter (“just for
fun”). In (3), the user also employs a fake name and presents himself by appeal-
ing to his sexual tendencies – which could also be fake – and his coolness. In any
case, s/he seems to have stricken the right cord among Twitter users, since s/he
has succeeded to attract the attention of more than 12,000 followers. As a Twitter
user, however, I have opted to present myself by using my professional identity
and adding a real photo:
(3) Lecturer in English and Linguistics at Complutense University of Madrid. Love my job,
reading, learning and languages, among many other things!
It is important to point out, however, that the difference between anonymity and
nonymity is more dependent on how users choose to present themselves than on
the type of ‘channel’ (e.g. Facebook, Twitter, WhatsApp, forums, etc.) even if
these ‘channels’ seem to boost anonymity or nonymity to different degrees. For
example, Facebook seems to encourage online relationships with users we also
know offline. In forums, on the other hand, relationships tend to be purely online
and frequently anonymous. Furthermore, the users’ reasons for using these social
media are also important. In other words, for users like those of examples (1)
and (2), Twitter is probably employed as a means to entertainment rather than to
keep in touch with users they also know offline (see also Zappavigna, Ch. 8, this
volume). In contrast, where relationships online may be also ‘anchored’ offline,
users tend to depict themselves according to acceptable social standards, as in
example 3. As Zhao, Grasmuck and Martin have argued (2008: 1815): “Depending
on the degrees of nonymity in the given situation, the level of conformity varies
accordingly”.
3
This user’s profile photograph depicts two pears, in clear allusion to the pun in Spanish,
which translates as “Albert Touches Pears” (colloquialism for the female breast). His
identity claim appears in English in the original. I personally know the user and know
that he is using this account anonymously.
4
This user’s profile photograph is a bearded yellow minion, the popular characters from
the children film Despicable me. As argued by Zhao, Grasmuck and Martin (2008:
1829), “public proclamations of non-mainstream or gay sexual orientations seemed to
be rare on Facebook” as opposed to more anonymous settings, like in this example.

Authenticated
On the whole, users can disclose personal information in two (often inter-
twined) ways: by “telling” and by “showing” (Zhao, Grasmuck and Martin 2008).
Some social media can also afford an interface where users disclose personal
information (e.g. in the case of Facebook via the lists of interests and hobbies, the
“about me” section, etc.). Self-disclosure, however, is far from static. Quite on the
contrary, it is a dynamic process which develops along five major interconnected
dimensions (cf. Kim and Dindia 2011; Yang and Brown 2015): breadth, depth,
positivity, authenticity and intentionality. The first four focus on the content of
the information disclosed whilst the last relates to the individual’s consciousness
while disclosing it. The following subsections will focus upon breadth, depth, pos-
itivity and authenticity while leaving off intentionality since it is firmly believed
that online self-disclosure is mostly and increasingly intentional, even if platform
owners search “to engineer and steer performance” into “unintentional expres-
sions of the self” (van Dijck 2013: 202) for commercial data-mining purposes.
2.1. Breadth
Breadth relates to the amount of personal information presented. Disclosing a great
deal of personal information can be positively valued by other members and recip-
rocally met by their own disclosure in return. However, disclosing too much can
also have counter effects both online and in face-to-face communication (Altman
and Taylor 1973). There seems to be a correlation between age and breadth. For
example, Bronstein (2014) found out that young Facebook users disclose personal
information more frequently than their older counterparts. They also post more
photographs –especially with friends, in an attempt to appear as socially desirable,
popular members of their group. Similar results were also reported in a previous
study by Nosko, Wood and Molema (2010).
However, it is important to avoid the stereotypical assumption that young-
sters disclose personal information less controllably than adults. In a contrastive
study between adolescents and adults (288 and 285 participants respectively),
Christofides, Muise and Desmarais (2009) revealed that adolescents displayed
more information because they spent more time on Facebook, which led to a
higher amount of disclosure, but both age groups were more similar than different
when protecting their privacy.
Facebook is probably one of the social settings which offers a comprehensive
built-in interface where users may choose how much personal information to dis-
close, i.e. hobbies, lists of interests, quotes, etc. However, most of the information
users choose to disclose is implicitly done by means of photographs, where users
are shown as socially desirable actors. These photographs have received a great
deal of scholarly attention, especially profile photos (Gibbs, Ellison and Heino
2006; Bond 2009; Mesch and Beker 2010; Leary and Allen 2011; Hum et al. 2011,
inter alia). As argued by Hancock and Toma (2009: 368), profile photographs are

Authenticated
“a central component of online self-presentation, and one that is critical for rela-
tional success” since they “identify themselves [i.e. users] within the entire net-
work” (Hum et al. 2011: 1828). Furthermore, most users may also be aware that
careful selection of their profile photograph may not only have a positive effect in
relational terms but also in professional ones since (potential) employers increas-
ingly seem to check their candidates’ online identity too. Back in 2010, Microsoft
reported that 70 % of employers abstained from hiring job candidates based on
information they had found online and nowadays this tendency is on the rise,5 with
a reported 92 % of recruiters using social media to get information about job appli-
cants, mostly LinkedIn but also Facebook and Twitter. Furthermore, inappropriate
visual self-presentation may have negative professional consequences (Beals 2010).
This certainly explains why most users seem to select carefully their presenta-
tion photo, as commented by the following Facebook users (Maíz-Arévalo 2013:
62):
Well, I usually upload those photos where I look pretty, haha. I’m not going to upload
one where I look ugly, am I? I’ve got my friends to do that, hahaha.
I upload photos where I like the way I look, for me it is like being on display and I want
people to like what they see.
As argued by Toma and Hancock (2011), individuals tend to idealize their self-pres-
entation – albeit to different degrees – at the cost of accuracy. In this respect,
Vázquez (2012) humorously points out that “No one is as ugly as they look in their
passport photograph or as attractive as they look in their profile photograph”.6
Thus, Toma and Hancock (2011) highlight the importance, especially in nonymous
mediums like Facebook, to strike a balance between the desire to enhance the self
and the need to present an accurate view of the self since doing the opposite, i.e.
posting an unfaithful photograph, may easily attract the other users’ mockery (jocu
lar or not), as illustrated by the following example from Maíz-Arévalo (2015: 308):
(4) (Context: User 1 has just changed her profile photograph).
U2 (f): Mate get rid of that profile pic, no justice mate – fat face syndrome in it !! Sorry
had to be honest, love ya xx
U1 (f): lol I wana die laughing and I’m in a fucking publc [sic] place!!! Thanks for
that [Addressing U2]!!! I’ll do it babe don’t worry! for all of you who were accidently
included in this, sorry about that she’s actually my best friend!! and as you can see
utterly honest!!
U2 (f): Lol, sorry everyone – it’s just how we are, all said with love I promise x
5
http://business.time.com/2012/08/08/does-not-having-a-facebook-page-make-you-
suspicious-to-em
6
The original Spanish is “Nadie es tan feo como en su DNI ni tan guapo como en su foto
de perfil”. Available at http://elpais.com/diario/2012/02/05/eps/1328426821_850215.
html. El País is one of the most prestigious newspapers in Spain.

Authenticated
In a revealing study on Facebook profile photographs, Hum et al. (2011) put for-
ward a taxonomy of profile photographs into six categories, i.e. sex, level of phys-
ical activity, candidness,7 appropriateness and the number of subjects depicted in
the photograph. Their results, after analysing a sample of 140 college students,
showed that most profile photographs followed the same pattern, presenting the
user as inactive, posed, appropriate and alone, regardless of gender. Their results
contradict former studies which reported gender differences such as the face-ism
index, i.e. the relative prominence of one’s face in an image, where male users
displayed higher scores than their female counterparts or the female preference for
posting profile photos with their partners, friends, or family (Bond 2009).
More recent studies have also shown cultural differences in the type of pho-
tographs Facebook users post. For example, in their contrastive study between
American and Asian Facebook users, Huang and Park (2013) found out that East
Asian users are more likely to deemphasize their faces compared to Americans,
favouring context inclusiveness in their profile photographs whilst American
users seem to prioritize their face at the expense of the background. Smith and
Cooley (2012) also carried out a contrastive analysis of American and Russian
Facebook users’ profile photographs, revealing significant distinctions in how
Russian women presented themselves in contrast to US women, with a larger gap
in face-ism presentation between Russian men and women than that of US men
and women. In both cases, differences in self-presentation were allegedly related
to major cultural differences.
Curiously enough, this desire to portrait a socially functional and popular per-
sona online produces contradictory effects. Hence, whilst it boosts self-esteem
and self-concept (Srivastava and Beer 2005; Christofides, Muise and Desmarais
2009; Gonzales and Hancock 2011; Bronstein 2014), it also seems to give rise
to depression as some viewers perceive others to have a happier and fuller life
than themselves. This concern about the negative effects of social networking sites
has led to more recent research among social psychologists in order to determine
whether the use of online networking is actually positive or not for individuals.
For example, Chou and Edge (2012) interviewed 425 Facebook users. They found
out the latter thought other people lived more happily than they did, which tended
to cause different degrees of personal dissatisfaction. Similar results have been
obtained by more recent psychological studies like Tandoc, Ferrucci and Duffy
(2015), Sagioglou and Greitemeyer (2014), Krasnova et al. (2013) or Kross et al.
(2013), whose empirical study alarmingly concluded: “On the surface, Facebook
provides an invaluable resource for fulfilling the basic human need for social con-
nection. Rather than enhancing well-being, however, these findings suggest that
7
Candidness measured the level of spontaneity displayed in the photo. They established
a cline whose extremes were candid (i.e. spontaneous) and posed.

Authenticated
Facebook may undermine it (http://journals.plos.org/plosone/article?id=10.1371/

journal.pone.0069841#s1).
2.2. Depth
This concept refers to the level of intimacy of the presented information, e.g. rela-
tionship status, sexual tendencies, etc. As rightly argued by Attrill (2012), different
settings will lead to different types of self-disclosure. In the case of Facebook, for
example, the interface provides users with the possibility to offer semi-automatic
self-disclosure by choosing from the available options: work and education, places
you’ve lived in, contact and basic information, family and relationships, details
about you and life events. Curiously enough, most options are kept open for users
to fill in with their personal information whilst the relationship tag provides them
with a cline including eleven ‘possible’ relationship statuses, from single to wid-
owed passing by the implicit ‘it’s complicated’. Other social media like Twitter do
not offer this possibility, leaving it for users to choose what level of self-disclo-
sure they wish to perform (see examples above).8 In other settings, like online
shopping, individuals may reveal intimate information such as their address out of
necessity whist other personal details like deep-seated feelings would be seen as
highly irrelevant (Attrill 2012).
As mentioned above, nonymity seems to play a crucial role in the kind of
personal information displayed by users. Thus, in a recent study, Bronstein (2014)
surveyed self-disclosure by more than one hundred Hebrew social networking site
users. Not surprisingly, most users admitted the information they disclosed was
rarely highly personal (almost 70 % of the sample) and that they avoided disclos-
ing embarrassing information (e.g. being drunk) since they were concerned the
self they wanted to portrait might get damaged if they did so. Interestingly enough,
63.2 % of the participants reported that when they did disclose highly personal
information, e.g. deep-seated feelings, it was consciously done.
Likewise, too much disclosure might be as counterproductive online as it
is in face-to-face encounters (e.g. Collins and Miller 1994) since “too much or
uneven disclosure can also produce disliking” (Weisbuch, Ivcevic and Ambady
2009: 574), as in example (5), where one of the Facebook users takes the common
joke too far and discloses his intimate sexual preferences, which produces explicit
rejection by another participant (“TMI, TMI”) and the end of the ‘conversation’
altogether:
8
As is well known, the disclosure of very intimate information can trigger extremely
negative consequences for the user – e.g. in sexting.

Authenticated
(5) [Context: several Facebook users are commenting on their team having lost the World
Cup already. They need to find another team to support and start humorous comments
on reasons for supporting one team or another. Just some extracts have been reproduced
given the length of the exchange]
User 1: I was living in Brussels for a week. Go Belgium!
User 2: That’s nothing, my hoover is German and I love it. Go Germany!!!
[…]
User 7: I’m a fervent supporter of Brazil, blame it on my grooming preferences, hehe
User 2: TMI [Name of User 7], TMI. [End of exchange]
In their seminal study, Zhao, Grasmuck and Martin (2008) identified taboo con-
tents which were rarely disclosed by users. These included sexual tendencies,
especially non-mainstream, religious preferences and, remarkably enough for their
university participants, academic interests. This marked absence was argued not to
form part of the participants’ ‘hoped-for selves’ they wanted to project.
In more anonymous settings, users can take advantage of anonymity to verbal-
ise what they might not do either in face-to-face or nonymous settings (Bronstein
2014). Talking about their hidden selves (Suler 2004), non-conventional tastes
(Rosenmann and Safir 2006), worries and personal concerns, difficult personal
situations, domestic violence, health problems, etc. seem to be pervasive in anon-
ymous settings like forums or blogs, where users may vent away their own life
situations in search for advice and support (Suzuki and Calzo 2004; for forums
see Arendholz, Ch. 5, this volume; for blogs Heyd, Ch. 6, this volume, and for
advice giving in social media Morrow, Ch. 24, this volume). Thus, taboo and very
intimate topics like illnesses, sexual relationships, personal relationships, etc. are
happily and profusely disclosed when done anonymously (Bonebrake 2002; Baker
2005).9
Besides (a)nonymity, age has also proven to play a core role in the amount
and quality of personal information users may display, with younger users dis-
closing highly personal information more often, even if they eventually regret
doing so in nonymous arenas like Facebook (Bronstein 2014). In his studies on
teenage blogs, García-Gómez (2009, 2010) provides examples of hyper-sexualised
accounts by teenagers, who have no problem to offer fully-fledged descriptions of
their love and sexual encounters. Similar results, i.e. displaying highly intimate
information, were also reported by Valkenburg and Peter (2009) in IM anonymous
encounters. The dangers of deep self-disclosure have been much warned against,
especially with regard to teenagers and younger users who might be placing them-
selves at great risk by revealing intimate personal information (cf. Nosko, Wood
and Molema 2010; García-Gómez 2010; Valkenburg and Peter 2009). Although it
9
Attrill (2012), however, points out that research on online self-disclosure (versus its
offline counterpart) has offered contradictory results.

Authenticated
is beyond the scope of this chapter to deal with the complex issue of sexting, its
dangers cannot be left unmentioned.
2.3. Positivity
Positivity involves users presenting themselves in a positive light by enhancing
their positive qualities, e.g. physical, social, psychological, etc. This enhancement
usually depends upon the users’ interactional intentions, such as e-dating, get-
ting Twitter followers, etc. In other words, positive comments about oneself try to
target the imagined audience (Marwick 2011) and to attract “contacts, contracts,
customers, or employers” (van Dijck 2013: 203). For example, this anonymous
Spanish Twitter user enhances her Twitter abilities by including her winning award
at the end of her presentation (in bold):
(6) Hola me llamo Puri y vengo a entregarte mi flor. Autora de TE DEJO es JÓDETE al
revés y La familia:alojamiento con tensión completa. Bitácoras ’15 Mejor tuitera
[Hello my name is Puri and I am giving you my flower. Author of I’M LEAVING YOU
is FUCK OFF the other way round and the family: lodgings of high pressure. Bitácoras
’15 Best twitterer]
Enhancing our positive qualities is not limited by anonymity. Nonymous Twitter
users also make reference to positive qualities in their profile presentations, where
more professional characteristics are balanced by positive personal features, as
follows (in italics, my emphasis):
(7) Teacher, teacher trainer, writer, raconteur and general bon viveur.
(8) English trainer 30+ yrs, coach 4 IELTS 7/8 and English 4 work. Gapfillers elearning,
total immersion courses UK, committed 2 helping people B the best they can B!
Prior studies on e-dating services, where users aim to obtain an eventual offline
relationship (e.g. Yurchisin, Watchravesringkan and McCabe 2005; Ellison, Heino
and Gibbs 2006; Gibbs, Ellison and Heino 2006) showed that users tended to
“stretch the truth a bit” (Yurchisin, Watchravesringkan and McCabe 2005: 742) by
presenting an improved version of themselves, e.g. by avoiding unwelcoming pho-
tographs and writing flattering self-descriptions. Curiously enough, there is a wide
array of links where users of e-dating services are advised how to write good e-dat-
ing profiles.10 Self-selling also applies to other social media like LinkedIn, where
professionals aim to get employed, as illustrated by example (9)11 (my emphasis):
10
Some examples are: www.edatingdoc.com/online-dating-profile-examples/, www.
sampleonlinedatingprofiles.com/sample-men-profiles, www.pickuplinesgalore.com/
online-dating/profiles.html, etc.
11
The users’ names have been removed to preserve their privacy.

Authenticated
(9) Telecommunication Engineer with broad experience in IT, Programming, Tester, Net-
works Call Center, Project Management and experience as customer service. Hard-
working, resourceful, leadership-gifted willing to join in a company where I can apply
my skills and willingness to learn new skills in order to build a successful career.
However, presenting a positive image needs striking a careful balance. In Yang
and Brown’s words (2016: 4): “to make themselves appear as an attractive social
partner, people need to strike a balance between desirability (positivity) and accu-
racy”. In the same line, West and Trester (2013: 137) argue that “doing positive
facework in posts is a bit of a balancing act; it can easily tip into a threat to a
member’s positive face – painting them as a braggart – if done incorrectly”. As
a result, presenting too positive a personal description can have the undesirable
effect of getting disliked, as in the following fake example,12 where a highly posi-
tive self-disclosure on a Facebook post provokes a disapproving reaction by other
users:
Figure 1. Facebook joke (quoted in West and Trester 2013: 142)
These jokes stress how positive self-disclosure needs mitigation and careful face-
work so as to avoid making a negative impression on other users. This can clearly
been seen in the following extract from Maíz-Arévalo (2015: 304) where a user
posts the news of his best employee prize-winning but mitigates the whole news
by saying he does not really deserve such a prize and thankfully compliments all
his colleagues (see also Placencia and Lower, Ch. 23, this volume):
(10) A real honour, I don’t think I deserve it, because all my colleagues are fantastic, but
it’s made me really happy. Thanks, thanks everyone!
12
West and Trester (2013) include Petri’s Facebook jokes available at fakebook.com as
evidence of what users recognize as undesirable behaviours in the social network site.

Authenticated
In sum, displaying our positive qualities is affected not as much by (a)nonymity as

by the imagined audience we are trying to impress, be it friends, potential partners
or employers.
2.4. Authenticity
In self-disclosure, authenticity relates to the degree to which the disclosed infor-
mation accurately reflects reality (Yang and Brown 2015). As already mentioned,
users may take advantage of anonymous settings to reflect their ‘socially unaccept-
able’ authentic self. However, in nonymous settings, and even though users claim
to be ‘authentic’ and post real information about themselves, it has already been
proved that most information is carefully crafted to enhance the positive over the
negative so as to present the user as ‘likeable’. Rather than lying, users are simply
economical with the truth. As Zhao, Grasmuck and Martin (2008: 1829) argue, “it
is equally instructive to consider what aspects of identity [are] not being stressed”
(emphasis in the original). In other words, presenting a pessimistic, apprehensive,
sad or negative persona is extremely rare in nonymous settings as opposed to dis-
closing happy or fun moments.
However, some ‘negative’ comments may occasionally make the user appear
as a socially desirable, feeling and committed person, as in the case of indignation
or anger at social injustice or when publicly expressing their condolences, as in the
following example by celebrity David Ghetta:
(11) Alan, I just can’t believe you’ve gone, my friend. Thank you for all you did for me.
I lost my production manager and the house community lost a true music lover. I’m
hurting and sending love to Lainie, family and friends.
To conclude, this section has shown that self-presentation or self-disclosure online

relies on a wide myriad of multimodal elements, e.g. profile photographs, sta-
tus, self-description, etc. Likewise, (a)nonymity plays a crucial role in how users
choose to present themselves. However, the difference between opting for an
anonymous or nonymous self-presentation is not as dependent on the social media
as on the users’ reasons for using them. Thus, even in the case of seemingly more
anonymous social media (e.g. Twitter or YouTube13), users may opt for employ-
ing their offline ‘real’ identity14 (including their real name) and hence follow a
13
In their policy, neither Twitter nor YouTube ask users to employ their real identity, as
opposed to Facebook.
14
Companies and celebrities may be included in this category, since they are boosting
their offline identity via their Tweets.

Authenticated
similar line of behaviour in their profile self-presentation photographs as in more

nonymous social media like Facebook. As for future research, more cross-cultural
studies are clearly needed, as well as ones which focus on age differences and how
different age groups choose to present themselves, since up to now most studies
have focused on teenage or university students while neglecting other age groups,
e.g. retired people.
3. Social expressivity
Social expressivity is defined as the “display of sociable interactivity” (Weisbuch,

Ivcevic and Ambady 2009: 575). Thus, whilst self-presentation and self-disclosure
are related to how online users construct a likeable persona, social expressivity is
more related to their interrelation and interaction with other users by paying atten-
tion to others rather than to the self. Given that this chapter is devoted to ‘getting
liked’, social expressivity will be understood in terms of positive politeness to
show other users they are liked and belong to the group, e.g. by reacting to other
users’ photos or comments, contacting others, wishing happy birthdays, re-twit-
ting, sharing, etc.
As already mentioned, human beings are social beings that need ‘getting liked’
and having a sense of belonging (Srivastava and Beer 2005). Roughly following
Goffman’s (1967) notion of “face”, Brown and Levinson (1987: 62) defined this
want to be liked and appreciated in the following terms: “Positive face: the want
of every member that his [sic] wants be desirable to at least some others”, “a desire
to be ratified, understood, approved of, liked or admired”.
Leech (1983) included our desire to get liked, supported and understood in
three of his politeness maxims: approbation, agreement and sympathy. Prior to
that, Lakoff (1973: 298) also acknowledged (as a rule of politeness) “to make the
addressee feel good, be friendly”. It is beyond the scope of this chapter to deal with
all the criticism, some of it more than reasonable, that these seminal politeness
theories have attracted throughout the years (see Gilks 2010 and Locher 2012 for a
comprehensive summary of main critiques to Brown and Levinson); however, the
overall notion of positive politeness might still prove useful from a methodologi-
cal point of view, providing a point of departure to ascertain what kind of positive
facework users are performing online.
In any case, there is a consensus that human beings expect having at least some
of their wants and needs fulfilled by others. Thus, because of this need to be rati-
fied, understood, approved of, liked, or admired, a YouTuber who uploads a video
expects being watched and offered positive comments by other YouTubers (see
also Johansson, Ch. 7, this volume). Likewise, a Twitter user feels ratified when
their Tweet is chosen as a favourite and shared with the rest of the community.
Anonymous creators of memes may also derive great pleasure from having their

Authenticated
original idea re-used by others.15 On e-dating sites, getting liked and contacted by
other members responds to fulfilling the user’s wants and needs. In fact, attracting
a positive response from other users often triggers a response back in a give-and-
take fashion (Christophides, Muise and Desmarais 2009; Bronstein 2014) since
“the construction and maintenance of interpersonal relationships is central to com-
puter-mediated interaction” (Locher, Bolander and Höhn 2015: 6). Miller (2008;
2015) even talks about the creation of an online ‘phatic culture’, where users’ main
aim is to establish phatic communion with other users.16
By their own intrinsic nature, social networking media (e.g. Facebook, Twitter,
YouTube, LinkedIn, to name just a few) build on the very notion of positive face-
work and reciprocity to boost group solidarity “regardless of their unquestionable
differences in terms of their formats, functions, and overall technological affor-
dances” (Theodoropoulou 2015: 24). They all share the fact that user-generated
contents can be shared and commented by other users, which is indeed the spirit
of Web 2.0. Being part of these social media, hence, demands the users’ effort to
keep up with their online social world in what has been termed “analytic labour”
(Karakayali and Kilic 2013: 61). Being the most popular social network, however,
Facebook has received most attention from different disciplines (see also Eisen-
lauer, Ch. 9, this volume).17
In this respect, it is worth mentioning a relatively recent article by West and
Trester (2013) with the very telling title of “Facework on Facebook”. The authors
focus on the social expectations users have when interacting on the social network
and which mostly relate to positive politeness. For example, when Facebook users
change their profile or cover photo, they expect other users’ feedback, most of all
compliments (cf. Placencia and Lower, Ch. 23, this volume; Maíz-Arévalo and
García-Gómez 2013). Not interacting, i.e. being a ‘lurker’ or a passive user, can be
negatively interpreted as “neglecting […] facework obligations” (West and Trester
2013: 141). Similar results were obtained by Maíz-Arévalo (2013: 63), whose
interviewees often complained about the lack of other users’ comments:18
15
It goes without saying that some very popular or viral memes can also provide their
creators with financial rewards.
16
Miller (2015: 3) defines phatic communion as the “speech which is used to express or
maintain connection with others in the form of shared feelings, goodwill or general
sociability, rather than to impart information exchange.” Among the six functions of
language distinguished by Jakobson (1960), he also included the phatic functions of
language to establish bonds with others.
17
Remarkably, the latest especial issue of the International Pragmatics Association (IPrA)
is precisely on relational work on Facebook and discussion boards; see Pragmatics
(2015) vol. 25.
18
For the sake of space, only the English translation is quoted.

Authenticated
(12) I don’t get offended although I do know people who have got offended because their
sister-in-law hasn’t given any opinion about her photo, or her cousin or someone. […]
I personally don’t mind. […] I understand there are people who check Facebook once
a month, once a week, once every three months, or once or twenty times a day, but you
know that.
West and Trester (2013: 141) also point out that the very existence of a Facebook
open group named “If you don’t write on my wall then WHY did you friend request
me?” stresses the “importance of paying attention to members’ positive faces”. In
other words, being part of the network entails certain social obligations which
mostly translate into positive politeness. Showing we like others can be explicitly
verbalized or by means of the interface itself, via the like button. Prior research has
shown that the like button proves pragmatically very convenient (Maíz-Arévalo
2013; Santamaría-García 2014), “however, if clicking the like button is a way to
demonstrate interest without making a huge effort, it might also carry the meta-
message of being a minimal effort response […]. After all, since a choice to like
can be a choice not to say anything more about it, like can embed the very choice
of having chosen not to comment” (West and Trester 2013: 145).
This is nicely illustrated by another joke from the above mentioned Petri’s blog
“Fakebook”:
Figure 2. Facebook joke (quoted in West and Trester 2013: 145)
In this respect, it could be argued that the counterproductive effects underlying

the choice of the ubiquitous ‘like’ button might have led its creators to develop the
new “Reactions”, which were released in October 2015 and piloted in Ireland and
Spain before being launched worldwide. Remarkably enough, reactions still do
not include a dislike button, which might go against positive facework, but focus
on what Ekman (2007) calls “basic emotions”. Users, hence, can “react” to other
users’ posts and comments by clicking on one of the following emoji’s, apart from
the ‘like’ button, which is maintained.
Figure 3. Facebook’s new “reactions”

Authenticated
It goes without saying that this innovation on the interface may be opening new
avenues for further research so as to ascertain whether reactions are adding to the
users’ interactional behaviour or constraining them to express their thoughts in a
non-textual way. Together with reactions, another popular affordance on Facebook
is the birthday reminder, which allows users to show positive politeness towards
other users by congratulating them on their birthday. In this line, it is worth men-
tioning Theodoropoulou’s recent article (2015), where she analyses cultural dif-
ferences when responding to birthday wishes, focusing on Greek responses in con-
trast to other languages like English. Following Watts’s (2003) model of politeness
and Interactional Sociolinguistics (Schiffrin 1994; Gumperz 1999), she argues
that responding to birthday wishes in Greek is a paramount example of positive
politeness reciprocity since thanking is not enough and needs to be accompanied
by individual re-wishing rather than a communal thanking, which seems to be
the norm in other languages, especially among younger users (cf. also West and
Trester 2013: 149). Theodoropoulou’s results open up interesting cultural con-
trasts that need further investigation.
Together with Facebook, the second larger social networking site is LinkedIn
(Papacharissi 2009) targeted at professionals. In his thought-provoking article, van
Dijck analyses how changes in the interface of LinkedIn are increasingly targeted
“to boost the site’s social networking functions” (2013: 208). Hence, although still
focused on promoting the professional side of its users, LinkedIn has also added
interactional features like “newsfeed” or “shares”, encouraging positive facework
by its users rather than simply displaying their curricula. In fact, “many companies
encourage the use of LinkedIn for communication between colleagues and with
outside contacts” (2013: 209) up to the extent of LinkedIn having been nicknamed
“Facebook in a suit”. The hidden agenda of such changes, however, is obtaining
‘true’ information about their (potential) employees’ behaviour.
Finally, I would like to close this section by focusing on what Brown and Lev-
inson (1987) consider a macro strategy within positive facework: claiming com-
mon ground. Indeed, claiming and sharing common ground with others, whether
knowledge, experiences, likes, dislikes, etc., is a powerful way to strengthen social
relationships and to get liked by those with whom we share that common ground
(Baym 2006). For example, using in-group identity markers like specific language
only in-group members can identify versus newbies strengthens the formers’ rela-
tionship. Likewise, Twitter has been shown to create ad hoc communities of TV
watchers who are no longer simply watching their favourite programme but shar-
ing the experience with other fans via Twitter simultaneous interaction (Androut-
sopoulos 2015). This is illustrated by example (13), where an ordinary Twitter user
invites others to live the same experience by means of inclusive forms (“here we
are”):

Authenticated
(13) Ladies. Gentlemen. Here we are. It is happening. WELCOME TO THE GREAT

FINAL OF TOP CHEF. Welcome to the …#facetofacetopchef.19
Other strategies boosting common ground among users are implicitness, i.e. by
presupposing common ground, and humour. Given their importance, the following
two sections will be devoted to each of them respectively.
4. Implicitness
As Zhao, Grasmuck and Martin (2008) showed in their seminal study on Face-
book identity construction, users seem to recur both to explicit identity claims
and to implicit ones. In other words, they “show rather than tell” by visual means
(2008: 1816). In a prior study of chat messages, Nastri, Peña and Hancock (2006)
also found a recurrent employment of implicitness when users abandoned the key-
board. More recently, Bolander and Locher (2010) carried out a pilot study where
ten Swiss Facebook users tended to construct their identity by implicit comments
on their status updates rather than by making explicit claims about themselves (see
also Locher and Bolander, Ch. 15, this volume). Thus, they quote the following
example as an instance of implicit identity construction, where the user presents
herself as a student without explicitly saying so:
(14) <F7> has got to start writing her first assignment for university now 
In a more recent study on mock impoliteness on Facebook, Maíz-Arévalo also

found out that users tend to be more implicit so as to avoid being (mis)interpreted
as bragging or ‘taking themselves too seriously’, which often leads to being jocu-
larly mocked by other users, as in the following example (2015: 305):
(15) (Context: U1 has posted a photograph of himself practising what looks like a yoga
position).
U1 (m): I didn’t know I was that flexible
U2 (f): Impressive :-0
U3 (m): I won’t be impressed till you levitate, man
U1 (m): haha, very funny
Implicit comments also seem a convenient way to disclose highly personal infor-
mation without making it too ‘public’ (see Section 3.2). Such implicitness may
also help boost solidarity and “in-group bounding” (Bolander and Locher 2010:
183) since only close relationships who already share that common knowledge are
likely to come up with the correct inference. Similar results have been obtained
by Maíz-Arévalo (work in progress), where Facebook users tend to avoid express-
19
Top Chef is a cooking reality show followed by millions of viewers in Spain. Every
programme becomes trending topic.

Authenticated
ing very personal – and especially negative – information in their status updates.
When they do so, they seem to prefer implicit renderings, as illustrated by the
following two examples of Facebook updates:
(16) If I had known, I wouldn’t have come …
(17) That’s the way you’ve made me feel [accompanied by the photo of a mouse getting
hunted by a cat]
(18) My first word faded away
Interestingly enough, (16) triggered plenty of worried comments and questions,

which the User chose not to reply (maybe regretting having disclosed her emo-
tional outburst in the first place). (17) got supporting comments from other users
(e.g. “he’s not worth it”, “his loss”, etc.) who presumably knew what was going
on. Finally, (18) was the status update written by a user the same day his mother
passed away. This preliminary study seems to point out to the use of implicitness
as a means to avoid disclosing too personal information in a semi-public milieu
whilst simultaneously boosting in-group bonding with anchored and offline rela-
tionships, able to infer correctly what the message is about.
Finally, another reason why cybernauts may choose to be more implicit is their
intention to draw other cybernauts’ attention, hence getting liked, by appealing to
human beings’ natural ‘epistemic’ curiosity (cf. Berlyne 1954 for a seminal study
on human curiosity). Thus, encountering opaque, cryptic messages seems to trig-
ger in addressees the need to know more about them. Such strategy, which is often
employed by advertisers, may explain why some users choose to imply rather than
explicitly say or do something.
This section has focused on the use of implicitness as a strategy to get liked
in CMC. However, the relative scarcity of studies on this topic calls for further
research with larger samples including a more varied sociological spectrum. For
example, Zhao, Grasmuck and Martin (2008) included just sixty-three US stu-
dents whilst Nastri, Peña and Hancock (2006) based their results on merely for-
ty-four US students. Opening up to other languages and other cultural groups (e.g.
Bolander and Locher 2010) as well as to other CMC media apart from Facebook is
hence much needed (but see Chiluwa 2011 on the use of implicitness in a political
discussion forum).
5. Humour
Besides implicitness, one of the most powerful ways to create and boost common
ground is humour. In fact, both Lakoff (1973) and Brown and Levinson (1978,
1987) include joke telling as a positive politeness strategy and “an invitation to
demonstrate membership and solidarity” (Norrick 1989: 118). A cursory look at
Web 2.0. shows the ubiquity of humour (e.g. YouTube funny videos, viral memes,

Authenticated
etc.). This section will first revisit the introductory, seminal studies on humour in
CMC before reviewing the most recent ones and closing with some of the gaps still
remaining in this subfield.
Back in 1995, in her seminal study on humour in CMC, Baym already pointed
out the need to analyse its role in CMC: “Analysis of humor is important because
CMC research has been slow to address the formation of group identity and soli-
darity, though such phenomena occur in on-line groups and are negotiated, in part,
through humor” (Baym 1995).
Thus, focusing on the analysis of one of Usenet newsgroups on soap operas
(what might now parallel forums, Facebook common interest groups or simultane-
ous TV-watching while Twitting) Baym’s study (1995) demonstrated that humour
is widely used to create group solidarity, group identity, individual identity and
helped users to ‘get liked’ by their fellow members. As most of her interviewees
reported “there are few posts (or posters) more appealing than those which make
them laugh” since humour is “an important way in which participants connect to
one another and create the group’s social environment” (Baym 1995;20 see also
Norrick 1993; Palmer 1994).
A few years later, Holcomb (1997) analysed the process of spontaneous joke
making by students interacting in a synchronous way, e.g. via classroom chat.
Although his results paralleled Baym’s and also pointed out the crucial role humour
plays in boosting group solidarity, Holcomb found out that it might also lead to
asymmetric relationships by leaving out those who (i) did not get the joke, (ii)
chose not to laugh and (iii) were laughed at rather than with (Holcomb 1997: 4).
In the case of CMC, word play and misspellings seem to be a trigger for
humour, as in the following example by Holcomb (1997: 8) where one student
has written Boccaccio’s name inappropriately and is met by others’ jokes about it:
(19) Erik (Message #17):
Erin – Tell me about this Coccacio guy!
Michael (Message #20):
Erik – Coccaccio is Boccaccio’s evil twin brother.
More recent studies have also revealed that misspelling is frequently utilized as a
source for humorous exchanges, working as corrective facework when self-laughed
at, as in this exchange from a Spanish Facebook common interest group, which
triggers a whole bunch of humorous answers or “constitutive laughter” (Hübler
and Bell 2003: 280), aimed at enhancing the group’s rapport (Maíz-Arévalo, forth-
coming):
20
Available online at http://onlinelibrary.wiley.com/doi/10.1111/j.1083-6101.1995.
tb00327.x/full.

Authenticated
(20) LVL: he puesto “catamos”, cantamos, joperrrrr, cantamos, catar podemos catar tam-
bién, pero me refería a cantar, jajajajajajajajaja21
LLR: Te entendemos igualmente … jjjjjjjjj
LVL: ok, jajajajajjaajajajaja.
LD: Jijiii..cataremos algún cubata mientras cantamos,
LLR: A eso me apunto !!!!!
VAB: Y yo también me apunto y más si tengo que cantar!!
LLR: Con dos cantamos todo el repertorio !!!!
LD: Yo con uno … […]
[I’ve written “taste”, sing, darnnnnn, sing, taste we can taste too, but I meant to sing,
hahahahahahahaha // We understand you the same …hahahaha // ok, hahahahaha-
hahaha // hehehe..we’ll taste some cocktail while singing // count me in!!!! // Count
me in too, especially if I have to sing!! // With two [cocktails], we’ll sing the whole
repertoire!!!! // I need only one …]
Despite the potential negative effects of humour, later research has also demon-
strated that humour “pervades the rhetorical process of forming and maintaining
online groups” (Hübler and Bell 2003: 278). In their study, Hübler and Bell set out
to explore the role played by spontaneous humour22 in an asynchronous university
writing centre mailing list, to find out that “discursive threads of laughter help to
constitute an online community when individuals come to share as a group the
values and knowledge implicit in an ongoing joke” (2003: 280). They also agree
with Norrick (1993) on the central role played by self-deprecating humour, which
is also pervasive in CMC and highly contributes to getting liked by others since:
“The goodwill shown in making oneself the butt of a joke de-emphasizes hierarchy
and at the same time subtly incorporates the other primary components of ethos
by conveying an intelligent resourcefulness and a modest character” (Hübler and
Bell 2003: 281).
In a more recent article, Maíz-Arévalo (2015) focuses on the use of jocular
mockery among two groups of Spanish and English Facebook users, respectively.
Amongst other results, her study reveals the importance of self-deprecating humour
– as well as of accepting others’ jocular mockery – as a way to boost group solidar-
ity and rapport. As for the linguistic encoding of humour, her results parallel those
by previous research where users resort to typographic variation (e.g. capitalisa-
tion, emoticons, letter repetition, etc.) but also to other strategies found in face-to-
face communication such as lexical exaggeration, fomulaicity, topic shift markers
or contrastiveness. As the author points out, despite some common patterns, there
emerge interesting cultural differences, worth further research. Such cross-cultural
21
In Spanish, the form “catar” (to taste) and “cantar” (to sing) only differ by a consonant,
which explains the user’s misspelling.
22
As Holcomb (1997), Hübler and Bell (2003) also make a distinction between spontane-
ous, unplanned humour and planned, canned humour, e.g. like memes (see below).

Authenticated
studies, however, are far from frequent. Future research might want to focus on the
role and realization of CMC humour in languages other than English.
The previous studies share their focus on textually-based, spontaneous CMC
humour. Nevertheless, recent investigation has partially shifted its interest to mul-
timodal examples of CMC humour such as GIFs23 or memes, increasingly perva-
sive and ‘viral’ in our everyday life.
Internet memes24 have been defined as “units of popular culture that are cir-
culated, imitated, and transformed by Internet users, creating a shared cultural
experience” (Shifman 2013: 367). Nowadays, memes have become a part of our
everyday routine. The phenomenon of Internet memes has itself attracted growing
public interest, with web sites such as knowyourmeme.com, memedump.com, or
memebase.com acting as archives of famous memes.
In one of the first and seminal studies on popular memes, Knobel and Lank-
shear already defined them as “catchy and widely propagated ideas or phenomena”
(2007: 201). They adopted a functional discursive perspective to analyse a corpus
of popular and successful memes so as to determine what made them successful,
i.e. what made them get liked, by other users. According to Knobel and Lankshear
(2007: 209), successful memes seem to share at least some of the following major
characteristics:
i. Some element of humour (e.g. eccentric, potty humour, parodies, irony, etc.).
ii. Intertextuality (e.g. wry cross-references to different everyday events, icons or phe-
nomena).
iii. Anomalous juxtapositions (e.g. incongruity between the image and the text).
Successful memes can become so replicated in terms of fecundity and popularity

that they can transcend our smartphones and social media to be even talked about
and exploited by mainstream media. For example, the Star Wars Kid (a video
posted by a YouTuber in 200225) became so viral that it was referred to in Time,
Wired, the New York Times, the BBC, etc. which testifies to its tremendous popular-
ity. Another interesting study on viral YouTube memes was carried out by Shifman
23
GIF stands for “Graphics Interchange Format” and is defined by the online Oxford
Dictionary as “a set of standards and file format for storage of digital colour images and
short animations”. Catchy GIFs can become viral and derive into memes themselves.
For the sake of space, the rest of this section will focus on the latter.
24
The original term was popularised by Richard Dawkins in his celebrated The Selfish
Gene (1976), where he defined memes as replicators of ideas, knowledge, and other
cultural information through imitation and transfer, leading to actual neurological
changes. Memes are characterised by three main features: fidelity (or replicability),
fecundity (or rate at which it is copied and spread) and longevity. In this chapter, how-
ever, I shall adopt the most popular and current definition of online memes (see above).
25
Available at https://www.youtube.com/watch?v=HPPj6viIBmU.

Authenticated
in 2012. After analysing 30 memetic videos, the author found out that these videos
shared a set of common features that got them liked (and shared and/or replicated)
by other users: all of them focused on ordinary people, flawed masculinity, made
use of humour, were simple, repetitive and of whimsical content.
Memes, however, are interesting not only as humorous and entertaining mul-
timodal messages but as a way to understand “mindsets, new forms of power and
social processes, new forms of social participation and activism, and new distrib-
uted networks of communication and relationship – among other social phenom-
ena” (Knobler and Lankshear 2007: 201). As argued by Bauckhage (2011: 42),
memes have also been embraced by professionals in public relations and adver-
tising, e.g. in viral marketing, as well as by political campaigning, where memes
can be used to shape public opinion. However, it is still relatively unknown what
factors enhance “virality” (Shifman 2013).
In this line, in a more recent article Milner adopted a critical discourse analysis
perspective to study the crucial role played by memes in social protests like OWS
(Occupy Wall Street), where memes were found “to facilitate political debate across
participatory media networks” (2013: 2359). In other words, memes can be a way
for popular culture to actively participate in public social debates, hence its impor-
tance (see also Shifman 2013). The interplay between humour and protest may also
become a powerful way to “strip away artifice, highlight inconsistencies, and gen-
erally challenge the authority of official political discourse” (Dahlgren 2009: 139).
However, and despite some exceptions (e.g. Shifman 2007, 2012; Bauckhage
2011; Börzsei 2013; Milner 2013; Rintel 2013) the phenomenon of memes is still
overwhelmingly under-researched. As already argued, their potential in multi
faceted spheres, marketing data mining, political opinion, etc., is increasingly
acknowledged. As argued by Shifman in, to date, the most comprehensive book
on digital memes:
If indeed memes are evolving as “the language of the Internet”, future research should
take on the ambitious mission of understanding how this language is used by diverse
groups. One of the main questions relating to this process has to do with global and
local identities: to what extent do Internet memes constitute a “global” language […]
and to what extent do specific cultures create their own discrete memetic vernaculars?
(2013: 173–174)
Furthermore, given memes’ potential to generate political and socioeconomic
changes, the question of their real ‘effect’ and how it can be actually measured
also opens future avenues for research. As Shifman also argues:
There are considerably more questions marks than exclamation points in the state of our
knowledge of Internet memes. […] Since memes are fundamental building blocks of
digital culture, understanding them means understanding ourselves. A comprehensive
tapestry of the ways in which memes interact with social, cultural, and political reali-
ties will require a combination of quantitative “big data” analysis and qualitative close
reading of texts, as well as comparative cross-cultural studies. (2013: 175)

Authenticated
6. Conclusions and pointers to future research
This chapter has attempted to offer a comprehensive review of the main strategies
online users may employ when trying to fulfil the human need to get liked by oth-
ers. These strategies can be summed up into four intertwined aspects, individually
presented here for the sake of clarity: self-presentation or self-disclosure, social
expressivity, implicitness and humour. All of them, either in combination or not,
may contribute to establish users’ online identity and, as a result, get liked by their
fellow users.
As this brief review has attempted to show, the rapid development of emergent
forms of communication on Web 2.0 has been enthusiastically tackled by research-
ers from a myriad of disciplines, including pragmatics. However, the swift advance
of digital communication, not necessarily mediated by computers but also by other
devices like Smartphones, requires research into other means of communication
such as WhatsApp or Snapchat.26 Remarkably enough, and despite a few excep-
tions (Calero-Vaquera and Vigara 2014; Sánchez-Moya and Cruz-Moya 2015),
these popular messaging applications are still under-researched, which opens up
new avenues for future studies.
Furthermore, a more multimodal focus is needed (Androutsopoulos 2015). As
Bourlai and Herring (2014) point out in their analysis of Tumblr posts, different
modes (e.g. visual, textual, etc.) are becoming increasingly employed for different
communicative purposes, with images expressing more emotion and valence pos-
itivity than posts containing only text. Their pioneer study opens up new avenues
for research, where special attention is paid to the visual code. Future research into
more multimodal ways of communication might also concentrate, for example, on
the phenomenon of viral messages such as memes and gifs. Likewise, it would be
interesting to carry out more extensive research on the influence of globalization
processes to find out whether they are bringing about homogenized behaviour
(Sifianou 2013). This would be closely connected to further research on cross-cul-
tural studies on Web 2.0 phenomena, to find out whether getting liked involves
the same in different cultures. These suggestions are merely an attempt – far from
comprehensive – to provide some potential aspects for future research. With the
rapid expansion of online technologies, the possibilities are unlimited.
26
Snapchat is a mobile messaging service that sends a photo or video to someone that lasts
only up to 10 seconds before it disappears. Its ephemeral nature has made it extremely
popular and is likely to have important effects in the way users interact.

Authenticated
References
Altman, Irwin and Dalmas Taylor

1973 Social Penetration: The Development of Interpersonal Relationships. New
York: Holt, Rinehart and Winston.
2006 Special Introduction: Sociolinguistics and Computer-mediated communica-
tion. Journal of Sociolinguistics 10(4): 419–438.
2015 Towards a ‘third wave’ of digital discourse studies: Audience practices on
Twitter. Plenary talk at ADDA I Conference, University of Valencia, 18–20
November. Available at http://adda.blogs.uv.es/plenary-speakers/jannis-an-
droutsopoulos/.
Aronson, Elliot
1984 The Social Animal. New York: NH Freeman.
Attrill, Alison
2012 Sharing only parts of me: Selective categorical self-disclosure across Internet
arenas. International Journal of Internet Science 7 (1): 55–77.
Baker, Andrea J.
2005 Double Click: Romance and Commitment among Online Couples. Cresskill,
NJ: Hampton.
Bargh, John A., Katelyn Y. A. McKenna and Grainne M. Fitzsimons
2002 Can you see the real me? Activation and expression of the ‘‘true self” on the
Internet. Journal of Social Issues 58(1): 33–48.
Bauckhage, Christian
2011 Insights into Internet memes. Proceedings of the 5th International AAAI Con-
ference on weblogs and social media, 42–49.
Baumeister, Roy F.
1982 Self-esteem, self-presentation, and future interaction: A dilemma of reputa-
tion. Journal of Personality 50(1): 29–45.
Baumeister, Roy F. and Mark R. Leary
1995 The need to belong: Desire for interpersonal attachments as a fundamental
human motivation. Psychological Bulletin 117(3): 497–529.
Baumeister, Roy F. and Dianne M. Tice
1984 Role of self-presentation and choice in cognitive dissonance under forced
compliance: Necessary or sufficient causes? Journal of Personality and Social
Psychology 46(1): 5–13.
Baym, Nancy
1995 The performance of humor in computer-mediated communication. Journal
of Computer-Mediated Communication 1(2). http://onlinelibrary.wiley.com/
doi/10.1111/j.1083-6101.1995.tb00327.x/full.
Baym, Nancy
2003 Interpersonal Life Online. In: Leah A. Lievrouw and Sonia Livingstone (eds.),
Handbook of New Media: Student Edition, 35–54. London: Sage.
Beals, Shawn R.
2010 Superintendent’s Facebook posts could be trouble. Hartford Courant. http://
articles.courant.com/2010-07-28/news/hc-windsorlocks-superintendent-

Authenticated
0728-20100727_1_superintendent-s-job-new-schoolsuperintendent-
social-media.
Berlyne, Daniel E.
1954 A theory of human curiosity. British Journal of Psychology. General Section
45(3): 180–191.
Börzsei, Linda K.
2013 Makes a meme instead: A concise history of Internet memes. New Media Stud-
ies Magazine, Issue 7. http://works.bepress.com/linda_borzsei/2/.
2010 Constructing identity on Facebook: Report on a pilot study. In: Karen Junod
and Didier Maillât (eds.), Performing the Self. SPELL: Swiss Papers in Eng-
lish Language and Literature 24, 165–187. Tübingen: Narr.
Bonebrake, Katie
2002 College students’ Internet use, relationship formation, and personality corre-
lates. CyberPsychology and Behavior 5: 551–557.
Bond, Bradley J.
2009 He posted, she posted: Gender Differences in self-disclosure on Social Net-
working Sites. Rocky Mountain Communication Review 6(2): 29–37.
boyd, Danah
2007 Why youth <3 social network sites: The role of networked publics in teenage
social life. In: David Buckingham (ed.), Youth Identity and Digital Media,
119–142. Cambridge MA: MIT Press.
Bourlai, Elli and Susan C. Herring
2014 Multimodal communication on tumblr: i have so many feels!. Proceedings of
the 2014 ACM Conference on Web Science: 171–175. http://delivery.acm.org/
10.1145/2620000/2615697/p171-bourlai.pdf?ip=93.219.186.163&id=2
615697&acc=OPENTOC&key=4D4702B0C3E38B35 %2E4D4702B-
0C3E38B35 %2E4D4702B0C3E38B35 %2E6656B80B51FFDE8C&CFID=
870339813&CFTOKEN=22735123&__acm__=1480501729_223f41582431
8477ed709576cf504c20.
Brandtzaeg, Petter B., Marika Lüders and Jan Håvard Skjetne
2010 Too many Facebook ‘friends’? Content sharing and sociability versus the need
for privacy in social network sites. Journal of Human-Computer Interaction
26(11): 1006–1030.
Bronstein, Jenny
2012 Blogging Motivations for Latin American Bloggers: A Uses and Gratifications
Approach. In: Tatyana Dumova (ed.), Blogging in the Global Society, 200–
215. Hershey, PA: Information Science Reference.
Bronstein, Jenny
2014 Creating possible selves: information disclosure behaviour on social networks.
Information Research 19(1): 1–14.
versity Press.
Calero-Vaquera, María Luisa and A. Vigara Tauste
2014 El discurso del WhatsApp: entre el Messenger y el SMS. Oralia 17: 87–116.

Authenticated
Chiluwa, Innocent
2011 On political participation: Discursive pragmatics and social interaction in pol-
itics. Studies in Literature and Language 2(2): 80–92.
Chou, Hui-Tzu Grace and Nicholas Edge
2012 “They are happier and having better lives than I am”: the impact of using Face-
book on perceptions of others’ lives. Cyberpsychology, Behavior, and Social
Networking 15(2): 117–121.
Christofides, Emily, Amy Muise and Serge Desmarais
2009 Information disclosure and control on Facebook: Are they two sides of the
same coin or two different process? CyberPsychology and Behavior 12(3):
341–345.
Collins, Nancy L. and Lynn C. Miller
1994 Self-disclosure and liking: A meta-analytic review. Psychological Bulletin
116: 457–475.
Cooley, Skye C. and Lauren Reichart Smith
2013 Presenting me! An examination of self-presentation in US and Russian online
social networks. Russian Journal of Communication 5(2): 176–190.
Cozby, Paul C.
1973 Self-disclosure: A literature review. Psychological Bulletin 79(2): 73–91.
Davey, Christopher G., Nicholas B. Allen, Ben J. Harrison, Dominic B. Dwyer and Murat
Yücel
2010 Being liked activates primary reward and midline self-related brain regions.
Human Brain Mapping 31(4): 660–668.
Dahlgren, Peter
2009 Media and Political Engagement: Citizens, Communication, and Democracy.
Eklund, Lina
2015 Bridging the online/offline divide: The example of digital gaming. Computers
in Human Behaviour 53: 527–535.
Ekman, Paul
2007 Emotions Revealed: Recognizing Faces and Feelings to Improve Communica-
tion and Emotional Life. London: Macmillan.
Ellemers, Naomi, Bertjan Doosje and Russell Spears
2004 Sources of respect: The effects of being liked by ingroups and outgroups.
European Journal of Social Psychology 34(2): 155–172.
Ellison, Nicole, Rebecca Heino and Jennifer Gibbs
2006 Managing impressions online: self-presentation processes in the online dating
environment. Journal of Computer-Mediated Communication 11(2): 415–441.
García-Gómez, Antonio
2009 Teenage girls’ personal weblog writing: Truly a new gender discourse? Infor-
mation, Communication and Society 12(5): 611–638.
García-Gómez, Antonio
2010 Disembodiment and cyberspace: Gendered discourses in female teenagers’
personal information disclosure. Discourse and Society 21(2): 135–160.
2011 Computer-mediated communication. In: Jan-Ola Östman and Jef Verschueren
(eds.), Pragmatics in Practice, 93–110. Amsterdam/Philadelphia: Benjamins.

Authenticated
Gibbs, Jennifer L., Nicole B. Ellison and Rebecca D. Heino

2006 Self-presentation in online personals: The role of anticipated future interac-
tion, self-disclosure, and perceived success in Internet dating. Communication
Research 33(2): 152–177.
Gilks, Kate
2010 Is the Brown and Levinson (1987) model of politeness as useful and influen-
tial as originally claimed? An assessment of the revised Brown and Levinson
(1987) model. Leading Undergraduate Work in English Studies 2: 2009–2010.
Goffman, Erving
1959 The Presentation of Everyday Life. New York: Anchor Books.
Goffman, Erving
1967 Interaction Ritual: Essays on Face-to-face Interaction. Oxford: Aldine.
Goleman, Daniel
1995/2006 Emotional Intelligence. New York: Bantam Dell.
Gonzales, Amy L. and Jeffrey T. Hancock
2011 Mirror, mirror on my Facebook wall: Effects of exposure to Facebook on
self-esteem. Cyberpsychology, Behavior, and Social Networking 14(1–2):
79–83.
Gumperz, John J.
1999 On interactional sociolinguistic method. In: Srikant Sarangi and Celia Roberts
(eds.), Talk, Work and Institutional Order: Discourse in Medical, Mediation
and Management Settings, 453–472. Berlin/Boston: de Gruyter.
Hancock, Jeffrey T. and Catalina L. Toma
2009 Putting your best face forward: The accuracy of online dating photographs.
Journal of Communication 59: 367–386.
Hodkinson, Paul and Sian Lincoln
2008 Online journals as virtual bedrooms? Young people, identity and personal
space. Young 16(1): 27–46.
Holcomb, Christopher
1997 A class of clowns: Spontaneous joking in computer-assisted discussions. Com-
puters and Composition 14(1): 3–18.
Huang, Chih-Mao and Denise Park
2013 Cultural influences on Facebook photographs. International Journal of Psy-
chology 48(3): 334–343.
Hübler, Mike T. and Diana Calhoun Bell
2003 Computer-mediated humor and ethos: Exploring threads of constitutive laugh-
ter in online communities. Computers and Composition 20(3): 277–294.
Huo, Yuen J. and Kevin R. Binning
2008 Why the psychological experience of respect matters in group life: An integra-
tive account. Social and Personality Psychology Compass 2(4): 1570–1585.
Hum, Noelle J., Perrin E. Chamberlin, Brittany L. Hambright, Anne C. Portwood, Amanda
C. Schat and Jennifer L. Bevan
2011 A picture is worth a thousand words: A content analysis of Facebook profile
photographs. Computers in Human Behavior 27(5): 1828–1833.
Jakobson, Roman
1960 Linguistics and Poetics. In: Thomas A. Sebeok (ed.), Style in Language, 350–
377. Cambridge, MA: MIT-Press.

Authenticated
Jourard, Sidney M.
1971 Self-disclosure: An Experimental Analysis of the Transparent Self. Oxford:
John Wiley and Sons.
Karakayali, Nedim and Aysenur Kilic
2013 More network conscious than ever? Challenges, strategies, and analytic labor
of users in the Facebook environment. Journal of Computer-Mediated Com-
munication 18(2): 61–79.
Keyes. Corey Lee M.
1998 Social well-being. Social Psychology Quarterly 61: 121–140.
Kim, Jinsuk and Kathryn Dindia
2011 Online self-disclosure: A review of research. In: Kevin B. Wright and Lynne
M. Webb (eds.), Computer-mediated Communication in Personal Relation-
ships, 156–180. New York: Peter Lang.
Knobel, Michele and Colin Lankshear
2007 A New Literacies Sampler. Vol. 29. New York: Peter Lang.
Kowalski, Robin M. and Mark R. Leary.
1990 Strategic self-presentation and the avoidance of aversive events: Anteced-
ents and consequences of self-enhancement and self-depreciation. Journal of
Experimental Social Psychology 26(4): 322–336.
Krasnova, Hanna, Helena Wenninger, Thomas Widjajas and Peter Buxmann
2013 Envy on Facebook: a hidden threat to users’ life satisfaction? Wirtschaftsinfor-
matik 92: 1–16.
Kross, Ethan, Philippe Verduyn, Emre Demiralp, Jiyoung Park, David Seungjae Lee,
Natalie Lin, Holly Shablack, John Jonides and Oscar Ybarra
2013 Facebook use predicts declines in subjective well-being in young adults. PloS
one 8(8): e69841.
Lakoff, Robin
1973 The logic of politeness: Or, minding your p’s and q’s. Papers from the Ninth
Regional Meeting of the Chicago Linguistic Society, Chicago Linguistic Soci-
ety: 292–305.
Leary, Mark R. and Roy F. Baumeister
2000 The nature and function of self-esteem: Sociometer theory. Advances in Exper-
imental Social Psychology 57: 1–62.
Leary, Mark R. and Ashley Batts Allen
2011 Self-presentational persona: Simultaneous management of multiple impres-
sions. Journal of Personality and Social Psychology 101(5): 1033–1049.
Leech, Geoffrey N.
1983 Principles of Pragmatics. London: Taylor and Francis.
Livingstone, Sonia
2005 On the relation between audiences and publics. In: Sonia Livingstone (ed.),
Audiences and Publics: When Cultural Engagement Matters for the Public
Sphere, 17–42. Portland, OR: Intellect.
Locher, Miriam
2012 Politeness research from past to future, with a special focus on the discursive
approach. In: Lucía Fernández Amaya, Maria de la O Hernández López and
Reyes Gómez Morón (eds.), New Perspectives on (Im)Politeness and Inter-
personal Communication, 36–60. Newcastle upon Tyne: Cambridge Scholars
Publishing.

Authenticated

2015 Humour in microblogging: Exploiting linguistic humour strategies for identity
construction in two Facebook focus groups. In: Marta Dynel and Jan Cho-
vanec (eds.), Participation in Public and Social Media Interactions, 135–155.
Locher, Miriam A., Brook Bolander and Nicole Höhn
2015 Introducing relational work in Facebook and discussion boards. Pragmatics
25(1): 1–21.
2013 “Just click ‘Like”’: Computer-mediated responses to Spanish compliments.
2015 Jocular mockery in computer-mediated communication: A contrastive study of
a Spanish and English Facebook community. Journal of Politeness Research
11(2): 289–327.
Manago, Adriana M., Michael B. Graham, Patricia M. Greenfield and Goldie Salimkhan
2008 Self-presentation and gender on MySpace. Journal of Applied Developmental
Psychology 29(6): 446–458.
Marwick, Alice E.
McKenna, Katelyn Y. A. and John A. Bargh
2000 Plan 9 from Cyberspace: the implications of the Internet from personality and
social psychology. Personality and Social Psychology Review 4 (1): 57–75.
Mesch, Gustavo S. and Guy Beker
2010 Are norms of disclosure of online and offline personal information associated
with the disclosure of personal information online? Human Communication
Research 36: 570–592.
Michikyan, Minas, Subrahmanyam, Kaveri and Jessica Dennis
2015 Facebook use and academic performance among college students: A mixed-
methods study with a multi-ethnic sample. Computers in Human Behavior 45:
265–272.
Miller, Vincent
2008 New media, networking and phatic culture. Convergence: The International
Journal of Research into New Media Technologies 14(4), 387–400.
Miller, Vincent
2015 Phatic culture and the status quo: Reconsidering the purpose of social media
activism. Convergence: The International Journal of Research into New
Media Technologies 1–19. doi: 10.1177/1354856515592512.
Milner, Ryan M.
2013 Pop polyvocality: Internet memes, public participation, and the Occupy Wall
Street Movement. International Journal of Communication 7: 2357–2390.
Nadkarni, Ashwini and Stefan G. Hofmann
2012 Why do people use Facebook? Personality and Individual Differences 52(3):
243–249.
Nastri, Jacqueline, Jorge Peña and Jeffrey T. Hancock
2006 The construction of away messages: A speech act analysis. Journal of Compu
ter-Mediated Communication 11(4): 1025–1045.

Authenticated
Norrick, Neal R.
1989 Intertextuality in humor. Humor – International Journal of Humor Research
2(2): 117–140.
Nosko, Amanda W., Eileen Wood and Seija Molema
2010 All about me: disclosure in online social networking profiles. The case of
Facebook. Computers in Human Behavior 26(3): 406–418.
Palmer, Jerry
1994 Taking Humour Seriously. New York: Routledge.
Papacharissi, Zizi
2009 The virtual geographies of social networks: a comparative analysis of Face-
book, LinkedIn and ASmallWorld. New Media and Society 11(1–2): 199–220.
Papacharissi, Zizi (ed.)
2011 A Networked Self: Identity, Community, and Culture on Social Network Sites.
London: Routledge.
Punset, Eduardo
2000 El Viaje a la Felicidad [The Trip to Happiness]. Barcelona: Destino libro.
Qian, Hua and Craig R. Scott
2007 Anonymity and self-disclosure on weblogs. Journal of Computer-Mediated
Reed, Adam
2005 ‘My blog is me’: Texts and persons in UK online journal culture (and anthro-
pology). Ethnos 70(2): 220–242.
Rheingold, Howard
1995 The Virtual Community: Finding Connection in a Computerized World. Lon-
don: Secker and Warburg.
Rintel, Sean
2013 Crisis memes: The importance of templatability to Internet culture and free-
dom of expression. Australasian Journal of Popular Culture 2.2: 253–271.
Rosenmann, Amir and Marilyn P. Safir
2006 Forced online: push factors of internet sexuality: A preliminary study of online
paraphilic empowerment. Journal of Homosexuality 51(3): 71–92.
Sagioglou, Christina and Tobias Greitemeyer
2014 Facebook’s emotional consequences: Why Facebook causes a decrease in
mood and why people still use it. Computers in Human Behavior 35: 359–363.
Sánchez-Moya, Alfonso and Olga Cruz-Moya
2015 “Hey there! I am using WhatsApp”: A preliminary study of recurrent discur-
sive realisations in a corpus of WhatsApp statuses. Procedia – Social and
Behavioral Sciences 212: 52–60.
Santamaría-García, Carmen
2014 Evaluative discourse and politeness in university students’ communication
through social networking sites. In: Geoff Thompson and Laura Alba-Juez
(eds.), Evaluation in Context, 387–412. Amsterdam/Philadelphia: Benjamins.
Schau, Hope Jensen and Mary C. Gilly
2003 We are what we post? Self-presentation in personal web space. Journal of
Consumer Research 30(3): 385–404.
Schiffrin, Deborah
1994 Approaches to Discourse. Malden/Oxford: Blackwell.

Authenticated
Shifman, Limor
2012 An anatomy of a YouTube meme. New Media and Society 14(2): 187–203.
Shifman, Limor
2013 Memes in a digital world: Reconciling with a conceptual troublemaker. Jour-
nal of Computer-Mediated Communication 18(3): 362–377.
Sifianou, Maria
2013 The impact of globalisation on politeness and impoliteness. Journal of Prag-
matics 55: 86–102.
Sifianou, Maria
Smith, Lauren Reichart and Skye C. Cooley
2012 International faces: An analysis of self-inflicted face-ism in online profile pic-
tures. Journal of Intercultural Communication Research 41(3): 279–296.
Strano, Michele M.
2008 User descriptions and interpretations of self-presentation through Face-
book profile images. Cyberpsychology: Journal of Psychosocial Research
on Cyberspace 2(2). http://www.cyberpsychology.eu/view.php?cislo-
clanku=2008110402&article=1.
Srivastava, Sanjay and Jennifer S. Beer
2005 How self-evaluations relate to being liked by others: integrating sociometer
and attachment perspectives. Journal of Personality and Social Psychology
89(6): 966–977.
Suler, John
2002 Identity management in cyberspace. Journal of Applied Psychoanalytic Stud-
ies 4(4): 455–459.
Suler, John
2004 The online disinhibition effect. Cyberpsychology and Behavior 7: 321–326.
Surratt, Carla G.
1998 Netlife: Internet Citizens and their Communities. New York: Nova Science.
Suzuki, Lalita K. and Jerel P. Calzo
2004 The search for peer advice in cyberspace: An examination of online teen bulle-
tin boards about health and sexuality. Journal of Applied Developmental Psy-
chology 25(6): 685–698.
Tandoc, Edson C., Patrick Ferrucci and Margaret Duffy
2015 Facebook use, envy, and depression among college students: Is facebooking
depressing? Computers in Human Behavior 43: 139–146.
23–45.
Toma, Catalina L., Jeffrey T. Hancock and Nicole B. Ellison
2008 Separating fact from fiction: An examination of deceptive self-presentation
in online dating profiles. Personality and Social Psychology Bulletin 34(8):
1023–1036.
Toma, Catalina L. and Jeffrey T. Hancock
2011 A new twist on love’s labor: Self-presentation in online dating profiles. Com-
puter-mediated Communication in Personal Relationships: 41–55.

Authenticated
Turkle, Sherry
1995 Life on the Screen: Identity in the Age of the Internet. New York: Simon and
Schuster.
Valenzuela, Sebastián, Namsu Park and Kerk F. Kee
2009 Is there social capital in a social network site?: Facebook use and college stu-
dents’ life satisfaction, trust, and participation. Journal of Computer-Mediated
Van Dijck, José
2013 ‘You have one identity’: Performing the self on Facebook and LinkedIn.
Media, Culture and Society 35(2): 199–215.
Valkenburg, Patti M. and Jochen Peter
2009 The effects of instant messaging on the quality of adolescents’ existing friend-
ships: A longitudinal study. Journal of Communication 59(1): 79–97.
Walker, Katherine
2000 ‘‘It’s difficult to hide it”: The presentation of self on Internet home pages.
Qualitative Sociology 23(1): 99–120.
Walther, Joseph B.
1996 Computer-mediated communication: Impersonal, interpersonal, and hyperper-
sonal interaction. Communications Research 23: 3–43.
Watts, Richard J.
Weir, Kirsten
2012 The pain of social rejection. American Psychological Association 43(4): 50.
http://www.apa.org/monitor/2012/04/rejection.aspx [accessed on November
2015].
Weisbuch, Max, Zorana Ivcevic and Nalini Ambady
2009 On being liked on the web and in the “real world”: Consistency in first impres-
sions across personal webpages and spontaneous behavior. Journal of Experi-
mental Social Psychology 45(3): 573–576.
West, Laura and Anna Marie Trester
2013 Facework on Facebook: Conversations on social media. In: Deborah Tannen
and Anna M. Trester (eds.), Discourse 2.0, 133–154. Georgetown: George-
town University Press.
Yang, Chia-chen and B. Bradford Brown
2016 Online self-presentation on Facebook and self-development during the college
transition. Journal of Youth and Adolescence 45(2): 402–416.
Yum, Young-ok and Kazuya Hara
2006 Computer-mediated relationship development: A cross-cultural comparison.
Yurchisin, Jennifer, Kittichai Watchravesringkan and Deborah Brown McCabe
2005 An exploration of identity recreation in the context of internet dating. Social
Behavior and Personality: an International Journal 33(8): 733–750.

Authenticated
22. Conflictual and consensual disagreement
Brook Bolander and Miriam A. Locher
Abstract: This chapter explores a series of foci which have emerged in the study
of disagreement in social media, notably in Web 2.0 environments. These include
comparisons between disagreement online and offline; the role played by social
and medium factors for the linguistic realisation and emergent meanings of dis
agreement; the recent upsurge in work on sociable disagreement and play; a focus
on disagreement, language and gender; the exploration of disagreement in con-
nection with polylogues and participation frameworks; increased interest in dis-
agreement in educational contexts; and a progressive move to include data from
language varieties other than English. To account for the fact that disagreement
is a move which can have various manifestations and meanings, we also include
discussion of “conflictual disagreement” and “consensual disagreement”. In the
course of the chapter, we further highlight the close relationship between changes
in the study of disagreement online and changes in the study of language use
online more generally; and we address contemporary research on the challenge
of context and the indexing of emotion as pertinent for the study of disagreement
online.
1. Introduction
This chapter focuses on conflictual and consensual disagreement in social media.

Social media is thereby broadly conceived of as “(the totality of) digitally medi-
ated and Internet-based platforms which are interactively used (by individual and
collective participants) to exchange, share and edit self- and other-generated tex-
tual and audio-visual messages” (Hoffmann, Ch. 1, this volume). In this entry, dis-
agreement is conceptualised as a ‘discourse move’ (Locher 2006) or contribution
used to express opposition with a position, stance, or view taken by another party/
ies and/or source (cf. e.g., Locher 2004; Baym 1996; Bolander 2012, 2013). In
exploring disagreement, we are particularly interested in its linguistic and semiotic
performance (achievement and effects), and in the multitude of social and medium
factors (Herring 2007) which influence its enactment online.
The beginning of research on disagreement online can be traced back to the
mid 1980s (cf. e.g., Sproull and Kiesler 1986), with systematic, empirical research
on disagreement emerging from the mid 1990s onwards (Baym 1996). Much has
changed during this time, with technological advances concurring with changes
in epistemology and approach. The shift from Web 1.0 to Web 2.0 has led to the
DOI 10.1515/9783110431070-022
Authenticated
608 Brook Bolander and Miriam A. Locher
emergence of new environments, modes and interactional affordances which offer

increased possibilities for participation. With respect to approach and epistemol-
ogy, we find a progressive move away from a view of disagreement as disruptive,
to one which acknowledges the influence of multiple social and medium factors
on issues of performance and perception. In other words, there is a shift towards
a more varied and differentiated understanding of the functions of and different
types of disagreement.
We begin our chapter by conceptualising disagreement against the backdrop
of changes in technology and epistemology (Section 2), before reviewing key lit-
erature on conflictual and consensual disagreement (Section 3). Subsequently, we
tackle two issues we view as pertinent to contemporary research on disagreement
(Section 4): the challenge of context and the indexing of emotion. Finally, we pro-
vide an overview of key points and arguments, and an outlook to future research
(Section 5).
2. Disagreement: Conceptualisation and development from Web 1.0 to

Web 2.0
In the first special issue to deal with disagreement offline and online, Angouri
and Locher (2012: 1551) argue for an approach which acknowledges four key
factors: 1) “expressing opposing views is an everyday phenomenon”; 2) “certain
practices are prone to contain disagreement so that this speech act is expected
rather than the exception”; 3) “disagreeing cannot be seen as an a priori nega-
tive act”; and 4) “the ways in which disagreement is expressed – and not only its
occurrence per se – will have an impact on relational issues”, with “expectations
about how disagreement is valued in a particular practice […] influenc[ing] what
forms participants choose”. In making these points, the authors draw attention to
the pervasiveness of disagreement; and to the importance of addressing context
and “frames” (or “structures of expectation”, Tannen 1993: 53) in connection with
how often and how disagreements are performed and evaluated. Disagreement
can, in other words, vary with respect to how conflictual or consensual it is; and
it can range from boosted, face-threatening, unexpected disagreement, through to
consensual, expected disagreement, to sociable disagreement, and play. Situations
where disagreement is not expected (and often dispreferred; Pomerantz 1984) and
formulated in a face-threatening manner (e.g., boosted) give rise to what we term
“conflictual disagreement”; whereas situations where disagreement is expected
and its manifestation pertains to the expected social norms give rise to “consen-
sual disagreement”. These are not, however, to be understood as dichotomous, but
rather as positions on a cline, with fuzzy boundaries, such that it is also possible
for consensual disagreement to become conflictual, or for the two to co-occur
within the same exchange.

Authenticated
Conflictual and consensual disagreement 609
In acts of disagreeing, both informational and relational aspects are of impor-

tance. As stated by Schneider (2010: 254), it is not possible to sever “the mere
transmission of information” from “properties due to the speaker and hearer”.
“[I]nformational properties” and “relational properties” overlap, “since both the
content of a speech act and the manner in which the content is expressed define a
relation between the interactants” (Schneider 2010: 254). Analogously a focus on
disagreement from the perspective of content needs to be complemented with an
interpersonal perspective (Locher and Graham 2010), which explores the interplay
between the sharing of information and “relational work” (Locher and Watts 2005,
2008), “relational practice” (Holmes and Schnurr 2005) or “rapport management”
(Spencer-Oatey 2000).
In the earliest studies of conflictual disagreement online, it was assumed that
both the form and subsequent meaning of the disagreement stemmed directly from
the immediate context of the online environment, which was understood with
respect to its technological properties. The perceived lack of “social context cues”
in online settings compared with offline ones, coupled with the notion that com-
munication online had “few widely shared norms” (Kiesler, Siegel and McGuire
1984: 1126) and was predominantly anonymous, were seen as linked to an increase
in uninhibited behaviour online; including both “increases in friendliness and inti-
macy” and “increase[s] in examples of aggressive and disrespectful behavior”
(Reid 1999: 111). With respect to the linguistic realization of disagreement, this
suspension of “norms governing disagreement in oral interaction” (Baym 1996:
323) in online settings suggested a prevalence of boosted as opposed to mitigated
disagreement. And this argument provided grounds for a view of online and offline
interaction as fundamentally different. While acknowledging that this view can be
enticing given examples of exaggerated disagreement, Baym (1996: 323) argues
for a more nuanced interpretation, which recognizes that a multitude of factors
influence the realisation and social meaning of disagreements and agreements. In
the case of her own research on Usenet (cf. Section 3.1 below), these include its
technologocial properties, as well as the context, topic, goals, norms and charac-
teristics of the participants (Baym 1996).
This shift from explanations inspired by “technological determinism” to ones
acknowledging that various co-occurring factors shape disagreement online, par-
allels a broader move from a “first wave” of scholarship on language use online to
a second and even third wave (Androutsopoulos 2006: 420–421). Without ignor-
ing the role played by technology, this latter research engages with the multiplicity
of ties between form and meaning, and with notions of performance. In recent
studies of disagreement, which will be reviewed in Section 3, it is thus typical to
find a wide range of factors influencing the performance and evaluation of disa-
greement in social media.
Concurrent with and related to these changes in approach, there has been a shift
from studies on Web 1.0 to Web 2.0. Web 2.0 is largely connected to the emer-

Authenticated
gence, in the 2000s, of “a new generation of websites [which] integrate applica-

tions for interpersonal communication and tools for the management of user-gen-
erated content” (Androutsopoulos 2010: 207). These typically include “social
networking and media-sharing sites” (Androutsopoulos 2010: 207), such as blog-
ging and microblogging services (including Twitter) and social networking sites
(SNS) like Facebook and Instagram, which encourage interactivity, participation
and exchange (Androutsopoulos 2010; Yus 2011; and also Heyd, Ch. 6, this vol-
ume for blogs, Zappavigna, Ch. 8, this volume for Twitter, and Eisenlauer, Ch. 9,
this volume for Facebook). As outlined by Zappavigna (2012: 2), Web 2.0 – or the
“social web” – underscores a view of the Internet as a space where relationships are
enacted (as opposed to one solely or primarily dedicated to information exchange).
The emergence of such sites is clearly of general interest to pragmatics (Yus 2011:
93). With respect to disagreement, a Web 2.0 environment can (at least theoret-
ically) also facilitate the performance of disagreement by participants who can
choose to adopt different roles and rely on a gamut of technological affordances for
the enactment of these moves (on both an ideational and relational level).
3. Literature review on conflictual and consensual disagreement
Viewed as typically dispreferred in early conversation analytic work on casual face-

to-face interaction (Pomerantz 1984), literature on conflictual disagreement both
offline and online has drawn attention to its potentially face-threatening nature.
Conflictual disagreement can have negative effects on one’s own or one’s interac-
tional partner’s face, or both. According to Goffman (1967: 5), face is understood
as “the positive social value a person effectively claims for himself [sic] by the
line others assume he has taken during a particular contact”. Conflictual disagree-
ment can thus potentially lead to relational tensions and influence the likelihood
for particular individual and/or group goals to be achieved. It is the potentially
face-threatening nature of conflictual disagreement which renders it conceptually
similar to flaming (cf. Hardaker 2010, and Ch. 18, this volume) and impoliteness
(Graham, Ch. 17, this volume; see also Dynel 2015: 339–340). These overlaps
are reflected in the literature, with scholars exploring both flaming and conflict
(cf., e.g., Bomberger 2004; Turnage 2007) or disagreement and impoliteness (cf.,
e.g., Angouri and Tseliga 2010) in various online contexts. Yet despite conceptual
overlaps, there are differences. Flaming is typically defined as hostile, aggressive
behaviour (Turnage 2007), yet disagreement can range in degree and intensity
(e.g., from conflictual to consensual); and whereas impoliteness can emerge in
situations where there is no opposition in viewpoint/stance, such opposition is
central to the conceptualisation of disagreement (Kakavá 1993).
A review of key literature on disagreement underscores at least seven main, yet
often related, foci. These include

Authenticated
1. the comparative study of disagreements online versus offline (in face-to-face and writ-
ten interaction);
2. research exploring the role and relationship between social and medium factors (Her-
ring 2007) for the performance and emergent meanings of conflictual and consensual
disagreement;
3. an increased focus on sociable disagreement and play, as connected to focus 2;
4. a focus on language and gender with respect to the performance of disagreement online;
5. an increased emphasis on disagreement in dialogic and polylogic interactions, often in
connection with questions of participation frameworks;
6. an upsurge in research on disagreement in educational/pedagogical contexts; and
7. increased incorporation of data from different languages.
These foci emerge in scholarship on various e-genres, including computer con-

ferencing (Reid and Hards 1998), email (Tanskanen and Karhukorpi 2008; Haugh
2010), chat (Ong 2011; Vandergriff 2013), blogs and microblogging (Bolander
2012, 2013; Luzón 2013; Anderson and Cermele 2014; Zhang and Kramarae
2014; Dayter 2016), Youtube (Reynolds 2011; Bou-Franch and Blitvich 2014),
online newspapers (Lewis 2005; Langlotz and Locher 2012; Angouri and Wodak
2014), newsgroups (Baym 1996), Facebook (Vraga et al. 2015), online courses
and learning environments (Peters and Swanson 2004; Lapadat 2007; Monteserin,
Schiaffino and Amandi 2010; Lu, Chiu and Law 2011; Maíz-Arévalo 2014), and
fora and discussion lists/boards (Hert 1997; Bomberger 2004; Guiller and Durn-
dell 2006; Graham 2007, 2008; Perelmutter 2010; Angouri and Tseliga 2010; Lan-
done 2012; Kádár, Haugh and Chang 2013; Shum and Lee 2013). In what follows,
these seven main foci will be illustrated with a selection of studies drawn from
these and other works.
3.1. The comparative study of disagreement online versus offline

(in face-to-face and written interaction)
Comparisons between disagreement online and offline go back to a general inter-
est in the relative similarities and differences of computer-mediated communica-
tion and offline communication prominent in early sociolinguistic research. Here
we review the first systematic, empirical study of “Agreements and disagreements
in a computer-mediated discussion” by Baym (1996), who sets out to compare
the use, nature and meanings of disagreements and agreements in a Usenet news-
group with disagreements and agreements in letters and face-to-face interactions.
For Baym (1996: 320), “Usenet interaction is a hybrid between oral, written,
interpersonal, and mass communication”; and its hybridity is important for the
meanings and realisation of disagreement and agreement. To explore the effects
of hybridity on these moves, Baym (1996) draws comparisons with Pomerantz’s
(1984) work on the relative preference/dispreference of agreements and disagree-
ments in casual face-to-face interaction, and with Mulkay’s (1985, 1986) research

Authenticated
on disagreements and agreements in letter exchanges on scholarly scientific

topics.
Based on a qualitative analysis of 70 agreements and 51 disagreements in a
Usenet newsgroup storyline (consisting of a total of 542 messages about the soap
opera All My Children), Baym (1996: 321) underscores salient similarities and
differences with respect to both the characteristics of the moves, as well as their
functions. A key similarity between online and offline contexts is that disagree-
ments tend to be “more complex and mitigated” (Baym 1996: 332) than agree-
ments. Mitigation is achieved through various strategies including the use of partial
dis/agreement prefaces (like but, though), the acknowledgment of the other par-
ty’s perspectives, and qualifications which “frame disagreements as resulting from
differences in subjective opinion, thus leaving room for the other’s viewpoint”
(Baym 1996: 337). Differences in the realisation of disagreements and agreements
in Usenet stem from both medium and social factors (Herring 2007), i.e., from the
technological affordances of the Usenet system coupled with the context framing
the interaction and participant characteristics, norms, relationships and goals.
With respect to technology, Baym (1996) highlights the role played by “quo-
tation”. Since the Usenet interface allows participants to signal responsiveness
via quotation, this becomes the predominant means participants use to mark what
message, or more often part of a message, participants are dis/agreeing with.
While disagreements are more complex than agreements, both disagreement and
agreement moves tended to contain “pervasive elaboration”, “reasoning” and
“qualification” (Baym 1996: 339). Baym (1996: 340) explains this by referring to
the characteristics of the audience, with “[e]laboration [emerging …] as a way to
increase a message’s interest value for a mass audience, meeting a wide reader-
ship’s needs while demonstrating one’s own competence at doing so”. In addition
to quoting and audience, she also underscores norms, topic, purpose, community
characteristics and gender as factors important to the form and functions of disa-
greements and agreements in her data.
3.2. The role and relationship between social and medium factors
In their study of two online communities of practice (CofP), Angouri and Tseliga
(2010) similarly highlight that a range of factors influence the performance of
disagreement and impoliteness. Drawing on data collected from two asynchro-
nous fora (Greek students and professional academics) and follow-up interviews,
Angouri and Tseliga (2010: 65) demonstrate the intrinsic role played by norms and
expectations for the performance and evaluation of “potentially face-aggravating
acts”, including “inappropriate and impolite language use”. By combining obser-
vation with interviews – participants were asked to “talk through” relevant threads
and postings “where disagreement is explicitly marked” (Angouri and Tseliga
2010: 65) – they are able to explore practices as well as emergent meanings. While

Authenticated
the authors rightfully caution from assuming that this provides them with access to
participant intentions at the time of the interaction, the interview data enables them
to “gather […] a more in-depth description of the context of interaction in general
and the escalation of disagreement in particular” (Angouri and Tseliga 2010: 66).
In addition to stressing community expectations, their research highlights the
social factors of participant identity and topic (with political topics, for example,
being linked to the prevalence of conflict). It also points to the role played by
the medium, notably the “lack of paralinguistic cues” (Angouri and Tseliga 2010).
In their data, the lack of paralinguistic cues becomes tied to the prevalence of
“[u]nconventional spelling and punctuation” (Angouri and Tseliga 2010: 77). As
“a means of accentuating emotions”, these spellings and punctuation become a
way to express “strong disagreement” and “aggravate face-threatening acts” (Ang-
ouri and Tseliga 2010: 77). Often, particularly in the forum to which students con-
tributed, the unconventional spelling and punctuation co-occur with the discourse
particle [re] (untranslatable), which functions as a means to “initiate disagree-
ment” (Angouri and Tseliga 2010: 77).
Other research underscoring the variable roles played by particular medium
and/or social factors includes Bolander’s (2012, 2013) study of disagreement, con-
flict and other discourse moves in a corpus of personal/diary blogs, and Luzón’s
(2013) research on conflict and identity in academic blog discussions. Importantly,
a comparison of this work shows striking differences with respect to the realisation
of disagreement, which is seldom mitigated in Luzón’s (2013) data, but tends to be
dispreferred in Bolander’s (2012, 2013) data. This may be linked to the fact that
the blogs belong to different sub-genres. Luzón’s (2013) academic blogs prompted
critical exchange of viewpoints as a norm, while Bolander’s diary blogs were more
oriented towards bonding and consensus. While both scholars acknowledge vari-
ous factors besides genre for the use of disagreement and conflict, this difference
in sub-genre also draws attention to the importance of going beyond a view of a
mode or a genre as per se similar to recognising possibilities for internal differ-
ences.
As indicated by the discussion of these works, the upsurge of systematic, empir-
ical research on disagreement online was also accompanied by the recognition that
the realisation and meaning of disagreement cannot be determined a priori, or on
the basis of looking only at form. Rather, ascertaining the meaning and function of
disagreement is contingent upon taking into account a whole range of social and
medium factors (Herring 2007), which have been shown to influence language
use online, including disagreement (cf. particularly Tables 6 and 7 in Bolander
2013: 57, 71). Of particular relevance here are norms and expectations. This is in
accordance with our claims about conflictual and consensual disagreement made
in Sections 1 and 2. By exploring the range of factors influencing disagreement
online, scholars have progressively highlighted the prominence of both conflict-
ual and consensual disagreement. Thus, while Baym (1996), for example, draws

Authenticated
attention to the fact that disagreements tend to be mitigated relative to agreement,

she also underscores that they are positively evaluated and seen as compatible with
the Usenet newsgroup’s goals. Indeed, since “[d]isagreement invariably involves
the contribution of different interpretive resources, [it] thus facilitates the group’s
primary goal” (Baym 1996: 341). In this sense, disagreements are both conflictual,
face-threatening or potentially face-threatening, as suggested by the strong pres-
ence of mitigation, and consensual, by being conducive to community formation
and maintenance.
3.3. Sociable disagreement and play

Recent research on social media has also begun to study sociable disagreement,
conceptualised here as a particular kind of consensual disagreement. The notion
of sociable disagreement can be traced back to research conducted in the early
1980s (cf. e.g., Tannen 1981; Schiffrin 1984). Put forward to account for and draw
attention to cross-cultural differences in the social meaning of disagreement, such
research draws attention to the possibility that discourse which has the form of
(strong) disagreement is not necessarily face-threatening to its participants. It can
instead be reflective of particular norms and relationships between participants
which render the apparent conflict sociable. As stated by Schiffrin (1984: 331),
“sociable argument” constitutes “a speech activity in which a polarizing form has
a ratificatory meaning”. Introduced by Schiffrin (1984) in connection with her
work on Jewish Americans in Philadelphia, sociable disagreement accounts for the
function of typically strong disagreements as resources and markers of solidarity
(Schiffrin 1984: 332). In addition to highlighting cultural differences in the rela-
tive preference of disagreements and agreements (without homogeneising across
cultures), “sociable argument” underscores the key role played by participant rela-
tionships. Sociable disagreements “seem designed to show that the interactants’
relationship is close enough to withstand what would be considered by outsiders
to be verbal assaults” (Schiffrin 1984: 331).
While demonstrating that disagreement and conflict can take on different forms
and functions, research on language use in social media has only recently engaged
with sociable disagreement. In doing so, it has extended and adapted the notion
to online settings, highlighting the role of sociability, participant relationships,
humour, mitigation and play, in particular. Perhaps the earliest of these studies is
Weber’s (2011) research on the socialisation of newcomers into a Usenet news-
group for sexual abuse survivors. Drawing on extensive data (over 2000 messages)
obtained during 1.5 years of regular observation and archiving, Weber hones in
on a dispute which took place during two weeks and which involved a total of 16
participants and around 50 messages. Part of the dispute involves two “regular”
participants, who seemingly engage in a conflictual disagreement. Yet as Weber
(2011) demonstrates, the participants draw on a variety of resources to mark the

Authenticated
dispute as non-threatening. These include “statements which refer to or imply a

durative relationship between the two combatants”, “the use of emoticons which
key playfulness throughout the argument” and “the presence of politeness behav-
iors and the absence of flame-like behaviors in messages posted in the same thread
by and to other participants” (Weber 2011). Based on a close examination of these
cues, Weber (2011) underscores that such “play” is unmarked and emerges as a
frame shaping the way these two participants interact with one another, although
it does not constitute an expected practice for the group as a whole. In this sense,
the example underscores the co-presence of consensual (sociable) as well as con-
flictual disagreement.
Bolander (2012, 2013) also draws attention to the role played by participant
relationships for the enactment of sociable disagreement in her study of eight per-
sonal/diary blogs. Here, too, the sociable disagreement involves two participants,
who use the comments section of a blog, written by a mutual friend, to play and
banter with one another. Drawing explicitly on Schiffrin’s (1984) notion of “socia-
ble argument”, Bolander (2012, 2013) explores the ways these participants make
manifest that their relationship is not fleeting (for example through references to
a shared offline workspace), while highlighting the creative use of paralinguistic
features for the performance of the sociable argument, and the presence of “bond-
ing humour”, which serves to mark the “disagreement as sociable and to index an
existing (close) relationship” (Bolander 2013: 67). For Bolander (2012, 2013),
the sociable disagreement exchange is especially salient because it gives rise to
reader-reader interaction, an otherwise less typical interactional frame in the blog
corpus (as discussed in Section 3.4 below).
Ardington (2013) adds to this discussion by underscoring the fluid relationship
between sociable and conflictual disagreement. Her focus is a collaborative online
report produced by staff members of a large Australian metropolitan university
who felt threatened by the Vice Chancellor’s video message announcing staff cuts,
and an ensuing multiparty asynchronous forum discussion. She is thereby particu-
larly interested in exploring the “relational aspects of conflict talk for this social
group (academic community of practice) in text-based CMC from an interactional
sociolinguistic perspective” (Ardington 2013: 170). Using the last paragraph of
the report and posts from the asynchronous discussion as data, Ardington (2013)
highlights a shift from sociable to conflictual disagreement in the course of the
discussion. The early stages of the discussion are characterised as sociable: “posts
reflect a playful aggressive frame” (Ardington 2013: 174) and are co-constructed
by posters (Ardington 2013: 174), who also take “enjoyment” (Ardington 2013:
179) from the creative mocking and critique of the Vice Chancellor and other
senior management personnel. Here humour plays a key role, notably “abbrevi-
ated address forms, mitigated forms, complimenting actions, irony, metaphor and
recycling of others’ words” (Ardington 2013: 185). Yet in the course of the discus-
sion, the frame shifts from a “playful” to a “serious, more adversarial frame where

Authenticated
co-present participants become direct targets of the talk” (Ardington 2013: 181).
This shift is marked by an increase in personal criticism – with both “direct and
indirect personal attacks” (Ardington 2013: 183) gaining in currency. Since those
critiqued are expected to respond, to “justify, defend, or counterattack”, the result
is an overall “spiraling of conflict” (Ardington 2013: 184).
This move from sociable to conflictual disagreement indicates that there is
often no straightforward means of deciding when “sociable argument” becomes
“serious conflict” (Ardington 2013: 172). While this point has been made for
research offline, Ardington’s (2013) contribution extends the conversation to
social media. When reviewing literature for this chapter, we too became aware of
the fuzzy boundaries between conflictual and consensual disagreement (includ-
ing sociable disagreement). While there are differences between these forms, “the
boundary between play and non-play is a contested and difficult space” (Arding-
ton 2013: 172). Addressing how this boundary is marked and enacted is reliant
on close empirical analysis of disagreement, particularly of latent and emergent
norms and frames, or expectations. Going beyond notions of preference and dis-
preference, this foregrounding of norms allows for and highlights the importance
of the parallel inclusion of different types of disagreement.
3.4. Language and gender

A further topic which has received widespread attention in research on disagree-
ment online is language and gender. It, too, stems from an interest in comparisons
between offline and online settings for the pragmatics of disagreement in various
modes of computer-mediated communication (cf. also Section 3.1). Fuelled by the
question of whether the Internet would bring about more gender parity compared
with offline environments, early research on language and gender online explored
differences between male and female communicative styles with respect to issues
of power and dominance (cf. e.g., Herring 1993, 2001). It then progressed to more
specific analyses of types of differences, including a focus on conflictual disa-
greement and flaming. Herring’s (1994) paper “Politeness in computer culture:
when women thank and men flame”, for example, is based on the assumption that
men and women not only behave differently online, but also differ with respect to
“ideas of what constitutes appropriate and inappropriate behavior on the net” (Her-
ring 1994: 278). Drawing on an analysis of nine discussion lists with varying per-
centages of female subscribers, and an anonymous survey on net etiquette, Herring
(1994: 291) argues that whereas women value “politeness”, men adopt an “ethic of
anarchic self-determination and vigorous debate”. As Herring (2003: 207) main-
tains, research on language and gender online highlights differences between male
and female users of CMC with respect to “verbosity, aggressiveness, use of pro-
fanity, politeness (and rudeness), typed representations of smiling and laughter,
and degree of interactive engagement”. Similar claims regarding differences in

Authenticated
male and female linguistic behaviour online have been made by other scholars, for
example, by Savicki (1996) with respect to Internet discussion groups, Arnold and
Miller (1999) for home pages, and Guiller and Durndell (2006) for educational
online discussion groups.
Paralleling a more general development in sociolinguistics (cf., e.g., Eckert
and McConnell-Ginet 1992), Rodino (1997) argues for the need to shift from a
view of gender online as a “stable category” towards one which acknowledges its
perpetual construction. She thereby critiques research which reifies “binary gen-
der”, calling instead for the exploration of the multitude of ways “gender is expe-
rienced and exhibited” in various contexts and through differing practices (Rodino
1997). In doing so, she does not reject the possibility for participants to “reflect
and recreate gender bifurcation” (Rodino 1997). As her work on communication in
Internet Relay Chat (IRC) makes manifest, her participants present and construct
gender in various ways, including both “stable representations” and “contradictory
performances” (Rodino 1997; see also Huffaker and Calvert 2005). While Rodino
(1997) does not deal with disagreement and conflict, her arguments clearly have
implications for the study of these moves with respect to gender online.
However, and as Rodino (1997) argues, this view of gender should not pre-
clude a focus on women’s oppression online. Thus, paralleling and largely also
motivating research is an interest in power relations and the continued enactment
of aggression against women online. As Anderson and Cermele (2014: 289) have
argued in a recent paper on “Public/private language aggression against women:
Tweeting rage and intimate partner violence”, verbal aggression against women
often “reproduce[s] a binary construction of gender that justifies women’s sub-
ordinate status”. To demonstrate the pervasiveness of verbal aggression against
women, they explore two different sources of data: a corpus of tweets (which
deal with a “critical analysis of gendered representations and the under-representa-
tion of female protagonists in video games” [@femfreq]; Anderson and Cermele
2014: 275) and a corpus of Civil Protection Order petitions (filed with the aim of
allowing victims of domestic abuse to seek protection from the civil courts). Yet
whereas overt verbal aggression which is gendered and which involves threats of
violence is the norm in the Civil Protection Order petitions, the “dominant pattern
of sexist verbal aggression in tweets makes use of subtle strategies that construct
women as weak, lesser, and other, while retaining the veneer of polite and reason-
able discourse” (Anderson and Cermele 2014: 289, emphasis in original; cf. also
Bou-Franch and Garcés-Conejos Blitvich 2014). These strategies include claims
that women are not as capable gamers as men, constructions of women as outsid-
ers (and thus less capable), language suggesting that @femfreq is weak, as well
as “projections” of sexism back onto the target (such that the target of abuse is
claimed to be acting in a sexist manner) (Anderson and Cermele 2014: 286–288).
As the authors underscore, both differences as well as the general prevalence of
aggression in the two sources of discourse can be linked to the ways the sets of

Authenticated
data are simultaneously public and private. Thus “tweets may be considered ‘pub-
lic’ speech that is assumed to be impersonal and protected by ideologies of free
speech, and private speech that is protected by the conventions that it is backstage
talk that should not be taken too seriously”; and the Protection Order petitions as
“public in that [they] become […] part of an official court record” while being
based on private encounters between partners (Anderson and Cermele 2014: 289,
emphasis in original). Encouraging a view which takes account of the complex
ways context and technology shape language use, this research demonstrates that
there is no tie between more covert, indirect forms of verbal aggression and level
of threat, such that both forms – overt and covert – serve to reinforce women’s
subordinate societal status (Anderson and Cermele 2014: 290).
3.5. Disagreements in dialogic and polylogic interactions

Recent research on disagreement has also progressively begun to explore perfor-
mance in connection with participation frameworks and broader questions of inter-
activity. This is linked to the shift to Web 2.0 (cf. Section 2), which has prompted
an upsurge in research literature on the complexity of participation and partici-
pation frameworks in social media (cf., e.g., the papers in Dynel and Chovanec
2015). As Chovanec and Dynel (2015) outline, such questions are linked to types
of participants, degrees and layers of participation, issues of audience, reach and
transmission, and (disappearing) boundaries between public and private. These are
all also pertinent issues for the study of disagreement.
Bolander’s (2012, 2013) research on language and power in eight personal/
diary blogs, for example, highlights ties between participation framework, interac-
tional patterns, and particular discourse moves, including disagreement and agree-
ment. With respect to the participation framework, quantitative and qualitative
analysis of 841 comments shows that bloggers emerge as dominant participants to
whom the majority of disagreements and agreements are directed, and who agree
and disagree with their readers (in responses to reader comments). Via a paral-
lel focus on how these disagreement and agreement turns are made responsive
(i.e., how interlocutors mark to whom they are directing their turn, cf. also the
discussion of Baym 1996 in Section 3.1), she highlights ties between participation
framework and types of responsiveness. Thus whereas readers tend to be more
explicit (e.g., naming, quoting, format tying) when directing disagreements and
agreements at other readers, when addressing the blogger they tend to rely on less
explicit markers of responsiveness. For Bolander (2012, 2013) this is reflective
of both the latent, structural properties of the blog, which give rise to an a pri-
ori division of rights and obligations between blogger and readers, and the emer-
gent participation framework, which reinforces the blogger’s privileged status, as
becomes apparent through the actual degree and emergence of interaction in the
comments.

Authenticated
For Luzón (2013: 118), whose research on academic blogs is discussed in

Section 3.2 above, conflict in blog discussions is constructed across exchanges
“with several interlocutors taking sides and responding to various comments in
the discussion”, a finding she links with the blog’s “medium affordances” which
facilitate “polylogues”. Research on online polylogues is typically traced back to
Marcoccia’s (2004) study of participation framework in three online newsgroups.
As Bolander (2012, 2013) argues, the emergence of polylogues in blogs is linked
to the types of discourse move that are used. In her data, particular moves, notably
disagreement, agreement and criticism, become associated with interactions going
beyond two or three turns and involving several participants. (Cf. also Reynolds
2011 for a discussion of how particular types of “uncontroversial questions” serve
to elicit certain answers in data on protests posted to YouTube.)
An exploration of polylogues is also central to Bou-Franch and Blitvich’s
(2014) research on conflictual disagreement in YouTube. Working with a corpus
of comments following a public service announcement on the topic of teenage
homosexuality (posted by a Spanish organisation), they explore the emergence and
development of conflict (beginning, middle, end/closing). Advanced for the study
of conflict in offline settings, existing models and taxonomies on stages of conflict
have not been applied to massive online polylogues. In YouTube, considerations
of anonymity and asynchronicity (the fact that messages are not exchanged in
realtime, but with a time lag), coupled with YouTube’s facilitating of multi-par-
ticipant interaction are central to understanding how conflict emerges, is made
meaningful and co-constructed (Bou-Franch and Blitvich 2014: 20, 21). By virtue
of the persistent status of text on YouTube, researchers are also provided with
“unprecedented access […] to the diachronic unfolding of conflict” (Bou-Franch
and Blitvich 2014: 21), which has not yet received systematic attention in studies
of language in social media (but see DuVal Smith 1999; Reid 1999). Combining
qualitative and quantitative methods, Bou-Franch and Blitvich (2014: 32) argue
for the importance of exploring both the synchronic and diachronic construction
of conflictual disagreement in YouTube. On the basis of their analysis they argue
for the need to problematize existing conceptualisations of “beginnings, mid-
dles and ends” (Bou-Franch and Blitvich 2014: 33). Rather than viewing these
stages as “reified constructs”, closer attention needs to be paid to the ways in
which multi-participation and identity facilitate the maintenance of a “continuous
middle”, and thus to the tendency for conflict to be “on-going and unresolved”
(Bou-Franch and Blitvich 2014: 33). To highlight the “multifunctionality of utter-
ances”, the “multi-sequentiality” of the conflict and the fact that conflict tends not
to be resolved, they propose an alternate, adapted model outlining how conflict
unfolds in massive polylogues (thereby building on previous research by Dobs and
Garcés-Conejos Blitvich 2013, and Bousfield 2007).
Questions of interactive patterns, participation and collaboration also feature
in Lewis’ (2005) study of argument in French political discussions, Taskanen and

Authenticated
Karhukorpi’s (2008) research on repair and the negotiation of affiliation in email,

Hodson-Champeon’s (2010) work on intertextuality in racially antagonistic dis-
course, Kleinke’s (2010) analysis of disagreements and “emergent networks” in
English and German-language public newsgroups, and Perelmutter’s (2010) study
of Russian complaint discourse and the performance of relational work across
conversations. Despite different foci, each of these papers points to the impor-
tance of empirically examining how particular technological affordances become
entangled with the ways participants use language to perform disagreement. By
highlighting the variety of ways participants use language creatively in an attempt
to achieve individual or group goals, they also underscore the need to move away
from a priori assumptions about straightforward ties between technological affor-
dances and resulting practices.
3.6. Research on disagreement in educational/pedagogical contexts

A further strand of research explores disagreement in learning environments
(Monteserin, Schiaffino and Amandi 2010; Roschelle et al. 2010; Lu, Chiu and
Law 2011; Chiu and Cowan 2012; Verecellone-Smith 2012; Vandergriff 2013;
Maíz-Arévalo 2014). Some of this work also has a cross-cultural focus, for exam-
ple, Maíz-Arévalo’s (2014) study of disagreement and pragmatic rules in Eng-
lish as a Lingua Franca discourse (cf. also Chiu and Cowan 2012). Choosing to
analyse disagreement by virtue of its typically face-threatening nature and “dis-
ruptive potential”, Maíz-Arévalo’s (2014) intercultural pragmatic approach aims
to ascertain how students from various cultural backgrounds and with different
degrees of proficiency in English enact disagreement. On the basis of an analysis
of asynchronous mediated discussion data, the research highlights that mitigated
disagreement is more prominent than strong disagreement, and that students with
a greater degree of proficiency in English draw on a wider range of strategies,
tending “to follow the same strategies as native speakers do in order to avoid face-
threat” (Maíz-Arévalo 2014: 220).
3.7. Incorporation of data from different languages

While research on disagreement has prioritised English-language data, as shown
in Sections 3.1–3.5, there seems to be a progressive focus on other languages (cf.,
e.g., Angouri and Tseliga 2010 for Greek; Perelmutter 2010 for Russian; Kleinke
2010 for German and English; Landone 2012 for Spanish; Enama 2014 for French
and English; and Kádár, Haugh and Chang 2013; Shum and Lee 2013; and Zhang
and Kramarae 2014 for different varieties of Chinese). As with the other themes
discussed in this section, this gradual inclusion of other languages and varities in
connection with disagreement parallels a broader shift in the study of language and
language use online. Marked explicitly by the publication of Danet and Herring’s

Authenticated
(2007) The Multilingual Internet, research has progressively done justice to the
fact that the Internet, while still dominated by English, is a multilingual space.
Thus, Kádár, Haugh and Chang (2013), for example, explore similarities and
differences in the ways Mainland Chinese and Taiwanese discussion board par-
ticipants variously enact verbal aggression and impoliteness. Focusing on nation-
alistic discourse, they examine “the role of perceptions of national identity and
national face in occasioning instances of aggression” (Kádár, Haugh and Chang
2013: 345), notably with respect to terms of address and reference. Combining
a quantitative and qualitative analysis of web-based text produced by Mainland
Chinese or Taiwanese, as well as discussion board data, the authors demonstrate
salient differences, with the former tending to use “associative identity practices”
(whereby pan identity claims are indexed; Kádár, Haugh and Chang 2013: 347)
and the latter “dissociative identity practices” (whereby distance to the nation is
underscored; Kádár, Haugh and Chang 2013: 348). When interacting with one
another, in other words, the Mainland Chinese are more likely to use the “ostensi-
bly polite forms tongbaomen (‘compatriots’) or xiongdijiemei (‘brothers and sis-
ters’)”, whereas the Taiwanese tend instead to employ “abusive terms, negative
metapragmatic discourse on the use of such pan-Chinese forms of address/ref-
erence, marked use of deferential forms, as well as sarcasm” (Kádár, Haugh and
Chang 2013: 367). Interpreting this difference with respect to “broader macro-dis-
courses on national identity”, the authors argue for the importance of studying face
not only at the “interpersonal”, but also at the “intergroup” level (Kádár, Haugh
and Chang 2013: 344).
4. Key issues: From the challenge of context to indexing emotion
The discussion in Section 3 highlights central developments in the study of disa-

greement in social media. In doing so, it draws attention to changes in technology
and approach, notably a shift to Web 2.0 and the progressive emphasis on the mul-
tifaceted nature of disagreement online, on conflictual and consensual disagree-
ment, and on performance and play. In this section, we review two issues which
have begun to receive more attention in studies of disagreement: context as related
to the complex relationship between online and offline spaces; and the indexing of
emotion through written language in social media.
The first of these concerns the need for social media scholars to tackle how
context is shaped and brought into being. This includes thinking about context both
in terms of technological affordances as well as with respect to participants whose
use of technology is shaped by various factors linked to but also going beyond the
immediate affordances of the particular site in question. As Jones (2004) main-
tains, issues of context are intricately linked to questions of the border between
“offline” and “online” and more specifically to its usefulness. In his (2004) paper

Authenticated
on “The problem of context in computer-mediated communication”, he argues that

there has been a tendency for research on computer-mediated communication to
treat interaction as if it “takes place in a kind of virtual vacuum with little connec-
tion to the material worlds of the people sitting in front of computer screens and
producing the words that analysts spend so much time dissecting and interpreting”
(Jones 2004: 21). If we acknowledge that the context for social media interactions
does not exist a priori, but is enacted and made into context by participants, the
boundary between offline and online/virtual and real begins to collapse (Jones
2004; cf. also Barton and Lee 2013; Bolander and Locher 2014).
For the study of language use and disagreement online, problematizing context
has both epistemological as well as methodological implications. Thinking about
disagreement, it means considering the possibility that disagreement can be real-
ised across modes of social media as well as across “lines” (offline/online), while
critically reflecting on the ways different sources of data can help answer various
research questions pertaining to the use and emergent meanings of disagreement.
While much research on these issues still needs to be done, Sections 2 and 3 show
scholars considering the interplay between social and medium factors, as well as
drawing on mixed methods and the use of different sources of data to explore dis-
agreement from multiple angles.
As shown in Angouri and Tseliga (2010: 66), for example, using interview data
in addition to the analysis of asynchronous fora data provides greater insight into
“the context of interaction”, and thus into how disagreement and impoliteness are
viewed by participants themselves. It allows them, in other words, to more fully
explore what these moves mean and how they are evaluated; and thus to reflect
upon the relationship between evaluations and practices. Similarly Bolander
(2012, 2013) works with interview data collected from bloggers on a range of
topics (pertaining, for example, to language use, and the bloggers’ roles and rela-
tionships with their readers), as a means to complement her discourse analysis of
the pragmatics of disagreement, conflict and other moves in blog posts and com-
ments.
More recently, Niemi (2014) highlights the fuzzy boundaries between online
and offline in her analysis of children’s classroom disputes and threats. Using
Goffman’s (1974) concept of frame and conversation analysis, she explores the
ways two young boys (best friends in the same class at school) shift between
different “pretend frames” – from video games as well as “real life” – in enacting
a dispute. Although the video game frames cannot be acted on in the classroom
setting, by drawing on them as a resource the boys bring them into the classroom.
This “video game activity”, in other words, “can be ‘done’ in the classroom”, a
finding Niemi (2014: 119) takes as indicative of occurrences in the video game
world influencing what happens in “the real world”, and vice versa. For the prag-
matics of disagreement, this article draws attention to the importance of consid-
ering the possibility that offline sites can also constitute or rather become, if we

Authenticated
think of context as something which is enacted and emergent, a key space for the
study of online language use and practices.
Returning to online spaces, context has also become more complex for inter-
action in social media as a result of what Marwick and boyd (2010: 122) term
“context collapse”. This notion refers to the fact that many types of social media,
notably social networking sites, “flatten […] multiple audiences into one”. Despite
the possibility to manage privacy by creating different accounts for different audi-
ences, context collapse means “participants must contend with groups of people
they do not normally bring together, such as acquaintances, friends, co-workers,
and family” (Marwick and boyd 2010: 122). As a result, users strategically employ
resources in connection with impression management, and as a means to navi-
gate between the “portray[al … of] an authentic self and an interesting personal-
ity” (Marwick and boyd 2010: 122). While typically explored in connection with
the construction and performance of online identity, and more recently language
choice and multilingualism (Androutsopoulos 2014), Vraga et al. (2015) extend
the discussion to disagreement online. Focusing on Facebook, they combine inter-
view with survey data of young adults to explore ties between the likelihood for
conflictual political disagreement on Facebook and variation in participants’ “per-
ceptions of the political climate on Facebook” (Vraga et al. 2015: 282). In address-
ing these questions, they pay close attention to the way audience heterogeneity (in
connection with “context collapse”) emerges as relevant for individual perceptions
of political disagreement on Facebook. While their focus is not pragmatic, their
results have implications for issues of how disagreements are realised, notably
their finding that perceptions of political disagreement on Facebook are strongly
linked to tone, as well as to the participants’ Facebook network structures and size
(Vraga et a. 2015: 288).
Another strand of recent research has begun to explore how emotion in disa-
greement is cued in written language. In doing so, such research demonstrates that
social media is not impoverished relative to offline interaction (cf. Section 2 for
a discussion of the cues-filtered out approach). Langlotz and Locher (2012), for
example, study ties between emotion, disagreement and relational work via a quan-
titative and qualitative analysis of 120 English postings from MailOnline. They
thereby demonstrate that emotion can be indexed via “expression” (including both
linguistic means like “name calling” and “verbalisation of emotional reaction”,
as well as graphic means like “smileys”), via “implication” (where implicature is
used to index emotional stance) and via “description” (through the use of “emotion
words” and “verbal descriptions/ascriptions of verbal states”; Langlotz and Locher
2012: 1600–1601). As these examples highlight, the possibility to index emotional
stance is varied, despite it being either linguistic or graphic (as opposed to visual
and/or aural).
In addition to highlighting participant creativity with respect to the use of writ-
ten language for emotional stance, this research demonstrates the polysemy of

Authenticated
emotive cues. As outlined, for example, by Vandergriff in her (2013: 1) article on

conflictual chat discourse, the meanings of CMC cues are highly context depend-
ent. Focusing on the use of cues amongst advanced foreign language learners of
German tasked with negotiating consensus on a moral dilemma on chat (Vander-
griff 2013: 2, 3), she argues for a move away from trying to map offline emotive
cues onto online ones, towards an exploration of “how CMC participants adopt the
semiotic system of chat to their communicative needs” (Vandergriff 2013: 2). Cen-
tral to her argument is the finding that “form-meaning correlations of CMC cues
only emerge in context” (Vandergriff 2013: 8), a result which is compatible with
Langlotz and Locher’s (2012) results discussed above. She thereby also demon-
strates that these cues tend not to emerge alone, but in interaction with other cues,
both verbal and nonverbal, in contexts of emotive communication. By highlighting
the ways emotion is indexed online through written language (linguistic, graphic,
paralinguistic), both Langlotz and Locher (2012) and Vandergriff (2013) remind
us of the importance of not bracketing out or ignoring the multifacetedness of writ-
ten language in light of the increased multimodality of social media.
5. Conclusions
This chapter begins by underscoring that there is no straightforward relationship

between form and function, such that looking solely at the linguistic manifestation
of disagreement cannot be taken as grounds for ascertaining what disagreement
might mean to participants involved in interaction in social media. Coupled with a
view of language as always informed by both informational and relational proper-
ties, we underscore the importance of paying close attention to how disagreements
are realised, to evidence of whether they are expected, and to possible implications
for face. The importance of treating disagreement as multifaceted, of recognising
that there is no straightforward form-function relationship, and that meanings of
disagreement can shift in the course of a single interaction are thus all points which
emerge from our review of disagreement. In Section 2 of the paper, we problema-
tize these issues via a discussion of a shift in epistemology in research on language
use online, via a focus on Web 2.0 environments, and via a cautious differentiation
between conflictual and consensual (including sociable) disagreement. The largest
section of the paper, Section 3, is then devoted to exploring seven foci highlighted
as pertinent in the literature we reviewed on disagreement in social media: the
comparative study of disagreements online versus offline (in face-to-face and writ-
ten interaction); research exploring the role and relationship between social and
medium factors (Herring 2007) for the performance and emergent meanings of
conflictual and consensual disagreement; an increased focus on sociable disagree-
ment and play; research on language and gender with respect to the performance of
disagreement online; an increased emphasis on disagreement in dialogic and poly-

Authenticated
logic interactions, often in connection with participation frameworks; an upsurge

in research on disagreement in educational/pedagogical contexts; and incorpora-
tion of data from various languages as part of enhanced recognition of the web as
a multilingual space.
Subsequently we turned to two issues we feel are of import for scholars inter-
ested in disagreement in social media: context as related to the multifarious rela-
tionship between online and offline spaces, and the indexing of emotion through
written language in social media. In many ways, these issues also constitute our
outlook. The first prompts for further teasing out of the implications of increas-
ingly blurring boundaries within different forms of social media as well as across
lines (online/offline) for pragmatic research on particular practices, including but
not limited to disagreement. And the second calls for increased studies on modal-
ity, via further analysis of the ways disagreement can be enacted through different
modes, along with careful consideration of how these modal affordances offer
additional layers for reflecting on social and medium factors shaping language use.
Acknowledgments
Many thanks to Aline Bieri for her assistance in compiling literature and address-
ing many style issues. For feedback on this chapter, we thank Wolfram Bublitz and
Christian Hoffmann. All remaining faults are our own.
References
2006 Introduction: Sociolinguistics and computer-mediated communication. Jour-
nal of Sociolinguistics 10(4): 419–438. http://www.blackwell-synergy.com/
doi/abs/10.1111/j.1467-9841.2006.00286.x.
2010 Localizing the global on the participatory web. In: Nikolas Coupland (ed.),
The Handbook of Language and Globalisation, 203–231. Malden, MA:
Wiley-Blackwell.
2014 Languaging when contexts collapse: Audience design in social networking.
Discourse, Context and Media 4–5: 62–73.
Anderson, Kristin L. and Jill Cermele
2014 Public/private language aggression against women: Tweeting rage and inti-
mate partner violence. Journal of Language Aggression and Conflict 2(2):
274–293. doi: 10.1075/jlac.2.2.05and.
Angouri, Jo and Miriam A. Locher
2010 Disagreement on and off line. Paper presented at the Fifth International Sym-
posium on Politeness, Basel.

Authenticated

2010 “you HAVE NO IDEA WHAT YOU ARE TALKING ABOUT!” From e-disa-
greement to e-impoliteness in two online fora. Journal of Politeness Research
6(1): 57–82.
Angouri, Jo and Ruth Wodak
2014 ‘They became big in the shadow of the crisis’: The Greek success story and the
rise of the far right. Discourse and Society 25(4): 540–565.
Ardington, Angela
2013 Negotiating shared perspectives that move in and out of sociability: Play and
aggression in technologically mediated communication. Journal of Language
Aggression and Conflict 1(2): 165–193. doi: 10.1075/jlac.1.2.03ard.
Arnold, Jill and Hugh Miller
1999 Gender and web home pages. Paper presented at the CAL99 Virtuality in Edu-
cation Conference, London. http://ess.ntu.ac.uk/miller/cyberpsych/cal99.htm.
Barton, David and Carmen Lee
2013 Language Online. Investigating Digital Texts and Practices. London: Rout-
ledge.
Baym, Nancy
Bolander, Brook
responsiveness. Journal of Pragmatics 44: 1607–1622.
Bolander, Brook
2013 Language and Power in Blogs: Interaction, Disagreements and Agreements
methodological issues. Discourse, Context and Media 3: 14–26. doi: http://
dx.doi.org/10.1016/j.dcm.2013.10.004.
Bomberger, Ann M.
2004 Ranting about race: Crushed eggshells in computer-mediated communication.
Computers and Composition: An International Journal for Teachers of Writ-
ing 21(2): 197–216. doi: 10.1016/j.compcom.2004.02.001.
Journal of Pragmatics 73: 19–36. doi: 10.1016/j.pragma.2014.05.001.
Carnevale, Peter and Tahira Probst
1997 Conflict on the internet. In: Sara B. Kiesler (ed.), Culture of the Internet, 233–
255. Mahwah, NJ: Lawrence Erlbaum.
Chiu, YiChing Jean and John Cowan
2012 Helping Eastern students to master Western critical thinking. Asia Pacific
Education Review 13(1): 103–111. doi: 10.1007/s12564-011-9189-2.
media. In: Marta Dynel and Jan Chovanec (eds.), Participation in public and
social media interactions, 1–23. Amsterdam/Philadelphia: Benjamins.

Authenticated
Collins, Mauri
1992 Flaming: The relationship between social context cues and uninhibited verbal
behavior in computer-mediated communication. http://www.mediensprache.
net/archiv/pubs/2842.htm.
Danet, Brenda and Susan Herring (eds.)
2007 The Multilingual Internet: Language, Culture, and Communication Online.
New York: Oxford University Press.
Dayter, Daria
2016 Discursive Self in Microblogging: Speech Acts, Stories and Self-praise
DuVal Smith, Anna
1999 Problems of conflict management in virtual communities. In: Marc Smith and
Peter Kollock (eds.), Communities in Cyberspace, 134–163. New York: Rout-
ledge.
Dynel, Marta
2015 The landscape of impoliteness research. Journal of Politeness Research 11(2):
329–354.
Dynel, Marta and Jan Chovanec (eds.)
2015 Participation in Public and Social Media Interactions. Amsterdam/Philadel-
phia: Benjamins.
Eckert, Penelope and Sally McConnell-Ginet
1992 Communities of practice: where language, gender, and power all live. In: Kira
Hall, Mary Bucholtz and Birch Moonwomon (eds.), Locating Power. Proceed-
ings of the Second Berkeley Women and Language Conference, 89–99. Berke-
ley, CA: Women and Language Group.
Enama, Patrick Rodrigue Belibi
2014 Information Structure, Agreement and Disagreement in English and French
Asynchronous Online Discussion. MA Thesis. Albany: State University of
New York at Albany.
1967 Interaction Ritual: Essays on Face-to-face Behavior. Garden City, NY: Anchor
Books.
Goffman, Erving
1974 Frame Analysis. An Essay on the Organization of Experience. Cambridge,
2008 A manual for (im)politeness?: The impact of the FAQ in electronic commu-
nities of practice. In: Derek Bousfield and Miriam A. Locher (eds.), Impolite-
ness in Language. Studies on its Interplay with Power in Theory and Practice,
Guiller, Jane and Alan Durndell
2006 ‘I totally agree with you’: Gender interactions in educational online discus-
sion groups. Journal of Computer Assisted Learning 22(5): 368–381. doi:
10.1111/j.1365-2729.2006.00184.x.

Authenticated
Hardaker, Claire
Haugh, Michael
tions of impoliteness. Journal of Politeness Research 6(1): 7–31. doi: 10.1515/
jplr.2010.002.
Herring Susan C.
1993 Gender and democracy in computer-mediated communication. Electronic
Journal of Communication 3(2). http://ella.slis.indiana.edu/~herring/ejc.doc.
Herring, Susan C.
1994 Politeness in computer culture: Why women thank and men flame. In: Mary
Bucholtz, A.C. Liang, Laurel A. Sutton and Caitlin Hines (eds.), Cultural Per-
formances: Proceedings of the Third Berkeley Women and Language Confer-
ence, 278–294. Berkely, CA: Berkeley Women and Language Group.
Herring, Susan C.
2001 Gender and power in online communication. Center for Social Informatics
Working Papers, no. WP-01-05. https://scholarworks.iu.edu/dspace/bitstream/
handle/2022/1024/WP01-05B.html?sequence=1.
Herring, Susan C.
Internet 4. http://www.languageatinternet.de/articles/2007.
Hert, Philippe
1997 The dynamics of on-line interactions in a scholarly debate. The Information
Society 3(4): 329–360.
Hodsdon-Champeon, Connie
2010 Conversations within conversations: Intertextuality in racially antagonistic
online discourse. Language@Internet 7(10). http://nbn-resolving.de/urn:n-
bn:de:0009-7-28206.
Homes, Janet and Stephanie Schnurr
2005 Politeness, humor and gender in the workplace: Negotiating norms and iden-
tifying contestation. Journal of Politeness Research 1(1). doi: https://doi.org/
10.1515/jplr.2005.1.1.121.
Huffaker, David A. and Sandra L. Calvert
2005 Gender, identity, and language use in teenage blogs. Journal of Computer-Me-
diated Communication 10(2). doi: 10.1111/j.1083–6101.2005.tb00238.x.
Jones, Rodney
2004 The problem of context in computer-mediated communication. In: Philip LeV-
ine and Ron Scollon (eds.), Discourse and Technology: Multimodal Discourse
Analysis, 20–33. Washington D.C.: Georgetown University Press.
Kádár Dániel, Z., Michael Haugh and Melody Chang Wei-Lin
2013 Aggression and perceived national face threats in Mainland Chinese and Tai-
wanese CMC discussion boards. Multilingua 32(3): 343–372. doi: 10.1515/
multi-2013-0016.
Kakavá, Christina
1993 Negotiation of Disagreement by Greeks in Conversations and Classroom Dis-
course. Doctoral Thesis, Washington D.C.: Georgetown University.

Authenticated

Psychologist 39(10): 1123–1134. doi: 10.1037/0003-066x.39.10.1123.
Kleinke, Sonja
2010 Interactive aspects of computer-mediated communication: ‘Disagreement’ in
an English and a German public news group. In: Sanna-Kaisa Tanskanen, Mar-
ja-Liisa Helasvuo, Marjut Johansson and Mia Raitaniemi (eds.), Discourses in
Landone, Elena
2012 Discourse markers and politeness in a digital forum in Spanish. Journal of
Pragmatics 44(13): 1799–1820. doi: 10.1016/j.pragma.2012.09.001.
Pragmatics 44(12): 1591–1606.
Lapadat, Judith C.
2007 Discourse devices used to establish community, incerase coherence, and nego-
tiate agreement in an online university course. Journal of Distance Education
21(3): 59–92.
Lewis, Diana M.
2005 Arguing in English and French asynchronous online discussion. Journal of
Pragmatics 37(11): 1801–1818. doi: 10.1016/j.pragma.2005.02.014.
Locher, Miriam A.
2004 Power and Politeness in Action: Disagreements in Oral Communication. Ber-
Locher, Miriam A.
2006 Advice Online. Advice-giving in an American Internet Health Column. Amster-
Locher, Miriam A. and Sage L. Graham
2010 Introduction to Interpersonal Pragmatics. In: Miriam A. Locher and Sage
Lambert Graham (eds.), Interpersonal Pragmatics, 1–13. Berlin/Boston: de
Gruyter.
9–33.
2008 Relational work and impoliteness: Negotiating norms of linguistic behaviour.
In: Derek Bousfield and Miriam A. Locher (eds.), Impoliteness in Language.
Studies on its Interplay with Power in Theory and Practice, 77–99. Berlin/
Boston: de Gruyter.
Lu, Jingyan, Ming Ming Chiu and Nancy WaiYing Law
2011 Collaborative argumentation and justifications: A statistical discourse analysis
of online discussions. Computers in Human Behavior 27(2): 946–955. doi:
10.1016/j.chb.2010.11.021.
2013 “This is an erroneous argument”: Conflict in academic blog discussions. Dis-
course, Context and Media 2(2): 111–119. doi: 10.1016/j.dcm.2013.04.005.

Authenticated
2014 Expressing disagreement in English as a lingua franca: Whose pragmatic
rules? Intercultural Pragmatics 11(2): 199–224. doi: 10.1515/ip-2014-0009.
Marcoccia, Michel
internet newsgroups. Journal of Pragmatics 36(1): 115–145. doi: 10.1016/
S0378-2166(03)00038-9.
Maricic, Ibolya
2005 Face in Cyberspace: Facework, (Im)Politeness and Conflict in English Dis-
cussion Groups. Växjö: Växjö University Press.
Marwick, Alice and Danah Boyd
2011, first published online on July 7, 2010 I tweet honestly, I tweet passion-
ately: Twitter users, context collapse, and the imagined audience. New Media
and Society 13(1): 114–133. http://nms.sagepub.com/content/early/2010/06/2
2/1461444810365313.
Monteserin, Ariel, Silvia Schiaffino and Analía Amandi
2010 Assisting students with argumentation plans when solving problems in CSCL.
Computers and Education 54(2): 416–426. doi: 10.1016/j.compedu.2009.
08.025.
Mulkay, Michael
1985 Agreement and disagreement in conversations and letters. Text 5: 201–227.
Mulkay, Michael
1986 Conversations and texts. Human Studies 9: 301–321.
Muntigl, Peter and William Turnbull
1998 Conversational structure and facework in arguing. Journal of Pragmatics
29(3): 225–256.
Niemi, Kreeta
2014 “I will send badass viruses.” Peer threats and the interplay of pretend frames
in a classroom dispute. Journal of Pragmatics 66: 106–121. doi: 10.1016/j.
pragma.2014.02.012.
Ong, Keng Wee
2011 Disagreement, confusion, disapproval, turn elicitation and floor holding:
Actions as accomplished by ellipsis marks-only turns and blank turns in qua-
sisynchronous chats. Discourse Studies 13(2): 211–234.
Perelmutter, Renee
2010 Impoliteness recycled: Subject ellipsis in Modern Russian complaint
discourse. Journal of Pragmatics 42(12): 3214–3231. doi: 10.1016/j.
pragma.2010.07.005.
Peters, Brad and Diana Swanson
2004 Queering the conflicts: What LGBT students can teach us in the classroom and
online. Computers and Composition 21(3): 295–313. doi: 10.1016/j.comp-
com.2004.05.004.
Pomerantz, Anita
1984 Agreeing and disagreeing with assessments: some features of preferred/ dis-
preferred turn shapes. In: J. Maxwell Atkinson and John Heritage (eds.), Struc-
tures of Social Action, 57–101. Cambridge: Cambridge University Press.

Authenticated
Reid, Elizatbeth
1999 Hierarchy and power. Social control in cyberspace. In: Marc A. Smith and
Peter Kollock (eds.), Communities in Cyberspace, 107–133. London: Rout-
ledge.
Reid, Fraser J. M. and Rachael Hards
1998 The effects of time scarcity on conflict and compromise in computer confer-
encing. Computers in Human Behavior 14(4): 637–656. doi: 10.1016/s0747-
5632(98)00028-4.
Reynolds, Edward
2011 Enticing a challengeable in arguments: Sequence, epistemics and preference
organisation. Pragmatics 21(3): 411–430. doi: 10.1075/prag.21.3.06rey.
Rodino, Michelle
1997 Breaking out of binaries: Reconceptualizing gender and its relationship to lan-
guage in computer-mediated communication. Journal of Computer-Mediated
Communication 3 (3). http://jcmc.indiana.edu/vol3/issue3/rodino.html.
Roschelle, Jeremy, Ken Rafanan, Ruchi Bhanot, Gucci Estrella, Bill Penuel, Miguel Nuss-
baum and Susana Claro
2010 Scaffolding group explanation and feedback with handheld technology:
Impact on students’ mathematics learning. Educational Technology Research
and Development 58(4): 399–419. doi: 10.1007/s11423-009-9142-9.
Savicki, Victor
1996 Gender language style and group composition in internet discussion groups.
Journal of Computer-Mediated Communication 2(3). http://jcmc.indiana.edu/
vol2/issue3/savicki.html.
Schiffrin, Deborah
1984 Jewish Argument as Sociability. Language in Society 13: 311–335.
Schneider, Stefan
2010 Mitigation. In: Miriam A. Locher and Sage L. Graham (eds.), Interpersonal
Shum, Winnie and Cynthia Lee
2013 (Im)politeness and disagreement in two Hong Kong Internet discussion forums.
Journal of Pragmatics 50(1): 52–83. doi: 10.1016/j.pragma.2013.01.010.
Spencer-Oatey, Helen
2000 Rapport management: A framework for analysis. In: Helen Spencer-Oatey
(ed.), Culturally Speaking: Managing Rapport through Talk across Cultures,
11–46. London: Continuum.
Sproull, Lee and Sara Kiesler
1986 Reducing social context cues: Electronic mail in organizational communica-
tion. Management Science 32(11): 1492–1512.
Tannen, Deborah
1981 New York Jewish conversational style. International Journal of the Sociology
of Language 30: 133–149.
Tannen, Deborah
1993 What’s in a frame?: Surface evidence for underlying expectations. In: Deborah
Tannen (ed.), Framing in Discourse, 14–56. Oxford: Oxford University Press.
Pragmatics 40(9): 1587–1600.

Authenticated
Turnage, Anna
2007 Email flaming behaviours and organizational conflict. Journal of Comput-
er-Mediated Communication 13(1): article 3. http://jcmc.indiana.edu/vol13/
issue1/turnage.html.
Vandergriff, Ilona
communication (CMC) cues. Journal of Pragmatics 51: 1–12. doi: 10.1016/j.
pragma.2013.02.008.
Vercellone-Smith, Pamela, Kathryn Jablokow and Curtis Friedel
2012 Characterizing communication networks in a web-based classroom: Cogni-
tive styles and linguistic behavior of self-organizing groups in online discus-
sions. Computers and Education 59(2): 222–235. doi: 10.1016/j.compedu.
2012.01.006.
Vraga, Emily K., Kjerstin Thorson, Neta Kligler-Vilenchik and Emily Gee
2015 How individual sensitivities to disagreement shape youth political expression
on Facebook. Computers in Human Behavior 45: 281–289. doi: 10.1016/j.
chb.2014.12.025.
Weber, H. L.
2011 Missed cues: How disputes can socialize virtual newcomers. Language@
Internet 8. http://www.languageatinternet.org/articles/2011/Weber.
Yus, Francisco
Zappavigna, Michele
iation on the Web. London: Continuum.
sion. Journal of Pragmatics 73: 66–81. doi:10.1016/j.pragma.2014.07.008.

Authenticated
23. Compliments and compliment responses
María Elena Placencia and Amanda Lower
Abstract: Compliments and the responses that they elicit have been widely stud-
ied in face-to-face interactions. Researchers are now turning to digital contexts,
particularly social media, where complimenting is a widespread occurrence. This
chapter gives an overview of the current state of the art of compliment and compli-
ment response studies on social media. After a brief exploration of the background
of compliment studies in face-to-face interaction, we provide a working definition
of compliments. We then examine studies to date of complimenting behavior on
social media, highlighting similarities and differences, and any emerging trends.
Next, methodological and ethical considerations are explored in this relatively
new area of study. Finally, we comment on directions that future research could
take.
1. Introduction and background
This chapter focuses on research on complimenting behavior in social media.

Complimenting is a commonplace social practice in many languages and sociocul-
tural contexts. For the past four decades, it has received a great deal of attention in
pragmatics research in relation to face-to-face interaction (cf. Chen 2010). Devel-
opments in technology and computer-mediated communication with Web 2.0
capabilities in particular have in recent years opened up new spaces for the study
of complimenting behavior (and other social practices) online. Indeed, compli-
menting and the related ‘likeing’ (or equivalent) action (cf. Maíz-Arévalo, Ch. 21,
this volume) in response to posting pictures on different social networking sites
(SNSs) seem to have become pervasive features of online social interaction. Pic-
tures, as suggested by Placencia and Lower (2013: 618) with respect to Facebook,
act as a kind of summons (Schegloff 1972 [1968]) inviting a response that can
take the form of a ‘like’ and/or of a comment which can be a compliment. Emer-
gent research (Section 3), while still incipient, is beginning to uncover the ways
in which complimenting behavior has been adapted to suit the online medium of
interaction and the ways some features of technology are exerting influence on
its realization. Traditional frameworks of analysis are having to be adjusted to
account for new phenomena in complimenting behavior online.
Interest in the linguistic study of complimenting behavior can be traced back to
a number of key works from the 1970s and 1980s that have been highly influential
and that are also exerting influence in current research on complimenting behavior
DOI 10.1515/9783110431070-023
Authenticated
634 María Elena Placencia and Amanda Lower
online. They include, among others, Manes and Wolfson’s ethnographic research
on compliments in American English (Wolfson and Manes 1980; Manes and
Wolfson 1981; Wolfson 1981; Manes 1983; Wolfson 1983); Pomerantz’s (1978)
conversation analytic work on compliment responses also in American English,
and Brown and Levinson’s (1987 [1978]) as well as Leech’s (1983) theories of
politeness. Manes and Wolfson adopted Hymes’s (1972 [1964]) ethnography of
speaking perspective and sought to identify the rules of speaking behind compli-
menting behavior among middle-class Americans. This involved finding out who
compliments whom, the form compliments take, the functions compliments fulfill,
as well as the topics that are the object of complimenting. Manes and Wolfson’s
research agenda, extensively explored in face-to-face contexts as we will see, has
acquired renewed interest in the study of complimenting behavior in virtual envi-
ronments.
Pomerantz (1978), in turn, attempted to explain from a conversation analytic
perspective how Americans go about responding to compliments in relation to
a system of preferences (e.g. preference for agreement over disagreement and
acceptance over rejection). Agreeing with a compliment runs counter to the con-
straint that speakers are faced with of avoiding self-praise. Speakers thus need to
employ a number of strategies to deal with this type of constraint. One of them is
agreeing with, but down-grading, the compliment. Pomerantz’s work has served as
the basis for various taxonomies that have been applied to numerous languages and
sociocultural contexts, Herbert’s (1986, 1989, 1990) and Holmes’s (1986) being
the most influential ones. However, the new forms of communication offered by
social media, with features such as multimodality, asynchronous communication,
and anonymity in some contexts, appear to be altering the way compliments are
responded to online (if at all), with factors linked to the medium and technology
influencing complimenting response behavior (see Section 3.2).
Finally, with respect to politeness theory, the interpersonal function of compli-
ments highlighted in Brown and Levinson’s (1987 [1978]) model (cf. Graham, Ch.
17, this volume), and other works has been and continues to be a topic of interest
in the characterization of complimenting behavior. According to Brown and Lev-
inson, for example, compliments appear to constitute a type of positive politeness
strategy aimed at attending to people’s needs for approval and appreciation. Com-
pliments fit nicely under their category, “claim ‘common ground’ (1987 [1978]:
102); however, Brown and Levinson note that they can also constitute face-threat-
ening acts in certain contexts (1987 [1978]: 66). Authors such as Kerbrat-Orec-
chioni (1997) regard compliments as intrinsically face-flattering, while Sifianou
(2001: 397) suggests considering them as gifts and, as such, links them to offers
and Brown and Levinson’s (1987: 102) strategy “give gifts to H (goods, sympathy,
understanding, cooperation)”. Jaworski (1995: 75), on the other hand, highlights
an instrumental function of compliments whereas Wolfson (1983: 91) emphasizes
their solidarity-creating/maintaining function. However, she also acknowledges

Authenticated
Compliments and compliment responses 635
the multifunctionality of compliments. Identifying the functions of compliments

in social media is an undertaking that is beginning to surface too (see Section 3.1).
Additionally, under the influence of more recent politeness theories such as
Locher and Watts (2005), appropriateness in complimenting behavior online is a
research topic that is gaining interest. What is emerging is that norms from face-
to-face contexts do not necessarily apply online (Section 3). A discursive struggle
(Watts 2003) about what is considered appropriate behavior can be observed in
some social media interactions and discussions. For example, a debate that has
surfaced on Twitter in recent years relates to the acceptability of re-tweeting the
compliments a person receives. For some it is a form of bragging that is unaccept-
able and should be sanctioned, but others disagree (cf. Kiefer Lee 21 Aug 2012;
Herstand 20 Aug. 2013). Additionally, from time to time, Twitter is one of the
battlegrounds for debates on gender practices and sexism behind complimenting
behavior in social networking sites and professional sites such as LinkedIn (see,
for example, Proudman 2015).
Complimenting behavior has also been examined from the perspective of other
research traditions, in addition to the ones considered above. They include cor-
pus linguistics (Jucker et al. 2008); ethnopragmatics (cf. Goddard and Wierzbicka
2014); relevance theory (cf. Ruhi and Doğan 2001) and variational pragmatics (cf.
Schneider 1999). With one or two exceptions, works from these perspectives have
not had as much influence on research on compliments online as the classic works
mentioned above.
Before considering the main findings of research on complimenting behavior
in SNSs available to date (Section 3), in Section 2 below, we offer a working
definition of compliments. Methodological and ethical issues are considered in
Section 4. Finally, in Section 5, some conclusions are provided and prospects for
the future are sketched out.
2. Compliments: A working definition
We adopt a broad definition of compliments as expressions of positive evaluation

(Wolfson 1981: 120) that attribute credit to the addressee (Holmes 1986: 492).
Prototypical expressions of positive evaluation examined in research on compli-
menting behavior are of the verbal type (e.g. “I really like your new kitchen!”).
However, this notion has been more recently extended to cover other phenomena
such as appreciatory sounds (Golato 2011) or (de)gustatory expressions (Gardner
2001), like mmmh, commonly used as compliments in assessments of food and
drink in face-to-face interaction in some sociocultural contexts. Along the same
lines, it is being extended to cover online phenomena such as the use of ‘like’ on
Facebook (cf. Placencia and Lower 2013) or equivalent functions (see Section 3)
and visual means of communication such as emojis (cf. Placencia 2015).

Authenticated
The definition of compliments as expressions of positive evaluation may seem

straightforward; however, determining what is a ‘positive evaluation’ can be prob-
lematic. Implicit compliments may go unnoticed, as it can be difficult to identify
them (cf. Boyle 2000; Sifianou 2001; Maíz-Arévalo 2012). On the other hand, what
appears to be a direct compliment in terms of its structure and lexis may not func-
tion as such in a given context. In face-to-face interaction compliments typically
form part of an adjacency pair; as such, the response or second pair-part can help
the analyst determine whether a given utterance was interpreted as a compliment
or not. However, this is not a tool that analysts can rely on in the examination of
compliments on SNSs like Facebook given that the absence of a reply on these sites
is not an uncommon phenomenon (see Section 3). Likewise, in the examination
of face-to-face interaction, prosodic features can come to the aid of the analyst. In
online interaction, these features are absent; however, the use of ‘prosodic spell-
ings’ (Androutsopoulos 2000: 521) such as capital letters and multiple exclamation
marks as well as emoticons (cf. Yus 2005), for example, can provide clues about
how a given utterance should be interpreted in an online context. Co-occurring
verbal forms such as interjections, address terms and expressions of affection can
perform the same function both online and offline (see Blum-Kulka, House and
Kasper 1989 and their notion of supportive moves). There are times, nonetheless,
when the value of certain utterances cannot be established as, for example, when
the utterances in question relate to in-group jokes or other in-group usage.
Yet another difficulty is that in some contexts it may not be easy to distinguish
compliments from other related communicative actions such as congratulations,
with which they share a pragmatic space (Jucker and Taavitsainen 2000: 92). In
most characterizations of illocutionary verbs, compliments are classified under
the category of expressives (Searle 1976), alongside congratulations, apologies,
and so forth (cf. Vanderveken 1990; Haverkate 1994). Interestingly though, for
Bach and Harnish (1979), for example, compliments constitute a subcategory of
congratulations, which, in turn, correspond to their category of acknowledgements
– acts that express “certain feelings toward the hearer” (1979: 51).
All in all, it can be problematic to identify compliments in naturally occur-
ring discourse. Some of these problems are shared across media; others are medi-
um-specific. However, there are resources such as compliment responses and/or
co-occurring elements that analysts can make use of to guide them in their cat-
egorization process. Using an independent rater can also help deal with unclear
cases as can interviews or focus groups discussions, for example, with members
of a particular SNS, although these options may not always be available. In any
case, we agree with Jucker et al.’s (2008: 292) observation that “[t]here is always a
subjective element in interpretation” and that “it is the context that is the deciding
factor”. In other words, careful attention needs to be paid to the local context of
the interaction, the nature of the activity in which the compliment is embedded,
and the co-text. The socio-cultural context needs to be taken into account too since

Authenticated
what is positive for one group may not be for another (see also below), and there
may even be variation across socio-cultural groups in terms of “what counts as a
compliment” (Wolfson 1981: 117). For instance, Achugar (2002) examined how
piropos, or compliments with an amorous or sexual tone, were interpreted by older
and younger Uruguayan women of the same socio-economic background. On the
whole she found that more older women, compared to younger ones, interpreted
piropos as positive evaluations.
3. Research on complimenting behavior in social media
There is a dearth of research on complimenting behavior in social media. With

respect to compliments, as can be seen in Table 1 below, works available exam-
ine compliments on Facebook among Americans (Placencia and Lower 2013)
and Spaniards (Maíz-Arévalo and García-Gómez 2013); on Orkut, among Ben-
galis (Das 2010); on Instagram, among Ecuadorians (Placencia 2015); in a virtual
region within Second Life (SL)1 where English is used as a lingua franca (Cirillo
forthcoming); on Twitter among Malaysians (Siti Yuhaida and Tan 2014), and in
blogs (Hoffmann 2013).2
Table 1. Available studies on compliments, presented according to language (variety)

examined and site, with basic details of the corpus employed
Language (variety Social media Author(s) Corpus employed

where known) site
American English Facebook Placencia and 1057 compliments;
Lower (2013) 1346 ‘likes’
Bengali Orkut Das (2010) 110 compliments from 79
dyads
British English and Facebook Maíz-Arévalo and 100 compliments
Peninsular Spanish García-Gómez (50 BrEng; 50 PenSp)
(2013)
1
Virtual worlds like SL may not be immediately associated with social media. They seem
to constitute a hybrid category in that they provide a ‘community’ space for socializa-
tion among SL members, albeit interaction is through digital personas in the form of
avatars, with elements of gaming.
2
We limit our overview to studies that focus on complimenting behaviour – either com-
plimenting or responding to compliments. There are a few studies such as Black et al.
(2015) where compliments are mentioned as a strategy in use, but are not dealt with in
detail.

Authenticated
Language (variety Social media Author(s) Corpus employed

where known) site
English Personal and Hoffmann (2013) 199 compliments
corporate blogs (100 from personal blogs;
99 from corporate blogs)
Ecuadorian Spanish Instagram Placencia (2015) 411 compliments
English as a lingua Second Life vir- Cirillo 74 compliments
franca tual world game (forthcoming)
(Malaysian) English Twitter Siti Yuhaida and 220 compliments
and Malay Tan (2014)
Concerning compliment responses (Table 2), Placencia, Lower and Powell (2016)
examine interactions on Facebook among Americans. Maíz-Arévalo (2013) also
looks at Facebook interactions but among Spaniards, whereas Eslami, Jabbari and
Kuo (2015) analyze interactions in Persian among Iranian users of Facebook. Siti
Yuhaida and Tan (2014), mentioned above, study responses to compliments in
(Malaysian) English and Malay on Twitter. Finally, Cirillo (2012) looks at com-
pliment responses in SL, with her study seemingly constituting the first attempt to
look at complimenting behavior in a virtual environment, where English is used
as a lingua franca..
Table 2. Available studies on compliment responses, presented according to language

(variety) examined and site, with brief details of the corpus employed
Language (variety) Social media Author(s) Corpus employed

site
American English Facebook Placencia, Lower Responses to
and Powell (2016) 1057 compliments
English as a lingua SL virtual Cirillo (2012) Responses to
franca world game 74 compliments
(Malaysian) English Twitter Siti Yuhaida and Responses to
and Malay Tan (2014) 220 compliments
Peninsular Spanish Facebook Maíz-Arévalo Responses to
(2013) 177 compliments
Persian Facebook Eslami, Jabbari Responses to
and Kuo (2015) 497 compliments

Authenticated
In the following sections we look at some features of complimenting behavior

across studies, sketching out commonalities and general patterns, as well as point-
ing out some differences in approach or results.
3.1. Some features of compliment realization in social media

Most authors examining compliment realization in social media take as a base-
line results from classic studies such as Manes and Wolfson (1981) on American
English (see Section 1) or Holmes (1986) on New Zealand English, exploring
potential variation regarding the structure of compliments, topics and functions of
compliments, who compliments and so forth. Characteristics of the medium and/
or situational and other factors that appear to have an impact on the realization
of compliments in socio-digital environments are also considered in the different
studies available.
The structure of compliments

Manes and Wolfson (1981) found that compliments in their American English cor-
pus were highly formulaic: they were realized by mainly nine syntactic patterns,
with three categories, 1–3 below, accounting for 85 % percent of all the compli-
ments in their corpus.3
1. NP {is/looks} (really) ADJ (53.6 %)
2. I (really) {like/love} NP (16.1 %)
3. PRO is (really) (a) ADJ NP (14.9 %)
4. You V (a) (really) ADJ NP (3.3 %)
5. You V (NP) (really) ADV (2.7 %)
6. You have (a) (really) ADJ NP (2.4 %)
7. What (a) ADJ NP! (1.6 %)
8. ADJ NP! (1.6 %)
9. Isn’t NP ADJ! (1.0 %)
(Adapted from Manes and Wolfson 1981: 120–121)
This finding led scholars to examine the formulaicity of compliments initially in

face-to-face contexts and online in recent years. Most of Manes and Wolfson’s
(1981) patterns have been found in social media studies, with some similarities
but also differences in frequency of use. For example, Hoffmann (2013), who
analyzed explicit compliments in two types of blogs, found instances of patterns
1–8 in personal blogs, and 1–4 and 6–9 in corporate blogs. In turn, Placencia and
Lower (2013) found patterns 1–4 and 7–8 in their Facebook corpus correspond-
ing to direct compliments in American English. Concerning frequency of use, in
3
NP = noun phrase; really = any intensifier; ADJ = adjective; PRO = you and demonstra-
tive pronouns; V = verb; ADV = adverb

Authenticated
both studies (as in Manes and Wolfson’s 1981 work) patterns 1 and 2 were the
most frequently employed. On the other hand, Manes and Wolfson’s pattern 8, a
reduced form, was found to be more frequently used by both Hoffmann (2013)
and Placencia and Lower (2013). Additionally, in both studies a new category for
another reduced form was put forward: (ADV) ADJ (+PP) (Hoffmann 2013: 349)
and HOW ADJ! (Placencia and Lower 2013: 631).
Moreover, Placencia and Lower introduced subtypes of some of Manes and
Wolfson’s (1981) categories. This was in order to accommodate elliptical forms
that amounted to a variation on the original category. The authors attribute the
(higher) occurrence of these elliptical forms to the informal, written medium of
Facebook. Furthermore, Placencia and Lower identify indirect realizations of
compliments, that is, compliments that require going through an inferencing pro-
cess to be understood as such. They also tallied Facebook ‘likes’ as a subset of
compliments, albeit with their own characteristics. On account of the word like,
they suggest that ‘likes’ can be regarded as direct compliments. However, they
also note that these are opaque forms in that what is being liked is often not clear
(e.g.: Is it the content of the photo or the photographic skills of the user? The pho-
to’s composition? etc.). In any case, ‘likes’ were found to occur more frequently
than compliments, possibly because they provide “a simpler way to connect with
others” (2013: 15); they are more impersonal and thus less committal, and yet,
they still serve an important interactional function (see below).
Reduced forms also appear in Cirillo’s (forthcoming) study on compliment-
ing behavior in SL. While Manes and Wolfson’s (1981) pattern 1 above was also
the most common in her study, it was followed by pattern 8 and not pattern 2.
Cirillo highlights the nature of the chat in SL as requiring short formulations. She
observes that SL users “pay compliments to satisfy a social function, but they do
it with minimum effort” (forthcoming: 10). Interestingly, she found 15 additional
complimenting patterns and explains such heterogeneity in relation to the popu-
lation of SL users, which can be quite diverse in terms of gender, age, nationality,
etc.
In examining complimenting behavior in languages other than English, clas-
sifications developed for English may not necessarily be (entirely) suitable.
Maíz-Arévalo and García-Gómez (2013) use a different classification to describe
the structure of explicit compliments in their contrastive study of compliment-
ing behavior on Facebook between speakers of British English and Peninsular
Spanish. They distinguish three main types of constructions: exclamative, declar-
ative, and elliptical (2013: 743) and assert that these syntactic forms have inher-
ent meanings. Specifically, exclamative sentences are expressions of a speaker’s
emotions, and declarative sentences convey an expression of fact. Elliptical forms,
on the other hand, involve a “co-construction” process, which “helps strengthen
the rapport and solidarity between both interlocutors” (2013: 752). Their results
show that Spanish participants gave exclamative compliments more frequently

Authenticated
than British participants and that those compliments covered physical beauty and
intelligence, while the British exclamative compliments tended toward admiration
for possessions. In terms of declarative compliments, they found that full declar-
ative compliments on appearance occurred more frequently in the British group,
and full declarative compliments on personality occurred more frequently in the
Spanish group. Concerning elliptical compliments, the Spanish participants used
ellipsis more frequently to evaluate appearance, while the British participants used
ellipsis more frequently to evaluate possessions.
Placencia (2015) also employs a somewhat different classification for her
study of compliments among Ecuadorian teenage girls on solo shots on Instagram.
As in Placencia and Fuentes Rodríguez (2013), she identifies six main verbal com-
pliment categories. The first five apply to direct realizations: exclamations with
qué, declaratives with an evaluative noun/verb/adjective, elliptical forms, inter-
jections, and (semi-) fixed expressions. The sixth category corresponds to indi-
rect/implicit compliments with no fixed syntactic structure (see also Placencia
and Lower 2013). The first three of the direct categories are roughly similar to
Maíz-Arévalo and García-Gómez’s (2013) (see above). Placencia proposes two
additional types to account for nonverbal and hybrid realizations of compliments,
reflecting the use of multimodality: pictorial compliments realized by emojis4
occurring on their own (e.g. for ‘princess’ or for ‘perfect’) (2015: 12) or in
combination with other emojis, and hashtags which state the positive evaluation in
the hashtag (e.g. Dany #salesbienguapa tu ‘Dany #youlookgorgeous’) (2015: 10)
and link the recipient with multiple (sometimes hundreds of thousands) images
with the same hashtag thus offering amplified, polyphonic compliments. Ellipti-
cal forms in Placencia’s study constitute the most frequently employed category,
unlike in Placencia and Lower (2013), for example. Placencia tentatively suggests
that this result could be related to the age of the Instagram users in her study.
Exclamations with qué and declaratives are the second and third most frequently
employed categories in Placencia (2015). With respect to indirect/implicit forms
(2015: 10), they include, for example, the use of irony. Placencia (2015: 14) also
considers the use of internal modification such as prosodic spellings (Androut-
sopoulos 2000) and external modification or supportive moves (cf. Blum-Kulka,
House and Kasper 1989; Placencia and Lower 2013) that can take the form of
interjections and emojis, for example.
4
Placencia (2015: 7) employs the term pictorial for compliments realized through emo-
jis, taking into account the origin of the term emoji in Japanese: ‘e’ for picture + ‘moji’
for letter or character (Oxford Dictionaries, 2016). Emoticons, in contrast, are “iconic
combinations of punctuation marks like [:-)]” that typographically, rather than through
pictures, represent facial expressions, “typically communicat[ing] a broad kind of emo-
tion” (Yus 2005: 167).

Authenticated
Interestingly, reduced verbal forms for compliments observed in online inter-

action were present already in face-to-face interaction more than 30 years ago (cf.
Manes and Wolfson 1981). What seems to have changed is the range of reduced
forms available and their frequency of use. However, caution is needed in draw-
ing comparisons with results from old studies. For example, reduced forms high-
lighted in online studies may also be more common in current face-to-face inter-
action than they were 30 years ago.
Topics of compliments
According to Manes (1983), topics of compliments reflect cultural values. She
observes that “people compliment one another time after time on the same things:
personal appearance, new acquisitions, good work” (1983: 98). These are objects
or actions “which any member of the speech community will recognize as posi-
tive” (1983: 98). Manes, however, does not provide specific figures for the three
main categories that she describes for American English compliments: appear-
ance, possessions, and ability/skills, as opposed to Holmes (1986), who provides
such figures for New Zealand English, offering a baseline for subsequent studies.
Table 3 below lists Holmes’s categories and results as well as the categories and
results described in social media studies available.5
As this table shows, similar topics are the object of complimenting in social
media and face-to-face interaction according to Holmes (1986), albeit with some
variation in frequency of occurrence and in relation to site. Appearance is clearly
the most frequent compliment topic across all studies, followed by possessions in
Placencia and Lower (2013) and Siti Yuhaida and Tan (2014). Possessions is not a
relevant category in Cirillo’s (forthcoming) study where the appearance of avatars,
including their clothing, is the main object of praise. Personality is complimented
on across all studies, but stands out particularly in Siti Yuhaida and Tan (2014)
where celebrities, well known offline, are the recipients of compliments.
A difficulty in comparisons on compliment topics is that authors may group
some topics differently. For example, Holmes (1986) combines personality and
friendship under one category, whereas Cirillo (forthcoming) and Placencia and
Lower (2013) keep them as separate categories. Even the same label may not cover
the same aspects. Placencia and Lower (2013: 637), for instance, classify com-
pliments on the appearance of children and pets under appearance and not under
possessions as in other schemes.
5
Maíz-Arévalo and García-Gómez (2013) also look at compliment topics and list three
categories: Appearance, personality and skills, and possession; however, they are not
included in Table 3 below as they do not provide exact figures that can be used for
comparison.

Authenticated
Table 3. Distribution of compliments according to topic in studies available
Holmes Cirillo Placencia and Lower Siti Yuhaida and

(1986: 496) (forthcoming: 8) (2013: 637) Tan (2014: 86)
Appearance Appearance Appearance Appearance
(50.7 %) (76.81 %) (75 %) (32.7 %)
Ability/ Ability/Skills/ Ability Performance/
Performance Personal taste (7 %) Ability/Skills
(30.6 %) (8.7 %) (21.4 %)
Possessions – Possessions Possessions
(11.2 %) (8 %) (25.5 %)
Personality/ Personality Personality Personality
Friendship (5.8 %) (5 %) (20.5 %)
(4.8 %)
– Name – –
(5.8 %)
– Friendship Friendship –
(1.5 %) (4 %)
– Place – –
(1.5 %)
Other – Other –
(2.7 %) (1 %)
Functions of compliments
Wolfson and Manes (1980: 391) identify both discourse and social functions of
compliments. Compliments can be employed, for example, to initiate a conver-
sation or to change the topic of conversation. Socially, compliments are used as
a means of establishing or reaffirming common ground (1980: 395), creating “a
sense of rapport” (1980: 397), or as “social lubricants” (Wolfson 1983: 89), in lieu
of greetings or preceding criticism. The authors propose that the formulaic nature
of compliments is tied to their solidarity-creating/reaffirming function in that com-
pliments need to be easily recognized to be able to fulfill this function more easily.
They also note negative uses of compliments that may involve “direct manipula-
tion” (Manes 1983: 97) in order to encourage a particular desired behavior (see
also Holmes 1988 and Jaworski 1995).
Studies on compliments in social media also report on a range of discourse
and social functions that verbal compliments appear to fulfill. In the context of SL
interactions, Cirillo (forthcoming: 8) claims that compliments “function mainly as
conversation starters, attracting other users’ attention” and “getting into their good

Authenticated
books”. She underlines their rapport-building function in a context “where words

are at the very basis of every relation” (forthcoming: 8).
Placencia and Lower (2013) suggest that when a user posts a photograph on
Facebook, it acts as a sort of summons, inviting a response (see Section 1). In
this context, they regard the main function of compliments as that of “keeping
the communication channel open” and “affirming and strengthening relation-
ships” (2013: 639), though they also observe that compliments constitute one of
the means of constructing “the kind of mundane, friendly conversation” that Wolf-
son and Manes (1980: 397) refer to in their study of compliments in face-to-face
interaction (2013: 640).
Siti Yuhaida and Tan (2014: 85) identify various functions of compliments to
celebrities on Twitter. Express[ing] admiration is the most salient one, followed
by establish[ing] solidarity. It is not clear though how these two categories can
be distinguished easily from each other. Other minor functions that they describe
include, for example, open[ing] conversation and reinforc[ing] desired behavior
(2014: 85).
Hoffmann (2013), on the other hand, finds that compliments appear to play
different functions in personal as opposed to corporate blogs. In the former, their
function as social lubricants (2013: 351) as well as attention grabbers and conver-
sations starters (2013: 352) seems to stand out. In corporate blogs, other functions
can be discerned: compliments as disclaimers that mitigate commentators’ criti-
cism, a “persuasive use of compliments for some personal gain”, and non-authen-
tic comments with “an excess of explicit personal compliments to boost the online
reputation of the blogger” (2013: 352).
Concerning the function of medium-specific tools, namely ‘likes’ and emoti-
cons and emojis in the context of complimenting behavior studies, there has not
been as much discussion of their occurrence with/as compliment responses (Sec-
tion 3.2). Starting with ‘likes’ on Facebook, Placencia and Lower (2013: 633)
regard them as expressions of approval “of the hearer, his/her comments, posses-
sions, taste or skills, etc.”. They propose viewing them as a subset of compliments
(see above), and, on the whole, as “a kind of phatic affirmation, after Malinowski’s
(1972 [1923]) notion of phatic communion” (2013: 639) through which interper-
sonal bonds are created or strengthened.
With respect to emoticons and emojis, both Placencia and Lower (2013) as
well as Placencia (2015) observe their use as supportive moves (Blum-Kulka,
House and Kasper 1989), that is, as moves that accompany verbal compliments,
reinforcing them. Placencia and Lower (2013: 635) point out that emoticons help
clarify “that the compliment was well intended in case the recipient has a doubt
about the complimenter’s intentions” (2013: 636). This interpretation is in line
with Dresner and Herring’s (2010: 255) observation that emoticons function as
“indications of the illocutionary force of the textual utterances that they accom-
pany” (see also Section 3.2). Placencia (2015) observes that, generally speaking,

Authenticated
the emojis employed by the teenagers in her Instagram corpus appear to intensify
the compliments as well as display creativity and often add a touch of playfulness.
Finally, as pointed out above, Placencia (2015) highlights the use of stand-alone
emojis that function as pictorial compliments.
Compliments and gender

Wolfson’s (1983: 92) tentative claim that women appear to give compliments more
frequently than men do is corroborated by Placencia and Lower’s (2013) study.
While their corpus contained only compliments received by females, they ana-
lyzed compliments given by both sexes. Women gave the majority of compliments
(91.1 %), a pattern that is repeated in the use of the ‘like’ function: 84.1 % of
‘likes’ in the study were made by women (2013: 637). Wolfson (1983: 92) also
claimed that women appear to receive compliments more than men. Siti Yuhaida
and Tan’s (2014: 86) results corroborate this claim: from the 220 compliments in
their Twitter compliments, 159 were addressed to women and 61 to men. Finally,
with respect to the topic of compliments, Manes (1983: 98) observed that compli-
ments on appearance “typically involve women as speakers or addressees”. Siti
Yuhaida and Tan’s (2014: 86) finding that compliments on appearance were pro-
duced more frequently by females compared with males (37.1 % vs. 21.3 %) is
along the same line as Manes’s claim. All in all, the impact of gender as a macro
social factor on complimenting behavior in social media is an area deserving more
attention in future studies (see also Section 1). This should include examining not
only who compliments whom and on what, but also compliment formulation and
functions in same sex and cross-gender interactions.
Compliments and social distance

Das (2010) examines Bengali speakers’ use of compliments, greetings, and expres-
sions of gratitude on Orkut, a now defunct social network, in terms of Wolfson’s
(1988) bulge theory. This theory suggests that dyads at the extreme ends of social
distance engage in fewer politeness acts than those in the middle social distances.
In this study, Das described the intensity of compliments on Orkut used between
intimate friends, friends, and acquaintances residing in a Midwestern United States
university town. For the bulge theory to hold true, fewer compliments should have
occurred between intimate friends and acquaintances, as these are at extreme ends
of social distance. Most compliments should have occurred among friends, who
occupy the middle ground of social distance. Das found that while intensified com-
pliments did follow the bulge pattern, unmarked compliments did not. Unmarked
compliments were given most often to intimate friends, and least to acquaint-
ances. Das suggests that this may be because intimate friends give compliments to
each other through other media; unmarked compliments then become appropriate
between intimate friends. He also assumes that although unmarked compliments
are by nature less intense than intensified compliments, they are still compliments,

Authenticated
and it is logical that people would give more compliments to people with whom
they share a closer relationship. The influence of microsocial factors such as social
distance in complimenting behavior in social media, as in Das’s (2010) study, is
another area that merits further exploration. Its importance surfaces in Placencia,
Lower and Powell (2016) in relation to compliment responses (see below).
3.2. Features of compliment responses in social media

Along with studies of compliments on social media, studies of responses to those
compliments are also emerging. The few studies available suggest that it is a rich,
burgeoning area that presents researchers with unique issues to address, related in
some cases to the affordances and limitations of technology. These issues include
how to classify compliment responses online, whether a response is relevant, how
users respond to compliments with social-media features only, and how social
media-specific features are used to respond to compliments.
Taxonomies of compliment responses

As in studies on compliment responses in a face-to-face setting, researchers of
compliment responses on social media look for the taxonomy or scheme that can
best account for the responses in their studies. The two most frequently used tax-
onomies in face-to-face interactions, Holmes (1986) and Herbert (1986, 1990),
have been used in research of compliment responses on social media. At present,
the number of studies following Holmes (1986) and Herbert (1986, 1990) seems
to be fairly evenly split: Placencia, Lower and Powell (2016) and Maíz-Arévalo
(2013) adopt Holmes (1986), as do Eslami, Jabbari and Kuo (2015) via Maíz-
Arévalo (2013), while Cirillo (2012) and Siti Yuhaida and Tan (2014) follow Her-
bert (1990). However, grouping studies into either followers of Holmes (1986)
or Herbert (1986, 1990) is not so straightforward: their face-to-face taxonomies
are adopted to varying degrees by different authors. As shown in Tables 4, 5 and
6 below, the authors that follow Holmes’s (1986) taxonomy have modified it in
order to fit their data. Applying an existing taxonomy to new data collected over
new mediums is open to interpretation. Different researchers have therefore fitted
similar data differently into the same taxonomy. Italics in the tables below denote
departures from Holmes’s original taxonomy. What stands out is that Holmes’s
original global strategies (accept, reject, evade/deflect) are present in all of the
modified versions. The differences arise in where to place the affordances of social
media technology, particularly how to classify the ‘like’ function and emoticons.

Authenticated
Table 4. Comparison of the compliment response taxonomies by Holmes (1986) and

Placencia, Lower and Powell (2016)
Holmes (1986) Placencia, Lower and Powell (2016)

Accept Accept
– Appreciation/agreement token – Appreciation/agreement token
– Agreeing utterance ‘Like’ + comment
–
– Downgrade/qualifying utterance Emoticon + comment
–
– Return compliment – Agreeing utterance
‘Like’ (without comment)
–
Emoticon (without comment)
–
– Downgrade/qualifying utterance
– Return compliment
Reject Reject
– Disagreeing utterance – Disagreeing utterance
– Question accuracy – Challenge sincerity
– Challenge sincerity
Deflect/Evade Deflect/Evade
– Shift credit – Shift credit
– Informative comment – Informative comment
– Ignore – Ignore6
– Legitimate evasion – Legitimate evasion
– Request reassurance/repetition – Request reassurance/repetition
No acknowledgement
Table 5: Maíz-Arévalo’s (2013) taxonomy of compliment responses

Explicit (written) response Implicit response No
response
Accept Reject Evade Emoticons ‘Like’
– Appreciation – Question – Shift credit function
– Agreement accuracy – Informative
– Downgrade – Challenge comment
– Enhance sincerity – Ignore
– Return
6
Ignoring a compliment is replying to a compliment with a comment, which neither
relates to the compliment nor explicitly answers another part of the conversation. Con-
trast this with no acknowledgement, which is no comment whatsoever.

Authenticated
Table 6. Eslami, Jabbari and Kuo’s (2015) taxonomy of compliment responses
Response No response
Verbal Non-verbal
– Accept ‘Like’
–
– Reject ‘Like’ +
–
– Evade emoticon
– Combination Emoticon
–
As the tables above demonstrate, compliment response taxonomies diverge on

the classification of new methods of responding to compliments. In other words,
unique elements of digital communication, such as ‘likes’ and emoticons, are
being handled in different ways. For example, while Placencia, Lower and Pow-
ell (2016) focus on the value of response utterances, including ‘likes’ and emoti-
cons as communicating acceptance, rejection or deflection/evasion, Maíz-Arévalo
(2013) re-categorizes these three main global response strategies as explicit and
proposes the category of implicit for ‘likes’ and emoticons. A difficulty with this
distinction is that verbal responses can also be formulated by implicit means, just
like compliments (see above); that is, restricting the category of ‘implicit’ to ‘likes’
and ‘emoticons’ is problematic. Also, Maíz-Arévalo’s categorization implies that
it not possible to attribute any value at all to responses that take the form of a ‘like’
or an emoticon in relation to Holmes’s global strategies. For Placencia, Lower
and Powell (2016), ‘like’ constitutes a positive evaluation even if conveyed in a
vague way, thus suggesting acceptance of the compliment. This interpretation was
confirmed by a group of college Facebook users that they interviewed. Placencia,
Lower and Powell (2016: 358) argue that ‘like’ could reasonably be interpreted as
a token of appreciation or an agreement. They interpret emoticons such as smileys
in a similar way (see also below).
Cirillo (2012) uses a slightly modified version of Herbert’s (1986) taxon-
omy. She breaks down Herbert’s (1986) categories into further sub-categories.
For instance, in the case of comment history, she distinguishes between agreeing
responses that are strictly comment history and those that are a combination of
comment history plus some other strategy (2012: 49). She also identifies a new,
non-agreement strategy that is not separately classified in Holmes’s (1986) study,
namely irony (2012: 50).
Siti Yuhaida and Tan (2014) also use Herbert’s taxonomy, though not its 1986
version but the modified 1990 version. They note that some types of responses
occurring in face-to-face interaction are not present in their corpus: they found
no incidences of praise upgrade, qualification, or requests for more information.
These absences seem to be explainable by the corpus examined: Siti Yuhaida and
Tan’s (2014) corpus consists of interactions with celebrities. It is unsurprising that
celebrities, as public figures, did not attempt to upgrade their compliments, nor did

Authenticated
they qualify or request more information about compliments received. This satis-
fies modesty impositions, which may be even more important for public figures.
Frequency and distribution of compliment responses

What is striking when comparing results across studies is that there appear to be
very few similarities. The only discernible similarity is that reject constitutes the
least observed response across the five studies available (see Table 7). Evade is the
second least used strategy across works, except for Cirillo’s (2012) study, where
no response occurred less frequently than evade.
Table 7. Distribution of compliment responses
Placencia Maíz- Eslami Cirillo Siti Yuhai-

et al. Arévalo et al. (2012) da and Tan
(2016) (2013) (2015) (2014)
Accept 15 % 19.7 % 79.7 % 54.1 % 77.27 %

Reject 1% 1.1 % 1% 4.1 % 1.81 %
Evade 3% 7.9 % 2.4 % 24.2 % 18.61 %
No response 81 % 30 % 16 % 17.4 % 2.27 %
Implicit response – 41.3 % – – –
Combination – – 0.6 % – –
However, as Table 7 shows, there is a great deal of variability concerning no

response and other categories. Placencia, Lower and Powell (2016) document an
81 % no response rate for their American English Facebook study. This figure
stands in stark contrast to the figures reported by Maíz-Arévalo (2013) in her work
on Facebook responses by Peninsular Spanish speakers (only 30 %) or Siti Yuhaida
and Tan’s (2014) study of responses by Malaysian celebrities on Twitter (2.27 %).
Concerning the latter, the low figure may be explained by taking into account that
celebrities may feel obliged to respond to fans as otherwise they may be perceived
as conceited or ungrateful. With respect to the other two studies, there could be
a gender element impacting on response behavior: Placencia, Lower and Pow-
ell (2016) only focus on responses by a group of women whereas Maíz-Arévalo
(2013) looks at responses from both males and females. It could also be attributed
to cultural differences although this would need further exploration. However, one
cannot discount the fact that the use of different data collection methodologies may
have also had an impact on the results. In any case, Placencia, Lower and Powell’s
(2016) results above are in line with their interview data findings that suggest that

Authenticated
a lack of response is generally not seen as problematic by college Facebook users

in their study and that responses may even be viewed negatively in some contexts.
Concerning the rest of the categories, and given the variation in classifications
used, it is more difficult to make direct comparisons. Nonetheless, Eslami, Jabbari
and Kuo’s (2015) as well as Maíz-Arévalo’s (2013) results, both on Facebook but
in different sociocultural contexts, are comparable as the former largely based
their categorization on the latter. It is striking that Eslami, Jabbari and Kuo (2015)
shows a much higher rate of acceptance than Maíz-Arévelo (2013), namely 79.7 %
vs. 19.7 %. It is possible that a good part of Maíz-Arévalo’s responses under the
category of implicit (i.e., 41.3 %) could be interpreted as representing acceptance.
This seems a likely explanation of the difference. Eslami, Jabbari and Kuo (2015)
classify instances of the ‘like’ function alone, or in combination with emoticons, as
an acceptance of a compliment (2015: 255–256). As such, Maíz-Arévalo’s figure
for acceptance could possibly be (much) higher. Overall, these results show the
need for further studies on compliment responses and for further discussion on
how medium-specific elements such as ‘likes’ (see also below) should be inter-
preted and classified. This requires the use of a combination of methods.
Social media-specific forms of compliment response

As mentioned above, social media provides tools for responding to compliments
that are not part of face-to-face interactions: emoticons/emoji and the ‘like’ func-
tion. How to interpret these affordances of social media technology is up for debate.
As pointed out in the previous section, Dresner and Herring (2010) highlight the
supportive function of emoticons. The studies discussed above demonstrate that
it is often the case that emoticons accompany a compliment or a compliment
response, thus lending credibility to Dresner and Herring’s (2010) view. They
assert that stand-alone emoticons can either be classified as “expressions of emo-
tion that map iconically onto body movements, such as smiling or frowning” or as
performances of illocutionary acts (2010: 258). The authors seem to suggest that
meaning can be assigned to emoticons. This is precisely what Placencia, Lower
and Powell (2016) propose when they interpret stand-alone emoticons such as smi-
leys as compliment acceptance tokens. Therefore, there does not appear to be the
need to create another global category to be added to Holmes’s (1986) taxonomy.
Cirillo (2012) and Siti Yuhaida and Tan (2014) used Herbert’s (1986, 1990)
taxonomies and did not add new categories to deal with emoticons. They do not
discuss how they treated emoticons that they encountered in their samples, so
one assumes that they were fitted into the taxonomy as supportive moves in line
with Dresner and Herring (2010). Maíz-Arévalo (2013) and Eslami, Jabbari and
Kuo (2015) give emoticons their own place within their taxonomies of compli-
ment responses, but outside Holmes’s global categories of acceptance (see Tables
5 and 6, above). In brief, it seems that the intention behind the use of emoticons in
responding to compliments on Facebook is open to greater debate and exploration.

Authenticated
Another method of interacting on social media that generates debate is the

‘like’ function (see 3.1 and also Maíz-Arévalo, Ch. 21, this volume). Placencia and
Lower (2013) suggest that using the ‘like’ function displays the ‘liker’s’ approval
or appreciation of the recipient. With respect to responses, Placencia, Lower and
Powell (2016) interpret the ‘like’ function as a form of compliment acceptance,
as ‘liking’ something is a positive evaluation of the object being assessed (2016:
348). Even though motivations behind the use of the ‘like’ function can be vague,
Placencia, Lower and Powell (2016) place its use as a response to a compliment
on Facebook in the compliment acceptance category. Maíz-Arévalo (2013) gives
Facebook’s ‘like’ function its own place in her taxonomy; however, she does not
classify it as an acceptance but as an implicit response or a non-verbal response. At
the same time, Maíz-Arévalo seems to recognize its value as an acceptance token:
she asserts that the use of ‘like’ on its own, in response to a compliment, is a polite
acknowledgement of the compliment and a demonstration of appreciation (2013:
64). As such it is unclear why this use of ‘like’ was not classified by Maíz-Arévalo
as a type of acceptance strategy. Similarly, Eslami, Jabbari and Kuo (2015) seem
to show some ambiguity in their categorization of ‘like’. As Maíz-Arévalo (2013),
they classify clicking ‘like’ together with emoticons as nonverbal strategies (2015:
254), which stand in contrast with the category of verbal strategies (i.e., accept,
reject, evade or a combination of these). On the other hand, Eslami, Jabbari and
Kuo (2015: 256) assert that when used alone or when combined with emoticons,
‘like’ is a form of compliment acceptance. In fact, ‘like’ appears unambiguously as
an acceptance strategy in their discussion of acceptance sub-strategies according
to gender (2015: 258).
4. Methodological, ethical and other issues in the study of

complimenting behavior in social media
As with any area of research, methodological and ethical issues must be taken into
account and given proper attention, together with practical considerations. This is
especially relevant in social media research, as it is a relatively new area of study,
and there are no tried and true medium-specific methods for researchers to fall
back on. Thus far, the majority of studies seem to involve a form of participant
observation that takes researchers into the realm of digital ethnography (see Sec-
tion 3 above).
A key methodological challenge is in developing a corpus. Leaving ethical
matters aside (see below), it may seem simple, given that the interactions unfold
in written form, i.e., it’s all there, waiting to be downloaded and analyzed. The
question then is how and how much. The answer is closely linked to the research
questions addressed. However, there are practical considerations that impact on
the research. How does a researcher extract data from a social media platform,

Authenticated
and, once extracted, where does the data go for analysis? If the data is copied and
pasted manually from a social media platform, this process is extremely time-con-
suming (cf. Placencia and Lower 2013) and ultimately restricts the amount of data
available for analysis. Storing data extracted from social media is also problem-
atic: if a researcher simply extracts the words from comments, there are few prob-
lems with storage capacity. However, without further context such as a photograph
in the case of Facebook or Instagram and sometimes the co-text, deciphering the
meanings of comments becomes more difficult and more time consuming, as a
researcher will have to go back to the original source. This is not always possible,
as social media users frequently delete photographs and comments. Therefore, it
becomes more advantageous for a researcher not only to extract word data, but
photographic data as well. This causes practical problems as photographs take up
vast amounts of storage space on a computer. Additionally, ethical issues may be
raised about the storage of photographs unless informed consent is obtained, as
participants may have reservations about identifying data potentially being leaked
if not stored properly, or identifying data being shared with other researchers or
used for other purposes in the future.
Given the vast amount of data available on social media, where does a researcher
draw boundaries for a given study? This is a difficulty highlighted by Hine (2009),
among others. For example, when conducting research on Facebook, a researcher
must decide the number of participants to extract data from and an end date. Face-
book users constantly upload new photographs, and other users constantly make
comments on photos. Researchers could be updating their data indefinitely unless
they draw clear boundaries for data extraction. On the other hand, if too little data
is extracted from too few participants, the data is not a very representative sam-
ple of a larger population. This is also a concern when data is extracted from one
person’s social network. Randomly selecting participants from one social network
goes some way to remedying this potential pitfall.
Once a researcher has determined how to extract data from social media, and
how much data to extract, they have to decide how to organize this data into a
meaningful and useful corpus. This is something that is usually not considered or
mentioned in published research. Placencia and Lower (2013) copied and pasted
word data into an Excel spread sheet. This did not include photographs but rather a
brief description of each photograph in the sample. The advantage of Excel is that
it makes the data easy to organize and search. It was also relatively simple to use
macros in order to extract counts of various occurrences of data from the spread-
sheet. The disadvantage of this method is that if researchers want to see the con-
text (i.e., photographs and co-text in some cases) and desire to use the data later
for other purposes, it is possible that some examples may have to be discarded
because the original context has disappeared from the social media platform.
If a researcher does devise a way to extract a large corpus from social media,
how then should the analysis proceed? For example, what kinds of searches should

Authenticated
be made? This is a problem similar to that described by Jucker et al. (2008), who
attempt to use automated methods to extract occurrences of various syntactic pat-
terns of compliments from a large corpus. This is a dilemma for analyzing social
media as well and is exacerbated due to the huge amount of orthographic variation
(including spelling mistakes as well as intentional spelling alterations) found on
social media, which makes automated analysis nearly impossible. The most accu-
rate way to analyze this sort of data is manually, which can be extremely time
consuming. Jucker et al. (2008) advocate a combined quantitative and qualitative
analysis when working with large corpora, but, due to the wide variation of lan-
guage usage on social media, automated quantitative analysis may not yield signif-
icant results. The need to analyze data manually no doubt restricts a researcher’s
ability to deal with large quantities of data.
Another issue in complimenting behavior research, not specific to the online
medium, relates to a lack of agreement concerning the analytical categories
employed. Comparisons between studies available or emerging can be hampered
by different ways in which researchers categorize compliments and compliment
responses (cf. Placencia, Lower and Powell 2016: 344 and above). With respect to
compliments, for example, different researchers use the terms direct and indirect
in at least two different ways. With reference to Searle (1975) and Blum-Kulka,
House and Kasper’s (1989) characterization of levels of (in)directness in speech
act research, direct compliments would be those where the positive evaluation is
made explicit through lexical choices that carry a positive semantic load (Jaworski
1995; Placencia and Lower 2013). By contrast, indirect or implicit compliments
are those that rely on “interpretive procedures” (Boyle, 2000: 35) as there is no
element in the utterance that conveys a positive semantic load (see Section 3).
On the other hand, for Kerbrat-Orecchioni (1994) direct compliments are those
that concern the hearer directly whereas indirect ones are those that concern a
third person, closely associated with the hearer. This distinction may derive from
a remark by Manes and Wolfson (1981) about indirect forms that they do not
develop further: indirect compliments are those that are “overtly addressed to one
person but actually compliment another party who is present” (Manes and Wolf-
son 1981: 122). They go on to point out that it is often the complimentee, and
not the addressee, who replies to what seems to be an overheard compliment.
Kerbrat-Orecchioni, like Holmes (1986), uses the terms explicit and implicit to
refer to Blum-Kulka, House and Kasper’s (1989) notions of direct and indirect
speech acts. A third use of the notion of (in)direct compliments is mentioned by
Hoffmann (2013: 346) with reference to Yuan (2001): the complimenter quotes a
second person who compliments a third party. These are cases where the “speaker
and complimenter are not one and the same person” (2013: 346).
Likewise, with respect to compliment responses, Placencia, Lower and Pow-
ell (2016: 344) note that returning a compliment is considered an acceptance by
Holmes (1986), whereas Herbert (1990) classifies this strategy under ‘other inter-

Authenticated
pretations’. Similar observations can be made about emerging categorizations of

certain compliment (response) phenomena online. For example, Maíz-Arévalo
(2013) regards the ‘like’ function as a nonverbal strategy, whereas Placencia and
Lower (2013) view it as a hybrid form given that it contains the word ‘like’ which
carries a positive semantic load; they therefore propose to classify it as a subset of
compliments, albeit with its own peculiarities (see above).
In addition to adapted methodological approaches required by research in
social media, ethical considerations must be considered afresh with this new
medium. The Association of Internet Researchers (AoIR), an academic associa-
tion that promotes scholarly Internet research, has published a set of guidelines
(originally published in 2002, updated in 2012) on conducting Internet research.
According to the AoIR, the following are the main considerations for researchers
in Internet-related contexts: 1) how the context is defined and conceptualized; 2)
how the context is being accessed; 3) who is involved in the study; 4) what is the
primary object of the study; 5) how the data is being managed, stored, and repre-
sented; 6) how the texts/persons/data are being studied; 7) how the findings are
presented; 8) what the potential harms or risks associated with the study are; 9)
what the potential benefit of the study is; 10) how the researcher recognizes the
autonomy of others (acknowledging that they are of equal worth to the researcher
and should be treated as such); and 11) what particular issues might arise around
the issue of minors or vulnerable persons. With respect to the first consideration
– the definition and conceptualization of the context – a further question is what
ethical expectations, particularly privacy considerations, users attach to the venue
in which they are interacting. The AoIR further explains that individuals’ expec-
tations of privacy are constantly changing, given the context. Therefore, expecta-
tions of privacy by users might differ across social media platforms, which means
that researchers need to be aware and must take into account the fluid and flexible
nature of privacy; it is not a one size fits all approach.
5. Conclusions and prospects for the future
As we saw in Section 3, studies on complimenting behavior in social media have so

far to a large extent focused on comparisons with face-to-face interaction. Results
point to certain similarities as well as some differences, the latter largely linked
to the feature of interaction and the technological affordances and constraints of
the different media. The influence of the medium can be seen, for example, in
the higher use of elliptical formulations of compliments on SNSs than in face-
to-face contexts, the use of prosodic spellings (Androutsopoulos 2000) and other
devices of “textual deformation” (Yus 2011) for emphasis or other purposes, as
well as visual means that include emojis and hybrid forms such as the ‘like’ func-
tion and hashtags. It also shows in the occurrence of non-response, rather unthink-

Authenticated
able in face-to-face contexts (cf. Valdés and Pino 1981). All in all, the results that
are emerging suggest that some new norms of appropriateness are developing to
suit the online medium. On the other hand, situational and social factors, like in
face-to-face interaction, also continue to play a role. For instance, non-technical
characteristics of a site such as the goals pursued by a given site’s creator (as
in Hoffmann’s 2013 study on personal and corporate blogs) can have an impact
on complimenting behavior. Likewise, macrosocial factors, such as gender (cf.
Siti Yuhaida and Tan 2014), and age (cf. Placencia 2015) also appear to play a
role. It is not surprising to find that particular patterns described for face-to-face
interaction (e.g. women receiving and offering more compliments than men, with
compliments on appearance standing out) have carried over to online contexts, and
now appear to include ‘likes’ too. They reflect deep-seated values and behaviors
that simply appear to be replicated in interactions mediated by new technologies.
Methodologically, researchers need to complement observation of behavior
online with other methods such as interviews and focus group discussions in order
not only to understand certain uses of emojis, for example, but also to access users’
perceptions of appropriateness and to gain insight into the different social and sit-
uational factors that influence complimenting behavior online.
Concerning prospects for the future, with developments in technology and
social media, new or altered spaces for the study of complimenting behavior will
continue to emerge. Researchers need to be flexible and change with the platforms
available for examination in order to fully exploit research possibilities. In this
respect, while it is paramount to be respectful of subjects’ autonomy and privacy,
the pendulum can swing too far in the direction of over-cautiousness if approval
for social media research is too difficult to obtain. Social media is an increasing
part of our lives and thus needs to be encouraged and supported, and old research
norms should not be forced where they do not fit. Also, research in the area so far
has represented good attempts at navigating a new course through unchartered
territory and we can only see it improving. Researchers need to use each other as
resources, to learn and build on past works, and to address areas in past research
that perhaps left a bit to be desired, specifically in terms of rigorousness of meth-
odology.
At the moment, compliment research on social media understandably draws
heavily on comparisons between online and face-to-face interactions; hopefully
the future will bring comparisons between different social media platforms, and
different studies on the same platforms. Of course, comparisons between online
and face-to-face are useful because they represent pertinent shifts in society, but
at some point it would be useful to see compliment research on social media as
a focus in its own right (cf. Das 2010). If anything, comparisons with face-to-
face interactions should ideally be current in order to best reflect societal norms.
There are multiple SNSs where complimenting behavior could be explored from
different perspectives: from purely pragmalinguistic to sociocultural and critical

Authenticated
approaches. Taking Instagram as one example, anonymous vs. non-anonymous

compliments could be analyzed. On Instagram, people can have thousands of fol-
lowers, so the majority of compliments come from strangers. Often, these clearly
cross lines of propriety. How do users perceive the crossing of these boundaries,
and how do they react? How do compliments influence or reinforce negative
behavior? These are some of the questions, among many others, that would be
interesting to explore. An interdisciplinary approach will be required to address
some of them.
Finally, while research in face-to-face contexts has shown cultural variabil-
ity in both compliment and compliment response behavior, a potential area of
exploration is the extent to which social media, with both enabling and constrain-
ing features that shape up interaction online, is resulting in more homogenous
complimenting behavior. With respect to compliment responses, the non-response
referred to above is found across studies, although the rate of incidence of this
category varies rather drastically. The factors influencing these results still need
to be teased out: does it have to do with the methodology employed, and/or the
degree of social distance between users, and/or the age of the participants and their
sociocultural background?
References
Achugar, Mariana
2002 Piropos: Cambios en la valoración del grado de cortesía de una práctica dis-
cursiva. In: María Elena Placencia and Diana Bravo (eds.), Actos de habla y
cortesía en español, 175–192. Munich: Lincom Europa.
2000 Non-standard spellings in media texts: The case of German fanzines. Journal
of Sociolinguistics 4: 514–533.
Bach, Kent and Robert M. Harnish
1979 Linguistic Communication and Speech Acts. Cambridge, MA: M.I.T. Press.
Black, Pamela J., Melissa Wollis, Michael Woodworth and Jeffrey T. Hancock
2015 A linguistic analysis of grooming strategies of online child sex offenders:
Implications for our understanding of predatory sexual behavior in an increas-
ingly computer-mediated world. Child Abuse & Neglect 44: 140–149.
Blum-Kulka, Shoshana, Juliane House and Gabriele Kasper
1989 The CCSARP coding manual. In: Shoshana Blum-Kulka, Juliane House and
Gabriele Kasper (eds.), Cross-cultural Pragmatics: Requests and Apologies,
273–294. Norwood, NJ: Ablex.
Boyle, Ronald
2000 “You’ve worked with Elizabeth Taylor!”: Phatic functions and implicit com-
pliments. Applied Linguistics 21: 26–46.
1987 [1978] Politeness: Some Universals in Language Usage. Cambridge: Cambridge
University Press.

Authenticated
Chen, Rong
2010 Compliment and compliment response research: A cross-cultural survey. In:
Anna Trosborg (ed.), Pragmatics across Languages and Cultures, Handbooks
of Pragmatics, Volume 7, 79–101. Berlin/Boston: de Gruyter.
Cirillo, Valeria
2012 The pragmatics of virtual environments. Compliment responses in Second
Life. Lingue e Linguaggi 7: 37–58.
Cirillo, Valeria
forthcoming “I never forget a pretty pixel”: Compliments in a virtual environment.
Das, Anupam
2010 Linguistic Politeness and Interpersonal Ties among Bengalis on the Social
Network Site Orkut: The Bulge Theory Revisited. Unpublished doctoral thesis.
Indiana: Indiana University.
Dresner, Eli and Susan C. Herring
2010 Functions of the nonverbal in CMC: Emoticons and illocutionary force. Com-
munication Theory 20: 249–268.
Eslami, Zohreh R., Nasser Jabbari and Li-Jen Kuo
2015 Compliment response behaviour on Facebook: A study with Iranian Facebook
users. International Review of Pragmatics 7: 244–277.
Gardner, Rod
2001 When Listeners Talk: Response Tokens and Listener Stance. Amsterdam/Phil-
Goddard, Cliff and Anna Wierzbicka
2014 Words and Meanings: Lexical Semantics across Domains, Languages, and
Cultures. Oxford: Oxford University Press.
Golato, Andrea
2011 Appreciatory sounds and expressions of embodied pleasure used as compli-
ments. In: Gisle Andersen and Karin Aijmer (eds.), Pragmatics of Society,
Handbooks of Pragmatics, Volume 5, 361–392. Berlin/Boston: de Gruyter.
Haverkate, Henk
1994 La cortesía verbal: estudio pragmalingüístico. Madrid: Gredos.
Herbert, Robert K.
1986 Say “Thank You” – or something. American Speech 61(1): 76–88.
Herbert, Robert K.
1989 The ethnography of English compliments and compliment responses: A con-
trastive sketch. In: Wieslaw Oleksy (ed.), Contrastive Pragmatics, 3–35.
Herbert, Robert K.
1990 Sex-based differences in compliment behavior. Language in Society: 19(2):
201–224.
Herstand, Ari
20 Aug. 2013 Why retwitting compliments is not bragging. Ari’s take. http://aristake.
com/?post=84 (accessed 10 December 2015).
Hine, Christine
2009 How can qualitative internet researchers define the boundaries of their pro-
jects? In: Annette N. Markham and Nancy K. Baym (eds.), Internet Inquiry:
Conversations about Method, 1–20. Los Angeles: Sage.

Authenticated
Hoffmann, Christian
2013 E(-lectronic) schmoozing? A cross-generic study of compliments in blog com-
ments. In: Katrin Röder and Ilse Wischer (eds.), Anglistentag Potsdam 2012.
Proceedings, 341–357. Trier: Wissenschaftlicher Verlag Trier.
Holmes, Janet
1986 Compliments and compliment responses in New Zealand English. Anthropo-
logical Linguistics 28: 485–508.
Holmes, Janet
1988 Paying compliments: A sex-preferential politeness strategy. Journal of Prag-
matics 12: 445–465.
Hymes, Dell
1972 [1964] Toward ethnographies of communication: The analysis of communicative
events. In: Pier Paolo Giglioli (ed.), Language and Social Context, 21–43.
Harmondsworth, Middlesex: Penguin.
Jaworski, Adam
1995 This is not an empty compliment!: Polish compliments and the expression of
solidarity. International Journal of Applied Linguistics 5: 63–94.
Jucker, Andreas H., Gerold Schneider, Irma Taavitsainen and Barb Breustedt
2008 Fishing for compliments. In: Andreas H. Jucker and Irma Taavitsainen (eds.),
Speech Acts in the History of English, 273–294. Amsterdam/Philadelphia:
Benjamins.
Jucker, Andreas H. and Irma Taavitsainen
2000 Diachronic speech act analysis: Insults from flyting to flaming. Journal of
Historical Pragmatics 1: 67–95.
1994 Les interactions verbales. Variations culturelles et échanges rituels. Volume 3.
Paris: Armand Colin.
1997 A multilevel approach in the study of talk-in-interaction. Pragmatics 7: 1–20.
Kiefer Lee, Kate
21 Aug. 2012 Retweeting compliments: Yea or nay? Forbes (Tech). http://www.forbes.
com/sites/katelee/2012/08/21/retweeting-compliments-yay-or-nay/ (accessed
10 December 2015).
Leech, Geoffrey N.
2005. Politeness theory and relational work. Journal of Politeness Research 1: 9–33.
2012 “Was that a compliment?” Implicit compliments in English and Spanish. Jour-
nal of Pragmatics 44: 980–996.
2013 “Just click ‘Like’”: Computer-mediated responses to Spanish compliments.
2013 “You look terrific!” Social evaluation and relationships in online compliments.
Discourse Studies 15: 735–760.

Authenticated
Malinowski, Bronislaw
1972 [1923] Phatic communion. In: John Laver and Sandy Hutcheson (eds.), Commu-
nication in Face-to-Face Interaction, 146–152. Harmondsworth, Middlesex:
Penguin.
Manes, Joan
1983 Compliments: A mirror of cultural values. In: Nessa Wolfson and Elliot Judd
(eds.), Sociolinguistics and Language Acquisition, 96–102. Rowley, MA:
Newbury House.
Manes, Joan and Nessa Wolfson
1981 The compliment formula. In: Florian Coulmas (ed.), Conversational Routine:
Explorations in Standardised Communication Situations and Prepatterned
Speech, 115–132. The Hague: Mouton.
Oxford Dictionaries
2016 Oxford University Press. http://www.oxforddictionaries.com/ (accessed 10
April 2016).
Placencia, Maria Elena
2015 Peer evaluation among Ecuadorian teenage girls on Instagram. Paper pre-
sented at the 14th International Pragmatics Conference, Antwerp, 26–31 July
2015.
Placencia, María Elena and Catalina Fuentes Rodríguez
2013 Cumplidos de mujeres universitarias en Quito y Sevilla: un estudio de vari-
ación pragmática regional. Pragmática Sociocultural. Revista Internacional
sobre Lingüística del Español 1: 100–134.
Placencia, María Elena and Amanda Lower
2013 Your kids are stinking cute. Complimenting behavior on Facebook among fam-
ily and friends. Intercultural Pragmatics 10: 617–646.
Placencia, María Elena, Amanda Lower and Hebe Powell
2016 Complimenting behaviour on Facebook: The case of compliment responses in
American English. Pragmatics and Society 7: 339–365.
Pomerantz, Anita
1978 Compliment responses: Notes on the co-operation of multiple constraints. In:
Jim Schenkein (ed.), Studies in the Organization of Conversational Interac-
tion, 79–112. New York: Academic Press.
Proudman, Charlotte
2015 How many women @LinkedIn are contacted re physical appearance rather
than prof skills? https://twitter.com/crproudman/status/640934811381706752
?lang=en-gb (accessed 10 December 2015).
Ruhi, Şükriye and Gürkan Doğan
2001 Relevance theory and compliments as phatic communication: The case of
Turkish. In: Arin Bayraktaroĝlu and Maria Sifianou (eds.), Linguistic Polite-
ness across Boundaries, 341–390. Amsterdam/Philadelphia: Benjamins.
1972 [1968] Sequencing in conversational openings. In: John Laver and Sandy Hutche
son (eds.), Communication in Face-to-Face Interaction, 374–405. Harmond-
sworth, Middlesex: Penguin.
Schneider, Klaus P.
1999 Compliment responses across cultures. In: Maria Wysocka (ed.), On Language

Authenticated
in Theory and Practice: In Honour of Janusz Arabski on the Occasion of His

60th Birthday. Volume 1: Language Theory and Language Use, 162–172.
Katowice: Wydawnictwo Uniwersytetu Ślᶏskiego.
Searle, John R.
1975 Indirect speech acts. In: Peter Cole and Jerry L. Morgan (eds.), Syntax and
Semantics 3: Speech Acts, 59–82. New York: Academic Press.
Searle, John R.
1976 A classification of illocutionary acts. Language in Society 5: 1–23.
Sifianou, Maria
2001 “Oh! How appropriate!” Compliments and politeness. In: Arın Bayraktaroğlu
and Maria Sifianou (eds.), Linguistic Politeness across Boundaries: The Case
of Greek and Turkish, 391–430. Amsterdam/Philadelphia: Benjamins.
Siti Yuhaida, Anniqah Mohd Yusof and Bee Hoon Tan
2014 Compliments and compliment responses of Twitter among male and female
celebrities. Pertanika Journal of Social Science & Humanities 22: 75–96.
Valdés, Guadalupe and Cecilia Pino
1981 Muy a tus órdenes: Compliment responses among Mexican-American bilin-
guals. Language in Society 10: 53–72.
Vanderveken, Daniel (with Kenneth MacQueen)
1990 Semantic analysis of English performative verbs. In: Daniel Vanderveken,
Meaning and Speech Acts, Volume 1: Principles of Language Use, 166–219.
Watts, Richard J.
Wolfson, Nessa
1981 Compliments in cross-cultural perspective. TESOL Quarterly 15: 117–124.
Wolfson, Nessa
1983 An empirically based analysis of complimenting in American English. In:
Nessa Wolfson and Elliot Judd (eds.), Sociolinguistics and Language Acquisi-
tion, 82–95. Rowley, MA: Newbury House.
Wolfson, Nessa
1988 The bulge: A theory of speech behavior and social distance. In: Jonathan Fine
(ed.), Second Language Discourse: A Textbook of Current Research, 21–38.
Norwood, NJ: Ablex.
Wolfson, Nessa and Joan Manes
1980 The compliment as a social strategy. Papers in Linguistics: International Jour-
nal of Human Communication 13: 391–410.
Yuan, Yi
2001 An inquiry into empirical pragmatics data-gathering methods: Written DCTs,
oral DCTs, field notes, and natural conversations. Journal of Pragmatics 33:
271–292.
Yus, Francisco
2005 Attitudes and emotions through written text: The case of textual deformation
in Internet chat rooms. Pragmalingüística 13: 147–174.
Yus, Francisco

Authenticated
24. Requesting and advice-giving
Phillip R. Morrow
Abstract: This chapter presents a survey of research on the speech acts of request-
ing and advice-giving in CMC. The two are related since both involve asking
someone to do something. Though both occur in all forms of CMC, research has
concentrated on e-mail requests and on advice-giving at health-related websites.
Research on e-mail requests has focused on identifying and explicating the use
of mitigating strategies and other factors related to request compliance. Another
focus has been the appropriateness of e-mail requests by non-native speakers and
their development of pragmatic competence. Studies of advice-giving at health-re-
lated websites have elucidated several aspects of advice-giving in CMC: the solic-
itation of advice, the relational work associated with advice-giving, the linguistic
structures used to express advice, the self-presentation of the advice-giver, and
how CMC is well-suited for advice-giving on sensitive topics.
1. Introduction
Making requests and giving advice are very common speech acts in daily interac-
tion, and they have received considerable attention from researchers in pragmatics
and related fields. Requesting and advising are similar speech acts in that both
involve asking someone to do something. Because of the potentially face-threat-
ening nature of these acts, individuals tend to adopt mitigating strategies when
performing them. However, there is considerable variation in how people mitigate
their requests or advice. Much of the research on both requesting (e.g. Biesen-
bach-Lucas 2007; C. Chen 2000, 2001; Economidou-Kogetsidis 2011) and advis-
ing (e.g. Locher 2006; Morrow 2006; Kouper 2010) has been concerned with
describing the variety of mitigating strategies people use, the reasons why certain
strategies are selected and used, and the effectiveness of their use, that is, whether
or not the users are able to achieve their communicative aims successfully through
their use. Thus, the aims of research on requesting and advising overlap to a large
extent.
As with other speech acts, how requests and advice are encoded and carried out
depends on the context in which they occur. Although the studies reviewed here
all deal with requesting and advising in the general context of Computer-Medi-
ated Communication (CMC), the realm of CMC is wide and encompasses diverse
forms of communication including, for example, e-mail, chatrooms, discussion
forums, social networks, and advice columns, each with its own distinctive char-
DOI 10.1515/9783110431070-024
Authenticated
662 Phillip R. Morrow
acteristics and norms of language use. Hence, the requests and advice that occur in
each of these forms of CMC can be expected to exhibit distinctive characteristics
associated with the form in which they occur. Sociolinguistic studies have shown
that people’s ways of using language depend in part on factors such as gender, age,
ethnic group membership, social class, social role, cultural beliefs, and values, and
these factors also can be expected to influence the ways individuals make requests
or give advice in CMC. Moreover, the way one uses language also depends on
one’s level of proficiency; those who use more than one language are not necessar-
ily highly proficient in all of the languages they use. Thus there are a host of con-
textual and other factors that potentially influence how requests and advice-giv-
ing are carried out. We can point to similarities and differences, but research on
requesting and advice-giving needs to carefully explicate how these speech acts
are carried out in each of the specific contexts in which they occur.
This chapter presents a survey of research on requests and advice-giving in
CMC; for an overview of issues and research related to requests in general see
Walker’s (2013) survey, and for advice-giving, Limberg and Locher’s (2012) col-
lected volume on advice in discourse. Much of the research on requests in CMC
deals with e-mail requests, and much of the research on advice-giving in CMC has
concentrated on advice-giving on health-related websites. Thus, most of the stud-
ies reviewed here are about research in these two areas. It should be stressed that
although e-mail and websites are, in a sense, classic forms of CMC, they are not
prototypical. The fact that most of the studies reviewed here are studies of these
forms reflects the reality that pragmatic research on social media is still somewhat
limited.
This survey is organized as follows: Following this introduction, Section 2
deals with e-mail requests in academic settings and Section 3 with requests in
professional settings. Section 4 is about requests in social networks and Section
5 about requests on health-related websites. In Section 6 I turn to advice-giving
in CMC, and in Section 7 to advice-giving at health-related websites. Section 8
covers advice-giving in other (non-health-related) websites and Section 9 presents
a conclusion to this survey.
2. E-mail requests in academic settings
Various forms of CMC are used in academia, but e-mail is one of the oldest and
is widely used, for example, for communication between university students and
faculty members. Students’ e-mail requests, like other requests, involve some level
of imposition and can be considered face-threatening acts in that they ask profes-
sors to perform actions they might not otherwise perform. When someone makes a
request they may need to mitigate the force of the request if it involves a substan-
tial imposition or if the request is being made to a person in a position of greater

Authenticated
Requesting and advice-giving 663
power or who is socially distant from the speaker. Although students’ requests
are often about mundane matters such as arranging appointments, they may also
involve weightier issues such as requesting professors to make exceptions to rules
or asking them to read and comment on lengthy papers. The relationship between
students and professors is an asymmetrical one in which professors have greater
power because of their institutional status. Moreover, the relationship may be a
distant one, as when a student is taking a professor’s course but does not know
the professor personally. Under these circumstances, students need to mitigate the
force of their requests in order to achieve their communication goals.
To choose appropriate strategies for mitigating requests, students need to
assess the level of imposition their requests involve, consider their distance or
closeness with the professor, and take account of the status or power difference.
This can be challenging for university students who often have considerable expe-
rience of using CMC with peers, but limited experience of using it to interact more
formally with individuals of higher status. The difficulty is amplified for non-na-
tive speakers (NNSs) who may lack not only the linguistic resources, but also the
cultural knowledge needed to assess the level of imposition and choose strategies
that express an appropriate level of politeness for their requests.
Given these difficulties, it could be expected that NNS students would avoid
using e-mail to make requests. However, Bloch (2002) reported that one of NNS
students’ four main uses of e-mail was to make formal requests. He also found that
many of their requests to faculty contained pragmatic infelicities that could lead to
the requests’ being negatively evaluated by professors and reduce the professors’
willingness to comply with them.
Numerous studies have investigated NNS university students’ e-mail requests
and they explore a number of issues related to the structure and use of such requests.
Many of these studies have a comparative focus, documenting how NSs and NNSs
differ in their production of requests (e.g. Chang and Hsu 1998; C. Chen 2001;
Biesenbach-Lucas 2005, 2007; Ford n.d.) or how different groups of NNSs (e.g.
Zhu 2012) or NSs (e.g. Merrison et al. 2012) vary. The following sections present
a survey of these studies, the approaches they have taken, their analyses of the use
of directness and mitigating strategies, their findings regarding the differences
between NSs’ and NNSs’ requests, and the implications of those differences for
NNSs’ acquisition of pragmatic competence.
2.1. Approaches to the analysis of student e-mail requests

A major study of requests was undertaken as part of the Cross-Cultural Speech
Act Realization Project (CCSARP), and as part of that project Blum-Kulka, House
and Kasper (1989) devised a framework for analyzing the content and structure
of requests. This framework is based on Goffman’s (1967) concept of face and on
Brown and Levinson’s (1978, 1987) theory of politeness and is particularly con-

Authenticated
cerned with the use of politeness strategies in mitigating face-threatening acts (cf.
Graham, Ch. 17, this volume). In the CCSARP framework, requests are analyzed
in three main ways: in terms of directness (whether they are direct or indirect),
in terms of the lexical and syntactic elements used to mitigate the requests, and
in terms of perspective (whether they adopt the hearer’s or the speaker’s point of
view) (Biesenbach-Lucas 2006).
There has been some criticism of the CCSARP analytical framework, particu-
larly in regards to the concept of directness, and the tendency to equate indirect-
ness with politeness and directness with impoliteness. Van Mulken notes that in
the CCSARP coding manual “[t]he criteria for indirectness […] remain inexplicit”
(1996: 692), and comments that hints, which are usually regarded as indirect, can
be interpreted as an impolite strategy (1996: 692). Furthermore, it has been shown
that in some non-western cultures, directness is not avoided in requests to superiors
as would be predicted by Brown and Levinson’s theory. In their study of request
strategies, Rue and Zhang found that “Chinese and Koreans were not necessarily
less direct to their superiors, or more direct to their close friends, than they were
to acquaintances” (Rue and Zhang 2008: 12). In the case of Korean, this is partly
because the language has an “effective and rich honorific system [which] mitigates
the illocutionary force of a direct speech act” (Rue and Zhang 2008: 297). Thus,
direct requests in Korean are not necessarily perceived as blunt or impolite.
Many of the analyses of e-mail requests in CMC have utilized the CCSARP
framework. However, this framework was developed for coding data obtained
through discourse completion tasks, and difficulties can arise in using it to code
naturalistic data. Biesenbach-Lucas (2006) found that some e-mail request data
could be coded into more than one category, and some other data did not fit well
into existing categories. This led her to adapt the framework. She reported that
different and even conflicting results were obtained when the data were coded with
the adapted version. When requests were coded with the original version, results
indicated that the cyber-consultations entailed directness, but when coded with the
adapted version, they indicated that cyber-consultations did not entail directness,
but rather led to greater indirectness. This difference arose in large part from the
coding of need and want statements; they were coded as indirect strategies rather
than as direct ones in the adapted version. The differing results underscore the fact
that there is a degree of subjectivity in the interpretation of speech act data, and
this can lead to different conclusions.
In analyzing the structure of requests, a distinction is made between the head
act, which expresses the request itself, and supporting moves, which can occur
before or after the head act. Supporting moves such as grounders, which support
the request by providing reasons, are referred to as external modification, while
the use of lexical items or syntactic structures in the head act is referred to as inter-
nal modification. Some studies (e.g. C. Chen 2000, 2001; Economidou-Kogetsidis
2011) have reported that NNSs show a preference for external modification. They

Authenticated
suggest that external modification strategies are easier to use and that their use
may reflect the influence of the NNSs’ L1. In many languages and cultures, it is
common to preface a request by grounders that give the reason for the request on
the assumption that people are more likely to comply with a request if they under-
stand why it is being made.
While most research on e-mail requests has been concerned with analyzing the
content and structure of request messages, another line of research explores the
cognitive processes that underlie the requests. For example, Y. Chen (2015a) inves-
tigated four aspects of the cognitive processes involved in Chinese EFL learners’
production of e-mail requests to faculty. Students were given a task that involved
writing two e-mail requests to faculty. While the students were doing the task
and after they had completed it, the researcher collected verbal reports that were
coded into four categories: intention, cognition, planning and evaluation. Besides
contributing to a better understanding of the cognitive processes that underlie the
production of requests, Y. Chen’s study is significant in that it demonstrated the
usefulness of verbal reports as a methodology for research in this area.
2.2. Directness and mitigation in student e-mail requests

The terms direct and indirect have often been used in describing students’ e-mail
requests, and there has been a general perception that NNSs’ requests are more
direct than those of NSs and lack appropriate expressions of politeness (Hart-
ford and Bardovi-Harlig 1996; Hassall 1999; Bloch 2002; Mohammadi and Zarei
2012). However, research studies do not always bear this out. C. Chen (2001)
found NNSs’ e-mail requests less direct than those of American students, and
in Ford’s (n.d.) study, NNS’s e-mail messages were rated as “too polite”. It is
worth reiterating here (see Section 2.1.) that using the concept of directness can
be problematic: the criteria for directness or indirectness are not always clear (von
Mulken 1996; Biesenbach-Lucas 2006), and the tendency to equate indirectness
with politeness is not cross-culturally valid (Rue and Zhang 2008). Nor are there
clear criteria for judging a request “too polite”.
One strand of research on students’ e-mail requests has been concerned with
identifying the factors that contribute to directness and exploring the relationship
between directness and politeness or impoliteness. A significant study in this area
is Biesenbach-Lucas’ (2007) comparative study of e-politeness by NS and NNS
students at an American university. Her results showed that both NS and NNS
students varied their request strategies according to the level of imposition, using
more direct strategies for lower imposition requests and more indirect strategies
for higher imposition ones. Also, more politeness devices were used with direct
request strategies and fewer with indirect strategies. However, although both NSs
and NNSs used the same general strategies, the two groups realized the strate-
gies differently and showed preferences for different types of linguistic devices.

Authenticated
Biesenbach-Lucas attributed this difference partly to NNSs’ “lack of linguistic

flexibility and idiomatic expressions, unawareness of letter conventions transfer-
able to email, and inability to select appropriate lexical modification” (2007: 74).
While the NNS students in Biesenbach-Lucas’ (2007) study could modify
their requests to adjust to the level of imposition, Economidou-Kogetsidis (2011)
reported a lack of modification in a similar study. Her study also dealt with the level
of directness and politeness strategies used in e-mail requests, and was based on
the analysis of 200 e-mails, written by 200 Greek Cypriot university students to 11
faculty members. The requests included both requests for information and requests
for action. Some were classified as high-imposition and others as low-imposition.
Economidou-Kogetsidis reported that “the NNS students resorted largely to direct
strategies (rather than conventional indirectness) both in the case of requests for
actions and for information, with the imperative (‘please + imperative’), direct
questions and want statements as the most preferred substrategies” (2011: 3206).
She also found that “the majority of the e-mail requests were bare of any lexical/
phrasal modification …which can also cause pragmatic failure by adding a coer-
cive tone” (2011: 3207) and added that some even included an upgrader such as
asking that the recipient comply “as soon as possible”.
As part of her study, Economidou-Kogetsidis (2011) surveyed British native
speakers to determine whether and to what extent they perceived unmodified and
direct requests as impolite. Based on their perceptions as well as her own, she
concluded that directness, lack of lexical/phrasal downgraders, and omission of
greetings and closings contributed to a brusque or impolite tone and sometimes
led to pragmatic failure. Similarly, in a study of e-mail requests by Dutch NNSs,
Hendriks (2010) found that underuse of more linguistically complex modification
could lead to NSs’ negative perception of a sender’s personality. On the other
hand, the use or non-use of single modifiers such as possibly, or could rather than
can did not seem to cause any negative perception.
Various other factors have also been identified as contributing to a message’s
being perceived as overly direct or impolite. Hartford and Bardovi-Harlig (1996)
found that requests that assumed a greater obligation of the faculty member to
comply were associated with negative affect. Furthermore, requests with less mit-
igation, less acknowledgement of the imposition involved, less tactful reference
to time constraints, and more personal rather than institutional reasons were also
evaluated negatively.
Salutations, length and organization may also affect how a request is per-
ceived. In Economidou-Kogetsidis’ (2011) study, faculty participants commented
negatively on the omission of a salutation, or the use of a name without Dear, Hi,
or Hello. Chang and Hsu (1998) reported a preference on the part of American NSs
for requests that were “short, to the point, direct, clear and which give the recipient
options” (1998: 145). However, the Chinese students in their study tended to use
an inductive ordering in which the actual request was prefaced by preambles and

Authenticated
reasons. Similarly, C. Chen (2000) noted that some discourse features used by Chi-
nese NNS students such as delayed introduction of requestive intention, facework
for positive politeness and pre-posed grounders, could have a negative effect on
communication with Americans.
While most studies of e-mail requests published in English deal with requests
in English, Chejnová (2014) analyzed the degree of directness, amount and use
of internal and external modification, and use of address terms in Czech students’
e-mail requests in Czech. She observed that for some professors, the use of pos-
itive politeness strategies rather than negative ones could lead to requests being
perceived as impolite:
Negative politeness strategies are expected in institutional settings; therefore, requests
written in Standard Czech, starting with a deferential form of address and containing
a great amount of syntactic modification and numerous apologies may be perceived as
the most appropriate. E-mails that are based more on positive politeness strategies (sub-
standard language expressions, phatic elements, narrative grounders including personal
information, excessive thanking, emotional expressions) may be perceived as polite by
some recipients but as inappropriate by some other recipients. (2014: 187)
Chejnová suggested that the lack of consensus on the appropriate way of express-
ing politeness may reflect a shift in communication norms.
It is not only in Czech that there is a lack of consensus about the appropri-
ate ways of expressing politeness in e-mail. A comparative study (Merrison et al.
2012) of e-mail requests by NS Australian and British university students revealed
significant differences in the ways that students constructed their requests and also
their student identities. The Australian students were writing to a faculty member
in an Australian university, and the British students to a faculty member of a uni-
versity in England. Merrison et al. attributed the differences in their requests to a
difference in orientation, with the British students being more oriented toward “def-
erential dependence” and the Australians toward “interdependent egalitarianism”
(2012: 1078). The British students’ deference and orientation to their institutional
status was manifested through their use of professional titles (such as Dr.), which
they used more than eight times as much as the Australian students. The British
also used more apologies and presented themselves as having a lower social status
and as not always able to manage their role as students. On the other hand, the
Australians were more oriented to their equal status in the wider society, and their
requests were more on the level of peer-to-peer, de-emphasizing status differences
and stressing the value of social equality. This was reflected in their greater use of
geniality (well-wishing, stressing closeness and common ground) and their use of
paid employment as mitigating circumstances. By referring to employment, they
presented themselves, like their instructors, as contributing members of an egali-
tarian society. These differences led Merrison et al. to argue that “knowing how to
do politeness in English is not enough – we also have to know how to do politeness
in a particular cultural and situational context” (2012: 1096).

Authenticated
While research findings have not always validated the assumption that NNSs’
e-mail requests are more direct than those of NSs, comparative studies have shown
that NNSs and NSs differ in the way they carry out the speech act of making e-mail
requests. This raises the question of how the differences are to be explained and
how students can acquire pragmatic competence in making requests.
2.3. Acquiring pragmatic competence in making requests

Lack of linguistic resources has been cited as a reason for pragmatic infelicities
in NNSs’ e-mail requests. Based on the results of questionnaires and interviews,
Tseng (2015) reported that the high- and low-intermediate level Taiwanese EFL
learners in his study believed that being indirect was the most important way of
expressing politeness, but that they lacked the linguistic resources and contextual
knowledge necessary for using appropriate strategies. Likewise, Biesenbach-Lu-
cas (2007) pointed out that although the e-mail format allows for time to plan,
compose, revise and edit, having extra time does not necessarily help NNSs since
they lack the linguistic resources to make their messages more polite.
In fact, it is very difficult for NNSs to acquire e-mail pragmatic competence
judging from C. Chen’s (2006) longitudinal case study of one Taiwanese graduate
student’s e-mail practice over a period of two and a half years. In the early part of
the study, the student tended to make requests using want statements (e.g. “I need
your suggestion …”), while later she used many query preparatory strategies
(e.g. “I’m wondering if you could …”). The student explained that she used want
statements to emphasize that the professor’s help was very important. However,
C. Chen noted that these could have a negative effect in that they were “likely to
project a negative image as a needy, helpless student” (2006: 44). Furthermore,
using unmitigated want statements could be perceived as coercive, emphasizing
her rights a student, without giving options to the professor.
C. Chen reported a marked change in the type of supportive moves used by
the student in the early and later periods. In the early period, she tended to give
supportive moves before the request, generally giving student-oriented reasons
and personal details that emphasized urgency or seriousness. Later, however, she
tended to use supportive moves related to institutional needs (e.g. a requirement
to consult an advisor before registering). She also started using types of mitigating
supportive moves that she had not used previously such as asking about the pro-
fessor’s availability or showing reluctance to impose. C. Chen observed that these
strategies are highly valued in Western cultures and suggested that these changes
could be attributed to a change in identity. She concluded that the student’s use of
lengthiness, inductive structures, want statements and self-humbling expressions
were related to the student’s image as a needy master’s student, and that replac-
ing these strategies was in keeping with a new identity as a competent doctoral
student.

Authenticated
C. Chen’s study points up the difficulties NNSs experience in learning to

express the appropriate level of politeness. She noted:
The difficulty arose from the fact that there were no models for her to imitate and no
explicit rules to follow either. She could not completely imitate the way her professors
wrote e-mails to her due to the power asymmetry and neither could she imitate the way
native speakers wrote their e-mails to their professors since she had no opportunity to
read those e-mails. Moreover, she was unable to find out what part of her e-mails was
problematic or inappropriate because her e-mail interlocutors did not tell her. (2006: 50)
Some researchers (e.g. Biesenbach-Lucas 2007; Ford n.d.) have called for peda-
gogical intervention to train students to carry out e-mail requests appropriately.
Findings from Ford’s (n.d.) study indicate that e-mail pragmatic features can be
taught to NNSs and that doing so enables NNSs to present their requests effectively
and thereby achieve their communicative goals. Ford (2003) provides a sample
lesson for teaching pragmatics. Y. Chen (2015b) also reported that the Chinese
students of English in her study made improvement overall in producing e-mail
requests after six hours of instruction. However, they improved more in framing
moves (subject, greeting and closing) than in content moves (request strategies
and request support). While some textbooks deal with e-mail and include sections
on writing requests (e.g. Chapman 2007), they have limited value for students as
they are aimed primarily at business professionals. Writing effective e-mails is
very much a concern of business and other professionals and has become a topic
of research. I turn to a review of this research in the following section.
3. E-mail requests in professional settings
Several studies of requests have dealt with the issue of how e-mail can be used stra-
tegically in professional settings to increase the likelihood of compliance with a
request. At a time when e-mail was still a rather new medium of communication in
organizations, Phillips and Eisenberg (1993) conducted a study to determine how
e-mail was being used. They identified four common strategies: direct requests,
indirect-implicit requests, carbon-copying, and message-forwarding, and found
that carbon-copying and message-forwarding were used in conjunction with direct
and indirect requests to increase the likelihood of compliance. By carbon-copying
(cc-ing) bosses or peers, and by forwarding prior messages – sometimes a copy of
a first request – senders put pressure on co-workers to comply with their requests,
thereby increasing the likelihood of compliance.
The success of an e-mail request normally depends on getting a response, and
the probability of getting a response depends in part on how the request is pre-
sented. One aspect of the presentation is whether it is addressed to a single recip-
ient or multiple recipients. Based on the results of an experiment, Barron and

Authenticated
Yechiam (2002) determined that “there are more responses to e-mails addressed to
a single recipient, that the responses are more helpful, and that they are lengthier”
(2002: 507). They explained this finding in terms of “diffusion of responsibility”,
that is, when a request is addressed to multiple recipients, such as an entire discus-
sion group, individuals may feel that others are able to provide assistance, and thus
may not feel obligated to do so themselves (2002: 515).
The success of an e-mail request is also related to who makes the request and
how they present themselves. Ho (2010) analyzed the strategies by which leaders
of a group of English teachers in Hong Kong constructed identities for themselves
through the e-mails they sent their subordinates. Ho’s analysis was based on the
“Self-Aspect Model of Identity” (Simon 2004: 45), according to which an individ-
ual has various self-aspects and by highlighting one or more of their self-aspects in
discourse, individuals can construct different identities for themselves in different
situations. The identities these teachers constructed included being “an accounta-
ble leader, a rational leader, an authoritative leader, an understanding, considerate
and polite leader, and a capable leader” (Ho 2010: 2255).
In another study, Ho (2011) examined the intertextuality and interdiscursiv-
ity functions of e-mail requests in professional communication. Citing Fairclough
(1992), Ho argued that intertextuality is present in any text because texts are “con-
stituted by elements of other texts” (Fairclough 1992: 102, cited in Ho 2011: 2535).
As for interdiscursivity, Ho quoted Candlin and Maley (1997: 212) to define it as,
“the use of elements in one discourse and social practice which carry institutional
and social meanings from other discourses and social practices” (Ho 2011: 2536).
Ho also noted that the intertextual and interdiscursive elements are not essential to
a request since the recipient would be able to interpret the message without them,
but that these elements can be used for special pragmatic functions. Ho concluded
that through effective use of intertextual and interdiscursive elements profession-
als could enhance their chances of achieving “the purposes of the communication
act more effectively – achieving request compliance in the present case – in both
direct and indirect ways” (2011: 2546).
Although e-mail requests have received much attention from researchers, they
are just one type of request in the ever-expanding realm of CMC. The following
sections survey research on requests on social networks.
4. Requests on social networks
More and more people are using social networks (SNS): “As of 2013, 67 % of
American Internet-using adults reported using a social network site, up from a mere
5 % of adults in February 2005” (Rainie, Smith and Duggan 2013, cited in Ellison
et al. 2013). Facebook is the most widely used SNS. Ellison et al. reported that
92 % of SNS users had Facebook accounts. These users often use status updates to

Authenticated
make requests. Using SNS to make requests has several advantages: responses are
tailored to the questioner, respondents are known and trusted by the questioner, and
by using SNS a questioner may receive emotional support (Ellison et al. 2013; see
also Eisenlauer, Ch. 9, this volume). Ellison et al. examined 20,000 public Face-
book status updates for instances in which requests were made for an action or
response. Requests were categorized as “social coordination, factual knowledge,
opinion/polling, recommendation/suggestion, and favor/request for action” (2013:
157). The results showed that requests for help, information or action constituted
approximately 4 % of the status updates. Among the requests, requests for favors
were the most common type, and requests for recommendations were the least. As
expected, “the majority of requests required the least effort to fulfill” (2013: 162);
perhaps costlier requests were made directly to those who could fulfill them. There
were differences in the types of messages posted by different groups of users: “users
who posted social coordination requests were significantly younger than those who
posted factual knowledge requests, recommendation/suggestions, and polls for
opinion” (2013: 163). Such differences can be regarded as reflecting different opin-
ions as to what types of requests are appropriate to post on Facebook. Ellison et al.
suggested that some users may not see Facebook as an appropriate place to make
requests for advice, opinions and favors, but see it as a purely social space.
4.1. Questions as requests in SNS

Searle has written that “asking questions is really a special case of requesting”
(1969: 69). Requests are often phrased as questions, and most questions, except
rhetorical questions, can be seen as requests for action, information, or support.
Researchers in information science, communication and other areas are interested
in the types of questions people ask, and the frequency and quality of replies they
receive on various internet sites. Morris, Teevan and Panovich (2010) conducted
a survey of 624 SNS users to obtain information about the kinds of questions they
asked and answered online. Their research “explored the relationships between
answer speed and quality, properties of participants’ questions (type, topic, and
phrasing), and properties of participants themselves (age, gender, and social net-
work use habits)” (2010). A later study by the same researchers (Teevan, Morris
and Panovich 2011) examined the factors affecting response quantity, quality and
speed for questions asked on Facebook. To investigate these factors, they con-
ducted a controlled study in which “282 participants posted variants of the same
question as their status message on Facebook” (2011: 630), and found that, “by
ending an information need with a question mark, explicitly scoping the audience,
and being succinct, a person can increase the likelihood of quickly receiving many
high-quality answers” (2011: 630).
Many other studies have investigated online questions and answers. For
example, Harper et al. (2008) examine the types of questions (e.g. conversational

Authenticated
questions and informational questions) asked and responses received on Yahoo!

Answers; Efron and Winget (2010) and Paul, Hong and Chi (2011) analyze ques-
tions on Twitter; Gray et al. 2013 examine the connection between bridging social
capital, question type and relational closeness to the usefulness of responses to
questions on Facebook; and Arguello et al. (2006) and Burke et al. (2007) inves-
tigate the strategies, including the use of requests, that lead to people receiving
answers to their posts in online communities.
4.2. Requests in web chats

Using methodology from conversation analysis, Golato and Taleghani-Nikazm
(2006) investigated the “preference organization of requests in online chats in
Germany by systematically analyzing their turn construction, the context in which
they occur, and their content” (2006: 302). They found that some features related
to preference organization that occur in ordinary conversation do not occur in
web chats; there are not silences before dispreferred turns, nor can self-repair be
observed in web chats. However, other features that occur regularly in ordinary
conversation can also be observed in web chats. One such feature is the common
use of pre-requests. Because requests are dispreferred actions, participants may
use a pre-request to elicit an offer, and thereby avoid having to make a request.
Other features that occur both in ordinary conversation and in web chats include
the use of reasons or justifications with requests, as well as the use of if-con-
ditionals and modal verbs to mitigate the request (2006: 308–309). Golato and
Taleghani-Nikazm observe that overall, “request sequences in chats […] appear
less ‘complex’ (they contain fewer insertion sequences) and are ‘cleaner looking’
(they contain neither repairs nor hesitations) than they are in ordinary conversa-
tion” (2006: 309).
One characteristic feature of web chat requests that does not occur in ordinary
conversation is the use of smiley faces and other emoticons. In fact, Golato and
Taleghani-Nikazm (2006) reported that all of the requests in their corpus con-
tained at least one emoticon. They see the use of positive emoticons as a mitigating
device.
5. Requests on health-related internet sites
Individuals frequently post messages to health-related discussion boards in order

to receive information, support, or advice. While asking for information is rela-
tively straightforward, eliciting support or advice is a delicate matter, and for that
reason advice is not usually requested directly. Rather, there is a tendency for peo-
ple to elicit advice by describing a problem or expressing uncertainty about how
to deal with some issue.

Authenticated
A number of studies have investigated the means people use to elicit support
or advice on health internet sites. These include (but are not limited to) requests.
Studies of ways of eliciting support generally analyze interaction on one website
that is related to a specific health issue, for example, Wang, Kraut and Levine’s
(2015) study of postings to the discussion board of a cancer support community,
Lippka et al.’s (2013) study of a bladder cancer discussion board, or Ballentine
and Stephenson’s (2011) study based on a survey posted to a Weight Watchers
Facebook page.
Wang, Kraut and Levine’s (2015) study investigated how individuals elicit and
provide social support in online communities. Prior research (Bambina 2007) indi-
cated that people ask questions to get factual information, but often relate personal
narratives to get emotional support. Wang, Kraut and Levine used more than 1.5
million messages on more than 67,000 conversational threads to test whether dif-
ferent elicitation strategies (personal self-disclosure vs. asking questions) did in
fact produce different kinds of support (emotional support vs. informational sup-
port). Their findings confirmed that different strategies did produce different kinds
of support, but their study is also of interest because of its methodology which
demonstrated both the feasibility and utility of automated coding for analyzing
large corpora of online texts.
Gender can also influence the form of requests. Lippka et al. (2013) found gen-
der differences in initial postings to threads dealing with bladder cancer on a dis-
cussion board: “[w]hereas both genders equally often ask for medical information,
women more often want to share their experiences and look for emotional support”
(2013: 1). These studies also raise the issue of identifying indirect elicitation as
a request, since the studies indicated that in the context of an internet health site,
individuals interpret self-disclosure as a request for support.
6. Advice-giving in CMC
Giving advice is similar to making a request in that it involves asking someone

to do something. However, as Searle (1969: 67) pointed out, there is a difference,
because when you give someone advice you are asking them to do something
because it is best for them, whereas this is not necessarily or usually true in the
case of requests. The studies of NNSs’ requests reviewed above demonstrated that
making a request can be a difficult and delicate task requiring a high level of
pragmatic competence, and this is also true for giving advice. As with requesting,
giving advice involves proposing that someone do something, and this can be
face-threatening. Seeking advice can also be face-threatening as it implies that one
lacks competence in handling one’s own affairs. Hence, advice-seekers often elicit
advice by indirect means rather than by requesting it directly.
Advice-giving occurs in various forms of CMC including advice columns, dis-

Authenticated
cussion boards, and e-mail among others. There is advice from experts, advice from
peers, free advice, paid advice, advice given privately, and advice given publicly.
There is now a substantial body of research on advice in CMC, but although advice
is given online on almost any topic, much of the research deals with health-related
internet sites. Therefore, the following sections deal primarily with research about
such websites since they are the ones which have been most studied.
7. Advice-giving about health-related issues
Some of the main aspects of health-related advice-giving that have been

investigated include: the topics of advice, the solicitation of advice, the discourse
moves in advice messages, the interpersonal aspects of advice-giving, the iden-
tity of advice-givers, and the differences between advice-giving in CMC and in
other mediums. The following sections present an overview of significant studies
and findings related to these aspects of advice-giving on health-related websites,
beginning with Locher’s (2006) major study of advice-giving in an internet health
column.
7.1. Expert advice-giving in an internet health column

A seminal work on advice-giving in CMC is Locher’s (2006) book on advice-giv-
ing in an internet health column. Locher conducted an in-depth and multi-faceted
analysis of expert advice-giving in an internet health column sponsored by an
American university. The health column is referred to by the alias Lucy Answers
and is written by a team of health experts. It deals with a wide range of topics: rela-
tionships, sexuality, sexual health, emotional health, drugs, fitness and nutrition
and general health. The column uses a letter format in which individuals describe
a problem in a letter and receive a response also in the form of a letter. This is a
familiar format, similar to that used in agony aunts columns in newspapers.
Each of the main chapters in Locher’s book deals with one major aspect of the
analysis. For instance, Chapter 5 presents an analysis of the content of response
letters in terms of the types and frequencies of the discursive moves used. Chapter
6 is about relational work in the response letters, including mitigation strategies.
Chapter 7 considers the personal and public dimension of advice-giving in the
response letters, including how a problem letter is selected for the column, how the
scope of the response letter is broadened and how it is addressed to a wider read-
ership in the column. Chapter 8 deals with the construction of the identity of the
fictive expert advice-giver. It describes how Lucy presents herself as competent
and knowledgeable, the way advice is presented, the tone of the language used,
and the use of humor. Beside analyzing advice in the response letters, Locher also
analyzed several aspects of the problem letters such as the number and type of

Authenticated
discursive moves in them, the forms of address used, the use of questions, and the
background information supplied.
The data for Locher’s study consisted of 2,286 question-answer sequences
comprising 990,036 words. For the purpose of qualitative analysis a smaller sub-
corpus was compiled with 40 question-answer sequences from each of the seven
topic categories. Some of the specific findings from Locher’s study will be related
in the following sections along with the findings of other studies. In later articles,
Locher (2010, 2013) provides an overview of the linguistic analysis of health-re-
lated Internet sites, especially advice columns, using examples and data from her
(2006) study of Lucy Answers.
Another study based partly on Lucy Answers is Harvey, Locher and Mulla-
ny’s (2013) study. This study illustrates how a corpus linguistics approach can be
utilized to investigate large corpora and identify significant patterns of language
usage. In Harvey, Locher and Mullany’s (2013) study, data were drawn from two
expert advice columns: Lucy Answers, the university-sponsored health advice col-
umn, and Teenage Health Freak, a column aimed at adolescents. A main focus of
this study was the verbs used by questioners and respondents to talk about HIV/
AIDS. The verbs that questioners used gave an indication of their level of under-
standing – or misunderstanding – of HIV/AIDS, and this type of finding is useful
to professionals providing health care to adolescents and college-age students.
Thus, this study demonstrates the value of linguistic analysis for dealing with real-
world health issues.
7.2. Topics of health-related advice

The range of topics on which online advice is given and received can be seen
by conducting a simple search using the keyword advice. Google searches with
advice yield huge numbers of hits. Online advice is available on virtually any topic,
for example, investing, dating, legal issues, consumer issues, and health issues
to name a few of the more popular ones. Some websites (e.g. Yahoo! Answers,
Google) deal with general questions on almost any topic, while others are spe-
cialized according to topic (e.g. TripAdvisor, where users can get information and
advice about topics related to travel, organized by destination). However, much of
the research on advice-giving in CMC has concentrated on health-related internet
sites. For almost any medical condition or relationship issue, there are numerous
sites at which a person can get information or advice from experts as well as other
sites where peer advice is available. The latter includes a large number of support
sites where individuals relate their experiences with a particular issue and give and
receive support and sometimes advice from others who have had similar experi-
ences. Some sites sponsored by large organizations have a number of specialized
discussion boards for particular medical conditions or issues. There are also news
groups, listservs, chatrooms and health advice columns that deal with a variety

Authenticated
of health-related issues. Thus, advice is solicited, given and received on a broad

range of health-related topics, and the advice-giving occurs in various formats and
takes various forms.
The internet is especially well-suited for advice-giving on topics people find
sensitive or embarrassing. This point is demonstrated in Suzuki and Calzo’s (2004)
study of teens’ use of two bulletin boards, a teens’ issues board and a board dedi-
cated to sexual health for teens. Their analysis showed that questions were asked
most frequently about romantic relationships and sexual health, and that the sexual
health board was more popular than the teens’ issues board. Regarding this result,
Suzuki and Calzo remarked that, “[w]hat makes this result interesting … is the
fact while teens are reluctant to seek face-to-face advice about sexuality from
physicians and others […], these were the most popular questions posted using an
online bulletin board format” (2004: 695).
7.3. Soliciting advice on health issues

When someone seeks advice they can either ask for it directly or elicit it indirectly.
Since asking for advice can be threatening to a person’s own face, people often
avoid asking for advice directly and elicit it indirectly instead.
In her study of Lucy Answers, Locher (2006) analyzed the discourse moves in
the problem letters to which Lucy replied with advice. She reported that “questions
are the main way in which advice is asked for in Lucy Answers, and added that all
but four of the problem letters in her study contained either questions or explicit
requests for advice (2006: 231). Others have also noted the use of questions to
solicit advice. Morrow (2006) reported that or-type questions (e.g. ‘Do I have
postnatal depression or am I just another bored housewife …?’) were often used
to elicit advice in his data from a discussion forum bulletin board on depression.
Heritage and Sefi (1992) explained the use of these questions about self-diagnosis
as reflecting a desire to appear competent in the management of one’s own affairs.
This can also explain the tendency, noted by Kouper (2010), for advice to be solic-
ited in a rather elaborate manner: In her data, solicitations for advice were often
embedded in long narratives that included a great deal of background information
and justification clauses providing reasons for asking for advice.
In their studies of online advice-giving, Kouper (2010) and Sillence (2013)
both categorized the types of advice-solicitation in their data. Kouper’s data came
from an online motherhood community on LiveJournal.com, and Sillence’s from
an online support group for breast cancer sufferers. Kouper analyzed 102 instances
of advice solicitation and categorized them as follows based on a typology devel-
oped by Goldsmith (2000): 1. Request for advice (18), 2. Request for opinion
or information (44), 3. Problem disclosure (38), 4. Announcement of a plan of
action (0), and 5. Advisor volunteers advice (2). The first four of these categories
were also used by Sillence, who obtained similar results: few requests for advice

Authenticated
and most advice solicitations presented as opinion/information requests (34 %) or

problem disclosures (35 %).
Sillence’s (2013) fifth category was different from Kouper’s. Sillence called
this category, “Anybody in the same boat?”, and 20 % of the advice-solicitations in
her data took this form. She pointed out that it provided a way for people “to find
very specific advice from very specific people, namely, those who match their own
medical experience as closely as possible” (2013: 483). Sillence also observed that
not only did users seldom ask for advice directly, but even when they did, they
did so in an equivocal way; when they used the phrase “Any advice?” they often
followed it up by saying that they did not want someone to tell them what to do
(2013: 482).
7.4. Relational work in advice-giving about health issues

Most of the research on advice-giving has dealt with the means people use to
mitigate the potentially face-threatening effect of giving advice. This aspect of
advice-giving has usually been discussed and analyzed in terms of face and polite-
ness theory (see also Graham, Ch. 17, this volume, and Locher and Bolander, Ch.
15, this volume). The concept of face was derived from the work of the sociologist,
Erving Goffman (1967), who used it to explain the “positive social value” (1967:
5) which people try to maintain in interaction. The concept was subsequently
adopted and modified by Brown and Levinson (1987), and it became a key con-
cept in their theory of politeness. They divided face into positive face and negative
face to distinguish two main types of needs that individuals have in interaction.
In brief, positive face refers to the need to be accepted, liked and appreciated by
others, while negative face refers to an individual’s desire not to be imposed upon,
to be free to do what one wants. Brown and Levinson also introduced the term
face-threatening act for acts, such as requests, that threaten an individual’s neg-
ative or positive face. Much of Brown and Levinson’s (1987) study is concerned
with identifying and describing the linguistic strategies that speakers use to miti-
gate potentially face-threatening acts.
Although these strategies are often referred to as politeness strategies, Locher
(2006) prefers the more general term relational work, and uses this term to include
not only mitigation strategies, but also involvement strategies such as bonding,
empathizing and praising. In her (2006) analysis of expert advice-giving in Lucy
Answers, Locher aimed at “showing tendencies of what kind of relational work is
expressed, and where relational work manifests itself within the answers” (2006:
125). From a close reading of the response letters, Locher identified seven cate-
gories of relational work: hedging, empathizing, boosting, bonding, praising, crit-
icizing and using humor. She found that different types of relational work tended
to be associated with different discursive moves. The involvement strategies of
empathizing, praising and bonding tended to occur with assessments, the face-sav-

Authenticated
ing strategy of hedging tended to occur with advice, and the face-threatening strat-
egies of boosting and criticizing tended to occur with the discursive moves of
advice and assessment, respectively. Locher identified a total of 1,351 relational
work moves in her data and found that although face-saving strategies (hedging)
were the most frequent (48 %), involvement strategies (37 %) and face-threatening
strategies (16 %) were also important.
Relational work is also a notable feature of advice from peers. Morrow (2006)
reported that almost all the advice messages in his data from the NetDoctor.co.uk
depression discussion forum contained some expressions of empathy, and that in
more than half of them, an expression of empathy served as an opener, (e.g. “i
really understand what you are going through …”). In addition to overt expres-
sions of empathy, which were very frequent, “the act of responding to a problem
message can in itself be seen as an expression of positive regard. Giving consid-
eration to someone else’s problem and writing a message about it demonstrates
a positive interest in the other person’s well-being and a recognition of his/her
problems as legitimate ones” (2006: 542–543). In Morrow’s study, advice-givers
also utilized face-saving strategies such as stressing the advisees’ competence or
assuring them that they were not responsible for their problems and therefore need
not feel ashamed of seeking or receiving advice. The closings of advice messages
frequently contained expressions of affect such as love, take care, all the best, and
“there were also requests for and offers of continued contact and support, […] (e.g.
‘please reply, and let me know if you’ll listen to some simple advice’)” (2006: 543).
7.5. Indirectness in advice-giving about health issues

Studies of advice-giving in CMC almost invariably note the use of indirectness.
Locher (2006) reported that hedging was the most frequent relational strategy in
the online expert advice column she studied, and indirectness is also a very notable
feature of advice from peers. Kouper (2010) analyzed the directness of peer advice
in her data from an online motherhood community using Hinkel’s (1997) set of
three categories of directness (Direct advice, Hedged advice and Indirect advice),
to which she added a fourth category (Description of personal experience). She
identified 132 pieces of advice in her data. They were distributed into the four
categories as follows: Direct advice (45), Hedged advice (30), Indirect advice (14)
and Personal experience (43). Thus, roughly two-thirds of the advice given was
not direct advice. Kouper observed that, “direct advice was common in situations
that were straightforward and could have relatively simple solutions, i.e., situa-
tions involving minimal face threat” (2010). For more sensitive matters, personal
experience was the preferred strategy: “Through personal example, advice givers
established their authority and yet avoided responsibility and shifted the agency
of advice, because they did not explicitly tell advice seekers what to do” (2010).
The use of narratives was also a preferred strategy in Harrison and Barlow’s

Authenticated
(2009) study of how people used politeness strategies when giving advice in an
online arthritis workshop. In this workshop participants had to develop weekly
action plans and give others feedback on their plans, which involved giving advice
about parts of others’ plans. Participants used a variety of strategies to mitigate
the face-threatening effect of giving advice, but the use of personal narratives was
particularly frequent and noteworthy. “The narrative feedback typically describes
how the writer of the message has addressed the same problem as the recipient,
and these narratives therefore also embody indirect advice” (2009: 93). They
concluded, “[t]his technique [using personal narratives] is successful in enabling
advice givers to avoid being prescriptive while at the same time demonstrating
empathy and shared concerns with the recipient, and leaves the recipient respon-
sible for their own decision about whether to follow the advice” (2009: 108).
Although the narratives could be interpreted as merely expressions of empathy,
there was evidence for interpreting them as advice since the recipients sometimes
responded to the narratives by expressing thanks for the “suggestions”. Morrow
(2006) also noted advice-givers’ use of narratives for establishing competence and
offering advice indirectly.
How advice-givers present themselves through narratives, or in the case of the
Lucy Answers advice column, how an identity is established for a fictional advice-
giver, is also an important aspect of advice-giving.
7.6. The health advisor’s self-presentation

Whether or not advice is accepted and followed depends in part on who is g iving
the advice. The advice column which Locher (2006) used for her study features
an expert advice-giver, Lucy. Lucy is not a real person. The advice given by Lucy
actually comes from a team of health advisors, but through the advice letters signed
by Lucy, the advice team gradually developed a persona or identity for Lucy. Lucy
needed a trustworthy persona that demonstrated expertise, but also an attractive
one that potential readers could feel comfortable with and relate to. Through the
linguistic strategies used in the advice letters published under her name a persona
was developed for Lucy by the team of advisors. Locher and Hoffmann (2006)
identified seven characteristics that contributed to the construction of Lucy’s advi-
sor identity:
1. Lucy’s name, self-reference and address terms.
2. Lucy presents herself as a competent and knowledgeable source of accurate information.
3. Lucy makes readers think and gives options when she presents her advice.
4. Lucy chooses an easily accessible, informal and inoffensive range of vocabulary.
5. Lucy has an opinion (positive and negative evaluations).
6. Lucy shows awareness of difficult situations (empathy).
7. Lucy has a sense of humor.
(Locher and Hoffmann 2006: 78)

Authenticated
In Chapter 8 of Locher’s (2006) book and also in Locher and Hoffmann (2006)
there are extensive examples and explanations illustrating how each of these char-
acteristics is realized through the linguistic choices and strategies used to present
advice in Lucy’s response letters.
The advice column Lucy Answers appears on a website sponsored by a univer-
sity health service. In this context, advice-seekers can expect to receive profes-
sional advice, and advice-givers may not feel obliged to establish their competence
(though Locher and Hoffmann see this as a characteristic of Lucy’s identity). By
contrast, in the discussion forums and other sites where peers offer advice, there is
a need for an advice-giver to demonstrate competence. Morrow (2006) and Harri-
son and Barlow (2009) have noted that one motivation for giving advice through
personal narratives is that by relating their own experience writers establish their
competence for giving advice to others.
7.7. Health advice in a Japanese discussion forum

While most of the research on advice-giving on health-related websites has been
concerned with advice in English, Morrow (2012) conducted a study of advice-giv-
ing on a Japanese discussion forum about divorce. Research on Japanese discourse
has shown that there are distinctive patterns and usages in Japanese discourse that
can be related to core cultural values that sometimes differ from those of western
cultures. Thus it could be expected that there would be differences between Japa-
nese and English in the way advice is given. Morrow’s data and methodology were
similar to those used by Locher (2006) in her study of Lucy Answers, and from
a comparison of results, Morrow identified some distinctive features of Japanese
advice-giving that could be attributed to differing cultural values. These included
the non-use of imperative forms, a strong tendency for advice-givers to express
empathy and urge advisees to be empathetic, the use of apologies, and a tendency
to refer to advisees’ social roles and to advise people to behave in ways appropri-
ate to their social roles. Some features resulted from the Japanese language itself,
particularly the use of interactional particles such as ne to set an empathetic tone
and create a sense of solidarity (2012: 276). This study illustrates how cultural
values influence the way in which advice is given and how structural features of a
language, such as interactional particles in Japanese, can serve an important func-
tion in advice discourse.
8. Advice-giving in other forms of CMC
While much of the research on advice-giving in CMC deals with advice-giving at

health-related sites, there has also been research on advice-giving in other CMC
contexts. As a part of her study on public participation media, Thornborrow (2015)

Authenticated
delves briefly into the form of advice in texts and e-mails written by participating
listeners to be read out on the air. While these texts exhibit some of the features
found in health-related advice messages, they also show differences. In particular,
[…] the many-to-one framework here on the one hand produces advice-giving that is
largely direct and unmitigated, on the other, it also displays a high degree of positive
face-work in terms of that affiliation. In their emails, texts and messages, advice-givers
often draw on their own similar experiences, taking up a position of situated expertise
at the same time as displaying empathy and strong alignment with the recipient. (2015:
169)
Another analysis of non-health-related advice-giving is that of Placencia (2012)

who studied peer-to-peer advice in Spanish using data from Yahoo!Respuestas. At
that website, users can ask questions on any topic and receive replies from other
users. A notable feature of the advice-giving on Yahoo!Respuestas was the infor-
mal style which was manifested in the use of informal pronouns and the “direct
formulations of advice” (2012: 302). While the discourse moves that were used for
giving advice were similar to those used in Locher’s (2006) study of Lucy Answers,
the peer advisors in Placencia’s study used a direct style in giving advice “rather
than, for example, asking questions in order to invite introspection […]” (2012:
302) as was often done in Lucy Answers (Locher 2006). Placencia suggested that
the use of this direct style could have been related to the topics, which were in the
general area of “beauty and style”, and that the peer advisors might have adopted
a less direct style if the topics had been more sensitive ones. She added that the
direct style could also reflect cultural influence (2012: 303).
Analysis of advice-giving was not a main focus of Arendholz’s (2013) study
of appropriate and inappropriate online behavior, but her analysis of some threads
containing advice offers an interesting contrast to the practice of advice-giving
on health-related websites. Whereas advice-giving at the health-related websites
is associated with relational strategies such as bonding and praising, Arendholz
found numerous instances of disagreement and confrontation in threads containing
advice on online message boards. However, in most of her examples, advice-giv-
ers did not engage in flaming or other inappropriate behavior, and overall there
was a lot of “thankfulness and sharing” (2013: 246). As in Placencia’s (2012)
study, it is quite possible that the use of confrontational language could be related
to the topics, which were not particularly personal or sensitive.
9. Conclusion
Although requesting and advising are treated separately in research studies, they
are related speech acts since both involve asking someone to do something. As
this is potentially face-threatening, those who make requests or give advice tend to

Authenticated
use strategies to mitigate the potential face-threat. The use of such strategies can
be crucial to carrying out these speech acts successfully, and therefore researchers
have been very interested in the strategies that are used in the performance of these
speech acts. There is great diversity in the types of requests that are made and
advice that is given, and there is also much diversity in the ways that these speech
acts are carried out and in the contexts in which they take place. It follows then that
there is also considerable diversity in strategies used to carry out these speech acts.
The diversity of contexts and strategies is very much a feature of requesting and
advising in CMC, as is attested by the studies reviewed in this survey.
While requesting and advising occur in all forms of CMC, research on these
speech acts in CMC has tended to concentrate on a few areas. With regard to
requests, there have been many studies on e-mail requests, particularly in aca-
demia but also in professional settings. As for advice-giving, much of the research
has investigated advice-giving on health-related websites. In these areas, various
methodologies and approaches have been used, and they have yielded a variety of
findings. They have also raised questions for future research.
Many studies of requests deal with e-mail requests by students to professors.
Making such requests might seem like a simple and straightforward matter, yet
NNS students experience considerable difficulty in formulating requests appro-
priately. Analysis of NNS students’ requests shows that making such requests is,
in fact, a complex task requiring a high level of pragmatic competence. Students
need to modify their request so as to express an appropriate level of formality and
they may need to provide appropriate supporting moves to justify their request.
The use of salutations, address forms, closings and the amount and sequencing
of information can also affect the recipient’s perception of the request and will-
ingness to comply with it. However, students often receive no training in writing
e-mails or making such requests, and seldom have models to follow. Furthermore,
there are not established standards for language use in e-mails and individuals may
have different expectations about what level of politeness is appropriate and how
it should be expressed.
Research on e-mail requests in professional settings has been concerned with
investigating how requests can be made strategically through e-mail, especially
how they can be formulated so as to get a reply. In this connection the use of car-
bon-copying (cc-ing), the number of addressees, the identity adopted by the writer,
the use of intertextuality and interdiscursivity have all been identified as having an
influence on the success of an e-mail request.
The studies of requests have produced interesting and insightful findings, but
even in the much-researched area of students’ e-mail requests questions remain.
For instance, how much do students modify their requests depending on the indi-
vidual professor to whom they are writing? Most studies have dealt with requests
made to one or two professors, but students might write differently to professors
depending on the professors’ gender and age, for example. Also, more longitudinal

Authenticated
studies along the lines of C. Chen (2006) could enlarge our understanding of the
process by which students acquire pragmatic competence in a second language.
This could have important implications for second language teaching.
Research has dealt with requests in SNS and other formats, but much of
the research has been on e-mail requests. Further research on different types of
requests by different groups of users in other forms of CMC is needed to identify
the characteristics associated with each form. The results from such studies could
be useful in identifying the general characteristics of requesting in CMC, if indeed
such general characteristics exist.
Similar observations can be made about research on advice-giving in CMC.
Particularly in the area of health-related issues, research has contributed signifi-
cantly to an understanding of some of the main aspects of advice-giving, such as
the solicitation of advice, the topics of advice, the relational work associated with
advice, the linguistic structures used to express advice, and the self-presentation
of the advice-giver. However, there has been much less research on advice-giving
on non-health-related topics. Research on advice-giving in other forms of CMC
is needed for a broader understanding of the practice of advice-giving and of the
shape and form of advice in CMC.
In CMC as in other forms of communication, advice is given in specific lan-
guages and in specific cultural contexts, and these linguistic and cultural factors
influence the practice of advice-giving. For example, although advice-giving has
been seen as face-threatening in English-speaking cultures, it is not necessarily
seen that way everywhere. Hinkel notes that “Chinese frequently view advice-giv-
ing as a mark of solidarity, concern, and sincere interest” (1997: 16). Thus, lin-
guistic and cultural factors need to be taken into account in examining the practice
of advice-giving. There is a need for further study of the influence of these factors
and for research findings to be integrated into explanations of advice-giving on
CMC.
Another issue pertaining to advice in CMC relates to what constitutes advice or
a request for advice. Researchers (e.g. Harrison and Barlow 2009; Morrow 2006)
have noted how personal narratives are used both to elicit and to give advice. Can
such narratives be considered indirect requests for advice or indirect advice? Under
what conditions do people interpret them in these ways? Similarly, the effect of
topic on advice-giving needs to be explored more deeply. CMC is well-suited to
advice-giving on sensitive topics because of its anonymity, but the sensitivity of
some topics can be expected to influence the form of advice associated with them.
Requesting and advice-giving are complex speech acts. The research surveyed
here has made significant progress in explicating how these speech acts are carried
out in CMC, but many questions remain, leaving considerable scope for further
study.

Authenticated
References
Arendholz, Jenny
Relations: Amsterdam/Philadelphia: Benjamins.
Arguello, Jaime, Brian Butler, Elisabeth Joyce, Robert Kraut, Kimberly S. Ling and Xiao-
qing Wang
2006 Talk to me: Foundations for successful individual-group interactions in online
communities. In: Proceedings CHI 2006, 959–968. ACM Press. https://ils.unc.
edu/~jarguell/ArguelloCHI06.pdf.
Ballentine, Paul D. and Rachel J. Stephenson
2011 Help me, I’m fat! Social support on online weight loss networks. Journal of
Consumer Behavior 10: 332–337.
Bambina, Antonina D.
2007 Online Social Support: The Interplay of Social Networks and Computer-Medi-
ated Communication. New York: Cambria Press.
Barron, Greg and Eldad Yechiam
2002 Private e-mail requests and the diffusion of responsibility. Computers in
Human Behavior 18: 507–520.
Biesenbach-Lucas, Sigrun
2005 Communication topics and strategies in e-mail consultation: Comparison
between American and international university students. Language Learning
and Technology 9(2): 24–46.
2006 Making requests in e-mail: Do cyber-consultations entail directness? Toward
conventions in a new medium. In: Kathleen Bardovi-Harlig, J. César
Félix-Brasdefer and Alwiya S. Omar (eds.), Pragmatics and Language Learn-
ing. Vol. 11, 81–107. Honolulu: University of Hawaii Press.
2007 Students writing emails to faculty: An examination of e-politeness among
native and non-native speakers of English. Language Learning and Technol-
ogy 11(2): 59–81.
Bloch, Joel
2002 Student/teacher interaction via email: The social context of Internet discourse.
Journal of Second Language Writing 11: 117–134.
Blum-Kulka, Shoshana, Juliane House and Gabriele Kasper (eds.)
1989 Cross-Cultural Pragmatics: Requests and Apologies. (Vol. XXXI in the series
Advances in Discourse Processes, Roy O. Freedle, ed.) Norwood, NJ: Ablex.
1978 Universals of language use: Politeness phenomena. In: Esther Goody (ed.),
Questions and Politeness, 56–324. Cambridge: Cambridge University Press.
versity Press.
Burke, Moira, Elisabeth Joyce, Tackjin Kim, Vivek Anand and Robert Kraut
2007 Introductions and questions: Rhetorical strategies that elicit response in online
communities. Pittsburgh, PA: Carnegie Mellon University. http://repository.
cmu.edu/hcii/89.

Authenticated
Candlin, Christopher and Yon Maley

1997 Intertextuality and interdiscursivity in the discourse of alternative dispute res-
olution. In: Britt-Louise Gunnarson, Per Linell and Bengt Nordberg (eds.),
The Construction of Professional Discourse, 201–222. London: Longman.
Chang, Yu-Ying and Yi-Ping Hsu
1998 Requests on e-mail: A cross-cultural comparison. RELC Journal 29(2): 121–
151.
Chapman, Rebecca
2007 English for Emails. Oxford: Oxford University Press.
Chejnová, Pavla
Chen, Chi-Fen Emily
2000 Interlanguage requesting behavior in e-mail: A study of Taiwanese students’
English written requests to U.S. professors. Paper presented at the 7th Interna-
tional Pragmatics Conference, Budapest, Hungary.
Chen, Chi-Fen Emily
2001 Making e-mail requests to professors: Taiwanese vs. American students. Paper
presented at the Annual Meeting of the American Association for Applied Lin-
guistics, St. Louis, MO. (ERIC Document Reproduction Service No. ED 461
299)
Chen, Chi-Fen Emily
2006 The development of e-mail literacy: From writing to peers to writing to author-
ity figures. Language Learning and Technology 10(2): 35–55.
Chen, Yuan-shan
2015a Chinese learners’ cognitive processes in writing email requests to faculty. Sys-
tem 52: 51–62.
Chen, Yuan-shan
2015b Developing Chinese EFL learners’ email literacy through requests to faculty.
Economidou-Kogetsidis, Maria
2011 “Please answer me as soon as possible”: Pragmatic failure in non-native
speakers’ e-mail requests to faculty. Journal of Pragmatics 43: 3193–3215.
Efron, Miles and Megan Winget
2010 Questions are content: A taxonomy of questions in a microblogging environ-
ment. Proceedings of the American Society for Information Science and Tech-
nology 47(1):1–10.
Ellison, Nicole B., Rebecca Gray, Jessica Vitak, Cliff Lampe and Andrew T. Fiore
2013 Calling all Facebook friends: Exploring requests for help on Facebook. Pro-
ceedings of the Seven International AAAI Conference on Weblogs and Social
Media, 155–164.
Fairclough, Norman
1992 Discourse and Social Change. Oxford: Blackwell.
Ford, Shawn
2003 “Dear Mr. Shawn”: A lesson in e-mail pragmatics. TESOL Journal 12: 39–40.
Ford, Shawn
n.d. The use of pragmatics in e-mail requests made by second language learners of
English. http://www2.hawaii.edu/~sford/research/jsls5.pdf.

Authenticated
Goffman, Erving
1967 Interaction Ritual: Essays on Face-to-Face Behavior. New York: Pantheon
Books.
Golato, Andrea and Carmen Taleghani-Nikazm
2006 Negotiation of face in web chats. Multilingua 25: 293–321.
Goldsmith, Daena J.
2000 Soliciting advice: The role of sequential placement in mitigating face threat.
Communication Monograph 67(1): 1–19.
Gray, Rebecca, Nicole B. Ellison, Jessica Vitak and Cliff Lampe
2013 Who wants to know? Question-asking and answering practices among Face-
book users. In: Computer Supported Cooperative Work and Social Com-
puting (CSCW) 2013. https://www.msu.edu/~nellison/GrayEllisonVitak-
Lampe2013CSCW.pdf.
Harper, F. Maxwell, Daphne Raban, Sheizaf Rafaeli and Joseph A. Konstan
2008 Predictors of answer quality on online Q&A sites. In: Proceedings CHI, 865–
884. http://www-users.cs.umn.edu/~harper/publications/harper-chi2008.pdf.
Harrison, Sandra and Julie Barlow
2009 Politeness strategies and advice-giving in an online arthritis workshop. Jour-
nal of Politeness Research 5: 93–111.
Hartford, Beverly S. and Kathleen Bardovi-Harlig
1996 At your earliest convenience: A study of written student requests to faculty. In:
Lawrence F. Bouton (ed.), Pragmatics and Language Learning. Monograph
Series Volume 7, 55–69. Urbana, IL: Division of English as an International
Language, University of Illinois.
Harvey, Kevin, Miriam A. Locher and Louise Mullany
2013 “Can I be at risk of getting AIDS?” A linguistic analysis of two Internet col-
umns on sexual health. Linguistik Online 59(2).
Hassall, Tim
1999 Request strategies in Indonesian. Pragmatics 9: 585–606.
Hendriks, Berna
2010 An experimental study of native speaker perceptions of non-native request
modification in e-mails in English. Intercultural Pragmatics 7(2): 221–255.
Heritage, John and Sue Sefi
1992 Dilemmas of advice: Aspects of the delivery and reception of advice in inter-
actions between health visitors and first-time mothers. In: Paul Drew and John
Heritage (eds.), Talk at Work, 359–417. Cambridge: Cambridge University
Press.
Hinkel, Eli
1997 Appropriateness of advice: DCT and multiple choice data. Applied Linguistics
18(1): 1–26.
Ho, Victor
2010 Constructing identities through e-mail requests. Journal of Pragmatics 42:
2253–2261.
Ho, Victor
2011 What functions do intertextuality and interdiscursivity serve in request e-mail
discourse? Journal of Pragmatics 43: 2534–2547.

Authenticated
Kouper, Inna
2010 The pragmatics of peer advice in a LiveJournal community. Language@Inter-
net 7, article 1 (urn: nbn:de:0009-7-24642). http://www.languageatinternet.
org/articles/2010/2464.
Limberg, Holger and Miriam A. Locher (eds.)
2012 Advice in Discourse. Amsterdam/Philadelphia: Benjamins.
Lippka, Yannick, Oliver Patschan, Tilmann Todenhöfer, Christian Schwentner, Andreas
Gutzeit, Axel S. Merseburger and Marcus Horstmann
2013 Bladder cancer discussed on the internet: A systematic analysis of gender dif-
ferences of initial posters on an online discussion board. SpringerPlus 2: 445.
Locher, Miriam A. and Sebastian Hoffmann
2006 The emergence of the identity of a fictional expert advice-giver in an Ameri-
can Internet advice column. Text and Talk 26(1): 69–106.
Locher, Miriam A.
2006 Advice Online: Advice-giving in an American Internet Health Column. Amster-
Locher, Miriam A.
2010 Health Internet sites: A linguistic perspective on health advice columns. Social
Semiotics 20(1): 43–59.
Locher, Miriam A.
2013 Internet advice. In: Susan C. Herring, Dieter Stein and Tuija Virtanen (eds.),
Pragmatics of Computer-Mediated Communication. Handbooks of Pragmatics
vol. 9, 339–362. Berlin/Boston: de Gruyter.
tion in Britain and Australia. Journal of Pragmatics 44: 1077–1098.
Mohammadi, Masoomeh and Gholam Reza Zarei
2012 In-directness and politeness in Iranian Persian and English electronic requests
to faculty. Elixir Literature 53C: 11785–11795.
Morris, Meredith Ringel, Jaime Teevan and Katrina Panovich
2010 What do people ask their social networks, and why? A survey study of status
message Q&A behavior. CHI 2010. Proceedings of the SIGCHI Conference
on Human Factors in Computing Systems, Atlanta, Georgia. http://dl.acm.org/
citation.cfm?id=1753587.
Morrow, Phillip R.
2006 Telling about problems and giving advice in an Internet discussion forum:
Some discourse features. Discourse Studies 8(4): 531–548.
Morrow, Phillip R.
Paul, Sharoda A., Lichan Hong and Ed H. Chi
2011 Is Twitter a good place for asking questions? A characterization study. Pro-
ceedings of the Fifth International AAI Conference on Weblogs and Social
Media, 578–581. http://www.aai.org/ocs/index.php/ICWSM/ICWSM11/
paper/viewFile/2813/3225.

Authenticated
Phillips, Steven R. and Eric M. Eisenberg

1993 Strategic uses of electronic mail in organizations. The Electronic Journal of
Communication 3(2). http://www.cios.org/EJCPUBLIC/003/2/003212.HTML.
Placencia, María Elena
2012 Online peer-to-peer advice in Spanish Yahoo!Respuestas. In: Holger Limberg
and Miriam A. Locher (eds.), Advice in Discourse, 281–305. Amsterdam/Phil-
Rainie, Lee, Aaron Smith and Maeve Duggan
2013 Coming and going on Facebook. Pew Internet and American Life Project,
Washington, DC. http://pewinternet.org/Reports/2013/Coming-and-going-on-
facebook.aspx.
Rue, Yong-Ju and Grace Qiao Zhang
2008 Request Strategies: A Comparative Study in Mandarin Chinese and Korean.
Searle, John R.
1969 Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge
University Press.
Sillence, Elizabeth
2013 Giving and receiving peer advice in an online breast cancer support group.
Cyberpsychology, Behavior, and Social Networking 16: 480–485.
Simon, Bernd
2004 Identity in Modern Society: A Social Psychological Perspective. Oxford:
Blackwell.
Suzuki, Lalita K. and Jerel P. Calzo
2004 The search for peer advice in cyberspace: An examination of online teen bulle-
tin boards about health and sexuality. Applied Developmental Psychology 25:
685–698.
Teevan, Jaime, Meredith Ringel Morris and Katrina Panovich
2011 Factors affecting response quantity, quality, and speed for questions asked via
social network status messages. Proceedings of the Fifth International AAAI
Conference on Weblogs and Social Media, 630–633.
Thornborrow, Joanna
2015 The Discourse of Public Participation Media: From Talk Show to Twitter.
London/New York: Routledge.
Tseng, Chia-Ti Heather
2015 “You must let me pass, please!”: An investigation of email request strategies
by Taiwanese EFL learners. Journal of ELT and Applied Linguistics 3(1):
11–28.
Van Mulken, Margot
1996 Politeness markers in French and Dutch requests. Language Sciences 18(3–4):
689–702.
Walker, Traci
2013 Requests. In: Maria Sbisá and Ken Turner (eds.), Pragmatics of Speech Actions.
Handbooks of Pragmatics, vol. 2, 445–466. Berlin/Boston: de Gruyter.
Wang, Yi-Chia, Robert E. Kraut and John M. Levine
2015 Eliciting and receiving online support: Using computer-aided content analysis
to examine the dynamics of online social support. Journal of Medical Internet
Research 17(4): e99. doi: 10.2196/jmir.3558.

Authenticated
Zhu, Wuhan
2012 Polite requestive strategies in emails: An investigation of pragmatic compe-
tence of Chinese EFL learners. RELC Journal 43(2): 217–238.

Authenticated
Authenticated
About the authors
Jenny Arendholz is a Senior Lecturer at the Department of Anglistik and Ameri-

kanistik at Ludwig-Maximilians University Munich, Germany. She holds a Ph.D.
in English Linguistics, which she received at Augsburg University in 2011 with her
monograph (In)Appropriate Online Behavior. A Pragmatic Analysis of Message
Board Relations published at John Benjamins. She is also a co-editor of the reader
The Pragmatics of Quoting Now and Then (de Gruyter Mouton, 2015). Her main
research interests include (interpersonal) pragmatics, identity construction, com-
puter-mediated communication, multimodality, (hyper-)text linguistics and syntax.
Monika Bednarek is Associate Professor in Linguistics at The University of Syd-

ney, Australia. She researches language use in the mass media, with particular
interests in news (Evaluation in Media Discourse, 2006; News Discourse, 2012,
with co-author Helen Caple) and television series (The Language of Fictional
Television, 2010; Telecinematic Discourse, 2011, with co-editors Roberta Piazza
and Fabio Rossi). She is currently working on a corpus linguistic project on tele-
vision dialogue, analyzing the Sydney Corpus of TV Dialogue. Another research
interest concerns the relationship between language and emotion (Emotion Talk
Across Corpora, 2008).
Birte Bös is Professor of English Linguistics at the University of Duisburg-Essen,

Germany. Her research interests include synchronic and diachronic pragmatics,
discourse analysis and media linguistics. She has investigated the communicative
practices of historical and modern news discourse and digital discourse. She is
the co-author of News as Changing Texts. Corpora, Methodologies and Analysis
(2015) and the co-editor of Changing Genre Conventions in Historical English
News Discourse (2015).
Brook Bolander is Assistant Professor in the School of English at the University

of Hong Kong. Her major research interests include language use online, the socio
linguistics of globalisation, language use and space, and English as a transnational
language. She has published a monograph on Language and Power in Blogs (John
Benjamins, 2013); and together with Miriam A. Locher, she has worked and pub-
lished on language and identity construction in Facebook, and on research meth-
odology for the study of language use online. She has also co-edited (with Miriam
A. Locher and Nicole Höhn), a special issue on “Relational Work in Facebook and
Discussion Boards/Fora” (2015, Pragmatics 25.1).

Authenticated
692 About the authors
Wolfram Bublitz is Professor Emeritus of English Linguistics at the University

of Augsburg, Germany. He received his doctorate from the University of Hamburg
and his habilitation from the University of Trier, Germany. He has also been Full
Professor at the University of Braunschweig (Germany), Exchange Professor at
Northern Illinois University (USA) and Visiting Professor at the Universities of
Pittsburgh (USA) and Hamburg (Germany). His research interests concern most
areas of text analysis, (constructivist) pragmatics and computer-mediated commu-
nication. His (edited) books include The Pragmatics of Quoting Now and Then
(with J. Arendholz and M. Kirner-Ludwig, 2015), Foundations of Pragmatics
(with N. Norrick, 2011), Englische Pragmatik (2nd ed., 2009), Metapragmatics
in Use (with A. Hübler, 2007), Coherence in Spoken and Written Discourse (with
U. Lenk and E. Ventola, 1999).
Marta Dynel is Associate Professor in the Department of Pragmatics at the Uni-

versity of Łódź, Poland. Her research interests are primarily in pragmatic and cog-
nitive mechanisms of humour, neo-Gricean pragmatics, the pragmatics of inter-
action, (im)politeness theory, the philosophy of overt and covert untruthfulness
(irony and deception), as well as the methodology of research on film discourse.
She has published internationally in linguistic journals and volumes, contributing
over 70 articles in the space of the past 10 years. She also authored Humorous Gar-
den-Paths: A Pragmatic-Cognitive Study (Cambridge Scholars Publishing, 2009)
and edited several Special Issues for JCR journals, and volumes: The Pragmatics
of Humour across Discourse Domains (John Benjamins, 2011), Developments in
Linguistic Humour Theory (John Benjamins, 2013), as well as Participation in
Public and Social Media Interactions (with Jan Chovanec, John Benjamins, 2015).
Volker Eisenlauer is a Postdoctoral scholar and lecturer at the University of Salz-

burg, Austria, Department of English Linguistics. Since 2016 he is also Research
and Development Manager at the higher education technology company xStudy SE,
Munich, Germany. His research focuses on interdisciplinary issues at the interface
between media linguistics, literacy research and pragmatics. He published widely
on Social Media Linguistics (Storytelling in Weblogs, 2010; A Critical Hypertext
Analysis of Social Media, 2013; Facebook as a Third Author, 2014) and mobile
learning (Multimodality in Mobile-Assisted Language Learning, 2014; Englisch
zum Selberbasteln, 2017) and edited a special issue on media participation (Par-
ticipation Framework Revisited: (New) Media and their Audiences/Users. Journal
of Pragmatics, 2014). In a joint project with the online study information portal
‘studieren.de’ he is currently working on the development and implementation of
a mobile app (module.org) supporting academic planning and study organization.

Authenticated
About the authors 693
Elisabeth Fritz is a Research Assistant and Ph.D. candidate in English Linguistics

at the University of Augsburg, Germany. Her research interests include contrastive
linguistics, translation and computer-mediated communication. She is completing
her doctoral dissertation on strategies and patterns of interactional topic manage-
ment in online discussion forums.
Maximiliane Frobenius is a post-doc at Hildesheim University, Germany. Her

main research areas are computer-mediated communication and learner language.
She has published on YouTube video blogs and their comments, Facebook com-
menting sections, and (written) blogs. Her work includes The Pragmatics of Mon-
ologue: Interaction in Video Blogs, 2014, “Beginning a monologue: The opening
sequence of video blogs”, 2011, “Tying in comment sections: The production of
meaning and sense on Facebook”, 2015, and “Pointing gestures in video blogs”,
2013. She edited Culinary Linguistics: The Chef ’s Special (2014, together with
Cornelia Gerhardt and Susanne Ley) and a special issue of the Journal of Pragmat-
ics: Participation Framework Revisited: (New) Media and Their Audiences/Users
(2014, together with Volker Eisenlauer and Cornelia Gerhardt). Her current work
involves the study of naturally occurring speech of advanced learners of English.
Cornelia Gerhardt works as a Lecturer of linguistics at the English Department of

Saarland University, Germany. Her main research areas include the appropriation
of media discourse as well as language at the interface with social phenomena like
football or food. She published a monograph Appropriating Live Televised Foot-
ball through Talk (2014) as well as two edited volumes with different co-editors:
The Appropriation of Media in Everyday Life (2012, together with Ruth Ayass)
and Culinary Linguistics: The Chef ’s Special (2014, together with Maximiliane
Frobenius and Susanne Ley). Also, she edited a special issue of the Journal of
Pragmatics: Participation Framework Revisited: (New) Media and their Audi-
ences/Users (2014, together with Volker Eisenlauer and Maximiliane Frobenius).
She is currently pursuing a project on “Multimodality and Embodied Interaction”
as principal investigator (together with Elisabeth Reber) financed by the German
Science Foundation (DFG).
Sage L. Graham is an Associate Professor of Linguistics in the Department of

English at the University of Memphis in Memphis, TN, USA. Using language as
an analytical lens, her research explores causes and effects of misunderstanding,
conflict, and impoliteness in digital contexts, patterns of identity construction and
gender in religious contexts, and professional communicative practices in on- and
offline medical and social work settings. Using the perspective of Interpersonal
Pragmatics (Interpersonal Pragmatics, Locher and Graham, eds., de Gruyter
2010), her recent research has begun to address multimodal elements of commu-
nication in digital contexts, examining the interplay between synchronous written

Authenticated
communication (e.g. chat interaction) and simultaneous audio and video feeds. In
addition to chapters published in several edited volumes, she has published articles
in the Journal of Pragmatics, Technical Communication Quarterly, and the Jour-
nal of Politeness Research.
Claire Hardaker is a Lecturer in Forensic Corpus Linguistics at Lancaster Uni-

versity, England. She researches online aggression, deception, and manipulation.
In particular, she has researched and worked alongside Twitter in pursuit of a safer
internet. Claire is currently working on a number of projects including the analysis
of online aggression within the ESRC-funded Centre for Corpus Approaches to
Social Science (CASS), the EC-funded Study on comprehensive policy review
of anti-trafficking projects funded by the EC, the development of tools to make
analysis of online data more accessible to non-programmers as part of the ESRC-
funded Twitter rape threat and the discourse of online misogyny project, and the
HMG-funded Native Language Influence Detection 6 project. She is writing two
books: The Antisocial Network (Palgrave) and Trolling: Aggression, Deception
and Manipulation Online (OUP).
Theresa Heyd is Deputy Professor of English Linguistics at the University

of Duisburg-Essen, Germany. She received her Ph.D. in English from Hein-
rich-Heine-Universität Düsseldorf with a work on Digital Folklore, published
with Benjamins in 2008. Since then, she has worked on different aspects of digital
linguistic practice, including linguistic innovation and change through mediated
discourse and the convergence of online/offline practice. Her more recent work
considers digital linguistic practice as one central factor in a broader framework
of the sociolinguistics of mobility. In a DFG (German Science Foundation) funded
project at the Albert-Ludwigs-Universität Freiburg, Germany, this connection was
analyzed in a large-scale corpus project on Digital Diasporas and the effect of dig-
ital communication on globalizing contact varieties such as Nigerian Pidgin. In her
current work, she explores effects of sociolinguistic convergence between online
and offline practice in Anglophone communities of Berlin, Germany.
Christian R. Hoffmann is Senior Lecturer for English Linguistics at the Univer-

sity of Augsburg, Germany. He has worked on narratives in old and new media
(Narratives Revisited, 2011) as well as on the pragmatics of digital communica-
tion (Cohesive Profiling. Meaning and Interaction in Personal Weblogs, 2012).
Currently, he is editing a collective volume describing new trends in the pragmat-
ics of telecinematic discourse. His research efforts are largely devoted to com-
piling a comprehensive, cross-generic corpus of annotated television transcripts
(How Hollywood speaks. A Film Corpus Project), while writing a monograph on
the pragmatics of film practice (How to Do Things In Films. The Pragmatics of
Telecinematic Discourse).

Authenticated
Marjut Johansson is a Professor of French Language at the School of Languages

and Translation Studies, University of Turku, Finland. Her current research inter-
est covers digital interaction and discourse, such as online discussions, digital
news discourse, and video discourse. Her theoretical frameworks are based on
sociopragmatics, interaction, and digital discourse analyses, and at the moment,
she is interested in multimodal, digital humanities, and mixed method approaches.
She has recently co-edited a special issue of Digital Agora (with S. Kleinke and
L. Lehti, Discourse, Context and Media, 2017), and a volume on Language in the
Net. Introduction to the Digital Interaction (in Finnish with Helasvuo and Tanska-
nen, 2014).
Sonja Kleinke is a Professor in English Linguistics at the English Department of

Heidelberg University, Germany. She has worked and published in the fields of
Cognitive Linguistics, English verb complementation, cognitive pragmatics, cog-
nitive metonymy and the pragmatics of computer-mediated discourse. Her current
research interests include Cognitive Linguistics, Pragmatics, and Internet Commu-
nication. She is co-editor of Cognitive Explorations into Metaphor and Metonymy
(2014).
Daniela Landert is a researcher in English linguistics at the University of Zurich,

Switzerland. Her research interests include historical pragmatics, corpus pragmat-
ics, stance, mass media communication and computer-mediated communication.
She is currently working on a project in historical corpus pragmatics, in which she
studies epistemic and evidential stance in Early Modern English. In this project,
she investigates stance expressions on the basis of several Early Modern English
corpora and she explores new methods for analysing pragmatic meaning in cor-
pora. Previously, she worked on public and private aspects of mass media com-
munication. As part of this research, she developed a model that differentiates and
integrates various dimensions of personalisation in online news (Personalisation
in Mass Media Communication, 2014).
Andreas Langlotz is Privatdozent (Reader) in English and General Linguistics

at the University of Basel, Switzerland and he teaches English and German at a
secondary school. His early research interests in cognitive linguistics and phra-
seology are reflected in the monograph Idiomatic Creativity (John Benjamins,
2006). More recently he has worked on integrating the social and cognitive dimen-
sions of discourse and interaction with a focus on interpersonal pragmatics, verbal
humour, conflict talk, and language and emotion. His socio-cognitive theory of
social meaning is presented in Creating Social Orientation Through Language
(John Benjamins, 2015).

Authenticated
Miriam A. Locher is Professor of the Linguistics of English at the University of

Basel, Switzerland. In her work she combines an interest in interpersonal prag-
matics, relational work, (im)politeness, and advice-giving in health and comput-
er-mediated communication (CMC) contexts. Her publications include Power
and Politeness in Action (2004), Advice Online (2006) and the edited collections
Impoliteness in Language (2008, with D. Bousfield), Standards and Norms of the
English Language (2008, with J. Strässler), Interpersonal Pragmatics (2010, with
S. Graham), Teaching and Learning (Im)politeness (2015, with B. Pizziconi) and
Narrative Matters across Disciplines in Medical Contexts (2015, with F. Gygax)
as well as special issues on CMC and politeness (Journal of Politeness Research 6,
Pragmatics, with N. Höhn and B. Bolander). She is currently working on an SNF
project on Language and Health Online and on the connection between emotional
display and relational work with A. Langlotz.
Amanda Lower is a PhD student in the Department of Cultures and Languages,

at Birkbeck, University of London, UK. Her research interests lie in sociocultural
pragmatics, variational pragmatics and computer-mediated discourse, with refer-
ence to social media in particular. Her thesis compares the complimenting habits
on Facebook of young people across two varieties of Spanish.
Carmen Maíz-Arévalo is Professor of Pragmatics and English at the Com-

plutense University of Madrid, Spain, having obtained her PhD in English Lin-
guistics in 2001. Her fields of interest are mainly pragmatics and intercultural
pragmatics; more specifically, speech act theory and verbal (im)politeness in
computer-mediated communication. Her most recent publications include the
articles: “Jocular mockery in computer-mediated communication: A contrastive
study of a Spanish and English Facebook community” in the Journal of Politeness
Research (2015), “Just click ‘Like”’: Computer-mediated responses to Spanish
compliments”, published in the Journal of Pragmatics in 2013, “‘Was that a com-
pliment?’ Implicit compliments in English and Spanish”, also in the Journal of
Pragmatics in 2012 or “‘You look terrific!’ Social evaluation and relationships in
online compliments”, co-authored with Antonio García-Gómez and published in
Discourse Studies in 2013. Besides her research and teaching, she is currently the
academic secretary of the Department of English Linguistics at the Complutense
University.
Phillip R. Morrow is a Professor of English and linguistics at Nagoya Gakuin

University in Nagoya, Japan. He received a Ph.D. in linguistics from the Uni-
versity of Illinois, USA, and has done research in the areas of discourse analysis,
World Englishes (particularly Japanese English), the teaching and use of Eng-
lish in Japan and corpus linguistics. His current research interests focus on the
discourse of advice, computer-mediated communication and discourse patterns in

Authenticated
Japanese. His research has been published in World Englishes, Discourse Studies,
English for Specific Purposes, and various Japanese journals.
Andreas Musolff is Professor of Intercultural Communication at the University of

East Anglia in Norwich, UK. His main research areas include Figurative Language
Use in Press Media and Computer-mediated Communication, Immigration Dis-
course, Intercultural Communication and the History of Pragmatics. He has pub-
lished monographs on Political Metaphor Analysis (2016), Metaphor, Nation and
the Holocaust (2010), Metaphor and Political Discourse (2004), edited volumes
on Metaphor in Intercultural Communication (2014) and Contesting Europe’s
Eastern Rim: Cultural Identities in Public Discourse (2010) as well as numerous
articles and book chapters on metaphor theory and discourse analysis.
Ruth Page is a Senior Lecturer in the Department of English Language and

Applied Linguistics at the University of Birmingham, UK. Her research inter-
ests lie in sociolinguistic approaches to narrative, language and gender, and new
media. She has published extensively in all three fields and is the author of Sto-
ries in Social Media (Routledge, 2012) and Literary and Linguistic Approaches to
Feminist Narratology (Palgrave, 2006).
Stephen Pihlaja is Senior Lecturer in Stylistics at Newman University, Birming-

ham, UK. His monograph, Antagonism on YouTube (Bloomsbury, 2014), looked
at the use of metaphor in dialogue among Christians and atheists on YouTube. His
current project investigates evangelical talk among atheists, Christians, and Mus-
lims on social media.
María Elena Placencia is Reader in Linguistics at Birkbeck, University of Lon-

don, UK. She has published extensively on language and social interaction in a
range of sociocultural contexts in the Spanish- and English-speaking worlds. Her
recent work explores features of relational work online and how communicative
practices from face-to-face interaction translate in virtual environments including,
for example, complimenting behaviour on Facebook and Instagram, advice-giving
on YahooAnswers, and bargaining in e-service encounters in MercadoLibre, a vir-
tual marketplace.
Christoph Schubert is a Full Professor of English Linguistics at the University

of Vechta, Germany. His main publications are a PhD thesis on the complex sen-
tence in English poetry (Peter Lang, 2000), a postdoctoral study on the cognitive
constitution of space in descriptive text types (Niemeyer, 2009), and a book-length
introduction to English text linguistics (Erich Schmidt, 2nd ed. 2012). He is co-
editor of a special issue of the Journal of Language and Politics titled Cogni-
tive Perspectives on Political Discourse (2014) and of two forthcoming collective

Authenticated
volumes, Variational Text Linguistics: Revisiting Register in English (de Gruyter

Mouton) and Pragmatic Perspectives on Postcolonial Discourse: Linguistics and
Literature (Cambridge Scholars). His main research areas are discourse analysis,
text linguistics, pragmatics, stylistics, and cognitive linguistics.
Michele Zappavigna is a Lecturer in the School of the Arts and Media at the
University of New South Wales, Australia. Her major research interest is the lan-
guage of social media as reflected in her most recent books: Discourse of Twitter
and Social Media (Bloomsbury, 2012) and, with R. Page, J. Unger and D. Barton,
Researching the Language of Social Media (Routledge, 2014). She is currently
working on a book on the discourse of social tagging which will be published in
2017.

Authenticated
Name index
Name index
A Antaki, Charles 409, 423, 527, 532, 540

Abbasi, Ahmed 498, 513 Ardington, Angela 615–616, 626
Abedin, Zainul 415, 424 Arendholz, Jenny 10, 24, 36, 54, 65, 85, 94,
Abrahamsen, Adele 355, 375 102, 110, 112, 125, 135–136, 140–143,
Abrol, Satyen 242 152, 173, 239, 287, 292, 296–302,
Achugar, Mariana 637, 656 307–308, 387, 397, 413–414, 423, 465,
Acquisti, Alessandro 91–92, 115, 231, 240 470, 472, 486, 494, 513, 583, 681, 684
Adami, Elisabetta 70, 74, 77, 179, 183, Arguedas, Maria Estellés 281, 314
190–191, 194, 318, 323, 327–328, 339, Arguello, Jaime 672, 684
531, 540 Armstrong, Arthur 448, 454
Adams, Brett 456 Arnold, Jill 617, 626
Adato, Albert 279, 281, 307 Aronson, Elliot 596
Adedoyin-Olowe, Mariam 227, 239 Artemeva, Natasha 202, 217
Adnan, Muhammad 203, 217 Arundale, Robert B. 410, 423, 511, 513
Agrawal, Megha 453 Atkinson, Jane 501, 513
Ahn, June 346, 375 Attiaoui, Dorra 516
Akehurst, Lucy 512, 513 Attrill, Alison 582–583, 598
Allan, Stuart 39, 54 Auer, Peter 183, 194
Allen, Ashley Batts 602 Auerbach, David 250, 267
Allen, Nicholas B. 579, 600 Augustson, Erik 218
Alloway, Ross G. 345, 375 Avgerinakou, Anthi 500, 513
Alloway, Tracy P. 345, 375 Ayaß, Ruth 72, 77, 547, 569
Almeida, Virgilio 218 Aysenur, Kilic 586, 600
Altman, Irwin 579, 598
Alzahabi, Reem 346, 375 B
Amandi, Analía 611, 620, 630 Bach, Kent 511, 513, 636, 656
Ambady, Nalini 582, 587, 606 Badger, Meredith 156, 168
Anand, Pranav 159, 168 Baker, Andrea J. 583, 598
Anand, Vivek 684 Baker, Paul 203, 217, 418, 423, 497–498,
Anderson, Jeffrey F. 250–252, 267 500, 502
Anderson, Kristin L. 611, 617–618, 625 Bakhtin, Michail M. 66, 77, 187, 194
André, Paul 514 Ballentine, Paul D. 673, 684
Andrejevic, Mark 92, 112 Bamberg, Michael G.W. 409, 423, 526, 540
Androutsopoulos, Jannis 2, 20, 24, 85, Bambina, Antonina D. 673, 684
109, 132, 142, 174–175, 179, 183–184, Bamford, Holden 378
194, 217, 227, 239, 256, 267, 286, 307, Bamman, David 203, 217
324, 339, 387, 397, 444, 448, 452, 495, Bansal, Ankush 493, 519
513, 530, 538, 540, 546–547, 552–554, Barab, Sasha 145, 399, 517
557–558, 563, 567, 569, 590, 597–598, Barcellini, Flore 301, 308
609–610, 623, 625, 636, 641, 654, 656 Bardovi-Harlig, Kathleen 6653–666, 686
Angeli, Charoula 498, 517 Bargh, John A. 577, 598, 603,
Angouri, Jo 22, 24, 135, 142, 388, 396–397, Bargiela-Chiappini, Francesca 410, 423
413–414, 423, 470, 486, 608, 610–613, Bar-Ilan, Judit 156, 171
620, 622, 625–626 Barker, Chris 548, 569

Authenticated
700 Name index
Barlow, Julie 417, 426, 678, 680, 683, Bernstein, Basil 216, 217,
686 Bernstein, Michael S. 496, 514
Barnes, Susan B. 10, 24, 89, 112 Berry, Sarah 548, 570
Baron, Naomi 65, 77, 94, 112, 206, 217, Bevan, Jennifer L. 601
247–249, 252, 267, 285, 308 Bhanot, Ruchi 631
Baronchelli, Andrea 220 Biber, Douglas 97, 113, 127, 141, 143, 169,
Barron, Anne 44, 264, 494, 516 320, 339, 437, 452–453
Barron, Greg 669, 684 Biesenbach-Lucas, Sigrun 661, 663–666,
Barsalou, Lawrence W. 351, 375 668–669, 684
Bartlett, Tom 43, 54 Bieswanger, Markus 94, 113
Barton, David 27, 100, 113, 118, 148, 212, Bijker, Wiebe E. 46, 54
214, 217, 221, 573, 622, 626 Bimber, Bruce 47, 54
Barton, Kristin M. 42, 54 Binning, Kevin R. 599
Bateman, Patrick 84, 86, 90, 92, 113 Binns, Amy 494, 514
Bauckhage, Christian 596, 598 Bins, Elmar 133, 143
Baughman, Linda 545, 548, 561, 566–567, Bizzell, Patricia 216–217
572 Black, Emily 264, 268
Baumeister, Roy F. 575–576, 598, 602 Black, Lisa 505, 514
Baym, Nancy 83, 86, 113, 133, 139, 143, Black, Pamela J. 637, 656
205, 217, 227, 239, 284–285, 308, 345, Black, Steven D. 285, 300, 308
366, 375, 413, 418, 423, 464, 486, 494, Blake, Megan 156, 168
514, 566, 568–569, 590, 593, 598, 607, Blasiola, Stacy 33, 39, 58
609, 611–614, 618, 626 Bless, Herbert 350, 375
Beals, Shawn R. 580, 598 Bloch, Joel 663, 665, 684
Beard, Fred K. 250–252, 267 Blommaert, Jan 538, 540
Beavin, Janet H. 407, 434 Blood, Rebecca 151, 153, 168
Bechtel, William 354, 375 Bloomfield, Leonard 247, 268
Becker, Mark W. 346, 375, 541 Blum-Kulka, Shoshana 636, 641, 644, 653,
Bedijs, Kristina 414, 423–424 656, 663, 684
Bednarek, Monika 17, 19–20, 24, 437, 452, Boardman, Mark 324, 339
455, 545, 547, 551, 559, 562, 564–570 Bocij, Paul 494, 503, 514
Beer, Jennifer S. 575, 581, 587, 605 Bolander, Brook 17, 19, 21–22, 37–38,
Beißwenger, Michael 132, 143, 250, 268, 54, 133–134, 137, 147, 158, 164, 168,
284–285, 308 204, 217, 236–237, 239, 389, 398, 407,
Beker, Guy 577, 601 413–414, 419–420, 424, 430, 444, 461,
Belair-Gagnon, Valerie 383, 397 471, 487, 489, 551, 588, 591–592, 599,
Bell, Allan 9, 68–69, 75, 77, 107, 113 603, 607, 611, 613, 615, 618–619, 622,
Bell, Diana Calhoun 593–594, 601 626, 677
Bellmore, Amy 218 Bolinger, Dwight 448, 452
Belmore, Nancy 97, 114 Bollen, Johan 436, 452
Benevenuto, Fabricio 218 Bolter, Jay David 46, 54, 495, 514
Bennett, Lucy 203, 220 Bomberger, Ann M. 610–611, 626
Benson, David C. 389, 397 Bond, Bradley J. 579, 581, 599
Benson, Phil 189, 194 Bonebrake, Katie 583, 599
Benwell, Bethan 138, 143, 461, 486 Bongard, Josh 355, 378
Berger, Charles R. 87, 113 Bonk, Curtis J. 498, 517
Berglund, Therese Örnberg 252–253, 268, Bonus, Sabrina 169
302, 314 Börzsei, Linda K. 596, 599
Bergmann, Jörg R. 182, 194 Bös, Birte 9–10, 43, 61, 83, 93–94, 113, 117,
Berker, Thomas 383, 397 135, 146, 299–302, 308, 387–389, 398,
Berlyne, Daniel E. 592, 599 412, 414, 428, 463, 488

Authenticated
Name index 701
Bou-Franch, Patricia 65–66, 70, 74, 78, 81, 309, 317–318, 321, 330, 340, 346, 365,
189–190, 194, 198, 284, 289–290, 302, 367, 370, 375
306, 308, 327, 339, 412, 414, 424, 444, Bucholtz, Mary 409, 411, 424, 426, 496,
452, 470, 489, 611, 617, 619, 626 514
Bourlai, Elli 597, 599 Buckels, Erin E. 393, 398, 494, 514
Bousfield, Derek 465, 487, 512, 514, 619 Buckley, Chloe 519
boyd, danah 69, 78, 82–86, 88, 91, 93, 111, Bühler, Karl 62, 78
113, 175, 188, 194, 198, 205, 211–212, Bull, Ray 512–513, 519, 522
217–218, 225–227, 234–235, 239, 241, Buni, Catherine 34, 55
318, 328, 339, 386, 394, 398, 401, 419, Burgess, Jean 69, 73, 78, 166, 168, 175–176,
424, 450, 456, 463, 487, 489, 533, 542, 178–179, 185, 195, 206, 208, 213, 215,
546, 548, 566, 568–569, 571, 576, 599, 218, 222–223, 457
623, 630 Burgoon, Judee K. 522
Boyd, Michael 63, 71, 78, 188–189, 194 Burke, Kenneth 408, 424
Boyd-Graber, Jordan 168 Burke, Moira 672, 684
Boyle, Ronald 636, 653, 656 Burkhalter, Byron 138, 144
Brabazon, Tara 494, 514 Burkhardt, Jean-Marie 308
Brake, David 69, 78, 581, 597 Burnett, Gary 448, 452
Brandtzaeg, Petter B. 597 Buschow, Christopher 549, 570
Brandel, Mary 505, 514 Butler, Brian S. 84, 86, 90, 92, 113, 684
Brennan, Susan 76, 78, 361, 365, 375–376, Button, Graham 279, 281, 295, 309
450, 453 Buxmann, Peter 602
Bretherton, Di 496, 515
Breustedt, Barb 658 C
Brinker, Klaus 277–278, 308 Caffi, Claudia 32, 55
Brock, Alexander 141, 143 Calero-Vaquera, María Luisa 597, 599
Brody, Nicholas 420, 432 Calixte, Shana 156–157, 165, 169
Brommer, Sarah 284–286, 310 Calvert, Sandra 165, 169, 617, 628
Brône, Geert 356, 379–380 Calvin, Angela 213, 218
Bronstein, Jenny 576–577, 579, 581–583, Calzo, Jerel P. 583, 605, 676, 688
588, 599 Cameron, Lynne 390, 398
Brown, Alexander 393, 398 Candlin, Christopher 670, 685
Brown, B. Bradford 575–576, 579, 585–586, Çapa-Aydin, Yesim 494, 521
606 Capps, Elinor 19, 26, 141, 148, 525, 527,
Brown, Barry 249–250, 272 532, 537, 543
Brown, Gillian 275, 278, 309, 317, 339 Carah, Nicholas 32, 55
Brown, Penelope 21, 126, 133–135, 138, 144, Cardoso, Gustavo 174, 195
410, 412, 424, 437, 444, 452, 464, 466, Carlson, Thomas 62, 68, 75–76, 79
475, 487, 511, 514, 587, 590, 592, 634, Carnevale, Peter 626
656, 663–664, 677, 684 Carter, Simon 213, 218
Brummernhenrich, Benjamin 414, 429 Casey, Neil 279, 281, 295, 309
Bruner, Jerome 524, 540 Čech, Claude G. 285, 309
Bruns, Axel 2, 24, 42, 54, 152, 168, 176, 195, Cermele, Jill 611, 617–618, 625
203, 205, 209, 212–213, 215, 218–219, Ceron, Andrea 436, 452
222–223, 389, 398, 436, 452, 457, 536, Chafe, Wallace L. 43, 55, 96, 113
540 Chaffee, Steven H. 40–41, 55
Brunst, Philip W. 389, 398 Chalfonte, Barbara 518
Bubel, Claudia 76, 78 Chamberlin, Perrin E. 601
Bublitz, Wolfram 42, 46, 55, 70–71, 78, Chandler, Daniel 323, 340, 495, 514
94, 101, 113, 235, 239, 248, 256, 268, Chandrasegaran, Antonia 138, 144
278–280, 282–283, 286, 291, 298–299, Chang, Wei-Lin Melody 72, 80, 138, 145,

Authenticated
702 Name index
290, 292, 297, 311, 413–414, 418, 427, Cölfen, Hermann 277–278, 308
611, 620–621, 628 Collins, Mauri P. 494, 515, 627
Chang, Yu-Ying 663, 666, 685 Collins, Nancy L. 582, 600
Chapman, Rebecca 669, 685 Collister, Lauren 259, 268
Chejnová, Pavla 470, 487, 667, 685 Collot, Milena 97, 114
Chemaly, Soraya 34, 55 Comarela, Giovanni 218
Chen, Chi-Fen Emily 661, 663–665, Conboy, Martin 44, 55
667–668, 683, 685 Condon, Sherri L. 285, 309
Chen, Gina Masullo 415, 424 Conrad, Susan 127, 141, 143, 339, 437, 453
Chen, Hsinchun 498, 513 Cooke, Maeve 87, 114
Chen, Rong 633, 657 Cooley, Skye C. 581, 600, 605
Chen, Yuan-shan 665, 669, 685 Cooper, Randolph 501–502, 518
Cheng, Justin 494, 515 Costello, Victor 545, 547, 570
Cherny, Lynn 284, 289–290, 294–295, 303, Cothrel, Joseph P. 499, 515
309 Cotten, Sheila R. 377
Cherry, Colin 202, 222 Couldry, Nick 174, 195,
Chester, Andrea 496–497, 500, 512, 515 Coulthard, Malcolm 189, 199
Chi, Ed H. 672, 687 Counts, Scott 453
Chiluwa, Innocent 159, 168, 438, 452, 592, Coupland, Nikolas 348, 377, 547, 570
600 Cowan, John 620, 626
Chin, Wynne 501–502, 518 Cox, Ana Marie 505, 515
Chiou, Wen-Bin 494, 515 Cox, Andrew 156, 168
Chiu, Ming Ming 611, 620, 629 Coyne, Richard 388, 402
Chiu, YiChing Jean 620, 626 Cozby, Paul C. 600
Chou, Hui-Tzu Grace 581, 600 Crair, Ben 250, 268
Chovanec, Jan 9, 24, 61–62, 66, 70, 72–73, Crawford, Jate 222, 457
75, 78, 184, 189, 195, 287, 309, 332, 340, Creeber, Glen 547, 570
618, 626–627 Crespo-Fernández, Eliecer 388, 398
Christensen, Christian 218, 452 Croft, William 353, 376
Christie, Bruce 379 Crow, Bryan K. 280, 310
Christofides, Emily 579, 581, 600 Cruse, David A. 353, 376
Chui, Hin Leung 418, 420, 430, 471, 489 Cruz-Moya, Olga 597, 604
Chun, Elaine 187–188, 195 Crystal, David 86, 93, 95, 99, 114, 125, 127,
Chung, Deborah S. 38, 40, 55 131–132, 144, 159, 168, 247–248, 268,
Cienki, Alan 353, 375 286, 294, 310, 318, 320, 324, 332–333,
Cimino, Richard 381, 398 335, 340, 383, 398, 449, 453
Cirillo, Valeria 637–638, 640, 642–643, 646, Culnan, Mary J. 370, 376
648–650, 657 Culpeper, Jonathan 465–466, 487, 512, 515
Cislaru, Georgeta 445, 452 Cunha, Evandro 214, 218
Clark, Andy 351, 355–356, 366, 375 Cunningham, Stuart 548, 571
Clark, Herbert 9, 15, 24, 62, 68, 75–76, Curini, Luigi 452
78–79, 82, 345, 347, 353, 356–358, Cutts, Katie 519
360–365, 368, 376, 450, 453
Claro, Susana 631 D
Cléirigh, Chris 220 Dąbrowska, Marta 328, 340
Coates, Jennifer 288, 309 Dahlberg, Lincoln 505–506, 511, 515
Cohen, Jonathan 547, 570 Dahlgren, Peter 596, 600
Cohen, Leor 538, 540 Dale, Rick 367, 378
Cole, Debbie 547, 570 Danescu-Niculescu-Mizil, Cristian 494, 515
Cole, Michael 355, 379 Danet, Brenda 93, 95–96, 114, 538, 540, 620,
Cole-Lewis, Heather 203, 218 627

Authenticated
Name index 703
Darics, Erika 249, 252, 268, 414, 425 Draucker, Fawn 332, 340
Das, Anupam 637, 645, 655, 657 Dresner, Eli 644, 650, 657
Dascal, Marcelo 281, 310 Dressler, Wolfgang Ulrich 317, 319–320, 340
Davey, Christopher G. 575, 600 Drew, Paul 250, 268, 279, 310, 312, 365,
Davidov, Dmitry 445, 453 367, 376
Davies, Bethan L. 489, 687 Du Bois, John 177, 195, 354, 356, 376
Davies, Bronwyn 236, 239, 409, 425, 526, Ducol, Benjamin 389, 398
540 Duffy, Margaret 581, 605
Davis, Wayne 511, 515 Duggan, Maeve 670, 688
Dawson, Clare 375 Duh, Kevin 203, 221
Day, Liz 519 Dunbar, Robin 230, 239
Dayter, Daria 165, 168, 203–204, 218, 414, Duranti, Alessandro 66, 79, 183, 196
418, 420–421, 425, 611, 627 Durndell, Alan 611, 617, 627
de Beaugrande, Robert-Alain 317, 319–320, Dürscheid, Christa 10, 25, 35, 45, 55, 57,
340 93, 95–97, 99, 107, 114, 116, 173, 197,
de Choudhury, Munmun 436, 453 229, 239, 241, 247, 250, 256, 269, 271,
de Fina, Anna 411, 425, 538, 540 283–286, 310, 312, 419, 428, 437
de Groot, Daphne 463, 489 Dutta, Soumitra 231, 240,
de Nooy, Juliana 137, 145 DuVal Smith, Anna 418, 425, 619, 627
de Oliveira, Sandi Michele 73, 79 Dwyer, Dominic B. 600
de Saussure, Ferdinand 62, 79, 247, 272 Dwyer, Paul 220
Debatin, Bernhard 89–91, 114 Dynel, Marta 8–9, 24–25, 32, 61–63, 67–69,
December, John 495, 516 71–76, 78–80, 179, 188–189, 195, 287,
Demiralp, Emre 602 309, 332, 340, 414, 425, 444, 453, 470,
Dennis, Jessica 603 487, 610, 618, 626–627
Deppermann, Arnulf 409, 425, 526, 540 Dziurzynski, Lukasz A. 453
DeSanctis, Gerardine 128, 139, 144
Desmarais, Serge 579, 581, 588, 600 E
Détienne, Françoise 308 Eagleton, Terry 382, 398
Deumert, Ana 536, 540 Ebersbach, Anja 226, 240
Deuze, Mark 61, 79, 183, 195 Eckert, Penelope 88, 114, 409, 425, 428, 617,
Develotte, Christine 173, 195 627
Dholakia, Nikhilesh 64, 82 Economidou-Kogetsidis, Maria 661, 664,
Dholakia, Ruby Roy 64, 82 666, 685
Dienlin, Tobias 90, 114 Edelsky, Carole 288–289, 310
Dill, Jody C. 494, 516 Edge, Nicholas 581, 600
Dill, Karen E. 494, 516 Edwards, Derek 354, 376
Dillon, Leevia 401 Efron, Miles 672, 685
Dindia, Kathryn 579, 602 Efthimiadis, Efthimis 212, 219
Dlala, Imen Ouled 494, 516 Eggins, Suzanne 449, 453
Doğan, Gürkan 635, 659 Eglin, Peter 409, 427
Dolan, Bill 202, 222 Ehrhardt, Claus 414, 425
Donath, Judith S. 139, 144, 295, 310, 494, Eichstaedt, Johannes 436, 453
496, 505–506, 511, 516 Eisenberg, Eric M. 669, 688
Donnerhacke, Lutz 133, 144 Eisenchlas, Susana A.139, 144, 414, 416, 425
Dontcheva-Navratilova, Olga 318, 340 Eisenlauer, Volker 9, 12, 25, 42, 49, 55–56,
Doosje, Bertjan 575, 600 65, 70, 72, 80, 94, 98–101, 110, 115, 165–
Dorval, Bruce 277, 281, 315 166, 168, 173, 179, 195, 225, 229, 234,
Douglas, Karen M. 497, 516 237–238, 240, 256, 321, 328–330, 340,
Dovey, Jon 81 356, 370, 376, 384, 414, 419, 422, 425,
Drasovean, Anda 438, 453 471, 525, 530, 541, 588, 610, 671, 692

Authenticated
704 Name index
Eisenstein, Jacob 203, 217 Florio, Susan 288–289, 315

Eklund, Lina 577, 600 Flowe, Heather D. 519
Ekman, Paul 512, 516, 589, 600 Flowerdew, John 318, 340
Ellemers, Naomi 575, 600 Ford, Sam 548–549, 561, 571
Ellison, Nicole 90–91, 93, 113, 115, 117, Ford, Shawn 663, 669, 685
175, 194, 226, 230, 239–240, 318, 328, Ford, Tamsin 297, 315
339, 419, 424, 576, 579, 584, 600–601, Forsyth, Carol 377
605, 670–671, 685–686 Fraser, Matthew 231, 240
Elmer-Dewitt, Philip 498, 516 Frehner, Carmen 318, 340
Emmens, Tobit 315 Frick, Karin 99, 114, 250, 256, 269
Enama, Patrick Rodrigue Belibi 620, 627 Friedel, Curtis 632
Enfield, Nick J. 356, 372, 376 Friedman, May 156–157, 165, 169
Enli, Gunn 39, 47, 56, 218, 452 Friginal, Eric 169
Eno, Cassie 154, 169 Fritz, Elisabeth 14, 73, 127, 253, 275, 385
Ensslin, Astrid 472, 487 Fritz, Gerd 278, 310, 318, 321, 338, 340
Epley, Nicholas 518 Frobenius, Maximiliane 13–14, 25, 66, 76,
Erdur-Baker, Özgür 494, 521 80, 156, 158, 163, 169, 182–183, 195, 245,
Erickson, Frederick 288–289, 315 256, 265–266, 269, 527, 530, 534, 543
Ermert, Karl 12, 25, 228, 240 Fröhlich, Uta 136, 144, 414, 425
Eslami, Zohreh R. 638, 646, 648–651, 657 Frosch, Caren A. 519
Estrella, Gucci 631 Fung, P. 519
Eysenbach, Gunther 111, 115
G
F Gaber, Mohamed 239
Fabrot, Bernard 133, 144 Gal, Noam 387, 399
Fairclough, Norman 44–45, 56, 382, 399, Gal, Susan 409, 426
670, 685 Galegher, Jolene 518
Fang, Yuguang 121 Gamberini, Luciano 263, 273
Farrell Whitwort, Kathleen 137, 144 Gamon, Michael 453
Fauconnier, Gilles 353, 376 Garcés-Conejos Blitvich, Pilar 65–66, 70,
Fayard, Anne-Laure 128, 139, 144 74, 78, 80–81, 189–190, 194, 198, 284,
Feldman, Valerie 242 289–290, 302, 306, 308, 327, 339, 412,
Ferris, Pixy 495, 516 414, 424, 444, 452, 465, 470, 487, 489,
Ferrucci, Patrick 581, 605 617, 619, 626
Fest, Jennifer 318, 342 Garcia, Angela C. 13, 25, 245, 249, 251,
Fetzer, Anita 116, 183, 195 258–259, 269–270, 279–281, 316
Feyaerts, Kurt 377, 379 García-Gómez, Antonio 22, 26, 418, 420,
Fichman, Pnina 494, 516 430, 576, 583, 588, 600, 637, 640–642,
Fiedler, Klaus 350, 375 658
Fillmore, Charles J. 353, 376 Gardner, Rod 635, 657
Finegan, Edward 339, 437, 452 Gardner, Roderick 277, 310
Finin, Tim 220 Garfinkel, Harold 13, 25, 245, 250, 269
Fiore, Andrew T. 685 Gawne, Lauren 498, 522
Firth, Alan 173, 197 Gee, Emily 632
Firth, John Rupert 216, 219, 438, 453 Gee, James Paul 174, 196
Fischer, Kerstin 356, 376 Gee, Matt 212, 214, 220
Fish, Robert 518 Georgakopoulou, Alexandra 158, 169, 414,
Fisher, Danyel 521 418, 426, 526, 530, 532, 540–541, 560,
Fitzgerald, Richard 390, 399, 409, 425 570, 600
Fitzsimons, Grainne M. 577, 598 Gere, Charlie 174, 196
Flanagin, Andrew J. 494, 519 Gerhardt, Cornelia 13–14, 245, 547, 569

Authenticated
Name index 705
Gerodimos, Roman 49, 56 Greenfield, Patricia 13, 25, 251, 256, 261,
Giardina, Natasha 93, 118 269, 603
Giaxoglou, Korina 539, 541 Greenhalgh, Greg 219
Gibbs, Jennifer 576, 579, 584, 600–601 Greenspan, Brian 537, 541
Gibbs, Raymond W. 511, 516 Greenwell, Christopher 219
Gibson, James 179, 196, 350, 376, 528, 541 Gregoriou, Christiana 20, 25, 552, 567, 570
Gibson, Paige L. 89, 91, 93, 118 Greitemeyer, Tobias 581, 604
Gibson, Will 248, 269 Grice, H. Paul 134, 138, 145
Giddings, Seth 81 Grieve, Jack 157, 169
Giles, David 204, 219, 547, 570 Griffith, Maggie 182, 196
Gilks, Kate 587, 601 Gronning, Anette 418, 433
Gill, Phillipa 219 Gross, Ralph 91–92, 115, 231, 240
Gillmor, Dan 156, 169 Grosseck, Gabriela 202, 219
Gilly, Mary C. 576, 604 Grozeva, Maria 137, 145
Giltrow, Janet. 173–174, 176, 181–182, 196 Gruber, Helmut 93–94, 108, 115, 228, 240,
Glaser, Markus 226, 240 286, 293, 295, 297–298, 310, 311, 438,
Goddard, Cliff 635, 657 453
Goffman, Erving 9, 21, 61–62, 64, 66–69, Grusin, Richard 46, 54, 495, 514
71–75, 80, 126, 130, 133, 135, 137–138, Gruzd, Anatoliy 449, 453–454
145, 184, 187–188, 196, 229, 240, 260, Guadagno, Rosanna 154, 169
264, 269, 289, 310, 409–410, 412, 426, Guiller, Jane 611, 617, 627
577, 587, 601, 610, 622, 627, 663, 677, Gumperz, John 182–183, 196, 590, 601
686 Gutwinski, Waldemar 317, 341
Gohl, Fabian 235, 238, 240 Gutzeit, Andreas 687
Golato, Andrea 258, 272, 635, 657, 672, 686
Golder, Scott 211–212, 218 H
Goldsmith, Daena J. 676, 686 Haddington, Pentti 183, 196
Goleman, Daniel 575, 601 Haenlein, Michael 106–107, 117, 174, 197
Gomes, Danielle 401 Hafner, Christoph 226, 241
Gonçalves, Marcos André 218 Hagel, John 448, 454
Gonzales, Amy L. 581, 601 Hagelmoser, Rebecca 170
Goodman, Simon 392, 399 Hagemann, Jörg 277, 308
Goodwin, Charles 63, 80, 183, 196, 365, 376 Häkkinen, Ari 187–188, 198
Gordijn, Ernestine H. 497, 520 Hall, Kira 409, 411, 424, 426, 496, 514
Graesser, Arthur C. 354, 377 Halliday, Michael A.K. 14, 25, 317–320, 325,
Graham, George 354, 375 327, 339, 341, 348, 377, 382, 399, 436,
Graham, Michael B. 603 439, 446, 454
Graham, Sage Lambert 17–19, 134, 162, Hambrick, Marion 203, 219
410–414, 426, 430, 437, 444, 455, 459, Hambright, Brittany L. 601
463, 465, 467–470, 482, 485, 487–488, Hampel, Elisabeth 416, 426
494, 512–513, 516, 609–611, 627, 629, Han, Rongbin 39, 56
634, 664, 677 Hancock, Jeffrey 206, 221, 420, 431, 576,
Grant, Iain 81 579–581, 591–592, 601, 603, 605, 656
Grasmuck, Sherri 232, 242, 419, 434, Hanna, Barbara 137, 145
576–579, 583, 586, 591–592, 606 Hara, Kazuya 606
Gray, Rebecca 115, 672, 685–686 Hara, Noriko 494, 498, 517, 520
Green, Joshua 69, 73, 78, 175–176, 178–179, Hardaker, Claire 18–19, 132, 139, 203,
185, 195, 548–549, 561, 571 219, 391–393, 399, 413–414, 426, 444,
Green, Mitchell 511, 516 468–470, 480, 488, 493–494, 506, 517,
Greenall, Annjo 161, 169 610, 628
Greene, Richard 393, 399 Hards, Rachael 611, 631

Authenticated
706 Name index
Hargittai, Eszter 85, 101, 110, 115, 450, 455 196–197, 202, 210, 219, 247–252,
Harnish, Robert M. 636, 656 2 56–257, 267, 270, 272, 283–286,
Harper, F. Maxwell 671, 686 289–290, 293–295, 299–300, 303–306,
Harper, Richard 256, 269, 527, 530, 534, 543 311–312, 318–319, 321, 324, 341, 389,
Harré, Rom 236, 239, 346, 355, 377, 409, 392–393, 399, 412, 414–415, 419, 427,
425, 526, 540 431, 438, 448–449, 454, 463, 488,
Harrington, Stephen 203, 219, 546, 548, 570 494–495, 497–498, 500, 505, 512, 517,
Harris, Kathleen 383, 399 523, 528, 538, 540–541, 597, 599, 607,
Harrison, Angela 221 611–613, 616, 620, 624, 627–628, 644,
Harrison, Ben J. 600 650, 657
Harrison, Sandra 259, 417, 426, 678, 680, Herrmann, Andrew F. 139, 145
683, 686 Herstand, Ari 635, 657
Harry, Drew 514 Hert, Philippe 611, 628
Hartford, Beverly S. 665–666, 686 Hester, Stephen 409, 427
Hartmann, Maren 397 Heuvelman, Ard 70, 82, 394, 401, 493, 519
Harvey, Kevin 139, 145, 417, 426, 673, 684 Hewis, Elaine 315
Hasan, Ruqaiya 14, 25, 317–320, 325, 327, Heyd, Theresa 10–11, 66, 151, 156, 162, 169,
339, 341 173, 184, 197, 335, 389, 471, 538, 541,
Hassall, Tim 665, 686 583, 610
Hatcher, Ruth M. 519 Hickman, Jonathan 558, 561, 569, 571
Haugh, Michael 72, 80, 138, 145, 290, 292, Highfield, Tim 203, 213, 218–219
297, 311, 413–414, 418, 426–427, 437, Hiippala, Tuomo 228, 240
454, 465, 467, 470, 488–489, 494, 511, Hilderman, Robert 445, 457
517, 611, 620–621, 628, 687 Hills, Matt 546, 549, 566, 570
Have, Paul ten 248, 259, 262, 269–270 Hiltunen, Turo 383, 399
Haverkate, Henk 636, 657 Hiltz, Starr Roxanne 512, 518
Hayashi, Reiko 288–289, 311 Himelboim, Itai 435, 454, 456
Haythornthwaite, Caroline 449, 454 Hine, Christine 652, 657
Heath, Christian 265, 270 Hinkel, Eli 678, 683, 686
Heatherly, Raymond 242 Ho, Victor 470, 488, 670, 686
Heekeren, Hauke R. 345, 378 Hobbs, Jerry R. 277–279, 281, 293, 303–304,
Heffernan, Virginia 505, 517 312
Heger, Dominik K. 242 Hockett, Charles 247, 270
Heidemann, Julia 277, 231, 240 Hodkinson, Paul 576, 601
Heigl, Richard 226, 240 Hodsdon-Champeon, Connie 620, 628
Heino, Rebecca 576, 579, 584, 600–601 Hoffmann, Christian R. 1, 9, 25, 31, 37, 42,
Helasvuo, Marja-Liisa 174, 196 49, 55–56, 63, 71–72, 76, 81, 83, 86, 94,
Held, Gudrun 412, 414, 423–424, 427 98–101, 110, 115–116, 151, 158, 165–166,
Helfrich, Uta 413, 427 168–169, 235, 239, 256, 268, 286, 293,
Hendriks, Berna 666, 686 296, 298–299, 309, 318, 324, 335, 341,
Henri, France 498, 517 367–377, 389, 399, 525, 541, 607, 637–
Hepp, Andreas 174, 195 640, 644, 653, 655, 658
Herbert, Robert K. 634, 646, 648, 650, 653, Hoffmann, Ludger 132, 143, 293, 296, 312
657 Hoffmann, Sebastian 417, 430, 677–678,
Heritage, John 676, 686 685
Hermerén, Lars 44, 56 Hofmann, Stefan G. 603
Hernandez, Ivan 435, 457 Hogan, Bernie 208, 219
Herring, Susan 3, 25, 46, 56, 63, 65, 69, 74, Höhn, Nicole 133, 147, 414, 430, 461, 489,
80–81, 88, 92–95, 97, 99–103, 106, 110, 588, 603
115–116, 132, 135, 139, 145, 151–152, Holcomb, Christopher 593–594, 601
155–156, 158, 169, 173–174, 181–182, Holly, Werner 228–230, 241

Authenticated
Name index 707
Holmer, Torsten 290, 295, 303, 312 J

Holmes, Janet 22, 25, 609, 634–635, 639, Jabbari, Nasser 638, 646, 648, 650–651, 657
642–643, 646–648, 650, 653, 658 Jablokow, Kathryn 632
Holotescu, Carmen 202, 219 Jacka, Elizabeth 548, 571
Holt, Elizabeth 279, 310, 312 Jackson, Don D. 407, 434
Holton, Avery E. 456 Jackson, Michele 63, 82
Honeycutt, Courtenay 202, 210, 219, 256, Jacobs, Jennifer B. 13, 25, 245, 249, 251,
270 258–259, 269–279
Hong, Lichan 672, 687 Jakobson, Roman 62, 81, 588, 601
Honkanen, Mirka 162, 169 Jakobsson, Ármann 504, 518
Hood, Susan 446, 454 Jang, Chyng-Yang 164, 171
Hoon, Tan Bee 660 Janney, Richard W. 32, 55
Hopkinson, Christopher 494, 517 Janoschka, Anja 41, 44, 56
Hopwood, Christopher J. 346, 375 Jaroenkitboworn, Kandaporn 472, 490
Horstmann, Marcus 687 Jarvis, Jeff 83, 116
Horton, John 375 Jasper, Dirk 126, 133, 145
Horvitz, Eric 453 Jaszczolt, Katarzyna M. 511, 518
Hössjer, Amelie 414, 428 Java, Ashkay 220
Houghton, David J. 116 Jaworski, Adam 348, 377, 634, 643, 653, 658
House, Juliane 636, 641, 644, 653, 656, 663, Jefferson, Gail 75, 82, 249–250, 258, 272,
684 279, 281, 312, 449, 457
Housley, William 390, 399 Jenkins, Henry 4, 26, 42, 56, 174, 178, 197,
Hsieh, Yin Feng 239 536, 542, 545–546, 548–549, 552, 561,
Hsu, Yi-Ping 663, 666, 685 566, 568–571
Huang, Chih-Mao 581, 601 Jenks, Christopher 173, 197
Huang, Jeff 212, 219 Jewitt, Carey 174, 197
Huang, Yan 6, 25 Jha, Sneha 453
Huo, Yuen J. 601 Job-Sluder, Kirk 145, 399, 517
Hübler, Mike T. 593–594, 601 Johansson, Marjut 11, 61, 65, 81, 85, 89, 93,
Huettner, Alison 498, 521 110, 116, 174, 179, 182–183, 186, 189,
Huffaker, David 165, 169, 617, 628 192, 196–197, 199, 388, 400, 444, 587
Hughes, Amanda 222 Johansson, Stig 324, 339
Hum, Noelle J. 579–581, 601 Johnson, Bobbie 83, 116,
Humphreys, Lee 205, 219 Johnson, Norman 501–502, 518
Hunston, Susan 17, 26, 437, 454 Johnston, Angus 88, 116,
Hunt, Daniel 417, 419, 428, 536, 542 Johnstone, Barbara 34, 57
Hurst, Matthew 498, 519 Joinson, Adam N. 90–92, 116
Hutchby, Ian 74, 81, 179, 197, 248, 255, 270 Jones, Ceri 519
Hutcheon, Linda 529, 542 Jones, Dave 221, 226
Hutcheon, Michael 529, 542 Jones, Quentin 448, 455
Hutchins, Edwin 355, 377 Jones, Ray 315
Hymes, Dell 488, 524, 528, 542, 634, 658 Jones, Rod 288, 312
Jones, Rodney 241, 249–250, 253, 270–271,
I 527, 537, 542, 621–622, 628
Iacus, Stefano M. 452 Jones, Steven 392, 400
Iedema, Rick 184, 197 Jonides, John 602
Ihlebæk, Karoline Andrea 388, 400 Joseph, John Earl 411, 428
Irvine, Judith T. 409, 428 Jourard, Sidney M. 602
Ito, Mizuko 263, 270, 546, 548, 566, Jousmäki, Henna 198
568–569, 571 Joyce, Elisabeth 684
Ivcevic, Zorana 582, 587, 606 Jucker, Andreas H. 10, 14, 22, 26, 35, 57,

Authenticated
708 Name index
61, 70, 81, 93–97, 99, 102, 107, 116–117, Khadilkar, Vaibhav 242
173, 197, 229, 241, 247, 271, 283–286, Khan, Latifur 242
312–313, 321–322, 341, 388, 400, 409, Khan, Shariq M. 203, 217
419–420, 428, 430, 437, 493–494, 500, Khiabany, Gholam 39, 57
518, 635–636, 653, 658 Khorasani, Manouchehr Moshtagh 134,
Jucks, Regina 414, 429 146
Juffermans, Kasper 538, 540 Kiefer Lee, Kate 635, 658
Junco, Reynol 346, 377 Kieran, Kelly 81
Justinussen, Jákup 49, 56 Kiesler, Sara 370, 379, 385, 400, 512, 518,
521, 607, 609, 628, 631
K Kiesling, Scott 411, 428
Kádár, Dániel Z. 138, 145, 413–414, 427, Kilian, Jörg 286, 312
437, 454, 465, 467, 488, 611, 620–621, Kilic, Aysenur 588, 602
628 Kim, Jinsuk 579, 602
Kakavá, Christina 610, 628 Kim, Tackjin 684
Kalviknes Bore, Inger-Lise 558, 561, 569, King, Joseph 168
571 Kiritchenko, Svetlana 445, 456
Kampf, Zohar 387, 399 Kirkham, Natasha Z. 367, 378
Kankaanranta, Anne 418, 433 Kirkup, Gill 165, 170
Kantarcioglu, Murat 242 Kirner-Ludwig, Monika 10, 24
Kanzashi, Deepak 493, 519 Kirschner, Paul A. 346, 377
Kaplan, Andreas 106–107, 117, 174, 197 Kirsh, Elana 495, 518
Kappas, Arvid 49, 57–58 Klein, Yvonne 318, 341
Karahalios, Karrie 295, 310 Kleinberger Günther, Ulla 131, 148
Karakayali, Nedim 588, 602 Kleinke, Sonja 9–10, 37, 43, 57, 61, 85,
Karhukorpi, Johanna 27, 260, 273, 611, 93–94, 113, 117, 135–137, 146, 299–302,
631 308, 387–389, 398, 412, 414, 428, 463,
Karpinski, Aryn C. 346, 377 488, 620, 629
Kasper, Gabriele 540, 636, 641, 644, 656, Klemm, Michael 553, 571
663, 684 Klier, Mathias 240
Katriel, Tamar 281, 310 Kligler-Vilenchik, Neta 632
Kavanagh, Barry 162, 170 Kluge, Bettina 413, 428
Kaveri, Subrahmanyam 25, 269, 603 Knieps, Melanie 519
Kayany, Joseph M. 494, 500, 518 Knight, Naomi 17, 26, 216, 220, 443–444,
Kaye, Barbara K. 47, 57 451, 455
Kaye, T. 512, 518 Knobel, Michele 595, 602
Kee, Kerk F. 575, 606 Knoke, Felix 226, 241
Keegan, Brian 220 Knox, John 210, 220
Kehoe, Andrew 212, 214, 220 Knupsky, Aimee C. 415, 428
Keller, Eric 276, 312 Koch, Michael 227, 241
Kellogg, Guy 137, 148 Koch, Peter 10, 26, 43, 57, 95–97, 117, 283,
Kelsey, Darren 203, 220 286, 312
Kelsey, Sigrid 158, 171 Köhnken, Günter 513
Kerbrat-Orecchioni, Catherine 66, 81, Kollock, Peter 130, 146
287–289, 312, 634, 653 Kong, KahMun Clara 138, 145
Kern, Margaret L. 453 Konstan, Joseph A. 686
Kern, Richard 173, 195 Kopytowska, Monika 389, 400
Kerr, Elaine B. 512, 518 Korenman, Joan 400
Kettrey, Heather Hensman 392, 400 Korolija, Natascha 279–280, 288, 313
Keyes, Corey Lee M. 602 Koteyko, Nelya 139, 145, 417, 419, 426, 428,
Keysar, Boaz 511, 518 536, 542

Authenticated
Name index 709
Kouper, Inna 116, 170, 399, 417, 428, 661, Langlotz, Andreas 15, 346–347, 350,
676, 678, 687 352–356, 358, 362, 369, 373, 377–378,
Kouzina, Joanne 239 413–414, 418, 429, 470, 488, 611,
Kowalski, Robin M. 576, 602 623–624, 629
Kramarae, Cheris 434, 444, 458, 611, 620, Lankshear, Colin 595–596, 602
632 Lapadat, Judith C. 611, 629
Krasnova, Hanna 581, 602 Largier, Céline 283, 290, 313
Kraut, Robert 494–495, 497, 518, 673, 684, Larsson, Anders Olof 47, 57, 218
688 Laster, Whiteney Nicole 392, 400
Kreiss, Daniel 47, 57 Latour, Bruno 229, 234, 241
Kreß, Beatrix 414, 428 Launspach, Sonja 136, 146
Kress, Gunther 136, 146, 174, 197–198, 323, Laursen, Ditte 254–255, 263, 271
328, 341, 351, 377, 529, 544 Lavé, Jean 128, 146, 466, 489
Krishnamurthy, Balachander 219 Law, Nancy WaiYing 611, 620, 629
Krishnamurthy, Sandeep 170 Lazaraton, Anne 259, 271
Kristeva, Julia 209, 220 Lazer, David 220
Krogh Hansen, Per 529, 542 Lea, Martin 371, 378, 463, 489, 494, 497,
Kroskrity, Paul V. 382, 402 500, 512, 519
Kross, Ethan 581, 602 Leary, Mark R. 575–576, 579, 598, 602
Kruger, Justin 498, 518 Leaver, Tama 213, 219
Krumsvik, Arne H. 388, 400 LeBlanc, Tracy 135, 147
Kumar, Aman 378 Lee, Benny P.H. 279, 313
Kumar, Krishan 84, 87, 103, 120 Lee, Carmen 100, 108–109, 113, 117, 220,
Kumar, Ravi 152, 170 328, 331, 341, 418–420, 429, 622, 626
Kunkel, Melanie 413–414, 429 Lee, Cynthia 414, 433, 611, 620, 631
Kunneman, Florian 445, 455 Lee, David Seungjae 602
Kuo, Li-Jen 638, 646, 650–651, 657 Lee, Hangwoo 494, 519
Kurtz, Andrew J. 293, 303, 311 Lee, Lillian 436, 456
Kytölä, Samu 109, 117, 198, 203, 220 Leech, Geoffrey 134, 147, 313, 339, 455,
464, 467, 489, 558, 602, 658
L Lemke, Jay 548, 571
Labarthe, Darwin R. 453 Lengel, Laura 132, 149, 284, 315
Labov, William 19, 26, 141, 146, 204, 220, Lenk, Uta 276, 313, 340
408, 429, 448, 455, 524–525, 542 Léon, Jacqueline 409, 429
Laclau, Ernesto 382, 400 Leppänen, Sirpa 57, 94, 110, 117, 174, 184,
Lakoff, Robin T. 134, 146, 464, 488, 587, 187–188, 198
592, 602 Leskovec, Jure 494, 515
Lambiase, Jacqueline 286, 292–295, Lester, Jessica 219
305–306, 313 Levin, James A. 308
Lampe, Cliff 90–91, 115, 117, 230, 240, Levine, John M. 673, 688
685–686 Levine, R. Mark 497, 520
Lampley, Jonathan Malcolm 42, 54 Levinson, Stephen 62, 67, 81, 126, 133,
Lamy, Marie–Noëlle 173, 195 138, 144, 182–183, 198, 356, 371, 378,
Landert, Daniela 5, 8, 10, 26, 41–45, 57, 61, 410, 424, 452, 464, 466, 475, 487, 511,
85, 93, 96–99, 102, 106, 117, 283, 313, 514, 587, 590, 592, 599, 634, 656, 677,
388, 400, 420, 428–429 684
Landherr, Andrea 240 Lewis, David 76, 81
Landone, Elena 137, 146, 611, 620, 629 Lewis, Diana M. 611, 619, 629
Langacker, Ronald W. 353, 356, 377 Leyda, Julia 563, 571
Lange, Patricia 88, 94, 117, 178, 198, 400, Li, Xigen 388, 400
414, 429, 444, 455 Licoppe, Christian 13, 26, 264, 271

Authenticated
710 Name index
Liebrecht, Christine 455 MacCulloch, John Arnott 504, 519

Lievrouw, Leah A. 174, 198 Macdonald, Clare 299, 315
Light, Ben 91–92, 117 MacQueen, Kenneth 660
Limberg, Holger 662, 687 Madianou, Mirca 99, 118, 186, 192, 198
Lin, Natalie 602 Magno, Gagno 218
Lin, Shu-Fang 64, 69, 82 Maher, Brenden 472–473, 489
Lin, Yu-Ru 215, 220 Mahlberg, Michaela 318, 340
Lincoln, Sian 576, 601 Mahrt, Merja 223
Linell, Per 182–183, 198, 247, 271, 279–280, Maíz-Arévalo, Carmen 418, 420, 430, 471,
288, 291–292, 313 489, 580, 585, 588–589, 591, 593–594,
Ling, Kimberly S. 684 603, 611, 620, 629, 633, 636–638,
Linnemann, Gesa 414, 429 640–642, 646, 648–651, 654, 658
Lippka, Yannick 673, 687 Mak, Bernie Chun Nam 418, 420, 430, 471,
Lister,Martin 74, 81 489
Litosseliti, Lia 410, 433 Maley, Yon 670, 685
Litt, Eden 69, 81, 450, 455 Malinowski, Bronislaw 644, 659
Livingstone, Sonia 69, 81, 174, 198, 576, Mallan, Kerry 93, 118
602 Maltby, John 494, 519
Locher, Miriam A. 329, 331, 342, 346, 352, Manago, Adriana M. 4–5, 26, 577, 603
378, 409–414, 416–420, 424, 429–430, Mandiberg, Michael 4–5, 26
434, 437, 444, 455, 459, 461, 465–466, Manes, Joan 22, 26, 634, 639–640, 642–645,
469, 471, 487, 489, 494, 512, 519, 653, 659–660
536, 542, 551, 587–588, 591–592, 599, Mao, Huina 436, 452
602–603, 607–609, 611, 622–626, 629, Marcoccia, Michel 36, 58, 66, 72–73, 82,
635, 658, 661, 674–681, 686–687 127, 147, 173, 189, 198, 252, 271, 284,
Lomborg, Stine 228, 241 287, 289, 295, 298, 302, 313, 630
Longley, Paul 203, 217 Marder, Ben L. 116
Loosen, Wiebke 89, 118 Margolin, Drew 220
Lorenzo-Dus,Nuria 62, 65–66, 70, 74, 78, Maricic, Ibolya 134, 147, 630
189, 194, 198, 284, 290, 302, 306, 308, Markman, Kris 247, 249, 254, 262, 271–272,
327, 339, 470, 489 292, 294–296, 303, 313, 318, 342, 383,
Losh, Elizabeth 230, 238, 241 400
Lotan, Gilad 211–212, 218 Marko, Georg 138, 147
Louw, Eric 32, 55 Markus, M. Lynne 370, 376
Love, Brad 456 Marshall, Catherine 76, 79
Lower, Amanda 585, 588, 633, 635, 637–644, Martell, Craig 168
646–654, 659 Martin, Arnaud 516
Lu, Jingyan 611, 620, 629 Martin, James 17, 26, 216, 220, 228, 232,
Luckmann, Thomas 182, 194 241, 342, 435–441, 449, 451, 455–456,
Lüders, Marika 41, 57, 599 525, 542
Luff, Paul 265, 270 Martin, Jason 242, 416, 434, 576–579, 583,
Luhmann, Niklas 40, 58 586, 591–592, 606
Lundby, Knut 546, 571 Martinson, Anna M. 286, 292–293, 303–305,
Lundmark, Vicki 518 315
Luzón, María José 162, 170, 414, 430, 611, Marwick, Alice 69, 82, 86, 88, 111, 113, 167,
619, 629 170, 185–186, 188, 198, 235, 241, 384,
Lyons, John 247, 271 394, 401, 450, 456, 463, 489, 533, 542,
576, 584, 603, 623, 630
M Maslen, Rob 390, 398
Maagaard, Cindie 536, 542 Mately, David 456
Maaß, Christiane 412, 414, 423–424, 430 Matthews, Heather 138, 147

Authenticated
Name index 711
Matthiessen, Christian M.I.M. 318–320, 341, Mohammadi, Masoomeh 665, 687

348, 377, 439, 446, 454 Molema, Seija 557, 583, 604
Maynard, Douglas W. 275, 279, 281, 313 Møller, Janus Spindler 117
McCabe, Deborah Brown 576, 584, 606 Mondada, Lorenza 361, 378
McCloud, Scott 529, 542 Monroy–Hernández, Andrés 514
McConnell-Ginet, Sally 617, 627 Monteserin, Ariel 611, 620, 630
McCosker, Anthony 494, 519 Moor, Peter 70, 82, 394, 401, 493, 519
McCourt, Claire 414, 434 Moore, Barbara 545, 547, 570
McCreery, Stephen 435, 454 Morel, Julien 13, 26, 264, 271
McDonald, Daniel 36, 58 Morris, Meredith Ringel 671, 687–688
McEnery, Anthony 203, 217 Morris, Merrill 64, 69–70, 82
McGarty, Craig 497, 516 Morrow, Phillip R. 23, 139–140, 147, 417,
McGlashan, Mark 203, 219, 391, 393, 399, 431, 583, 661, 676, 678–680, 683, 687
493–494, 517 Mouffe, Chantal 382, 400
McGrath, Kathy 91, 117 Moulitsas, Markos 505, 519
McGuire, Timothy W. 385, 400, 512, 518, Mroczek, Kristine 97, 99, 120, 174, 199, 419,
521, 609, 628 429, 433
McKenna, Katelyn Y.A. 577, 598, 603 Mühleisen, Susanne 165, 168
McLaughlin, Margaret L. 132, 147 Muise, Amy 579, 581, 588, 600
McNeill, Laurie 151, 170 Mukopadhaya, Tridas 518
McQuail, Denis 40, 58 Mulholland, Joan 27, 249, 253, 261–262,
Medina, Richard 227, 242 272
Mehan, Hugh 308 Mulkay, Michael 630
Meheula, Lyn 378 Mullany, Louise 410, 431, 686
Mehus, Siri 355, 379 Muntigl, Peter 630
Mei, Qiaozhu 223 Müller, Marion G. 49, 57–58
Meibauer, Jörg 235, 241 Münz, Stefan 133, 147
Memon, Amina 512, 519 Murgia, Madhumita 47, 58
Mendelson, Andrew I. 86, 110, 118 Murthy, Dhiraj 201, 204, 221
Mendoza–Denton, Norma 409, 411, 431 Musolff, Andreas 16, 388–389, 401
Merchant, Raina M. 453 Myers, Greg 153, 157, 160, 162–163, 170,
Meredith, Joanne 249–250, 258–259, 272 173, 181, 199, 336, 342, 471, 490, 538,
Merrison, Andrew John 470, 489, 663, 667, 542
687 Myrick, Jessica Gall 436, 456
Merseburger, Axel S. 687
Mesch, Gustavo S. 579, 603 N
Meshi, Dar 345, 378 Nadkarni, Ashwini 603
Meyer, Jonas Ivo 170 Nagarajan, Meenakshi 211, 221
Meyerhoff, Miriam 408, 431, 448, 456 Nagourney, Adam 47, 58
Michikyan, Minas 603 Nagy-Bell, Natalie M. 415, 428
Millard, William B. 494, 500, 519 Nakrasova, Tanya 169
Miller, Carolyn R. 181, 199, 389, 401, Naper, AnjaAaheim 47, 56
Miller, Daniel 99, 118, 186, 192, 198 Naper, Ida 486, 490
Miller, Hugh 617, 626 Naraine, Ryan 505, 519
Miller, Lynn C. 582, 600 Nass, Clifford 378
Miller, Vincent 221, 588, 603 Nass, Matthew 378
Mills, Sara 437, 454 Nastri, Jacqueline 206, 221, 420, 431,
Milner, Ryan M. 596, 603 591–592, 603
Misoch, Sabina 133, 147 Neo, Loo Seng 393, 401
Moe, Hallvard 209, 218 Neubig, Graham 203, 221
Mohammad, Saif 445, 456 Neumann, Stella 318, 342

Authenticated
712 Name index
Neurauter-Kessels, Manuela 37–38, 58, 388, P

401, 412–414, 431 Page, Ruth 1, 19, 27, 43, 58, 94–95, 97,
Newbury, Elizabeth 219 99–101, 103, 106, 108, 118, 133, 141, 148,
Ng, Zhi-Wen 518 156, 170, 173, 177, 188, 199, 203–204,
Nguyen, Hanh Thi 137, 148 211, 213–215, 221, 416, 419–420,
Nguyen, Thin 449, 456 431–432, 438, 471, 490, 525–528, 530,
Nielsen, Rasmus Kleis 47, 58, 154, 170 533–537, 539, 543, 569, 571
Niemi, Kreeta 622, 630 Painter, Clare 438, 456
Nigam, Kamal 498, 519 Palen, Leysia 222
Nilsson, Bo 165, 170 Palmer, Emma J. 519
Nishimura, Yukiko 414, 431, 470, 490 Palmer, Jerry 593, 604
Nitin, Chanderwal 493, 519 Pang, Bo 436, 456
Nix, Carole G. 286, 293, 295, 303–304, 312 Panovich, Katrina 514, 671, 687–688
Noë, Alva 355, 378 Panteli, Niki 498, 520
Norman, Donald 234, 241 Panyametheekul, Siriporn 251–252, 257, 267,
Nørreby, Thomas Rørbeck 117 272, 415, 431
Norrick, Neal R. 592–594, 604 Paolillo, John 158, 169, 399, 415, 427
Nosko, Amanda W. 579, 583, 604 Papacharissi, Zizi 9, 27, 33, 39, 42, 58,
Nöth, Winfried 323, 342 84–89, 91, 93, 108, 110, 118, 182, 196,
Novak, Jasmine 170 435, 456, 576–577, 590, 604
Nunamaker, Jay F. 522 Pappert, Steffen 277–278, 308
Nünning, Ansgar 165, 170 Park, Denise 575, 581, 601
Nussbaum, Miguel 631 Park, Gregory 453
Nystrand, Martin 448, 456 Park, Jiyoung 602
Park, Namsu 606
O Parker, Jason 518
O’Driscoll, Jim 470, 490 Parker, Richard 287, 314
O’Halloran, Kay 174, 199 Patschan, Oliver 687
O’Hara, Agi 497, 512, 515 Patterson, Michael 518
O’Keeffe, Anne 62, 74, 82 Paul, Sharoda A. 672, 687
O’Reilly, Michelle 519 Paulhus, Delroy L. 393, 398, 494, 514
O’Reilly, Tim 535, 543 Paulus, Trena 219
O’Shea, T. 519 Pavlidou, Theodossia-Soula 288, 314
O’Sullivan, Patrick B. 494, 519 Pea, Roy 346, 378
Oard, Douglas W. 168 Peirce, Charles S. 360, 378
Ochs Keenan, Elinor 278, 280, 293–294, 303, Pellicer, Régine 547, 570
313, 409, 431, 437, 456 Peña, Jorge 206, 221, 420, 431–432,
Ochs, Lisa 19, 26, 141, 148, 525, 527, 532, 591–592, 603
537, 543 Pennacchiotti, Marco 203, 221
Oesterreicher, Wulf 10, 26, 43, 57, 95–96, Penuel, Bill 631
117, 283, 286, 312 Perelmutter, Renee 413–415, 432, 470, 490,
Ogan, Christine 64, 69–70, 82 611, 620, 630
Ogden, Richard 250, 268 Pérez-Sabater, Carmen 94, 118
Okabe, Daisuke 263, 270 Perlroth, Nicole 233, 242
Okdie, Bradley 154, 169 Perrin, Andrew 85, 118
Ong, Keng Wee 611, 630 Peter, Jochen 583, 606
Orlikowski, Wanda J. 498, 522 Peters, Brad 611, 630
Osborne, Kerry K. 132, 147 Petersen, Eric 159–160, 164, 170
Östman, Sari 186, 199 Petersen, Line Nybro 552–553, 567,
Ovadia, Steven 212, 221 571
Owens, Christabel 315 Peterson, Leighton C. 498, 522

Authenticated
Name index 713
Petronio, Sandra 86, 91, 119 158–159, 161, 163, 167, 170–171, 212,
Peuronen, Saija 198 222–223, 335–336, 342, 444, 456
Pfeifer, Rolf 355, 378
Phillips, David J. 497, 512, 520 Q
Phillips, Steven R. 669, 688 Qian, Hua 576–577, 604
Phillips, Whitney 494, 520 Quan-Haase, Anabel 27, 107, 110, 121, 208,
Phung, Dinh 456 219
Pike, Jacqueline C. 84, 86, 90, 92, 113 Quinn, Clark N. 308
Pino, Cecilia 655, 660 Quiring, Oliver 110, 121
Pittam, Jeffrey 13, 27, 249, 253, 260–262,
272 R
Piwinger, Boris-A. 133, 143 Raban, Daphne 686
Placencia, María Elena 22, 27, 414, 417, 432, Rabinowitz, Peter 529, 543
585, 588, 633, 635, 637–655, 659, 681, Raclaw, Joshua 262–263, 272
688 Rafaeli, Sheizaf 36, 59, 686
Placks, Simon James 494, 520 Rafanan, Ken 631
Planalp, Sally 281, 314 Raghavan, Prabhakar 170
Planchenault, Gaëlle 413–414, 416, 432, 470, Rainie, Lee 670, 688
490 Rampton, Ben 538, 540
Plato 497, 520 Rance, Marcus 378
Plum, Gunther 525, 542 Ranta, Michael 529, 543
Poell, Thomas 174, 195 Rappoport, Ari 445–446, 453, 457
Poff, Michele 415, 434 Rauch, Shannon M. 392, 402
Pojanapunya, Punjaporn 472, 490 Recanati, Francois 511, 513, 520
Pomerantz, Anita 608, 610, 630, 634, 659 Reddy, Michael J. 356, 378
Pons Bordería, Salvador 281, 314 Reed, Adam 576, 604
Poole,Marshall 63, 82 Reed, Darren 219
Popescu, Ana-Maria 203, 221 Reeves, Stuart 249–250, 272
Porro, Giuseppe Porro 452 Reich, Zvi 388, 402
Portwood, Anne C. 601 Reicher, Steve 520
Posada, Susana 218 Reichman, Rachel 277, 280, 314
Posner, Roland 228, 241 Reid, Elizabeth 413, 418, 432, 609, 619 630
Postmes, Tom 370, 378 Reid, Fraser J.M. 611, 631
Potts, Amanda 211, 221 Reinecke, Leonard 89–90, 93, 108, 120
Potts, Liza 211, 221 Rentel, Nadine 413, 420, 432
Povolná, Renata 318, 340 Resnik, Philip 168
Powell, Hebe 638, 646–651, 653, 659 Rettberg, Jill Walker 157, 165–167, 171
Poynton, Cate 436–437, 456 Reynolds, Edward 611, 619, 631
Presley, Ifukor 452 Rheingold, Howard 4, 27, 88, 119, 241, 448,
Preston, Jesse Lee 435, 457 457, 576–577, 604
Prior, Paul 216, 221 Rhodes, Catherine C. 531, 544
Probst, Florian 240 Riboni, Giorgia 47, 59
Probst, Tahira 626 Richardson, Daniel C. 367, 378
Proferes, Nicholas 91, 121 Richardson, Kay 549, 559, 562–563, 565,
Proudman, Charlotte 635, 659 567, 571
Pugatch, Jillian 218 Richet, Bertrand 470, 490
Punie, Yves 397 Richter, Alexander 227, 241
Punset, Eduardo 575, 604 Rintel, Sean E. 13, 27, 249, 253, 260–262,
Purohit, Hemant 211, 221 272, 596, 604
Puschmann, Cornelius 46, 58, 85–86, Ritter, Alan 202, 222
94, 103, 106, 108–109, 119, 154–156, Ritter, Ryan S. 435, 457

Authenticated
714 Name index
Robertson, Scott 227, 242 Savicki, Victor 617, 631

Rocamora, Agnès 156, 171 Sawyer, Sheryl 138, 148
Rodino, Michelle 617, 631 Scannell, Paddy 62, 74, 82
Rodríguez, Catalina Fuentes 641, 659 Schaefer, Edward 9, 24, 62, 68, 75–76, 79
Rodriguez, Henrik 295, 298–299, 315 Schandorf, Michael 247, 272
Rodzvilla, John 157, 171 Schank, Roger C. 277–278, 281, 314
Roloff, Michael E. 87, 113 Schanz, Kimberley 392, 402
Roman, James 547, 571 Schat, Amanda C. 601
Rosch, Eleonor 355, 379 Schau, Hope Jensen 576, 604
Roschelle, Jeremy 620, 631 Scheckler, Rebecca 145, 399, 517
Roscoe, Timothy 69, 82 Scheepers, Helana 545, 566, 572
Rosenberg, Åsa 111, 119 Schegloff, Emanuel 75, 82, 235, 242, 245,
Rosenmann, Amir 577, 583, 604 249–250, 258, 261, 272, 279, 281, 291,
Roskos-Ewoldsen, David R. 87, 113 294–295, 314, 365, 378, 449, 457, 633,
Rosson, Mary Beth 202, 223 659
Rowe, Lottie 399 Scheible, Jeff 214, 222,
Rubin, Victoria L. 512, 520 Scheidt, Lois Ann 156, 158, 165, 169, 171,
Ruble, Racheal A. 140, 148 399
Rüdiger, Sofia 414, 420, 425 Scheidt, Louise 116
Rudolf von Rohr, Marie–Thérèse 417, 432, Scherlis, William 518
434 Schiaffino, Silvia 611, 620, 630
Rue, Yong-Ju 664–665, 688 Schieffelin, Bambi B. 278, 280, 293–294,
Ruhi, Sükriye 635, 659 303, 313, 382, 402, 437, 456
Runkehl, Jens 132, 142, 148, 269 Schiffrin, Deborah 115, 342, 348, 377, 379,
Rupp, Jan 170 517, 532, 543, 590, 604, 614–615, 631
Rusaw, Erin 472, 490 Schilling, Caroline 235, 238, 240
Ruston, Scott 537, 543 Schlesselman-Tarango, Gina 213, 222
Rutledge, Pamela 46, 59 Schlobinski, Peter 132, 142, 148
Ryan, Marie–Laure 525, 529, 543 Schmid, Hans-Jörg 104, 120
Schmidt, Jan 155, 165, 171
S Schmidt, Jan-Hinrik 87, 89, 107, 110, 119,
Saarikoski, Petri 199 222
Sachs-Hombach, Klaus 136, 148 Schneider, Beate 549, 570
Sack, Warren 308 Schneider, Gerold 635, 658
Sacks, Harvey 13, 27, 75, 82, 245, 249–250, Schneider, Klaus P. 659
258, 261, 272, 279, 281, 291, 314, 409, Schneider, Stefan 609, 631
432, 449, 457 Schnoebelen, Tyler 203, 217
Safir, Marilyn P. 577, 583, 604 Schnurr, Stephanie 416, 430, 609
Sagioglou, Christina 581, 604 Schober, Michael 62, 68, 82
Şahin, Reyhan 239 Schönfeldt, Juliane 258, 272, 296, 314
Salimkhan, Goldie 603 Schrader-Kniffki, Martina 414, 432
Samu, Kytölä 117, 198, 220 Schubert, Christoph 14–16, 257, 275, 286,
Sánchez-Moya, Alfonso 597, 604 317–320, 325, 342
Sanders, Amy 218 Schultz, Tanjev 39, 59
Sanderson, David W. 498, 520 Schütte, Wilfried 96, 119, 127, 131, 135, 148,
Sandner, Philipp G. 457 290, 314
Sanfilippo, Madelyn Rose 494, 516 Schwämmlin, Eva 464, 490
Santamaría-García, Carmen 589, 604 Schwartz, Hansen Andrew 453
Sap, Maarten 453 Schwarz, Mary 218
Saul, Jennifer 511, 520 Schwentner, Christian 687
Sauter, Theresa 213, 222 Scott, Craig R. 576–577, 604

Authenticated
Name index 715
Scott, Derek 494, 520 Sinha, Chris 356, 380

Scott, Kate 213, 222, 457 Sirois, Patricia 277, 281, 315
Seargeant, Philip 1–2, 5, 27, 203, 222, 242, Siti Yuhaida 637–638, 642–646, 648–650,
392, 402, 551, 572 655, 660
Searle, John R. 511, 520, 636, 653, 660, 671, Skjetne, Jan Håvard 599
673, 688 Skogerbo, Eli 218
Seedhouse, Paul 246, 273 Skovholt, Karianne 254, 273, 418, 433
Sefi, Sue 676, 686 Slade, Diana 449, 453
Seidel, Beate 141, 143 Sloan, Luke 1, 27
Seitzinger, Joyce 221 Smith, Aaron 670, 688
Sellen, Abigail 265, 273 Smith, Christopher 381, 398
Selwyn, Neil 227, 238, 242 Smith, Lauren Reichart 581, 600, 605
Senft, Theresa M. 185, 199 Smith, Marc 130, 146, 435, 454, 516, 521
Sengupta, Somini 233, 242 Smith, Matthew J. 370–371, 380
Severinson Eklundh, Kerstin 295, 298–302, Smithson, Janet 297, 315
315 Song, Xiaodan 220
Shablack, Holly 602 Spagnolli, Anna 263, 273
Shachaf, Pnina 494, 520 Sparks, Colin 44, 59
Sharkey, Siobhan 315 Spears, Russell 371, 378, 463, 489, 497, 519,
Sharma, Sanjay 206, 222 575, 600
Shaw, Frances 222, 435–436, 457 Spencer-Oatey, Helen 410, 433, 511–512,
Shea, Virginia 468, 490 521, 609, 631
Shema, Hadas 156, 171 Sperber, Dan 353, 356, 372, 379
Shepherd, Dawn 389, 401 Spiegel, Carmen 131, 148
Sheth, Amit 211, 221 Spilioti, Tereza 158, 169, 263, 273, 414, 433
Shi, Priscilla 401 Sprenger, Timm O. 457
Shifman, Limor 184, 199, 387, 399, 595–596, Sproull, Lee 370, 379, 607, 631
605 Squires, Lauren 151, 171, 217, 546–547, 571
Shin, Jiwon 494, 520 Srivastava, Sanjay 575, 581, 587, 605
Short, John 370, 379 St.Amant, Kirk 158, 171
Shultz, Jeffrey J. 288–289, 315 Stæhr, Andreas 99, 117, 119
Shum, Winnie 414, 433, 611, 620, 631 Stahl, Frederic 227, 239
Shuman, Amy 532, 543–544 Starbird, Kate 202, 222
Siegel, Jane 385, 400, 497, 500, 512, 518, Stefanone, Michael 164, 171
521, 611, 628 Stein, Dieter 3, 25, 63, 81, 95, 97, 112,
Siever, Torsten 132, 142, 148, 269 115–116, 132, 145, 173–174, 176,
Sifianou, Maria 203, 222, 444, 457, 471, 490, 181–182, 196–197, 319, 341, 438, 454
576, 597, 605, 634, 636, 660 Stein, Laura 32, 34–35, 59,
Siles, Ignacio 165, 171 Steinfield, Charles 90–91, 115, 117, 230, 240
Sillence, Elizabeth 417, 433, 676–677, 688 Stephenson, Rachel J. 673, 684
Simeon, Credell 445, 457 Sternberg, Janet 496, 521
Simha, Aneesh 378 Stillerman, Benjamin 378
Simm, Will 221 Stivale, Charles J. 494, 521
Simmons, Jason 219 Stockdale, Rosemary 545, 566, 572
Simon, Bernd 606, 670, 688 Stöckl, Hartmut 136, 148
Simpson, James 288–290, 315 Stoerger, Sharon 414–415, 427, 463, 488
Sinclair, John 189, 199, 548, 571 Stokoe, Elizabeth 138, 143, 249–250,
Sindoni, Maria Grazia 13, 27, 86, 93–94, 258–259, 272, 461, 486
96–97, 99–101, 103, 107, 109, 119, 178, Stommel, Wyke 219
199, 265, 273, 324, 326, 342 Storrer, Angelika 132, 143, 318, 321–322,
Singer, Jane B. 389, 402 342

Authenticated
716 Name index
Strack, Fritz 350, 375 Teevan, Jaime 671, 687–688

Strangelove, Michael 178, 199 Tench, Sara 217
Strano, Michele M. 605 Tepper, Michele 494, 505, 521
Streeck, Jürgen 355, 379 Tereick, Jana 174–175, 179, 183–184, 194
Strom, Paris S. 494, 521 Terkourafi, Marina 465, 467, 490
Strom, Robert D. 494, 521 Thaler, Verena 414, 433
Stromer-Galley, Jennifer 286, 292–293, Thelwall, Mike 83, 90, 92, 119, 156, 171
303–305, 315 Theodoropoulou, Irene 238, 242, 420, 433,
Subasic, Pero 498, 521 471, 490, 588, 590, 605
Subrahmanyam, Kaveri 13, 25, 251, 256, Thompson, Clive 505, 521
261, 269, 603 Thompson, Evan 355, 379
Suchman, Lucy 335, 379 Thompson, Geoffrey 17, 26, 437, 454
Sudweeks, Fay 36, 59 Thompson, John B. 85, 120
Şükriye, Ruhi 659 Thompson, Marshall 217
Suler, John 131, 148, 388, 402, 534, 544, Thompson, Robert J. 511, 521
577, 583, 605 Thornborrow, Joanna 288, 312, 680, 688
Sun, Jinyuan 121 Thornton, Katherine 212, 219
Sun, Tao 223 Thorsen, Einar 39, 54
Sundén, Jenny 407, 433 Thorson, Kjerstin 632
Sunderland, Jane 410, 433 Threadgold, Terry 228, 242
Suominen, Jaakko 175, 199 Thuraisingham, Bhavani 228, 242
Surratt, Carla G. 576, 605 Thurlow, Crispin 33, 39, 46, 59, 85, 92, 97,
Suzuki, Lalita K. 583, 605, 676, 688 99, 120, 132, 149, 174, 199, 284, 315, 415,
Svennevig, Jan 254, 273 419, 433
Swales, John 448, 457 Thurnherr, Franziska 417, 434
Swann, Joan 410, 433 Tice, Dianne M. 598
Swanson, Diana 611, 630 Tiittula, Liisa 276, 315
Swigger, Nathaniel 393, 402 Till, James E. 111, 115
Todenhöfer, Tilmann 687
T Tolson, Andrew 74, 82
Taavitsainen, Irma 22, 26, 493–494, 500, Toma, Catalina L. 605
518, 636, 658 Tomasello, Michael 356, 379
Tagg, Caroline 1–2, 5, 27, 203, 208, 222, Tomic, Alice 132, 149, 284, 315
234, 242, 328, 335, 342–343, 392, 402, Tomkins, Andrew 170
438, 453, 551, 572 Topçu, Çigdem 494, 521
Taleghani–Nikazm, Carmen 672, 686 Tracy, Karen 281, 310, 314
Tamir, Diana I. 345, 378 Tran, Truyen 456
Tan, Bee Hoon 637–638, 642–646, 648–650, Trapani, Gina 250, 273
655, 660 Trapnell, Paul D. 393, 398, 494, 514
Tan, Jethro 401 Traverso, Véronique 288–291, 315
Tandoc, Edson C. 581, 605 Tredinnick, Luke 32, 59
Tanna, Vanita 255, 270 Trepte, Sabine 89–90, 93, 108, 114, 120
Tannen, Deborah 43, 59, 83, 115, 119, 318, Trester, Anna Marie 318, 331, 339, 343, 585,
342–343, 517, 608, 614, 631 588–590, 606
Tanskanen, Sanna-Kaisa 13, 27, 131, 142, Tsagkias, Manos 213, 218
149, 174, 196, 260, 273, 317, 338, 343, Tseliga, Theodora 22, 24, 135, 142, 388,
611, 631, 695 396–397, 413–414, 423, 470, 486, 610–
Tauste, A. Vigara 599 613, 620, 622, 626
Taylor, Charles 138, 149 Tseng, Belle 220
Taylor, Dalmas 579, 598 Tseng, Chiao-I. 318, 343
Tazzyman, Sarah 519 Tseng, Chia-Ti Heather 668, 688

Authenticated
Name index 717
Tsur, Oren 445–446, 453, 457 Veen, Wim 346, 379

Tudini, Vincenza 250, 273 Venkatesh, Alladi 64, 82
Tulloch, John 44, 59 Venkatesh, Svetha 456
Tumasjan, Andranik 436, 457 Vercellone-Smith, Pamela 632
Turkle, Sherry 495, 521, 532, 544, 576–577, Verduyn, Philippe 602
606 Verhagen, Arie 356, 379
Turnage, Anna 413, 434, 610, 631 Verleur, Ria 70, 82, 394, 401, 493, 519
Turnbull, William 630 Verschueren, Jef 62, 67, 82, 382–383, 396,
Turner, Graeme 177, 185, 199–200 402
Turner, Mark 353, 376 Viégas, Fernanda 295, 310
Turner, Tammara Combs 505, 521 Vieweg, Sarah 222
Turtiainen, Riikka 199 Vinagre, Margarita 463, 491, 496, 522
Twitchell, Doug 522 Virkar, Shefali 494, 522
Tynan, Dan 92, 120 Virtanen, Tuija 3, 10, 25, 27, 63, 95, 97, 116,
Tyworth, Michael 399 140, 145, 149, 173–174, 181, 197, 319,
341, 429, 438, 454
U Vitak, Jessica 115, 683–684
Ueberheide, Simon 549, 570 Vogelstein, Fred 384, 402
Unger, Johann W. 27, 118, 148, 221, 571 Vraga, Emily K. 611, 623, 632
Ungerer, Friedrich 104, 120 Vrakking, Ben 346, 379
Unsworth, Len 456 Vrij, Aldert 512–513, 519, 522
Upadhyay, Shiv 414, 434, 470, 491 Vygotsky, Lev 355, 379
Utz, Sonja 494, 505, 522
W
V Wagner, Earl 168
Vaccari, Cristian 47, 58 Waletzky, Joshua 19, 26, 141, 146, 204, 220
Vachek, Josef 247, 273 Walker, Katherine 577, 606
Valdés, Guadalupe 655, 660 Walker, Traci 250, 268, 662, 688
Valenzuela, Sebastián 575, 606 Wall, Melissa 156, 171
Valkenburg, Patti M. 583, 606 Walters, Keith 187–188, 195
van den Bosch, Antal 455 Walther, Joseph B. 110, 120, 250–252, 267,
van Dijck, José 174–177, 195, 536, 544, 372, 380, 606
576, 579, 584, 590, 606 Wang, Xiaoqing 673, 684
van Dijk, Teun A. 277–278, 280, 315, 353, Wang, Yi-Chia 688
356, 369, 379, 382–383, 402, 527, 544 Wang, Yingmin 401
van Leeuwen, Theo 15, 27, 136, 146, 174, Ward, Katie 397
198, 323, 325, 341, 343, 529, 544 Waskan, Jonathan A. 354, 380
van Mulken, Margot 455, 664–665, 688 Watchravesringkan, Kittichai 576, 584, 606
van Schie, Emil G.M. 494, 522 Watson, Rod 247, 273
Vander Wal, Thomas 212, 222 Watts, Richard J. 126, 135, 147, 410–411,
Vandergriff, Ilona 414, 434, 462, 491, 611, 430, 461, 465–467, 489, 491, 590, 606,
620, 624, 632 609, 629, 635, 658, 660
Vanderveken, Daniel 636, 660 Watzlawick, Paul 407, 434
Varela, Francisco J. 355, 379 Weber, H.L. 614–615, 632
Vargas, Greg 514 Webster, James G. 64, 69, 82
Varghese, Arun 218 Weckerle, Andrea 494, 522
Vasalou, Asimina 116 Weerkamp, Wouter 213, 218
Vásquez, Camilla 438, 457, 528, 536, 544 Weidenhöffer, Jessica 20, 24, 546, 552–554,
Vatrapu, Ravi 227, 242 557–558, 563, 567, 569
Vaughan, Jill 498, 522 Weintraub, Jeff 84, 87, 97, 103, 120
Veale, Tony 373, 379 Weir, Kirsten 575, 606

Authenticated
718 Name index
Weisbuch, Max 582, 587, 606 Wortham, Stanton 531, 544

Weller, Katrin 206, 223 Wright, Elija 116, 169, 399
Wellman, Barry 448, 457 Wyatt, Nancy 388, 400
Welpe, Isabell M. 457
Welsch, Peter 399 X
Welser, Howard T. 521 Xing, Jianyu 511, 521
Wenger, Etienne 128, 146, 216, 223, 357, Xu, Jun–Ming 218
380, 448, 458, 466, 489 Xu, Zhi–Wei, 227, 242
Wenninger, Helena 602
Werry, Christopher C. 251, 273, 303, 316 Y
Wesch, Michael 450, 458 Yaghlane, Boutheina Ben 516
West, Candace 279, 281, 316 Yang, Chia-chen 575–576, 579, 585–586, 606
West, Laura 223, 331, 343, 585, 588–590, Yang, Lei 215, 223
606 Yang, Steven 378
Westinen, Elina 109, 117, 198, 203, 220 Yao, Mike Z. 84, 91, 121
White, Peter 17, 26, 435–441, 448, 451, 455, Yates, JoAnne 498, 522
458 Yates, Simeon 65, 82
Whittle, Jon 221 Ybarra, Oscar 602
Whitty, Monica T. 494, 522 Yechiam, Eldad 670, 684
Wickham, Phil 547, 572 Young, Alyson L. 107, 110, 121
Widdicombe, Sue 409, 423, 527, 532, 540 Yu, Ning 399
Widjajas, Thomas 602 Yuan, Elain 392, 403
Wiegman, Oene 494, 522 Yuan, Yi 653, 660
Wierzbicka, Anna 657 Yücel, Murat 600
Wikström, Peter 213–214, 223, 445, 458 Yule, George 275, 278, 309, 317, 339
Willems, Herbert 409, 434 Yum, Young-ok 606
Williams, Ederyn 379 Yun, Christopher 218
Williams, Lawrence 414, 434 Yurchisin, Jennifer 576, 584, 606
Williams, Raimond 84, 120 Yus, Francisco 93, 99, 101, 108, 110, 121,
Wilson, Deirdre 353, 356, 372, 379 127, 149, 155–156, 160–161, 171, 173,
Wilson, Jack J. 489, 687 200, 226, 232, 234, 236, 242, 318, 328,
Wilson, Samuel M. 498, 522 332–333, 335, 343, 389, 403, 412, 414,
Winget, Megan 672, 685 434, 610, 632, 636, 641, 654, 660
Wiszniewski, Dorian 388, 402 Yusof, Anniqah Modh 660
Wittgenstein, Ludwig 139, 149
Wodak, Ruth 611, 626 Z
Wodzicki, Katrin 464, 490 Zappavigna, Michele 1, 11–12, 17, 19,
Wolfe, Alan 88, 121 27–28, 39, 47, 59, 65, 118, 148, 173,
Wolff, Carolin 242 201–203, 205, 209–210, 212–216,
Wolfson, Nessa 22, 26, 634–635, 637, 220–221, 223, 256, 318, 332, 343, 347,
639–640, 642–645, 653, 659–660 384, 420–422, 434, 436, 438, 441, 444,
Wollis, Melissa 656 450–451, 458, 471, 535, 544, 549, 562,
Wood, Andrew F. 370–371, 380 568, 571–572, 578, 610, 632
Wood, Eileen 577, 581, 604 Zarei, Gholam Reza 665, 687
Wood, Helen 553, 572 Zarsky, Tal. Z. 497, 512, 522
Wood, Megan M. 545, 548, 561, 566, 567, Zdenek, Sean 497–498, 522
572 Zeng, Xiaojun 436, 452
Woodward–Kron, Robyn 36, 58 Zerdick, Axel 230, 242
Woodworth, Michael 656 Zhang, Chi 89, 121
Woolard, Kathryn A. 382, 402 Zhang, Grace Qiao 664–665, 688
Wortham, Jenna 233, 242 Zhang, Ming 223

Authenticated
Name index 719
Zhang, Wei 416, 434, 444, 458, 611, 620, Zhu, Hongqiang 224
632 Zhu, Wuhan 663, 689
Zhang, Xing-Zhou 227, 242 Zhu, Xiaoyan 121, 213, 218
Zhang, Yi 536, 544 Ziegele, Marc 110, 121
Zhao, Dejin 202, 223 Ziem, Alexander 351, 380
Zhao, Shanyang 232, 242, 419, 434, Zima, Elisabeth 353, 356, 380
576–579, 583, 586, 591–592, 606 Zimmer, Michael 91, 121
Zhivov, Jacob 545, 566, 572 Zimmerman, Don H. 506, 532, 544
Zhou, Lina 494, 522 Zlatev, Jordan 356, 380
Zhou, Michael 378 Zywiczynski, Przemyslaw 135, 149

Authenticated
Authenticated
Subject index
Subject index
A audience 41, 94, 100, 159, 166, 205, 212,

accessibility 68, 84–85, 93–97, 106, 111, 328, 385, 497, 545, 558, 567
247 ~ design 107, 109, 155, 328, 411, Ch. 3
addressee 23, 31–32, 39, 41, 43–44, 66, 68, types of ~ 20, 31, 38, 40, 44, 46, 93, 107,
72–73, 96, 254, 257, 328, 362, 368, 450, 155, 175, 188, 386, 394, 464, 450, 533,
635 546, 565, 612, 623
administrator see moderator authenticity 93–94, 101, 109–111, 152,
advice (~ giving, ~ seeking) 36, 94, 133, 165–167, 185, 394, 396, 407, 411,
139–142, 162, 237–238, 297, 416–417, 416–417, 422, 532–534, 579, 586
512, 536, 583, Ch. 24 authority 139–140, 383, 388, 391, 678
affordance 106–108, 216, 234, 248, 251, 350,
367, 370 B
(socio)cognitive ~ 160, 365 blog 34, 37–38, 72–73, 86, 98, 108–110,
communicative ~ 212, 267 249, 251, 259, 335–339, 382, 389, 414,
contextual ~ 246 470–471, 525, 583, 613, 615, 618–619,
functional ~ 236 638–639, Ch. 6
interactional ~ 608 corporate ~ 46, 638–639, 644
media ~ 523, 527–528, 532, 534, 539, 619 personal ~ 321, 618, 638–639
(multi-)modal ~ 100, 265, 625 video ~ 66, 71, 99, 177, 182, 185, 190–
semiotic ~ 351, 373 191, 251, 265, 328, 346
technical/technological ~ 89, 91, 100, 104, breadth (of information) 493, 578
179, 182, 232, 255, 257, 259–260, 266,
347, 394, 419, 588, 590, 610, 620, 646, C
650, 654 code of conduct 135, 291, 475
agreement 320, 560, 634, 647–648, Ch. 22 coherence 129, 135, 251–252, 256–257,
see also dis~ 259–260, 275, 286, 299, 302, 317, 534
aggression 130, 385, 412–413, 415, 466, 473, sequential ~ 256, 293–294
502, 609–610, 615–617 topical ~ 127, 130, 275, 283, 293–294
sexual ~ 203, 391 see also incoherence
verbal ~ 617–618, 621 cohesion 275, 303, Ch. 12
aggressive see aggression common ground 76, 159, 279, 297, 300,
affiliation 408, 421, 438, 443, 448, 539, 568 357–364, 368, 371, 374, 386, 391, 450,
ambient ~ 39, 201, 216, 332, 436, 451 590–592, 634, 643
social ~ 210, 213, 435 communicative setting 49, 97, 141, 147, 181,
anonymity 31, 40, 44, 62, 64, 69, 71, 73, 76, 229, 235, 238, 276, 461
90, 98, 102, 108–109, 190, 226, 392–393, community
396, 407, 413, 463, 496–498, 510–511, ~ building 47, 225, 234, 266, 389, 407,
528, 532, 534, 577–578, 583–584, 411, 418, 422, 470, 660
586–587, 609, 619, 634, 656, 683 ~ of practice 43, 88, 128, 157, 357, 533,
anonymous see anonymity 615
appraisal/Appraisal Framework 436–440, offline ~ 186, 533
442–443, 447 online ~ 36, 133, 153, 162, 226, 232,
appropriation 547, 552, 562–564 418. 448, 468, 594
asynchronicity see synchronicity compliment 588, 615, Ch. 23

Authenticated
722 Subject index
compliment response see compliment F

conditional relevance 246, 253–254, 261, Facebook 34–35, 45–53, 153, 205, 208,
296 256–258, 302, 322–323, 328–332,
conflict 37, 126, 133–135, 157–158, 162, 338–339, 381, 384–386, 393–396, 411,
164, 189–190, 413, 418, 444, 463, 497, 419–421, 471, 527, 529–531, 533–535,
499, 505–506, 508, Ch. 22 538–53, 562, 566, 577–583, 585, 588–594,
see also disagreement 610–611, 623, 633, 636–640, 644,
contextualization cue 101, 317, 339, 418 649–652, 670–673, Ch. 9
coordination device 358–364, 368–369, 372, facework 133–136, 140, 161–162, 535,
374 587–589, 593, 667, Ch. 15
coupling (of ideation and experience) 421, see also relational work
436, 443–444, 451 fan/fandom 42, 69, 188, 533, Ch. 20
cyberbullying 494, 503–504 FAQ 131, 135, 466–468, 472
cyberstalking 494, 503–504 flaming 126, 12, 4–235, 190, 260, 394, 413,
468–470, 477, 610, 681, Ch. 18
D forum (discussion ~) 94, 108, 127–129,
deception 132–134, 141–142, 256, 284–286,
identity ~ 139, 512 296–297, 306, 385, 388–389, 414–415,
online ~ 494 417, 470, 499, 503, 506, 534–536, 538,
narrative ~ 512 577–578, 610–612, 622, 680
nefarious ~ 511 see also message board
deictical see deixis
deixis 106, 109, 157, 162–163, 209, 299, G
336, 538 genre 46, 65–66, 68, 83, 94, 96, 98,
depth (of information) 90, 109, 557, 582 100–102, 108, 141, 151–158, 165–167,
diary 37, 151, 155–156, 165, 182, 204–205, 174, 181–184, 191, 202, 227–228, 267,
335, 613, 615, 618 388–389, 449, 473, 524–525, 548, 559,
disagreement 37, 135, 164, 389, 565, 634, 613
681, Ch. 22 getting liked see like
see also agreement grounding 76, 361–363, 372
see also conflict
(self–)disclosure 90–93, 102, 104, 106–107, H
109–111, 164, 185, 232, 421, 496, 534, hashtag 35, 212–216, 322, 421, 444–448,
576, 579, 582–583, 585–587, 673 534–535, 539, 641
discussion forum see forum hashtagging see hashtag
distance see immediacy hypercohesion see cohesion
hyperlink 49, 265, 322, 335
E
email 44, 164, 249–250, 254, 260, 286, 418, I
470, 485, 502–503 identity 92, 98, 101, 133, 139, 164–165, 326,
emoji 462, 498, 529, 589, 635, 641, 644–645, 386, 496, 508, 532–533, 538, 561, 565,
650 577–578, 586, 593, 668, Ch. 15
emoticon 32, 76, 96, 162, 284, 330–331, 346, ~ construction/formation 109, 126,
356, 387, 418, 438, 462, 474, 557, 594, 137–138, 141–142, 165, 182, 186–188,
615, 636, 641, 644, 647–648, 650–651, 226, 231, 237, 266, 322, 373, 511, 531,
672 546, 568, 591
evaluation 141, 204, 297, 388, 494, 524, 530, immediacy 86, 93–97, 100, 112, 248,
534–535, 557–558, 622, 635–637, 648, 256–257, 283
651, 653, 665, Ch. 16 linguistic/language of ~ 43–45, 106
impoliteness 135–136, 390, 396, 412–413,
512, 591, 610, 621–622, 664–665, Ch. 17

Authenticated
Subject index 723
Instagram 213, 233, 471, 531, 536, 637–638, netiquette 131–133, 135, 383, 413, 468, 485,
641, 656 500
interactivity 42, 49, 61, 64, 70, 83, 99, 174, non-anonymous see nonymity
205, 411, 499, 525 nonymity 578, 582–584, 586–587
Internet forum see forum nonymous see nonymity
involvement 278, 547, 677–678, Ch. 2
O
L other-presentation see presentation
like 35, 40, 130, 234, 236–237, 331, 384, other-representation see presentation
461, 471, 545, 635, 640, 646–648,
650–651, 653, Ch. 21 P
linking see like participant 85–86, 93–94, 98, 100, 102, 106,
LinkedIn 225, 227, 231–233, 576, 580, 584, 109, 128, 158, 164, 179, 188–189, 264,
590 287, 289–290, 307, 365, 367, 369, 444,
484, 607, 610–611, 618–619, 625, Ch. 2,
M Ch. 3
medium/media multiparty ~ 251–252, 276–277, 285, 293,
~ factor 102, 528, 530–532, 607–608, 296, 619
611–613, 622, 624–625 participation framework see participant
mass ~ 31, 40–41, 43–44, 47, 63, 97, 230, participatory media see medium/media
450, 545–547, 559 politeness 133–139, 161–162, 410–414, 422,
meso ~ 230, 235 444, 511, 587–590, 592, 615–616, 634,
meme 70–71, 213, 215, 381, 387, 393, 645, 663–669, 677, 682, Ch. 17
420–421, 498, 552, 562, 587–588, polylogue 65–66, 277, 300, 320, 327, 415
595–597 polyloguicity see polylogue
message board 35–36, 152, 226, 552, Ch. 5 presentation
see also forum self- ~ 111, 130, 165, 181–182, 185, 226,
microblogging 152–154, 332, 338, 421, 444, 232, 234, 236–237, 339, 361, 372, 576–
611 577, 580–581, 586–587, 597, 679, 683
see also Twitter producer 62–65, 67, 69–76, 93, 97, 153–154,
moderator 37, 127, 134, 477–479, 481, 486 188, 228–229, 249, 349, 367–369, 372,
monetization 151–152, 156, 158, 167 374, 546, 557
monotopical 277, 290–292, 295, 305 production (role) see producer
monotopicality see monotopical pseudonymity 496, 505
multimodal 49, 74, 76, 86, 93–94, 98–104,
106, 108–110, 112, 132, 153, 156, 163, Q
166, 174–175, 178, 182, 190–191, 204, quote see quoting
229–230, 246–247, 256, 265–266, 321, quoting 130, 256, 298, 303, 386, 562–563,
323–325, 328, 331, 334, 336, 351, 618
356–357, 360, 387, 422, 459, 470, 472,
486, 498, 523, 527–531, 536, 549, 559, R
567–568, 576–577, 586, 595–597, 624, receiver 42, 44, 62–76, 188, 249–250,
634, 641 255, 276, 289, 295, 297, 301, 362–363,
multimodality see multimodal 368–370, 372, 374, 554–555, 557,
multi-party interaction 65–67 669–670, 679, 682
reception (role) see receiver
N recipient see receiver
narration 43, 94, 98, 101, 141, 156–158, relational work 133, 141, 162, 410–411,
165–166, 203–204, 297, 512, 547, 414, 465, 470–472, 485, 609, 620, 623,
557–561, 676, 678–680, 683, Ch. 19 677–678, 683
narrative see narration see also facework

Authenticated
724 Subject index
repair 246, 258–260, 333, 620, 672 troll 139, 189, 382, 393–394, 413, 468–469,
rudeness 135, 465–466, 616 480, Ch. 18
see also impoliteness trolling see troll
troller see troll
S Tumblr 156, 233, 249, 552–553, 565,
searchable talk 201, 212, 332, 549 567–568, 597
self–disclosure see disclosure turn-taking 63–65, 70, 246, 251, 257, 449
self–presentation see presentation Twitter 35, 94, 188, 233, 251, 256, 320–322,
sequential organization 246, 248, 251–252, 332–335, 338, 345, 347–348, 367–374,
254, 265, 280 382, 384, 391–393, 419–422, 444, 450,
Social Network(ing) Site (SNS) 83, 86, 471, 528, 530–531, 533, 536, 539, 549,
90–91, 153, 249, 256, 318, 328, 394, 411, 552, 558, 568, 577–578, 582, 584, 587,
419, 471, 497, 534, 588, 610, 623, 633, 590, 635, 644–645, Ch. 8
670–671, Ch. 9
social tagging 212, 436, 438, 444, 449 V
spam 189, 468–469, 477, 479–486, 503 video blog see blog
spammer see spam virtual reality see virtual world
spamming see spam virtual world 240, 472, 474, 637
synchronicity 64, 131, 248–249, 251, 277, vlog see video blog
283–285, 295, 366, 462–463, 471, 486,
619 W
Web 2.0 61, 91, 98–99, 102–103, 151–152,
T 306, 319, 384, 407, 535, 575, 588, 592,
template 130, 153, 155, 205, 234, 237, 286, 597, 607–610, 618, 621, 624, 633
298, 303, 339, 370, 525, 530, 562 weblog see blog
thread 36, 128, 130, 132, 134–136, 286, YouTube 35, 94, 188, 233, 351, 256,
289, 294–300, 302–303, 305–307, 332, 320–322, 332–335, 338, 345, 347–348,
338–339, 386, 413, 499, 535, 594, 681 367–374, 382, 384, 391–393, 419–422,
threading see thread 444, 450, 471, 528, 530–531, 533, 536,
topic 38, 43, 65, 73, 97–98, 102, 106–107, 539, 549, 552, 558, 568, 577–578, 582,
109–110, 127–128, 130, 132, 155–156, 584, 587, 619, Ch. 7
159, 214, 253, 332, 335–338, 445, 526,
535, 594, 613, 639, 642–643, 675, Ch. 11

Authenticated

Hoffmann, C. & Bublitz, W. (2017) - Pragmatics of Social Media

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hoffmann, C. & Bublitz, W. (2017) - Pragmatics of Social Media

Uploaded by

Copyright:

Available Formats

Pragmatics of Social Media

Library of Congress Cataloging-in-Publication Data

Bibliographic information published by the Deutsche Nationalbibliothek

© 2017 Walter de Gruyter GmbH, Berlin/Boston

Preface to the handbook series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

1. Log in: Introducing the pragmatics of social media

I. The nature of social media

2. Participation as user involvement

3. Participation as audience design

4. Publicness and privateness

II. Social media platforms

9. Social network sites/Facebook

III. Social media and discourse

10. Discourse and organization

11. Discourse and topic

12. Discourse and cohesion

13. Discourse and cognition

14. Discourse and ideology

IV. Social media and identity

15. Facework and identity

17. Politeness and impoliteness

18. Flaming and trolling

V. Social media and functions

21. Getting “liked”

22. Conflictual and consensual disagreement

23. Compliments and compliment responses

24. Requesting and advice-giving

About the authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691

The series Handbooks of Pragmatics, which comprises thirteen self-contained

Brought to you by | Cambridge University Library

(which we consider a decided asset), we suggest a definite structure, which gives

– it operates with a wide conception of pragmatics, dealing with approaches that

Brought to you by | Cambridge University Library

Brought to you by | Cambridge University Library

Augsburg and Berlin, May 2017

Brought to you by | Cambridge University Library

Abstract: This introduction to the handbook of the Pragmatics of Social Media

1. The aim of the handbook

This handbook provides a comprehensive overview of state-of-the-art pragmatic

may share a mutual focus on computer-mediated communication, they are differ-

Brought to you by | Cambridge University Library

celebrity Facebook profiles, academic tweets. However, (rather futile) attempts to

Brought to you by | Cambridge University Library

2. Defining social media

Brought to you by | Cambridge University Library

In contrast to ‘classic’ Internet websites then, social media encourage Internet

Brought to you by | Cambridge University Library

Brought to you by | Cambridge University Library

4. Overview of the handbook

Brought to you by | Cambridge University Library

speech act functions, e.g. disagreeing, complimenting, requesting. Each chapter

4.1. The nature of social media

Brought to you by | Cambridge University Library

Brought to you by | Cambridge University Library

4.2. Social media platforms

Brought to you by | Cambridge University Library

ical benchmarks of social media interaction. Describing the difficulties of defin-

Brought to you by | Cambridge University Library

straints and options, before turning to interdisciplinary and pragmatic research

Brought to you by | Cambridge University Library

4.3. Social media and discourse

Brought to you by | Cambridge University Library

Brought to you by | Cambridge University Library

Figure 1. Visualisation of involvement strategies in a three-dimensional model