Chovanec. Pragmatics of Tense and Time in News

Pragmatics of Tense and Time in News
Pragmatics & Beyond New Series (P&BNS)

Pragmatics & Beyond New Series is a continuation of Pragmatics & Beyond and
its Companion Series. The New Series offers a selection of high quality work
covering the full richness of Pragmatics as an interdisciplinary field, within
language sciences.
For an overview of all books published in this series, please see
http://benjamins.com/catalog/pbns
Editor Associate Editor

Anita Fetzer Andreas H. Jucker
University of Augsburg University of Zurich
Founding Editors
Jacob L. Mey Herman Parret Jef Verschueren
University of Southern Belgian National Science Belgian National Science
Denmark Foundation, Universities of Foundation,
Louvain and Antwerp University of Antwerp
Editorial Board
Robyn Carston Sachiko Ide Deborah Schiffrin
University College London Japan Women’s University Georgetown University
Thorstein Fretheim Kuniyoshi Kataoka Paul Osamu Takahara
University of Trondheim Aichi University Kobe City University of
Miriam A. Locher Foreign Studies
John C. Heritage
University of California at Los Universität Basel Sandra A. Thompson
Angeles Sophia S.A. Marmaridou University of California at
University of Athens Santa Barbara
Susan C. Herring
Indiana University Srikant Sarangi Teun A. van Dijk
Cardiff University Universitat Pompeu Fabra,
Masako K. Hiraga
Barcelona
St. Paul’s (Rikkyo) University Marina Sbisà
University of Trieste Yunxia Zhu
The University of Queensland
Volume 253
Pragmatics of Tense and Time in News
From canonical headlines to online news texts
by Jan Chovanec
Pragmatics of Tense and Time
in News
From canonical headlines to online news texts
Jan Chovanec
Masaryk University
John Benjamins Publishing Company

Amsterdam / Philadelphia
TM
The paper used in this publication meets the minimum requirements of
8
the American National Standard for Information Sciences – Permanence

of Paper for Printed Library Materials, ansi z39.48-1984.
Library of Congress Cataloging-in-Publication Data
Chovanec, Jan.
Pragmatics of tense and time in news : from canonical headlines to online news texts /
Jan Chovanec.
p. cm. (Pragmatics & Beyond New Series, issn 0922-842X ; v. 253)
Includes bibliographical references and index.
1. Pragmatics--Data processing. 2. Mass media and language. 3. Mass media--Social
aspects. 4. Speech acts (Linguistics) 5. Internet--Social aspects. I. Title.
P96.L34C58 2014
070.401’41--dc23 2014028182
isbn 978 90 272 5658 4 (Hb ; alk. paper)
isbn 978 90 272 6932 4 (Eb)
© 2014 – John Benjamins B.V.

No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any
other means, without written permission from the publisher.
John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands
John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa
Table of contents
Preface ix
Acknowledgements xi
List of abbreviations xiii
List of tables and figures xv
Chapter 1
Introduction1
1.1 Goals and objectives 2
1.2 Approach 6
1.3 Data 6
1.4 Overview of chapters 7
Part I. Temporal deixis in print and online news
Chapter 2
Theoretical foundations 15
2.1 Pragmatics 15
2.2 Functionalism and Halliday’s metafunctions 17
2.3 Heteroglossia 20
2.4 News discourse analysis 23
Chapter 3
Temporal deixis and news discourse 25
3.1 Deixis and interaction 26
3.2 Deictic centre 29
3.3 Deictic projection 34
3.4 Deictic projection in news texts 37
3.5 Time adverbials and shared temporal context 39
3.6 Deictic and non-deictic time expressions 41
3.7 Time expressions in news texts 43
3.8 Deictic centres in print newspapers 46
3.9 Pre-emptiveness of deictic time adverbials 48
3.10 Modelling deictic projection in news texts 51
3.11 Temporal deixis and tenses 56
vi News and Time
Chapter 4
Temporal deixis in online newspapers 59
4.1 Hypertextuality and the double textual level of online news 60
4.2 Temporal anchorage points in online newspapers 67
4.3 Temporal anchorage on the home page 71
4.4 Temporal anchorage in article previews 74
4.5 Temporal anchorage on article web pages 77
4.6 Hypertextuality and temporal mapping in online articles 79
4.7 Temporal deixis and internal hyperlinks 86
Part II. Textual rhetoric of headlines
Chapter 5
Temporal deixis in headlines 97
5.1 Material for analysis 97
5.2 Headlines and the expression of time 99
5.3 Adverbials of time in headlines 99
Absence of adverbials of time from headlines 100
Presence of adverbials of time in headlines 103
5.4 Expressing the setting and location of the story 106
5.5 Verbal tenses in headlines 109
5.6 Tense in headlines in the data 110
5.7 Headline conventions 117
Grammatical features 119
Lexical features 120
Non-linguistic features 120
Chapter 6
The present tense in headlines 123
6.1 The defaultness of the present tense in news headlines 123
6.2 Deictic and non-deictic tense 125
6.3 Present time reference of the simple present tense 127
State present 127
Habitual present 128
Instantaneous present 128
6.4 Past-time reference of the simple present tense 131
Semantics of the past-time reference of the present tense 132
Historic present 134
Tense as an evaluation device 135
Deictic centre projection 137
6.5 Future time reference of the simple present tense 140
Table of contents vii
6.6 Potential ambiguity of the simple present tense in headlines 141

Manipulation of temporal deixis 143
Chapter 7
Other tenses in headlines 145
7.1 Expressing futurity 146
To-future 147
Modal auxiliaries 149
Will-future 151
Lexically expressed future 152
7.2 The present perfect 154
Headlines marking trends and changes 155
Heteroglossic headlines 156
Dual headlines 157
7.3 The simple past tense: From heteroglossia to information
flow management 158
The past tense in the non-authorial accessed voice 159
The past tense in the paper’s authorial voice 160
The past tense as a marker of non-recency 161
The past tense, subordination and information flow 165
The past tense as a marker of accessed voice 171
Other uses of the simple past tense – the non-factive presupposition 173
Other uses of the simple past tense – satellite articles 174
Chapter 8
Auxiliaries in headlines: Ellipsis and (non)-finiteness 179
8.1 Ellipsis of auxiliaries 179
8.2 Potential ambiguities 182
8.3 Explicit use of auxiliaries 188
Accessed voice and reported speech 189
Semantic specification 191
Foregrounding of major news stories 195
8.4 Complex headlines with subordinate clauses 198
8.5 Concluding remarks on temporal deixis in headlines 201
Part III. Textual rhetoric of news texts
Chapter 9
The textuality of news texts 205
9.1 Textual segments: The headline and beyond 206
9.2 Cohesion analysis 207
viii News and Time
9.3 Information chaining 208

9.4 Process chains 210
Non-cohesiveness of the present perfect tense 213
9.5 Double tense shift pattern 215
9.6 Cohesion and the three metafunctions 218
Towards a functional model of temporality in process chains 219
Variations of the idealized pattern 220
9.7 Patterns of cohesion and co-referentiality in online news texts 225
Non-permanence of home page article previews 226
The structural template for online news 227
9.8 The double tense shift pattern and its variations 230
Example 1. The triple tense pattern 231
Example 2. Complex chain involving nominal transformation 234
Example 3. Variations on the triple tense pattern 239
Example 4. The triple tense pattern as a cohesive structure 241
9.9 Concluding remarks on tense and textuality 244
Chapter 10
Temporal structure of news reports 247
10.1 Non-chronology and the narrative structure of news stories 248
10.2 Temporal structure of news stories 254
10.3 Modelling the internal structure of news texts 260
Thematic structure of news texts 260
Conceptual structure of the news story & event frames 262
Orbital organization and the interpersonal dimension of news time 265
10.4 Final remarks 268
Chapter 11
Conclusion269
11.1 Tense shifts 269
11.2 Temporality and the textual rhetoric of headlines 270
11.3 Temporality and the textual rhetoric of news texts 273
11.4 Temporality in online news 274
References277
Index285
Preface
The increasing attention that has been paid over the past couple of decades to the
study of language in the media attests to linguists’ fascination with a field that
is immensely diverse as regards communicative situations, discourse phenomena
and the communicating participants. The globalized transfer of information and
the almost continuous exposure to diverse communication channels means that
the media have a more central role than ever in the lives of almost all individuals.
Mass media messages have become omnipresent, and the public sphere is increas-
ingly encroaching upon private communicative situations, providing the interloc-
utors with topics, standpoints, opinions, catchphrases, etc.
This book aims to be a contribution to our understanding of some of the
linguistic phenomena that operate in the public, institutional discourse of the
media. It is primarily interested in how the traditional print and the new online
newspapers enhance the interpersonal dimension of communication by relating
to readers through certain conventional linguistic means, most notably the shift
of grammatical tenses. The overall focus on temporal deixis and its explanation as
an audience-oriented strategy is indicative of the classification of the present book
as a pragmatic analysis.
The book is the result of a long process that ultimately goes back to my doctor-
al dissertation defended at Masaryk University in Brno in 2001. While the thesis,
based on a diachronic analysis of a set of news reports on a single topic, remained
unpublished, most of its novel findings, such as those concerning the temporal
aspects of news headlines (including the conventional deictic projection), and the
personalization patterns observable in the changing forms of reference to social
actors, were never even presented at conferences since my research interests drift-
ed in different (though not unrelated) directions.
When it was eventually decided to present some of the ideas in a book form, it
became evident the whole issue needs to be radically reworked. In order to assure
thematic consistency, the topic became limited solely to temporal deixis in news
stories, with a particular focus on news headlines. The earlier thesis provided little
more than a general structural framework for what is an entirely new work that is
based on new, up-to-date material.
Brno, June 2014

Acknowledgements
The production of the present book would not have been possible without the
inspiration from many people I have met over the past 15 years. They are mostly
colleagues I met at various conferences and during many short teaching stays at
partner universities abroad. Their expertise and enthusiasm have provided me
with constant encouragement and a sense of direction in the diverse research
projects that I have been involved in. I am grateful to them for helping to shape –
directly as well as indirectly – my linguistic thinking about media discourse and
related concepts. They include Richard Janney and Hans Sauer from Munich,
Andreas Jucker from Zurich, Joanna Thornborrow from Brest, Josef Schmied
from Chemnitz, Roswitha Fischer from Regensburg, Isabel Ermida from Braga,
Piotr Cap and Marta Dynel from Lodz, María Martínez Lirola from Alicante,
Eva Lavric from Innsbruck, Martin Conboy from Sheffield, Marie Krčmová from
Brno and many others. Special thanks also to Walt Wolfram from NCSU and
Ronald Butters from Duke University and their colleagues for enabling me to
briefly experience the American academic culture at a time when I was finishing
the manuscript of this book.
I would also like to thank my colleagues from the English Department at
Masaryk University in Brno, in particular the late professor Josef Hladký, a func-
tional linguist and lexicologist, without whose encouragement I would not have
turned towards linguistics, and Professor Ludmila Urbanová, who was the super-
visor of my doctoral dissertation and who has been a very good and supportive
colleague for over a decade. It was in her graduate seminars on spoken discourse
and pragmatics that I got acquainted with some current approaches to the study
of language and became truly fascinated with the analysis of the systematic nature
of human communication across various domains. She also read and commented
on a draft version of this book and offered some valuable feedback. I also owe a
debt to Don Sparling, who was a great model teacher to me during my undergrad-
uate years, and the current head of the English department, Jeff Vanderziel, who
has been instrumental in creating a good working atmosphere.
Most of all, I am greatly indebted to my family for providing me with an envi-
ronment that made it possible to engage in this project and eventually see it to its
xii News and Time
meaningful end. Thanks to Vítek and Zuzanka for bringing so much joy into my
life and Barbora for her endless patience and everything else.
I gratefully acknowledge the help of Matthew Nicholls, who proofread the
main part of the manuscript, and voice my appreciation for the suggestions and
constructive comments made by the two anonymous reviewers. I am aware that
there are many different perspectives on the data but hope that readers will find
some inspiration in my approach to understanding the complexity of news texts –
and will do so despite any shortcomings of the present account, for which I take
the sole responsibility.
List of abbreviations
BC body copy
CDA critical discourse analysis
CT coding time
DD direct discourse
ET event time
H headline
H1 headline in article preview (home page)
H2 headline in news article (article web page)
IC image caption
ID indirect discourse
L lead
L1 lead in article preview (home page)
L2 lead in news article (article web page)
PDT publication date time
PT publication time
PrT printing time
RT receiving time
S subhead
S2 subhead in news article (article web page)
SFG systemic-functional grammar
STR speech and thought representation
List of tables and figures
Tables
Table 5.1 Clausal vs. nominal headlines
Table 5.2 The proportion of syntactically complex headlines
Table 5.3 The proportion of tenses in headlines
Table 5.4 The proportion of tenses in complex headlines
Table 5.5 The proportion of tenses in finite headline segments
Figures
Figure 3.1 The relationship between the coding time (CT) and the receiving time (RT)
Figure 3.2 The sequence of distinct temporal categories (and deictic centres)
Figure 3.3 Projections and simultaneity of time frames in a daily newspaper
Figure 3.4 Temporal reference of the time adverbial yesterday
Figure 4.1 The major macro-structural textual segments of online news
Figure 4.2 The structural expansion of online news stories: the double headline/double lead
pattern (The Telegraph; 3 October 2012)
Figure 4.3 Elaboration of information in two levels of leads (The Telegraph; 3 October 2012)
Figure 4.4 Repetition of headlines and minimal expansion of leads/subheads in online news
stories (The Telegraph; 16 November 2012)
Figure 4.5 The dateline as an explicit anchorage point for temporal reference in an online
newspaper (The Independent; 18 January 2013)
Figure 4.6 Temporal metadata about an article: recording multiple article versions
(The Guardian; 20 February 2013; downloaded at 10.20 GMT)
Figure 4.7 Temporal anchorage of online newspapers – home page
Figure 4.8 Temporal anchorage in online newspapers: articles in news text clusters
(The Times; 21 February 2013)
Figure 4.9 Temporal anchorage and attributions in article bylines
Figure 4.10 The temporal map of a single topic. A chain of hyperlinked articles forming a
dense intertextual network (Electronic Telegraph, February – November 1997,
articles on the case of Louise Woodward; Chovanec 2000: 271)
Figure 4.11 Temporal orientation and directionality of internal links
Figure 4.12 Online article types according to the nature of connecting hyperlinks
Figure 4.13 Placement and pattern of internal links in the 1997 data: absence of temporal
indication for accompanying articles and text-internal links (Woodward’s
lawyers ask for manslaughter sentence, Electronic Telegraph,
www.telegraph.co.uk, 4 November 1997)
Figure 4.14 Placement and pattern of internal links in the 1997 data: calendrical date +
headline
xvi News and Time
Figure 4.15 Placement and pattern of internal links in the 2012–2013 data (British climber
says ‘dream come true’ to scale Antarctica’s Ulvetanna; The Telegraph;
31 January 2013; article retrieved on 31 January 2013)
Figure 4.16 Live content on an archived article page in online newspapers (British climber
says ‘dream come true’ to scale Antarctica’s Ulvetanna; The Telegraph;
31 January 2013, retrieved on 22 February 2013)
Figure 5.1 The scale of the relative importance of headline constituents
Figure 6.1 The universal temporal reference of the present time (Quirk et al. 1985: 176)
Figure 6.2 Present tense in past time territory (Huddleston and Pullum 2002: 130–131)
Figure 6.3 Double projection of deictic centres: the simple present tense in headlines
referring to past events
Figure 7.1 The anteriority (non-recency) meaning of the past tense in headlines
(A = receiving time RT (readers’ “now”), B = past temporal watershed,
e.g. the last publication date)
Figure 9.1 The transformations of the temporal and notional components in process chains
(sample news texts)
Figure 9.2 Encoding and decoding of the double tense shift pattern
Figure 9.3 Temporal re-evaluation within a co-referential process chain
Figure 9.4 Functions of lexical and temporal components in process chains: the case
of lexical repetition and the double tense shift
Figure 9.5 Functions of lexical and temporal components in process chains: lexical
repetition and single tense shift
Figure 9.6 Multiple levels of textual segments in online news (H – headline, L – lead,
S – subhead, BC – body copy)
Figure 9.7 Multiple summaries on the newspaper home page and the article web page
(The Guardian; 30 January 2013)
Figure 9.8 The linear sequence of cohesive items within the co-referential chain
Figure 9.9 The parallel segmentation of online news texts on the newspaper home page
and the article web page (The Guardian; 30 January 2013)
Figure 9.10 The progression of the process chain from the home page to the article web page
Figure 9.11 Non-finite ellipted headlines and the simple past in the subhead (The Guardian;
31 January 2013)
Figure 9.12 An article with a triple tense pattern (The Guardian; 31 January 2013)
Figure 9.13 Interaction between a process chain and a selection of other co-referential chains
Figure 10.1 Narrative analysis of a sample news story and its simplified time structure (after
Bell 1991, The Independent; 16 November 2012)
Figure 10.2 A complex timeline of a news story distinguishing between ‘events’ and ‘states’
(Delin 2000: 21)
Figure 10.3 The temporal sequence of events and states in the order of occurrence in the
‘Record BP Penalty’ story
Figure 10.4 Article timeline across the main textual segments
Figure 10.5 The model structure of a news text (Bell 1991: 171)
Chapter 1
Introduction
In this book, I consider how temporality is encoded in news texts and how shifts
of grammatical tense, first in the news headline and then in the rest of the article,
are pragmatically motivated by the need to negotiate a shared temporal context
with the readers. The book explores a topic that is very central to media discourse
because of the media’s preoccupation with time: news needs to be both current
and delivered to the audience as quickly as possible, subject to the news produc-
tion cycle of a given media channel. It is hardly surprising that recency is one of
the top news values associated with the nature of the news story.
The aim of the book is to reveal the operation of the textual rhetoric of tradi-
tional written media by considering the relatively straightforward phenomenon
of temporal deixis, i.e. the way the temporal anchorage of past-time events is en-
coded. The starting point is the interpretation of the conventional non-deictic
present tense found in English headlines as an interactive device. Instead of pro-
viding temporal anchorage of the main event, it is oriented towards establishing
the illusion of a shared discourse space. Against this background, other tenses
are analysed with the aim of describing and explaining some of the patterns that
arise out of the combination of diverse textual and intertextual features, the most
important of which is heteroglossia. The interpretation of tense in headlines as an
interpersonal device is supported by considering the differences in how temporal
deixis is encoded in different segments of news stories.
The book also develops our current understanding of the pragmatic nature
of news texts by extending its approach to cover online news that is much more
dynamic and immediate than its print counterpart. Hypertextuality marks the
transition from traditional news texts as relatively independent units to a situ-
ation in which a given news text is not only connected with other news items
and multi-modal materials in a complex news cluster on a single topic, but also
interlinked with previous news texts in the news archives. The compression of
time resulting from the ease with which readers can move back in time affects the
increased need for explicit encoding of time in the news story’s attribution and
byline. In addition, the presentation of news previews on the papers’ home pages
has brought an additional textual level that needs to be, as argued in this book,
taken into account in the analysis of online news texts.
2 News and Time
It is believed that the analysis will shed new light on some traditional con-
cepts analysed in media discourse analysis, as well as offer novel insights into the
nature of online news texts.
1.1 Goals and objectives
One of the most striking and noticeable linguistic phenomena characterizing

modern English news discourse is the use of the present tense in headlines to refer
to past-time events. This convention is so notoriously known that there is hardly
any speaker, native or non-native, who would not be able to recognize, name and
correctly interpret such brief texts as the following:
Six killed as 8.2 magnitude earthquake hits northern Chile
(The Telegraph; 2.4.2014)
However, as is often the case, some phenomena that appear as relatively simple
and self-evident may turn out to embody a surprising extent of complexity that
involves not only the phenomenon at hand but also a host of other related issues.
Thus, if we want to understand how the above-mentioned headline operates and
what mechanisms allow us to correctly interpret the temporal situation encoded
in it, we may have to consider more than just the conventionality of the present
tense (‘hits’) and the operator deletion in the passive construction (‘Six killed’),
which are the salient formal features that trigger our instant recognition of the
text in question. While these conventional forms are explicable in pragmatic
terms, as the result of a complex projection of deictic centres, a true understand-
ing of temporal deixis in news texts – with the present tense in headlines being
a classic example – requires us to consider temporal deixis in a broader context.
Thus, for instance, the coverage of the event initially broken in the headline
above is continued in the subhead with the present tense providing the initial
description of the event (cf. ‘triggers’ and ‘alerts’):
Powerful quake triggers tsunami waves of two metres on Chilean coast and
alerts along 3,000 miles of coastline in Central and South America
(subhead; Six killed as 8.2 magnitude earthquake hits
northern Chile; The Telegraph; 2.4.2014)
However, the temporal situation then changes in the first paragraph of the text
(the lead) where a switch into the simple past tense occurs, as indicated by the
forms of verbs that refer to the same past-time moment (cf. ‘hit’ and ‘lashed’):
Chapter 1. Introduction 3
A powerful 8.2-magnitude earthquake hit Chile’s Pacific coast on Tuesday,

killing at six people as tsunami waves of more than two metres lashed the
shore. (lead; Six killed as 8.2 magnitude earthquake hits
northern Chile; The Telegraph; 2.4.2014)
While the repetition of some of the key information about the event derives from
cyclical presentation of content occasioned by the inverted pyramid structure, we
can see that the verb phrase undergoes a significant transformation in its gram-
matical marking of tense. This indicates that temporal deixis is deployed very dif-
ferently in the different segments of news texts. The different functions of verbal
tense in the headline as opposed to the other structural segments can be account-
ed for in pragmatic terms, as the result of a specific textual rhetoric of news texts.
Let us conclude these introductory observations by briefly contrasting the
above-mentioned example with the coverage of the same event in a different
newspaper, with the aim of hinting at some aspects of systematic variation in
temporal deixis.1
Tsunami warning and evacuation of thousands after earthquake in
Chile
Deaths reported and people on coast spending night in the hills after mag-
nitude 8.2 undersea quake triggers emergency
An earthquake of magnitude 8.2 has jolted northern Chile, triggering a tsu-
nami alert and the evacuation of thousands of people from coastal areas. At
least five people were killed and more than 300 women escaped from a coastal
prison. (The Guardian; 2.4.2014)
While the headline is non-finite (and, thus, not explicit about the real-time place-
ment of the event, leaving it up to the reader to infer), the subhead uses the typical
headline syntax by ellipting the operator (‘Deaths reported’) and including the
present tense (‘triggers’). Arguably, in the absence of verbs, the effect of the earth-
quake is presented in the headline more as a state than an action (cf. ‘warning’ and
‘evacuation’ in the Guardian example versus ‘hits’, ‘triggers’ and ‘alerts’ in the Tele-
graph example). As a consequence, the Guardian article communicates a stronger
sense of current relevance, which is further enhanced by its use of the present per-
fect tense in the lead (‘has jolted’). The appearance of this tense is made possible
1. Both articles were retrieved from the online versions of the newspapers on the day of
the event and at the same time. Sources: http://www.theguardian.com/world/2014/apr/02/
chile-earthquake-sparks-tsunami-warning-and-evacuation-of-thousands and http://www.
theguardian.com/world/2014/apr/02/chile-earthquake-sparks-tsunami-warning-and-
evacuation-of-thousands.
4 News and Time
by the avoidance of any adverbial of time that would, as in the first example (‘hit
… on Tuesday’), establish an unequivocal past-time anchorage.
Clearly, journalists and editors are able to skilfully exploit discourse patterns
and genre conventions by selecting meaningful forms from the set of options that
are available to them at a given point of occurrence. The relatively straightforward
situation observable in the examples above is frequently complicated by not only
the non-chronological structure of news texts but also their heteroglossic compo-
sition when diverse accessed voices are cited. Moreover, online news texts appear-
ing on 24/7 news websites that provide continuous coverage are not only updated
with newly emerging information but are also much closer – in terms of real
time – to the events reported than corresponding news items in print newspapers.
All that makes it worthwhile to explore the systematic patterns of the coding of
temporality in news texts and some of the specific features found in online news.
The analysis of temporal deixis in news texts offered in this book revolves
around three main research questions that are reflected in the division of the book
into its three parts. Together, these research questions aim to provide a compre-
hensive account of temporal deixis in news texts.
Thus, the first part of the book seeks to answer the following question:
– How does temporal deixis operate in traditional news texts, and what are
the specificities of temporal anchorage in online news articles?
This question addresses the nature of the encoding and decoding processes
involved in the production and reception of news texts and what it implies for
the linguistic coding of time. By viewing these phenomena from the perspec-
tive of traditional pragmatics, I consider how a pragmatic model of deictic
centre projection can represent this situation in print news and how online
news, on account of its hypertextuality and process-rather-than-product na-
ture, stands out from traditional print news.
The second part of the book focuses on the explanation of specific linguistic phe-
nomena observed in the analysed data set, seeking to address the following issue:
– Which verbal tenses are used in headlines, and what is the motivation for
the variation?
This question deals with the variation of verbal tenses in a ‘paradigmatic’ sense,
by viewing them as a set of options available to the editor when formulating
headlines. Thus, the investigation seeks to explore what factors – semantic as
well as pragmatic – motivate the choice between deictic and non-deictic tense
in headlines and how the convention is affected by the utilization of accessed
external voices. The issues related to this dimension of news texts form what
I refer to as ‘textual rhetoric of headlines’.
Finally, the third part of the book adopts a more global view of news texts, by
considering and explaining temporal deixis and tense variation as a systematic
macrostructural pattern. This idea is captured in the final research question:
– What is the temporal organization of news texts beyond headlines, and how
can the regularities in some of the organizing principles be explained?
This question elaborates the notion of temporal variation by developing the
analysis along the ‘syntagmatic’ axis, i.e. by considering the linear progression
of news texts across their individual structural segments. With reference to the
inverted pyramid structure of news texts, the individual segments (including
the headline) are documented to have different functions that are systemat-
ically reflected in the patterns of temporal deixis. The issues related to this
dimension of news texts form what I refer to as ‘textual rhetoric of news texts’.
Each of the broad reseach questions requires the exploration of a number of issues,
such as the role of time adverbials when providing specific temporal anchorage,
the function of heteroglossia to access external voices, and the multi-functional-
ity of the verb phrase in different structural segments of news texts. While those
more specific phenomena are analysed in independent chapters, they are all re-
lated to the overall aim of the book and its structural framework in terms of the
three parts identified above.
The discussion of the issues outlined above aims to show that temporal deixis
is a crucial pragmatic phenomenon that is conventionally used in newspaper dis-
course (both printed and online) in a relatively complex way in order to develop
the interpersonal dimension of media communication. The fundamental princi-
ple on which news texts operate is the projection of certain temporal deictics to
the anticipated time of the texts’ reception, whereby the impression of a shared
context is constructed. Temporal deixis is understood here in the broad sense
as the textual specification of time, realized through verbal tense and adverbials
of time, rather than in the narrow sense of pragmatic anchorage of a particular
proposition or act with respect to the moment of utterance.
The aim of the book is also to offer a pragmatic explanation for a set of tense
shifts observable in news texts. The shifts, used by editors when formulating the
news texts, refer to the relation between the event and the grammatical tense cho-
sen to encode the event. Editors formulating headlines can perform two switches,
each for a different purpose. The first switch is towards the conventional present
tense to refer to a past-time event, by means of which the proper genre charac-
teristics of the text at hand can be emphasized. The second switch operates in the
opposite direction: away from the conventional present tense and towards other
tenses, which – paradoxically – can result in the seemingly normal situation of
semantic correspondence, i.e. the use of the past tense for past-time events. As a
6 News and Time
result of the second shift, headlines become more personal, authentic, and con-
versational. Another dimension of tense shift operates not on the paradigmatic
level of tense choice within the headline but on the transition from the headline
to the lead and eventually to the body copy, where we find the typical pattern of
tense switch from the non-deictic simple present to the semi-deictic present per-
fect and eventually the deictic simple past tense.
1.2 Approach
News texts are approached in this study from the socio-pragmatic perspective.
The central pragmatic issue, namely temporal deixis, is treated as a phenome-
non through which the writer/speaker positions a text with respect to his or her
audience. In that sense, it is an interpersonal element that – together with other
linguistic features – co-constructs the interactive nature of the text. While prag-
matics is understood in its broad sense as the use of language in social settings, it
is the interactive aspect that is central in our use of the term. When analysing the
data, the findings are interpreted in a multidisciplinary way in connection with
some of the central concepts of the metafunctional framework employed in sys-
temic-functional linguistics, and in close contact with the stylistic/narratological
concept of heteroglossia. The presence of non-authorial external voices is shown
to have a crucial role in shaping the formal composition of the message and affect-
ing the temporal and deictic situation in news texts.
1.3 Data
The main part of the analysis is based on 837 hard news headlines from the on-
line versions of four British daily newspapers: The Guardian, The Independent,
the Telegraph, and The Times. The data set, retrieved from the newspapers from
autumn 2012 through spring 2013, is described in more detail in Chapter 5. The
analysis is predominantly qualitative, striving to detect patterns of tense use and
offer explanations for the various forms identified. Quantitative analysis is used in
Chapter 5 in order to provide a general description of the composition of the data,
with a focus on the proportion of the individual grammatical tenses used and the
finite vs. non-finite and nominal character of headlines. For illustrative purposes,
some additional data from two earlier studies are used.
Although the headlines from the four newspapers in the sample are pro-
cessed together, the data set contains a significant degree of variation. There are
noticeable differences between as well as internal variability within the individual
newspapers. In general, the different styles found in many newspapers are partly
related to the characteristics and the socio-economic status of their target audi-
ences (cf. Jucker 1992; Lennon 2004). While the newspapers analysed here are
all classifiable as broadsheets, their audiences do differ in terms of their social
background and political preferences. The style adopted by particular papers can
also be used as a marker of their distinct identity. Thus, for example, the dual
tautological structure observed in many headlines in The Independent (see Sec-
tion 7.3) has not been found among the other newspapers in the data set. It is
also evident that each of the newspapers has adopted a different editorial practice
as regards the way online news stories are presented on their home pages. The
Guardian, for instance, tends to have quite short headlines that fit on a single line.
The Telegraph’s online presentation of news differs in that its headlines are longer,
and The Independent’s headlines often trail across several lines in the rather nar-
row textual columns on the paper’s web page. While all newspapers provide brief
article previews for most of the news stories, The Times provides a more extensive
introduction for one or two of the major news stories of the day.
While it is not the aim of the present book to document the variability within
the sample and establish any sociolinguistically relevant patterns of correlation
between the language forms used and the readership targeted, the internal diver-
sity should not be entirely disregarded, whatever its motivating factors. It is held
that the internal heterogeneity in the sample – within the limits of what are con-
sidered as mainstream newspapers – does not prevent the identification of some
common patterns in the way temporality is expressed because the newspapers
operate, after all, with the same or very similar tools of trade. These consist of the
shared norms of the journalistic profession, namely the general conventions of
headlinese, and the preferred textual structure of hard news items.
1.4 Overview of chapters
As outlined in the specification of the main research questions in Section 1.1, the
book is organized into three sections that cover the following broad areas: the
pragmatic description of temporal deixis in news texts and online news (Part I);
the temporal situation in headlines according to the editor’s choice of tense
(Part II); and the temporal situation beyond headlines (Part III). The individual
sections partially overlap because the primary aim of the book is to demonstrate
that the choice of tense – with respect to news texts in general and headlines in
particular – is the outcome of a complex process that involves not only the implic-
it conventions of the genre but also the way in which diverse heteroglossic voices
are incorporated. It is argued that the choice of tense not only has a local function
8 News and Time
with respect to the textual segment of the news text in which it is located, but that
it is also inextricably intertextually linked to the forms and functions of the other
segments as well as other related texts.
The basic theoretical foundations are briefly expounded in Chapter 2, which
locates the book at the intersection of pragmatics and functionally-oriented lin-
guistics and explains the notion of heteroglossia – borrowed from narratology –
that is eventually used in order to explain the operation of headline conventions
and the resulting choices of verbal tense in news texts. Although temporal deixis
is approached as a pragmatic phenomenon, it is argued that integrating perspec-
tives from several disciplines can bring novel insights into our understanding of
the complexity of temporal relations in news texts.
The main theoretical account is provided in Chapter 3, which deals with tem-
poral deixis in news texts. The account starts by discussing the issue of deictic
centre as the crucial construct for interpreting spatial and temporal positioning
through deixis. It goes on to detail the deictic projection that is involved in tense
shifts. A model for the projection of deictic centers in news texts is proposed, tak-
ing into account the main temporal situations involved (event time, coding time,
publication time, and receiving time). Since the projection results not only in the
shift of tenses but also the reformulation of the relevant adverbials of time, the dis-
cussion includes deictic adverbials as well because their form serves as additional
evidence of the deictic perspective from which a news story is encoded/decoded
(cf. the projection of the adverbial today into yesterday in traditional print news).
Chapter 4 examines temporal deixis in online newspapers. It notes the dynam-
ic presentation of news content in terms of such features as the multiple textual
levels of news texts, the hypertextual organization of online news content, and the
presentation of news as an open, evolving coverage of events. The analysis distin-
guishes between the news text on the newspaper home page, consisting of an arti-
cle preview (headline + lead), and the news text on the article web page, consisting
of the entire article. Adopting such a two-level approach, we can then approach
the news text as a complex conglomerate of textual segments through which a
reader moves when processing the news text. Attention is also paid to the tempo-
ral anchorage of online news texts and the different pragmatic implications of the
possible ways of indicating the publication time of news articles on newspapers’
home pages and article web pages. Finally, it is shown how hypertextuality – as one
of the defining features of the online environment – has transformed the relatively
isolated news texts found in print media into textual conglomerates consisting
of mutually interlinked synchronic article clusters and diachronic news chains,
enabling readers to move through a given story’s development in time from the
latest update back to its first appearance in the media. Last but not least, the online
media can not only offer live coverage of events but also include live content that is
appended to archived news texts and is frequently personalized to specific readers.

In this way, there is always some current content on the online page.
The second part of the book opens with Chapter 5 on temporal deixis in head-
lines. It notes that information that is circumstantial to the main event – which
typically includes location and time – tends not to be encoded through adverbi-
al prepositional phrases. While location is typically expressed through adjectival
premodification, giving rise to complex nominal groups, the temporal anchorage
of the event is usually not provided at all. Instead, it needs to be inferred on the
basis of the presumption of recency that is at the core of the newsworthiness of
almost any event reported in the media. Temporality is not normally expressed
through verbal tense in headlines because tense is not used deictically. Instead,
it is conventionally shifted, which is one of the defining features of the style of
English headlines.
The shift of tense in headlines towards the non-deictic present tense is the
focus of Chapter 6. This tense, which encodes past-time events in headlines, can
serve a number of grammatical functions and can be considered as atemporal be-
cause of its semantic properties. With reference to the semantics of what is known
as the historic present, the past-time reference of the present tense is interpreted
as adding vividness to the event reported. In pragmatic terms, it constructs the
impression of temporal co-presence between the event and the anticipated time
of the news text’s reception. Since the spanning of the two time zones is essentially
reader-oriented, it is understood as enhancing the interpersonal dimension of the
communication.
Chapter 7 completes the picture of the temporal situation in headlines by
considering less canonical situations such as the expression of the future and the
occurrence in headlines of other tenses than the simple present. The future is
encoded through the present tense, to-future and will-future, the latter typically
with the modal meaning of expressing either volition or prediction. The present
perfect is found to enhance the impression of ‘current relevance’, particularly in
narrative headlines and headlines summing up trends and developments, where
it can co-occur with adverbials of time. The most significant tense, however, is
the simple past, which constitutes a marked choice of tense in headlines. It oc-
curs in a number of specific functions. Most importantly, it identifies heteroglos-
sia, i.e. stretches of text formulated in non-authorial accessed voice as direct/
indirect speech. When present within the paper’s authorial voice, the past tense
also becomes either an explicit marker of non-recency (since recency is encoded
through the present tense) or a device that can help manage information flow
in the headline by backgrounding a particular proposition whenever the focus
of the news story is on some other, more newsworthy element. The chapter also
describes the appearance of the past tense in satellite articles giving background
10 News and Time
information within clusters of articles on related topics, and elaborates on the

diverse forms of heteroglossia in various types of headlines as well as on the ways
in which the single or multiple switch of voices can be indicated. The ultimate
effect of heteroglossia is that it brings about a second shift of tense in headlines:
the first shift, represented by the encoding of a past-time event in the present
tense, is supplanted by the second shift, with the conventional present tense not
used in headlines and, instead, replaced by the past tense. While there appears to
be no surface discrepancy between the event (in the past) and the tense encoding
the event (the past tense), the double tense shift pattern removes the imperson-
ality associated with the block language of headlinese. The past tense – enabled
through the presence of the external voice – thus makes the relevant section of the
headline closer to ‘normal’ conversational language, i.e. language as unaffected by
headline conventions.
Chapter 8 supplements the discussion of the temporal situation in headlines
by considering the role of auxiliaries that are ellipted in a significant proportion of
headlines. The non-finite verb form is typical in passive constructions. It may also
give rise to ambiguities if the non-finite past participle is read as a finite verb form
in the past tense, i.e. a situation that is not – as shown in the previous chapter – so
unusual. It is argued that the absence vs. the presence of auxiliaries in the verb
phrase is, in a sense, parallel to the contrast of the present tense vs. the past tense
in headlines, since the former option is interpreted as unmarked (i.e. in harmony
with headline conventions) and the latter as marked (i.e. possible under some
special circumstances that justify the deviance from the norm). It appears that
elliptible auxiliaries in headlines are present for the sake of emphasis, as a marker
of accessed voice, in order to provide a semantic specification and, potentially, to
foreground major news stories.
The final part of the book moves beyond headlines into considering the tem-
poral situation in the lead and the body copy. Chapter 9 addresses the issue of
the textual recurrence of verbal elements that refer to the same event in the three
main structural segments, approaching it from the point of view of cohesion anal-
ysis. This enables the tracking of cohesive elements across the three central textual
segments of news stories – headlines, leads and the body copy – and reveals the
existence of a systematic pattern of tense shifts across these segments. The indi-
vidual cohesive elements (linguistic forms or, more generally, semantic concepts)
can be assembled into matched sequences of several parallel co-referential chains,
with the individual elements gradually developing the main proposition. The ex-
pression of temporality is mapped across the three structural segments of news
stories because in each segment – the headline, the lead and the body copy – it is
carried by the relevant verb phrase that forms the grammatical core of the respec-
tive proposition. Although the situation in specific news articles tends to be rather
complicated (e.g. on account of the presence of subheads, the existence of nomi-

nalizations, and the appearance of diverse transformations of the verb phrase), it
appears that the three structural segments are organized on the basis of the ide-
alized pattern of two tense shifts, namely ‘present tense → present perfect → past
tense’. In online news, which contains more summarizing segments on account of
the existence of the home page preview, the double tense shift pattern tends to be
partially replicated in the preview. It is argued that verbal tense serves different
functions in the three structural segments: it is used interpersonally in the head-
line (where the tense is conventionally shifted towards the present tense) and the
lead (with a shift into the present perfect to express ‘current relevance’), and ide-
ationally in the body copy (where the past tense is used in its deictic function to
anchor the event in the past time). The lexical component of the verb phrase then
articulates the textual function, i.e. it provides for the necessary cohesion between
the three structural segments.
The discussion of the temporal structure of news texts concludes with Chap-
ter 10, which addresses non-chronology as a phenomenon that arises from the in-
verted pyramid structure of news texts. There are several complementary ways of
approaching news texts: we can analyse them in terms of the narrative elements of
the news story, the thematic structure of the news event in terms of event frames,
and the cyclical pattern of re-expressing content throughout the news text. It is
argued that the narrative element of Abstract, which is the category with the cru-
cial discourse-organizing role in headlines and leads, holds the key to one of the
major macrostructural features of news texts, namely the cyclical repetition of
the basic information in various textual segments. The orbital organization of the
content of the news story is reflected in the temporal structure, where repeated
references to subevents located in the past and the future are made. Owing to the
orbital structure in the body copy and across the three main structural segments,
the non-chronology found in news texts is not haphazard but has a perceptibly
patterned format.
To sum up, the present book offers an interpretative approach to the analysis
of print and online news texts, hoping to elucidate the complex nature of tempo-
ral deixis and offer a plausible explanation for some of the most noticeable and
best known phenomena of news discourse – such as the present tense in headlines
and its relation to the inverted pyramid structure of news texts. It also aims to
identify other elements that, though less known, are equally fascinating, such as
the double tense shift found across the main structural segments. The book sets
out to ready the ground for further systematic pragmatic investigations into the
area of temporality in news texts, whether in regard to their traditional printed
format or diverse modern forms such as online news.
Part I
Temporal deixis in print and online news

Chapter 2
Theoretical foundations
This section sets out the major theoretical approaches that have influenced the
perspective adopted in the present book. The approach draws on the overlapping
traditions of pragmatics and discourse analysis, adopted with the aim of under-
standing specific pragmatic phenomena and their grammatical, discursive and
rhetorical manifestations. Language use is also related to social interaction and
practice and explained against the background of the functional approach that
stresses that language has an important relational aspect realized along the inter-
personal dimension. The discussion is complemented with an outline of several
studies of mass media discourse which have provided an inspiration to the so-
cio-pragmatic interpretation of news discourse in the chapters to follow.
2.1 Pragmatics
Deixis, as the main phenomenon studied here, is a linguistic category that con-
nects the user, the specific context and the communicative event. The concept has
been studied in traditional pragmatics alongside such concepts as presupposition,
implicature, speech acts, politeness, etc. that often claim universal application and
have been considered as falling within the scope of ‘utterance-pragmatics’ or ‘mi-
cropragmatics’ (Cap 2011: 51). However, in my approach, I adopt a broader sense
of pragmatics – one that is common in the European tradition, which goes beyond
the philosophical and intercultural aspects of pragmatics and emphasizes the so-
cial dimension of language and language use (cf. Levinson 1983; Verschueren 1999;
Verschueren et al. 2003). As Taavitsainen and Jucker (2010: 5) note, this concep-
tion of pragmatics “takes a sociologically-based approach and wants to understand
the patterns of human interaction within their social conditions”. It is evident that
Europeans treat pragmatics differently – as a broader socio-interactional concep-
tualization of language use, not just the inferential-cognitive interface that is more
common in Anglo-American approaches to pragmatics (cf. Jucker 2012: 198). This
also entails that the methodologies of pragmatic analysis of language data can be
quite diverse and will inevitably reflect a researcher’s specific aims and theoretical
background applied to language interaction between speakers.
16 News and Time
Rather than providing for a set of some clearly defined methodological tools
and distinct theoretical constructs, the broad conception of pragmatics is more
of a shared general outlook on language in use that seeks to understand the rela-
tionship between speakers, language form, discourse structure and the variety of
contexts in which interactions are embedded (social, cultural, historical, person-
al). Pragmatics describes regularities across speakers and genres as well as idio-
syncrasies in language use in particular situations. It draws on various established
disciplines (discourse analysis, conversational analysis, critical discourse analysis,
systemic-functional grammar, interactional sociolinguistics) rather than standing
in contrast to them. For instance, Locher and Graham (2010: 2) – when defining
‘interpersonal pragmatics’ – state broadly that it is a perspective on “the interper-
sonal side of language use” that considers “facets of interaction between social ac-
tors that rely upon (and in turn influence) the dynamics of relationships between
people and [that looks at] how those relationships are reflected in the language
choices that they make”.
In its broad sense, pragmatics is thus increasingly understood as a linguis-
tic approach to human communication, as opposed to structurally and formal-
ly-oriented research traditions that operate with static and decontextualized
abstractions. In this way, Bublitz and Norrick (2011: 4) extend the definition of
pragmatics to cover “the scientific study of all aspects of linguistic behavior”, ar-
guing that it is “fundamentally concerned with communicative action in any kind
of context”. As regards actual pragmatic analysis, the micro- and the macro-levels
are generally studied in a close relationship, although pragmatics considers the
linguistic form as a point of entry for understanding more general patterns. Thus,
“specific concrete linguistic events (and their contexts) are only relevant in as far
as their properties and constraints can be integrated under analogous conditions
into a general concept of language and communication” (Bublitz and Norrick
2011: 4).
The broad perspective view of pragmatics was described in an early study by
Leech (1983: 10) as ‘sociopragmatics’. According to Weisser (2013), this view “sees
pragmatics as a function of language that influences the other levels and incorpo-
rates a larger situational context that also includes sociolinguistic factors”. While
Leech’s use of the term essentially points out the general social orientation of the
discipline, the latter definition might give the impression of a more narrow scope.
However, the term ‘sociolinguistics’ itself suffers from a referential vagueness sim-
ilar to ‘pragmatics’ in that it can be conceived of either in a narrow sense (in the
tradition of large-scale variation studies) or in a broad sense (including any study
of language use in society and in a social context, cf. Downes 1998: 15). It may
be difficult to draw an exact boundary between pragmatics and sociolinguistics
(cf. Levinson 1983: 29) because there is a great deal of overlap in the ‘grey zone’
Chapter 2. Theoretical foundations 17
that the two disciplines share. As argued by Culpeper (2011: 2), “a focal point
for sociopragmatics is the way in which speakers exploit more general norms to
generate particular meanings, take up particular social positions, and so on”. In
his view, sociopragmatics “is not simply concerned with mapping regular patterns
of usage in interaction, as might characterize much work in sociolinguistics, but
with understanding how those regular patterns are used and exploited in particu-
lar interactions” (2–3).
It is this conception of pragmatics that underlies my approach to the phenom-
ena studied in this book. The primary concern for an essentially micro-pragmatic
phenomenon (temporal deixis) maps this pragmatic phenomenon as having a
number of context-specific and conventional instantiations on the level of lin-
guistic form, yet it is interpreted in this book in a constructivist way on a broad-
er contextual level as articulating a particular form of relationship between the
producer of the text and its recipients. Temporal deixis is thus explored on the
interface between language use and language users who are brought together as
participants in the production and reception of mass media texts.
In my socio-pragmatic approach to the data at hand, however, I strive to in-
tegrate several other approaches in order to provide a multifaceted explanation of
the operation of temporal deixis in news texts.
2.2 Functionalism and Halliday’s metafunctions
Although the present work is pragmatically-oriented, it is also influenced by

the functional understanding of language as a system and a purposeful activity
through which meaning is constructed and interpreted in specific sociocultural
contexts. In a classic study, Halliday describes language as a ‘social semiotic’. In
this definition, language “consists of text, or discourse – the exchange of meanings
in interpersonal contexts of one kind or another” (Halliday 1978: 2). The relation-
ship between language and situation can be described as a predictive potential:
language is “what the speaker can do”, i.e. what he or she “can mean” (27–28) in a
given situation. In this sense, language is viewed as a “meaning potential”.
At the same time, however, the potential is inextricably linked to a specific
communicative situation that provides the limits within which interlocutors can
be expected to operate with respect to their linguistic behaviour. With his con-
cept of ‘register’, Halliday proposes a method of relating the typical lexico-gram-
matical features that are characteristically associated with particular situations
of language use – hence the predictive effect of extralinguistic situations on the
language used.
18 News and Time
Although the concept of register itself is not without problems, its framework
has proved influential by inspiring many to conceptualize language and language
use along the three crucial dimensions of field, tenor, and mode. In Halliday’s
theory, there are three corresponding language meta-functions: the ideational
(language as content), interpersonal (language as interaction), and textual (lan-
guage as text). In the present work, the functions are taken as one of the points
of departure for considering the socio-interactional interpretation of news texts
and their structural components. The ideational function is crucial in the jour-
nalist’s construction of meaning in a news text, while the interpersonal function
frequently motivates the choice of specific lexico-grammatical forms with the aim
of developing the intersubjective dimension – it contributes to the positioning of
the text’s author as well as its recipients. The textual function is important for our
understanding because the value of specific linguistic forms partly derives from
their realization in a particular textual (discursive) context. Since textuality con-
cerns the construction of coherent and cohesive stretches of text beyond the sin-
gle utterance/sentence, meaning also arises from the consideration of a linguistic
form beyond its physical point of occurrence. An analysis from the textual per-
spective enables us to trace how genre conventions work and look for reasons for
any deviations from or violations of predicable patterns. The textual dimension
also applies to situations in which individual independent texts are organized into
wholes on a hierarchically higher level, e.g. in the form of news articles in a single
newspaper forming a text colony (cf. Hoey 2001), or where hypertextuality in
the online environment results in the complex linking of a specific article with a
number of other news and non-news texts in an extensive hypertextual network.
The three dimensions are interconnected: the author of a text constructs a
discursive representation of an event (ideational dimension), while simultaneous-
ly negotiating a certain position towards the intended recipients (interpersonal
dimension), and does so by drawing on various text-forming as well as intertex-
tual resources (textual dimension). If we see text construction as such a multi-di-
mensional act performed by a speaker, then the three dimensions can, arguably,
be integrated in the broad view of pragmatics subscribed to here. In other words,
pragmatics is not, as the more traditional or narrow conception would have it,
seen as an additional linguistic level on a par with and separate from semantics
and grammar. It is an approach rather than a strict methodology, an approach
that emphasizes the social dimension of language use and its orientation to users.
However, the role of pragmatics needs to be explored with slightly more detail,
particularly with a view to some of the principles of systemic-functional gram-
mar (SFG), as it developed from Halliday’s early social-semiotic ideas (cf. Halliday
1978; Halliday and Matthiessen 2004). The relationship between these disciplines
is far from simple or free from controversy. While SFG combines register and
genre theory (Martin and Rose 2007), it has also strongly focused on interpersonal
meanings and evaluation, particularly as embodied in the more recent Apprais-
al Theory (Martin and White 2005; Bednarek 2006; Martin and Rose 2007). The
approach is based on the analysis of the positions that speakers adopt and negoti-
ate with respect to each other through alignment/disalignment, the attitudes they
express, and the evaluations they build into their utterances. From the European
perspective, these are issues that may be treated within the scope of interaction-
al pragmatics, since they are related to the dynamic meaning-making work per-
formed by language users in their interactions and communicative encounters.
Dealing with the same phenomena, pragmatics – as the more general approach –
may be more inclined to subsume the perspective of SFL than vice versa. That
appears to be so, particularly with respect to the more orthodox view under which
SFG declares itself as not compatible with pragmatics (as long as the latter is seen
as a distinct linguistic level rather than a more general approach), since SFG prior-
itizes the role of grammar. Thus, as Thibault and van Leeuwen (1996: 567) argue,
“the whole range of meanings is constructed through systems of choices in lexico-
grammatical and other semiotic codes”.
My approach is not intended to add a pragmatic perspective to SFG but the
other way round: to enrich and extend the pragmatic explanation of concrete
linguistic phenomena with reference to some concepts and tools of SFG, in the
belief that such cross-fertilization may enhance our understanding of the nature
of news texts.
Thus, for instance, I share the view propounded by SFG that texts are consid-
ered as dialogic because they construct intersubjective meaning: the expression
of an evaluative stance is an act of discursive positioning that places the author
and the recipient in specific roles. That is particularly acute in case of mass com-
municative texts because they “construct for themselves an ‘envisaged’, ‘imagined’
or ‘ideal’ reader, since it is with this putative addressee that the speaker/writer is
presented as more or less aligned/disaligned” (Martin and White 2005: 95). With-
in pragmatics, discussions of discursive positioning can be found, for instance,
under the rubric of the much-researched notions of mitigation and hedging (cf.
Caffi 2007; Kaltenböck et al. 2010). For SFG, the traditional understanding of
hedging as the expression of the author’s epistemic stance towards the truth value
of the proposition is considered limiting (cf. White 2011: 21). By contrast, the
dialogic perspective shifts the focus from the relationship between the author and
the propositional content (as found in traditional pragmatic accounts of hedg-
ing) towards the more general idea of discursive and intersubjective positioning
in SFG. In harmony with its emphasis on lexico-grammar, the Appraisal Theory
goes on to postulate an exhaustive catalogue of criteria and linguistic forms that
provide an account of evaluation as ‘interpersonal functionality’, i.e. as a system
20 News and Time
that enables speakers to take stands and “engage with socially-determined val-
ue-positions” (White 2011: 14) that eventually serve as a basis of speakers’ align-
ment to or disalignment from other individuals.
Within the SFG framework, the dialogic approach has been systematically
applied to various aspects of media discourse, such as news texts (Iedema et al.
1994), online news (Knox 2010), and the verbo-visual interface in print news
(Bednarek and Caple 2012). Some authors also operate with the notion of ‘voice’
inspired by Bakhtin’s dialogism, paying attention to monoglossic and heteroglos-
sic formulations in news texts (Martin and White 2005: 100) as well as distin-
guishing various evaluative ‘keys’ between such dialogistic positionings as the
‘reporter voice’, ‘correspondent voice’ and ‘commentator voice’.
The notion of voice also forms a crucial theoretical underpinning in the ac-
count offered in this book, because, by constructing a heteroglossic text, voice
is centrally related to the function of a linguistic form: it constitutes the local
micro-context which makes relevant interpretation possible. Although it is a no-
tion that also appears in SFG, voice is a concept that has been subject to the most
systematic treatment in stylistics and narratology, originating in linguistically-in-
spired analyses of literary works. Since issues involving voice and reported dis-
course significantly affect the coding of temporality in news text and, thus, have
a central role in my account of temporal deixis, let me go on to briefly outline the
way the notion of heteroglossia is applied in this work.
2.3 Heteroglossia
The present work operates with the premise that the value (meaning) of a linguis-
tic form is intimately based not only on the macro- and micro-level situational
context represented by the conventions of the relevant genre and the point-of-oc-
currence of a given formal feature, but also on the attribution of the form to a
particular voice. In media contexts, there is a fundamental difference if a particu-
lar lexical encoding or syntactic representation of an event is uttered through the
institutional voice or through some other external voice. The mere quantification
of occurrences of various formal features in media discourses is always potentially
risky because the disregard of the exact local context – including the identity of
the voice responsible for a given form – may obscure the fact that the same forms
are often used with contrary meanings. In media utterances, what matters is not
the denotation of a linguistic form but, rather, ‘who says what to whom with what
effect’. Such broader contextual considerations, including the attribution of utter-
ances to the relevant voices, need to be reflected in any socio-pragmatic account
of textual data.
The observation that all texts are ultimately heteroglossic derives from
Bakhtin’s reflections on the nature of language (Bakhtin 1981), which have been
extremely influential not only in literary studies but also among linguists. For
Bakhtin, the core of any linguistic utterance is dialogue. This is because dialogue
is inevitably oriented to other people while, at the same, it carries its discursive
history with it. In his philosophical view, “the event of existence has the nature
of dialogue … there is no word directed to no one” (Holquist 1990: 24). Thus,
even a monoglossic text has a dialogic nature. Heteroglossia in Bakhtin’s original
conception refers to the distinct varieties that are coexistent in a given text. It is,
thus, possible to trace the multiplicity of voices found in the author’s narrative.
In a similar way, Fowler (1991: 60) locates modes, registers and dialects in texts,
pointing out the texts’ frequent heterogeneity.
The Bakhtinian polyphony of voices finds its most basic linguistic counter-
part in the distinction between direct and indirect speech. The classic stylistic
framework proposed by Leech and Short (1981) elaborates the distinction by
identifying five different modes in which voices can be represented: direct speech,
indirect speech, free direct speech, narrative report of speech acts, and free in-
direct speech. Fowler (1996) develops the traditional narratological notions of
internal and external narration into a taxonomy of types of narration, as does
Simpson (1993) in his model of the point of view in texts, as well as others who
work in the area of cognitive stylistics (cf. McIntyre 2006).
In the area of media discourse, Fairclough (1995: 55) preserves the traditional
duality of what he calls Direct and Indirect Discourse. He pays attention, among
other, to the conversion of DD into ID, which occurs through subordination, shift
of pronouns, shift of deictics and backshift of tense. Fairclough further identifies
‘slipping’ as a special subtype of DD and a fourth category that he terms ‘unsig-
nalled mode’. While the former concerns the incorporation of the direct discourse
mode in a stretch of indirect speech, the latter is secondary discourse that “ap-
pears in primary discourse without being explicitly marked as represented dis-
course” (1995: 55) – a use that can be observed in some headlines (cf. Fairclough’s
example Mrs Thatcher will not stand for any backsliding).
The present account does not aim to develop a taxonomy of the diverse types
of heteroglossic discourse in news texts. Rather than seeing the ‘direct’ and ‘indi-
rect’ categories as a matter of discourse representation, I consider them as indi-
cations of the diversity of voices. Indirect speech/discourse thus opens the text to
external voices: the author relinquishes the responsibility for the linguistic encod-
ing of the content, yet preserves the gate-keeping power of structuring the entire
speech event and arranging the heteroglossic components. There are two basic
categories: the authorial voice (AV) and the contrasting external accessed voice
(EV). The basic duality is not meant to disguise the fact that multiple external
22 News and Time
voices are commonly present in news texts; what matters, however, is the shift in
the point of view – the defocalization from the authorial discourse to some other,
albeit mediated, encoder of the content. The consideration of the impersonal au-
thorial voice of the paper and the heteroglossic elements (attributed or unattrib-
uted to particular sources) is deemed crucial for explaining some of the deictic
shifts observable in news texts, particularly in headlines.
The unceasing interest of linguists in the notion of heteroglossia and its role
in news discourse can help us to better understand the complex ways in which the
media represent reality and pass judgement, yet manage to assign responsibility
to outside sources. In a recent study, for instance, White (2012) has engaged the
concept of heteroglossia in connection with Appraisal Theory in order to explain
how evaluation is expressed in the news through other voices than the reporter’s
voice. There is much promising research to be done in this area, e.g. about the
implicit construction of meaning that arises from the juxtaposition of multiple
accessed voices and the authorial voice, where the latter provides the seemingly
objective background against which the more openly subjective external voices
are deployed.
My pragmatically-oriented analytical framework considers the role of heter-
oglossia to be more fundamental than just being a matter of the internal struc-
turation of texts, i.e. a structuration that ultimately results in the diversification
of texts into a conglomerate of multiple voices. While that is certainly a true de-
scription of the situation, such a static or formal view is not fully adequate, since it
captures only one dimension of the concept of heteroglossia when there are really
two. Heteroglossia is also a text-forming strategy applied by a speaker with the
aim of pursuing some specific goal (e.g. the coding of evaluation as in Apprais-
al Theory). By pursuing that goal, the speaker deploys diverse linguistic forms
while simultaneously juxtaposing the authorial voice with various accessed voic-
es. Thus, the recipients of the text may be invited to draw relevant inferences both
about the propositional content thus communicated and about the motivations
that the author may have had for chosing some specific forms to communicate
the intended content.
Arguably, this more dynamic conception allows us to go beyond the mere ob-
servation, description and possibly taxonomic organization of the forms of heter-
oglossic utterances. Instead, we want to seek an explanation of the phenomenon
by pointing out how it squares with the possible aims and purposes of the speaker.
This constitutes one of the premises on which this book bases its interpretation of
the relationship between heteroglossia and temporal deixis in news texts: the for-
mer can be used strategically by journalists to affect a change in the conventional
coding of the latter. In simple terms, heteroglossia can be a tool for bypassing
some of the strict conventions of the genre, particularly in headlines.
2.4 News discourse analysis
There is one more area of research that the present study cannot ignore. This is
the long tradition of discourse analysis that has centred on news texts and that
has embraced various (socio)cognitive-structural and pragmatic-interactional
approaches. Much of that work is directly relevant for the purposes of this book,
in that it essentially reflects the textual and interpersonal preoccupations of the
present analysis. The (socio)cognitive-structural perspective derives, among oth-
ers, from the early studies by Teun van Dijk and Allan Bell, in which they laid
the ground for a systematic linguistic analysis of news discourse. Van Dijk’s so-
ciocognitive study (1988) and Bell’s sociolinguistic and ethnographic account of
media practices and texts (1991) are central for the textual dimension of the anal-
ysis presented here. Both authors deal with the structural organization of news
reports – and some of its linguistic implications – by considering the top-down
principle of relevance that is used to organize the presentation of topics. As a
result, the difference between the structure of the extralinguistic event and its
linguistic representation needs to be accounted for since the top-down model
of topic presentation results in non-chronology (Bell 1995, 1998; Ungerer 2000;
Jucker 2005) and a cyclic presentation of topics (van Dijk 1988; Iedema et al. 1994;
White 2003).
The pragmatic-interactional strand follows the idea that a news text is an arte-
fact that connects the participants in the speech event – it is a textual site of inter-
action between its producer and recipient. This means that the text’s producers, in
addition to attending to the content side of their textual creations, need to be aware
of the relationship between themselves, the texts and the audience. Ultimately,
they must take the recipients into account and relate to them – by fashioning the
style and the linguistic means available to them. In this connection, Fowler notes
that the media create “an illusion of conversational style” (1991: 65) and adds that
“[i]n newspaper discourse, deixis provides important cues to the oral mode” (64).
The outcome is interactiveness of form and interactivity of content that compen-
sates for the constraints blocking real interaction (either impossible or severely
limited and delayed in the printed media). Here, a pragma-linguistic analysis can
map the ways the producer-audience relations are discursively constructed and
how the participants are positioned. This has implications, for instance, for some
of the work done in critical discourse analysis (cf. Richardson 2007), because such
positioning can be associated with ideologies that are shared, implicitly expressed
or presupposed as ‘common sense’ (cf. Chovanec 2012).
The recipients, however, are far from being passive recipients of ‘prepackaged’
messages (cf. the ‘conduit’ metaphor of media communication): they are actively
involved in producing meaning as well. They make sense of the texts not only
24 News and Time
by processing and inferring information (ideational dimension) and by being

complicit discourse participants who are shaped and positioned by the author
(interpersonal dimension), but also by drawing on the textual resources that the
readers need “to bring to a text in order to make sense of it” (Talbot 2007: 46)
(textual dimension). Along the interpersonal axis, the notion of text as dialogue,
the dialogical positioning of discourse participants, and the internal heteroglossic
structuration of news texts thus cooperate in a highly complex way in order to
facilitate the transactional aspect of the communication.
These assumptions provide us with a starting point for the investigation of
temporal deixis in news texts. We will address the forms which temporal deix-
is assumes and the roles that it can serve. The analysis will start by considering
headlines which, while standing at the macro-level, make news texts so distinc-
tive. The analysis of temporal relations beyond the headlines will be limited to
tracing the transformations that the verbal phrase undergoes when the same
content is cyclically repeated in the individual textual segments of news stories.
While temporal deixis is considered as a crucial, though conventional element for
enhancing the interactivity of texts (cf. deictic projection and shifts of tenses), it is
the heteroglossic composition of news texts that helps us to understand not only
the temporal complexity of the texts as such but also the effect that accessed voice
has on such convention-bound texts as news headlines.
Chapter 3
Temporal deixis and news discourse
The pragmatic category of deixis significantly contributes towards the construc-

tion of the interpersonal dimension of discourse. In general, it draws on various
lexico-grammatical means in order to articulate a shared context between the
participants in a speech event, as well as to define, reinforce and modify their per-
sonal relations. Defined as the encoding of personal, spatial, temporal and social
contexts and the subjective experience of the encoder (cf. Green 2006: 415), deixis
constructs a cognitive focus of reference that centres on the relevant speaker(s).
The category of deixis has an important role in pragmatically-oriented investiga-
tions not only because of the flexibility with which deictic centres can be relocated
(shifted) but also because the consideration of deixis is crucial for the study of a
range of related phenomena, such as cohesion and coherence (e.g., the various
types of reference), the social relations between the participants (e.g., terms of
address, honorifics), as well as the cognitive construction of temporality (e.g., the
use of tenses and other temporal indicators).
While some of the lexico-grammatical features through which deixis finds its
way into texts are relatively micro-level phenomena, others are related not only
to the construction of texts (by means of providing for their textuality) but also
to more extensive textual macro-structures. News stories, for instance, describe
events in a non-chronological manner; such a macro-structural pattern of dis-
course organization is one of the typical characteristics of modern news discourse
(cf. Bell 1998; Ungerer 2000) and underlies the common journalistic format of the
‘inverted pyramid’, despite the fact that it is a relatively recent development origi-
nating at the beginning of the 20th-century (cf. Jucker 2005). Through the use of
verbal tenses and adverbials of time, temporal deixis has a significant role in the
analysis of the complexity of events reported and encoded in harmony with the
conventions of the ‘news story’ genre.
The present chapter deals with temporal deixis, focusing on how it is used as
an interpersonal element in newspaper discourse. The aim of the chapter is two-
fold. First, it provides a general theoretical background to deixis. Second, it postu-
lates a model of deictic centre projection applicable to news texts. While it serves
as a point of reference for the analysis of actual data in the remainder of the book,
the model is illustrated in this chapter with examples documenting (a) temporal
26 News and Time
deixis in headlines (focusing on adverbials, tenses, and auxiliaries) and (b) tem-
poral deixis beyond headlines (concentrating on tenses and correspondences be-
tween tenses and adverbials in headlines, leads, and the body copy).
The chapter outlines some of the issues that are eventually described in detail
in the individual chapters. Particular attention is, thus, paid to the projection of
temporal relations, manifested in the conventional shift of tenses in headlines,
which is subsequently documented with actual data in Chapters 5, 6 and 7. The
projection, which makes the verbal tenses correspond to the readers’ deictic cen-
tre built around the receiving time, is interpreted as enhancing the interpersonal
dimension of communication. Thus, while the verbal tense in headlines and leads
typically behaves interpersonally, the same grammatical category has a different
meaning in the other structural segments of news stories, particularly the body
copy, where tense is ideational and textual (i.e., deictic and cohesive) – this point
is then argued in Chapters 9 and 10. The same interpretation holds for adverbials
of time, which are discussed in Sections 3.5 to 3.7 and 5.3 to 5.5: while absent
from headlines altogether (on account of being more specific and thus ‘ideational’
than verbal tenses), some non-calendrical adverbials of time implicitly presup-
pose a shared temporal context in leads because of their ‘pre-emptive’ character.
3.1 Deixis and interaction
Deixis is a means of connecting language, the communicative situation, the ex-

tralinguistic reality (the context) and the speakers. In an early definition of the
term, Levinson (1983) highlights the connection between language and context
as follows: “Essentially, deixis concerns the ways in which languages encode or
grammaticalize features of the context of utterance or speech event, and thus also
concerns ways in which the interpretation of utterances depends on the analysis
of that context of utterance” (1983: 54). In a similar vein, Huddleston and Pullum
describe as deictic those expressions whose “reference is determined in relation
to certain features of the utterance-act: essentially, when and where it takes place,
who is speaking to whom, [and] the relative proximity of entities to the speaker”
(2002: 68).
Both of these traditional pragmatic and grammatical definitions, however, are
somewhat static since they do not explicitly take into account the fact that speak-
ers use deixis to construct reality with respect to other interlocutors, be they pres-
ent or absent, and real or fictional. Pragmatics, of course, considers human beings
to be at the core of the meaning-making process realized through language. In
pragmatic theory analysing the indexical and symbolic meanings of deictic terms,
this conception is also reflected in the more recent shift from their “meaning and
Chapter 3. Temporal deixis and news discourse 27
reference […] in possible contexts to consideration of the cognitive methods that

addressees employ in the interpretation of utterances” (Green 2006: 417).
A more sociolinguistic definition that points out – in very direct and simple
terms – the link between language and the interlocutors is provided by Fowler
(1991) in his classic study of news discourse from the perspective of early critical
linguistics: “Deixis in language … consists of the devices which link a text with
the time and place of communication and with the participants; which ‘orient’
speaker and addressee in relation to the content of the discourse” (1991: 63). In
this conception, language is regarded as an intermediary between the context of
utterance, the content of the discourse, and the participants. It is seen dynamically
as a process and interaction rather than a mere formal system of signs. While lan-
guages possess specific forms and structures to encode the features of the context
of utterance, it is the participants who use the deictic means available in order to
encode and decode often highly context-specific meanings.
Deixis is thus a category that has a strong underlying interpersonal basis.
When considering indexical expressions, Mey (2001: 55) notes how speakers can
anticipate and reorient themselves to the points of view of other people, by chang-
ing the perspective and encoding the reality from some alternative deictic centre.
Obviously, when the relationship between the participants and the subjective/
objective content of the discourse is not certain, the participants can consciously
negotiate the deictic reference and, for instance, actively redefine their points of
view. This is, arguably, also the case in the conventional utilization of temporal
deixis in news stories.
A pragmatic approach needs to explicitly acknowledge that deixis is an im-
portant linguistic device for constructing a meaningful discourse that can be
understood and interpreted by the participants. Deixis constitutes a pragmatic
phenomenon through which some specific aspect of the communicative situation
shared by the participants in a particular instance of communication is encoded
through language. It is essentially exophoric (as opposed to text-internal anaph-
ora, with which it is sometimes contrasted), and subjective because its operation
relies on the discursive organization of the discourse space into deictic centres
that typically coincide with the speakers who hold the floor at particular mo-
ments. Yet, deixis has an intersubjective dimension in that it allows for the op-
eration of deictic shifts and manipulations that take into consideration the other
interlocutors, e.g. by coding the message with respect to the anticipated time and
place of its reception. In this sense, deixis can be used as an interactive device.
As mentioned above, deixis in general – and temporal deixis in particular –
plays a role in the development of the interpersonal dimension of texts since it
constructs a shared context between the discourse participants – it postulates
their co-presence within a deictic field. In news texts, the fictitious co-presence
28 News and Time
is achieved in a number of ways. In the present account, we will pay attention to

those means that create the impression of a shared frame of temporal reference.
These means, which are understood to increase the interactiveness of news texts,
include the following representative examples:
(a) The social networking site resets passwords for 250,000 users following one
of its biggest ever cyber attacks. (Hackers target 250,000 Twitter
passwords; The Telegraph; 2.2.2013; lead)
(b) A planet with conditions that could support life orbits a twin neighbour of the
sun visible to the naked eye, scientists have revealed.
(Scientists find ‘habitable’ planet; The Independent; 19.12.2012; lead)
(c) Sebastian James said he agreed with the views of John Lewis boss Andy Street
who warned on Wednesday that rivals such as Amazon would be able to use
their tax position to “out-invest” and “out-trade” UK companies.
(Dixons Retail chief Sebastian James backs Amazon tax warnings;
The Telegraph; 16.11.2012; lead)
All three examples report past-time events that are located in the recent past, un-
derstood here as a relatively brief period of time preceding the publication date.
For ease of comparison, they are all taken from the same textual segment of news
texts, namely the lead, which typically provides the summarization of the news
story. The individual leads in the examples above use different tenses to refer to
the individual past events: the simple present, the present perfect, and the simple
past, respectively. At the same time, all three news items have headlines with the
verb in the present simple, referring to the relevant events located in the past
time. There is more to the three examples than merely the representation of some
newsworthy events; their encoding takes into account the recipients by relying
on a shared temporal context of reference, although each of the texts does so in a
somewhat different way.
In the sections to follow, temporal deixis is dealt with in connection with an
analysis of selected surface linguistic features in news texts by offering a com-
plex grammatical description supplemented with a pragmatic interpretation of
the relevant phenomena. The present tense in headlines is thus theorized within
its grammatical and semantic classification and explained pragmatically against
the routines of the news production process. The present perfect is identified as
another means of conventionally enhancing the current relevance of the main
event and constructing a temporal context shared with the readers. Finally, the
interplay between tense and adverbials of time is addressed as well, since even a
deictic use of tense (such as the past tense in example (c) above) can combine with
a non-calendrical adverbial of time – and such a combination inevitably relies for
its successful interpretation on the presupposition of a temporal context shared

with the readers.
3.2 Deictic centre
Central to our discussion of temporal deixis in news texts is the notion of the
deictic centre, which provides the focal point towards which verbal tenses and
time adverbials are related. The deictic centre is understood pragmatically as a
changeable construct that can be actively modified and constructed by interlocu-
tors in order to serve their communicative needs.
The idea of the deictic centre goes back to Bühler’s conception of the commu-
nicative event (1934), in which he identifies the centre of reference (‘origo’) as the
intersection of personal, spatial and temporal dimensions located in the speak-
er’s fixed point of ‘I-here-now’. He notes that the origo is not inescapably linked
to a certain place but can be transposed to someone else’s location. Since then,
the issues of deixis and deictic centres have attracted the attention of scholars
from various disciplines, ranging from pragmatics to narratology and cognitive
science. Literary studies, for instance, have traditionally distinguished between
various types of narration, with the basic alteration between first-person and
third-person narrators. Here, we must not only distinguish between the narra-
tor and the author but also consider the issue of focalization of the third-person
narrative through a particular character. On the level of linguistic form, this has
implications for distinguishing various types of narratives, such as free indirect
discourse and others (Fludernik 1993, 2009), and tracing the properties of speech
and thought representation (STR; Vandelanotte 2009).
Apart from narratology, the idea of the transposition of deictic centres (or
points of view, depending on one’s perspective), also features in cognitive science.
As a part of a larger research project into the structure of discourse and narra-
tive, Segal (1995), for instance, proposes a theoretical framework called the ‘deic-
tic shift theory’, where he approaches the issue from the perspective of cognitive
and computational science. Under this theory, readers locate themselves in the
‘story world’ – the mental model they create of the narrative. In doing so, “readers
and authors shift their deictic center from the real-world situation to an image
of themselves at a location within the story world” (1995: 15). The manipulation
of the deictic centre is thus related to an individual’s ability to creatively distance
themselves from their here-and-now: it is a special kind of cognitive displacement.
Sanders and Redeker (1993) consider the shift of deictic centres together with a
number of other phenomena related to the construction of perspective in nar-
rative texts through quotation and focalization. Looking at how these linguistic
30 News and Time
devices represent a subjective point of view in short news stories, their empirical
research indicates that while “focalization is judged as an atypical element in news
texts … [and is] primarily a dramatizing device” (82), implicit viewpoints that are
signalled by such linguistic forms as “tense shifts, focalizing verbs, marked refer-
ences, deictic viewpoint and other subtle linguistic devices” appear “more accept-
able as news texts than the neutralized version” (86).
Deictic terms and indexicals have been subject to extensive treatment in ref-
erential semantics and pragmatics (cf. an overview in Levinson 2004). Levinson
(1983: 64) refers to “unmarked anchorage points” that constitute the deictic cen-
tre. In addition to the person, time and place of the speaker, he mentions the
discourse centre (defined as “the point which the speaker is currently at in the
production of his utterance”) and the social centre (understood as the “the speak-
er’s social status and rank, to which the status or rank of addressees or referents is
relative”). It is generally understood that the deictic centre is “located within the
context of utterance by the speaker” (Brown and Yule 1983: 52) and it shifts in a
conversation as the individual interlocutors take turns.
A question might arise about the primacy of either speaker-centred or hear-
er-centred deixis. Jones (1995) objects to the egocentricity of such accounts,
arguing that communication is a social act and the centre of reference is jointly ar-
ticulated by the interlocutors. Reader/hearer-centred and writer/speaker-centred
deixis is, in his view, seen as a departure from the harmonious situation of social
interaction between individuals, though that departure is typically conventional
and related to particular activities. Basing his argument on Lyon’s observation of
deictic simultaneity (1977: 685), Jones suggests that such a deictic as now is not
subjectively orientated (or speaker/hearer-centred). Rather, it is sociocentric be-
cause the temporal zero-point is the same for the speaker and the hearer (1995: 38).
A somewhat different approach is adopted by Koyama (2009), who sees the
deictic centre as the fluid ‘here and now’ origo that constantly moves through dis-
course, being the centre of discursive interaction. He argues that the point of view
of referential (propositional) pragmatics should be supplemented by more focus
on “non-referential, social-indexical pragmatics, concerning not ‘what is said’
[…] but ‘what is done’, i.e., the communicative practices that ‘index’ the socio-
cultural power-relations and identities of the communicative-event participants”
(Koyama 2009: 80). In this dynamic approach, the shifting origo, here understood
as the ‘here and now’ of discourse, also affects such phenomena as information
structure, where ‘new’ (rhematic) and ‘old’ (thematic) information are indexical
notions, relative to the phase of discourse at which the information is introduced.
The shifting deictic centre then stands at the core of all socio-cultural phenom-
ena: “the sociocultural universe, in its entirety, is indexically anchored onto the
ever-shifting deictic center of discursive interaction” (2009: 90).
From a cognitive pragmatic perspective, Marmaridou (2000) defines a deictic

expression as “one that builds up a mental space in which the speaker and the ad-
dressee are co-present at a given moment in time” (2000: 100). Temporal deixis is
seen – in line with personal and spatial deixis – as egocentric, i.e. referring to the
participant roles in the speech event (2000: 82). Deixis constitutes an act whereby
the speaker draws the addressee’s attention “to an entity in terms of its (spatial)
relation to the agent” (2000: 100) and uses certain grammatical constructions
for that purpose. Importantly, “[t]he use of these constructions by a particular
speaker automatically authorizes her as the deictic centre, the source of the act.
In other words, deictic constructions construct the speaker as the deictic centre”
(2000: 100).
In recent years, the idea of deictically-anchored mental spaces has inspired
some interesting research in cognitively-oriented critical discourse analysis. The
practice of intentionally shifting the deictic centre closer to the addressee, for
instance, has been documented from the domain of political discourse as a strat-
egy of enhancing the perlocutionary effect of one’s utterances, e.g. by boosting
the perception of an (alleged) threat. Chilton’s model of political communica-
tion (2004) localizes some entities within a deictic centre, while certain other,
alien entities are perceived as being outside that centre and are ascribed a negative
agency role with respect to the central entities. The deictic centre is understood as
“the implied ‘anchoring’ point that utterers and interpreters construct or impose
during verbal interaction” and that “depends on cognitive frames that embody
conventional shared understandings about the structure of society, groups and
relations with other societies” (Chilton 2004: 56). The model articulates the ex-
istence and possible projection of deictically centred ‘discourse spaces’ that are
defined through the positioning of abstract vectors (Chilton 2005).
In this approach, the deictic centre is understood to be located at the inter-
section of spatial, temporal and modal/axiological axes, with cognitive shifts
achieved through so-called proximization (Cap 2008). The concept of proxi-
mization itself is defined as “a pragmatic-cognitive strategy that relies upon the
speaker’s ability to present events on the discourse stage as directly affecting the
addressee, usually in a negative or threatening way” (Cap 2010: 119). For instance,
pre-war and pro-war rhetoric will typically construe the discourse stage in proxi-
mal spatial-temporal terms, i.e. emphasizing that the threat is present and current
(in the ‘here and now’). The effect of such persuasive discourse on the recipients is
higher if the threat is discursively constructed as being close and real rather than
distant and merely hypothetical. Such discourse space conceptualizations may be
traced through specific lexical and grammatical forms that represent the surface
manifestations of the proximization strategy.
32 News and Time
If we adopt a general pragmatic perspective on the elements that are related

to deictic centres of interlocutors in the case of face-to-face speech events, it is
an undisputed fact that the temporal and spatial settings of the individual in-
terlocutors are identical. This is so despite the inevitable shifts of deictic centres
between the interlocutors when they take turns as speakers, thereby affecting the
use of personal pronouns. Obviously, the overlap of the spatio-temporal settings
of the interlocutors is the result of the speaker’s and the hearer’s physical co-pres-
ence and the fact that their contexts of situation (time and location) coincide. In
this sense, personal deixis is more dynamic because any time speakers change
turns, there is a switch in deictic centres; spatial deixis is less dynamic though still
changeable, e.g. reflecting the physical locations of the speakers (cf. the encoding
of movement and directions in the case of such verbs as come and go); and tem-
poral deixis is comparatively the most static since all interlocutors have the same
temporal co-presence.
However, many discourses take place in displaced communicative situations –
in split contexts of utterance (Fowler 1996: 112; Crystal and Davy 1969: 119). This
occurs when the interlocutors communicate despite not sharing the same physi-
cal environment: they are either located in different places (as when communicat-
ing by phone or synchronous online chat), or they communicate across time (cf.
the classic example of a note pinned on the door and saying “Out of office today”).
In most cases of written communication, the interlocutors are separated both in
space and time – this situation also holds for computer-mediated communica-
tion of the asynchronous type. In mass media contexts, the split spatio-temporal
context is the rule in both written media and spoken broadcasts, as long as they
are pre-recorded and not aired live.2 Needless to say, modern online journalism is
also finding ways to overcome the temporal displacement traditionally inherent
in the production and consumption of news texts. One such example consists of
the innovative patterns of live text reporting, i.e. the production of online texts in
real time, almost contemporaneously with the unfolding events (Chovanec 2010).
The traditional contrast between the spoken and the written modes (and, by
extension, the distinct spoken and written genres) reflects deictic situations that
are based on either the existence of the physical co-presence of the interlocutors
or the lack thereof. In spoken communication, the temporal and spatial context
is shared either fully or partially. Thus, in face-to-face conversation, the speakers
share all of the deictic variables, while in such speech situations as telephone con-
versations, radio phone-in programmes and video conferences, the interlocutors
2. The process of pre-recording such programmes typically takes into account the anticipated
time when the programmes are scheduled to be broadcast. The encoding then operates with a
projection of deictic centres to the future moment of the texts’ reception.
share the temporal framework only. In the latter case, the spatial context may – if
needed – be expressly negotiated and specified by the speakers.
By contrast, written communication traditionally operates in a split context of
situation and is characterized by both temporal and spatial displacement. Thanks
to advances in modern technology, the temporal displacement can be partially
overcome in such genres as synchronous online communication (electronic chat),
Internet email conferences, and live online text commentary. In these situations,
the participants in the speech event share the context of time in spite of the use
of the written medium.3 Interestingly, there are also communicative situations
defined by the users’ presence in a shared virtual space, despite their possible
temporal and physical dislocation. Hence, shared space can be construed as a
non-physical entity, defined as the textual environment in which some commu-
nicative interaction occurs, despite the existence of a split context between the
interlocutors.
However, the permanence of the written medium also allows for reception to
be non-simultaneous and to occur beyond such a temporal frame of co-presence.
Thanks to the preservability of written texts, the textual record can be processed
(as well as reacted to) asynchronously, i.e. with a time delay after the texts are pro-
duced and perhaps received by the majority of recipients, as may be the preferred
way of consuming the relevant texts.
Thus, the temporal and spatial frames of reference involved when communi-
cating in the written mode need not be shared by the writer and the reader. The
interlocutors, however, will typically find some ways of anchoring their deictic
situations because temporal and spatial deixis needs to be fixed in an unequivocal
manner in order to prevent ambiguities and to enable unproblematic communi-
cation in such displaced settings.
Before considering the issue in greater detail, let us state that not all time
expressions are deictic. For a time expression to be considered deictic, some tem-
poral reference point needs to be established. This is most commonly the moment
of speaking, i.e. the coding time (cf. Fillmore 1997: 53; Marmaridou 2000: 100).
For instance, a deictic time adverbial such as yesterday becomes meaningful to the
recipient only when it is properly anchored to a specific temporal reference point.
3. ‘Sharing the context of time’ may also be understood not as the exact temporal overlap be-
tween the moments of production and reception but, rather, the co-presence of the participants
in the course of the text’s production in a more durative sense. People sending text messages
and making other brief electronic communications can be considered as sharing the same tem-
poral context as long as they process the messages at the time of receiving them – despite the
fact that the typing of the text message is carried out in private. Strictly speaking, temporal
co-presence is realized in the moment of clicking and sending the message.
34 News and Time
It locates the past-time moment referred to only in relation to some other mo-
ment – the speaker’s deictic centre, i.e. the moment of utterance (cf. Huddleston
and Pullum 2002: 1452). Thus, the deictic adverbial yesterday can function only
as long as the speaker and the recipient operate within a shared temporal context.
3.3 Deictic projection
Since the deictic centre is not some unique and fixed entity, speakers can shift
it to various possible points along the time line. They can then formulate their
utterances from a point of view anchored to such a point that is detached from
the speakers’ own here-and-now. The most common situation occurs when the
speaker negotiates their point of view between two deictic foci for deictic refer-
ence. One deictic centre exists on the part of the speaker/writer and another on
the part of the addressee.
If we consider the more specific issue of temporal deixis, two distinct times
are typically distinguished that correlate to the two contrasting participant roles
of the speaker/writer and the addressee: the coding time (CT) and the receiving
time (RT). While the former concerns “the moment of utterance (or inscription)”
(Levinson 1983: 73), the latter is the moment the encoded message is processed by
the recipient. In ordinary, face-to-face communication, CT and RT are identical
because the speaker and the hearer are co-present in space and time. As Levinson
(1983: 73) observes, “… in the canonical situation of utterance, with the assump-
tion of the unmarked deictic centre, RT can be assumed to be identical to CT” – a
situation referred to by Lyons (1977: 685) as deictic simultaneity. The coding time
CT and the receiving time RT are thus co-temporal (Jaszczolt 2002: 195).
Some accounts of deixis tend to operate with a three-layer system of time
relations: in addition to coding time and receiving time, the level of content time
is also distinguished (Green 1995: 22). Here, content time relates to the semantic
content of the proposition. In that sense, it is ‘internal’ to the proposition, as op-
posed to the coding time and the receiving time, which are ‘external’ to it.
Grammatical accounts of tense and time operate with somewhat different
categories, e.g. orientation time and situation time, which are temporally related
to speech time by means of tense forms (cf. Declerck 2006: 95). Huddleston and
Pullum (2002: 125–126) distinguish between four temporal categories:
Tsit – time of situation
Tr – time referred to (identified with Tsit when aspectuality is perfective)
Td – deictic time (normally the time of utterance)
To – time of orientation (identified as Td in the default case)
The mutual relationship between time referred to (Tr) and the time of orientation
(the time of utterance, To) divides the time line into three time zones in the fol-
lowing way:
Past time: Tr anterior to To Tr < To
Present time: Tr simultaneous with To Tr = To
Future time: Tr posterior to To Tr > To
Without elaborating on the issue of the deictic centre, Huddleston and Pullum
(2002: 1453) note how it can be switched in conversation from one speaker to
another as well as to the absent addressees or some other persons. They give the
example of a notice in a public toilet (Now wash your hands.) and note how “now
refers not to the time of writing the notice but to the time of reading it – and
hence is constantly shifting with the passage of time.”
Thus, an account of temporal deixis in displaced communication contexts
needs to operate with the two fundamental frames of reference, clearly distin-
guishing between the coding time and the receiving time. There are many sit-
uations when the CT and the RT are not identical (letter writing, pre-recorded
media broadcasts, etc.). Because of the split context, speakers need to provide for
a different anchorage of their temporal, personal and spatial deixis, since deictic
simultaneity does not obtain automatically.
In split contexts, there are essentially two possibilities: the deictic centre can
either:
1. correspond to the encoder (speaker/writer); or

2. be projected onto the receiver in order to coincide with the RT.
These possibilities are present in the following two sentences (taken from Levinson
1983: 74), which reflect the two different points of view embodied in the choice of
verbal tense: I write this letter while chewing peyote. vs. I wrote this letter while chew-
ing peyote. In the former sentence, temporal deixis is related to the point of view
of the writer and the coding time CT, while in the latter it is related to the point of
view of the addressee and the receiving time RT.
For Huddleston and Pullum (2002: 126), the concept of coding time CT is
identified as ‘time of orientation To’, i.e. the time of utterance. With the prima-
ry tense (the preterite and the present), “To is normally the time of speaking or
writing”. The term deictic time is used “to allow for the fact that in special circum-
stances it can be the time of decoding rather than that of encoding”. To illustrate
the change of perspective, they give the following two examples:
I am writing this letter while the boys are at school
[To is time of encoding]
36 News and Time
You are now leaving West Berlin. [a written notice]

[To is time of decoding]
Noting that the non-projected time of orientation is the default interpretation in

written texts, Huddleston and Pullum specify that in both utterances the tense has
a deictic function:
In ordinary speech the time of encoding and the time of decoding are identical,
but in writing they can be different. Where this is so, the default identification of
To, as in [the first example], is with the time of decoding, the addressee’s time.
The difference between these is not marked linguistically in any way and the term
deictic time covers both cases: it is defined by the linguistic event itself.
(Huddleston and Pullum 2002: 125)
What these examples illustrate is the well-documented semantic-pragmatic no-

tion of ‘deictic projection’ (Lyons 1977), which is also known, in a more general
sense, as a ‘shift in points of view’ (cf. Fillmore 1975; Fillmore 1997: 122). We
clearly see that “[s]peakers, or writers, do have the option of transferring the deic-
tic centre to the hearer’s, or reader’s, spatio-temporal situation in which the text
will be encountered” (Brown and Yule 1983: 53).
It is up to the speaker to decide whether or not to project the deictic centre in
anticipation of the context in which the recipient is most likely to encounter and
consume the text. The projection – realized by means of modifying the temporal
indicators to code deictic simultaneity/non-simultaneity – has the effect of objec-
tifying the relevant utterance for the addressee. This is because the utterance then
indicates more clearly the actual relationship between the coding time CT and
the receiving time RT (i.e. non-simultaneity: cf. I wrote this letter while…). On
the other hand, where the speaker/writer in a displaced communication context
decides to retain the deictic centre and not to verbalize the message with respect
to the receiving time, this may have the result that the distinction between the
CT and the RT becomes blurred for the addressee. The deictic centre will remain
non-projected, rendering a more subjective effect with the message deictically
anchored to its producer. In this sense, a text with non-projected deixis may re-
quire an increased effort on the part of the recipient since the reader may need to
work out the relationship between the CT and the RT. The reader may thus need
to interact relatively more with the text to account for the two unbridged deictic
centres. If, however, the writer chooses to utilize the former type, i.e. to project
his/her deictic centre onto the receiving time RT, this conscious act will position
the text more as a product produced for the benefit of the addressee – in other
words, the text will be ‘pre-processed’ for the reader as an account of something
that has already been done. In the opposite case, the text will have a more univer-
sal application and will appear topical each instance it is accessed. This contrast
of effects upon the addressee is apparent, for instance, from the following two
verbalizations of the same event:
I wrote this letter to tell you that…
(CT projected to RT; ‘objective’)
I am writing this letter (now) to tell you that…
(CT not projected; ‘subjective’)
There are two components involved in any temporal situation: the event and the
embedding of the event in space and time, relative to the time of utterance (cf.
von Stutterheim et al. 2003: 98). We thus need to enrich the grammatical account
by distinguishing more explicitly between the encoding of a verbal account of an
event and the occurrence of the event itself (as has been traditionally the case in
narratology, cf. Fludernik 1993).
3.4 Deictic projection in news texts
The need to specify the event time (ET) becomes evident when we consider a
similar kind of projection that occurs in news discourse and that was hinted at
in the brief discussion of the examples of news stories presented above. In news
articles, we typically encounter the present tense in headlines that conventional-
ly refers to newsworthy events that happened in the past. The headline present
tense (e.g. Scientists find ‘habitable’ planet) conventionally indicates deictic
simultaneity between the centres that revolve around the event time (ET) and the
receiving time RT. The coding time CT is not explicitly stated, but it is implicitly
present as the intervening frame in which the verbalization of the event is pro-
duced. The temporal projection in the verb makes the utterance more ‘subjective’:
the event time is not projected into the deictic centre of the receiving time for the
benefit of the reader.
By contrast, where the past tense and adverbials of time are used in the leads
or the body copy of news articles (e.g. A court ruled on Thursday that Blær … can
be legally used as the girl’s first name), they are indicative of the placement of the
event into the readers’ anticipated receiving time RT, thus rendering the sentence
as more ‘objective’ on account of placing the past event into the past zone. This
type of projection needs to operate with additional deictic centres and is further
detailed below. The shift of tenses that conventionally occurs at the transition
from the headline to the lead (and eventually the body copy) lies at the core of our
understanding of deictic projection, indicating that the most important temporal
indicators in news stories have a strong interpersonal orientation.
38 News and Time
It should be pointed out that the expressions ‘objective’ and ‘subjective’, when
used to refer to the presentation of an event with respect to the relevant time zone,
are relative. They relate to the perception of an event from the point of view of
the recipient of an utterance as opposed to that of the encoder. In other contexts,
the projection of deictic centres can serve quite contrary purposes. In direct and
indirect reported speech, for instance, a direct speech quote will typically access
a ‘subjective’ time frame and a ‘subjective’ (i.e. non-projected) CT. Yet, it may
be used as a clearly objectifying device by virtue of its construction (or rather,
re-construction) of direct, unmediated, first-hand discourse. Conversely, indirect
reported speech, which may have an ‘objectified’ time frame and an ‘objective’ (i.e.
projected) CT, may appear as subjective by providing a secondary, mediated ac-
cess to the original voice – which then becomes re-interpreted via someone else.
The re-interpretation of deictic centres is also apparent in the system of English
grammar, where tenses and deictic adverbials of time and place are convention-
ally backshifted in indirect reported speech. Indeed, this is one of the signs of the
conversion of direct into indirect discourse (cf. Fairclough 1995: 55).
There are three ways in which the coding time CT and the receiving time RT
can be related. The first concerns deictic simultaneity between the speaker and the
recipient, when the two times are identical. In displaced communication, howev-
er, where the two deictic centres are distinct, there are two possibilities. First, no
projection of the speaker’s coding time CT is implemented, with the result that
the encoding and decoding situations are construed as independent. The second
possibility is the situation of projection, when the coding time CT is shifted to the
future moment of the anticipated reception of the text, resulting in purposefully
constructed deictic simultaneity (cf. Figure 3.1).
The reasons for projection may include the author’s need to construct a shared
temporal context in order to erase the discursive trace of the coding time CT,
which effectively stands as a barrier between the event itself and the recipients of
its encoded verbalization. What may likewise be involved is the author’s intention
to establish a degree of complicity between the participants, as is commonly the
case in literary texts, where “the narrator pretends that there is a common point of
Simultaneous communication Deictic centre overlap CT = RT

(genuine deictic simultaneity)
Displaced communication a) No projection CT ≠ RT

b) Projection CT → RT
(constructed deictic simultaneity)
Figure 3.1 The relationship between the coding time (CT) and the receiving time (RT)
time, a ‘present’, which he and the reader share” (Fowler 1996: 79). Obviously, this
is a strategy for increasing the appeal of texts, particularly those in displaced and
depersonalized situations. In printed media discourse, the issue of deictic centre
projection finds its manifestation in the specific usage of time adverbials and the
conventional manipulation of tenses.
In news texts, temporal deixis is modified in two distinct ways. First, the
projection of deictic centres affects the use of specific adverbials of time that are
related to the discourse space centring on the receiving time RT. This is, in the
case of English, accompanied by the second phenomenon, namely the existence
of headline conventions that provide for the rather specific uses of some tenses,
mainly the present simple and the present perfect. These phenomena are treated
in Section 5.7 and Chapters 6 and 7.
Temporal shifts in news texts are closely related to the fact that, in order for
temporal deixis to be fully operative, the anchorage of deictic adverbials and tenses
to specific points in time is necessary. Such anchorage enables the adequate inter-
pretation of time deictic expressions and tenses that are encountered in news texts.
The encoders of written texts in split contexts then need to indicate a common
frame of temporal reference.
There are two ways in which such a frame can be established:
1. by means of an explicit mention of a definite calendrical point that serves to

anchor the time deictics to the temporal axis; and
2. by means of implicit anchorage that is realized through the pragmatic as-
sumption that the encoder and the decoder share the same temporal frames
of reference (while no such assumption exists in the former case).
Clearly, whenever the verbal exchange occurs in real time, as in face-to-face and
synchronous spoken (as well as written) communication, there is no need to
establish such a temporal frame, whether explicitly or by implicit assumption,
because the common frame of temporal reference obtains naturally as a default
situation.
3.5 Time adverbials and shared temporal context
The following examples illustrate some of the typical uses of adverbials of time
that can be found at the beginnings of news articles. While it is rare for headlines
to specify when the reported event happened, it is almost a rule to include this
information in either the lead (the first paragraph) or the next paragraph of the
body copy. This placement is motivated by the top-down structure of news sto-
ries (the ‘inverted pyramid’), and the requirement that the lead should contain
40 News and Time
all the basic information about a given piece of news. This concerns information
which can be expressed by means of the classic 5 W’s (‘who’, ‘what’, ‘where’, ‘when’,
and ‘why’; although ‘how’ is sometimes included as well, the manner in which
a newsworthy event happened is often subject to elaboration in the body copy
rather than the lead since the significance of this explanatory aspect of the event
is typically overshadowed by the occurrence of the main event itself).
(a) A train crashed into a parade float carrying veterans in the US state of Texas
on Thursday, killing four people and injuring 17 others.
(Four dead as train slams into parade for wounded veterans;
The Telegraph; 16.11.2012; subhead/lead)
(b) A record of 57 new legal highs have been detected so far this year, with the
EU’s early warning system reporting the appearance of more than one new
psychoactive drug on the market every week.
(Huge rise in legal highs alarms drug experts at European
monitoring centre; The Guardian; 16.11.2012; lead)
(c) Two women accused of profiting from insider information supplied by a
two-timing German investment banker they both believed to be their boy-
friend were cleared by a jury yesterday.
(Two-timing banker’s girlfriends cleared of insider trading;
The Times; 16.11.2012)
(d) Sally Bercow was taking legal advice tonight as Lord McAlpine prepared to
issue a raft of defamation claims over internet paedophile smears.
(McAlpine prepares to sue wife of Speaker Bercow;
The Times; 16.11.2012; lead)
(e) Another major BBC figure of the last generation, Dave Lee Travis, was
arrested today in connection with alleged sexual offences by detectives
investigating a welter of claims following the exposure of Jimmy Savile as a
predatory paedophile. (Dave Lee Travis is latest held over
sex abuse allegations in Savile inquiry; Independent; 16.11.2012; lead)
The successful interpretation of all the underlined expressions is possible thanks

to the assumption of a shared context of temporal reference. The adverbial on
Thursday is interpretable within the context of a given week, while the adverbial
this year obviously constructs a much longer shared context. The last three ex-
amples – yesterday, tonight and today – are the most specific in that the shared
temporal context is construed to be a single day. Let us add that the examples
are taken from the online versions of newspapers, where news reports are added
continually. Hence, such adverbials as tonight and today can be meaningfully used
to refer to events that have already happened, i.e. to events grounded in the past
time. In printed versions of newspapers, there is a time lag – due to the process
of newspaper production – between the occurrence of an event and its verbalized
mediation to the readers, which leads to the inevitable reformulation of such ad-
verbials into different forms that go beyond the shared time frame of a single day,
most characteristically the time adverbial yesterday. (The operation of the shift is
detailed below in the section on deictic centre projection.)
3.6 Deictic and non-deictic time expressions
Before elaborating on how certain time expressions in news texts help to con-
struct the impression of a shared temporal context, an overview of some of the
differences between deictic and non-deictic time expressions would be helpful.
This is needed in order to lend support to the argument that the encoding of time
by means of deictic expressions entails the encoder’s choice to sacrifice referential
precision in favour of other considerations.
In general, there are two kinds of temporal indicators of events. Temporal
expressions either relate to the speaker’s time (or the hearer’s time in the case of
deictic projections) or identify the time absolutely in terms of calendrical time
(cf. Fillmore 1975). The contrast between the two types can be illustrated with the
following two examples:
A train crashed yesterday, killing four people.
A train crashed on 15 November 2012, killing four people.
While the former utterance encodes the time of the event as relative to the speak-
er’s time, the latter expresses absolute calendrical time that does not presuppose
or construct any link with the coding time CT or the receiving time RT. The ad-
verbial phrase 15 November 2012 “is interpreted calendrically, in terms of a point
in time which is specified on an absolute time measurement scale” (Marmaridou
2000: 82). In a similar fashion, Fillmore (1997) distinguishes between non-cal-
endric and calendric measures of time. They are described, respectively, as “time
measure periods taken only as units of measure, and those that have “fixed start-
ing points in absolute time” (Fillmore 1997: 49). Some expressions of time can be
used in both ways: e.g. year is both a non-calendrical unit of measure, equivalent
to any period of 365 days, and a specific calendrical unit describing the period
from 1 January to 31 December of a given year. Both can be used deictically,
however, relating to the moment of utterance (cf. the contrast between a unit of
measure and a calendrical unit in the phrases ‘over the past year’ vs. ‘this year’).
Calendrical expressions are either positional (i.e. indicating a position in a
sequence, such as morning, Tuesday, July) or non-positional (e.g. week, month,
42 News and Time
year). Time deixis is encoded through tense and deictic adverbials of time (Huang
2007: 145); the calendrical expressions are, strictly speaking, non-deictic because
they identify the exact time of an event without relating it to the moment of
utterance.
In everyday communication, there is a general preference to use adverbials of
time that are relative to the moment of utterance. By contrast, calendrical means
that locate the event exactly in time are dispreferred. Deictic adverbials of time
are said to be pre-emptive – they take priority over the calendrical means (Huang
2007: 145). In other words, it is more natural for speakers to say ‘The crash hap-
pened yesterday’ than ‘The crash happened on Tuesday’ if it is Wednesday today.
However, as O’Keeffe et al. (2011: 52) note, what is decisive is the use of the adver-
bials rather than their inherent formal classification. A time adverbial such as on
Tuesday is formally non-deictic, but – when used in combination with a deictic
expression such as ‘this’ (this Tuesday) – the whole combination becomes deictic.
And it may be argued that even the formally non-deictic on Tuesday has a deictic
interpretation as long as the interlocutors share a temporal frame of reference and
the actual relationship of the adverbial to the moment of utterance can be easily
inferred. Deixis is thus not a property of certain forms, but a function of the usage
of linguistic forms in context.
Huddleston and Pullum observe that location in space and time can be either
deictic or non-deictic. Thus, in the sentence He went to Spain last week, the time
adverbial is deictic because it is “identified in relation to the utterance-act”, while
in He went to Spain the day after the exam, the time adverbial is non-deictic be-
cause “it identifies the day in relation to the time of the exam, not to the time of
speaking” (2002: 1452).
Similarly, Declerck (2006: 591–592) distinguishes between deictic (anchored)
and nondeictic (unanchored) time-specifying adverbials. Deictic adverbials are of
two types: absolute deictics (e.g. yesterday, this morning, tomorrow, tonight, three
weeks ago) and relative deictics (e.g. the day before, the same day, that morning,
the next day, two days earlier). The relational nature of deictic expressions (both
absolute and relative) is crucial because they presuppose a context shared with
the recipient. Their appearance in the discourse indicates that the author chose to
encode some aspect of the situation as shared (temporal) common ground:
Deictic time-specifying adverbials like yesterday or in those days are ‘definite’ re-
ferring expressions, i.e. the speaker assumes (or pretends) that the referent time
is identifiable to the hearer, or that the hearer is at least familiar with the referring
expression. […] In other words, the use of such a deictic time-specifying adverbi-
al is assumed to be sufficient for the hearer to know or infer the temporal location
of both the ‘anchor’ and the Adv-time that is anchored to it. (Declerck 2006: 592)
Evidently, linking the time adverbial to the coding time CT or the receiving time
RT in this way is more interactive than the use of ‘scientifically precise’ and ob-
jectifying calendrical expressions. While the presence of a deictic time-specifying
adverbial in an utterance is related to the coding time CT, in texts that are intend-
ed for consumption at a later time, such adverbials inevitably take into account
the presence of the reader, presupposing the reader’s perspective from the point
of view of the receiving time RT. In that sense, they attend to the interpersonal
dimension of the text.
3.7 Time expressions in news texts
Since participants in split temporal contexts do not share their deictic centres, any
deictic, non-calendrical time adverbial is interpretable only as long as the recip-
ient of the utterance can work out the correct relationship between the receiving
time RT and the coding time CT. In written news discourse, this is possible as
long as a given reader of a daily newspaper reads the articles and contextualizes
the information contained in them on the basis of the assumption that his/her
frame of temporal reference (i.e., the receiving time RT) coincides with the pub-
lication date of the newspaper. This is the preferred mode of consumption of this
type of text – it is something that the media take into account when formulating
their stories. At the same time, periodical publications such as daily newspapers
may be read even beyond their anticipated shelf life. When that happens and the
same daily newspaper is read several days, weeks or even years later, the basic
underlying assumption will not be in operation. Nevertheless, as long as the read-
er knows he/she is reading a past issue, other time relational processes will take
place with the effect of ensuring that any linguistic forms expressing temporal
deixis are processed successfully (see below).
Of course, the normal practice is that readers of periodical press read news
texts assuming that the texts are recent. Where that assumption turns out to be
false and a reader finds out halfway that the newspaper he/she is reading was
published several days before, the text may cease to be relevant and the reader
may turn away from it with the unpleasant realization that he/she is reading stale
news. An entirely different situation occurs when readers knowingly read dated
periodical publications: in such cases, they typically have other motivations for
doing so than the intention to obtain current and relevant information. This rea-
soning applies to, for instance, patients reading old magazines in doctors’ waiting
rooms, scholars researching issues of historical interest in old newspapers, etc. By
doing so, however, they violate the default assumption of a basic shared frame of
temporal deixis that spans the moment of production as well the processing of
44 News and Time
such texts. Commenting from the perspective of relevance theory, Dor (2003: 719)
observes that when readers read old newspapers merely to kill time, their cogni-
tive style of reading is different from the traditional reading practice that is based
on the assumption of co-temporality. In such situations, readers knowingly read
news texts that bring them small or – in rare cases – even zero relevance, i.e. when
the contextual effects are minimal despite the high processing effort involved in
reading the texts.
Thus, the assumption of recency has two dimensions: it needs to be present
on the part of both the text’s recipients and the text’s producers. In order to be able
to construct the shared context in which the reference of temporal indicators can
be anchored deictically, the paper’s publishers have to rely on the readers to read a
current issue of the paper and not any other, older copies. This assumption is ena-
bled by the mutually shared implicit understanding that a daily newspaper is pub-
lished every day and that each new copy renders the preceding one out-dated. The
temporal deixis of a copy of a newspaper from the previous day becomes dated
in exactly the same way. This situation arises from the news cycle which printed
newspapers undergo. Jucker (2005: 17) represents this situation as follows: “The
printed newspaper report is clearly anchored in a twenty-four-hour cycle of pub-
lication. It looks back on yesterday’s stage and forward to today’s stage.”
This seemingly trivial operation of the periodical press – where today’s hot
news turns into history tomorrow and becomes forever relegated into the news-
papers’ archives with the next edition of the paper – affects the operation of tem-
poral deixis. Interestingly, the implicit assumption of recency necessarily differs
in the case of papers published with different periodicities. Weeklies and month-
lies, for example, will need to encode temporal deixis in a somewhat different
way. The assumption of recency spans the period since the appearance of the last
issue; and where there is a gap in the publication frequency of a daily newspaper,
for example on account of a national holiday or the weekend, the assumption of a
topical frame of temporal deixis is extended to cover an additional day. A Monday
issue of a newspaper will cover events happening over the weekend, i.e. a period
of two days, all being equally newsworthy since they have not been reported in
any previous issue of the paper.
The assumption of recency, which impacts the recipient’s interpretation of
temporal deictics, affects the encoder’s decision to project temporal indicators
onto the receiving time. As noted by Green (1996: 21), “[D]epending again on
the intentions of the speaker, if there is a significant interval between the time of
utterance and the time of (assumed) receipt of the utterance, the time reference of
a tensed sentence may be partially a function of the anticipated time of receipt”.
More explicitly, it is argued that adverbials – although they are related to the
moment of utterance – “could, under the circumstances described…, be indexed

to the assumed time of reception” (Green 1996: 21).
The indexing of a news text to the anticipated time of its reception allows for
the use of non-calendrical adverbials of time. By extension, depending on the
context and the publication frequency of a given press medium, the assumption of
recency affects even the category of verbal tense. The past tense, commonly used
in news stories, has an indefinite reference and normally needs to be anchored
to a particular moment in time by means of an adverbial of time. In some hard
news stories in daily newspapers, however, this is not necessarily so because of the
implicit assumption of recency of the given news event, which is understood to
have happened within the period from the preceding issue of the newspaper, i.e.
the time zone that can be identified as ‘yesterday’.
Let us turn to a sample text to illustrate the situation when no time adverbial
referring to the main event is present in the news story. Although adverbials of
time may locate certain aspects of the news story in time, here the temporality of
the event itself is based solely on verbal tenses. This is evident from the following
example of the beginning of a news report in an online newspaper:
Squeezed motorists being ripped-off, the AA warns
Hard-pressed motorists are being ripped off by petrol retailers as the falling
cost of fuel on the wholesale markets is not being reflected on forecourts, the
AA has warned.
[photograph, attribution, and link to comments omitted]
The motoring group said that the wholesale price of petrol has fallen by
around 9 pence a litre over the past month. However, average prices on UK
forecourts have fallen by less than 4 pence a litre.
Edmund King, the AA’s president, said it “beggars belief ” that motorists
are paying over the odds. He called on George Osborne to take action to
help struggling drivers. The Chancellor is under increasing pressure to drop
January’s planned increase in fuel duty, which will add a further 3 pence to a
litre of petrol.
The AA said that across Europe, the wholesale price of petrol – which is the
price that retailers pay for fuel – fell from around 54 pence a litre at the begin-
ning of October to around 45 pence a litre by the end.
With VAT, this should have knocked average UK petrol prices down by 10p
to 11p a litre, the AA said.
However, the group’s analysis found that petrol pump prices fell by less than
4p a litre between mid-October and the start of this week.
46 News and Time
In the absence of any explicit temporal anchoring of the main event reported (‘a
warning issued by AA’), the event is relegated into the temporal context of ‘yes-
terday’ by means of the default reference of the simple past tense in the lead (The
motoring group said…), although the event could, in fact, be several days old.4 The
news is based on a statement issued at a particular time although it is the content
of the warning rather than the warning itself that is newsworthy – the message of
‘motorists being ripped off ’ is then more difficult to pinpoint in temporal terms
since it refers to a general time, roughly equivalent to ‘at present’. In addition,
while the readers are not given any specific temporal expression contextualizing
the statement itself, they do encounter numerous temporal expressions specifying
the additional information contained in the text (… has fallen by around 9 pence
a litre over the past month; January’s planned increase; fell from around 54 pence
a litre at the beginning of October to around 45 pence a litre by the end; prices fell
by less than 4p a litre between mid-October and the start of this week). All these
adverbials of time partly substitute for the absence of a specific temporal reference
point for the main event.
3.8 Deictic centres in print newspapers
Having turned our attention to news discourse, the issue of deictic centre pro-
jection may require the specification of several additional deictic centres. These
individual centres can be arranged sequentially as points in time located on a time
axis. They include: event time (ET), coding time (CT), publication time (PT) and
receiving time (RT). Although pragmatic and semantic analyses of deixis com-
monly operate with a tri-fold distinction of temporal relations, identifying such
4. There are some events whose newsworthiness is guaranteed by other criteria than recency
(e.g. popular scientific articles about researchers’ findings). As a result, such news texts can
be published later than the day on which the information is released through the agencies or
obtained in some other way by the newspaper. Interestingly, in many national newspapers that
monitor international news sources, one can find a practice that serves as evidence of the loose
temporal anchorage of certain news items to a specific moment in time. This concerns the
situation when a news item published in English in an international (e.g. British) online daily
will be recycled a few days later in a slightly modified version in some national newspaper in
the local language. In such cases, temporal indicators anchoring the main event to a specific
time are – understandably – missing. The default processing of the news text as having ‘current
relevance’ prevents the readers from questioning the lack of temporal indicators and seeing the
absence for what it sometimes is: a strategy of ‘selling off ’ an old news item on an unsuspecting
audience. In this way, newspapers can pre-plan some of their content and ensure the balance of
topics that they write about.
News News News

occurrence production consumption
ET CT PT RT
event time coding time publication time receiving time
Figure 3.2 The sequence of distinct temporal categories (and deictic centres)
moments as point of speech, point of event and point of reference (or time of
utterance, topic time, and time of situation; cf. von Stutterheim et al. 2003), it
appears useful to distinguish an additional centre in order to do justice to the
complexity of news texts.
The timeline with the individual categories arranged in a linear sequence can
be illustrated by means of Figure 3.2.
Thus, the event time ET is constituted by the temporal frame of the event that
is reported and turned into a news story; the coding time CT refers to the moment
the news text is written; the publication time PT is the date actually inscribed on
the paper’s front page; and the receiving time RT is the time when the news is
consumed. Since most print newspapers operate on a daily publication basis, the
publication time PT and the receiving time RT constitute temporal units in the
length of one day. Moreover, the two units overlap. Their coincidence is the result
of the conventional cultural expectation that newspapers bring texts of current
relevance and have a limited shelf life: news texts are meant to be consumed on
the day of the newspaper’s publication. Beyond that relatively brief ‘window of
currency’, news items are dated and lose their informativeness. News is produced
and formulated on the assumption that the publication time PT and the receiving
time RT will coincide. However, as we saw in the previous section, news is some-
times read even beyond its anticipated life, which might call for the need to make
a more finely tuned distinction between ‘assumed receiving time’ and ‘actual re-
ceiving time’. This distinction, however, is marginal for the purpose of the present
account and does not warrant the treatment of the two zones as two independent
categories. Receiving time RT is taken here to be the preferred time of reception
that the newspaper anticipates for the periodical publication, although the inter-
nal complexity of that time zone is hereby acknowledged as well.
The other categories are not without some problems, of course. The event
time ET, for instance, sometimes not only includes actions of prolonged duration
but also disguises the internal temporal complexity of many news events (cf. Bell
1998). Similarly, the conceptualization of the publication time PT combines the
48 News and Time
time of the physical production of the printed copies and their distribution, al-
though such moments are, strictly speaking, quite distinct.
What matters, then, is how fine-grained the distinctions one wants to work
with: whether to meticulously separate all the possible temporal moments that
are involved in the processes of the production and consumption of news stories,
or be content with a certain level of abstraction, using more fuzzy-edged catego-
ries. Not all possible deictic centres, however, are equally important: only the four
basic categories participate in deictic shifts and projections and leave linguistic
traces in the news texts.
For instance, the publication time PT is a part of the news production pro-
cess, and, as such, it rarely leaves an explicit verbal trace in the news text. The time
is normally projected into the anticipated frame of the text’s consumption. How-
ever, the publication time can become prominent, for instance, when a news story
takes an unexpected twist during the period between the coding time CT and the
receiving time RT. At such moments, there may even arise the need to index the
publication time by means of some metalinguistic comment such as ‘after going
into print’, etc. Similar phrases underlie the media’s awareness of the developing
nature of the story from the time of its encoding (CT) and production (PT) until
the moment of reception (RT).5
3.9 Pre-emptiveness of deictic time adverbials
Let us now propose a model of deictic projection as it operates in periodic news

texts. When reading a daily newspaper, a reader is typically processing the text
with the belief that he/she is reading a current issue. In pragmatic terms, this
is the underlying assumption of temporal simultaneity that obtains between the
5. One is reminded of the occasional need to distinguish between printing time (PrT; related
to news production) and publication date time (PDT; related to news reception and printed
in the newspaper’s masthead) on those occasions when a major event happens in between the
printing time and the publication date time (i.e. the receiving time/date of publication), ren-
dering the entire news text obsolete. Such a situation occurs, for instance, when a paper prints
an interview with a celebrity in a weekend supplement and the celebrity dies a day before the
already printed supplement is distributed nationwide. Rafferty (2008: 219) reports on a head-
line that he saw changed during the proofs of the first edition from Labour rebels face ex-
pulsion into Labour rebels are expelled on the grounds that “it was ‘more or less done and
dusted’ ”. However, the change of perspective, where the predicted future outcome was worded
with the actuality modality of the present tense instead of the lexically expressed modality of
the verb ‘face’, turned out to do a disservice to the story because a court injunction was issued
after the paper went into print, bringing an end to the possibility of the rebels’ expulsion.
receiving time RT and the publication time PT (RT = PT; the same assumption
exists on the part of the publisher). The assumption allows for the use of non-cal-
endrical means of temporal deixis, such as the time adverbial yesterday. In tra-
ditional print journalism, however, the written accounts of current stories are
usually composed the day before in order to allow for the overnight printing and
early morning distribution of the actual issues. In other words, the coding time
will coincide with the event time (CT = ET), which both precede the publication
time and the receiving time (PT = RT) by one day, in order to meet the daily dead-
lines that regulate the production of news (cf. Bell 1991: 201).
The distinction between PT and RT helps to account for the fact that a read-
er can properly understand the temporal deixis in an old issue of a paper. The
comprehension is possible thanks to the reader’s ability to link such deictic ex-
pressions as yesterday, over the next few days, last Tuesday, etc. to the publication
date PT rather than the receiving time RT. This is because it is the publication date
PT – and not the RT – that provides the standard time frame for locating an event
in time. Otherwise, temporal deixis could not work since it needs to be linked to
the non-deictic, calendrical time frame.
At this point, let us recall the distinction between two types of adverbials of
time. The category of time can be linguistically coded through either (1) non-deic-
tic or (2) deictic expressions (cf. Yule 1996: 14). The non-deictic expressions are
context-independent: they include calendrical expressions and clock time, such
as on 15 September 2012, at 11.55, etc. They refer to relatively fixed points in time,
though they may presuppose some level of shared context (e.g. a given year, if
formulated as 15 September, or a given day, if a definite time expression – such as
at 11.55 – is used without any further specification). By contrast, deictic expres-
sions are indexed with respect to the moment of utterance; in that sense, they are
more subjective since they articulate a specific point of view. The same reality, of
course, can be coded through either non-deictic or deictic means, cf. the division
of the time zone through such expressions as 14 September / 15 September / 16
September vs. yesterday / today / tomorrow. Clearly, the writer has the choice from
the pair of locally non-deictic and deictic expressions with the same reference (14
September – yesterday; 15 September – today; 16 September – tomorrow). While
the choice of one over the other does not affect the propositional content of the
utterance in any major way, the interpersonal effect is rather different. The lat-
ter (yesterday / today / tomorrow) constructs a shared temporal context between
the encoder and the recipient while the former (i.e. the calendrical expression)
appears more impersonal and objective on account of its comparative lack of an-
chorage to the presumed moment when the message is received.
50 News and Time
It is the deictic expressions that tend to be preferred, at least in most face-to-

face situations. This is because they are pre-emptive with respect to the non-deic-
tic indicators of time. As noted by Levinson,
the deictic words yesterday, today and tomorrow pre-empt the calendrical or ab-
solute ways of referring to the relevant days. Thus the following, said on Thursday,
can only be referring to next Thursday (or perhaps some more remote Thursday),
otherwise the speaker should have said today: I’ll see you on Thursday. The same
holds if it is said on Wednesday, due to pre-emptive tomorrow.
(Levinson 1983: 75)
The following example illustrates a similar situation of pre-emptiveness, with the

only exception that the reference is retrospective towards the past rather than
prospective towards the future:
A court ruled on Thursday that Blær (which means “gentle breeze”) can be
legally used as the girl’s first name. (Icelandic girl wins right
to be called gentle breeze; The Guardian, 31.1.2013)
The phrase on Thursday appeared in a lead to a news article published, inciden-

tally, on a Thursday as well, i.e. the news item presented is one week old. It is
noteworthy that the encoding does not use the adjective ‘last’ (*last Thursday).
The pre-emptiveness of the deictic words yesterday, today and tomorrow over the
calendrical adverbials prevents the possible interpretation of the phrase on Thurs-
day in the example above as co-referential with any of the retrospective adverbials
(both yesterday and today, with the latter common in combination with the past
tense in online news).
Continuing his account of pre-emptiveness, Levinson goes on to remark that
“[p]erhaps this pre-emptive nature of pure deictic words is a general tendency: it
takes special conventions to make it appropriate for a speaker to refer to himself
by name, and it would be strange to say Do it at 10.36 instead of Do it now, when
now is 10.36” (1983: 75ff.).
While Levinson does not specify what the ‘special conventions’ might be, one
could certainly imagine some suitable contexts. After all, providing more infor-
mation that is needed in a given situation, where some information is normally
presupposed, constitutes a conversational implicature on account of a violation of
the conversational principle, primarily the maxim of quantity (‘Do not make your
contribution more informative than is required’; Grice 1975: 45). The speaker’s
preference for the non-deictic calendrical and clock-time expressions over the
pre-emptive deictic ones would then motivate the hearer to look for additional
interpretations. The hearer might conclude, for instance, that the speaker is not
aware of the current time or the day of the week, a situation that happens not that
infrequently.
Without adequate anchorage to some point on the time axis, deictic expres-
sions could not operate effectively in communication. As Brown and Yule observe,
“all you require to arrive at an interpretation are values for expressions being used
to refer” (1983: 43). The presupposition of a specific time at which a text is to be
consumed does provide such an anchorage, by assigning a specific value for a
relevant temporal expression of a deictic type. This holds for such expressions as
last week, last month, last year, two years ago, next week, next month, next year, on
Tuesday, etc., whose time reference is always anchored in relation to the produc-
tion time PT or, more specifically, to different aspects of PT – a particular day,
month, and year, respectively.
Consequently, if a news story published on a given day – say 15 September –
uses the deictic time adverbial yesterday, that expression will then be taken une-
quivocally to refer, from the point of view of both RT and PT, to the event time
(ET) of the preceding day, i.e. 14 September. Its value is fixed to that particu-
lar day – though only for one single day due to the ‘limited shelf life’ of news
as a highly perishable commodity (cf. Bell 1991: 201). As mentioned above, the
anchorage is achieved through either (a) the implicit assumption of the paper’s
recency (with the precise anchorage point explicitly stated on the newspaper’s
front page or its masthead) or (b) by means of a dateline attached to a particular
article. The latter method was relatively common in the past, and it has now again
become the rule in online versions of news reports in electronic newspapers.
3.10 Modelling deictic projection in news texts
Considering the production schedule of daily print media, the four temporal
frames postulated above could be related in the way represented schematically
below. The chart operates with coincidence between the publication time PT and
the receiving time RT, where PT is taken to be the date of publication inscribed in
a printed newspaper (despite the fact that the physical printing is done overnight
and coincides, rather, with the date of the coding time CT):
ET = CT → PT = RT
Figure 3.3 Projections (→) and simultaneity (=) of time frames in a daily newspaper
52 News and Time
The figure shows that the event time ET coincides with the coding time CT (ET =
CT). This is the case when the newspaper reports events that happen on a given
day and are written up for the next day’s print version of the paper. The journalists’
deictic centre, coinciding with the text production, is projected to the publication
time, i.e. the next day (CT → PT). The publication date establishes an anticipated
temporal frame for the consumption of a given media text, which – in the case
of daily newspapers – is one day. These two dates coincide (PT = RT) since it is
assumed by the news producers that the text will not normally be processed be-
yond the window opened up for the consumption of periodical publications, i.e.
beyond their (limited) shelf life. The model, as represented in the figure, achieves
a remarkable compression of temporal space: thanks to the projection, a range of
times spanning the event time ET and the receiving time RT can be conceptuali-
azed together.
The model is applicable also to periodicals operating on other than a daily
news production basis. The extent of the relevant frames of temporal reference
merely needs to be appropriately extended in order to account for a longer period
of time during which the news texts can be considered as current and non-stale.
Modern online journalism, particularly the advent of live text reporting, howev-
er, has effaced the need for the almost automatic projection of the coding time
deictic centre to the publication time, since news items now appear online on the
same day, often with a minimal delay after the occurrence of a newsworthy and
reportable event and increasingly even simultaneously with unfolding events.
The following Figure 3.4 illustrates the effect of deictic projection with a con-
crete example. It explains two issues: first, the difference between the reference
of the adverbials today and yesterday in ordinary non-projected contexts (the
upper part of the chart). The difference incorporates the two possible points of
view (namely the encoder’s and the recipient’s) that are reflected in the different
linguistic representations of time. The second issue demonstrated in the figure
concerns the reference of those time adverbials in a daily print newspaper which
realize the projection of CT to PT (the bottom part of the chart). The shading in
the box illustrates the common temporal space created by the projection of the
coding time into the future; a clear line of separation is drawn between the ET
and the CT.
It is evident that whatever constitutes ‘today’ for the encoder will, by virtue
of projection, have to be verbalized as yesterday and whatever is ‘tomorrow’ will
have to be expressed as today. In the case of weeklies and monthlies, as men-
tioned above, the general scheme remains the same; only, it concerns temporal
deictic expressions of a higher rank. Obviously, the expression yesterday could
not be sufficiently anchored in a weekly newspaper since the possible receiving
time RT spans seven different consecutive days and, on each of them, a different
Real date
15 September 2012 16 September 2012
(non-deictic)
Time frames
(without projection) ET = CT PT = RT
Deictic adverbial
(a) from CT point of view today tomorrow
(b) from RT point of view yesterday today
Temporal frames
(with projection) ET CT → PT = RT
Deictic adverbial
yesterday today
(printed in newspaper)
Figure 3.4 Temporal reference of the time adverbial yesterday
reading of yesterday would ensue. For periodicals appearing less frequently than
on a daily basis, the adverbial yesterday is not applicable since it becomes mean-
ingless on account of the extent of the anticipated length of the receiving time RT.
Instead, the appropriate expressions which participate in the deictic shift and the
anticipated receiving time include such higher-level expressions as last week, on
Tuesday, and next month. For weeklies/monthlies, essentially the same situation
obtains: the publication time PT provides a watershed moment separating two
distinct one week/month periods: (a) the one-week/month period of the antici-
pated receiving time RT (towards which the coding time CT is aligned and pro-
jected) that follows and (b) the one-week/month period of the event time ET and
the real-time (non-projected) coding time CT that precedes.
To sum up, temporal deictic expressions differ in the scope of the tempo-
ral frame which they presuppose as shared. Expressions such as on Tuesday, last
Monday etc. can be successfully interpreted only on the assumption that both the
writer and the reader share the temporal frame of reference in the extent of one
week. Unlike yesterday/today/tomorrow, whose real-time reference shifts on a dai-
ly basis as they become replaced by new, formally identical expressions, the adver-
bials of the on Tuesday type retain their valid reference for several days. The same,
of course, holds true for months (and to a lesser degree also for such (incom-
plete) calendrical expressions like February 11, where the time reference, shared
by the writer and the reader, is actually the current year). The shared contexts of
54 News and Time
time directly derive from the medium’s periodicity (day/week/month).6 While a

paper published on a daily basis enables the application of shared temporal con-
texts on the level of days (yesterday, on Tuesday), weeks (last week), months (next
month), and years (this year), a paper published on, for instance, a monthly basis
enables only the application of shared temporal contexts of the same or higher
level – months and years. Because of the extended receiving time RT (spanning a
month), a shared temporal context of days and weeks cannot be invoked.
The temporal situation could also be explained with reference to relevance
theory (Sperber and Wilson 1986, 1995) in that the reader seeks to minimize
his/her processing effort and maximize the contextual effects arising from an ut-
terance. In that sense, once a plausible context for interpretation is established
(i.e., the reference of the adverbial yesterday to the day immediately preceding
the expected publication date of the newspaper and hence the assumed time of
preferred reception), it would not be economical to try to work out any other
contexts (e.g. cancelling the presupposition of the conventional projection of the
journalist’s coding time CT). This corresponds, in other words, to Sperber and
Wilson’s presumption of optimal relevance. Such a theoretical interpretation is, of
course, related to what Brown and Yule (1983: 59) referred to earlier as the princi-
ple of local interpretation, which “instructs the hearer not to construct a context
any larger than he needs to arrive at an interpretation”.
The assumption of recency leads the readers to infer the correct frame of
temporal reference for adverbials of time. In the case of daily print media, the
presumption is so strong (cf. Bell 1991: 202) that readers will understand a hard
news item to have happened the day before even in the absence of an explicit time
adverbial such as yesterday. Commenting on why many leads do not specify time,
Bell (1991: 180) argues that “[w]e interpret these as ‘reported within the past 24
hours’ in the light of the recency criterion” but he warns that this is not always
the case since “in fact some of these items could be days-old news and we readers
would be none the wiser”. The omission of time adverbials can thus be a ploy set
up not to identify certain news stories as ‘stale news’ – the presumption of recency
will be extended, by default, to all articles printed in a given issue of a newspaper.
A different situation occurs when an event happens past the deadline or if its
relevance is missed by the editors of the daily papers. The event is then reported
two days later, even though a more up-to-date perspective needs to added on the
story. As Cotter observes,
6. One can operate with even longer periods, though it is less imaginable that such a situation
would be applicable for periodical mass media publications. Annual reports, for instance, con-
strue of time with the assumption of shared reference with the extent of one year; statisticians
and historians may even operate with decades and centuries.
the second-day lead has to convey the importance of the first-day (as it is, osten-
sibly, the first time readers will see it), without suggesting that it is the first day
(or else it would be misleading), incorporating what is new about the second-day
with the relevance dimension of the first-day.
(Cotter 2010: 157, original emphasis)
As a result, there may be no explicit temporal indicator anchoring the original

event in time so as to avoid the impression that it is old news. A similar strategy
is used by weekly papers that will mention the newest aspect of the original event,
because their aim is not to “suggest that the story happened yesterday, as it were –
just that it is freshly reported and up-to-the-minute” (Cotter 2010: 158). While all
the information about the ‘first-time’ story is typically given, the temporal place-
ment of the event is buried deeper beyond the lead in the article.
Last but not least, let us mention that there are occasions when some of the
temporal frames involved in the production and reception of news become fore-
grounded or when some individual level needs to be explicitly differentiated from
others. For example, a deictic centre built around the event time ET can be in-
voked in direct speech quotations as well as in the conventional use of the present
tense in headlines (the tense shift pattern – see Chapter 6). Similarly, the coding
time CT may involve the writer reflecting on the state of affairs as of the moment
he/she composes the story (cf. verbalizations such as at the time of writing). The
receiving time RT may, on occasion, be presented as distinct from the publication
time PT or become textually highlighted, for instance when the writer makes a
direct reference to the receiving time RT (cf. the following possible formulations:
*… by the time the reader reads this on the way to work tomorrow morning…;
*… when somebody reads these lines ten years from now…, etc.).
All in all, the projection of the writer’s deictic centre that results in the cre-
ation of a deictic simultaneity with the reader’s centre, as discussed extensively
in the present section, is a powerful interpersonal phenomenon. A newspaper
story that uses the deictic terms yesterday, today, etc. is, arguably, more interactive
because it involves the reader to a greater extent by predicating a common frame
of temporal reference. By contrast, the ‘objectifying’ calendar designations used
for indicating the fixed (i.e. unequivocal) temporal placement of events (Septem-
ber 15 instead of yesterday) call for the reader’s involvement to a comparatively
lesser degree. Indeed, if the ‘objectifying’ September 15 with the meaning of yes-
terday was to be used in a story, a reader would most likely conclude – by virtue
of (a) the operation of the Gricean maxim of quantity (‘Do not make your con-
tribution more informative than is required’), (b) the pre-emptive nature of deic-
tic time adverbials, and (c) the subconscious quest for optimal relevance – that
September 15 was too overt, i.e. more informative than was required, and that, as
56 News and Time
a consequence, a special reading applied. Such a reading would most likely lead to
the conclusion that September 15 did not, for that particular reason, correspond
to yesterday (relative to the publication time PT; see above) but that it most likely
concerned some previous date (i.e. ‘not-yesterday’). In some other genres, such
as police logs, detective stories, transportation/navigation contexts, etc., the use
of calendar designations functions as a stylistic device: the almost scientific pre-
cision in recording time, place, and person constitutes a text-forming strategy
that will immediately cue the readers’ recognition of such genres. In newspaper
discourse, by contrast, it is not scientific precision in recording temporal referenc-
es that is sought; instead, it appears that the guiding principles, at least as far as
the articulation of the temporal placement of news items is concerned, are partly
motivated by the need to engage the reader and operate with the presumption of
a shared context.
3.11 Temporal deixis and tenses
After an adverbial of time fulfils its basic function by providing the anchorage of
events to the time line – either absolutely or in relation to the time of speaking – it
need not be re-expressed in subsequent utterances unless a change in the tem-
poral ordering of events is required. Adverbials of time may then have a rather
sporadic presence in a text – unlike the other means of indicating temporal deixis,
namely tenses, whose presence is an obligatory feature in every finite verb form
in a sentence. Not all tense forms, however, serve the primary (deictic) purpose of
indicating the temporal placement of events; non-deictic uses are also common.
(One such situation concerning the conventional present tense in news headlines
is described in Chapter 6.)
The connection between adverbials of time and tensed verbs is, to a signifi-
cant extent, predictable. The past tense, for instance, correlates with certain time
adverbials (both calendrical expressions and those relating to the moment of ut-
terance, such as on Tuesday, last year, yesterday). Besides the simple past tense,
of interest is the present perfect; i.e. a tense that describes a past-time event with
consequences or results for the present and that frequently appears without any
adverbials of time in news texts. Cf. the following example from an online news-
paper, which includes the ellipted present in the headline, the present perfect in
the lead, and the simple past in the body copy:
‘Very silly’: Professor convicted of scrawling polite graffiti
A university professor has been convicted of scrawling polite graffiti on luxury
cars causing thousands of pounds worth of damage.
Stephen Graham, 47, who is a specialist in cities and urban life at Newcastle
University, today pleaded guilty to the wrecking spree.
He used a screwdriver to scratch words such as “very silly”, “really wrong” and
“arbitrary” on to vehicles in the Jesmond area of Newcastle.
Graham admitted four counts of damaging cars, at Newcastle Crown Court,
and asked for 23 similar offences to be taken into consideration.
At an earlier hearing the value of the damage stood at almost £18,000. The
cars included a Mercedes, an Audi, a Volvo and a Mitsubishi.
The spree took place in August last year when residents of Northumberland
Gardens woke up to find the polite phrases etched into the paintwork of their
cars.(The Independent; 18 January 2013; emphasis added)
Let us compare the effect of choice of tense in the article. In general terms, the
present perfect is a more interactive tense than the simple past. This is because
the present perfect highlights the current relevance of events that occurred in the
past, thus linking the past and the present, while the simple past merely locates an
event in the past time, without providing for such a link. In that sense, the present
perfect presupposes the receiving time RT, while the past tense does not. In news
stories where the present perfect occurs it is used without adverbials of definite
time (cf. A university professor has been convicted of scrawling polite graffiti… in
the example above). The absence of time adverbials with the present perfect in-
creases the news values of relevance and topicality because of the implicit refer-
ence to the present import of the event.
By contrast, the first occurrence of the simple past in the second paragraph
is contextualized with the use of an expression of definite time that adequately
locates the event in past time (Stephen Graham, … , today pleaded guilty to the
wrecking spree). In print news, the most characteristic adverbial indicating past
time is the adverb ‘yesterday’, which reflects the daily production process of tra-
ditional media. Online news, by contrast, publishes news as it becomes known
from agency copy on the same day the reported event occurs, and thus favours the
past-time orientation of the adverbial ‘today’ (i.e., the use that is almost altogether
absent from print news, save for occasional metacomments).
Where the past tense is used without an adverbial of time, it refers simply to
the past time, without making any explicit link to the present. This situation arises
when the temporal frame is already established and firmly anchored to a specific
point in the past. The past tense without a time adverbial appears in this function
in the fourth paragraph (Graham admitted four counts of damaging cars… and
asked…). The actions of ‘admitting’ and ‘asking’ are clearly related to the temporal
zone of ‘today’, more specifically the man’s court appearance. However, this par-
ticular sentence is preceded in the previous paragraph by a similar use of a past
tense without the accompaniment of a specifying adverbial of time (He used a
58 News and Time
screwdriver to scratch words…). The tense does not require any explicit anchorage
in the past even though it refers to a different (earlier) time zone from the imme-
diately preceding past tense in the second paragraph (… today pleaded guilty…).
This is because readers are able to clearly differentiate between the two time zones
on the basis of (1) the text’s internal cohesion (the sentence provides details of
the ‘wrecking spree’ mentioned before) and (2) its coherence based on the readers’
knowledge of the relevant scripts and schemata (readers can be expected to know
that a plea in court relates to a previous transgression of legal regulations and thus
are able to infer the causal, as well as temporal relations between the two prop-
ositions). A non-calendrical adverbial appears only in the final paragraph (The
spree took place in August…) when the background to the main event is provided
already for the second time.
As the example shows, adverbials of time interact interestingly with verbal
tenses. The text uses an adverbial of time (today) that is deictic, relying on the
same context of temporal reference between the encoder and the addressee – a
context that, as illustrated above, is established through deictic centre projection.
The comparative lack of interactivity of the simple past – at least with respect to
the present perfect found in the lead – is thus compensated for through the choice
of the time adverbial.
Overall, the internal temporal structure of the article is non-chronological
and quite complex, since the narration of the news story jumps from one zone to
another across the individual sentences of the news text. This narrative structure
and non-chronology, as well as the grammatical operation of the simple past tense
and the present perfect in news texts, are discussed in more detail in the sections
to follow.
Chapter 4
Temporal deixis in online newspapers
The discursive management of temporality in traditional print journalism dif-

fers significantly from modern journalistic practices found in online newspapers.
Print journalism is governed by the necessity to observe the strict deadline by
which all texts must be composed, edited, typeset and prepared in order to be dis-
patched to print. By contrast, online newspapers lack such a watershed moment.
Instead, the web pages of major newspapers are updated at irregular intervals
and new items and stories are added one by one as they are produced by the
journalists and approved by the editors. Moreover, existing stories are commonly
updated with new content once more up-to-date information becomes available –
the texts are thus not static and stable but subject to modification and editing,
sometimes giving rise to several versions of a single article.
The distinction between traditional print journalism and online journalism
could be described as ‘product’ vs. ‘process’: while the former is an activity that
culminates in the final copy of a newspaper, the latter is an ongoing process in
which the website constantly evolves, develops and metamorphoses. The 24-hour
rolling news cycle found in online media is putting new pressures on journalists.
One of its effects, for instance, is the proliferation of the textual lifting of stories
from rival newspapers. Since the media increasingly monitor each other’s con-
tent, this has inevitably led towards a homogenization of their content (cf. Phillips
2010: 96, 101).
In this chapter, I analyse three related issues. First, I argue for an extended
view of what constitutes a modern news text in other than print media. Thanks to
hypertextuality, the online news text consists of the article preview and the arti-
cle itself. While the preview provides one of the possible points of entry into the
article, its component parts (headline + subhead/lead) often differ from those of
the article itself. Online news, thus, has an additional textual level on which to ar-
ticulate temporality, even though the article preview is, in fact, merely temporary,
since it is not, unlike the article, preserved in the newspaper’s archives. Another
issue discussed in this chapter concerns the temporal anchorage of online news
articles. Because of rolling news coverage, some of the pragmatic assumptions
governing the logic of the projected deictic centres are different in online news
than in print news, with individual papers stressing the ‘current relevance’ of their
coverage. Last but not least, hypertextuality results in the mutual interlinking of
60 News and Time
related articles across time. The intertextual network, in which a reader can easily
navigate the history of a topic in a chain-like manner, weakens the position of
the single article. While traditional print news, which lack such easy access to its
textual predecessors, enjoy a more independent existence, online news are much
more easily accessible beyond the temporal context of their online publication.
As a result, the indication of temporal anchorage may need to be strengthened
by more explicit reference, e.g. we increasingly see that hyperlinks to previous
articles tend to be realized as a single unit consisting of a headline and the date of
publication.
4.1 Hypertextuality and the double textual level of online news
The major textual features of printed and online news are identical: headlines are
used and the news texts are structured in the familiar top-down manner. Similar-
ly, there is a tendency to cluster several articles on the same topic in the ‘package
approach’ to devising newspaper content (cf. Ungerer 2000). Online news has
been found to be more fragmented (focusing on “the minimalist data chunk”, cf.
Lewis 2003: 97) and layered, with a news topic “developed as a cluster of dynam-
ic, related, hierarchically-structured texts, like overlapping groups of concentric
circles” (Lewis 2003: 97).
Online news style combines features of the traditional news article with the
properties of hypertextuality (cf. Lewis 2003: 99). Thanks to hypertextual links,
news texts participate in an extensive intertextual network of textual and non-tex-
tual resources. The hyperlinks in online news articles can be of two basic types:
external and internal. While external links transfer the reader outside of the news-
paper’s web site, e.g. to source documents, other institutions, the paper’s adver-
tisers, etc., internal links provide for the construction of a broad textual network,
transporting the reader to related articles, commentaries, analysis, video clips and
other content within the same medium – a densely interlinked multimodal hy-
pertext. This situation is described by Jucker as follows:
The term “internal link” refers to those links that connect textual elements within
the same hypertext; and the term “external link” to those that connect to textual
elements of different hypertexts. Internal links can be further subdivided into
“intratextual links” for links within one textual element (or e-text), and “inter-
textual links” for links between different textual elements of the same hypertext.
(Jucker 2002: 40)
Intratextual links are typically not applicable in the case of online news since a
single textual element (an online article) is usually presented in its entirety on a
Chapter 4. Temporal deixis in online newspapers 61
single web page and without any internal signposting. By contrast, intertextual
links are very frequent. The two distinct kinds of intertextual links – internal and
external – can result in an intertextual network of news reporting on a single
topic spreading across an extensive period of time (for an example of a graphic
representation of such a hypertext, see Figure 4.10 below) and a truly boundless
hypertext if the external links are followed to other online content on media-ex-
ternal websites. It is not only that hypertextuality allows readers to access other
sources; also, it may lead to fragmentation of the online news article into what Yus
(2011: 82) calls “link-mediated chunks”.
As regards the nature of temporal deixis, online news differs from printed
news in two important respects. First, the temporal frame of reference is less stat-
ic. While news content in printed newspapers has a fixed anchorage point in the
form of the anticipated receiving time RT (identical with the publication date of
the paper), online news sites usually update their content round the clock and the
anticipated receiving time is not ‘the next day’ but, less definitely, the ‘time period
following the online publication’. Second, where printed news articles stand as
isolated textual monoliths, online news articles form, thanks to hypertextuality,
an extensive inter-text constructed through temporally organized internal links.
The genre of online news is also particular in that there is an additional tex-
tual level to the news article: the front page news preview.7 This consists of a
headline and (typically) a summarizing lead/subhead that will take the reader, by
means of a hyperlink, to a page with the full text of the article. What we then find
on the home pages of newspapers are brief newsbites, very often punctuated with
thumbnail images that function interpersonally, building up expectations about
the text (Knox 2009: 164). The combination of the headline and the lead appear-
ing on the home page constitutes an autonomous text that is often processed by
readers without them actually reading the rest of the news story on the article
web page. This situation is reminiscent of the way traditional headlines operate:
independently of the body copy of the relevant articles as relatively stand-alone
units (cf. Ifantidou 2009: 702).8
7. Some printed newspapers also include brief article previews on their front pages. However,
that concerns only a very limited number of the most important news items reported beyond
the front page. Thus, some select print news can be subjected to a two-level analysis similar to
that proposed here for online news.
8. It could be argued that the journalistic lead is inextricably linked to the body copy: the func-
tional value of that segment rests in the fact that it is followed by the article proper. However, the
lead can also operate as a stand-alone unit, existing only in conjunction with the headline – as is
the case with some very brief news items. That is the use which the newspaper home page usage
resembles. In the absence of a body copy, the whole news text can be constituted merely by the
62 News and Time
The front page preview is one of several ways in which an article can be ac-
cessed. Additional points of entry to the same article include sections such as
‘Most viewed’, ‘Most commented’, and ‘Most shared’ (in the Independent) and
‘Breaking news’, ‘Most viewed (Last 24 hours)’, and ‘Latest’ (in the Guardian). Sev-
eral ways of accessing news content may be distinguished: (a) access via an article
preview; (b) access via other internal/external hyperlinking that may be various-
ly textually introduced; and (c) direct access through the article’s unique URL
with little or no textual introduction. Through hyperlinking at various other sites,
readers are brought to the same text from different directions. As a result, they
will have different expectations and will be reading the text in different textual
and temporal contexts.
Once the hyperlink contained in the preview is followed, the reader is taken
to a specific page with the full text of the article, complete with a headline and
links to other textual and possibly non-textual content. The full text article is suf-
ficiently complex, at least as appears from the data analysed for the purpose of this
study, which does not confirm the trend of online news to present minimalist data
chunks as mentioned by Lewis (2003).
The hyperlinked transition from the home page to the article web page, how-
ever, is interesting for another reason: namely the textual transformation of the
article segments. The headline and the lead/subhead on the article’s own page are
frequently not identical with the corresponding textual segments on the home
page. This indicates that the front page serves as an additional level on which
information can be mediated to readers through the headline and the lead in a
gradual way, ideally with the aim of avoiding repetition and adding further rel-
evant/newsworthy information as well as the newspaper’s evaluative perspective
on a story.
The framework for analyzing online news thus needs to take into account this
new macro-structural textual level, where the processing of the text progresses in
several steps. The brackets in the following flowchart (Figure 4.1) indicate poten-
tial elements that are not realized in all articles.
The first level comprises the article preview on the newspaper’s home page. It
is through the hyperlinked headline that the readers can access the full text of the
article on the article web page. In this way, the preview plays a crucial function
as providing the point of entry into the article itself. On the other hand, the arti-
cle preview is a non-permanent textual feature since it is not archived. With the
combination of the headline and the lead. Similarly, with some genres of brief news items such
as live tickers on TV, a mere headline can constitute a complete and self-contained news text. It
almost seems a paradox that the textual segments (H/L) that normally abstract the body copy
are actually fully self-sufficient even without the text they summarize.
HEADLINE 1
Article preview Location: home page

(transient level) (SUBHEAD 1 /
LEAD 1)
HEADLINE 2
Full text article
(permanent level) Location: article web page
(SUBHEAD 2)
LEAD 2
Figure 4.1 The major macro-structural textual segments of online news
gradual replacement of article previews on the home page with more current
news, a given headline and a subhead will disappear as well, while the related ar-
ticle web page will be preserved in the newspaper’s archives and may become the
landing page for hyperlinks present elsewhere. By contrast, the article preview,
though initially important, is not hyperlinked (either internally or externally) and
is thus marked with transience.
Let us illustrate some of the pragmalinguistic consequences of this phenome-
non with a specific example. Figure 4.2 presents a news story that is headlined dif-
ferently on the newspaper’s home page and the article’s own web page. The arrows
indicate the linear reading path through which readers ideally progress towards
the content of the story in the body copy. The corresponding textual segments on
the home page and the article web page are aligned to appear next to each other.
This form of presentation enables easy comparison, while indicating the different
hierarchical placement of the two textual levels.
The figure intentionally leaves out visual elements such as photographs and
hyperlinks to related articles, which are commonly included in the complete news
package on the home page and/or the article web page. Figure 4.2 also sets aside,
for the time being, the technical distinction between the lead and the subhead.
While both structures have a summarizing function (which is why they are treat-
ed together here), the lead encapsulates all the main factual constituents of the
news event (who, what, when, where, how). The subhead, by contrast, may only
concentrate on some aspect of the news story, e.g. by elaborating or mentioning
consequences, and is not driven by the requirement of factual completeness to
such an extent as the lead, since it can complement the headline. Some linguistic
differences can be found as well: the lead is typically a full sentence, often specify-
ing the time of the event by means of a time adverbial. The subhead, by contrast,
64 News and Time
may lack complete sentence structure, retaining some of the typical features of
headlinese, as well as, occasionally, the lack of punctuation at the end. Both the
subhead and the lead tend to be typographically marked from the headline and
the body copy, though that is not always the case. The distinction between the
subhead and the lead becomes particularly relevant when an article contains both
elements (cf. the analysis in Chapter 9). In other cases – as in the article preview
on the newspaper’s home page – it may be actually difficult to identify the element
precisely; what matters there, however, is that the element, beyond its precise clas-
sification, has a summarizing function.
As Figure 4.2 indicates, the headline on the web page (H1) is shorter than the
headline on the article web page (H2). While the headline on the second-level
Home page Article web page

(level 1) (level 2)
Headline Troika demand even Debt crisis: troika
tougher Greek austerity demand even tougher
austerity on stricken
Greece
↓ ↓
Lead/Subhead Debt inspectors said to want Greece’s international
cut in minimum wage and creditors are demanding the
pensions. imposition of even tougher
austerity measures despite the
delivery this week of Antonio
Samara’s hard-won €13.5bn
package of cuts.
↓
Body copy --- On the second day of
negotiations in Athens, the
troika – officials represent-
ing the European Union,
European Central Bank
and International Monetary
Fund – reportedly pushed for
Greece to make deeper cuts
to the minimum wage and
pensions, while imposing
longer working hours. (…)
Figure 4.2 The structural expansion of online news stories: the double headline/double
lead pattern (The Telegraph; 3 October 2012)
textual level repeats the core of the proposition from the first headline (this is
the conceptual category of the main event in the article’s narrative structure;
cf. Troika demand even tougher Greek austerity), it also expands the in-
formation by placing it within the context of a long-term topic (Debt crisis)
and supplementing it with additional information in the form of a prepositional
phrase (on stricken Greece).9 By contrast, the leads on the two levels are tex-
tually significantly different. Not only do they formulate the basic elements of the
story in quite different ways but they also bring a different amount of information
to the reader. The lexical transformations and the expansion of the story in L2 are
illustrated in Figure 4.3.
In other cases, the textual segments on the two levels (the home page and the
article web page) are either identical or contain only very minor modifications.
The example in Figure 4.4, taken from the same newspaper, documents the mere
repetition of the headline and the minor alteration of the lead.
Unlike the previous example, the second-level textual segment (i.e. leads/sub-
heads) is formulated almost identically on the paper’s home page and the article’s
own web page. What is particularly noteworthy, however, is that – apart from the
ellipsis of Zlatan Ibrahimovic in the article preview, most likely motivated to save
space and prevent repetition of the information from the web page headline –
the two versions of the segment differ as regards temporal deixis. On the home
page, the present tense is used in the article preview. On the level of the web page,
Lead 1 Lead 2
(home page level) (article page level)
debt inspectors → Greece’s international creditors

said to want → are demanding
the imposition of even tougher austerity
cut in minimum wage and pensions →
measures
despite the delivery this week of Antonio
Ø →
Samara’s hard-won €13.5bn package of cuts
Figure 4.3 Elaboration of information in two levels of leads (The Telegraph; 3 October
2012)
9. The web page headline may be considered as a further abstraction of the article’s headline.
This process can thus result in a triple summary pattern, with summaries found in the lead,
the article headline, and the article preview headline. This issue is elaborated in more detail in
Section 9.2 on the narrative structure of news stories.
66 News and Time
Home page Article web page

(level 1) (level 2)
Headline Arsene Wenger: I tried to Arsene Wenger: I tried to
sign Zlatan Ibrahimovic sign Zlatan Ibrahimovic
for Arsenal for Arsenal
↓ ↓
Lead/Subhead Arsene Wenger hails wonder Arsene Wenger hailed
goal against England – and Zlatan Ibrahimovic’s wonder
claims he once tried to sign goal against England – and
the enigmatic Sweden striker. claimed he once tried to sign
the Sweden striker.
↓
Body copy --- Ibrahimovic scored all four
of Sweden’s goals in their 4–2
win over England in Stock-
holm, but it was his stunning
fourth – a 25-year overhead
kick – which has dominated
discussion since the moment
it hit the net. (…)
Figure 4.4 Repetition of headlines and minimal expansion of leads/subheads in online

news stories (The Telegraph; 16 November 2012)
the same verbs re-appear, but their tense marking is shifted into the simple past
(hails → hailed; claims → claimed).
The modified wording of the corresponding textual segments in the article
preview and the article itself potentially afffects how the lead/subhead is per-
ceived by the readers in each case. This segment appears to have more of a head-
line-quality on the newspaper web page, where it functions as a subhead rather
than a lead, while it is used as more of a lead on the article web page, where the
simple past tense indicates a disassociation from the conventions of the genre of
headlines. Semantically, it repeats rather than develops the information contained
in the headline – a characteristic found in leads and less so in subheads. Yet, the
segment on the article web page is placed under the headline above the accom-
panying picture and the attribution. It is thus separated from the body copy of
the article in a position that seems to indicate the text segment’s allegiance to the
headline rather than the body copy.
4.2 Temporal anchorage points in online newspapers
One of the advantages of the online environment is that news items can be pre-
sented to the audience one by one as they are prepared by the editorial office.
News does not need to wait for the ultimate deadline by which all stories are
simultaneously released to readers, as is the case with any printed publication
that brings all the news to the audience at one go in a single issue. This dynamic
continual presentation of news has an impact on the definiteness of the anchorage
point for the stories: although online articles certainly operate under the same
assumption of recency as their print counterparts, each article is released inde-
pendently of each other. Also, it is frequently the case that home pages include
relevant articles from the previous day as well, without identifying them explicitly
as ‘yesterday’s news’. This results in the increasing importance of the byline of each
article, since it specifies not only the author of the text but also the time the article
is released. As a consequence, each online article – unlike traditional news texts in
printed newspapers – explicitly includes the anchorage point for temporal deixis
used in the article. Figure 4.5 illustrates the byline of an article with the explicit
date that serves as such an anchorage point.
Figure 4.5 The dateline as an explicit anchorage point for temporal reference
in an online newspaper (The Independent; 18 January 2013)
68 News and Time
In general, the explicit anchorage point compensates for the fact that a series
of news items cannot be assumed to be temporally co-occurrent if they appear on
the newspaper’s home page. This situation has several reasons:
– The lack of the ultimate deadline for online news items

In the case of printed news, such a deadline provides a definite cut-off point
for editorial work. Owing to the lack of this unifying temporal feature, online
newspapers may find it more problematic to operate with the projection of
deictic centres. In addition, newspaper home pages increasingly depend on
live reporting, thus removing the traditional barrier between the event time
and the coding/receiving time – as a result, the home page is constantly in a
state of flux and development.
– Increased importance of news flow
The information is processed and posted online as it appears, i.e. real time
becomes more important in online news than in printed newspapers, where
all the events from the previous day are ‘pooled’ into the same one-day period
consisting of ‘yesterday’. This means that the distinction between the project-
ed coding time and the anticipated receiving time is disappearing since news
items tend to be consumed on the day they are encoded.
– The changeability of texts
Online articles are frequently updated as the editorial office receives new in-
formation or obtains comments on the events from various external sources.
As a result, an article may have its distinct history, having gone through sev-
eral versions and modifications. It is only the latest version that is available,
rather than the previous versions. Nevertheless, some online media equip
their articles with specific metadata about the articles’ history, thereby con-
firming the potentially changeable nature of the texts.
Figure 4.6 illustrates the fluid nature of online news texts and the way metadata
about any textual modifications are recorded in order to provide temporal an-
chorage for the gradually changing versions accessed by readers at different times.
The attribution line identifies the names of the journalists (James Meikle, Kate
Connolly, and Peter Newlands), references the source in which the article is be-
ing published (The Guardian) and provides objective anchorage of the news text
in terms of absolute calendrical time (Tuesday 19 February 2013 23.52 GMT).
The dateline specifies the time quite independently of the anticipated time of the
text’s reception because it provides temporal anchorage in absolute terms without
presupposing any shared temporal context.10 The article is supplemented with
10. Some newspapers may occasionally opt not to use the objective time reference but instead
encode the time through the adverbial of time today, which relies on the assumption of a shared
Figure 4.6 Temporal metadata about an article: recording multiple article versions
(The Guardian; 20 February 2013; downloaded at 10.20 GMT). The article was subse-
quently updated again – as late as 19 June 2014.
a column on the right, containing action links by means of which readers can
mediate the article and disseminate it to others by means of email and such cur-
rently popular social networking sites as Facebook and Twitter, etc. There are also
three ‘service buttons’ for additional functions (printing the text, sharing it, and
contacting the editorial office). Underneath, there is an active text segment that
reads Article history. A click on the text will open a special scroll window that
reveals the temporal history of the article (underlining in the example indicates
hypertextual links):
This article was published on guardian.co.uk at 23.52 GMT on Tuesday 19
February 2013. A version appeared on p11 of the Main section section of
the Guardian on Wednesday 20 February 2013. It was last modified at 00.35
GMT on Wednesday 20 February 2013. It was first published at 14.27 GMT
on Tuesday 19 February 2013.
The article history enables the readers to duly contextualize the text in case they
are confronted with several versions. The data may also be useful in situations
when it is crucial, e.g. for legal purposes, to document the existence of different
versions resulting from a series of updates of the latest information available at
time frame between the text’s author and its recipients. Where newspapers use an adverbial of
time such as Tuesday independently, such a use is deictic since it presupposes a shared frame of
temporal reference.
70 News and Time
different times of the day. From the point of view of the news production process,
the article history records a textual trace of the changes through which a news
text goes before it acquires the final form in which it is eventually transferred into
the newspaper archives. The actual changes in the formulation of the news text
are not documented; it is only the article’s metadata that remind the readers that
the online news text is potentially changeable – it is much more a text in process
than a text as a finished product. Online news texts, at least on media sites that do
not release articles as unchangeable final end-products that are never revisited by
the editors, are processed with the knowledge that the article is current as of the
moment of reading and that the state of affairs may change just minutes later. The
reader is thus accessing a current version of the news story that is public, though
possibly not yet final.11
This contrasts with print newspapers, where – for obvious reasons – only the
final version is publicly available and the entire news production process, with
possible changes and additions to the news texts prior to the final print dead-
line, are totally invisible. By explicitly documenting the article history in terms of
time, online newspapers can reveal an aspect of the dynamics of the news produc-
tion process that has previously been entirely hidden from readers. Close textual
analysis of the individual versions of such developing news articles holds much
potential for further work in discourse analysis since it may lead to interesting
findings about social, ideological, and other motivations for subtle micro-level
modifications and shifts of meaning between the divergent formulations and dif-
fering contents sometimes found in the individual versions.12
The article history section, used only by The Guardian from among the on-
line newspapers analyzed here, indicates that the various versions have distinct
temporal anchorage points. The part of the article web page containing the at-
tribution and the dateline provides only the last (i.e. the latest) anchorage point
coinciding with the moment when the article was released. When an article goes
through several versions, it is customary to include a note such as “Last updated
11. The constant updating of news articles and websites inevitably results in a change in the
process of news reception and consumption. Readers may find it relevant to revisit a given news
media site at later times – not only will new items be available for additional consumption but
also the old ones may be recast in view of the latest developments of the news story. This stands
in a stark and obvious contrast to print newspapers that are usually discarded after a reader
glances through its pages. Online news sites have the potential to become addictive in the sense
that any moment the reader accesses them, there will be something new and the constant de-
velopment can be an inherent attraction to many readers.
12. The news production process has been systematically studied by scholars associated with
the Ghent group (cf. Jacobs 1999; van Hout and Jacobs 2008; Cotter 2010).
on 12.01AM, January 31, 2013” in the attribution section of the article (cf. Fig-
ure 4.7 below).
Before addressing the ways in which online papers implement temporal an-
chorage of the individual articles, let us turn our attention briefly to the mac-
ro-level of the home page. As mentioned previously, this additional level provides
a preview of the news texts, consisting of the obligatory headline and, most typi-
cally, the lead as well. It appears that two distinct types of temporal anchorage can
be distinguished on the home page:
1. anchorage of the entire page (homepage-specific); and

2. anchorage of the individual articles (article-specific).
4.3 Temporal anchorage on the home page
As regards the way the temporal framework is established for the entire page (news
site), the current date and the day of the week are indicated at the top of the home
page. This form of temporal anchorage essentially stands as a direct parallel to
print newspapers since their mastheads bear very similar information. The date-
line takes the form of a calendrical expression of time, typically in combination
with the day of the week (Thursday 21 February 2013). Since the day of the week is
closely attached to the date, it is not – strictly speaking – used deictically, since it
does not depend for its anchorage on the anticipated time of reception of the news.
Figure 4.7 documents the different datelines used by the newspapers analysed.
There are some minor differences between the individual papers as regards
the way they indicate the date on their home pages. Where The Independent uses
the combination of the day of the week and the full date (Thursday 21 February
2013), The Times does not include the year (Thursday, February 21) and relies on
a shared frame of temporal reference with the readers opting not to provide an
absolute anchorage in terms of calendrical time. With the other two newspapers,
however, we see a deviation from the static format of the traditional masthead.
Both The Telegraph and The Guardian provide a mention about the last update of
the home page (Last updated 3 minutes ago; and Last updated less than one minute
ago, respectively). This text is automatically generated on the screen and related
to the reader’s time of reading. The temporal anchorage in the masthead is thus
realized both by absolute, calendrical means, and deictically, i.e. with respect to
the time of reading, or, more specifically, the time at which the web page is loaded
or refreshed. It is worth noting that the very first thing that appears on the reader’s
screen when the home page of The Guardian is loading is the name of the newspa-
per and the phrase Last updated less than one minute ago. Depending on the speed
72 News and Time
Figure 4.7 Temporal anchorage of online newspapers – home page
of the web browser, the actual news content of the page appears with a slight time
delay. The phrase thus underlies the newspaper’s effort to provide highly current
and up-to-date information and, what is even more important, inform the readers
of the paper’s immediacy.
As regards the individual articles on the newspapers’ home pages, most of
them lack any explicit temporal anchorage. The absence of temporal indicators is
partly related to spatial constraints and partly to the operation of the presumption
of recency. Under the recency criterion, all news texts are assumed to be topi-
cal, i.e. co-temporal with the publication date inscribed in the electronic paper’s
masthead and the day on which the page is accessed. Obviously, there may be a
discrepancy between the date that is automatically inscribed in the masthead and
marking the time the home page is accessed, and the time of publication of the
individual articles. This discrepancy is resolved where the individual article pre-
views contain a time indicator, e.g. in the form of a time stamp prefacing the lead.
From among the newspapers under analysis, The Independent and The Tel-
egraph do not include temporal indicators in any of their article previews ap-
pearing on the home pages. The Guardian occasionally mentions that a story has
been updated (e.g., the phrase Last updated 15 minutes ago prefaces one of its
home page stories on 21 February 2013). The Times, by contrast, appears to use
the following strategy most frequently and systematically: temporal indicators are
included in the most important news of the day, placed at the top of the page in
a section titled ‘Latest news’. The other sections on the same page (‘UK News’,
‘World News’, ‘Business’, ‘Sport’, ‘Best of the Blogs’, ‘Arts’) do not include any ex-
plicit indication of the recency of the news items. The temporal expressions are all
provided in terms of absolute time, i.e. without being linked to the moment the
web page is accessed by the reader. This pattern is represented by the following
three time lines (time stamps) accompanying the most important articles of the
day in The Times:
Last updated at 2:23PM, February 21 2013
This particular format of the temporal indicators has a dual function: the cal-
endrical expressions anchor the version of the text objectively with respect to a
certain point on the time axis, and the descriptive phrase Last updated at… serves
to enhance the current relevance of the event as well as the developing nature of
the news story. A phrase such as this also subtly communicates to the readers the
paper’s commitment to follow up the news stories and modify the texts of the ar-
ticles for the readers’ benefit rather than writing them up once and for all.
74 News and Time
Figure 4.8 Temporal anchorage in online newspapers: articles in news text clusters
(The Times; 21 February 2013)
4.4 Temporal anchorage in article previews
There is another interesting aspect concerning the way time is indicated in indi-
vidual articles in The Times. When the links on the home page are followed to the
individual articles, the newspaper displays a column on the right-hand side with
accompanying articles (usually titled as ‘Behind the story’) and a cluster of other
thematically-related brief article previews (i.e. headlines and leads, referred to as
‘newsbites’ by Knox 2009: 149) from the relevant section (variously titled such as
‘More from Politics’, ‘More from Crime’, ‘More from Retailing’, ‘More from Faith’,
etc.). All these articles include time captions, providing temporal anchorage for
the news stories. Two such sections are included in Figure 4.8.
These date lines indicate that there are several distinct ways of encoding
the publication time when an article appears online. The following types can be
isolated:
1. Calendrical means of indicating specific time in a static way (Published at

12.01AM, February 20, 2013). This method provides for a one-time temporal
anchorage, indicating a static news text that has not undergone any editorial
modification.
2. Calendrical means of indicating specific time in a dynamic way (Last updated

at 4.39PM, February 20, 2013). This method emphasizes the developing na-
ture of the text, indicating that several versions of the text were available to
readers at various times.
3. Deictic means of indicating non-specific time in a dynamic way (Published 5
minutes ago). This method provides for a shared temporal frame of reference
with the reader in a highly individualized manner since the time of publica-
tion is calculated automatically with respect to the moment the reader access-
es the web page.
Article clusters such as those documented in the figure above also show there
may be some variability in indicating time within the individual article previews
since they do not use a single method for the coding of time. Within a themati-
cally-related cluster of articles, the newspaper applies various types of temporal
anchorage. In this way, the part of the article preview bringing temporal metadata
is enriched – objective time reference can be interspersed with more interperson-
al ways of claiming common ground with the readers.
In addition, there is another phenomenon that can be observed here: the for-
mat of news presentation as shown in Figure 4.8 results in the placement, side by
side, of articles that may be recent but not necessarily originating on the same day.
The second news cluster in the figure above, for instance, brings together events
published on three distinct days. Three of the four articles in the cluster are dated
to February 20 (each with a different time of publication) and one to February
18. In addition, since the cluster accompanies a specific article (not shown in the
example above) on its own web page, the date of that main framing article also
needs to be taken into account: it is datelined as February 21, 2013.
Paradoxically, this type of online news presentation bears resemblance to the
way time was encoded in early news discourse, i.e. before the arrival of technolog-
ical innovations (such as the telegraph) at the end of the 19th century facilitated
the transmission of information across large distances. Previous to that, a news-
paper published on a certain day (which provides the primary temporal frame,
just as the masthead of the online papers today) could include – within the same
issue – articles datelined to various days (and locations) that would go beyond the
publication frequency of the periodical. Such temporal marking reflected the time
that it took a report on some newsworthy event to arrive at the editorial office.
The differential treatment of news items on the home page seems to indicate
that several general categories of news could be distinguished in online newspa-
pers according to how they are contextualized in terms of their temporality:
76 News and Time
– ‘breaking news’ – usually with a ticker running at the top of the page. A tem-
poral indicator is absent because the running text is an iconic indication of
emerging news. This format of news presentation is used by some TV news
channels, sometimes even with multiple tickers presenting several news items
simultaneously;
– ‘latest news’ – the time line indicates that the news stories are just minutes old;
– ‘live news’ – a special genre for events that are covered in a minute-by-minute
fashion by reporters in the field; and
– ‘common news’ – all other recent news. This category may include news that
survives on the news site from the previous day, although – in the absence of
a temporal indicator on the newspaper’s home page – this is not immediately
obvious to the reader.13
Interestingly enough, the broadsheets analyzed here indicate the times of the ma-
jor news stories only; none of them uses another possible strategy – found in
some tabloids – of indicating the times of the stories on a more consistent basis
on the home pages. The British paper The Sun, for instance, specifies time for all
of its brief article previews on the paper’s home page. It provides an extra line
above the headline that classifies the category of news (‘News’, ‘Showbiz’, ‘Brits’,
‘F1’, ‘Fun’, ‘Features’, ‘Politics’, etc.), with a temporal specification added to it. The
time is expressed with either deictic means (1 hr ago; 7 hrs ago) or calendrical
expressions (21/2/2013). Although the classification and the publication time of
the news items are given in a smaller font, their omnipresence on the web page
indicates that the paper considers temporal indicators to be very important. They
may increase the newsworthiness of the item, underlining its current and up-to-
date nature.
Last but not least, the categorization of news with respect to their temporality
needs to take into consideration the fact that some online news texts are more
dynamic than others. There are news articles that are ‘updateable’ – they are open
to subsequent modifications. The updates are textually recorded through deictic
and non-deictic temporal indicators such as Updated 10 minutes ago and Last
13. Should further distinctions be needed, the category of ‘common news’ could be extended to
include, more explicitly, ‘recent news’ and ‘non-recent news’. This subdivision does justice to the
fact that news sites sometimes include non-recent content that is only gradually pushed off the
home page by the addition of more current news. This phenomenon is observable e.g. at times
of national holidays when the influx of news is reduced as is the number of staff preparing the
news content. Occasionally, one can come across items that could even be designated as ‘stale
news’ – non-recent news that is obviously old and may result in the reader’s momentarily won-
dering about the reasons for including the item on the news site and its non-relegation into the
newspaper’s archive.
updated at 12.23AM. By contrast, some news articles are static and do not un-
dergo any updating after they are published online. From the perspective of the
reader, however, there is no way of distinguishing between such open and closed
types of texts.
4.5 Temporal anchorage on article web pages
The analysis of temporal descriptors attached to individual articles on their own

web pages needs to begin by considering the macro-structural pattern of online
news that is articulated on two distinct textual levels. As mentioned above, most
article previews on the paper’s home page are without any temporal indicators
and only some include a dateline or a timeline indicating the publication of the
article. All of the articles in the papers analyzed, however, do include complete
temporal specification on the second textual level, i.e. on the article’s own web
pages. In terms of the macro-structural temporal organization, we can thus ob-
serve two basic patterns:
1. no temporal anchorage on the home page → temporal anchorage on the arti-

cle page
2. temporal anchorage on the home page → temporal anchorage on the article
page
The most common situation is that the newspaper home page includes only the
headline and the lead of the news story, without any time line or attribution. That
information is specified only on the article page, cf.:
‘Fat finger’ causes Mumbai market crash (H1 in article preview on home
page)
↓
‘Fat finger’ causes Mumbai market crash (H2 article web page)
Robin Pagnamentan
Last updated at 1:13PM, October 5 2012
When the time of publication is specified on both textual levels, there are several
possibilities. The following headline is from an article preview on the home page
of The Times from 5 October 2012:
Cole in four-letter rant at FA after Terry ruling
Updated 43 minutes ago
When the hyperlink to the full text of the news text is followed to the article web
page, exactly the same temporal indication is found to be repeated within the
78 News and Time
article’s attribution line. The deictic reference indicates that this is a live news
story potentially subject to further development and coverage:
Cole in four-letter rant at FA after Terry ruling
Ashling O’Connor Sports News Correspondent
Updated 43 minutes ago
Once the news article becomes final and does not undergo any further changes,
the deictic temporal reference relating to the time of reception is changed into
calendrical means, enabling the article to be placed in the newspaper’s archives.
Thus, the same news text has another possible format of temporal reference, cf. the
headline and the attribution line taken from the same text several months later:
Ashley Cole in four-letter rant at FA after he is criticised in Terry
ruling
Ashling O’Connor Sports News Correspondent
Last updated at 3:30PM, October 5 2012
The temporal indicators thus provide anchorage for three possible varieties of the
same news story: the textual preview on the home page (Updated 43 minutes ago),
the current news story on the article web page (Updated 43 minutes ago), and the
same, archived news story on the same article web page later on (Last updated at
3:30PM, October 5 2012).
As mentioned previously, the presence of the temporal indicator on the level
of the paper’s home page is not the rule for all news texts, most likely because all
articles on the home page are presumed to be current news. This is the result of
the general assumption of the recency of news texts. The above example is taken
from the ‘Latest news’ section of The Times, which tends to include three or four
news items accompanied with visuals. Other sections on the home page of the
newspaper (‘UK News’, ‘More UK News’, ‘Business’, and ‘World News’) do not
provide any temporal anchorage. A specification of time is then provided only on
the articles’ own pages and is performed through calendrical means.
On the article web pages, all articles are consistently bylined with a speci-
fication of the time of publication. The screenshots in Figure 4.9 document the
patterns used by the individual papers.
By indicating the calendar date, the articles locate the news stories in ob-
jective time. In some cases, the metadata are supplemented with a note on the
last update of the text (Last updated at 12.01AM), which provides an additional
anchorage since the phrase operates with the presumption that the time of the
day applies to the date mentioned on the same line. There is a general preference
for the objective method of framing the text rather than relating it to the reader’s
The Times
The Independent
The Telegraph
The Guardian
Figure 4.9 Temporal anchorage and attributions in article bylines
receiving time. The latter, attested in phrases such as Updated 43 minutes ago, are
used less frequently. The following are sample formulations of mixed temporal
anchorage that use both calendrical time and deictic reference to the individual
reader’s receiving time:
4 October 2012 | Last updated less than one minute ago (The Guardian)
Last updated 1 minute ago | Thursday 04 October 2012 (The Telegraph)
The page then provides the objective time (i.e. the date, possibly in conjunction
with the day of the week) as well as the subjective time measured from the mo-
ment the individual reader accesses the web page. In the latter case, the temporal
deictic centre is located with the reader and the moment of accessing the page
constitutes the benchmark from which the ‘freshness’ of the text is measured (cf.
1 minute ago). The advantage of this way of indicating temporality is that the read-
er can immediately see that the news site is dynamic, with readers having access to
content that is entirely up-to-date with respect to their receiving time.
4.6 Hypertextuality and temporal mapping in online articles
Probably the most interesting phenomenon in online news is the existence of in-
ternal hyperlinks that connect a given article to other content within the newspa-
per’s own web site. In the case of topics and issues that are in the media’s focus of
attention for a prolonged period of time, the editorial office also tends to provide
links to previous articles from within the newspaper’s history. Hyperlinks can be
used to reconstruct a network of relations between individual articles, thus ena-
bling the reader to backtrack a particular issue over months or even years.
80 News and Time
Figure 4.10 illustrates a very complex network for a set of 64 news articles on
a single topic published in an online newspaper over the course of ten months,
as analysed in Chovanec (2000). It is a temporal mapping of all texts published in
the online version of the British daily newspaper The Telegraph. The first article in
the network (7.2.97) is presented in the top left-hand corner. Subsequent articles
are placed horizontally, snaking their way through the chart from left to right and
from right to left, with the last pair of articles in the bottom left-hand corner (both
appearing on 13.11). Multiple articles on a single day are shown as separate circles
and indicated with numbers (number 1 always refers to the main article). While
this pattern of hyperlinking, giving rise to a chain of interlinked texts, also char-
acterizes modern online news, there is a difference in the manner of presentation
of archived content in the current data: the archived web page is no longer a mere
‘snapshot of the past’ but has live content as well, interlinking the archived text
with more recent and up-to-date news on related issues (see below).
The interlinking of multiple articles into such a hypertextual network dimin-
ishes, in a sense, the traditional position of the newspaper article as a relatively
stand-alone unit. In printed newspapers, the individual articles are relatively in-
dependent. Sometimes, clusters are formed when several articles, usually on the
same page, address several different aspects of the same issue, e.g. providing com-
ments from news actors, interviews with authorities, background information on
similar events, etc. Ungerer (1999: 20) notes that articles in print newspapers form
text networks and text ensembles that – when diagrammed – can reveal some
unexpected relations, but his analysis is limited to such a self-contained unit as
a specific newspaper issue. In printed newspapers, the articles will be read only
within the context of each other and within a particular issue of the paper, not
with respect to previous news reports on the same story – simply because the
textual history is not physically available.
The situation is different in online news. Although it is theoretically possible
for a reader to backtrack an issue in past copies of printed newspapers (however
rare that practice may be), it is the technological affordances of online media and
hypertext that promote the construction of hypertextual networks along the dia-
chronic axis and increasingly allow the readers to access past, i.e. non-recent news
texts. In printed newspapers, it is the reader who navigates between the individ-
ual articles on the newspapers’ pages, and does so without following a discursive
path pre-defined by the editors. With respect to the diachronic hypertextual net-
work of articles shown in Figure 4.10, it is worth noting that while some articles
turn out to be quite central in the diachronic chain (14.2, 5.3, 31.10/1), others are
rather isolated since they are referred to by only one other news text (6.3) or none
at all (6.5, 1.11/4,5,6,8,9,10 – i.e., accompanying articles to the main news text).
7.2.97 11.2./1 12.2. 14.2. 16.2. 22.2./1

(5.10.96)
11.2./2
(11.2.97)
4.4. 13.3. 8.3 6.3. 5.3. 22.2./2
4.10.
6.5. 15.7. 19.7. 2.10. 3.10.
7.10
31.10./1 29.10. 28.10. 17.10. 13.10. 11.10

2x
1.11./3
31.10./2 1.11./1 1.11./2 1.11./4 2.11./1 2.11./2
1.11./5
1.11./6
1.11./7
1.11./8
1.11./9
1.11./10 2x
3x
(1.5.96)
7x
5.11./3 5.11./2 5.11./1 4.11./1 4.11./1 3.11.
5x (8.11./2) 4.11./3 5x
5x
6.11. 8.11. 9.11. 10.11./3 10.11./2 10.11./3
5x 2x 12.11./2 11.11./2
13.11./1 12.11./3 12.11./3 11.11./3 11.11./1
3x 12.11./4 (poll) 11.11./4 (6.2.97)

(4.10.95)
2x
13.11./2 12.11./5 5x 11x
12.11./6
Each cell corresponds to one article

Internal link A two-way internal link
External link
Figure 4.10 The temporal map of a single topic. A chain of hyperlinked articles forming
a dense intertextual network (Electronic Telegraph, February–November 1997, articles on
the case of Louise Woodward; Chovanec 2000: 271)
82 News and Time
In the current data, we can see some interesting developments in the hyper-
textual organization of related articles, particularly as regards the inclusion of
some live content on web pages displaying archived news content. After accessing
an article in the newspaper’s archive, the reader will not only be able to backtrack
a given issue through the hypertextual network (such as in Figure 4.10 above)
but also access some of the latest articles on the issue. The Guardian, for instance,
includes a special section entitled ‘Related’, which provides current links related
to the topic of the archived article.
It is one of the defining properties of hypertext that it is made up of various
textual and non-textual elements that are integrated into a single whole. As a re-
sult, hypertexts frequently do not have clear limits. As observed by Jucker,
A link in an electronic text does not so much refer the reader to another text, but
it integrates the other text into the original document. The boundaries between
one text and all the other texts it is linked with become fuzzy. The integration of
visual, auditory and graphic information into the verbal texts turns the hypertext
into a multi-modal hypertext or hypermedia. (Jucker 2002: 35)
A schematic analysis of the intertextual network made up of all interconnected

articles on a given issue indicates that there are two dimensions for classifying
internal links to other content supplied by a given news site. The links can be
classified on the basis of temporality and directionality (mutuality). While the
temporal orientation of the link leads to the distinction between synchronic vs.
diachronic links, the dimension of mutuality concerns the direction of the hyper-
link, giving rise to one-way and two-way links (see Figure 4.11 below). As a result,
there are several possible ways in which any two articles interlinked within the
temporal map can be related:
– synchronic bi-directional linking

– synchronic uni-directional linking
– diachronic uni-directional linking
Synchronic linking concerns articles published on the same day (or in the same
electronic ‘issue’ of news sites with non-daily periodicities) and typically concerns
clusters of articles on the same topic. Most often, such articles are mutually hyper-
linked, cross-referencing each other. The nature of their internal link is thus bi-di-
rectional. At the same time, articles can provide a one-way link to related content
from the same day, typically where such texts provide merely supplementary,
tangential information. Diachronic linking concerns a reference back in time to
some previous news article; such linking can be only uni-directional because the
earlier text cannot anticipate, at the time of its creation and eventual transfer to
the inactive texts contained in the newspaper’s online archives, any future text
that might refer to it.
At the same time, however, it needs to be acknowledged that the distinction
between synchronic and diachronic links in online news is not absolute. Mutually
linked articles on the same topic are typically published throughout the day. They
are, in fact, diachronically arranged, although they may be presented within one
and the same article cluster, which is viewed here, for the sake of simplicity, as a
batch of texts that are synchonically linked. In terms of the pragmatics of news
production, this raises a number of interesting questions, such as what effect the
traditional publication frequency of print news (daily) has on online news, and
at what point an article or an issue becomes archived rather than further devel-
oped by gradual updates (cf. Figure 4.6 above) and additional articles within a
cluster. In other words, it remains to be seen when an online news text turns
into a ‘finished product’ and, subsequently, when it loses its ‘current relevance’. It
appears that online news may not be totally independent of the traditional daily
publication cycle, if only because the linearity and continuity of time in the online
environment is punctuated by the realities of human life, where ‘every morning
marks a new day’.
The relationship between articles in terms of synchronic and diachronic link-
ing is schematically represented in Figure 4.11. The grouping of synchronically
related texts, indicated by means of the dotted ellipsis, represents their co-pres-
entation, typically in an article cluster. Technically speaking, the individual ar-
ticles within the cluster may be added at somewhat different times; but, since
readers are typically confronted with them as a batch of texts placed side by side,
they are treated as contemporaneous in the figure. While the scheme represents
the relations between related articles in online news, the same conceptualization
is actually applicable for the intertextual links between articles in a current issue
of a printed newspaper (i.e., a synchronic relationship coded through vertical ar-
rows) and articles in past issues (i.e., a diachronic relationship coded through
backward-pointing horizontal arrows).
The fact that diachronic internal links pointing to previous articles on a top-
ic are, by necessity, uni-directional and that past articles cannot anticipate and
link to future texts has a significant consequence for the way readers can navi-
gate through the entire intertextual network. Within the intertextual network of
articles, it is thus possible to move backwards but not forwards. When moving
from link to link, a reader is transported back in time, exploring the diachrony
of a given topic and following an intertextual path signposted by the newspaper’s
editorial staff. The network remains potentially open because other texts can be
added to it if the editors decide to link current texts to any of the articles within
84 News and Time
Earlier text Article cluster
Text 2
1
3
Text 1 Text 3
Text 4
Past issue(s) Current issue

TIME AXIS
Types of links (represented by arrows):
1 – synchronic bi-directional linking
2 – synchronic uni-directional linking
3 – diachronic uni-directional linking
Figure 4.11 Temporal orientation and directionality of internal links
the hypertextual network. This is related to the nature of uni-directional links: the
number of such links that emanate from a given article is obvious since the links
form a part of the article’s overall structure; the other articles that refer to a given
article cannot, however, be easily identified since there is no physical indicator of
the fact that a text is referred to by some other text.
Last but not least, not all articles within the diachronically organized chain
of hyperlinked texts are of equal importance. As argued in Chovanec (2000: 12),
three distinct types can be distinguished:
1. ‘hub’ articles – these radiate explicit links to a number of other articles;

2. ‘key’ articles – these achieve their intertextual prominence as a result of being
frequently referred to through internal links in subsequent articles; and
3. ‘pendant’ articles – these are attached to other texts (usually by means of a
two-way hyperlink) but they do not typically feature much in the intertextual/
hypertextual structure.
The distinction between the three types can be schematically represented in Fig-
ure 4.12. While hub articles cross-refer a great deal, key articles tend to be referred
to. Pendant articles are comparatively much less prominent, being relatively mar-
ginal within the hypertextual network. They cross-refer and are referred to only
• refers to 8 other texts

hub article
• referred to by 6 articles
• refers to 4 other texts

key article
• referred to by 8 articles
• refers to no other text

pendant article • referred to by 1 article
Figure 4.12 Online article types according to the nature of connecting hyperlinks
(vertical arrows indicate synchronic links to/from other current articles; horizontal
arrows indicate diachronic links either to past articles or from future articles)
occasionally because they provide information that is merely tangential to some

aspect of the main event.
Differences can also be expressed in terms of the article’s centrality with re-
spect to the hypertextual network connecting a large number of articles: while a
hub article is central immediately upon publication, a key article achieves central-
ity only as a result of subsequent links. A pendant article is non-central: it stands
isolated within the hypertextual network, being appended to some other, more
important article.
While the relative number of hyperlinks attached to a particular article will
immediately enable a reader familiar with a given news site to identify ‘hub’ arti-
cles, ‘key’ articles will emerge as central only with the passage of time as a result
of a significant number of other articles linking to them. It is not unusual for
hyperlinks to bypass some less important articles within the intertextual chain
(although that may also be the strategy of a given paper, based on the decision to
foreground some stories at the expense of others; the bypassing of some articles
in the chain is evident in Figure 4.10 above). By contrast, ‘pendant’ articles are
relatively marginal, even though they can become ‘key’ articles later on when the
focus shifts to a side issue and the relevant article starts to be increasingly hyper-
linked. If that happens, then they function, viewed retrospectively, as ‘dormant’
key articles.
86 News and Time
4.7 Temporal deixis and internal hyperlinks
Temporality plays a crucial role in hypertextual structuration since it provides

the guiding principle for organizing the entire network into a linearly sequenced
chain of articles. The connections between the articles are realized through inter-
nal hyperlinks that are either attached to or incorporated into the relevant news
texts. The internal links frequently open with a calendrical time expression, which
locates the hyperlinked article from the newspaper archive on a fixed point along
the time axis. The time stamp serves as an explicit indicator that – upon following
the hyperlink – the deictic centre will change and, consequently, that such deictic
temporal expressions as ‘yesterday’ will have to be re-anchored to the specific
calendrical time expression provided within the hyperlink.
Needless to say, not all online media furnish the reader with such explicit
re-anchorage markers: sometimes they merely provide the link and leave it up
to the reader to work out whether the hyperlinked text is stale news retrieved
from the newspaper’s archive or a currently topical accompanying article. This
situation occurs, for instance, when sentences/words are hyperlinked directly in
the text or when the headline of the hyperlinked article, without any time stamp,
is included under the text.
Where no temporal anchorage point is present in the hyperlinked article’s
byline (or if the reader fails to take it into account), the text may be processed
by the reader under the mistaken presumption of recency. The moment the true
deictic framing of the hyperlinked text is understood as being located in the past,
i.e. as non-recent, the reader may experience a feeling of ‘being tricked’ since he
or she needs to reassess the assumption of the text’s recency on which its process-
ing was initially based. No such surprise effect should result if the calendrical time
expression is included in the hyperlink.
A diachronic perspective reveals some interesting developments over the past
15 years. In the 1997 corpus of data from the Electronic Telegraph, there were in-
ternal links of three distinct types:
1. Internal links to accompanying articles within an article cluster

(see Figure 4.13)
2. Internal links arising from news content being hyperlinked through
selected words and phrases in the text (see Figure 4.13)
3. Internal links to past articles from the newspaper’s online archive
(see Figure 4.14).
There is no time stamp indicating the currency status of the accompanying arti-
cles (type 1) or the news texts hyperlinked through underlined phrases (type 2).
The latter are particularly interesting because they take the readers to past articles
Figure 4.13 Placement and pattern of internal links in the 1997 data: absence of tempo-
ral indication for accompanying articles and text-internal links (Woodward’s
lawyers ask for manslaughter sentence, Electronic Telegraph, www.telegraph.co.uk,
4 November 1997)
Figure 4.14 Placement and pattern of internal links in the 1997 data: calendrical date +
headline. For a schematic representation of links, see Figure 4.10
88 News and Time
in the newspaper’s archive without signposting the temporal shift in any way –
unlike the third type of internal links which are date-stamped, cf. Figure 4.14.
The 1997 data also indicate the consistent placement of the last type of link at
the very ends of articles, with the body copy presented as a single uninterrupted
mass of text. It is with this type of internal link that we find the pattern date +
headline, which anchors the hyperlinked story firmly to a specific day in the past,
cf. Figure 4.14. Interestingly enough, the bottom of the page used to provide a
link to another – unrelated – story published in the newspaper on the same day
(Next report: Paedophile suspects commit suicide in the next figure). This feature
shows that online editors used to set a clear reading path for the readers. The latter
could progress through the current issue in a linear manner, moving from one
story to another and from one page to another. This seems to indicate that while
the 1990s versions of online newspapers experimented with some of the technical
affordances of the Internet, they were still little more than just electronic copies of
the print versions of the major papers.
The 2012–2013 data from The Telegraph, however, show a markedly differ-
ent situation. Not only has body copy become fragmented with the insertion of
visuals, highlighted quotes, unincorporated textual segments, external links and
advertisements, but there is also a difference in the placement of the internal links
to previous articles: the links no longer appear exclusively at the ends of articles
(cf. Figure 4.15). Hyperlinks to previous articles on a given topic are provided in a
shaded box titled ‘Related Articles’ that is placed within the body copy. The links
have also changed their structural pattern: the former pattern of ‘Date: Headline’
(reminiscent of the common two-part headline structure of Topic: Comment)
is reversed into the combination of Headline + Date, with the two components
typographically distinguished by means of font size and colour.
While the 1997 data illustrate a period when online newspapers were basi-
cally electronic versions of the printed issues (mutually related in the form of a
hierarchical dependence), the later data come from a period when online news
had already established itself as a distinct genre of news that is independent of
printed newspapers. Nowadays, some news offices keep separate desks for the
online and the print editions (cf. Le 2012), which helps to better target their au-
diences. The shift from the earlier conception of online news presentation to the
modern practice of online journalism seems to have occurred in the first half of
the 2000s. Knox (2009), for instance, documents a similar shift in the design of an
online newspaper over the course of a few years by illustrating the changing role
of verbo-visual thumbnail images.
All this is indicative of the quick adoption by the media of the novel online
environment and utilizing its potential for not merely conveying news content
but also relating to readers by encouraging reader participation through online
forums and fashioning customized web page content, e.g. as regards the spon-
sored content appearing alongside the editorial content. As pointed out by Hall
(2008: 209), many newspapers were also quick to realize the market value of their
archives and started making them available for searches, even though they were
not (yet) hyperlinked with the current news content (the New York Times, for
instance, nowadays appends links to each published article to related content in a
collection of its online materials going back to 1996; prior to this date, the online
archive stretching back to 1851 can be accessed for a fee).
Extensive hypertextuality and multimodal storytelling have become the rule
in modern news. Figure 4.15 also illustrates the disruption of the body copy of the
news text by the insertion of a special section with internal hyperlinks to previous
Figure 4.15 Placement and pattern of internal links in the 2012–2013 data
(British climber says ‘dream come true’ to scale Antarctica’s Ulvetanna;
The Telegraph; 31 January 2013; article retrieved on 31 January 2013)
90 News and Time
articles (as well as a date stamped photo gallery and a video) on related topics.
Further disruption of the body copy results from the placement of the photograph
immediately after the ‘Related Articles’ section.14
This particular article is characterized by an unusual degree of fragmentation
of the body copy. The news text on the article web page is composed of the follow-
ing sequence of elements: headline; subhead; photograph; photograph caption;
attribution; 2 paragraphs of body copy (lead + first paragraph); photograph; pho-
tograph caption; 2 paragraphs of body copy; related articles section; photograph;
photograph caption; 2 paragraphs of body copy; photograph (with attribution but
no caption); 4 one-sentence paragraphs of body copy; photograph (with attribu-
tion but no caption). The segmentation of the text and its juxtaposition with visual
elements is such that the text resembles a magazine article rather than a more
traditional piece of news. It attests to the extent multimodality is used in modern
online media and the trend towards supplying news content in increasingly small
textual segments, even though they may constitute a part of a larger whole.
Beyond the text of the actual article, the article web page contains hyperlinks
to other news content. Figure 4.16 reproduces a screenshot from the article ana-
lyzed above, showing the textual block placed under the article. There are two
types of links here: internal links to other stories from The Telegraph and exter-
nal links that take the reader beyond the online newspaper’s site. Although the
screenshot was taken several weeks after the article’s publication, the links are up-
dated: the newspaper automatically generates up-to-date headlines with links to
articles relevant at the time the archived page is accessed, which holds also for the
external links on the right (‘More from the web’).15 The hyperlinked articles are
not related in any way to the static, archived text of the article and they are offered
to the readers as a motley mix of mostly recent articles but also news content that
can be several months old.
As suggested above, there is a clear difference in the way archived articles
were presented in 1997 and in 2013. The early versions were entirely static texts –
perfect snapshots of the articles from the day of publication. The current practice
is much more dynamic – the archived, static text is embedded within the rest of
the web page that contains more up-to-date and live content on the website. The
modern approach that mixes static article content with such a dynamic system
of surrounding links means a different reading experience: one that is much less
historical and that realizes a certain compression of time between the original
14. The full text of the article is available at: <http://www.telegraph.co.uk/news/worldnews/

antarctica/9839097/British-climber-says-dream-come-true-to-scale-Antarcticas-Ulvetanna.
html>.
15. Whenever the article web page is refreshed, new content for this section is generated.
Figure 4.16 Live content on an archived article page in online newspapers (British
climber says ‘dream come true’ to scale Antarctica’s Ulvetanna; The Telegraph;
31 January 2013, retrieved on 22 February 2013)
article and the reader’s processing of it. It is, also, a much more individualized ex-
perience: in the infancy of online newspapers, all readers accessing archived texts
retrieved the same web page; nowadays, each reader processes the same archived
text within a different textual environment that is generated by sophisticated al-
gorithms taking into account the reader’s online history and the time of retrieval.
The other newspapers operate a similar system of links, although they do not
incorporate the internal links to related articles on the same topic (i.e. those that
form the intertextual network) within the body copy of the article as The Tele-
graph does. The internal links are usually listed in a separate column alongside the
body copy of the article, so no fragmentation of the text results. This is the case
with The Guardian, for instance, which also has a section at the ends of articles
that is almost identical to that found in The Telegraph: it is titled ‘More from the
Guardian’ and ‘More from around the web’.16
The different treatment of internal links is illustrative of the development
of online news as a distinct genre. While in the late 1990s, online news essen-
tially existed as merely electronic versions of printed content and drew modest-
ly and tentatively on the new possibilities offered by hypertext, a decade later
16. The content in the right-hand column consists, in both newspapers, of paid-for links pro-
vided by a special provider of this service (Outbrain). As an explanatory note on the Guardian’s
web page mentions, the links “may or may not be relevant to the other content on this page”.
The function of these links is described as “driving traffic to [one’s] content”.
92 News and Time
online news has become a distinct form that uses the multimodal and hypertex-
tual potential of the new media to the fullest extent. The text of the article, which
was previously presented as a textual monolith, becomes fragmented with the
insertion of hyperlinked content and images, sometimes even to the extent that it
becomes difficult to read the primary verbal coverage without much interruption.
Hypertextuality in online news has a number of repercussions. On the textual
level, online news articles are characterized by fragmented textuality because the
linear flow of text tends to be interrupted with relevant links, photographic and
video material, and advertisements, etc. In some cases, the degree of disruption
may be such that the textual component loses some of its primacy and recedes
into the background. As regards intertextuality, hypertextual links give rise to
a highly complex network of texts through which the reader can navigate. The
intertextual network can then be represented by temporal mapping, since – by
moving from one text to another – the reader moves back and forth along the
time line as well.
Hypertextual analysis of online news also needs to distinguish between in-
ternal links and external links. Although the latter may be the editors’ selections
leading to additional newsworthy content, sources of the news, etc., the links to
further informational content may also have an economic basis, being sponsored
and placed by advertisers. Online newspapers are treading a thin line between
promoting internal links and including (or tolerating) external links because of
their primary concern for readership. While internal links secure the readers’
further attention to the media site, external links may result in a loss of control
over the reader, who may follow external links and not return to the website of
the online newspaper. It has been observed that some media tend not to favour
the use of external links on their pages, preferring to perform a gatekeeping role
instead. Other media may invite audience participation and interaction instead,
particularly in the form of allowing readers to comment on news stories (Redden
and Witschge 2010: 177, 183). Where the paper’s source of income is linked to
the revenue generated from mediating access to external sites, the selection and
presentation of news may also be geared – at least partly – towards directing the
readers to external links and guiding them to the advertisers’ sites.
Last but not least, the online environment may also work counter to the con-
ception of the news story as a stand-alone unit expressed in a single article or a
cluster of several related news texts. As Lewis observes,
There is less pressure in hypertext to identify discrete news ‘events’. News ele-
ments are embedded in and linked to wider content. A summary outline of one
news item can simultaneously be a detail of another. A news topic is no longer de-
veloped in a series of static texts emitted at regular intervals with implicit links to
other texts. It is developed as a cluster of dynamic, related, hierarchically-struc-

tured texts, like overlapping groups of concentric circles. (Lewis 2003: 97)
Individual news stories may then relinquish their independence and exist as a part
of the entire network of texts that cover a given issue in its globality, often giving
the reader a chance to follow the content in various directions. Lewis (2003: 97)
argues that this is a novel news genre that can be described as “a theme-based
group of news objects held together graphically, overlapping with other such
groups, and undergoing progressive updating”. The home page then becomes a
complex sign in which various multi-modal components interact in a coherent
way (Knox 2007: 23).
Part II
Textual rhetoric of headlines

Chapter 5
Temporal deixis in headlines
The first analytical part of this book, which seeks to explore the issue of tem-
poral deixis in news stories, focuses on temporality in headlines. This particu-
lar structural element deserves an independent treatment because it is a textual
macro-structure that opens the entire news story and is the culmination of the
news-editing process. The debate includes temporal deixis in headlines in gen-
eral terms (Chapter 5), the present tense in headlines (Chapter 6), other tenses
in headlines (Chapter 7), and the ellipsis of auxiliaries in headlines (Chapter 8).
The second analytical part builds on the four chapters by considering the role
of temporal deixis beyond headlines. Thus, it traces the use of tenses and adver-
bials of time in leads (subheads) and the body copy (Chapter 9) and proceeds to
offer a functional interpretation of the three main textual segments of news sto-
ries (Chapter 10). It argues that the expression of temporal deixis through more or
less conventional means is motivated by the media’s attempt to construct a shared
discursive space with their audiences.
While deixis is understood as a semantic-pragmatic phenomenon, the anal-
ysis tackles the grammatical forms through which it is expressed, mainly verbal
tenses. The interpretation of the whole situation, particularly with respect to the
conventional shift of tenses in headlines and the triple tense pattern found in
news stories, is cast from the perspective of functional linguistics. The grammati-
cal realization of the underlying pragmatic concern for the news media’s audience
is motivated by interpersonal considerations, i.e. the media’s attempt to attend to
the interpersonal metafunction.
5.1 Material for analysis
The data discussed in this chapter are based on a set of newspaper articles re-
trieved from the online versions of several British broadsheet daily newspapers
(The Telegraph, The Guardian, The Independent, and The Times) from autumn
2012 till spring 2013. The core sample consists of 837 articles published from Sep-
tember to December 2012 and includes texts that appeared on the newspapers’
home pages. Because of this particular focus, articles published in more special-
ized sections accessible through tabs on the home pages were not considered in
98 News and Time
order to limit the number of articles processed to a manageable amount. The ma-
terial for the corpus was drawn from hard news items appearing in different sec-
tions of the home pages and classified mostly as domestic (UK), international and
business news. This method of data collection aimed at achieving compactness of
the data rather than all-inclusiveness, since the focus of the analysis in this sec-
tion is on the interpretation of typical temporal patterns in hard news headlines
rather than on revealing all the potential formal variations. For this reason, news
on culture, lifestyle articles and other sections available on the newspapers’ home
pages were intentionally left out in order to ensure a relatively homogenous sam-
ple of headlines in their most canonical form. Since soft news (as well as articles
from the popular press) tends to have a more personal and narrative form, their
inclusion would inevitably have skewed the sample. By contrast, selected articles
from the sports sections of the newspaper were included, as long as they provided
information rather than opinion.
The articles were manually downloaded from the Internet and subjected to
quantitative and qualitative analyses. In the case of article clusters, only the first,
arguably the most important, article was included in the sample, with the satel-
lites left out of the corpus. The reason for their omission was that the additional
articles on a given story – as opposed to the main article – tend not to include a
subhead/lead on the home page of the online news site. Moreover, their headlines
are not independent, stand-alone units because they are related to the main news
item. It was deemed that the hierarchical dependence, which makes it possible to
distinguish between first-order and second-order headlines, could affect some of
the results of the quantitative analysis, yet it does have its place in the qualitative
analysis of the data.
At the same time, the study draws, for some of its conclusions, on two smaller
sets of data: a corpus of 250 hard news articles on diverse topics from the on-
line paper The Telegraph from 2002, and a set of 64 thematically related articles
retrieved from the same source in 1997 that were used for initial explorations
of temporal deixis in and beyond headlines in several earlier studies (Chovanec
2000, 2003a, 2003b, 2005b). The earlier data are useful in that they were compiled
specifically in order to include clusters of articles published on a single topic on
the same day and thus complement the current sample focused on the main arti-
cles only. As mentioned above, the earlier data sets led to the finding that the clus-
tering affects the form of headlines because the main article inevitably provides a
contextual background within which the subsequent headlines of satellite articles
and background stories are framed. The main article serves as a pivot to which the
other texts are connected. An article cluster can thus be an island of news texts
among which some information can be presupposed more easily.
Chapter 5. Temporal deixis in headlines 99
5.2 Headlines and the expression of time
News headlines are textual macrostructures that have numerous functions. Tra-
ditionally, they have been seen as having semantic and pragmatic functions (van
Dijk 1988; Bell 1991), although such a conception is not without its problems
(Dor 2003; Ifantidou 2009). Headlines form a part of the rhetoric of news texts,
typically offering a framing and potentially evaluative perspective on the news
(van Dijk 1988: 143; Bednarek and Caple 2012: 100; Molek-Kozakowska 2013)
and maximising the news value (Harcup and O’Neill 2001). They can be ap-
proached in terms of the narrative structure of the news stories they abstract (Bell
1998), which underlies their semantic (summarizing) function. Narrative analysis
reveals the non-chronological organization of news texts (Bell 1991) and the dis-
crepancy between the event structure and the narrative structure (Schokkenbroek
1999: 67) – issues that are discussed in an independent section.
The pragmatic function of headlines to relate to the readers means that head-
lines are used not only ideationally, i.e. to provide a summary or highlight some
detail of the story, but also interpersonally. Contact with the readers can be es-
tablished, for instance, through the choice of expressive lexis (Schaffer 1995), the
simulation of immediacy through various forms of reported speech, word play and
humour (Chovanec 2005a, 2008), and – more subtly – a reliance on background
knowledge and assumptions that form a shared context between the newspaper
and the recipients.
The expression of temporality in headlines is among such phenomena that
may be interpreted as being interpersonally-oriented. This concerns, above all,
(a) the presumption of recency manifested through the absence of explicit tempo-
ral indicators, such as adverbials of time; and (b) the conventional shift of tenses.
Other characteristic features of headlines, e.g. the ellipsis of auxiliaries, are in-
volved in this process as well.
5.3 Adverbials of time in headlines
Time adverbials are quite infrequent in headlines. An analysis of the 837 headlines
in the data set yielded only 19 time adverbials (2.3 per cent). Out of this number,
there are only a few examples when the adverbial specifies the time of the main
event, such as BBC executives to discover their fate today (The Telegraph;
19.12.2012). The vast majority of the adverbials are prepositional phrases that
add some newsworthy meaning, e.g. by specifying periods of time for long-term
trends (since Job’s death; for the first time in 18 years; in last decade; until 2015).
That is also the case in the following example:
100 News and Time
Hollywood hacker who leaked nude Scarlett Johansson photos

jailed for ten years (The Independent; 18.12.2012)
Here, the adverbial phrase for ten years relates to the length of the sentence. It
does not specify the time when the main event actually occurred (i.e., when the
sentence was issued ‘jailing’ the hacker).
The rare occurrence of time adverbials means that the conceptual category
of time in headlines is indicated almost exclusively by means of the grammatical
category of verbal tense, at least as long as the relevant headline contains a finite
verb form. However, tense is not provided explicitly and thus is not used deicti-
cally in the majority of headlines; the temporal placement of a given news story
has to be inferred by the reader. Time is most typically present in headlines as an
implicit category.
There are two issues relating to time adverbials in headlines. First, we need to
account for their absence from headlines, which appears to be the norm. The sec-
ond – and equally interesting – point concerns those rare situations when adver-
bials of time actually do occur in headlines. In these less common cases, it appears
that time adverbials actually do not provide the temporal anchorage of the event
reported; instead, they are semantically crucial for the proposition articulated in
the headline.
Absence of adverbials of time from headlines
There appear to be several reasons why adverbials of time tend to be absent

from headlines. Given the traditional spatially-constrained format of this text
type, which may motivate its condensed and telegraphic style (cf. Quirk et al.
1985: 845), one can identify several pragmatic, structural and grammatical moti-
vations for the lack of time adverbials.
On the pragmatic level, there is the general assumption of a shared temporal
framework that news production operates with. As a result, the expression of time
in headlines is redundant. Since newspapers have a very limited shelf life, the as-
sumption of recency is very strong and applies automatically to all texts contained
within a given issue. The explicit temporal anchorage for the stories is provided
in the masthead of the newspaper, although this method becomes meaningful
only when the readers check whether they are actually reading a current issue of
the paper. Otherwise, the texts are processed on the assumption that the stories
reported happened within a period of time that corresponds to the publication
frequency of the paper, i.e. ‘yesterday’ (or ‘since the last issue’).
From the structural perspective, the typical headline is a textual superstruc-
ture that provides an abstract of the news story – it functions as an initial summary
(van Dijk 1988: 53; Bell 1991: 150). However, it is a second-level summary since it
is formulated on the basis of the lead paragraph, which provides all the important
facts about the event (who, what, where, when, how). Since the headline is (in the
canonical situation) an abstract of an abstract, some information may have to be
sacrificed in the interest of formulating brief and succinct headlines that focus on
the main event or some other aspect of the story (cf. Ungerer 2000). If the story
needs to be abstracted to a minimal chunk of information, then it is more than
likely that temporal specification will not be included. The most important aspects
of the news story involve the news actors and the event (who and what) rather than
the temporal and – to a lesser degree – also the spatial setting (when, where). As
van Dijk (1988: 150) notes about the cognitive structure of headlines, “the initial
description of a setting is not part of an action sequence”.
The expression of time through adverbials of time is also redundant from the
point of view of sentence structure. Adverbials complement the meaning of other
elements contained in the clause and are merely optional in the basic grammatical
structure of the clause. None of the grammatical patterns of clauses in English
contains adverbials; the obligatory sentence constituents in the syntactic struc-
ture of the clause are subject and verb (S+V) for the simplest sentence pattern,
and direct/indirect object and complement for the other sentence patterns. The
grammatical nature of the predication therefore indicates that all adverbials, in-
cluding adverbials of time, are dispensable elements within the clause structure.
Their minimal presence in the spatially-constrained headline is little surprising.
Bell (1991) argues that the minimal headline has the structure of partici-
pant + event. With clausal headlines, this translates into the basic sentence struc-
ture of S+V. Non-clausal headlines can combine the two elements differently,
e.g. through the possessive construction or the colon structure. However, there
are also headlines that do not specify any participant because they focus sole-
ly on the event. While such headlines are not attested in the data set, they can
occasionally be found in the popular papers.17 The existence of some one-word
headlines is also enabled through their close link with accompanying visuals, i.e.
multimodality. Photographs often provide visual anchorage for the verbal com-
ponent, e.g. by identifying the main participant (or even the event itself), with the
17. Examples include –15° (Daily Mirror, 28 February 2006), Dough! (Daily Mirror, 24 Febru-
ary 2006), Expendable (Daily Mirror, 24 August 2003), Found (Daily Mirror, 4 March 2003),
Charged (Daily Mirror, 19 June 2003), etc. Cf. also one-word headlines that form a novel or wit-
ty combination of several elements, e.g. Hoondunnit (Daily Mirror, 22 July 2003), Heroo! (Daily
Mirror, 22 June 2004) and also the classic headline Gotcha! (The Sun, 4 May 1982; cf. Chovanec
2008). Similar minimalist headlines can occur within article clusters, where previous articles
establish a local shared context, within which the second-order headlines of accompanying
articles can presuppose information contained in the main, first-order headline/article.
102 News and Time
one-word headline encoding some other element of the story (as well the paper’s
standpoint).
Yet another explanation for the absence of adverbials of time can be found in
some of the pragmatic repercussions of headline rhetoric. Headlines refer to past-
time events by means of the conventional present tense, in what is a non-deictic
use of verbal tense. A non-deictic tense refers to a different time zone than when
the same tense is used deictically. In the event that a non-deictic use of tense (e.g.
the present tense encoding past-time events) combines with a definite adverbial
of time localizing the event into a specific moment in time (e.g. into the past), a
possible contradiction between the two would result. This may be the reason why
the combination of a non-deictic tense and a deictic time adverbial is avoided in
a single sentence. Where both are deictic, the verbal tense identifies the time zone
in a general manner and an adverbial of time provides a more precise temporal
specification.
Nevertheless, Huddleston and Pullum (2002: 126) admit the possibility of
the co-occurrence of a non-deictic tense and a deictic time adverbial in a single
sentence, a combination that would normally appear as incongruent. When dis-
cussing photograph captions – i.e. a text type that comes close to news headlines –
they note that “photographs and drawings can give permanence to what would
otherwise be a transient historical occurrence, and captions then use the present
tense”. As one of the examples to illustrate the non-deictic tense, they give the
following caption: Aboriginal protesters occupy part of the old Parliament House in
Canberra yesterday. Their explanation of the phenomenon goes as follows:
Note the contrast between the tense and the time adjunct […]. The tense reflects
the permanence of the photographic record while the adjunct yesterday (like the
dates in [entries in chronicles]) gives the time when the occupation actually took
place. (Huddleston and Pullum 2002: 126)
Although no such temporal contrast between the non-deictic verbal tense and
an adverbial of time is attested for the headlines in the analyzed sample, the the-
oretical possibility of that particular combination occurring cannot be entirely
discounted.
The combination of the historic present with past-time adverbials is, of course,
also possible, although this is not the usage we find in news headlines. Leech
(1971: 7) mentions the example Last week I’m in the sitting room with the wife,
when this chap next door staggers past and in a drunken fit throws a brick through
the window, explaining that “[s]uch utterances are typical of a highly-coloured
popular style of oral narrative, a style one would be more likely to overhear in the
public bar of a village inn than in the lounge of an expensive hotel”.
Normally, however, an adverbial of time commits a writer to the choice of

a corresponding tense so that a temporal contradiction between the two is pre-
vented. It could be argued that the absence of adverbials of time facilitates the
non-deictic use of verbal tenses in headlines. This view is supported by Declerck
(2006: 191), who notes that the absence of temporal indicators in some contexts
enables the use of the present tense to refer to a past situation: “In a few very
clearly defined contexts, the present tense may be used when it is not important
to give explicit information about the temporal location of a situation. Newspaper
headlines provide an illustration: Israel strikes back”.
Presence of adverbials of time in headlines
As regards the time adverbials in the corpus, they form two broad groups. There
is a difference between those adverbials that are crucial to the nature of the news
story and those that merely provide a general temporal anchorage of the event.
The first category typically includes reports that provide a summarization of
statistical and similar data over a period of time. The adverbial of time thus has
a central role in the story itself. The time when the report is released is not men-
tioned – that time can be assumed to be in the recent past (e.g. ‘yesterday’) and
need not be encoded in the headline. Cf. the following example:
Home ownership hits lowest level in 24 years
This type of adverbial tends to occur in conjunction with superlatives and ordinal
numerals such as the first, the second, etc. In some cases, this combination leads
to a dual expression of time: the superlative or the ordinal numeral appear in
phrases that have temporal meanings (cf. for the first time and the longest run in
the examples below):
Tesco profits fall for first time in 18 years (The Telegraph; 3.10.2012)
Gold price suffers longest run of monthly falls in 16 years
Occasionally, the factual or statistical information is attributed to some source

mentioned in the headline. Although it is the release of the data through the
source that makes the news, the headlines tend to focus on the content rather
than the source because it is the former that is newsworthy. The source is usually
specified in the lead (e.g. …according to a new report in the first example above,
and Tesco said pre-tax profits fell… in the second). The source is not represented in
104 News and Time
the headlines, unlike the following example, where the word census does indicate
the source of the data:
Census: 4m UK Christians fell away in last decade
(The Times; 11.12.2012)
The point is that the adverbial phrase indicating time forms a part of the em-
bedded proposition that is structurally subordinate to the attribution informing
us of the release of that some report/study/data. Syntactially, the leads (or some
other textual segments in the body copy) accompanying such headlines consist
of a reporting clause (which is the primary news event, i.e. the release of some
information) and a reported clause (which specifies the content of the news and
is newsworthy by itself). Hence, the lead that elaborates the last headline above:
“The number of Christians living in England and Wales has fallen by four million
in the past decade, the 2011 Census reveals”. Neither the lead nor the headline
provides a temporal specification of the time of release of the 2011 census. That
information can be inferred to be ‘yesterday’ or ‘over the past few days’.
This pattern is similar to the use of temporal indicators in syntactically sub-
ordinate clauses, such as the object clause in the next headline, where the time
adverbial three months ago stands at the centre of the item’s newsworthiness, be-
cause of its distinct focus on the past time (cf. also the verbs in the simple past):
Plebgate: Cameron knew Mitchell evidence was suspect three
months ago (The Independent; 21.12.2012)
Adverbials of time are also used in stories where time is of the essence, particular-
ly for future plans and predictions. These refer to the future time of some event,
even though they may be ultimately based on statements and plans made in the
recent past (cf. particularly the last example below).
4,000 troops home from Afghanistan next year (The Times; 19.12.2012)
BBC executives to discover their fate today (The Telegraph; 19.12.2012)
Lagarde wants Greek deal next week (The Guardian; 16.11.2012)
In the following example, the temporal specification five years on actually refers to
the present time, meaning ‘now’ or ‘still’. The present tense expresses a continued
state that is applicable for a prolonged period of time, rather than referring to
some identifiable past-time event:
Sub-prime haunts Wall Street five years on (The Telegraph; 3.10.2012)
There are several additional interesting phenomena related to the presence or

absence of time adverbials in headlines. Some adverbials of indefinite time can,
for instance, be present only implicitly and be expressed through other formal
means, e.g. as adjectives or verbs. Once again, the comparison of online headlines
on the paper’s home pages and article web pages can reveal noteworthy contrasts
in the formulation of temporality in headlines. In the following pair of headlines,
for instance, the former lexicalizes the future-oriented temporal element through
the adjective new, while the latter articulates that particular aspect of meaning
through the time adverb again.
(a) Dale Farm families face new eviction
(The Guardian; 19.12.2012, home page headline)
(b) Dale Farm travellers face eviction again
(The Guardian; 19.12.2012, article page headline)
A very similar situation is documented in the following two examples, where the
article page headline expands the information provided in the home page head-
line by adding an adverbial phrase of time (for the first time), which provides a
newsworthy aspect of the story that is emphasized through its inclusion in the
headline:
(a) Prime Minister admits Britain might decide to leave EU
(The Independent; 17.12.2012)
(b) Prime Minister admits for the first time that Britain might decide
to leave EU (The Independent; 17.12.2012)
Because of the general tendency of temporal indicators to be absent from head-

lines, online news is not significantly different from traditional print news. There
are, however, differences in leads that arise from the rolling 24-hour news cycle of
online news. Where print news needs to project the deictic centre of the coding
time into the anticipated receiving time, resulting in the use of such retrospective
adverbials of time as ‘yesterday’, online news does not perform such projection
because it is intended to be received on the same day, i.e. contemporaneously
with the occurrence and the editorial processing of the event. In this connection,
let us point to the tendency in online leads to use the time adverbial ‘today’ (and
even ‘earlier today’), often even in combination with the simple past tense. This
structure is something of a blend: the past tense is the conventional choice for
encoding a past-time occurrence in hard news beyond the headline, while the
adverbial of time expresses a temporal connection with the reader. Needless to
say, the specification of time is frequently missing from leads because they operate
with a similar assumption of recency as headlines, allowing them not to encode
time explicitly.
106 News and Time
5.4 Expressing the setting and location of the story
The category of time belongs to the semantic category of setting, together with the
specification of place, realized through a structural feature known as attribution
(Cotter 2010). Let us briefly discuss the nature of localization in news headlines
because the category of setting has, as seen above, a diminished role in headlines.
In contrast to time adverbials, phrases that specify place are more frequent (cf.
Bell 1991: 189), despite being expressed through similarly dispensable sentence
constituents such as adverbials of place and prepositional phrases specifying oth-
er sentence elements. An analogous observation is made by Bednarek and Caple
(2012: 101), who, however, do not elaborate on the possible reasons for this situa-
tion. Some typical examples include the following:
Saved by his safety rope on Ben Nevis. Killed by team who came to
rescue him? (The Independent; 28.2.2013)
Net migration to the UK falls by a third (The Telegraph; 28.2.2013)
In headlines, the localization of news stories appears to be relatively important.

Bell argues that:
Place is sometimes specified, but never time. […] place is the most common third
category in the headline and may even substitute for actor. Evaluation and other
commentary categories are absent. (Bell 1991: 189)
When combined with the two core elements of headlines, namely the event and
the participant, the individual headline constituents may be placed on a cline re-
flecting the different degrees of their relative importance. The event is arguably
the most important element, expressing some newsworthy state, action, result,
etc. The category of the participant identifies the social actor who is involved; this
category can occasionally be the most important, e.g. when the event is known
or expected and the identity of the social actor makes up the most newsworthy
element (e.g. the name of a sportsperson or a team that wins a championship).
The location and the time are optional elements with a decreasing level of relative
importance, since they represent circumstantial information – see Figure 5.1.
As regards the linguistic realization of the semantic categories of time and
place, they need not be expressed solely through adverbial phrases. There appears
to be a trend to express the elements through syntactic constituents that are either
indispensable in the syntactic structure of the sentence or placed very close to
obligatory sentence constituents. Thus, the category of time is expressed in every
finite verb form (even though tense tends not to be used deictically in headlines),
and the category of place finds its frequent expression through some other formal
Highest Lowest
Event / participant (→ location (→ time ))
Figure 5.1 The scale of the relative importance of headline constituents
means than the adverbial phrase. This is most commonly adjectival premodifica-
tion, as in the following examples:

Fukushima cancer risk ‘played down’ (The Times; 28.2.2013)

Egypt balloon pilot fights for life (The Guardian; 1.3.2013)
Pound falls as UK manufacturing shrinks (The Guardian; 1.3.2013)
Ryanair cuts Stansted capacity by 1m (The Telegraph; 28.2.2013)
In all of the examples, location is expressed through premodification in complex

nominal phrases (cf. Fukushima cancer risk; Egypt balloon pilot). The lo-
calization of the news stories could be re-expressed by means of adverbial phras-
es, but that would result in the syntactic expansion of the headline (e.g. Stansted
capacity → *capacity at Stansted). The addition of an optional sentence con-
stituent would also increase the length of the headline as well as decrease the
lexical density. The inclusion of locative elements within noun phrases, which
significantly increases their complexity, is at least partly motivated by the head-
lines’ need for brevity.
Yet, many headlines contain full adverbial phrases of place. The following ex-
ample illustrates the use of a descriptive locative adverbial phrase in a headline.
The phrase conveys newsworthy information in a headline segment that is in the
position of a syntactically subordinate relative clause. As such, it also contains the
simple past tense because the relevant proposition provides no more than back-
ground information to the main event mentioned in the main clause:
London Underground fined over runaway train that hurtled
through seven stations narrowly avoiding ‘terrible tragedy’
Commonly, the location is also expressed through the figure of speech of meton-
ymy, or when the name of a country appears in a specific syntactic position such
as the subject:

Washington gridlocked as US spending cuts threaten recovery

China executes four foreign nationals (The Guardian; 1.3.2013)
108 News and Time
This finding is in harmony with Bell’s observation that the category of location
can also be inferred from the description or the representation of the relevant
news actor: “Place is occasionally expressed in the canonical prepositional phrase
[…] More often we infer a location for events from the description of a main ac-
tor” (1991: 180). Moreover, the localization of the story is sometimes dependent
on the placement of the news item in a particular section (e.g. UK/national news)
or on the readers’ inferences ultimately based on their cultural knowledge. Cf. the
following examples in which the expression Tory unequivocally locates the for-
mer news item to the United Kingdom (newspaper section on domestic politics)
and the name of the Pope in the latter headline places the story in the Vatican:

Tory chastised over ‘abusive’ messages (The Times; 1.3.2013)
‘Sede vacante’ as Benedict XVI tells world: ‘Thank you and good
night’ (The Telegraph; 28.2.2013)
In online news, the situation is more complex owing to the existence of the two
levels of headlines, namely the article preview on the home page and the article
on its own web page. The headlines may undergo various transformations be-
tween these two levels. For instance, where the home page encodes the location
by means of a premodification of an obligatory sentence constituent, the headline
may re-express the information by means of a syntactically optional adverbial of
place. This is the case in the following example, where the nominal phrase China
landslide, functioning as the subject of a subordinate clause in the newspaper’s
home page headline, encodes the location of the story (China landslide buries
school) by means of premodification. The headline accompanying the story on the
article web page, however, re-expresses the same content in the adverbial phrase
in China landslide. The latter has a quasi-locative nature because, in addition to
place, the phrase also specifies the cause of the main event:
(a) 18 confirmed dead after China landslide buries school
(The Times; 5.10.2012)
(b) Eighteen children confirmed dead in China landslide
(The Times; 5.10.2012)
Interestingly, the pattern is continued in the leads accompanying the two respec-
tive headlines: the home page lead in the article preview encodes location through
the adjective Chinese (see (a) below), while the lead on the article web page pro-
vides a full localization of the event (in a mountainous part of southwestern Chi-
na), in combination with its temporal placement (yesterday; see (b)):
(a) A Chinese school that stayed open during holidays to make up for classes
missed because of an earthquake was buried in a landslide
(b) All 18 primary school children who were buried in a landslide in a moun-
tainous part of southwestern China yesterday have been confirmed dead after
rescue attempts failed to recover any survivors.
It appears that the various ways of encoding location can be exploited in online
news for the sake of ensuring some diversity in the opening segments of news
articles. As mentioned in Chapter 4, the same information is, as a rule, repeated
in online news several times across the the multiple textual segments encountered
on the newspaper home page and the article web page. While some elements may
remain stable (e.g., the process expressed in the verb, the identity of the news
actors), some variety is needed in order to avoid the impression of excessive struc-
tural (syntactic) as well as lexical (semantic) repetitiveness. The variable encoding
of locality through adjectival premodification and in adverbial phrases of place
adds such a minimal degree of variability.
5.5 Verbal tenses in headlines
As shown in the previous sections, the category of time is encoded in headlines

only infrequently through adverbials. In the absence of adverbials of time, the
category of time can be expressed in headlines through verbal tense. However,
verbal tense in English headlines is often not deictic: its choice is not motivated
on the basis of the link between the event referred to and some specific temporal
anchor (provided by the coding time or the receiving time). Tense in headlines
is thus primarily non-deictic and its nature arises from the operation of headline
conventions (cf. Section 5.7).
In my approach to the pragmatic interpretation of the conventionalized use
of the non-deictic present tense and some other tense shifts, I rely on the expla-
nations of English verbal tense found in the major academic grammars. While
they mostly explain the semantic properties of tense, they also make frequent
observations on pragmatic uses.
In the traditional definition of tense, based on morphological marking, the
English tense system contains two primary tense categories: present and preterite
(Huddleston and Pullum 2002), both realized by means of verbal inflection. The
secondary tense system distinguishes between preterite (i.e. the simple past tense)
and the perfect, both considered as past tenses because “they both express the
temporal relation of anteriority” (Huddleston and Pullum 2002: 139). Unlike the
preterite, the perfect is usually non-deictic.
110 News and Time
In this account, the following categories are applied: the present tense (also
commonly known as ‘the present simple’ or ‘the present’); the past tense (also
known as ‘the past simple’); and the present perfect. I am less interested in the
mutual relationship of these tenses on the basis of various grammatical criteria
than in the differential meanings that are associated with their use in headlines
and the implications which the choice of one of these tenses – as opposed to the
others – entails for the recipients. Being aware of the complexity of expressing fu-
turity in English (cf. Chapter 7.1), I also treat the will-future as a separate category
(the future tense, also known as ‘the future simple’), although its classification as a
distinct tense in the English tense system is generally understood as problematic.
Technically speaking, the form expresses a range of modal meanings and is for-
mally co-classifiable with the other modals (could, might, etc.).
In any case, while the future appears to be a relatively marginal category in
headlines (see below), my data indicate that it is the mutual contrast between
the present tense, the past tense and the present perfect that holds the key to our
understanding of temporal deixis in headlines. Shifts between these tenses are
motivated by semantic as well as pragmatic factors. The tenses can be used deic-
tically (to provide temporal anchorage of an event in a specific time zone) as well
as non-deictically (resulting in a conflict between the temporal placement of an
event and the verbal tense chosen to encode it). The choice of one over the other
entails the encoder’s decision to design the discourse space in a particular way,
e.g. by projecting deictic centres and negotiating the limits of the time zone of
‘recency’, which is crucial for the coverage of news events (cf. Section 7.3). Though
the present perfect is infrequent in headlines and, thus, has only a limited role in
the textual rhetoric of headlines, its significance emerges in the lead paragraph,
indicating that its true function is to be sought in the textual rhetoric of entire
news texts.
5.6 Tense in headlines in the data
Headlines can be categorized in various ways, e.g. according to their grammatical

complexity or syntactic or semantic structure. Obviously, any categorization crite-
ria derive from the specific goals of one’s analysis. Since the aim here is to primar-
ily address the ways tense is realized in headlines, the classification of headlines
is based on the headlines’ basic grammatical structure leading to the general dis-
tinction between nominal and clausal headlines. These categories can be usefully
extended into another grammatical contrast between finite and non-finite head-
lines, which reflects the nature of the predication. At the same time, the pragmatic
interpretation of temporal deixis in English headlines presented here means that
the analysis avoids an elaborate description of the formal types of headlines in

terms of precise sentence patterns that would capture the syntactic complexity
of this text type (cf. Straubmann 1935).18 Also, since the semantic coding of time
and temporal relations through other than verbal means is not germane to the
pragmatic explanation of verbal tense in news texts, this issue is not pursued here.
From the semantic point of view, the formal division of headlines into nomi-
nal and verbal (i.e., clausal) forms may itself be justifiably found to be problematic
(Schneider 2000: 46), particularly because, in spite of their nominal form, many
headlines do have strong verbal qualities. This is because they refer to actions
that are represented through nominalizations, (deverbal) nouns and other parts
of speech rather than the verbal phrase. For the sake of illustration, let us consider
several examples of this phenomenon:
Cardinal O’Brien admits sexual misconduct (The Guardian; 3.3.2013)
Cardinal O’Brien issues apology over sexual conduct
Cardinal sorry for sexual conduct (The Times; 3.3.2013)
‘Sexual conduct’ admission by cardinal Keith O’Brien
The examples, describing an identical news story appearing in several newspa-

pers on the same day, can be formally classified into three headlines that have
a clausal structure (The Guardian; The Independent, The Times) and one with a
nominal structure (The Telegraph). At the same time, one of the clausal head-
lines is non-finite since the copula verb ‘is’ has been conventionally ellipted (The
Times). Formally, the headline is clausal and non-finite, though the full form of
the missing verb is easily retrievable. The most interesting from among the four
headlines, however, is the nominal headline in The Telegraph. Despite its nominal
form, it encodes a proposition with a strong verbal (actional) quality (cf. the noun
admission). In fact, it communicates the same content as the most ‘actional’ head-
line in The Guardian, only it dispenses with the clausal structure and condenses
the predicate (i.e. the sequence of the verb and the object: …admits sexual mis-
conduct) into a complex noun phrase (‘Sexual conduct’ admission…). Similarly,
the verbo-nominal phrase in the Independent’s headline (…issues apology…) is
expressed in the Times’ headline by means of an adjective that describes a state
rather than an action (Cardinal sorry…). The headlines, while being formally
quite distinct, are thus semantically much less clear-cut, at least as regards the
18. A comprehensive framework for headline typology is offered by Studer (2003), who distin-
guishes between major, minor, integrated, combined and embedded headlines.
112 News and Time
Table 5.1 Clausal vs. nominal headlines

Headlines No. %
Clausal 762 91
Nominal 75 9
Total 837 100
expression of dynamic actions that tend to be most directly encoded through

verbs. In terms of their dynamic actional character, the headlines could be more
easily arranged on a scale rather than compartmentalized into a distinct set of
(binary) categories.
Since the focus of the present work is on the expression of time and temporal-
ity, the traditional formal distinction into clausal (verbal) and nominal (non-ver-
bal) headlines is retained. The distinction is motivated by the need to account
for the temporal situation that is directly encoded through the tensed verb form,
regardless of whether the form is used deictically or not. Nevertheless, it is being
acknowledged that the grammatical transformation of verbally expressed dynam-
ic actions into static nominalized forms brings about the deletion of the temporal
specification of the action, conveying a sense of ‘timelessness’ on the correspond-
ing nominal form.
Table 5.1 summarizes the composition of the analyzed data as regards clausal
and nominal headlines. The former group is made up of all headlines that have a
distinct clausal structure. It also includes those headlines where the clausal ele-
ment appears in a subordinate syntactic position, typically specifying a nominal
element (e.g. Flood warning as getaway begins; The Times; 20.12.2012). The
table thus classifies as nominal only those headlines that consist of an unattached
(absolute) nominal, as well as those that use nominal elements in conjunction
with prepositional and adverbial phrases (cf. Cole in four-letter rant at FA
after Terry ruling; The Times, 5.10.2012). Nominal headlines thus exclude all
cases of evident ellipsis, i.e. where no finite verb can be unequivocally inserted.
As the table indicates, clausal headlines outnumber nominal ones by the ra-
tio of approximately ten to one. The clausal group is made up of headlines with
a clausal structure, regardless of whether the verb group is finite or non-finite.
Non-finite structures are also classified among clausal headlines as long as they
contain verbal groups with ellipted finite verb forms that can be readily substitut-
ed into the full clausal structure (most typically the verb ‘to be’, cf. Tory chair-
man’s ‘get-rich-quick’ site investigated; The Times, 5.10.2012).
The classification of headline types as clausal vs. nominal, however, fails to
account for the internal syntactic complexity of many headlines. The problem
arises where a non-finite clause or nominal phrase is further specified by means of
Table 5.2 The proportion of syntactically complex headlines

Headlines No. %
Clausal – single 647 77.3
Clausal – complex 115 13.7
Nominal 75 9.0
Total 837 100.0
a finite clause. In this analysis, both headline subtypes are classified under clausal
headlines since the presence of a finite verb, even though on a syntactically subor-
dinate level, is taken to be decisive. For the sake of illustration, let us consider two
characteristic examples of such syntactically complex clauses. The first situation
concerns a headline containing a clausal structure that is modified by another
clausal element (18 confirmed dead after China landslide buries school;
The Times, 5.10.2012). This includes cases of reported speech when the reporting
clause contains a finite verb form. The second case is when a subordinate clause
modifies a nominal group that does not have a clausal structure, or a clause is
attached by way of specification to a nominal element (Flooding threat as
heavy rain hits UK; The Guardian, 5.10.2012). In an earlier study that made a
distinction between these headline subtypes (Chovanec 2003b: 86), it was found
that the combination of two finite (tensed) verb forms in a single headline is rel-
atively frequent, amounting to 12.5 per cent of headlines in the researched sam-
ple from the online version of The Telegraph. Out of that number, approximately
70 per cent were combinations of two verbs in the simple present tense, followed
by the combinations of the present simple and the past simple (18 per cent), and
the double use of the past simple (12 per cent). As shown in Table 5.2, the current
corpus includes 115 headlines (13.7%) with such complex structures. Out of this
number, 80 (9.6% of the total) consist of two (or more) finite clauses and 35 (4.1%
of the total) include a nominal phrase or a non-finite clausal structure with an
ellipted operator + a specifying finite clause (cf. Table 5.3).
As regards the expression of temporal deixis, it is headlines with a clausal
structure that are central to our purpose because their verbal forms allow for
the identification of grammatical tenses. Non-finite clausal headlines are treated
separately because they arise, most frequently, out of the conventional ellipsis of
auxiliaries that are typically easily recoverable. Table 5.3 presents the composition
of the sample as regards the tenses used. The category ‘other’ subsumes headlines
with other modal auxiliaries (regardless of their temporal orientation).
As Table 5.3 indicates, approximately one in six clausal headlines consists of a
complex sentence. The next table presents the proportion of tenses in such com-
plex sentences for the set of 115 headlines found in the sample. The total number
114 News and Time
Table 5.3 The proportion of tenses in headlines

Headlines Tense No. %
Clausal (762)
Single clause (647) Present tense 428 51.1
Past tense 30 3.6
Present perfect tense 8 1.0
Will-future 9 1.1
Other 10 1.2
Ellipted 162 19.3
Complex sentence (115) Clause + clause 80 9.5
Nominal/ellipted + clause 35 4.2
Nominal (75) Nominal 75 9.0
Total 837 100.0
Table 5.4 The proportion of tenses in complex headlines

Tense in clause unit No. %
Present tense 143 60.9
Past tense 29 12.3
Present perfect tense 3 1.3
Will-future 6 2.6
Other 19 8.4
Ellipted 25 10.6
Nominal 10 4.2
Total 235 100.0
of units analysed is 235 (rather than 230) because five headlines in the corpus
consist of three clauses. The category ‘ellipted’ includes non-finite clauses where
the copula verb is omitted, while ‘nominal’ comprises those headlines where the
finite clause appears in a position syntactically subordinate to a nominal fragment
(cf. ‘Plebgate’ arrest as PM hints at comeback; The Times; 20.12.2012).
It is clear from the data that the present tense is very dominant; Table 5.4 indi-
cates that a verb phrase in the present tense is found in 143 clausal units (60.9%),
which sharply contrasts with the past tense (29 cases, i.e. 12.3%), the will-future
(6 cases, i.e. 2.6%) and the present perfect tense (3 cases, i.e. 1.3%). Modals (other
than will, which is included in the previous category) account for 19 cases (i.e.
8.4%). Moreover, since the 143 occurrences of the present tense are spread across
103 of the 115 headlines in the subset, with the remaining twelve headlines con-
taining some other tense forms, the percentage of the present tense in complex
sentence headlines reaches a stunning 89.5 per cent.
The verb phrase is realized as non-finite in 25 cases (10.6%), which is a sig-

nificant proportion that conforms to the traditional perception of ‘headlinese’ as
a form of language use with extensive ellipsis of grammatical words. Since the
ellipsis affects the auxiliary verb in passive and progressive structures, the non-el-
lipted versions would – should the missing auxiliaries be recovered – most likely
be construed in the present tense as well, because that is the tense that is con-
ventionally expected in headlines. In a sense then, the ellipted headlines could
be seen as indirectly contributing to the dominance of the present tense. Cf. the
following headline, where the non-finite clause has a strong implication of an
ellipted present tense:
‘I’m not real, it’s your mum and dad’: Bad Santa suspended for saying
too much (The Independent; 21.12.2012)
The dominance of the present tense is also apparent from Table 5.5, which com-
bines the proportion of tenses in single clause headlines and complex sentence
headlines. Since the nominal units that form a part of some complex headlines
are excluded from this sum, the percentages differ slightly from those in the pre-
vious table – they reflect the proportion of only fully clausal headline segments
(i.e., the total in Table 5.5 includes 225 finite clausal units, as opposed to the total
of 235 finite and non-finite units in Table 5.4). Thus, where a headline consists of
two segments that are evidently clausal (e.g. 18 confirmed dead after China
landslide buries school; The Times, 5.10.2012), this is represented twice in the
table (in the example above, once in the ‘ellipted’ category and once in the ‘pres-
ent tense’ category). Complex sentence headlines with an initial nominal element
(e.g. Flooding threat as heavy rain hits UK; The Guardian, 5.10.2012) are
counted only once because the headline evidently contains only one verb phrase.
Table 5.5 The proportion of tenses in finite headline segments

Tense Headline form
Single clause Complex sentence Total
No. % No. % No. %
Present tense 428 66.2 143 63.6 571 65.5
Past tense 30 4.6 29 12.8 59 6.8
Present perfect 8 1.2 3 1.3 11 1.3
Will-future 9 1.4 6 2.7 15 1.7
Other 10 1.6 19 8.4 29 3.3
Ellipted 162 25.0 25 11.1 187 21.4
Total 647 100.0 225 100.0 872 100.0
116 News and Time
The data indicate that while the frequency of the present tense in the two
types of headlines is very similar (66.2% and 63.6%), single clause headlines use
ellipted auxiliaries more than twice as often. By contrast, complex headlines show
a higher frequency of the past tense and other forms (the will-future and modals).
The occurrence of the present perfect tense is minimal in both single clause and
complex headlines (see Table 5.5).
There are several observations to be made in connection with headlines con-
taining multiple tensed forms of verbs. The most noticeable feature is the compar-
atively more frequent occurrence of the simple past, as well as the future and the
‘other’ category. Thus, the simple past is used for events that qualify as non-recent
(cf. Section 7.3):
Fugitive McAfee feared he would be traced by phone
More importantly, however, the past tense appears in clauses that are formulated
in the non-authorial accessed voice of some news actor. The headline then essen-
tially has a dual structure consisting of (1) a reported clause in the external voice
(in direct, free direct or indirect speech) and (2) another clause in the authorial
voice of the paper, one that functions either as the reporting clause or a summary
of the main event, cf.:
Topless photo was a joke, protests ‘Agent Shirtless’
(The Times; 16.11.2012)
‘I thought I was going to die’: Blind man shot with 50,000-volt Taser
sues police (The Independent; 22.10.2012)
This pattern also occurs with other tenses, such as the present perfect and the
future, cf.:
Reports of my death have been greatly exaggerated: Fidel Castro
proves he’s alive and… gardening (The Independent; 22.10.2012)
We will sue to get asylum, say fearful Afghan interpreters
(The Times; 19.12.2012)
The greater variety of tenses in the complex sentence headlines most likely stems
from the fact that such headlines are structurally predisposed to the more explicit
use of the heteroglossic combination of two voices, where the external accessed
voice correlates with tenses other than the present tense. This phenomenon is
discussed in detail in Section 7.2.
As regards Table 5.5, there is another noteworthy pattern: the combination of

an initially placed imperative and a subsequent reporting clause formulated in the
present tense. The imperative can be variably addressed to the readers (cf. the first
two examples) or other news actors mentioned in the headline or the subsequent
text of the article, cf.:
Meet Yoda, the worm who lives 1.5 miles beneath waves
Hold onto your seats: American Airlines cancels 94 flights
Stay tough on Europe, Tory Right warns PM (The Times; 5.10.2012)
Ride to the rescue of nation’s cyclists, Cameron is urged
(The Times; 5.10.2012)
Get tough on banks or else, Osborne’s own advisers insist
Regardless of the method of counting the occurrences – in either whole headlines

or actual realizations of verb phrases – the quantitative findings provide a general
perspective on the proportion of the individual tenses that will be described qual-
itatively in Chapters 6, 7 and 8. This description will be based on the grammatical
meanings of the individual tenses and the pragmatic implications of their use in
headlines. Some of the recurrent patterns in complex sentence headlines will also
be commented on in a special subchapter (Section 8.4).
5.7 Headline conventions
The study of language in context will inevitably have to deal with the issue of
language variation and its patterned predictability. Register and genre-oriented
research has pointed out that there exists systematic variation that correlates with
particular situations of usage. For instance, the concrete configurations of the
three central parameters of domain (field), tenor and mode, which can be drawn
upon to describe registers, are associated with particular formal features that are
conventionally linked with specific contexts. Speakers and writers will draw on
conventional linguistic resources – be they lexico-grammatical, structural, prag-
matic, etc. – in order to create texts that are situationally appropriate and effective
in serving the purposes for which they are produced.
The conventionality of language is thus inevitably related to efficient commu-
nication. Efficiency typically ensues when the recipients’ attention is directed to
118 News and Time
the propositional content rather than the form of the message, i.e. when the ut-
terance is unmarked with respect to the expectations placed on its form. This fact
does not discount the possibility that formally marked (deviant, inappropriate,
unconventional, innovative or creative) utterances can also be used for efficient
communication, since deviation from the norm is typically not accidental. It can
serve as a means whereby the speakers/writers can trigger the hearers’/readers’ in-
ferencing processes and invite them to search for relevant interpretations beyond
what appears on the surface of the message. Familiarity with such shared conven-
tions forms a part of the background cultural knowledge shared by language users
or members of particular discourse communities. It can be actively, as well as stra-
tegically, used by speakers and writers towards pursuing their discursive goals.
Headlines constitute a well-known type of text (register) that is highly char-
acteristic because of its conventional linguistic form. Thanks to the very distinct
and easily discernible style of English headlines, the textual form characterizing
‘headlinese’ is almost instantly recognizable by native as well as non-native speak-
ers of English and exists as a shared discursive resource.
The telegraphic style of English headlines is culture- and time-specific. Head-
lines in other languages have their own conventions that may be quite distinct
from those of headlines in English. In Czech, for instance, headlines are compar-
atively more ‘narrative’, not manifesting any major grammatical deviations that
would make them perceptibly different from the corresponding non-headline
statements encoding the same content. Czech headlines do not utilize the con-
ventional manipulation of temporal deixis, as is customary in English headlines.
There is no projection of deictic centres: past-time events are rendered through
the past tense, not the present tense as in English. Consequently, the textual rheto-
ric of Czech headlines and news texts is markedly different from the English mod-
el outlined in the present book. Similarly, Thogmartin in his comparative analysis
(1991: 260) notes the greater range of tenses in French headlines as opposed to
English headlines, and Harder (1995: 494) mentions the much less frequent use
of the headline present in Danish newspapers. As regards the time-specificity of
headlines, the presently existing conventions of English headlines are a result of
the development of this textual format across several centuries: the headline de-
veloped from a form that was originally much more narrative and operated on
multiple levels that conveyed all the main elements of the story (cf. Bell 2002;
Schneider 2000: 54).
Various authors have mentioned the existence of a set of linguistic conven-
tions that characterize headlines (for the most recent account, see Bednarek and
Caple 2012: 101). Let us summarize the main headline conventions observable in
English in order to appreciate the set of forms that we typically find in these brief
and constrained textual segments. This is crucial for our understanding of how
tense variation operates in, among others, heteroglossic headlines (cf. Chapter 7),
since most of the grammatical and lexical features outlined below are applicable
only to the authorial voice:
Grammatical features
– conventional shift of tenses (past/present perfect → present; future → non-fi-

nite verbal group)
Tobacco giants lose Australian packet fight (The Times; 5.10.2012)

Belfast riot police arrest children (The Guardian; 18.12.2012)

Hong Kong to probe interbank rates (The Times; 20.12.2012)
– non-finite passive constructions (ellipsis of finite operators)

‘Shameful, lying’ police chief sacked for misconduct
(The Times; 5.10.2012)
Travellers hit by Heathrow train chaos (The Times; 21.12.2012)
– ellipsis of definite/indefinite articles and determiners

Boy in cancer legal batter has surgery (The Guardian; 20.12.2012)
French steelworkers make last stand (The Guardian; 5.10.2012)
– ellipsis of the definite forms of the verb ‘to be’ (both as an operator and a cop-
ula verb)
Tory chairman rocked by ‘get rich quick’ probe
Man in court on murder charge over Cardiff hit-and-runs
– unattached nominals – noun groups stating the mere existence of a

phenomenon
Pressure on BBC to publish Savile report (The Times; 5.10.2012)
Boost for Obama as US jobless rates falls (The Telegraph; 5.10.2012)
Scott discovery (The Times; 19.12.2012)
– condensed quotations: replacement of verbs of speech with the colon preced-

ed by a designation of the relevant news actor, used for attributing and hedg-
ing statements
Freddie Starr: I have never touched an underage girl
120 News and Time
– avoidance of modalized statements, lexical expression of modality (e.g.

through the verb ‘to face’)
Britain faces risk of blackouts, warns Ofgem (The Telegraph; 5.10.2012)
– shift in modality from possibility towards certainty (i.e. from modally hedged
utterances in the lead and the body copy to the presentation of the relevant
proposition in the headline as factual and without hedging or modality)
Lexical features
– specific headline lexis

Fears for patient safety as 60,000 NHS jobs face axe
– tendency towards monosyllabic expressive lexemes (particularly in the popu-

lar press)
– evaluative and emotional expressions
– propensity to word play, fostered by the homonymy potential of monosyllabic
words
– nominalizations and complex nominal phrases (including heavy premodifi-
cation of noun phrases), resulting in high information density (and placing
high demands on recipients)
Cole in four-letter rant at FA after Terry Ruling
(The Times; 5.10.2012)
Royal prank hoax nurse suicide tragedy: Husband’s fury at DJs and
hospital bosses over Jacintha’s death (The Mirror, 10.12.2012)
– lexical creativity (ad hoc creations, nonce formations)
Non-linguistic features
– visual prominence (typography, font size, layout, use of colour)

– interplay between visual and textual components (this also affects stories that,
though mutually unrelated, are placed side by side and brought into mutual
contact through the prominent presentation of the textual and the visual ele-
ments in the respective stories)
– other aspects of visual grammar that carry communicative value (framing,
salience, positioning on the page, preferred reading paths, etc.; cf. Kress and
van Leeuwen 2006)
While some of the features identified above are specifically associated with news
headlines, others are seen as instantiations of a style of language that has a more
general application. In the broader conception, the conventional form of headlines
is aligned to the language of telegrams, public notices, slogans, notes and other
short texts that are of an informative or regulatory nature and whose linguistic
characteristics (the typical ‘telegraphic style’) are seen as being motivated by cer-
tain external criteria such as lack of space. In this context, Halliday (1994: 392)
talks of the “grammar of little texts”, arguing that “some registers do have fair-
ly distinct grammars” and “differ in certain respects from the grammar of other
registers of English not constrained by such limitations”. Such ‘little texts’ share
certain general tendencies. For instance, they use nominals and verbals without
deixis, omit otherwise obligatory clause elements, evade mood specification, and
include unattached nominals. Various subtypes of little texts, however, may also
differ in some of their linguistic features. Thus, for instance, the tendency towards
monosyllabic expressive lexis and heavy premodification is not found in public
notices, which are, by contrast, sometimes marked by a noticeably formal and
detached style absent from, for example, news headlines, photo captions, etc.
As some of the examples above indicate, the situation in headlines becomes
more complicated in cases of direct, semi-direct and indirect speech. Whenev-
er another voice is accessed, the operation of headline conventions can be mo-
mentarily suspended since the headline provides a ‘window’ to another, mediated
discourse. While the resulting heteroglossia and intertextuality make headlines
distinct from other little texts, the suspension of their linguistic conventions can
be interpreted as a reader-oriented attempt to bridge the discursive gap existing
between the institutional, convention-bound voice of the paper and the individu-
al readers (see Chapters 7 and 8).
The existence of headline conventions, particularly as regards the central role
of the simple present tense and commonly occurring non-finite verbal groups,
becomes crucial for understanding the nature of temporal deixis in news arti-
cles. This is even more so because the headline is not a stand-alone unit, as is
sometimes argued in accounts focusing on the formal aspects of ‘headlinese’, but a
component of a larger whole, as is proposed here. The choice of tense is therefore
not simply a matter of representation: it forms a part of the discourse structure
of the entire article. As McCarthy and Carter (1994: 96) note, “the choice of tense
and aspect can be seen to have a discourse dimension, in that the choices are not
determined purely by semantic factors relating to ‘objective’ time”. They further
point out that “tense and aspect choices have become part of the conventions of
122 News and Time
the genre”. Given the commonly accepted (though not entirely unproblematic)19
understanding of headlines as serving both a semantic function and a pragmatic
function (Iarovici and Amel 1989; Bell 1991), the conventional presence of cer-
tain tenses in headlines (as well as leads) is connected to the pragmatic function
of relating to the readers. The headline present tense creates an increased level of
immediacy, relevance, and recency and thus contributes to the greater and more
active engagement of the reader.
Bearing in mind the overall situation as regards temporal deixis and tense us-
age in headlines, as outlined against the background of the conventional nature of
‘headlinese’, let us now proceed with an account of the individual tenses, starting
with the present tense in the next chapter.
19. Cf. Dor (2003), who provides for an alternative explanation of the function of headlines
from the point of view of relevance theory, and Ifantidou (2009), who uses the same approach
to justify the perception of headlines as autonomous texts, i.e. as separate from the body copy
of news texts. This approach is justified by the fact that readers very often do not process news
texts beyond the initial headlines.
Chapter 6
The present tense in headlines
This chapter deals with the reference of the present tense in headlines, mapping
its correlation with the three time zones: the present, the past, and the future.
The discussion of the semantic properties of the present tense relies on standard
grammatical accounts of verbal tense and leads to the necessity of distinguishing
between several subtypes of deictic tense usage. The non-deictic use, primarily
the reference of the present tense to the past time, is discussed in connection with
the conventional effects of the ‘historic present’ and the switch of tense to indicate
evaluation in narrative, although there are clear differences between those par-
ticular usages and the headline present tense. Adopting a pragmatic perspective,
the headline present tense is interpreted as a form resulting from a double pro-
jection of deictic centres. It is argued that the semantic properties of the simple
present tense make it a particularly suitable choice for enhancing the interper-
sonal dimension of communication in news headlines. The chapter concludes by
mentioning the issue of the potential ambiguity of some present tense headlines
and the disambiguating role of the ensuing context.
6.1 The defaultness of the present tense in news headlines
Since the present simple is conventionally used in news headlines, it is not sur-
prising that it is also the tense with the highest frequency of occurrence. In gen-
eral terms, this may be little surprising because the present simple is the most
fundamental tense in the system of English tenses (Huddleston and Pullum 2002)
and is commonly referred to as ‘atemporal’ or the ‘default’ tense because it de-
scribes a situation without necessarily locating it in time (Declerck 1991: 69).
As mentioned previously (see Tables 5.3–5.5), the number of headlines in the
present tense (535, i.e. 428 in single-clause and 103 in complex-sentence head-
lines) is relatively high, amounting to 63.9 per cent of all headlines in the corpus
(n = 837). The figure reaches 70.2 per cent if nominal headlines are excluded
from the set (n = 762), 61.4 per cent if all clausal headline segments are consid-
ered (n = 872), and 78.1 per cent if non-finite (ellipted) segments are excluded
(n = 685). Only 21.9 per cent of clausal headline segments contain other tenses
124 News and Time
(the simple past, the future, the present perfect and modal auxiliaries). The high
frequency of the present tense has been noted by others as well; for example,
Bednarek and Caple (2012: 103) note that the present tense prevails over the past
tense, often without tensed auxiliaries. This frequent use of the present tense
stands in contrast to the way verbal tenses are used throughout the bodies of
news reports – as Biber et al. (1999: 11) note, “because its overriding purpose is
to report past events, almost all the verbs in the news report are in the past tense.”
Apart from print news, the present tense is also used in broadcast headlines, al-
though some object that it is a cliché, particularly when it describes killings or
injuries (Thompson 2010: 147).
Non-finite headlines with a clausal structure and ellipted auxiliaries are also
likely to be understood as resulting from the ellipsis of auxiliaries that would oth-
erwise be cast in the present tense. This presumption is justified on account of the
existence of the generally shared conventions of English headlinese, which predis-
pose readers to seek the present tense as the default category of tense in headlines.
Given this understanding, non-finite headlines can be validly construed to have
originated from the ellipsis of the present tense. Thus, for instance, the reinstitu-
tion of the relevant full forms is possible in the following headlines (the ellipted
forms include are; are; and says, respectively):
Police suspended over man’s death (The Guardian; 1.3.2013)
Prescription charges to rise by 20p under new changes
Boris: EU bank bonus curbs are ‘deluded’ (The Telegraph; 28.2.2013)
Technically speaking, the present perfect tense is also among the possibilities for
referring to the present situation since its semantic properties align it with the
present situation time, showing affinity with the present tense. The present perfect
locates the event in the pre-present that “‘contains’ the situation time in terms of
inclusion or coincidence” (Declerck 2006: 212). In this way, the tense reflects “the
speaker’s concern with NOW” (Declerck 2006: 213). However, the present perfect
in headlines is so distinct that it forms a special category that will be treated sep-
arately rather than together with the present tense. Thus, the following headline
is untypical in that it describes the present result (i.e. the focus on the situation
time of NOW) of a long-term trend. It has a distinctly narrative feel to it, which is
rather unlike canonical hard-news headlines:
How Christmas television has changed (The Telegraph; 21.12.2012)
Chapter 6. The present tense in headlines 125
Focusing on the simple present tense, however, let us proceed by outlining some
of its semantic properties and proposing an explanation for its conventional usage
in news headlines.
6.2 Deictic and non-deictic tense
The present tense can be used in two distinct ways, depending on whether it refers
to the temporal situation at the moment of the utterance or not. This leads to the
distinction between the deictic and non-deictic uses of the present tense. (This
semantic distinction generally applies to other tenses as well.) When used deicti-
cally, the present tense refers to the present time. Non-deictic present tense refers
to a time that is either located in the past or in the future. The present tense then
does not correlate with the time of utterance. The past-time reference of the pres-
ent tense often underlies the perception of the event as having some present rele-
vance, while the future-time reference appears, among others, in the semi-modal
verbal forms ‘be to’, ‘be going to’, etc.
The first two examples below illustrate the deictic use of the present tense. The
latter example is non-deictic, referring to past time.
– Deictic present tense (present time reference):

Tory ministers plot rights act repeal (The Guardian; 3.3.2013)
NHS board backs under-fire chief (The Guardian; 28.2.2013)
– Non-deictic present tense (past/future time reference):

Ex-con breaks back in to prison (The Independent; 3.3.2013)
The present tense is the unmarked form in the system of tenses. It is, in this sense,
asymetrical (cf. Jakobson 1932). Biber et al. (1999: 457) point out that when cast
in the present tense, a verb form has “all-inclusive time reference” since the tense
“can be used to refer to events in the past, to present states, to present habitual
behaviour, or to future events”. This property is not shared by other tenses, most
notably the past tense, which is “used primarily to refer to states or events exist-
ing at some past time, excluding present and future”. The interpretation of the
present tense as the unmarked tense is also justifiable on morphological grounds
(Quirk et al. 1985). Such use of the present tense has been described as ‘atem-
poral’ (Dušková 1988: 219) or ‘timeless’ because the present tense is “used as a
default term in the primary tense system” (Huddleston and Pullum 2002: 130).
From another perspective, the present tense can express ‘universal time’, referring
126 News and Time
[now]
past time future time

[preceding now] [following now]
present time
[including now]
Figure 6.1 The universal temporal reference of the present time (Quirk et al. 1985: 176)
to an event whose existence is temporally unanchored. As Quirk et al. (1985: 175)

note, “something is defined as ‘present’ if it has existence at the present moment,
allowing for the possibility that its existence may also stretch into the past and
into the future”.
Thus, the present tense has a property which allows it to express other tem-
poral reference beyond the immediate present moment. This particular property,
which is also conventionally utilized in tense manipulation in headlines, is sche-
matically represented in Figure 6.1.
This universal reference of the present tense allows for its use in nonsimulta-
neous situations. Simultaneity, as understood by Declerck (1991: 70), is “the rela-
tion ‘holding at’ [and it] is the unmarked of the temporal relations.” He goes on to
say that “this explains why nonsimultaneous situations are often represented by
verb forms expressing simultaneity… A typical example of a shift of perspective
that is motivated by the desire to use the default tense is the substitution of the
present tense for any other tense in newspaper headlines”.
The present tense in headlines is associated with events that happened in the
recent past. The recency criterion is crucial to our understanding of the effect
of this grammatical form in news texts (cf. Quirk et al., who observe that “The
simple present is usually used in newspaper headlines to report recent events”
(1985: 182)). However, as regards the semantics of the tense, the recency has a
dual character. It concerns either (1) past events completed in the past with the
simple present used as a foregrounding device to highlight their current rele-
vance, or (2) past events that have brought about a change in the present time, the
present tense being justified on account of a stronger focus on the present result.
This issue is further discussed in connection with the ‘state present’ and other uses
of the present tense below.
The specific usage and effect of the present simple in headlines is closely re-
lated to its grammatical function. A review of its functions as they are postulat-
ed in traditional grammar is therefore complemented with illustrative examples
from the corpus. First of all, I will deal with the reference of the present tense
to the present time, and then cover its reference to the past time and the future.
The discussion illustrates that ‘the headline simple present’ is a relatively complex
phenomenon that combines semantic properties of deictically used tense on the
one hand and pragmatically-motivated tense shifts with a reader-oriented focus
on the other.
6.3 Present time reference of the simple present tense
As made clear above, not all headlines containing the present tense can be regard-
ed as representative of the classic shift of tenses that characterises the conventions
of headlinese. Where the simple present tense refers to the present time, there is
no significant discrepancy between the event time and the time expressed in the
proposition through verbal tense. However, the headline convention of ‘present
simple tense for past events’ draws on the effect created by the basic semantic
properties of the present simple that combine the three distinct subtypes of the
present tense: (a) the state present; (b) the habitual present; and (c) the instanta-
neous present (Quirk et al. 1985: 179–181).
State present
The state present concerns timeless statements (sometimes called ‘eternal truths’).
It includes “proverbial, scientific, or mathematical statements … [which] rep-
resent the extreme of temporal universality” but also “geographical statements
[such as] … examples of the ‘timeless present’” (Quirk et al. 1985: 179). The state
is either temporary or “persists for a longer time” (Biber et al. 1999: 453). The
timeless present can be illustrated with the following examples from the corpus:
Rats are ‘telepathic’: Scientists show rats can communicate using
brain-to-brain (The Independent; 28.2.2013)
MPs want new laws on police spies (The Guardian; 1.3.2013)
The first headline includes a timeless verb (Rats are ‘telepathic’…) that – while
expressing a general truth (a scientific finding) – contrasts with another instance
of the present tense in the same headline that has a clear past-time orientation (…
Scientists show…). The second headline above describes an event (MPs want) that
can be construed as relatively timeless and certainly valid at the time of reception,
although the article is based on information released in the recent past about a
statement issued by the Parliament.
128 News and Time
Habitual present
The habitual present is similar to the state present in that it implies “an inher-
ently unrestricted time span” with the dynamic verb referring to “a whole se-
quence of events, repeated over the period in question” (Quirk et al. 1985: 179) or
to “present habitual behaviour” (Biber et al. 1999: 453). The present tense verb in
the previous example (want) can be construed in this way since it can be read as
expressing a prolonged state rather than a momentary wish. Cf. also the following
example, where the semantic aspects of the habitual present and the present state
of affairs coincide:
Doctors don’t trust own hospitals (The Telegraph; 1.3.2013)
The habitual present is not very common in headlines, most likely because repeat-
ed events and habitual actions are less newsworthy in comparison with one-off
events whose singular occurrence has a stronger impact (cf. Bell 1991). Some
headlines that qualify for this category on account of referring to ‘a sequence
of events’ are border-line cases that could also be seen as instances of the state
present (cf. the first two examples below) and the instantaneous present (the last
example):
Apocalypse mania grips Russia (The Independent; 18.12.2012)
Floods cause chaos around Britain (The Guardian; 20.12.2012)
Anger builds in Italy as old guard plots fresh technocrat takeover
Instantaneous present
The third use of the present tense is the instantaneous present, which “refers to a
single action begun and completed approximately at the moment of speech [but]
because it implies that the event has little or no duration, it does not occur out-
side some rather restricted situations” (Quirk et al. 1985: 179). These situations
include, among others, sports commentaries and performative utterances (Austin
1962). The simultaneity can be subjective rather than objective since the event
does not usually happen exactly at the moment of utterance (cf. Leech 1971: 3).
Such subjective simultaneity can be established through the projection of deictic
centres and the shift of the temporal perspective. The following headline reports
on a performative action, so it contains a semantic element associated with the
instantaneity of the action, but because of the evident temporal distance between
the time of the event and the situation time, the whole headline might be read,
rather, as a description of the resulting state: the fact that an apology has been
issued:
BBC and ITV apologise to McAlpine (The Independent, 18.12.2012)
There is no genuine occurrence of the instantaneous present in the corpus. How-

ever, objective simultaneity does occur in some headlines – in a new category of
headlines found in online newspapers and accompanying live news.20 Many on-
line newspapers increasingly provide coverage of events as they are unfolding in
real time. The headlines of such news items, often prefaced with the word ‘LIVE’,
will then contain instances of genuine uses of the present tense with an instanta-
neous present meaning, cf.:
EU lifts arms embargo to rebels – live updates (The Guardian; 28.5.2013)
FTSE 100 rebounds as investors pin hopes on more stimulus
Nevertheless, it is the semantic category of the present tense that describes an

event as “an action ongoing at the time” (Biber et al. 1999: 454) that can help to
explain the effect of the conventional shift of tenses. With the discursive construc-
tion of the subjective simultaneity of the event time and the receiving time, the
events reported (which the reader knows have already happened) are presented
as if they were happening at the very moment of reading. Taking into account the
complex nature of deictic shifts in newspaper discourse, subjective simultaneity
actually applies at any moment the text is read and processed by individual read-
ers. The final effect is comparable to the vividness associated with the usage of this
tense in live spoken sports commentaries (i.e., the register of ‘sports announcer
talk’, cf. Ferguson 1983).
The subjective simultaneity present in some uses of the instantaneous present
indicates that the tense can be used metaphorically. For such uses, Quirk et al.
mention that the tense lends a certain theatrical quality to an utterance:
[t]here is bound to be simultaneity between the event described and the speech
event itself. In other cases, although such simultaneity does not obtain in any
exact sense, there is an implication of simultaneity which gives the utterance
with the instantaneous present a somewhat theatrical quality. Against the routine
ordinariness of the present progressive in Carlos is winning, we may place the
20. Live news headlines were not included in the compilation of the corpus for the present
study. In many cases, live stories are appended as satellite articles within an article cluster com-
plementing a main news story that is described statically. Some events, such as parliament
sessions, sports competitions, disasters, etc. are covered live on the Internet.
130 News and Time
dramatic air of Carlos wins! which pinpoints the final and climactic moment of
victory. (Quirk et al. 1985: 181)
Similarly, Leech (1971: 8) explains the present tense in headlines as having “some-
thing of the dramatic quality” of the instantaneous present. The present tense
then essentially functions not only as an attention-getting device but also as an
internal evaluation device whereby speakers can point out some aspect of their
narration as the most important (cf. Schiffrin 1981; for more details see below).
The dramatic air and the theatrical quality of an utterance encoded in the instan-
taneous present is something that can be found in many of the conventional and
highly stereotypical headlines that refer non-deictically to the past time, as in the
following examples:
Woman gives birth in helicopter (The Telegraph; 11.12.2012)
Pound falls as UK manufacturing shrinks (The Guardian; 1.3.2013)
In news headlines, the potential for such pragmatically-oriented use of the simple
present tense combines with the structural composition of the news article and its
top-down organization. The headline is, structurally, the most important element
of the news story, providing its encapsulation in the most basic form. While there
are various types of headlines, the most typical ones include the summarizing
headline that abstracts the story (and the story itself is, in turn, abstracted in
the lead). From this angle, it is perhaps little surprising that the simple present
should be used in the headline – the spotlight function of the tense has become
conventionalized within this textual macrostructure as a focusing device applied
by the writer to give prominence to the crucial aspect of the story. Unlike ordinary
narrative texts, where such a culmination usually occurs at the end, the climactic
moment of the news story comes at its very beginning in the form of the text’s
opening segment: the headline. The present tense pinpoints the key message of
the story, doing so by constructing a seemingly shared context of temporal refer-
ence linking the readers and the event.
Huddleston and Pullum (2002: 131) note that the present tense “is widely
used in news headlines (spoken, or written) for dynamic situations in the recent
past”. Explaining that such a choice of tense is shorter and more vivid than past
tenses, they regard this phenomenon as “a metaphorical extension of the use of
the present tense in commentaries”.
It appears that the present tense is a device whereby printed news discourse
may construct co-temporality – a feature that appears to be very strong in mod-
ern communication media (cf. Durant and Lambrou 2009: 17). The present tense
conventionally creates the illusion of a fictitious co-presence between the dis-
course participants and the event. In spoken broadcasts, a similar illusion can
be achieved through the use of time-shifted programmes and live transmissions;

in the temporally displaced context of printed texts, the effect can be assisted
through the manipulation of temporal deixis.
6.4 Past-time reference of the simple present tense
The present tense in headlines is most typically understood to refer to events that
happened in the past. Owing to the operation of the pragmatic presumption of
recency (which is also one of the news values motivating the choice of news sto-
ries), the past time is implicitly understood in a rather narrow sense as the period
of time coinciding with the publication frequency of the newspaper or journal.
The newsworthy period of past time thus extends from the coding time of the
previous issue until the coding time of the current issue of the paper, although
older stories are occasionally included as well, as are some other news items that
are tied to that particular brief period of time to a lesser degree (e.g. popularizing
reports on scholarly research).
Huddleston and Pullum (2002: 130–131) mention the general property of
the present tense as giving “a timeless, permanent perspective to what could also
have been conceptualised as past occurrences”. They give three specific situations
where the present tense “encroaches into past time territory”:
Meaning of tense Example

1. The historic present in narrative There was I playing so well even I couldn’t believe it
and along comes this kid and keeps me off the table
for three frames.
2. ‘Hot news’ in headlines UN aid reaches the stricken Bosnian town of
Srebrenica.
3. The past evidential use with verbs Your mother tells me you’re off to Paris tomorrow.
of communication I hear we’re getting some new neighbours.
Figure 6.2 Present tense in past time territory (Huddleston and Pullum 2002: 130–131)
Such extensions of the present tense into past time territory are considered as
metaphorical uses – conventional devices for:
a. making “the narrative appear more vivid by assimilating it to the here-and-

now of the speech act” (Huddleston and Pullum 2002: 131) (as in type 1
above);
132 News and Time
b. adding vividness to dynamic situations in the past (as in type 2); and
c. foregrounding the content of the proposition and backgrounding the com-
munication process itself (as in type 3).
From a pragma-linguistic perspective, there are three main issues to consider with
respect to the past-time reference of the simple present: the grammatical (seman-
tic) properties under which this phenomenon occurs; the pragmatic explanation
of the headline present tense as an instantiation of deictic centre projection; and
the interpretation of this usage as an internal evaluation device whereby the inter-
locutors can emphasize the relevance of a particular segment of their narratives.
Semantics of the past-time reference of the present tense
The use of the present tense with a past-time reference occurs in several common
situations. Grammars typically mention the non-present use of the tense with
verbs of communication, where “the implication of the present tense seems to
be that although the communication event took place in the past, its result – the
information communicated – is still operative” (Quirk et al. 1985: 181). Leech
(1971: 7) argues that the present tense occurs in those contexts where we typically
expect the present perfect and that it is motivated by a shift in perspective: “the
verbal meaning has been transferred from the initiating end to the receiving end
of the message. The communication is still in force for those who have received it,
and so the Present Tense is allowed”.
The second major use is the so-called historic present, which “describes the
past as if it is happening now: it conveys something of the dramatic immediacy of
an eye-witness account” (Quirk et al. 1985: 181). The historic present occurs most
commonly in colloquial speech (also with verbs of communication; cf. Biber et al.
1999: 454), and is central to some speech events and text types, such as the telling
of jokes, where it is partly a conventional linguistic form embedded in the joke
structure and partly a performance phenomenon by which the speaker increases
the suspense and immediacy of the joke telling. In such communicative settings,
the present tense represents a conventional choice of tense (cf. McCarthy and
Carter 1994: 95).
Because of the existence of headline conventions, the vast majority of finite
verb forms, projected through the institutional voice of the newspaper, are ren-
dered in the present tense in headlines. This concerns verbs of communication,
i.e. reporting verbs that introduce the verbal comments of other social actors, as
well as verbs that describe the relevant events. The following examples contain
some of the typical reporting verbs in headlines such as tell, say, admit, hail, urge,
warn, claim etc. In these examples, we can see a similar effect as in conversational
narratives containing speech-act verbs in the present tense that add to the vivid-
ness of the description (cf. Biber et al. 1999: 454–455):
Submariner admits meeting ‘spies’ (The Guardian; 13.11.2012)
European leaders hail Syria coalition (The Guardian; 13.11.2012)
Gore urges Obama to act on climate (The Guardian; 13.11.2012)
‘Sede vacante’ as Benedict XVI tells world: ‘Thank you and good
night’ (The Telegraph; 28.2.2013)
Hague slams ‘delusional’ Assad and hints at further UK involve-
ment in Syria (The Independent; 3.3.2013)
Huddleston and Pullum (2002: 131) describe a related, though distinct use of the
present tense for past-time events: the past evidential use with verbs of communi-
cation (cf. their example Your mother tells me you’re off to Paris tomorrow). They
argue that the main clause with the speech-act verb is in fact backgrounded and,
for that reason, it usually contains neither any temporal specification nor any
elaboration by adjucts. In their view,
It serves to background the communication occurrences themselves and to
foreground their content, expressed in the subordinate clause. The main clause
provides, as it were, the evidence for believing or entertaining this content. The
primary purpose is therefore normally to impart this content – or to seek confir-
mation of it. (Huddleston and Pullum 2002: 131)
The examples below indicate a similar pattern: the reporting verbs, despite show-
ing the characteristic headline shift of tenses towards the present tense, in fact
provide information that is of secondary importance. The reported clause is more
newsworthy, constituting the core of the story. The reporting clauses, which are
post-positioned, merely provide background information. Their function in-
volves an aspect of metalinguistic commentary: the reporting clauses in the
present tense indicate that the preceding textual segment is actually an utterance
ascribable to an external source, thus making the headline heteroglossic:
Wake up and cut taxes, Tories tell Chancellor
Tories may axe Human Rights Act, says Justice Secretary
(The Times; 3.3.2013)
The dramatic, eye-witness effect of the present tense with other than speech-act
verbs is apparent in other examples, cf.:
134 News and Time
Deadly snake eggs hatch in wardrobe (The Guardian; 21.12.2012)

Court quashes Hillsborough verdicts (The Guardian; 20.12.2012)
The present tense can be read as adding or emphasizing the relevance of the re-
ported information. Quirk et al. (1985: 183) identify some other contexts of its
use, such as stage directions, where “the present is used by convention, as if to
represent the idea that the events of the play are being performed before our eyes
as we read the script. A similar convention is used in summaries of narratives”.
Huddleston and Pullum (2002: 126) note that the present tense also conven-
tionally appears in the captions of newspaper photographs, where it “reflects the
permanence of the photographic record”. Declerck (2006: 191) adds that this hap-
pens when a photograph represents a certain “historical situation (e.g. The Queen
visits Saint Mary’s college in 1991)”. This is one of the rare situations when a defi-
nite adverbial indicating past time can combine with the simple present tense.
In some other contexts, the usage of the present tense in captions alternates with
the past tense. As Huddleston and Pullum (2002: 126ff.) note, “Captions to il-
lustrations in books often use the timeless present tense too (Dick Purfoy seeks
his fortune; Mr Meeking administers consolation), but a preterite is equally possi-
ble, to match the narrative text (D’Artagnan began to renew the acquaintance with
Groslow).” Viewed from this perspective, news headlines also represent a sort of
‘snapshot record’ of reality – a record of the momentary situation at the moment
of occurrence of the relevant action or event.
Historic present
Exploration of the systematic use of the present tense to refer to past-time events
cannot fail to include a more detailed consideration of the issue of the historic
present. The historic present applies, strictly speaking, to two situations: (1) the
narration of past events as if they were happening at the moment of utterance; and
(2) the narration of fictional events.
While the former concerns a relatively common phenomenon encountered
in personal narratives, the latter is described by Leech (1971: 12) as reference to
‘imaginary present time’, i.e. a fictional use adopted by novelists and writers ap-
parently “in imitation of the popular historic present of spoken narrative”. Such
fictional use is contrary to normal expectations and creates foregrounding effects.
Leech explains that “transposition into the fictional present is a device of dramatic
heightening; it puts the reader in the place of someone actually witnessing the
events as they are described” (1971: 12).
Regardless of its possible distinction into real and fictional narratives, the
historic (or ‘historical’) present is associated with the vivid description of events
(Declerck 2006) and is interpreted as stylistically marked. The phenomenon has,

of course, attracted much attention in literary stylistics and narratology (e.g.
Wright 1987; Semino 1995; Fludernik 2003; Kazunari 2009) because a change of
tense typically reflects a change in the perspective from which a story is narrated.
The alternating tense may thus accompany the distinction between the mimetic
and diegetic narrative levels, i.e. the direct enactment (showing) of an action as
opposed to its verbal description through a narrator. Mimesis may then be asso-
ciated with the present tense and diegesis with the past tense, reflecting the ‘inter-
nal’ as opposed to the ‘external’ perspective on events. Arguably, the present tense
transforms narration into ‘performance’ (Fleischman 1990: 60).
Literary scholarship offers an insightful view on the phenomenon of the
grammar of tense because what it considers to be crucial – with a view to the do-
main-specific aims of narratology – is not so much the choice of a particular tense
as the fact that a change of tense occurs. As argued by Wolfson (1978, 1979), oc-
currences of what she calls ‘the conversational historical present’ cannot be easily
replaced by the deictic past tense (unlike verbs in the true historical present). Oc-
casionally, the present tense also does not correlate with the most dramatic points
of narratives, as is often claimed. Given this state of affairs, the switch of tense in
narrative appears to be “a structural marker of a segmentation of events” (Wolfson
1979: 174). The switch need not be only from the past tense to the present, but
also the other way round: the speakers divide events in narrative by means of the
switch, not by means of the choice of a particular tense. The tense alternation “is
used to focus attention on the new action” (Wolfson 1979: 178) and is a perfor-
mance phenomenon that occurs in story-telling interactions (Wolfson 1978). The
speaker’s/writer’s choice of a specific tense – particularly one that realizes a switch
from the tensed forms used in previous sections of the text or narrative – thus
assumes an evaluative role.
Tense as an evaluation device
Thus, the mechanism of verbal tense switch enables speakers to indicate their
presence in the text, i.e. to identify certain actions as being in their subjective
centre of attention. Apart from this, there is another – more structural – sense in
which tense switch is used for evaluation: it characterizes a specific move with-
in the narrative structure of stories. Within the framework developed by Labov
and Waletzky (1967), the historic present (or rather the switch from the past into
the historic present and vice versa) constitutes one of the means through which
speakers focus their story-telling by emphasizing the key sections of their narra-
tives. The evaluation is not lexicalized (i.e. as external evaluation) but conveyed
136 News and Time
through the manner of the presentation (i.e. what Schiffrin 1981 refers to as in-
ternal evaluation).
Since tense switch is primarily oriented to the discourse participants them-
selves, it can be considered a pragmatic device. As McCarthy and Carter observe,
[t]he historical present operates as one of Labov’s ‘internal evaluation’ devices,
heightening the drama of events and focusing on particularly significant points
in the story. It is a perfectly coherent choice of tense, even though the events it
recounts occurred in the past. It brings the listener directly into the action with
the teller, and is thus a signal of interpersonal intimacy from the teller.
(McCarthy and Carter 1994: 94–95)
For this reason, the historic present tends to appear in those sections of narra-
tives that contain the narrative category of the complicating event (McCarthy and
Carter 1994: 94) because this is where the narrative account becomes dynamic.
The complicating event becomes foregrounded by means of the tense choice,
which – apart from its focusing function – draws the reader into the action. The
choice of tense is then motivated by the need to attend to the interpersonal func-
tion at both ends of the communicative dyad: the speaker uses the present tense
(or the tense switch) to highlight the crucial element in the story and, while doing
so, engages the hearer in the action by recounting the story with an in medias res
effect.
The use of the historic present in narrative is conventional, i.e. regular, and is
found in some typical registers, such as in joke-telling (cf. McCarthy and Carter
1994: 95). The conventionality of expressing the past-time event in the more in-
teractive present tense constitutes a part of the speakers’ common knowledge – a
text that conforms to implicitly shared and generally recognized discursive con-
ventions is interpreted as regular and unmarked.
However, the tense-shifting pattern found in newspapers is somewhat differ-
ent from that in oral narratives. As Harder (1995: 494) notes, “instead of first plac-
ing the story in relation to the time of speech and then switching to ‘story time’,
the headline describing the main event is often in the present, while in the article
itself the same event is in the past tense”. Within the cognitive theory of mental
spaces proposed by Langacker (1991), this is a deviation from the usual pattern
because normally “the deictic centre of the process of communication shifts from
the actual space to the narrative space”; in other words, one starts with tenses
that relate to the actual situation and switches the tense only after the narrative
space is established (Harder 1995: 483). On the textual level, the tense thus has a
foregrounding function.
In the conventional rhetoric of headlines, the phenomenon of tense switch is
also related to several other tense choices, which can be described as ‘vertical’ and
‘horizontal’. Vertical selection follows the paradigmatic axis of selection: the tense
in headline comes as a result of the editor’s choice when formulating the headline.
This shift has only one realization but two possible stages: (1) the shift into the
canonical present tense that is so dominant in headlines, and (2) the shift away
from the canonical present towards some other tense. The second pattern is as-
sociated with the violation of headline conventions and is the rule in heteroglos-
sic headlines (see Chapter 7). By contrast, horizontal tense selection follows the
syntagmatic axis: the tense choice in the headline is related to the tense choices of
the relevant verb in the lead and the body copy of the news article, i.e. the textual
segments which extend and develop the information provided in the headline.
While the verbs form a cohesive chain, their tenses differ – the linear progression
and processing of a news text from the headline through the lead to the body copy
thus incurs a shift of tense (see Chapter 10).
While these issues are treated in detail in separate chapters, let us proceed by
first explaining the effect and implications of the ‘vertical’ (paradigmatic) kind of
tense switch that results in the present tense in headlines, making that particular
textual segment more engaging and involving for the reader on account of the
construction of the illusion of a shared deictic centre.
Deictic centre projection
Let us consider the headline present tense and its effects with regard to the prag-
matic notion of deictic centre projection. In general, the presentation of events
in headlines in the simple present constructs the events as if happening in me-
dias res, lending them some of the semantic feel of the instantaneous present. If
the general news production processes, as discussed in Chapter 3, are taken into
account, then the tense actually operates on the basis of a double projection of
deictic centres (see Figure 6.3). First, there is the projection of the journalist’s cod-
ing time CT to the receiving time RT, which accounts for the shared reference of
such adverbials of time as yesterday in the main body of the text. This projection,
arising as a result of the way traditional news stories are formulated one day ahead
of the publication date of the print version of newspapers, ensures that the stories
can be viewed retrospectively from the anticipated time of reception.
By contrast, the second projection is headline-specific. In this step, the coding
time CT and the receiving time RT (already temporally unified by means of the
first projection) are projected back to the event time ET. As a result, a grammati-
cal transformation needs to be carried out that affects the tense choice: where the
first projection yields an utterance in the prototypical form ‘X happened yester-
day’, the second projection realizes the same semantic content by means of the
(quasi)-instantaneous present as ‘X happens’. Since an adverbial of time would
138 News and Time
lead/body copy
ET = CT → PT = RT
headline
Figure 6.3 Double projection of deictic centres: the simple present tense in headlines
referring to past events (arrows indicate projections; ET – event time, CT –
coding time, PT – publication time, RT – receiving time; PT and RT are construed
as contemporaneous)
unequivocally locate the event in the past, it is omitted in order to enable the
universal operation of the present tense. The tense is not contradicted by any ad-
verbial that would determine the actual time of the event and provide concrete
deictic anchorage. We see that the second projection metaphorically transfers the
reader to the event, giving rise to the above-mentioned effects of immediacy and
dramaticality.
The figure shows how the headline present tense maps onto the phenome-
non of deictic centre projection that normally operates in the process of news
discourse production. As mentioned previously, the event time ET typically co-
incides with the coding time CT on account of the daily publication frequency of
the periodical press (this temporal co-incidence is indicated by the dotted circle).
The coding time CT is then projected into the publication time PT and the re-
ceiving time RT (which are, under normal circumstances, co-temporal as well).
This projection affects the deictic coding in the article and accounts for the verbal
tense and adverbials of time in the lead and the body copy. The simple present
tense in headlines, however, redefines the entire situation and projects it back-
wards towards the event time since it transfers the reader’s centre momentarily to
the event time ET.
The headline present tense in fact achieves the unity of all the four times that
are distinctly differentiated in the news production process: the individual times
are compressed into the seemingly instantaneous present that constructs the il-
lusion that the event is unfolding at the time of reading. This impression is, of
course, merely a fiction that collapses in the immediately following structural seg-
ment (the lead), where tense is used deictically to specify to the reader the actual
temporal placement of the event.
The backward projection of the event time ET in headlines towards the deic-
tic centre is also enabled by the absence of adverbials of time. It could be stat-
ed that this absence enables verbal tense to switch its function from deictic to
non-deictic because the non-inclusion of an adverbial of time liberates the tense

from its semantic function. Under default circumstances, tense in headlines is not
associated with expressing temporality.
There appear to be little empirical data available to establish when this con-
vention in English-language headlines came into existence. In the 19th century,
headlines and the earlier category of article heads tended not to use the present
tense to refer to past-time events. Verbal tense in headlines appears to have served
a deictic function until the early 20th century, when the rise of the present tense
in headlines can be attested.21 More diachronically-oriented studies are needed
in the area of historical news discourse analysis in order to establish the gradual
process through which the present tense came to predominate in hard news head-
lines in English.
In his early study on the meaning of the English verb, Leech (1971) men-
tions some interesting uses of the historic present that may throw some light on
the possible emergence of the conventional present tense in headlines. The most
relevant appear to be recapitulations of previous instalments of serial stories on
radio, on TV, and in magazines that were – around the middle of the 20th centu-
ry – formulated in the present tense. Leech speculates about the possible origin of
such formulations as follows:
Perhaps this convention, now often superseded by the Past Tense summary, came
from the habit, adopted by novelists such as Smollett and Dickens, of giving
chapter summaries in the Present Tense in place of simple chapter titles: Chap-
ter XXI. Madame Mantalini FINDS Herself in a Situation of some Difficulty, and
Miss Nickleby FINDS Herself in no Situation at all. (Nicholas Nickleby).
(Leech 1971: 13)
The summarizing function of such utterances essentially corresponds to the tra-

ditional, semantic function of headlines to sum up the basic information about a
given news story. Just like headlines, chapter summaries are textual macrostruc-
tures that provide an overview of the main text to come. The two are very similar
because they function as introductions to what follows, be it the book chapter
proper or the text of the actual news article (i.e. the lead and the body copy).
This consideration suggests the hypothesis that such semantic summaries actually
bear allegiance to the ensuing text rather than the event itself – they operate as a
sort of textual ‘flashforward’. The present tense is then seen as an expression of the
21. This is based on an informal sampling of the full-text Times Digital Archive 1785–1985.
Studer (2003: 20) observes that summary headlines with verbal components were 19th centu-
ry innovations. For more information on 18th century headlines, see also Studer (2005) and
(2008).
140 News and Time
prospective orientation of the summarizing utterance towards the full text that
follows, rather than providing a temporal retrospective onto the event described.
There is a slight difference between the two types of initial summaries. With news
stories, readers are aware that the events reported are located in the still-rele-
vant near past, while fictional narrated events in books lack the retrospective el-
ement. In that sense, readers expect news stories with their headlines to be more
event-oriented, while chapters of fictional works are more about the act of telling
than the ex post description of past events.22
The prospective textual orientation of an initially-placed summary points
to another possible semantic aspect contributing to the meaning of the headline
present tense, namely futurity. The present tense may be related to what the reader
is about to find out about the event in the subsequent text. Apart from this, the
present tense stresses the factuality of the event, as well as the general validity of
the statement and the lack of any time restriction on this.
6.5 Future time reference of the simple present tense
The last possibility is the reference of the present tense to the future. In head-
lines, futurity tends to be expressed either through semi-modals such as ‘be to’
(cf. Carter and McCarthy 2006: 636) or certain lexical verbs such as ‘face’ (cf. Sec-
tion 7.1). These verbs, despite their present tense form, semantically refer to the
future time, encoding a present state that has a significant future orientation. The
will-future is avoided (although it does occur in headlines stressing the semantic
feature of volition and/or formulating the proposition through indirect quotes).23
The preference for the ‘be to’ future and for the lexical expression of futurity is
most likely associated with the fact that the present tense in headlines is already
laden with the additional conventional function of referring to past-time events.
Headlines are processed by readers on the basis of their background knowledge
of the conventions; hence, the occurrence of the present tense will be interpreted
as referring to a past event – unless additional contextual factors discount such an
interpretation (e.g. readers’ knowledge of the relevant facts obtained extra-textu-
ally from other media, intertextuality in the case of continuing stories, etc.).
22. There is another affinity between headlines and chapter summaries – many of the early
headlines in the 19th century were quite different from the modern minimalist information
chunks: they were descriptive narrative tellings that provided all the essential information
across several decks of headlines (cf. Bell 1995).
23. The semantic need to express volition or to create the rhetorical illusion of an accessed
voice thus overrides the headlines’ conventional succinctness motivated by spatial constraints.
6.6 Potential ambiguity of the simple present tense in headlines
Occasionally, genuine future-time reference of the simple present may clash with
the readers’ expectations about the local meaning of the tense in headlines. As
a result, ambiguities may arise. Let us illustrate this with the following example
(described in Chovanec 2000), which uses the conventional present tense in a po-
tentially ambiguous way:
Murder case nanny tells of TV ordeal in US jail
When processing the headline, the reader may be misled to assume that the pres-
ent tense – in harmony with the conventions of headlinese – refers to an action
located in the past. This assumption may be supported by the fact that the article
provides a current update on a developing story that has been in the focus of
the media’s attention for some time. Nevertheless, what can seem as a relatively
straightforward case of past-time reference of the verb (and, moreover, a speech-
act verb classifiable as a verb of communication) eventually turns out quite differ-
ently: it is a present tense verb with a future meaning. The reader will discover the
incongruity only when continuing with the text and reading the lead – the next
structural segment of the article. The lead specifies the temporal deictic situation
of the headline verb by imposing a future time interpretation on it (cf. Louise
Woodward … will be heard on television tonight…).
Obviously, we have something of a garden-path effect here, since the reader
needs to abandon the initial interpretative guess (the retrospective orientation
of tells towards a past action) and replace it with a contrary interpretation (the
prospective orientation of the verb to an action that is yet to happen). This re-in-
terpretation occurs upon encountering the will-future in the lead and becomes
confirmed with the time adverbial tonight. The future reference in the relevant
headline could easily be expressed by means of the semi-modal ‘to be’, which is
usually encountered in such a context (cf. Quirk et al. 1985: 846). That would ren-
der the surface form of the headline as *Murder case nanny to speak/is to speak of
TV ordeal in US jail.
Yet, there is a reading in which the choice of the present tense would be jus-
tified: the newspaper does, after all, report on an act of telling that happened in
the past because it is that particular verbal account which serves as the basis for
the TV programme advertised. In terms of the actual structure of the event, the
‘telling’ occurred in the past and will occur again in the near future when the pro-
gramme is aired for the benefit of the wide public. The act of telling thus has a dual
nature and the headline writer may have felt allegiance to the former aspect of the
event by coding it in the present tense, and thus implying the past-time location
142 News and Time
of the event. Intuitively, the immediacy associated with the present tense – rather
than the to-be-future – increases the newsworthiness of the story.
However, the headline is not a stand-alone unit: it abstracts the lead, which
in turn abstracts the rest of the article. In the canonical situation when the head-
line has a summarizing function, the same information is provided three times,
though with a degree of variation. The same holds for the temporal location of
the main event described: it is expressed in the headline, the lead, and the body
copy. The crucial segment for determining the real-time location of the story is
the lead – the tense used in this segment is truly deictic, unlike the headline tense,
which is conventionally shifted and non-deictic. Thus, since the lead in this par-
ticular example specifies the time of the ‘telling’ as future (‘will be heard … to-
night’), the corresponding verbal phrase in the headline (‘tells’) is likely to have
the same future reference despite the fact that it is given in the present tense.
The future reading of the verbal chain spanning across the three main struc-
tural segments of the story seems to be supported by the relevant formulation
in the body (‘[she] describes how fellow inmates forced her to watch a television
programme about children being abused’). The present tense (‘describes’) in the
text is related rather to the act of telling within the TV programme, of which the
newspaper gives a textual preview, than to her original speech act. The present
tense on the verb in the body copy of the article is thus another illustration of the
present tense with a future reference, or the habitual present used for relatively
timeless statements (cf. Quirk et al. 1985: 179).
Let us contrast the example above with another headline that uses exactly the
same form of verb in the present tense:
Kidnapped US reporter Richard Engel tells of dramatic escape in
Syria (The Times; 19.12.2012)
In this news headline, the verb ‘tells of ’ refers to an event in the past. It describes
a past-time telling of the reporter’s experience during a previous televised broad-
cast. The text of the article even provides a precise temporal anchorage by means
of the deictic time adverbial ‘yesterday’ (After his dramatic escape, Mr Engel told
NBC Today show yesterday that he and his team were captured…). This points
to an interesting behaviour of the verb ‘tell of ’ in headlines: in one of its uses, it
is associated with the media appearance of news actors, mediating their verbal
comments that have either been made in the past or are about to be aired by other
media outlets. In another of its uses, it has an evidential function involving an
implicit subjective evaluation of two phenomena that are presented as related.
The approximate meaning of the verb is then ‘indicate’, ‘stand as evidence’, etc.
(cf. the headline Emails tell of fears over EADS payments; Financial Times;
14.8.2012).
Manipulation of temporal deixis
As the last examples show, there is some potential for the manipulation of tem-
poral deixis in various segments of news articles. Bell describes the phenomenon
in leads, where adverbials of time are frequently modified in order to enhance the
recency of the story:
The time and place deictics which occur in the lead paragraphs of international
stories – here and today – are traps for the unwary. Copy editors can be tempted
to enhance the recency of slightly stale news by taking liberties with the time
specification. I have seen stories where time adverbials have been updated from
last night to today to late today as they were edited… with each version claiming
greater recency as the news in fact receded into the past. (Bell 1991: 226)
Such an intentional enhancement of recency is one of the ways in which the writ-
er may increase the appeal of a story in the process of its discursive construction.
As a primarily reader-oriented strategy, it is motivated by increasing the rele-
vance of the story for the reader. Interestingly, such deliberate temporal manip-
ulations can be seen in a similar light to the conventional change of the past (or
the present perfect) tense into the simple present tense in headlines. On a certain
level, this strategy even surpasses the need for the paramount value in journal-
ism, namely objectivity in the sense of faithfulness to the known facts about an
event. The interpersonal dimension becomes dominant, while the ideational one
is backgrounded.
As shown above, the choice of the present tense in headlines is related to the
editor’s redefinition of the discourse space, with the tense being used non-deicti-
cally in order to create the conventional impression of a shared temporal situation.
However, the modification of tense affects not only temporality but also modality.
In general, verbal tense combines the deictic function, whereby an utterance is
located “in the history of the discourse”, and a modal function indicating the “‘re-
ality-, certainty-, actuality-status’ of the ideational part of the utterance” (Kress
1977: 43). While the deictic temporal anchorage of tense is a matter of the ide-
ational function, the modality component concerns the interpersonal function.
As argued in this chapter, the deictic function of tense is weakened in headlines –
as well as in other situations where conventional shifts of tenses are present. By
contrast, the modal component of tense, particularly the actuality-status of the
utterance, is enhanced by the use of the present tense. This is obviously related to
another linguistic feature of headlines: the absence of epistemic modality markers
expressing the various degrees of probability. By rendering the proposition ex-
pressed in the headline through the conventional present tense, the modality of
certainty is implied even with respect to such propositions whose reality status is
144 News and Time
not certain. In other words, some modally hedged possibility utterances, which
can be found in leads and the body copy of articles, are expressed as factual cer-
tainty utterances in headlines as a result of the levelling effect of the simple pres-
ent tense. This conforms to Conboy’s observation (2007: 64) about the delayed
introduction of modality in many news stories, and it also attests to the existence
of a specific headline rhetoric: the newspapers’ unwillingness to commit to the
certainty of some propositions is actually overruled in headlines, at least as long
as modality is not expressed through some accessed, non-authorial voice. Thus,
the pragmatic function evidently prevails over the semantic one.
This issue is revisited in Section 8.3 on modals and in Chapter 10 on tempo-
ral deixis beyond headlines, because modality, in addition to temporal deixis, is
one of the linguistic categories that undergo a transformation at the transitions
between the three major structural components of news stories, namely the head-
line, the lead, and the body copy.
Chapter 7
Other tenses in headlines
After documenting the forms and explaining the functions of the present tense
in headlines in the previous chapter, this chapter complements the temporal sit-
uation in English headlines by considering other tenses and temporal domains.
First, the linguistic coding of futurity will be discussed, followed by the present
perfect tense and the past tense, which stand in contrast to the headline present.
This arrangement is adopted partly because the expression of futurity anticipates
some of the phenomena that we will see in full operation with the other tenses,
and partly because the complex ways in which the future temporal zone is re-
ferred to overlap with such issues as the expression of modality briefly mentioned
in the previous chapter.
In the rhetoric of headlines, which privileges the position of the simple present
tense, the choice of some other tense than the present typically implies a marked
selection from the options available. As will be argued again in the section on
the present perfect tense and the past tense, deictic use of tense can have either a
semantic function (when it encodes the pastness and non-recency of some event)
or a pragmatic function (when it is realized as a result of the encoder’s selection
from several possibilities). In some situations, the choice of tense also appears to
have a more macro-structural motivation, being related to the mutual hierarchy
of articles within an article cluster.
The main argument offered here rests on the contrast between the situations
in which the conventional present tense, the past tense and the present perfect
tense occur, particularly in combination with other instances when the conven-
tions of headlinese are broken. As documented in the present chapter, a crucial
role is played by heteroglossia: it is argued that access to other voices in headlines
provides an opportunity to discontinue the conventions and turn to tenses other
than the present simple. Such a secondary tense switch – away from the headline
conventions – is likewise interpreted as having a pragmatic motivation, although
it marks the switch from non-deictic present tense (as the default category in
headlines) to the deictic past tense.
146 News and Time
7.1 Expressing futurity
While headlines need to refer to the entire temporal conceptual domain, refer-
ence to future time is comparatively less important than reference to past time.
Yet, futurity is encoded because the media do not merely present information and
pass judgment about news stories that have already happened but they also make
predictions and provide information about upcoming events that are planned.
Reference to the future can be realized in a number of different ways. Being a
problematic category in the English tense system, the future is typically realized
through several conventional means. In addition to the present tense with a fu-
ture meaning that was discussed in the previous chapter, these means include the
will-future, semi-modals and full modals, as well as certain verbs that express fu-
turity lexically and thus partly overlap with the future-orientation of some head-
lines in the present tense. The following examples illustrate some of the forms
typically used in headlines to express futurity:
Drunk Tweeters will escape court (The Telegraph; 19.12.2012)
Floods set to continue after Christmas (The Guardian; 21.12.2012)
Prescription charges to rise to £7.85 (The Independent; 1.3.2013)
Makes a change from sugar lumps: Man faces trial for feeding
police horse a sausage roll (The Independent; 16.11.2012)
The relatively rare occurrence of the will-future in headlines is one of the most
noticeable features of headlinese. In my data set, there are only 9 instances of
will-future in single-clause finite headlines (i.e., 1.7%, n = 647), though the form
appears to be more frequent in complex sentence headlines, with 6 occurrences
(i.e., 5.2%, n = 115; the figure drops to 2.7% if the total number of finite clausal
units in such complex headlines is counted, i.e. n = 225). Regardless of headline
type, the 15 combined occurrences make up 1.8% of all headlines in the data set
(n = 837).
One of the reasons for the avoidance of the modal auxiliary may be sought
in the analytical character of this construction (commonly known as the simple
future tense). Because of the traditional spatial constraints resulting in the highly
compressed form of headlines, there is a pressure to reduce the verbal phrase to a
single word since the combination of two verbs (the auxiliary will and the infini-
tive of the lexical verb) takes up valuable space. The lexical component of the verb
phrase is, in terms of information content, more important than the modal auxil-
iary, which conveys temporal and modal specifications that can be conventionally
presupposed or conveyed through some other means (this feature is shared with
the passive, which is, likewise, commonly ellipted in headlines). As discussed
Chapter 7. Other tenses in headlines 147
previously, time need not be specified in headlines because of the assumption of

a shared temporal framework – an assumption that is based on the anticipation
of the recency (topicality) of a given issue of a newspaper. Another reason for the
avoidance of the will-future may consist in the fact that the will-modal retains a
strong aspect of the original meaning of the lexical verb from which the modal
developed. The meaning is connected with the intersubjective expression of the
speaker’s volition, promise, instant (unpremeditated) decision etc. The will-future
thus, in most cases, fails to convey pure futurity, which renders it less suitable for
use in headlines, unless, of course, the subjective meanings related to volition etc.
are needed. The will-future is then likely to appear – and it does, particularly in
those textual segments that constitute reported speech. The reappearance of the
will auxiliary is another effect of heteroglossia on the form of headlines (cf. Sec-
tions 7.2 and 7.3 below).
To-future
The subjective aspects of the will-future can be partly avoided by the conventional
reference to future time with such semi-modals as ‘be to’. As Carter and McCarthy
observe, “Be to is common in specific styles such as newspaper headlines and news
reports. In headlines it often occurs in an ellipted form” (2006: 636). Grammatical-
ly, the form is cast in the present tense (e.g., ‘The government is to announce…’)
but semantically it refers to the future. Since ellipsis of the finite forms of the verb
‘be’ is the rule in news headlines (there is no occurrence of the finite form of the
construction ‘be to’ in the data), this phrase is typically realized formally as a
non-finite construction with an underlying clausal structure. The corresponding
finite clause can be easily reconstituted by inserting the missing auxiliary, which
results in a full clausal structure. Cf. the following example where the non-finite
headline can be easily recast in its full non-ellipted form (‘…is to spy…’):
Government to spy on computers of the jobless
As a result of the ellipsis of the finite verb form in verbal phrases expressing the
future with the ‘be to’ structure, the main lexical verb is realized as an infinitive.
The expression of temporal subsequence is then taken over by the infinitival par-
ticle, which is a common grammatical device for referring to the future (cf. Quick
et al. 1985: 846, who note that “to is commonly used to express the future or a
predicted arrangement”).
However, the ellipsis of the finite verb form turns the ‘be to’ construction into
a non-finite clausal sentence structure that has to be distinguished from other
148 News and Time
possible uses of the infinitive in headlines, most notably as a form that either com-
plements an immediately preceding nominal phrase or forms part of an extended
verbal phrase. The first situation is represented in the following nominal headline:
Merkel threat to scupper summit over Cameron veto
(The Times; 22.10.2012)
Here, the headline starts with the nominal phrase Merkel threat, which refers to
a newsworthy event located in the recent past: it is the prior announcement by
Merkel that provides the actual basis for the news story. The nominal phrase is
followed with the infinitive to scupper, which refers to a future result of Merkel’s
previous verbal behaviour (i.e. ‘Merkel will (possibly) scupper a summit’). At the
same time, the infinitive deletes the hypothetical nature of Merkel’s future act
by expressing the modality of possibility through the noun threat rather than by
means of a modal verb prefacing the verb that would clearly refer to a future pos-
sibility (*‘Merkel may scupper summit’). In this way, not only is the speech-act
nature of the past event emphasized but the newspaper also has the opportunity
to select an expressive word that can maximize the impact of the story by increas-
ing its news value. It is also worthy of note that the formally nominal headline
could have been easily formulated as a clausal structure, which would add further
urgency to the proposition because the original speech act would be reported as a
dynamic process (cf. *Merkel threatens to scupper…) rather than a static product
(Merkel threat to scupper…).
A similar combination of a nominal phrase (calls for) with an infinitive ex-
pressing a future wish (to step down) is present in another sample headline that
is – once again – fully nominal in structure and can be semantically interpreted
as an existential proposition:
Calls for Georgia president to step down (The Independent; 3.10.2012)
In addition to nominal constructions, the to-infinitive is commonly used to refer

to the future in complex verb phrases and other constructions. Cf. the following
headline:
Russians told to curb corrupt phrases (The Guardian; 7.3.2013)
Here, the infinitive complements the past participle in what is technically a con-
densed non-finite object clause, since the phrase to curb corrupt phrases is the
direct object of the preceding verb. Since the indirect object of the verb ‘tell’ (Rus-
sians) is placed in the subject position as a result of the passive transformation,
the semantic agent of the proposition (i.e. the subject of the corresponding active
construction) is left unexpressed. Since the passive is conventionally ellipted into
the residual non-finite past participle, the proposition is syntactically condensed

even further. The headline thus expresses two events: the past-time event of ‘Rus-
sians being told’ and the future-time event of ‘curbing corrupt phrases’, both ex-
pressed through non-finite means. Such a combination of two events, located in
different temporal zones, is the rule in both nominal and verbal phrases followed
by an infinitival construction because the infinitive expresses the posteriority of
one event after another. In all the examples above, the infinitive particle ‘to’ acts
as a sequencer, i.e. a “sign of action after a specified point” (Kress and Hodge
1979: 129).
Modal auxiliaries
As some of the examples above indicate, futurity can also be expressed by means
of modal verbs denoting various deontic and epistemic modal meanings, typical-
ly various degrees of possibility of future events. Though the appearance of modal
verbs in headlines is less common, it is not exceptional.
First of all, let us mention the situation when a proposition formulated in
a headline is not modally hedged, although its reality-status concerns a future
possibility rather than a fact. This involves many headlines formulated in the
conventional present tense, which connotes – through its default grammatical
meaning – the modality of certainty. The reading of possibility in such headlines
is revealed in the subsequent textual segment (i.e., the lead), which contains a
parallel formulation with a modally hedged utterance. The rendering of the cat-
egory of possibility as certainty in headlines is a part of headline rhetoric. The
reality-enhancing present tense is used in order to increase the relevance and im-
pact of the story with respect to the reader, even though it misrepresents the real
state of affairs. It is thus, once again, associated with enhancing the interpersonal
dimension of headlines. This time, however, the enhancement is not in addition
to, but at the expense of the ideational dimension because an important aspect of
meaning is lost in the process.
However, there are also situations when newspapers formulate their head-
lines with a view to the ideational rather than the interpersonal axis of communi-
cation. This means that the semantic content of the relevant headline outweighs
its pragmatic function of increasing the relevance of the story to the reader. As a
result, modal auxiliaries are included in some headlines in order to specify the re-
ality status of the proposition expressed in the opening segment of the news story.
The following examples include modal verbs that express epistemic modality, i.e.
specify the degree of the possibility of some action occurring. In each case, the
relevant degree of modality is assessed by an external observer whose identity is
150 News and Time
not revealed in the headline, even though the propositions appear to express a
‘general’ possibility, cf.:
Ranking may harm ‘difficult’ patients (The Telegraph; 18.12.2012)
Italy could be forced to leave euro (The Telegraph; 19.12.2012)
UK may never recoup £66bn spent to save banks
There are also headlines expressing deontic modality, i.e. obligation, necessity, etc.
Utterances that contain this kind of modality are typically personal, emanating
from a speaker who has some authority or power over the process contained in
the modally hedged proposition. Since the following headline does not assign the
proposition to an external source, it might appear to be expressing the view of the
paper, i.e. it is as if the paper was formulating its policy towards the issue at hand:
Britain must not ‘undo’ EU (The Telegraph; 28.2.2013)
Yet, the inverted commas enclosing the lexical verb (undo) in the headline indi-
cate that there is a trace of an external voice accessed. Indeed, the deontic modal-
ity in the headline is not expressed by the voice of the paper but an external news
source (namely the EU’s commissioner for economic and monetary affairs). That
is made clear in the subhead to the article:
Britain is better off inside the European Union, Olli Rehn has said, as he urged
the UK to support the rebuilding of the 27-nation bloc rather than seek to “undo
our Community” by heading for the exit.
However, the situation is even more complex than that. Although the external
source did express the subjective opinion on the advisability of Britain’s actions
with respect to the EU, the source actually did not use the modal verb must in his
utterance. Rather, the deontic modality expressed in the headline is the paper’s
own interpretation of a statement issued, in a more oblique manner, by the exter-
nal voice. The formulation of the headline then consists, after all, of the paper’s
words, though they are ultimately ascribed to an external source. The external
voice is not, however, to be held accountable for the exact formulation of the
headline – that is, in harmony with headline rhetoric, understood to be a part of
the paper’s licence taken when formulating the story’s summarizing abstract.
The presence of modal verbs in headlines is thus one of the signs of heteroglos-
sia, i.e. the mixing of voices within a news text. Heteroglossia can be either im-
plicit, as in the example above (save for the tell-tale sign of the inverted commas),
or explicit. The latter occurs when the external source is either openly acknowl-
edged in the headline (e.g. through naming or some other representation) or made
evident in the way the headline is formulated, e.g. as a recognizable extract of di-
rect/indirect speech. Cf. the following examples:
Risky banks should be split-up, says parliamentary Commission
Mensch: Corby by-election would be my fault
Prime Minister admits Britain might decide to leave EU
Judge overrules mother: son aged seven must have cancer opera-
tion (The Independent; 19.12.2012)
In all of the examples above, the modal verbs occur within the reported clause
that is clearly attributable (and attributed) to the external accessed voice, which
is thus made explicitly responsible for the expression of modality within the sub-
ordinate clauses. The first two examples represent standard ways of introducing
external voices through the speech act reporting verb ‘say’ (…says parliamenta-
ry Commission) and the colon structure with the same function (Mensch: Corby
by-election…). The third example lexicalizes the reluctance on the part of the ex-
ternal voice (cf. admits) to articulate the modally hedged proposition in a head-
line, where the modality of possibility itself constitutes the crucial element of the
whole news story (Prime Minister admits Britain might decide to leave EU). The last
example above is remarkable as well since it is representative of a headline-formu-
lating strategy clearly observable in The Independent over the period of analysis:
the headline has a dual structure where both parts express essentially the same
content. Here, the first part of such an equational headline provides a summary
of the news story (Judge overrules mother) and the second part is a quasi-quote
that can be construed as an encapsulation of the judge’s decision (son aged seven
must have cancer operation). Once again, we can see that the key to understanding
of the grammatical composition of the headline rests in its heteroglossic nature,
since the presence of the other voice allows for the introduction of language forms
that would be very unlikely to appear in the authorial voice of the paper.
Will-future
While all of the modal verbs in the previous section have a future interpretation,
the future is probably most commonly thought of as being expressed by means of
the auxiliary verb will. The structure is known as the future simple tense, although
the auxiliary retains a modal meaning, typically expressing prediction. The fol-
lowing examples contain the future simple tense but they do not express pure
152 News and Time
futurity: they refer to predicted future action. In the first example, the auxiliary
will seems to convey a sense of an almost deterministic inevitability; in the second
it is associated with prediction constituting the genre of the weather forecast:
Ash ban will get a frosty reaction from stove owners
(The Times; 19.12.2012)
Respite, then more rain will hit UK (The Independent; 21.12.2012)
In some uses, the auxiliary will is associated less with prediction than with a fu-
ture course of action likely to ensue as a result of some decision that forms the
basis of the news story. Such a resultative use of will can be observed in the fol-
lowing headlines:
Drunk tweeters will escape court (The Telegraph; 19.12.2012)
‘Shameful’ chief constable will keep his pension
In all of the examples above, the modals are used within headlines that are cast
as the voice of the paper. It appears crucial to distinguish, when considering the
formal make-up of headlines, between the authorial and non-authorial voice (see
below) because the access in the headline to an external (non-authorial) voice has
the effect that headline conventions are suspended (or discontinued). As a result,
any segment of the headline that is cast in the non-authorial voice – be it formu-
lated as a direct or indirect speech quote – may use linguistic features that are not
associated with headlinese. As illustrated above, the heteroglossia affects the use
of modal auxiliaries, as well as the will-future that is used in the external voice to
refer to the future time instead of the stereotypical ‘to be’ future or infinitives, cf.:
We will sue for asylum, say Afghan interpreters
(The Times; 19.12.2012)
Britain will be weaker without EU, says USA (The Telegraph; 19.12.2012)
Weidmann: Greece will need more debt relief
Lexically expressed future
Futurity can also be expressed lexically rather than grammatically. There are sev-
eral verbs that serve this function in headlines, such as ‘face’, ‘set out’, and ‘(be) set
to’. These verbs are particularly interesting because the brief monosyllabic words
satisfy the headline requirement for vivid language and boost the immediacy of
the reported event. Semantically, they achieve the compression of the discourse
space by means of the shift from the future into the present, cf.:
Carney faces grilling over holiday with Brison (The Times; 18.12.2012)
Floods set to continue to Christmas (The Guardian; 21.12.2012)
Many of the news reports are also based on some past-time event, typically some
speech act of announcement, etc. That aspect of the story, however, is back-
grounded because what is newsworthy is actually the potential future event, as in
the examples above.
Another situation when the temporal domains are interconnected concerns
headlines with verbs that describe the initial stage or the commencement of some
action, such as ‘start’, ‘launch’, ‘open’, ‘set off ’, etc. While the actions they denote are
typically located in the past (and the use of the present tense for the coverage of
the past-time events is thus fully regular and conventional), they co-occur with
some other event that is to appear in the future or is already in progress in the
present. This is the case in the following headline, where the noun probe has such
a future reference:
Government launches probe into Comet collapse
The nominalization in the headline (probe) refers to a future course of action that
is newsworthy just as is the fact of the commencement of the formal procedure
(Government launches). The example indicates that headlines can sometimes be
complex in terms of the temporal relations they express, particularly concerning
which aspect of the story they focus on. While the headline above foregrounds the
process verb ‘launch’ in the present tense (referring to a past event) and simulta-
neously backgrounds the future process contained in the nominal form ‘probe’,
the lead and the body copy of the article have a complex temporal structure that
expresses various temporal aspects of the news story: present (Vince Cable’s de-
partment has launched an investigation into the demise of Comet…), past (…as it
emerged that OpCapita and its backers charged Comet millions for fees despite the
chain racking up losses.), and future (The probe will be conducted through the Com-
panies Investigations Branch at the Insolvency Service, which is part of Mr Cable’s
Department for Business, Innovation and Skills.).
The following examples illustrate yet another way of expressing futurity lex-
ically, this time with the phrase ‘be up for st’. The phrase is realized in the first
headline as non-finite since the copula verb is conventionally ellipted. The second
headline expresses the future time lexically with the phrasal verb ‘set to’. The two
154 News and Time
headlines refer to the same story, with the shorter version appearing on the news-
paper’s home page with the article preview and the longer version on the web page
with the full text of the article itself:
Arise, Sir Wiggo: Cyclist up for knighthood (The Telegraph; 18.12.2012)

Tour de France and Olympic champion Bradley Wiggins set to be
knighted in New Year Honours list (The Telegraph; 18.12.2012)
Last but not least, future time can also be referred to implicitly by means of
non-finite clauses. In the following example, there is no trace of the verbal phrase,
which has been fully ellipted. Readers are able to reinstitute the copula verb ‘be’ in
the grammatically correct form (‘are’ or, even better, ‘will be’), and assign future
time reference to the sentence fragment on account of their background cultural
knowledge. The ‘Afghan withdrawal’ describes a proposed future action to be tak-
en by the UK administration:
British troops ‘most vulnerable’ during Afghan withdrawal
For the sake of illustration, let us complement the headline with the lead of the
article, which explicitly specifies the future temporal placement of the action by
means of the will-future, cf.: British forces will be “at our most vulnerable” to a Tal-
iban attack as they carry out David Cameron’s withdrawal from Afghanistan, the
Chief of the Defence Staff has said.
7.2 The present perfect
The present perfect makes for a particularly interesting choice of tense in head-
lines, even though it is rather infrequent – the data set reveals only 11 occurrences,
i.e. 1.3 per cent of all finite clausal structures. Out of that number, 8 cases are used
within the authorial voice, i.e. those segments of headlines that are uttered in the
voice of the paper, and 3 involve heteroglossic headlines. One of the reasons for
the low frequency may be sought in the length of this analytical structure: as is the
case with other analytically formed tenses, the verbal phrase in the present perfect
includes two or three verbs, only one of which conveys lexical meaning. If the verb
phrase was to be included in the headline in full, the inclusion of auxiliaries that
convey merely grammatical meaning would go counter to the convention of head-
linese that requires economy of expression. Space in traditional headlines is almost
too precious to be wasted on an extended analytical verb phrase since the presence
of auxiliaries – unless conveying information that is essential to the story – would
dilute the semantic compactness of the headline. However, as shown in Chap-

ter 10, the present perfect tense has a paramount importance in other segments of
news texts, particularly the lead, where it often features as a tense into which the
headline simple present becomes transformed. Thus, although the present perfect
has a marginal role in headlines, it is clearly a component part of the rhetoric of
entire news texts, i.e. when the choice of tense is considered syntagmatically across
the individual structural components of news articles.
Headlines marking trends and changes
The situations in which the present perfect appears in headlines can be classified
into several types. In the following examples, the full verb phrase is needed be-
cause the tense expresses newsworthy information: the headlines report on the
change of some situation, contrasting the past and the present.
Chocolate bars really have shrunk as makers cut costs
Machynlleth: a pretty market town that has lost its innocence
In the first text, the contrast is emphasized with the adverb really, anticipating the
readers’ assumption of the current situation. The second example uses the present
perfect in a relative clause that supplies a characterization of a provincial town,
without revealing the precise reason for the transformation of the town. The lat-
ter headline exemplifies an indirect strategy to formulate headlines that defocus
from the main event to some other issue, in this case the impression of the ‘loss
of innocence’.
Another type consists of headlines that report on long-term trends. They
come very close to the type identified above because they are concerned with the
present relevance of a particular situation that is being contrasted with a previous
state of affairs. Once again, the present perfect is used to discursively map change
over time. This is in harmony with the grammatical meaning of the perfect aspect,
which refers to a situation (or result) that “began in the past and continues up to
the present” (Biber et al. 1999: 460):
How Apple has performed since Steve Jobs’ death
How Christmas television has changed (The Telegraph; 21.12.2012)
It appears that headlines with the present perfect have a tendency to be more
narrative. The two headlines above, for instance, are formulated with the question
156 News and Time
word ‘how’ to indicate that the news story will not only describe the changed
situation but also discuss the causes and reasons. The same syntactic pattern and
narrativity occur in another headline in the data set. Here, the present perfect
encodes a single event rather than a prolonged trend, although it does contain an
implicit contrast between the current and the past situations:
Why Strictly has floored the X factor this year
Heteroglossic headlines
The narrative nature of the headlines above links them to the circumstances un-
der which headlines show an increased incidence of such forms as modal verbs,
non-present tenses, etc. that depart from the conventional forms of headlinese
under the influence of heteroglossia (cf. Section 7.3). The present perfect is also
used in non-authorial textual segments, i.e. those that provide access to some ex-
ternal voice by means of a quote or through an intertextual reference to the verbal
reaction of some news actor. The paper thus breaks the convention of the present
simple with important interpersonal consequences: On the one hand, the present
perfect allows the writer to refer to the past time and, on the other, it indicates a
direct relevance for the present time by retaining the present-time status of the
event. The direct quote (or pseudoquote) marks the access to another voice, pro-
viding the illusion of an authentic utterance produced by a direct witness. This re-
sults in the evocation of a seemingly unmediated discourse where conventionally
non-shifted tenses may be used:
Freddie Starr: I have never touched an underage girl
The present perfect allows the speaker to talk about an event located in past time
by pointing out its direct relevance for the present time. This is the ‘current rele-
vance’ meaning of the tense, commonly identified in grammars (cf. Huddleston
1969: 783; Huddleston and Pullum 2002: 143; Declerck 2006: 347). The present
perfect is an important tense because its main focus is on the present – the choice
of tense therefore anticipates the speaker’s subjective presence in the frame of the
speech event: “The present perfect involves reference to both past and present
time: it is concerned with a time-span beginning in the past and extending up to
now” (Huddleston and Pullum 2002: 143). By contrast, the past tense excludes the
‘now’ component because it does not construct the temporal pre-present zone in
which the reported event and its situation time are located.
Dual headlines
The present perfect constitutes a departure from the strict headline conventions
requiring the tense to shift towards the simple present. Hence, it makes headlines
more narrative and naturalistic. This tendency is observable in yet another sub-
type of headlines that use the tense in headlines with a dual structure. One of the
parts in such headlines is clausal and cast in the present perfect (potentially in an
external or quasi-personal voice), the other part is clausal or nominal and pro-
vides an extension of (or introduction to) the other part. The headline can then
have an internally dialogic structure, counterpoising the voice of the paper and
some other – even unacknowledged – voice within the headline. Consider these
two examples:
Has Dettori’s career fallen at the last? Jockey faces lengthy sus-
pension after drug test (The Independent; 13.11.2012)
Now the boss has logged out: mystery surrounds Windows chief’s
Microsoft departure (The Independent; 13.11.2012)
Both examples are remarkable in that the second parts of the headlines use the
canonical present tense and both could stand quite independently as well-formed
and complete headlines (Jockey faces lengthy suspension after drug test and Mystery
surrounds Windows chief ’s Microsoft departure, respectively). The initial clauses
that use the present perfect merely set the scene: the first is a question speculating
on the outcome of the main event, and the second is an utterance that establishes
a general temporal context for the main event by using a witty instance of word
play (cf. the metaphorical use of ‘log out’ as a synonym for ‘leave a company’). The
more personal narrative nature of the initial segments is also reflected in the use
of determiners (cf. the definite articles in at the last and the boss). Such headlines
are thus further evidence of not only the dynamic dual pattern of headlines that is
increasingly found in some newspapers, but also the papers’ trend towards greater
colloquialization. Instead of providing the mere facts in headlines, they provide
the facts with an interpersonal twist: a little speculative commentary, some per-
sonal involvement, a touch of humour, etc.
Let us add that the quasi-narrative use of the present perfect in headlines
should be distinguished from the so-called hot news interpretation that ap-
plies to slightly different situations. The hot news reading of the present per-
fect concerns sentences where “the bygone actualization of the situation in
question [is presented] as very recent and as having high current relevance”
(Declerck 2006: 782), i.e. where the simple past tense could be used. In news texts,
the hot news interpretation applies mainly to the present perfect in leads, where
158 News and Time
it provides a grammatical bridge between the present tense in the headline and
the past tense in the body copy, with all tenses referring to the same event (see
Chapter 9).
7.3 The simple past tense: From heteroglossia to information flow

management
The simple past tense enjoys a similar status as the present perfect: it constitutes a
marked choice of tense in headlines, where the default (unmarked) situation pre-
fers the simple present. The past tense typically entails some special interpretation.
It significantly correlates with the presence of external voices since heteroglossic
stretches of headlines result in the discontinuation of headline conventions (cf.
Chovanec 2005b: 73).
The following examples illustrate some of the typical uses of the headline past
tense that serve as the basis for the identification of several distinct functions.
These include: pointing out the non-recency of the event (the first example),
backgrounding a non-recent aspect of a story (the second example), marking an
accessed voice (the third example), and identifying less important headlines in
satellite articles (the fourth example). The past tense is thus connected not only to
heteroglossia but also to the management of information flow in that it correlates
with non-central news stories.
Report alerted government to West Coast flaws
The thieves who stole Christmas at Great Ormond St: Presents
destined for sick children stolen (The Independent; 19.12.2012)
We overcame the odds, says relieved Clegg (The Times; 1.3.2013)
Catherine Deveney on how she broke the story
(The Guardian; 3.3.2013; an accompanying article to the main news item
which is headlined Cardinal O’Brien admits sexual misconduct)
The categories rely on a shared semantic feature associated with the past tense: the
indication of the non-recency of the event reported. The analysis shows, however,
that there are several subtle differences in how the non-recency is conventionally
exploited, which warrants the identification of several distinct categories. At the
same time, there is another dimension that cuts across the categories identified,
namely voice. As mentioned earlier, it is crucial to specify the identity of the voice
behind the actual use of the tense in headlines. The choice of tense has different
implications if the relevant tensed verb is uttered in the paper’s own institutional
voice (which tends to impart the information in an impersonal manner) or in the

non-authorial accessed voice of some other social actor describing the events in a
direct/indirect quote or a pseudo-quote. Let us first briefly illustrate the differenc-
es between voices in headlines and then discuss the semantic and pragmatic cat-
egories that capture the strategic use of the past tense in English news headlines.
The past tense in the non-authorial accessed voice
The simple past tense is a marked choice of tense in headlines that are cast in the
authorial voice of the paper. By contrast, however, there is noticeable regularity
in its usage in those stretches of headlines that are clearly attributable to some
external voice that is explicitly or implicitly acknowledged by the newspaper. The
externality can be marked by means of inverted commas in the case of direct
speech quotations, or through the presence of reporting clauses and other devices
(such as the colon structure) that introduce indirect speech. Let us illustrate some
of the means of indicating accessed voices with specific examples:
– Directly accessed voice

Hillsborough force ‘faked miners’ strike evidence’
(The Times, 22.10.2012)
Philpott wife ‘performed sex act on co-accused to keep him onside’
after fatal house fire (The Independent; 1.3.2013)
– (Quasi)-direct speech in dual headlines

WikiLeaks whistleblower comes clean: I just wanted to start
debate about military (The Independent; 1.3.2013)
– Indirect speech with external voice attribution (prepositioned / postpositioned)

MP claims Smith abused young boys (The Guardian; 13.11.2012)
Moran fiddled expenses, jury rules (The Guardian; 13.11.2012)
– Echoic use of the external voice (e.g. through scare quotes or the verbatim
representation of the encoding of some event – cf. the formulation ‘went well’
in the last two examples)
House fire parents ‘rehearsed’ how they would save children
Surgery ‘went well’ for cancer boy in legal dispute but mother
fights against radiotherapy (The Independent; 20.12.2012)
Life saving surgery on brain tumour boy ‘went well’
160 News and Time
The examples illustrate the main formal means through which a segment of text
can become attributed to an external source. Since heteroglossia is central to our
understanding of the temporal situation in headlines, it is discussed in a special
subsection below, where the past tense is interpreted as a specific marker of an
accessed voice. Prior to this, however, let us deal with the following situations: the
occurrence of the past tense in the authorial voice of the paper, its operation as a
marker of non-recency, and its function as a backgrounding device in structural-
ly more complex headlines. Although there are significant overlaps between the
categories, they are listed separately for the sake of clarity.
The past tense in the paper’s authorial voice
Less commonly, the past tense is used within the authorial voice of the paper. In
contrast to the previous subtype, the relevant headlines are entirely monologic
and monoglossic. In the following texts, the past tense is used for an emphatic
assertion of the reality status of the main event (Moran DID cheat in the first
headline) and as a way of defocusing away from the process encoded through the
verb (killed and died in the other two examples).
Ex-MP Margaret Moran DID cheat £53,000 of expenses
Drink-driver killed model (The Times; 13.11.2012)
Baby died after GP’s ‘lack of urgency’ (The Times; 1.3.2013)
Arguably, the past tense helps to refocus the headline towards some other aspect
of the story, which – in the headlines above – is the fact that the victim was a
model (the second case) and the negligent cause of death (the third example).
The action expressed by the verb is thus not the most newsworthy element of the
story; the most important aspect of the story consists, rather, of the circumstances
of the action. This defocusing function of the past tense is particularly evident in
the case of such verbs as ‘die’, which are very typically cast in the present tense in
headlines since they are at the core of many news stories whose primary focus is
the mere fact that particular individuals pass away at a certain age or that acci-
dents bring about large numbers of deaths. The interpretation of the past tense
as a backgrounding device in the last example, however, is also motivated by the
non-recency of the death, since the main event (GP’s ‘lack of urgency’) was only
revealed in a subsequent inquest. In the next example, the death does fall within
what constitutes the recent past, and the relevant verb could genuinely have been
formulated in the present tense:
Public schoolboy died inhaling laughing gas with friends

(The Times; 5.10.2012)
It is evident that the editor who formulated the headlines decided not to use the
conventional present tense on the verb ‘die’, which would have been a perfectly
legitimate choice of tense in that context, and instead opted for the simple past –
most likely because of the presence in the headline of another element that may
have been perceived as locally more important. The story is thus not newsworthy
merely on account of the tragic death of the schoolboy (Public schoolboy died…)
but, rather, the manner in which the death occurred (cf. …inhaling laughing gas
with friends). Clearly, explanations for the use of some forms have to be sought
after considering the options available to the author as well as the possible moti-
vations that are consciously or unconsciously involved when users of language try
to achieve certain effects through particular forms.
The past tense as a marker of non-recency
While the present tense in headlines conventionally encodes references to the

recent past, emphasizing the current relevance of the event, the past tense artic-
ulates the opposite meaning. It indicates that an event is firmly grounded in the
past. The past-time grounding of the event can mean that the event happened in
the more distant past or – more simply – that its occurrence preceded the time
span that coincides with the publication frequency of the newspaper (usually one
day), i.e. the time frame that defines events as ‘relevant’ for the current issue. The
past tense in headlines thus has a certain past-perfect feel to it in that it can indi-
cate the anteriority before another moment in the past, e.g. the previous issue of
the paper.
It is this function of the past tense to encode newsworthy but non-current
events that is pointed out to journalists when they are advised on the practice
of headline writing: “The past tense occasionally will be used to signal that the
newspaper has just learned of an event that occurred more than a day or two ago,
as in ‘Nixon taped all White House conversations’” (Rooney and Witte 2000). This
function is also confirmed by other manuals for journalist – for instance, they link
it to events that occurred “beyond the immediate past” (Russial 2004: 143).
Let us comment on several examples to illustrate the non-recency associated
with this tense form. The first example is a headline introducing a news text on
the discovery of the body of the English King Richard I, reporting on the results
of a scientific inquiry into the corpse:

Revealed: ‘Brutal’ Richard I’s heart was soaked in holy balm to ease
his passage into heaven (The Independent; 28.2.2013)
162 News and Time
The past tense of the verb (was soaked) is not explicitly anchored to a specific
point in time but the time of situation can be inferred to be the time of Richard’s
death over 800 years ago. It is noteworthy how the newspaper increases the news
value of the story – ultimately based on an event that has been non-recent for
centuries – by opening the headline with the past participle of the verb ‘reveal’.
This verb partly reports the scientific discovery (i.e. the ‘revelation’ of the finding
by scientists) and partly draws on tabloid rhetoric to make the headline maximal-
ly effective. The multifunctional past participle revealed has a dual orientation:
ideational (conveying information referentially) and interpersonal (enhancing
the news value for the readers). Without the past participle, the headline could
hardly be considered as newsworthy, since it would merely provide historical in-
formation. In a sense, the tabloid-style participle even compensates for the real
danger that the information in the rest of the headline could be found to be of
little current relevance.
When non-recency is expressed in headlines, adverbials of time can occa-
sionally be used as well in order to increase the perception of non-recency, but
only as long as that boosting function contributes to the news story. Thus, in the
following examples, non-recency associated with the past tense forms (knew and
fell away, respectively) is enhanced by the presence in the headline of an adverbial
of time that specifies the past event in a definite way (three months ago in the first
case) and as a less specific period of time (in last decade in the second).
Plebgate: Cameron knew Mitchell evidence was suspect three
months ago (The Independent; 21.12.2012)
Census: 4m UK Christians fell away in last decade
(The Times; 11.12.2012)
The time adverbials may be anchored, as in the two headlines above, to the an-
ticipated receiving time at which the news story is likely to be read. In this sense,
the anchorage of the past-time event to the readers’ deictic centre is a sign of
the indirect presence of the assumed readers in the text, since both ago and last
are relational adverbs linked to the time of utterance (projected into the reader’s
deictic centre).
The ultimate form of non-recency most likely consists of the encoding of
an event in the past perfect that indicates anteriority before another past-time
event. In the following example, the past perfect is needed in order to indicate
the non-chronological sequence of the two events (i.e. the hit-and-run incident
occurring after the mother’s murder):
Mass hit-and-run victim had lost mother to killer
(The Times; 22.10.2012)
Temporal zone PAST FUTURE
Status of temporal
non-recency recency
currency
Tense convention past tense present tense
B A
Figure 7.1 The anteriority (non-recency) meaning of the past tense in headlines
(A = receiving time RT (readers’ “now”), B = past temporal watershed, e.g. the last
publication date)
Because of the anteriority meaning of the past perfect, readers can infer what the
other past-time event is even though it is not expressed through a verbal phrase. It
appears as an attribute in the complex nominal phrase in the form of a lexicalized
attribute (mass hit-and-run victim), i.e. as a syntactically condensed predication.
The situation that the contrast between the past tense and the present tense
constructs is graphically represented in Figure 7.1. The figure outlines the con-
ceptual mapping of the past zone as being divided into two distinct temporal
subzones that differ in terms of the status of the temporal currency of the events
located within these subzones, namely recency vs. non-recency.
The figure divides the horizontal time line into several distinct sections whose
boundary points are delimited by two vertical lines. The vertical line A denotes
the receiving time RT. This is the time into which the deictic centre of the en-
coding time ET is projected; it also coincides with the publication time PT. The
anticipated receiving time provides the time with respect to which temporality
in news stories is anchored. Technically speaking, the receiving time constitutes
a point in time. In reality, however, the anticipated receiving time of print news-
papers forms a period of time since it is presumed that the individual readers of
print newspapers will be able to use the RT as an anchorage point for the duration
of the publication frequency of the paper, i.e. usually, for the length of one day.24
24. The conception of RT as a point or period reflects one’s perspective: if news is approached
with a view to its day-by-day production, then ‘one day’ constitutes such a single point equiv-
alent to ‘now’ – it is a unit of measure that is incrementally added to other units (points). On
a different level of abstraction, the unit of one day can be viewed as a period of time within
which the individual readings of the paper will occur at multiple yet different times of the day
by different readers. The lower level need not be relevant for print newspapers, but it is crucial
164 News and Time
From the point of view of the reader, all past-time events are located in the
past. However, some past-time events are represented as more current than oth-
ers. Events that are located in the pre-present zone (again typically coinciding
with the publication frequency, i.e. one day) have a high relevance and qualify
as current news. They are defined by their recency. Correspondingly, the verbs
representing such events in headlines will appear in the conventional present
tense. By contrast, events that are located beyond the time zone of recency or are
backgrounded or explicitly anchored to a definite past time, are represented as no
longer current and non-recent. Hence, they are represented in the past tense in
order to indicate their disassociation not only from the present time but also from
the zone of recency that is – despite being located in the past time – encoded in
the present tense.
The central section identified in Figure 7.1 as the zone of recency is the crucial
one for the majority of newsworthy events reported in the media. It denotes a pe-
riod that spans the time between two consecutive issues of the newspaper: the last
one and the latest one. The events that occur within this time span are, by default,
considered as being of current relevance and are conventionally encoded in head-
lines by means of the present tense, which emphasizes their current relevance at
the encoding time ET (and the receiving time RT).25
The time span between two consecutive issues of a given periodical publica-
tion defines what its conceptual category of ‘recency’ is implicitly understood to
be. With traditional print publications, the scope of the ‘recency’ category is typi-
cally one day in case of dailies (or two or more days whenever there is a gap in the
publication frequency) and one week for weeklies. Although there is some flexi-
bility in the scope in order to take into account the occasional irregularity of pub-
lication arising from national holidays, weekends, etc., the period is fairly clearly
defined. This is, no doubt, facilitated by the audience’s background knowledge of
the role of the media in bringing current news that updates the state of knowledge
in an incremental fashion. There is a shared awareness of a definite cut-off point.
The temporal situation is less clear-cut in online news media, which provide
an instant flow of news stories as the events happen and as the news becomes
known through news agencies or other news-gathering sources. The category of
‘recency’ in some online newspapers corresponds to the one-day time span of
print newspapers, particularly where the online and print versions are mutually
for online news content that is periodically updated and changed rather than released online in
a single batch.
25. This applies to the encoding of recent events through the authorial voice of the paper. As
observed earlier, accessed voices can employ the simple past even when encoding such recent
events.
dependent and the online version is filled with new content once a day. Where the
newsdesk updates the web page with current content in a story-by-story fashion,
it is the latest news stories that appear (usually) among the top stories in the rele-
vant sections. The addition of new stories then shifts the earlier stories down the
page, pushing them into the background – almost in a manner reminiscent of the
way posts are added in online chat and some other genres of computer-mediated
communication. The earlier stories may be pushed from the newspaper’s page
gradually, rather than being replaced at one go when the web page is updated with
new – and more recent – content. The crucial point here is how online news sto-
ries are relegated to the newspaper’s archives (i.e., how they become non-recent
and recede into the past): either all of them at once or one by one gradually.
Thus, while a very distinct boundary between the categories of ‘recency’
and ‘non-recency’ exists in print media, online media that update their content
throughout the day tend not to have such a clear cut-off point. Rather, they op-
erate with a temporal ‘grey zone’ which includes stories that – though no longer
the latest stories of the day – are still current, yet are gradually shifting into the
‘non-recency’ zone. Where the print media operate with the notion of ‘yesterday’
as the period of time corresponding to the recency criterion, the same time in
online media may already qualify into the non-recency zone because online me-
dia operate on a different cycle (e.g. 24 hours, reflecting their 24/7 production of
news). The publication and delivery of news in traditional print media (as well as
broadcast media) within a single batch constitutes the boundary point between
the categories of ‘recency’ and ‘non-recency’: the publication (or the airing of a
news bulletin) shifts all previously ‘recent’ events into the ‘non-recency’ zone and
opens a new period of recency for events that will occur until the publication of
the next issue of the paper.
The past tense, subordination and information flow
As we saw in the discussion of the historic present in Section 6.4 above, verbal
tense may serve the purpose of foregrounding or backgrounding certain events.
In this sense, the simple past tense functions as a grammatical device for the man-
agement of information flow within the headline. A rather specific case of the past
tense as a marker of non-recency occurs in situations when it identifies a subevent
that serves as a background to the main event. While the main event is obviously
the focus of the news story, the subevent, which receives the past tense, often de-
fines the newsworthiness of the whole story. In this way, the choice of tense assists
in indicating the hierarchical importance of the individual components of an event
for the benefit of the reader. The tense therefore serves a discourse-organizing
166 News and Time
purpose. In terms of the Hallidayan framework, the past tense assists in articulat-
ing the textual metafunction.
The identification of this use of the past tense as a special category is based
on the common occurrence of a specific structural pattern identified in the data.
In these instances, the past tense occurs in subordinate clauses specifying some
other elements, most commonly simple nominal phrases. It is not unusual for the
nominal phrase to be realized through a single word. Semantically, the headline
then encodes a state that obtains in connection with, or as a result of, the situation
described in the clausal component. Syntactically, the subordination of the clausal
element is a grammatical expression of the subordinate function of the proposi-
tion formulated in the clause that becomes mirrored through the use of the past
tense. In many examples, it is connected to the expression of anteriority – the past
tense expresses not only non-recency but also the precedence of the proposition
coded through the simple past tense before the event or state expressed in the
nominal phrase.
Let us illustrate how the past tense can be related to the hierarchical organ-
ization of information within the headline. In the following headline, the most
important (newsworthy) element is encoded through the main clause. The subor-
dinate clause with the verb in the past tense merely provides the necessary back-
ground by specifying the identity of the subject:
Police who dragged man behind van are suspended
(The Times; 1.3.2013)
The simple past tense in the subordinate clause downgrades the importance of the
specifying information, allowing the other section of the headline to be in focus.
At the same time, the past tense encodes the ‘non-recency’ of the backgrounded
proposition. It is worth noting that the past tense in the relative clause cannot be
readily reformulated into the present tense. The present tense might appear, how-
ever, if the subordinate clause was re-expressed as the main clause, whereby the
status of the proposition would be raised from mere background to a more news-
worthy element (cf. *‘Police drag man behind van and are suspended’), though the
presence in the headline of two such equally important propositions would be less
likely. A more likely reformulation in the present tense, however, concerns the
temporal-causative subordinator ‘after’ (*‘Police [are] suspended after they drag
man behind van’; though, in this particular case, it would be more natural to use
the -ing form, namely *‘after dragging man’).
The basic grammatical meaning of relative clauses is, of course, to provide
additional background information. However, the situation with other subordi-
nators is less clear, even though they evidently organize the whole sentence into
hierarchically arranged units. Let us briefly contrast the predominantly semantic

use of the present tense in relative clauses with two examples using the subordina-
tor ‘after’ (both examples are described in Section 8.4 in a different context). These
differ from the possible reformulation above in that there the subjects of the two
clauses are different, and no syntactic condensation by means of the -ing form is
possible. As a result, the subordinate clause receives a finite (tensed) verb form:
Nine-year-old girl in Halloween costume shot after relative mis-
takes her for a skunk (The Independent; 22.10.2012)
Government to probe Comet collapse after backers took £13m in
fees (The Telegraph; 18.12.2012)
When the two examples, structurally almost identical, are juxtaposed, the former
appears to assign a greater degree of newsworthiness to the reason specified in
the subordinate clause on account of the present tense (‘after relative mistakes her
for a skunk’). In the latter example, by contrast, the past tense places the subevent
into the zone of ‘non-recency’, thus defocusing from it to the benefit of the main
event (‘after backers took £13m in fees’). Should the past tense in the latter example
be substituted with the present tense, this would result in a headline that is per-
fectly correct and natural (*‘after backers take £13m in fees’). The only difference
between the two forms is that the present tense assigns more importance to the
subevent. Thus, at least in some types of subordinate clauses, the contrast between
the present tense and the past tense can be exploited as a strategy of foreground-
ing and backgrounding certain propositions and can be used pragmatically.
Let us discuss several more examples of relative clauses where the past tense
has a semantic motivation, providing information that, despite its newsworthi-
ness, remains in the background to some other proposition. The next example
combines this function with heteroglossia in the main clause. Here, the headline
accesses an external voice, indicated through the use of quotation marks, and
does so in the part that contains the verb phrase in the main clause, i.e. where the
most important event is to be sought:
Fire that killed six ‘was rehearsed’ (The Guardian; 28.2.2013)
Although there are two occurrences of the simple past tense in the headline, each
of them has a different function. The former (killed) occurs in the syntactically
subordinate relative clause and relegates the information provided by it into the
position of mere background. By contrast, the latter (‘was rehearsed’) appears in
the main clause and is part of the headline rhetoric that enriches news texts by
incorporating external accessed voices. Consequently, the accessed voice enjoys
the benefit of not having to follow headline conventions. With the headline in
168 News and Time
question, moreover, the use of the present tense would not be entirely in place
because the proposition refers to an event that happened in the non-recent past,
i.e. before the temporal watershed moment that consisted of a previous report
revealing the fact that ‘six people were killed in a fire’. Here, the past tense in the
accessed voice has an additional function: it also expresses the non-recency of the
event, but in this particular case the newspaper editor has decided to encode the
non-recency through the quasi-direct speech of an unacknowledged social actor
rather than in the paper’s own voice.
What we then see in many of these headlines is that they are composed of
a combination of two conceptual categories: ‘recency’ and ‘non-recency’. While
the focus is certainly on the aspect of the story that qualifies its inclusion under
‘recency’, it is the element of the story that is non-recent that adds the crucial
newsworthiness to the more recent phenomenon. In terms of cognitive temporal
structure, such headlines are thus marked by significant complexity.
The association of the past tense with a subevent in the background is not
affected by the non-finiteness of the verb in the main clause. Thus, despite being
the only finite verb form in the headline, the simple past tense does not express
the most newsworthy aspect of the story, i.e. it is not a key to ‘what the news item
is about’. Cf. the following example:
Hollywood hacker who leaked nude Scarlett Johansson photos
jailed for ten years (The Independent; 18.12.2012)
The headline consists of a complex sentence, with a subordinate relative clause

formulated with a finite verb in the past tense. The clause specifies an introduc-
tory nominal phrase (Hollywood hacker) by establishing that social actor’s claim
to newsworthiness (who leaked nude Scarlett Johansson photos) – in this way, it
serves to justify the media’s attention to the whole story. The rest of the main
clause is a non-finite clause (jailed for ten years) that uses the conventional ellipsis
of the operator in a passive construction. The entire headline is cast in the paper’s
authorial voice, so there is no motivation to use any other verb forms than the
conventionally ellipted present tense. Thus, although the main clause retains its
non-finite structure, it describes the main event of the story, while the relative
clause with the finite verb form in the past tense merely provides past background
and, potentially, links the story to readers’ previous knowledge of the case as well
as the paper’s previous coverage.
The defocalising effect of the past tense in such relative clauses needs to be
considered in close connection with other forms of managing the flow of infor-
mation in headlines, such as the practice of headline writers to condense informa-
tion into complex nominal phrases through premodification and compounding,
thereby significantly increasing the lexical density of headlines. The relative clause
is an alternative to such complex nominal groups: it is cognitively less demanding
because the underlying proposition is expressed in full and does not need to be
unravelled by the readers (unlike, for instance, the nominal phrase in the headline
Poppy-burning teenager meets forces; The Guardian; 20.12.2012). Thus, in
contrast to such complex nominal phrases, the past tense in relative clauses has the
advantage of clearly indicating the anteriority or non-recency of the proposition.
The following example shows a somewhat different situation in which the
past tense is used in a relative clause even though the proposition in the clause
does not express the background to any other, more newsworthy or newer, aspect
of the story:
The thieves who stole Christmas at Great Ormond St: Presents
destined for sick children stolen (The Independent; 19.12.2012)
The first part of the headline, which contains the past tense in the relative clause,
is, in fact, a nominal fragment: it merely presents the existence of ‘thieves’ and
identifies them with the relative clause. However, it is followed by a colon struc-
ture that introduces another non-finite clausal structure that mediates the most
newsworthy aspect of the story – the theft of gifts for children.
The colon has several functions in headlines, one of which is the manage-
ment of information flow. Colon structures divide headlines into two parts, with
the first segment expressing a general topic and the second describing an update
on the story. This division essentially highlights the dualistic distinction between
background information and new information, creating two distinct zones that
could be labelled, for instance, as thematic and rhematic (Firbas 1992). This pat-
tern is characteristic of stories that are covered repeatedly on several subsequent
days and thus become established as topics in the public consciousness. The first
part then serves as a sort of ‘mental short cut’ that triggers the relevant frames and
activates the readers’ background knowledge. The second part then presents the
essence of the news item – the latest development in the story.
The above headline bears a surface resemblance to the headline colon pattern
but is somewhat different in that both parts are semantically almost identical. It
is within the first part, whose topic-establishing function is rather weak in this
particular instance, that the past tense is used in order to categorize the relevant
social actors. Thus, not only does the past tense background the specification pro-
vided by means of the relative clause, but the past tense phrase stole Christmas also
functions as an allusion, potentially resonating with cultural meanings referring
to the well-known children’s book How the Grinch Stole Christmas by Dr. Seuss.
170 News and Time
A similar dual structure in a headline is also present in the next example. The
headline uses the past tense in a subordinate temporal clause introduced with the
conjunction as, which normally tends to be accompanied with the present tense:
Killed as they tried to save lives: Pakistan polio workers targeted
Here, however, the past tense is used as a result of a combination of two of its
functions: (1) the temporal clause in the first part of the headline provides the
background; and (2) the proposition it expresses (‘they tried to save lives’) is pre-
sented as anterior to the act of ‘killing’. This usage then combines the information
flow management and semantic functions of the tense.
Once again, the structure of the headline is essentially repetitive since it com-
municates very similar information twice. Both parts refer to the same news ac-
tors and both specify the nature of the negative action which involves the social
actors in the semantic role of patients, although they do so with a varying degree
of specificity. In the first part of the headline, the medical workers are represented
as victims (through the implication of the past perfect form killed), while their
profession has to be inferred from the subordinate clause (as they tried to save
lives). The second part is co-referential and refers to the victims by means of a
category label (polio workers). Although the main process (‘the killing’) is referred
to twice as well, the transition from the more specific killed in the first part to
the more general targeted in the second part makes it possible to interpret the
headline as developing the flow of communication rather than being tautological.
In this reading, the particular killing constitutes the background topic (themat-
ic element), which extrapolates into the more general issue that deadly violence
is starting to be used against medical workers – an issue that would constitute
the focus of the report as such. Indeed, the story continues with a report on the
suspension of the vaccination programme and the condemnation of the assault
rather than a description of the killing itself. Although the two parts of the head-
line are related to each other as the linear sequence of ‘specific → general’, this
rhetorical pattern is mapped onto the ‘thematic/known → rhematic/new’ format
of information flow management stemming from the use of the colon structure.
Thus, it appears that some subordinate clauses tend to encode past-time
events in the past tense on account of semantic reasons. Relative clauses, for in-
stance, seem to block the possibility of using the present tense because they are
strongly associated with providing additional information that serves as some
prior background to a more newsworthy event. As a result, the proposition can-
not be pragmatically reevaluated and enhanced through the present tense – the
deictic past tense is then the only actual option. In some other types of subor-
dinate clauses, such as those introduced with ‘after’, both tenses are sometimes
equally possible. This indicates that the editor’s final wording is, in some cases,
motivated not only by semantic but also by pragmatic reasons, since the choice
of the present tense allows for the subjective enhancement of a particular aspect
(subevent) of the main event. All this points to the importance of a multi-dimen-
tional analysis of news headlines that takes into account the interplay between
tense choice, syntactic structure, semantic/pragmatic motivations, and the man-
agement of information flow.
The past tense as a marker of accessed voice
As mentioned above, one of the most significant uses of the past tense in head-
lines is in quotations. Headlines frequently contain segments of reported speech,
whereby an additional voice enters the headline. As a result, many headlines be-
come heteroglossic because they place, side by side, the anonymous, institutional
voice of the paper and the external accessed voice of some news actor. A formal
indicator of the accessed voice is constituted by the presence of inverted commas
(quotation marks), though they may be absent in heteroglossic headlines formu-
lated as indirect and free direct speech. In this case, some other formal means,
such as the use of the past tense, may serve as indicators of the fact that the rele-
vant textual section is, in fact, an external voice different from the authorial voice
of the paper.
The external voice can be accessed in headlines in several ways that differ in
their degree of explicitness and acknowledgement of the intertextual origin of the
quote. Sometimes, the whole headline is made up of a direct speech quote that is
enclosed in quotation marks, cf.:
‘I hit girlfriend to save her from self-harming’ (The Times; 3.10.2012)
In cases like these, any textual attribution of the direct speech segment to a spe-
cific external source is typically missing, though the identity of the speaker may
be revealed through visual means and is almost invariably present in the lead/
subhead. The headline is, thus, fully monoglossic. The example above refers to a
statement made by the English comedian Justin Lee Collins during a trial follow-
ing accusations of harassment from his former girlfriend. However, direct quotes
in headlines are to be treated with a degree of licence since they arise out of the
need to compress often complex utterances into brief textual snippets that are
both informative and newsworthy. The ensuing verbal context in the body copy
usually reveals that such headlines are often pseudo-quotes rather than verbatim
citations. This is also the case with this particular headline since the original ut-
terance admitting responsibility was formulated somewhat differently.
172 News and Time
What is crucial about such quotes and quasi-quotes is that they recreate the
impression of direct speech, the representation of which is generally understood
to be fictional rather than authentic in headlines. While the pseudo-quotes add
a sense of immediacy – as if providing the reader with an in medias res glimpse
at the event reported – the past tense provides a further authentication of what
appear to be direct speech utterances.
Another type of headline containing accessed voices contains a mixture of the
institutional voice and the accessed voice. Unlike the previous example, which is
formally monoglossic, this headline type contains two distinct parts that are clas-
sifiable into the two different voices (see examples below). The utterances quoted
range from brief textual segments consisting of a single word or phrase to com-
plete clauses.
(a) Isolated word citations

Let us first discuss the issue of the external voice accessed in a minimal way, con-
sisting of a direct speech quote of one or more isolated words rather than com-
plete utterances.
Life saving surgery on brain tumour boy ‘went well’
In the example, the inverted commas indicate that the relevant words are ascribed
to an external source – they are not encoded through the voice of the paper. The
inverted commas also function as boundary markers not only separating the two
voices that are present in the headline but also realizing the switch from the con-
ventions of headlinese to a more conversational style of language.
The practice of using inverted commas to separate off isolated words and
phrases is very common in news discourse. It is used as a strategy for indicating
the paper’s non-compliance with a given representation, which is usually ascrib-
able to some external source (Cotter 2010: 174). Since they mark critical detach-
ment and evaluation, they have been called ‘scare quotes’. By putting expressions
into scare quotes, authors “warn the reader that these expressions are problematic
in some way. [This strategy] dissociates the writer from these expressions and
makes it clear they belong to someone else” (Fairclough 2001: 74). While this us-
age is commonplace, the example above, with the quotation marks enclosing the
phrase ‘went well’, does not represent the paper’s critical stance. Here, the quota-
tion marks are used merely to signal an accessed voice.
(b) Full sentence citations

The second type of heteroglossia in headlines includes the citation of a complete
utterance worded through an external accessed voice – it is a form of speech rep-
resentation. The utterance mediates a seemingly direct, eye-witness account of
an event that is, once again, often formulated in the past tense. Cf. the following
example:
‘I thought I was going to die’: Blind man shot with 50,000-volt Taser
sues police (The Independent; 22.10.2012)
Similar to the pattern identified earlier, the headline consists of two parts that
refer to the same event: while not exactly tautological, the two segments repre-
sent the event from two complementary points of view: (1) the direct participant
(news actor) cited in direct speech, and (2) the newspaper providing a factual
account of the incident. The colon structure is used here for the juxtaposition of
the external accessed voice and the internal institutional voice of the paper. In
terms of style, the former textual segment is presented in personal language (cf.
the personal pronouns and the past tense) while the latter follows the conventions
of headlinese (ellipted operator, absence of determiners, present tense).
Through heteroglossia, the convention-tied headlines can be freed to em-
brace a broader range of sentence types, tenses, and evaluative lexis, etc. than
would typically be present in monoglossic headlines. The accessed voice, thus,
has far greater impact on the form of news headlines than merely enabling the
occurrence of the simple past tense.
All in all, the practice of using the past tense in parts of headlines formulated
in an accessed voice constitutes an important strategy whereby the editor weakens
the dominant position of the present tense in headlines. Typically, headlines with
the past tense in the accessed voice can be reformulated with the verb in the pres-
ent tense, once they are recast in the authorial voice of the paper. It, thus, appears
that there is a certain complementarity, on the one hand, between the attraction of
the authorial voice and the present tense, and, on the other, the accessed external
voice and the past tense.
Other uses of the simple past tense – the non-factive presupposition
Before concluding this section on the past tense in headlines, let us make some
final observations related to the fact that the past tense appears in several other
types of headlines as well. Sometimes, the tense is required in order to represent a
state of affairs that did not materialize, as in the following headline which encodes
174 News and Time
a non-factive presupposition through the verb ‘plan’. Here, the past tense is need-
ed since it renders the plan inoperative:
Trio planned ‘to kill more than 7/7’ (The Guardian; 22.10.2012)
The headline is formulated so as to highlight the pastness of the plan. The past
tense emphasizes that the plan was thwarted. What makes the story newsworthy
is then not only the significance of the planned event (i.e., its comparison to the
7/7 attacks in London) but also the fact that it is no longer applicable: the seman-
tic status of the event is that of non-validity (Kress and Hodge 1979: 128).
The past tense is in place because of the semantic content of the verb ‘plan’.
The headline could certainly be reformulated in the present tense but a different
verb would have to be used to mark a shift in focus (e.g. *Trio tells of plan …).
However, should the process verb ‘plan’ be realized in a nominalized form, the
transformation would remove the temporal and modal marking that is an oblig-
atory component of the finite verb phrase. If nominalized, the original process
(‘planning’) would become atemporal. Paradoxically, that would also remove the
non-validity of the original process because the nominal form would not make
that particular aspect of meaning immediately obvious. It would shift the status
of the proposition into the sphere of validity (future possibility), i.e., the default
reading of the noun ‘plan’.
A similar situation would obtain if the verb ‘plan’ was, in harmony with head-
line conventions, shifted into the present tense (*Trio plans…). The present tense
would, once again, cancel the past possibility meaning of the semantic process
and reinstitute its reading as an event with a possible current applicability. Thus,
the utterance requires the past tense because the conventional present tense or a
nominalization of the process would actually communicate an entirely different
(and false) message.
Other uses of the simple past tense – satellite articles
Apart from the non-factive presupposition of the past tense of certain verbs, there
is another systematic situation when the past tense appears in headlines. In this
case, however, headlines must not be assessed on an individual basis. They need
to be approached intertextually as mutually related structures, some of which are
hierarchically organized.
This situation arises in the media coverage of stories that are reported in ar-
ticle clusters. The main article, breaking the story and providing the main infor-
mation, is accompanied by several supplementary articles that provide additional
or background information. As a result, the headlines of accompanying stories
cannot be treated as entirely independent, stand-alone units because they operate

within the context established by the central article on a given issue. This means
that there is a hierarchy of importance between news headlines, with some being
central and others more peripheral, i.e. subordinate. Any study of news discourse
should then take into account the fact that the hierarchical arrangement of stories
within article clusters (or ‘packages’) affects the way the articles and their head-
lines are formulated. The past tense, for instance, is one of such textual effects
observable in satellite articles providing background information.
This ‘package approach’ to the presentation of news content has been noted as
one of the characteristic trends observable in the modern media (Ungerer 2000).
It means that newspapers increasingly do not provide coverage of a given topic
in a single article only. Instead, they present the news content in several related
articles. The story is thus broken up into various aspects that are presented in a
piece-meal fashion.
The package approach to the construction of news stories and the presenta-
tion of news content in clusters of related articles significantly affect the nature
of textuality in newspapers. Traditionally, newspapers used to print articles that
were – despite being grouped into relevant sections – relatively independent of
each other. The genre of the newspaper functioned as a typical text or discourse
colony (Hoey 2001) since the individual and independent articles could be re-ar-
ranged without much regard to any logical connections between them: they were
independent texts rather than component parts of a higher textual unit.
The package approach has changed this relative independence of articles. The
components of article clusters derive a part of their meaning from their place-
ment alongside other articles that cover different aspects of the same story. As
regards headlines within an article cluster, this means that they are no longer just
stand-alone units placed at the beginnings of their respective news texts. Rather,
such headlines provide linkage between the central article on a given topic and
the content within the accompanying article. The headlines retain the opening
function since they provide the point of entry to the relevant articles but, at the
same time, they function as intertextual mediators linking the articles they open
with the central article and the other accompanying articles within an article clus-
ter. It is therefore important to determine whether a given headline has the cen-
tral position of opening the main story – and thus the whole article cluster – or
whether it is hierarchically subservient to some higher-level textual unit (another
headline, an article preview, or the whole text of the news article enjoying the
central position within the cluster).
Since the headlines analysed for the purpose of this study were intentional-
ly taken only from central articles and the accompanying articles were exclud-
ed from data compilation in order to ensure a comparable status for all texts, a
176 News and Time
quantitative description of this phenomenon is not applicable with the data at

hand. However, in order to illustrate the impact which the package approach has
on the use of the past tense, let us consider the following cluster of articles on the
same topic, originally mentioned in Chovanec (2005b: 79–80):
Soldiers gunned down after fury over searches
We won’t send more troops, Blair tells Commons
‘We would not take dogs into Iraqi homes’
Sergeant refused to abandon his trapped comrades
Resistance may grow to avenge wounded pride
The articles appeared in the online version of The Times in a package: they form
an article cluster and describe various aspects of a single major news story of
the day. In this way, the event is retold several times in a model of news report-
ing that has also been referred to as the nucleus-satellite structure (Iedema, Feez
and White 1994: 104; van Leeuwen 2009: 354). The central article (nucleus) in the
cluster provides information about the most crucial event – in this case the killing
of British soldiers (Soldiers gunned down) – and its cause (after fury over
searches). What follows is a group of four accompanying articles (satellites, also
known as ‘sidebars’) that focus on specific aspects of the event and its aftermath.
The individual headlines include:
– a free direct speech comment from the British prime minister quoted in a het-
eroglossic headline with an acknowledged and attributed source (We won’t
send more troops, Blair tells Commons);
– a direct speech comment citing an unnamed British soldier in what is a for-
mally monoglossic headline that does not attribute the external words to a
specific accessed voice (‘We would not take dogs into Iraqi homes’);
– a monoglossic description, in the authorial voice of the paper, of the poten-
tial heroism of one of the soldiers (Sergeant refused to abandon his
trapped comrades); and
– the possible future implications of the act, formulated in a monoglossic head-
line in the authorial voice of the paper (Resistance may grow to avenge
wounded pride)
We can see that what is reported in the individual articles is: (1) the main event
itself, (2) the verbal comments of elite news actors (politicians), (3) the verbal
comments of direct participants (soldiers) concerning the possible cause, (4) a
background event that becomes newsworthy in connection with the major event,
and (5) speculation on the future development of the case. The headlines, thus,
directly encode some of the components of the conceptual structure of the news
story – main event, verbal reaction, cause, background, consequences
(cf. Ungerer 2000; van Dijk 1988) – focusing on each of them in a separate news
text. Iedema, Feez and White make a similar observation by stating that “each
satellite provides a new set of details about the … event, including details about
the past or details about the consequences” (1994: 168).
The temporal deixis within the satellite headlines includes reference to future
events (We won’t send…, Resistance may grow…) as well as past events (‘We
would not take dogs…’, Sergeant refused to abandon). The first of the two
latter references is in the past tense because it is formulated as a direct speech in
the accessed voice of a direct participant in the event. However, the second ref-
erence (Sergeant refused to abandon his trapped comrades) is the usage
that interests us most here. Unlike the previous monoglossic and heteroglossic
examples, provided earlier in this chapter, of accessed voices containing utteranc-
es with the past tense, the whole headline is evidently encoded through the voice
of the paper. Since it is plausible to imagine the same headline in the present tense
referring to a one-off past-time event (*Sergeant refuses to abandon…),
the past tense is obviously not required for semantic reasons, e.g. to express the
non-validity of the event. Its function is likewise not to provide an unequivocal
anchorage of the event in the past time in order to indicate the non-recency of the
event and, thus, avoid a potential conflicting interpretation associated with the
use of the present tense. The present tense would indeed have provided an ade-
quate representation of the event in harmony with the conventions of headlinese.
The selection of the past tense is thus a matter of the editor’s choice rather than a
semantic necessity arising in that particular local context.
Although the past tense does encode non-recency in the headline, the expla-
nation for its preference over the present tense is related to the backgrounding
function of the past tense identified earlier. However, the backgrounding does not
operate within the headline itself (i.e. surrendering the focal place in the headline
to some other element) but with respect to the headline of the central article with-
in the article cluster. Arguably, the past tense indicates the relative non-promi-
nence of the event and allows the central article to keep its centrality. The present
tense is semantically possible but it may be dispreferred since it could deflect the
readers’ attention from the central event. This explanation does not, of course,
aim to deny that the present tense is, in actual fact, also commonly found in such
satellite articles, but it serves to underline the fact that editors have a relatively
simple choice between grammatical structures that can, most likely intuitively,
be used for the effective management of information flow. The internal relations
of hierarchy within a cluster of related articles thus evidently affect the choice of
tense. Although the past tense in the accompanying headlines is deictic, it is also
associated with the expression of the subordinate status of the individual articles
with respect to the main story encoded in the main headline of an article cluster.
178 News and Time
This chapter has shed further light on the dominant position of the present
tense in headlines by considering the contexts in which other tenses find their
presence in headlines. It appears that there are both semantic and pragmatic rea-
sons explaining their occurrence. Thus, the use of the will-future and the past
tense is often motivated by the need to provide clear temporal anchorage for a
reported event. The specific complementary relationship of the past tense with
respect to the present tense in news headlines has led me to postulate the di-
vision of the ‘past’ temporal zone into two segments, the zones of recency and
non-recency. While the former is expressed by the conventional present tense, the
latter is associated with the past tense, which expresses the anteriority of one past-
time event before some other past-time event that is already encoded through the
present tense and located in the zone of recency. At the same time, there are also
pragmatic factors for the use of tenses other than the present tense, most notably
heteroglossia (i.e. the author’s conscious importation into the text of an exter-
nal voice and an external perspective), which renders the application of headline
conventions as invalid. As a result, headline language may come to approximate
ordinary language, being liberated from the necessity to use the conventional
headline present tense. Heteroglossia offers the possibility for an alternative for-
mulation of the same content and using a greater range of tenses than the domi-
nant present tense.
Chapter 8
Auxiliaries in headlines
Ellipsis and (non)-finiteness
The discussion of temporal relations in headlines would not be complete without

considering the issue of clausal finiteness. The contrast between nominal and ver-
bal (clausal) headline types is complicated by the fact that many clausal headlines
do not contain a complete verb phrase since the auxiliary is typically ellipted, as is
customary in headlinese and other forms of block language. This leads to the need
to distinguish between finite and non-finite verbal headlines. In the case of the
latter, the headline has an incomplete clausal structure and the ellipted auxiliary
can usually be easily inferred and reinstituted by the reader. Despite the omission
of various grammatical elements, it is thus possible to reconstruct the pre-ellipted
(full) version of the relevant headlines. That property distinguishes verbal (claus-
al) headlines realized in a non-finite manner from fully nominal headlines that do
not contain any syntactic trace of a possible finite verb form.
In the present section, the focus is on the linguistic expression of finiteness
within the verbal phrase in headlines. The discussion revolves around two main
issues: the potential ambiguity of the non-finite phrase resulting from the ellip-
sis of auxiliaries, and the non-ellipted presence of auxiliaries that could easily
be ellipted. This is supplemented with a discussion of modal auxiliaries that are
typically non-elliptible. The data indicate that non-ellipted auxiliaries and modals
tend to appear – with some exceptions – in textual segments classifiable as stretch-
es of accessed voice. In that sense, the grammatical category of finiteness of the
verb phrase appears to be treated, in the case of elliptible auxiliaries, in a manner
similar to the way the simple past tense is treated in news headlines, namely as
a marker of accessed voice and a means of discontinuing headline conventions.
8.1 Ellipsis of auxiliaries
One of the requirements of block language is to omit sentence elements with low
information value (Halliday 1994: 392; Quirk et al. 1985; Dor 2003), typically
those that convey grammatical rather than lexical information. This results in the
omission of such elements as articles and auxiliaries, making both the nominal
180 News and Time
and the verbal groups incomplete. However, while the ellipsis of articles merely
affects the explicit coding of the grammatical category of definiteness, the ellipsis
of auxiliaries turns the verb phrase into a non-finite structure that dispenses with
the explicit specification of temporality and modality. As a result, headlines gain
compactness but, at the same time, relinquish the explicit expression of temporal
and modal exponents of verbs.
The non-finite nature of the resulting headline contributes to the telegram-
matic style of the genre (Crystal 1988: 388), shared with other forms of ‘block
language’ such as labels, titles, headings, notices and advertisements. As Quirk
et al. (1985: 845) specify, “some forms of block language have recognizable clause
structures. Those forms deviate from regular clause structures in omitting closed-
class items of low information value, such as the finite forms of the verb BE and
the articles, and other words that may be understood from the context”.
The following headlines exemplify the non-finiteness of the verbal phrase
with conventionally ellipted auxiliaries, arising from the ellipsis of the appropriate
finite verb forms of the verb ‘be’:
– Constructions in the passive voice:

Passengers hurt as coach overturns (The Independent; 7.1.2013)
Investors spooked by UK Energy Bill (The Telegraph; 7.1.2013)
– Constructions in the active voice and the progressive aspect:

NHS using dementia beds for norovirus patients (The Times; 7.1.2013)
– Constructions with the semi-modals ‘be to’ and ‘be set to’ expressing future:
Canadian PM to meet First Nations (The Guardian; 7.1.2013)
White House set to fight with NRA (The Guardian; 7.1.2013)
NHS inquiry to demand tighter control of managers
(The Times; 7.1.2013)
– Constructions in which the ellipted verb ‘be’ is a full verb (copula) rather than
an auxiliary:
Smith: Miliband ‘wrong on benefit cap’ (The Telegraph; 7.1.2013)
Astronomers in hunt for alien life (The Guardian; 7.1.2013)
As the examples with the ellipsis of the auxiliary ‘be’ show, the lexical verb is
retained in its non-finite form (present/past participle, infinitive). One of the rea-
sons for the omission of auxiliaries and the retention of the lexical verb can be
found in the dual nature of the verb phrase. Every verb phrase – even when made
up of a single verb – consists of two components: the lexical component and the
Chapter 8. Auxiliaries in headlines 181
temporal and modal exponents (TMEs). While the former convey semantic (lexi-
cal) information, the latter encode grammatical and attitudinal information.
The temporal and modal exponents are usually relatively less important for
the development of the message (cf. Firbas 1992: 71) than the lexical component.
There are several reasons for this. First, the temporal anchorage is often stable
over several utterances and verbal tense does not change – it ensures cohesion of
the text and does not contribute much new information. The grammatical com-
ponent of the verb becomes important when, for instance, a tense switch occurs,
indicating a rearrangement of the temporal situation. The relative importance
of the temporal coding also increases in echoic utterances, repetitions and brief
answers, where the lexical component may be repeated or re-expressed and the
local focus of the communication is actually on the negotiation of the temporal
situation. In that case, special focus may be placed on TMEs, e.g. by intonation,
typographical means, juxtaposition of forms, etc. The second reason for the di-
minished importance of the temporal and modal exponents vis-à-vis the lexical
component in verbs is the fact that the former can be re-expressed and specified,
if need be, through other parts of speech and sentence constituents that indi-
cate temporal and modal information as well (cf. yesterday; this year; perhaps;
likely; etc.).
Another argument for the diminished relative importance of TMEs is the
very existence of headline conventions, particularly the shift of tenses resulting in
the increased use of the non-deictic present tense. It should be noted that the shift
of tenses in headlines affects only the temporal and modal exponents of the verb
and the verb phrase, while the lexical component remains intact (except for cases
in which the specific headline rhetoric of a particular newspaper prefers expres-
sive and monosyllabic lexis that are subsequently re-expressed in the lead and the
body by means of their stylistically neutral counterparts).
An additional factor that contributes to the lower comparative significance of
the temporal and modal exponents is the overall context of situation. The genre of
news reporting is based on the expectation of recency of the news items report-
ed. The temporal context need not be explicated since it is presupposed, which
favours the non-expression of the category of time in headlines. No such presup-
position can exist on the part of the lexical meaning.26
Last but not least, headlines are not stand-alone units (see Chapters 9 and 10).
While they can have numerous functions, headlines provide a point of entry into
26. Various elements in the headline can, however, be linked to the visual elements accompa-
nying the news text. The visual elements may then provide the point of entry into the news text.
The headline may be formulated with the assumption of the readers’ prior familiarity with some
news content that is presented visually (cf. Bednarek and Caple 2012).
182 News and Time
the news story, often acting as an initial summary. It is a general editorial prac-
tice that, for the sake of newsworthiness, epistemic modality expressing various
degrees of possibility, speculation and non-commitment to the facticity of the
proposition is conventionally rendered in headlines through the modal certainty
of the present tense. It follows that under such conditions, the modal value of the
verb phrase in the headline is not necessarily an entirely true and objective guide
to the state of affairs reported in the news story.
Thus, it is perhaps little surprising that auxiliaries, which convey grammatical
information through the encoding of temporal and modal aspects, may be sacri-
ficed in headlines without any substantial loss of information. At the same time,
their absence may help to highlight the prominence of other elements crucial to
the story and forming the core of the headline summary.
8.2 Potential ambiguities
Readers can reconstruct most full forms of headlines with ellipted elements on
the basis of their knowledge of the grammatical rules of the English language and
the conventions of the genre. There are headlines, however, where the ellipted
words are not always “precisely recoverable” (Quirk et al. 1985: 884). As a result,
ambiguities may arise. Regardless of whether the ambiguities are produced inten-
tionally or not, they reinforce the inherent interactiveness of the headline because
the recipients face the necessity of disambiguating them. In this connection, van
Dijk observes that
articles and verb auxiliaries are often lacking, so that headline reading is impaired
on a number of points. Compared to full sentences they may be more ambigu-
ous, vague or syntactically complex. This means that much processing is left to
semantic interpretation. (van Dijk 1988: 144)
Headlines may also contain semantically underdetermined information that can

suggest several plausible interpretations. As argued by Ifantidou (2009: 701), such
underspecification of the contextual information can be viewed as a reader-ori-
ented text-production strategy that is particularly effective in cueing the readers’
interests. The respondents in Ifantidou’s study into the processing of headlines
appeared to express “a clear preference for headlines which are open to more than
one interpretation and seemed to reject the explicitly informative type” (704).
The ambiguities may be resolved visually (e.g. through the multimodal rela-
tionship between the headline and an accompanying photograph) or textually,
with readers retrieving the relevant meanings in the ensuing verbal context (i.e. by
continuing to read the lead and/or the body copy of the article). Some ambiguities
may not be immediately apparent and may lead to garden-path effects. Readers
may be forced to re-interpret their initial understanding of the content communi-
cated in the headline once the ensuing co-text forces them to abandon the initial
interpretation in favour of another. Not surprisingly, such headlines – authentic
as well as fictional – are widely circulated and frequently appear in collections of
language humour (cf. Bucaria 2004; Dynel 2009: 99).
Such re-interpretation of an initially false assumption about the intended
meaning may occur, for instance, in the case of lexical ambiguity residing in com-
plex nominal phrases. The property of nominal phrases to mask the syntactic
argument structure of the underlying predication has been noted by various crit-
ical linguists (cf. Fowler 1991, Richardson 2007). To illustrate this point, let us
focus on the nominal phrase nanny murder case in the sample headline Nanny
murder case feeds parents’ paranoia (The Telegraph; 16.2.1997). On the first,
entirely decontextualized, reading of the headline, the phrase nanny murder case
is ambiguous because it disguises – through the complex condensation of infor-
mation achieved by nominalization – the semantic roles of the relevant social
actors in the underlying predication (‘murder’). It is only the subsequent verbal
context (or, alternately, the readers’ prior knowledge of the case and hence their
background knowledge) that helps to resolve the ambiguity and assign proper
semantic roles that the nominalization of the relevant process (‘murder’) hides.
Thus, readers can establish that the category label nanny is the semantic agent
rather than the semantic patient of the proposition centring on the corresponding
actional predication condensed into the noun murder. In other words, the phrase
describes a case of ‘a murder committed by a nanny’ rather than ‘the murder of
a nanny’.
As regards the ellipsis of auxiliaries, this phenomenon may lead to ambigu-
ity concerning the syntactic structure of the headline, particularly the potential
confusion between sentence structures in the active voice and the passive voice.
However, the possibility is mostly theoretical. In actual practice such seeming
ambiguity is rare because the syntactic structure becomes disambiguated either
through the readers’ knowledge of headline conventions or their awareness of the
heteroglossic nature of a given headline. Quirk et al. comment on the role of the
context in the resolution of ambiguity in headlines (1985: 845) by observing that
headlines “can generally be analysed in terms of clause structure, though frequent-
ly scrutiny of the text below a headline obliges us to reinterpret the structure”.
The core of the potential ambiguity consists in the interpretation of the verb
phrase as either an ellipted passive voice structure or a complete clausal structure
in the active voice and with the verb in the past tense. That situation is possible
when the verb phrase consists of a non-finite past participle that is formally iden-
tical with the past tense form of the same lexical verb and the other constituents
184 News and Time
in the sentence structure admit both interpretations. The ambiguity arises in the
following example as a result of two possible readings of the verb ‘left’:
Second au pair left with three children (The Telegraph; 12.11.97)
The formal identity of the grammatical forms of the past tense and the past parti-
ciple means that the headline has two possible readings. They differ in their tran-
sitivity patterns and assign different semantic roles to the constituents depending
on whether the verbal form left is analyzed as (a) a complete verb phrase with a
finite verb in the simple past tense and the active voice, or (b) an incomplete verb
phrase with the non-finite past participle serving as a residue after the ellipsis of
the relevant auxiliary from a passive voice construction, cf.:
(a) X left … [semantic agent + simple past tense of intransitive ‘leave’]
(b) X is/was left … [semantic patient + past participle of transitive ‘leave’]
Thus, the subject performs the role of the semantic agent of the dynamic intransi-
tive verb in the active voice reading in (a) and the role of the semantic patient of
the stative transitive verb in the ellipted passive voice reading in (b). As suggested
above, the correct interpretation depends on several context-based considerations:
1. The headline is processed with an awareness of the conventions of the genre;

namely on the basis of the presumption that auxiliaries and other grammati-
cal words are omitted. As a result, the headline will be interpreted as a non-fi-
nite clausal structure with a passive meaning.
2. In the absence of any situational context identifying the utterance as an in-
stance of the genre of news headlines, the reader is likely to interpret the text
as formulated in the active voice. However, the same interpretation may ap-
ply even if the text is properly contextualized: the reader may still read the
headline on the presumption that the verb form left is a complete, finite verb
phrase expressing the past tense of the verb ‘leave’. That reading is possible if
the reader interprets the headline as uttered in the mediated voice of some
other news actor than the newspaper itself. As we saw in the previous section,
this is not impossible since the change of voice in headlines means that the
conventional style can change as well, bringing about the possibility of using
the deictic past tense form rather than the non-deictic present tense to refer
to events temporally located in the past time. As the external voice may be
accessed through a direct speech quotation or indirect speech and the head-
line may be a modified utterance rather than a verbatim representation of the
speech of an external actor, the presence or the absence of inverted commas
in the headline is not a decisive criterion for identifying whether a headline is
part of an accessed voice or not. Nevertheless, if the active voice reading was
to be encoded through the voice of the paper, it would most likely be formu-
lated not in the simple past but the simple present tense as *X leaves with three
children.
3. The reader may also be inclined towards the correct interpretation as a result
of his/her prior knowledge of the event; in that case, the information con-
tained in the headline serves to retrieve information from the reader’s mind
that is optimally relevant (Sperber and Wilson 1986) and that fits into the
previously constituted frame. The ambiguity is then unlikely to occur since
it would be cognitively less cost-effective to search for an additional inter-
pretation as long as the initial frame triggered is sufficiently plausible for in-
terpretation – the establishment of such a context stops further inferencing
processes on the part of the reader. In the case of the headline at hand, this
initial frame is made up of the knowledge that an au-pair in charge of several
children was undergoing a criminal trial (this is in harmony with the passive
reading). The second possible frame is more costly in terms of cognitive pro-
cessing: it involves the reinterpretation of the headline as encoding a change
in the au-pair’s location (corresponding to the active reading and the dynam-
ic meaning of the verb ‘leave’). Hence, it is less likely to be involved in the
process of interpretation.
4. Additional contextual and extra-textual information may also be of help in
unravelling the potential ambiguity in the headline. The cognitive frame may
be co-constituted by the verbal context, i.e. the utterances and texts placed in
the physical vicinity of the relevant headline. This type of intertextuality has
two aspects: the extra-textuality involved in the additional articles within a
cluster, mainly the most important article within the set, and the intra-textu-
al context of the actual news article opened by the headline, since potential
ambiguities are resolved in the lead and the body copy of the news text. That
is, indeed, the case with the headline at stake: the article is a part of a cluster
of articles on the same topic and it provides background information on an
aspect of the main event (‘the death of a child in the au-pair’s care’). The back-
grounding function, however, does not lead to the use of the simple past tense
(i.e. *Second au pair was left…). Instead, the headline conventions requiring
the ellipsis of the operator in the passive voice are observed here.
There are also headlines where the adverbials and the sentence structure do permit
an alternate reading, even though arriving at such an interpretation may prove to
be too cost-effective and thus rather unnatural, since the context usually resolves
the ambiguity. Let us, for the sake of argument, consider the following headline:
186 News and Time
Maria Miller reported to MP over £90,000 expenses claims

The potential ambiguity in the headline (obtaining from a decontextualized

reading) concerns, once again, the most salient effect of the passive transforma-
tion: the determination of the semantic role of the subject and, thus, the subject’s
involvement in the predicated action. In an active construction, the subject is
presented as the semantic agent of the action of ‘reporting’, while in a passive con-
struction, the subject is presented as the semantic patient. In the case at hand, the
remaining two constituents – both adverbial phrases – do not immediately help
to disambiguate the exact nature of the verb phrase, thus failing to reveal whether
the expression reported should be read as a non-finite past participle or a finite
past tense form.
The non-resolution of the ambiguity of the verb form is partly related to the
ambiguity of one of the two adverbial phrases themselves. The phrase to MP, serv-
ing the function of the indirect object of the verb ‘report’, denotes the beneficiary
in both the active and the passive reading. The only difference is that the phrase
implies a negative interpretation in case of the passive reading on account of the
negative prosody of the verb ‘report’ (cf. *Maria Miller has been reported to MP).
In the corresponding active clause (*Maria Miller reported to MP), the context is
positive or at least neutral. (The grammatical deviance of the active structure, aris-
ing from the non-expression of the obligatory direct object is discussed below.)
Similarly, the preposition over in the second phrase may – in the extreme
case – be understood in two different ways:
1. as an adverb related to the numeral; in this sense, it has the meaning of ‘more
than’;
2. as a preposition related, rather, to the verb, expressing the causative meaning
of ‘because of ’, ‘concerning’, ‘on account of ’ or ‘in connection with’.
In the first sense, the phrase over £90,000 expenses claims constitutes the direct
object of the transitive verb ‘report’, which corresponds to the active sentence in-
terpretation (the headline would then have the structure {X reported Y to Z}, with
X, Y, Z being the arguments required by the valency of the verb ‘report’). In the
second sense, the same phrase would constitute the reason for which the action of
‘reporting’ happened – that interpretation tallies with the passive understanding
of the verb phrase that does not specify the semantic agent of the action. In that
reading, the grammatical subject functions as the semantic patient, or as the di-
rect object of the corresponding active voice (*Maria Miller has been reported to
MP… → ‘X reported Maria Miller to MP …’).
More realistically, however, the above account is less a description of an actual

ambiguity than a theoretical speculation of the potentiality of the headline. There
are several reasons for why the ambiguity is unlikely to arise in this particular
case. First, when reading the headline as a monoglossic utterance issuing from the
voice of the paper (in the absence of any evidence to the contrary), the reader is
likely to process the occurrence of the verb phrase reported following the subject
as a non-finite fragment of a passive structure with an ellipted auxiliary. This is
based on the default understanding of past forms of verbs in headlines as non-fi-
nite past participles as long as they are presented through the authorial voice of
the paper. The status of the same form is quite different when presented in an
accessed voice – in that case, the default meaning is that of the past tense.
Second, in the event that the phrase over £90,000 expenses claims should be
interpreted as the direct object of the active verb ‘report’ in the past tense, the
placement of the indirect object (to MP) before such a direct object would be less
likely, bordering on ungrammatical. Moreover, the utterance would turn out un-
natural in terms of information flow, with a relatively marginal element (to MP)
located in a prominent position preceding the direct object.
Finally, the ambiguity is unlikely to arise since readers will, in all likelihood,
approach the headline with some prior contextual knowledge: not only does the
news story mention a British politician who is assumed to be known to the audi-
ence (as indicated by the paper’s representation of this social actor through nom-
ination rather than categorization) but it also fits into the cultural frame of the
misuse of public finance by politicians that has been a long-running topic in the
British media. The current socio-cultural context as well as the general awareness
of the operation of the political system will then predispose the reader towards
the passive interpretation.
All those considerations point to the conclusion that little actual ambiguity
is likely to appear in the processing of the headline as far as its active vs. passive
reading is concerned. The passive reading is most likely to dominate. This con-
clusion is eventually supported by the immediately ensuing verbal context of the
news text. The lead of the article essentially mirrors the syntactic structure of the
headline, this time with the complete finite verb phrase in the present perfect that
tends to substitute the headline present in the leads:
Maria Miller has been reported to the MP’s standards watchdog after it emerged
she claimed more than £90,000 from the taxpayer for a second home where her
parents lived.
The absence of auxiliaries, as well as the omission of other constituents from

headlines, thus places extra demands on readers since they may need to draw on
additional interpretative resources. The processing of headlines may then require
188 News and Time
more semantic interpretation, as suggested in the citation by van Dijk (1988: 144)
reproduced earlier in this section.
In this sense, the conventional and genre-defining practice of ellipting aux-
iliaries has an important interpersonal dimension. The absence of the auxiliary
implies that the writer does not necessarily commit to a particular representa-
tion, because the ellipsis of the grammatical word removes the explicit marking
of modality from the verb phrase. The readers are drawn more actively into the
process of communication because they have to reconstruct some of the missing
grammatical information on the basis of the sentence structure as well as the lex-
ical words present in the headline. This aspect of block language simultaneously
impacts the interpersonal and the ideational dimensions of the text. On the one
hand, the lack of explicitness may be viewed as interpersonally-oriented since it
contributes to the interactivity of the headline – the need to supply the missing
words increases the readers’ involvement with the text. On the other, the ellipsis
of auxiliaries (and other grammatical words, for that matter), may help to focus
the readers’ attention on the semantic content of the headline. The ideational (se-
mantic) and the interpersonal (pragmatic) functions of headlines are thus in a
very close connection.
8.3 Explicit use of auxiliaries
While headline conventions call for the omission of items of low information
value such as articles, auxiliaries and other grammatical words, there are also sit-
uations when such grammatical words are actually present in headlines. While
the absence of auxiliaries is unmarked with respect to the genre, their presence
constitutes a violation of the unstated rules and needs to be considered as marked.
The aim of the present section is to analyse the data with respect to the explicit use
of auxiliaries and suggest an explanation to account for their usage, as illustrated
in the following examples:
Balotelli is going nowhere, insists Mancini (The Times; 11.12.2012)
Cameron: Britain would fight another war to keep the Falkland
Islands (The Independent; 7.1.2013)
Mandela is back to health, says Zuma (The Independent; 7.1.2013)
Reflecting on the practice of headline writing in print newspapers, Rafferty

(2008: 215) notes that the inclusion of auxiliaries is sometimes motivated by the
editors’ need to fill space in vertical headlines that spread across several decks.
The auxiliaries may thus, at least in some cases, be present merely for technical
reasons connected with the typographical presentation of headlines. Despite that

qualification, however, there appears to be sufficient evidence among print and
online headlines to allow us to formulate several possible linguistic explanations
for the presence of auxiliaries in headlines.
The data indicate that there are at least three broad types of partially overlap-
ping situations when auxiliaries are retained. These include (a) headline segments
constituting accessed voice (both direct and indirect quotations); (b) headlines
where semantic specification and precision is sought (e.g. to prevent ambiguity);
and (c) headlines that report either major news events or culmination points in
stories developing over an extended period of time (i.e. in continued coverage).
It is argued that in all cases, the retention of auxiliaries serves the purpose of en-
hancing the interpersonal dimension of the headlines because their presence can,
in all three types, be explained as essentially reader-oriented.
Accessed voice and reported speech
As mentioned previously, heteroglossia is a very powerful phenomenon affecting

the formal composition of headlines. Not surprisingly, those headline sections
that represent the content as encoded from other points of view than the paper’s
typically include auxiliaries and modal verbs. This affects textual segments pre-
sented in both direct and indirect speech.
The following examples illustrate the usage of non-ellipted auxiliaries and
modal verbs in selected direct and indirect speech quotations. They include het-
eroglossic headlines combining the voice of the paper with the external voice (the
first headline), as well as monoglossic headlines where the accessed voices make
up the entire headline without any juxtaposition between the external voice and
the authorial voice of the paper (the second headline).
Mandela’s wife: his spirit and sparkle is fading (The Times; 11.12.2012)
‘Britain is becoming a nation of renters’ (The Telegraph; 11.12.2012)
To those examples, we may add a similar heteroglossic situation – when an ac-

cessed voice, presented as an indirect quote, retains the finite verb form of the
copula verb ‘be’, which is commonly found to be ellipted as well:
Internet access is a right, judges rule (The Telegraph; 13.11.2012)
The differential treatment of the verb phrase depending on whether it is located in

the voice of the paper or the accessed external voice is apparent from the follow-
ing two examples. The former is a non-finite sentence fragment arising from the
190 News and Time
conventional ellipsis of the auxiliary verb, and the latter is a structurally identical
headline with a complete and finite verb phrase containing a non-ellipted auxiliary.
NHS using dementia beds for norovirus patients (The Times; 7.1.2013)
‘China is blazing a trail for clean nuclear power’
Heteroglossia means that two (and exceptionally more) discourses are present
in the headline. The text thus becomes plural since different subjectivities can be
traced in it (cf. Fowler 1996: 197). The switch between the discourses is explicitly
indicated to readers through the use of inverted commas, which also serve to
mark the discontinuation of headline conventions. The retention of auxiliaries
has an objectifying function in that it provides access to a (seemingly) authentic,
unmediated utterance – one that is not regulated by the textual conventions of the
genre in which the fragment is recycled. It also makes the headlines more dynam-
ic and interactive because the readers may experience the feeling of being direct
participants, as if overhearing a fragment of an authentic discourse. What matters
is the effect of creating a make-believe reality rather than a verbatim representa-
tion of the actual words belonging to the discourse of another news actor. As van
Dijk observes: “That quotations are seldom fully correct contextually is irrelevant.
They should merely suggest that they are true, hence their rhetorical function and
effect” (1988: 87).
The significance of direct speech quotations also rests in the fact that they
open a space within the otherwise impersonal hard news headline to the expres-
sion of personal opinion and evaluation in all its forms (cf. White 2006; Bednarek
2006). While in the textual segments moulded in its authorial voice the newspa-
per must strive for impersonality and objectivity, there is no such limitation in
the textual segments cast in the accessed voice. Indeed, newspapers often use het-
eroglossia as a strategy for making evaluative statements that they would not be
able to put forward as their own representations of the reality, being constrained
due to professional, ethical, legal etc. reasons. Heteroglossia allows the media to
project the responsibility for the evaluation onto the accessed voice of the rele-
vant news actors. This is nowhere better seen than in the case of particular lexical
choices whose heteroglossic origin underlies their highly evaluative or potentially
controversial nature. Consider the two headlines that occurred in The Times side
by side on the same day:
Silbury Hill ‘scarred’ by trespassers (The Times; 8.1.2013)
Debt-ridden NHS trust should be ‘broken up’ (The Times; 8.1.2013)
By using the quotes, the newspaper externally manifests its non-identification

with the evaluation expressed. However, such a usage need not imply that the pa-
per is distancing itself from the evaluative proposition through what is commonly
known as the scare quotes – it may merely be exercising its encoding power to
express certain content through the accessed voice of someone else and enjoying
the benefit of not being held responsible for such a representation. Such eviden-
tial or hedging use of quotes is markedly different from the main function of
scare quotes to create a distance between the external opinion expressed and the
editorial line of the paper. Although it may be hard in individual cases to draw
the boundary between true detachment and the release from responsibility (the
‘exemption’ function of quotes), the role of heteroglossia needs to be affirmed in
the way the paper can use the external voice to navigate the area of evaluation and
ideological positioning.
The presence of the external voice has another ideological consequence: it
serves to underline the distance between the journalist and the story. As a re-
sult of its juxtaposition with a clearly subjective and personal external voice, the
‘neutrality’ of the authorial voice of the paper (or the journalist) can be indirectly
enhanced. As noted by Cotter,
[q]uoting someone or using another’s words allows a distant or ‘distal’ position-
ing of the reporter with respect to the content and the source of information.
[…] The discursive purpose of attributing a source from the news community’s
point of view is to render the reporter’s voice neutral, to position the reporter
as a conduit, or an objective party that is not taking a stance about the topic
being covered but is merely conveying a range of views. Thus attribution and
quotes themselves, whether direct or indirect, are important to front the speaker
or source and his or her stance and to take the opinion of the reporter out of the
equation. (Cotter 2010: 146)
Semantic specification
The accessed voice is sometimes used to provide a semantic specification because

the presence of the voice enables the paper to include modal verbs and auxiliaries.
While in the last example above (Debt-ridden NHS trust should be ‘broken
up’), the modal and the non-finite auxiliary (should be) are cast in the authori-
al voice of the paper, in the following headline, the grammatical words have a
different status: they are enclosed within the accessed voice. This is a case of an
unattributed direct speech quotation encoding an evaluative standpoint. Inverted
commas are needed in order to indicate that the proposition conveyed is to be
192 News and Time
associated with some external news actor, without the paper assuming responsi-
bility for the claim:
NHS trust ‘should be dissolved’ (The Guardian; 8.1.2013)
In general, heteroglossic headlines of this type are interactive in that they include
evaluative and personal utterances, though they frequently do not provide any at-
tribution. In this way, they attract readers’ interest and potentially motivate them
to follow up the headline into the lead and/or the body copy with the aim of
identifying the source of the quote. Since they omit attribution, i.e. information
that is to be found only in the subsequent textual segments of the news story,
headlines of this type are incomplete and leave a clear semantic gap that readers
may feel needs to be filled. Unattributed and visually unanchored heteroglossic
headlines of this type thus actually go counter to the widespread belief that head-
lines are stand-alone semantic units. Quite on the contrary: such headlines are se-
mantically incomplete since they fail to disclose the identity of the external voice
(unless the attribution is achieved, for instance, through multimodal, i.e. visual,
anchorage).
This means that heteroglossia can weaken the relative textual independence of
the headline by increasing its reliance on the subsequent textual segment, namely
the lead, subhead or the body copy of the article (or the accompanying visual
element, as the case may be). Let us illustrate this idea with another example that
has the full verb ‘be’ in the past tense and with a part of the headline marked off in
inverted commas as an accessed external voice. Here, the past tense indicates the
non-validity of the proposition, i.e. there are specific semantic reasons why the
full verb is not in the conventional present tense in the headline:
Waterfall tragedy tourist was ‘living my dream’ (The Times; 8.1.2013)
The headline of this story uses a fragment of an accessed voice with the personal
pronoun ‘my’. However, the exact reference of the pronoun is not entirely clear.
The pronoun is part of a text in quotes and thus belongs to some external voice
whose identity is not only left unspecified but also cannot be inferred from any
other signals. There is no context provided and – in the absence of a picture – the
pronoun cannot be contextualized through visual anchorage either. Readers can
hardly be expected to have some prior knowledge of the individuals involved,
even though some minimal familiarity with the topic may be assumed.27 Thus,
27. The same newspaper broke the story the previous day under the headline Briton falls to
death over Australian waterfall (The Times; 7.1.2013). Thanks to the previous story, the
paper is able to use the complex nominalization Waterfall tragedy tourist in its headline on the
next day, which indicates that some prior background knowledge of the incident may in fact be
the crucial place where more information can be obtained is the next textual seg-
ment – the lead:
Josh Furber, 20, tweeted that Australia had ‘my heart and soul’ weeks before he
plunged 300 feet to his death in Australia.
The lead establishes the semantic relation of co-reference between ‘tourist’ and
‘my’ in the headline and rules out the possible interpretation that the quote medi-
ates the words of some other person commenting on the tourist, rather than the
tourist commenting on himself. The situation in the headline is tricky because the
paper could have reformulated the utterance in its own voice (*Waterfall tragedy
tourist was living his dream) rather than resorting to the less usual use of a self-
quote by a deceased person: the heteroglossia may give out the impression that
some other person is involved, even though the headline is technically unambig-
uous. The body copy of the article makes it further clear that the accessed voice
actually constitutes the content of the dead tourist’s recent tweet (note also the
imprecise citation in the headline which leaves out the adverbial particle ‘out’):
A young English tourist in Australia tweeted that the country had his “heart
and soul” just weeks before he plunged to his death from a cliff-top in the Blue
Mountains national park.
Josh Furber, a 20-year-old student, said that he was “living out my dreams” in
Australia shortly before he fell from the top of the Wentworth Falls, a popular
tourist site, on Sunday afternoon. (The Times; 8.1.2013)
As noted earlier, the accessed voice tends to explicitly encode auxiliaries rath-
er than ellipt them, thus providing a semantic specification of the proposition.
Apart from operators, this affects modal auxiliaries, which increase the presence
of evaluative features in the headlines. Once again, while absent from headline
segments in the paper’s voice, they are commonly included in direct and indirect
quotations, cf.:
Vaz: Libor trial ‘must be held in UK’ (The Times; 21.12.2012)
Noah’s Ark and Great Flood ‘may have happened’
An auxiliary may occasionally be emphasized through typographic means (font

size, underlining, etc.) in order to indicate that the relevant meaning associated
with the auxiliary is central to the news value of the story. That is also the case of
the modality of certainty associated with the indicative mood in the next example:
assumed – the nominal phrase is used as a thematic element since the focus of the story is on
the tourist’s prior enjoyment of his Australian experience.
194 News and Time
Ex-MP Margaret Moran DID cheat £53,000 of expenses

The capitalization of the auxiliary emphasises the reality status of the entire prop-
osition that can, in most likelihood, be expected to be known to the readers from
their exposure to previous media coverage. The entire headline – except for the
auxiliary – thus provides previously known (thematic) information.
An auxiliary may also be present in order to prevent a possible incorrect in-
terpretation of a headline. That appears to be the motivation for using the auxilia-
ry ‘are’ in the following example:
South African officers involved in death of man dragged by police
van are suspended (The Guardian; 1.3.2013)
Should the auxiliary be ellipted, as is usual in passive constructions, the past par-
ticiple form might be interpreted as being related to the immediately preceding
nominal group (…police van suspended). The possibility of such a reading is inev-
itably ruled out by the resulting semantic incongruity as well as the general pro-
cessing of the headline, where readers expect the provision, later in the headline,
of some newsworthy information concerning the news actors mentioned at the
beginning (South African officers). It is also not insignificant that the sentence
already includes two non-finite past particles: both of the forms involved and
dragged are residues of conventionally condensed and ellipted relative clauses.
Adding a third ellipted form might make the headline syntactically too opaque.
A very similar situation appears in the next headline that uses the full finite
form of the verb phrase (are found) following two previous non-finite past par-
ticiple verb forms. By formulating the most important verb phrase with a finite
verb form, the headline writer avoids the possibility of a ‘semantic overload’ pos-
sibly arising from a cluster of non-finite past participles in a syntactically complex
headline:
‘This is insane!’ Couple feared abducted in Peru are found – kayak-
ing up the Amazon (The Independent; 1.3.2013)
Intuitively, the ellipsis of ‘are’ would appear to be possible if the headline was for-
mulated without the dash (*…Couple feared abducted in Peru are found
kayaking up the Amazon). In that case, the semantic centre of gravity of the
headline would be the circumstances under which the couple was found (kay-
aking up the Amazon). By contrast, as the headline stands, the most impor-
tant element is arguably the information about the finding of the couple, with the
circumstances added almost as an afterthought to increase the news value and
justify the evaluative quote that opens the headline (‘This is insane!’).
As mentioned previously, the explicit presence of auxiliaries is not dependent

on the formal marking of textual segments in headlines as belonging to the ac-
cessed external voice. The discontinuation of headline conventions also applies to
indirect speech and semi-direct speech, be it attributed to the source by means of
a reporting clause or the colon structure that conventionally removes the speech
act verb. Cf. the following heteroglossic headlines that are clearly made up of two
distinct textual segments, yet lack any formal marking of the two voices that are
merely placed side by side:
Labour: benefit cuts will hit women (The Guardian; 7.1.2013)
Gay marriage will be illegal in CoE, minister says
(The Times; 11.12.2012)
The interpersonal dimension of headlines is thus strongly enhanced by the inclu-

sion of auxiliaries and modals since they are associated with conveying personal
evaluative attitudes. Their association with heteroglossia, as well as their evalua-
tive character, serve to indicate that the news story is mediated through a textual
account that is based, to a certain extent, on a subjective perspective on the event.
By extension, they help us to see that the presentation of news is not a matter of
a transmission of some depersonalized collection of objective facts. Clearly, the
presentation of news content through headlines proves to be a complex phenom-
enon standing at the intersection of various voices and subjectivities.
Foregrounding of major news stories
The previous subsection has touched on an interesting situation that arises when
explicit auxiliaries appear within textual segments classifiable as the paper’s au-
thorial voice (i.e. not within accessed external voices). This means that the author
has a choice of either ellipting the auxiliary – in harmony with the headline con-
ventions – or preserving it despite the conventions. The presence of the auxiliary
thus stands in contrast to its potential absence, as can be attested by the usual
omission of the same auxiliaries in headlines containing similar structures (typ-
ically the passive). Such an auxiliary – elliptible yet present within the paper’s
voice – occurs in the following examples:
IQ tests are ‘fundamentally flawed’ and using them alone to meas-
ure intelligence is a ‘fallacy’ (The Independent; 20.12.2012)
It appears that the full verb phrase may be used in those headlines where there
is a need to put some emphasis on the news story. The presence of the auxiliary
has, by virtue of its elliptibility, a certain foregrounding effect. This is because the
196 News and Time
auxiliary – as a grammatical word rather than a lexical one – makes only a mini-
mal semantic contribution to the headline and is thus comparatively dispensable,
unless the importance of the grammatical information conveyed through the aux-
iliary becomes crucial for the given proposition (e.g. to express specific degrees of
epistemic modality). It is because of its potential redundancy that the presence of
an auxiliary becomes noticeable and marked, thereby attracting extra attention.
It has been documented in a previous study that the foregrounding use of
elliptible auxiliaries can also be attested along the diachronic dimension of the
coverage of some news event. Based on the observation that articles are not in-
dependent stand-alone units but form an intertextual chain that builds up and
develops a complex story of a single event over time, Chovanec (2000) analysed a
series of all articles from one paper on a single international topic over the period
of several months. The material, obtained from an online newspaper (The Tele-
graph), revealed that elliptible auxiliaries appeared only twice within the authorial
voice of the paper. However, each of those occurrences coincided with a crucial
moment in the story, marking its major turning points, cf.:

British au pair is charged with baby murder (12.2.1997)

Louise Woodward is freed (10.11.1997)
In each headline, the full form of the verb phrase can be interpreted with re-
spect to the diachronic development of the story. In the Woodward case, there
were three major moments in the reporting that marked particular points of cul-
mination. They were: first, the formal accusation on February 11 (with its legal
implications of an ensuing court trial, marking the end of media speculations
and the beginning of a scheduled, institutionalized story); second, the verdict and
conviction on October 30 (the key point in the court proceedings); and third, the
judge’s highly unexpected decision to overturn the jury’s verdict and, thus, give
freedom to Woodward on November 9. On two of those occasions, the points of
culmination in the continued coverage were reported in headlines by means of
complete verb phrases that included the elliptible auxiliaries, suggesting a possi-
ble link between the foregrounding function of the explicitly expressed auxiliaries
and the function of indicating a crucial update on a prolonged news story. The
third occasion was encoded without the finite form of the copula verb ‘be’, which
was ellipted (Au pair guilty of murder, 31 Oct.). (A graphic representation of
the series of articles in provided in Figure 4.10, which also indicates the hypertex-
tual centrality of the respective articles.)
The interpretation of the full verb forms as being associated with a turning
point in the continued coverage of a story is also based on the existence of several
instances of structurally very similar, monoglossic headlines in the paper’s own
voice that omitted the auxiliaries (e.g. English nanny accused of harming
baby (7 Feb.), Au pair accused of first degree murder (6 Mar.), Baby’s fam-
ily set to sue Woodward agency (4 Nov.), etc.). Although those headlines
brought some important and newsworthy information about the case, they did
not constitute such major watershed moments as to deserve the full use of the
auxiliary.
It seems that the auxiliary may serve to emphasize the present relevance of
the event (on account of the ‘actuality modality’ that it encodes, cf. Kress 1977)
with the stress being placed on the finalizing aspect of the story, almost as if there
was no possibility for a potential change. The operation of the emphasizing func-
tion (not necessarily corresponding to the spoken usage with a prominent stress
on the auxiliary) is enabled through the implicit assumption that an unmarked
headline (i.e. non-emphasizing, non-perfective and primarily factually informa-
tive) does not use auxiliaries on account of the conventional requirement of block
language to ‘omit auxiliaries if they can be omitted’. On this count, all three poten-
tial categories in which auxiliaries are retained (direct/indirect reported speech;
semantic grounds/disambiguation; emphasis of key diachronic moments) would
then be justified as representing the marked usage because in all of them, a special
interpretation follows from the use of auxiliaries.
The actuality modality, encoded through the finite form of the auxiliary ‘be’ in
verb phrases formulated in the passive, disappears in headlines with ellipted aux-
iliaries. Instead, it can be presupposed. The re-appearance of the auxiliary (other-
wise suppressed there because of its relatively low information value) constitutes,
vis-à-vis the headline conventions, a foregrounded assertion of the actuality mo-
dality. This interpretation supports the conclusion about the ‘watershed moment’
use of auxiliaries in headlines.
The explanation proffered here does not aim to be universal. Although the
writing of headlines is a skill learnt and acquired by editors through practice as
well as instruction, the formulation of actual headlines is, to a significant extent,
an intuitive process with the editors trying to adhere to the conventions of the
genre to make their headline maximally effective. Some encoding options may
come as a result of conscious choice motivated by the requirements of particular
media, e.g. the space available within a newspaper column for the headline (cf.
Rafferty 2008: 215).
While the headlines analysed here are all taken from hard news, where we find
the most canonical materializations of block language, there are other, less canon-
ical types. Thus, headlines introducing articles in other sections of newspapers,
particularly in such areas as editorial comments, lifestyle, culture, travel, as well as
sports, tend to be cast in a more personal style. The application of the conventions
198 News and Time
of headlinese is strongest in hard news headlines and correlates with the internal
voice of the paper since it is connected with the impersonal presentation of infor-
mation. For the sake of variety, however, newspapers will occasionally include –
within the hard news section – even such headlines that are formulated in a more
personal style and that bear features of less conventional language (e.g. headlines
in the form of questions and conversational utterances), cf.:
What have the Mayans ever done for us… apart from predict the
end of the world? (The Independent; 19.12.2012)
It did seem rather a lot of trouble to go to just to acquire a new
set of place mats (The Independent; 19.12.2012)
The ellipsis of auxiliaries in headlines is thus governed by a number of factors

that affect the editor’s eventual choice of whether to omit or preserve elliptible
auxiliaries. In addition to the presence of an external voice, the need for semantic
specification and the highlighting of some event, the choice may be related to the
section of the paper the story belongs to, the variety of headline types already
included in the section, as well as other potential factors such as the nature of a
visual element accompanying the story or simply the space available to the editor
when typesetting the headline within newspaper columns.
There appears to be much promising ground for further research in this area,
particularly concerning the cross-genre comparison of headlines, their grammat-
ical and heteroglossic composition and the degree of the textual presence of the
author of the text.
8.4 Complex headlines with subordinate clauses
Let us briefly turn attention to syntactically more complex headlines that come as
a combination of two clauses or contain subordinate clausal elements. It appears
that such headlines are also favourably inclined to express auxiliaries in their full
forms.
In some headlines, the main newsworthy event is encoded by means of a
non-finite clausal structure that describes some end-result situation represented
as a state. The subordinate clause that supplies additional information by elabo-
rating on the circumstances of the main event is, however, often formulated in a
more dynamic way by means of a process verb that retains its finite form. Cf. the
following examples:
Nine-year-old girl in Halloween costume shot after relative mis-

takes her for a skunk (The Independent; 22.10.2012)
Santa sacked for telling children he isn’t real
Both headlines encode the main event in a static, non-finite way (cf. Nine-year-
old girl in Halloween costume shot and Santa sacked). The temporal clause in the
first headline (after relative mistakes her for a skunk) and the non-finite structure
expressing reason in the second (for telling children he isn’t real, NB the embedded
finite object clause with a trace of the news actor’s accessed voice) express cir-
cumstances that preceded or directly led to the main event constituting the news
story. However, the main event is not sufficiently newsworthy in itself: it is only
in combination with the circumstances that had led to it that it becomes worthy
of inclusion in the paper. As a result, the headline constitutes a mini-narrative,
rather than focusing on just a single component of the news story.
The order in which the two events originally occurred is reversed: the head-
line opens with the later event, i.e. the eventual result. Grammatically, this reverse
placement is required on account of the nominal form in which the main event
is encoded; a bare nominal form cannot be post-positioned after an element that
modifies it.
As Biber et al. observe, the reverse ordering of a series of events is common
in news articles beyond headlines as well, where the subordinator ‘after’ “can also
have strong overtones of cause” (1999: 844). Similar temporal-causative reading is
also apparent in the following example:
Government to probe Comet collapse after backers took £13m in
fees (The Telegraph; 18.12.2012)
Moving on to other ways in which the conjunction between two elements is pre-
sented, the placement of two events alongside each other is particularly common
in clauses introduced with the subordinator ‘as’. Similar to the conjunction ‘after’,
clauses with ‘as’ can also express various semantic relations ranging from tempo-
ral to causative (cf. Biber et al. 1999: 846). In the following examples, ‘as’ expresses
simultaneity (the first two headlines) and introduces a clause specifying the cause
for the main action (the latter two examples), though in some cases the two func-
tions may actually overlap, cf.:
– As for expressing simultaneity

Flood warnings as getaway begins (The Times; 20.12.2012)
Flooding threat as heavy rain hits UK (The Guardian; 5.10.2012)
200 News and Time
– As for expressing causativity

‘Plebgate’ arrest as PM hints at comeback (The Times; 20.12.2012)
Ex-traders charged as Hong Kong joins probe into UBS Libor rig-
ging (The Telegraph; 20.12.2012)
Clauses introduced with ‘as’ do not necessarily encode events located in the past;
they also refer to present situations as well as future predictions. Although the
tensed clause is syntactically subordinate to the initial nominal or non-finite el-
ement, it conveys information that seems to be slightly more newsworthy than
the simple provision of circumstantial background information (cf. the earlier
examples with the subordinators ‘after’ and ‘for’).
We can see a clear pattern emerging of a combination of either a nominal
element or a non-finite clause with an ellipted operator, and temporal/causative
clause that is temporally subordinate yet finite. The pattern is also attested with
examples where the subordinate clause is in the full, non-ellipted passive that
retains the relevant operator (the present tense of the verb ‘be’):
OTF left red-faced as Tesco cheese ruling is overturned
Accidental death verdict quashed as new inquest into Hillsborough
disaster is ordered (The Independent; 20.12.2012)
However, the full finite verb phrase in the subordinate clause is not the rule. The
operator can, of course, be ellipted as well, giving rise to two parallel non-finite
propositions expressing concurrent events. The absence of the operator might
result in some potential ambiguities. Thus, in the following example, the subor-
dinate clause could be interpreted as being in the active rather than the passive
voice, thus rendering a slightly different meaning to the headline. The possibility
of real ambiguity is, of course, minimized by the semantic interplay between the
propositions, cf.:

Cameron to halve Afghan troop numbers as early exit revealed
The combination of more than two events in a headline is possible but rare be-
cause it significantly increases the complexity and length of what is already a spa-
tially severely constrained text. Semantic complexity is normally not an issue in
headlines since complex multiple nominalizations (cf. the noun phrase site de-
bunking end of the world myths below) and other non-finite elements commonly
condense numerous processes and underlying propositions in non-clausal ways.
The following headline is unusual in that it includes three tensed verbs in what
is essentially an extended variation on the dual-structure headline with the sub-

ordinator ‘as’. The extension consists of the addition of a main clause with the
coordinator ‘and’:
Calls to Nasa triple and site debunking end of the world myths
gets 4.6m views as anxious US prepares for Armageddon
What we can thus observe is that headlines manifest a significant variability in

the way temporal deixis is expressed. Tensed forms alternate with non-finite verb
phrase fragments while elliptible elements are sometimes omitted and, at oth-
er times, retained. It appears, however, that the variability is not haphazard but
clearly patterned and, to a certain extent, predictable.
8.5 Concluding remarks on temporal deixis in headlines
All in all, the analysis of the various aspects of temporal deixis in headlines in-
dicates that the expression of temporality has a strong interpersonal dimension.
There are several pragmatic considerations involved in this situation. Most im-
portantly, verbal tense is not used deictically to refer to the relevant time of oc-
currence of an event that is being reported. There is a conventional system in
operation that re-codes the temporality expressed through the verb phrase by
promoting the present tense into the default tense choice in headlines. This phe-
nomenon can be explained as the result of a complex projection of the deictic
centres involved, mainly on the levels of the occurrence of a news event, the en-
coding of the news item into a news story, and the eventual reception of the news
story by the readers. The tense shift process that privileges the present tense partly
draws on the universal meaning of the tense and partly serves to compress the
discourse space between the news event and the reader, creating an impression of
immediacy. Shifted verbal tense is thus not geared towards satisfying the ideation-
al (referential) function of providing information. Instead, it is reader-oriented
and enhances the interpersonal dimension of the text.
At the same time, however, the newspaper extensively relinquishes its au-
thorial voice to a number of external voices. When such accessed voices appear
in headlines, they are included as pieces of discourse not governed by the con-
ventions of headlinese. Regardless of whether they are cast in direct or indirect
speech, such headline segments with accessed voices use verbal tense deictically,
i.e. without the conventional shift of tenses. Tense in accessed voice is more objec-
tive than in the paper’s authorial voice because it is used deictically: it refers to the
time of occurrence of a given event. Although the deictic use of tense in accessed
202 News and Time
voice puts a distance between the news event and the reader’s decoding time,
heteroglossia has the effect of discontinuing headline conventions. As a result,
heteroglossic headline segments communicate in a style that is more personal
and conversational than the typical headline block language. That conclusion is
confirmed by the increased presence of the deictic past tense, modal verbs and
complete (unellipted) verb phrases in headlines.
Part III
Textual rhetoric of news texts

Chapter 9
The textuality of news texts
While the headline – as the most prominent structural feature of news articles –
certainly deserves the extensive attention paid to it in the previous chapters, it is
only the consideration of this element in the context of the entire news story that
can give full justice to the deictic situation in the headline and beyond. The pres-
ent approach to analysing the structure of news stories distinguishes three basic
textual components that reflect the organization of the news text into the inverted
pyramid pattern. Those components, present in most hard news stories, are the
headline, the lead, and the body copy. The explanation of the transformation of
the tensed verbal phrase across these segments constitutes what I refer to as the
‘textual rhetoric of news texts’, as distinct from the situation in headlines only.
The present chapter concentrates on the linguistic realization of the main pro-
cess chain that starts with the finite verb phrase in the headline. The progression of
the chain is then traced across the initial structural segments of news stories that
provide a summary of the main event and repeat the basic propositional content
expressed by the verb. Special attention is paid to the new textual model of online
news, which articulates news content on two levels, with the article preview on
the home page adding two more textual elements to the news story. It is argued
that while the temporal deixis of the verb phrase has a pragmatic function that
arises from the choice of tenses, the notional component of the verb phrase is not
only ideationally but also textually motivated since its repetition contributes to
the coherence of the news text. The chapter documents the double tense shift pat-
tern and some of the variations that the pattern undergoes in various news texts.
The macrostructural perspective adopted here can help us to identify another
pattern of tense use that is typical of news stories. In addition to the conventional
shift of the present tense in headlines described in Chapter 6, there is the temporal
pattern consisting of the three-move structure ‘present + present perfect + past’
that is spread out across the three textual segments of news articles. Although this
double tense shift pattern impairs the traditional function of verbal tense as a co-
hesive device, it does not affect the overall coherence of the text, which is formally
achieved through notional components of verbs and pragmatically through the
reader’s expectation of local coherence. The discussion offers an explanation of
the phenomenon within the framework of the three language metafunctions by
206 News and Time
identifying their role within the verb phrase. It is shown that the dual structure
of the verb, consisting of the combination of temporal/modal exponents and the
notional component, has the following main consequences: (1) the lexical com-
ponent is utilized for the ideational purpose of information transmission; and (2)
the temporal component can serve for the negotiation of the interpersonal space
between the text and the intended recipients. In that sense, the verbal phrase
combines the semantic and the pragmatic functions.
9.1 Textual segments: The headline and beyond
Before tracing the development of the verbal process chain and the transforma-
tion of tense, let us briefly outline the mutual relationship between the three main
structural segments of news texts that I distinguish: the headline, the lead, and
the body copy. The distinction stems from the functional differentiation of the
three segments. The headline provides the textual opening to the article, which is
a function that can be carried out in various ways, e.g. by summarizing the story,
by quoting a source, by commenting on a selected aspect of the event, etc. (cf.
van Dijk 1988; Bell 1991: 188). The lead has a predominantly semantic function:
by summarizing the story, it provides all the basic information about the news-
worthy event (who, what, when, where and how), usually in a single sentence.28
Finally, the body copy, as distinct from the summarizing lead, consists of the nar-
rative verbalization of the newsworthy event. It is typically recounted in a com-
plex non-chronological manner.
Modern news stories contain headlines and leads that can be classified into
various types – summarizing headlines, quotation headlines, summary leads,
delayed leads, second-day leads, etc. There are also some newer discourse-level
units, such as the ‘nut graf ’, which has developed as an additional textual summa-
rizer for attracting the readers’ attention. Placed in the body copy of the article,
it is defined as “a compendiary or synopsis paragraph […], much like the brief
summary that reporters are trained to give editors” (Cotter 2010: 160). The pres-
ent approach consciously simplifies the complex situation in actual news texts in
order to focus on the ‘idealized’ traditional format of the hard news story as con-
sisting of the three structural segments outlined above, with the headline and the
lead taken as initial summaries of the main event. My focus on the summarizing
28. Bell (1991) has found the average lead to be 30 words long, while van Dijk’s (1988) data
indicate the length to be 25 words. In soft news and features, other types of leads may be found,
e.g. the delayed lead that does not reveal all the information and postpones the most newswor-
thy element (Cotter 2010: 154).
Chapter 9. The textuality of news texts 207
headlines is linked to the need to work with linguistic forms that recur in all three
of the segments; that situation, however, does not apply in the case of many quo-
tative headlines or headlines that, instead of summarizing the main event, open
the story with some other component of the news story, such as consequences,
verbal reaction, etc.
The analysis of temporal deixis across the three segments is closely related
to the repetition and re-expression of information in the headline, the lead and
the body copy. Because of the double summary pattern, the verbal phrase can be
easily investigated in terms of cohesion analysis as constituting a co-referential
chain. Together, the units that make up such a chain represent a single structural
and semantic-pragmatic macrofield which enables us to trace the way the tempo-
ral information about the event is negotiated in the three elements, as long as, of
course, the chain is realized through some tangible formal units within all those
elements.
9.2 Cohesion analysis
As suggested above, the interpretation of the function of verbal tense in the lead
(as well as in any other structural segment of news texts) is incomplete without
considering the actual choice of tense within the broader context provided by the
surrounding segments of the headline and the body copy. This macrostructur-
al perspective on the issue lends itself well to cohesion and coherence analysis,
which looks at the way connected texts hold together (Halliday and Hasan 1976;
Hasan 1984; Halliday and Hasan 1985) and how they form hierarchically organ-
ized structures (Hoey 2001). The advantage of an approach inspired by cohesion
and coherence analysis is that it pays close attention to the formal realization of
the news text on the level of individual words, yet it does not consider the indi-
vidual items in isolation.
The aim of cohesion analysis is to identify formal links existing between the
individual expressions within texts. The links between a cohesive item and some
previous item is called a cohesive tie. In the classic account, cohesive resources are
classifiable into several types, such as reference, ellipsis, substitution, conjunction
and lexical cohesion (Halliday and Hasan 1976; cf. Martin 2001). The tie is the
relationship between items, which are classified together on the basis of co-ref-
erence. The sum of mutual relations between items can be viewed as a chain or a
string that stretches throughout a text. The whole text is then composed of a series
of co-referential chains that run parallel to each other and come to interact in
complex ways. Cohesion functions as a text-forming strategy that helps to create
coherence by signalling the unity of texts (cf. Tanskanen 2006: 13).
208 News and Time
At the same time, cohesion analysis is not just a mechanistic parsing of texts
into their formal constituents across sentence boundaries. It combines with co-
herence analysis, tracing how the logical connectedness of the text is formed. The
analysis thus also documents how a co-referential chain gradually develops and
how it expands by acquiring additional aspects of meaning. It also considers the
grammatical transformations that the individual cohesive items undergo within
the chain. The chains should then be seen as being less ‘lexical’ and more ‘con-
ceptual’, since the items that form a link may be realized in quite diverse ways
and even through textual absences and omissions (cf. the ellipsis of the semantic
agent in passive constructions, which leaves a textual gap that is not realized on
the surface of discourse, yet its absence may be significant in the relevant chain
that runs throughout the text). A concept initially lexicalized as a noun can be
transformed into a different part of speech at a later position in the chain, e.g.
into a verb, an adjective, an adverb, etc. Whenever verbal predications are trans-
formed into nominalizations, such transformations can be analysed as instances
of grammatical metaphor (Halliday 1994) that may make the cohesive relations
somewhat more opaque.
9.3 Information chaining
The cohesive elements within chains not only signal coherence to the reader but
they also interact with each other. The chains can thus be brought into various mu-
tual relationships, become connected, intertwined, etc. (cf. the notion of cohesive
harmony analysis elaborated by Hasan 1984, the linear analysis of patterns pro-
posed by Hoey 2001, and the chains of cohesion developed by Tanskanen 2006).
The relationship between chains is most evident in the way the individual
expressions combine within a clause. Each lexical unit is then linked with oth-
ers in two ways: first, it relates syntagmatically to the surrounding expressions,
e.g. on account of its syntactic role in the clause, and second, it is a paradigmat-
ic realization from among a set of possible expressions that could appear in the
same position within the clause as a part of a particular co-referential chain.29
The chain runs through the text alongside other chains with which it comes into
contact whenever the particular cohesive items from such chains are brought into
a mutual syntactic relationship. From a more critical discourse perspective, it is
29. The paradigmatic set is not a matter of the language system as such. It is the domain of the
‘parole’ rather than the systematic structure of language – it is made up of those expressions that
a speaker may decide to use within a particular position in the sentence, so it is a matter of the
speaker’s performance.
also significant which chains are brought into mutual contact, whether the con-
nection is made repeatedly, which chains are purposefully kept apart and how the
relevant entities are represented within the chains. Where a chain does not have
any explicit realization, the absence of the representation of the relevant entity
may likewise result from various motivations.
Hard news is characterized by a very specific cohesive pattern in which several
co-referential chains may repeatedly interact in an almost identical manner. This
phenomenon, which is common in the lead and the headline (particularly its sum-
marizing type), comes as the result of the summarizing function of the two textual
segments operating as narrative abstracts summing up the news story. While the
lead typically expands the information provided in the headline, the body copy
frequently opens with the same syntactic construction and reiterates the same
content for a third time. The triple repetition of the semantic proposition in those
three structural segments of news texts and the parallel placement of co-referen-
tial items that make up those propositions allow us to trace how the information is
gradually developed in an incremental fashion. The headline is the starting point
for cohesion analysis in news texts since it opens several co-referential chains
whose elements are brought into a mutual relationship for the first time.30
Such parallel mapping of information recurring across the different structur-
al segments of news stories as a result of the mutually interacting co-referential
chains is by no means exceptional. However, it is very rare to find a news text that
repeats the same lexical form of a verb in the headline, the lead, and the opening
of the body copy. Arguably, such repetition could diminish the news value of the
story for the reader. Thus, rather than expressing the same information in exactly
the same way, newspapers use several strategies to sustain the newsworthiness of
the item despite the structural parallelism and recurrence of the same concepts
arising from the double abstract summarization pattern. The lead and the body
copy then enrich the content encoded in the (summarizing) headline by means of:
– expansion through the addition of new information; and

– variation through the reformulation of content from previous segments, e.g.
syntactic transformation (active into passive voice), nominalization, attribu-
tive use of nouns, synonymy and other sense relations, etc.
The structural pattern across the three segments is not given over to a total refor-
mulation of all the elements that feature in the respective co-referential chains.
Some cohesive items are repeated verbatim or reformulated as different parts of
30. As noted later in this chapter, the cyclic pattern in which information is provided in the
opening sections of a news story can have up to five distinct levels.
210 News and Time
speech while preserving the lexical stem (for some examples see Figure 9.1 over-
leaf). It seems that the repetition of at least some lexical forms across the three
structural segments provides a basic skeleton structure that keeps the balance
between fully contextually known information (i.e. the repeated items) and the
expanded and reformulated elements that place a higher cognitive strain on the
reader’s processing because the elements bring new aspects of information, fur-
ther developing the message. The expression, re-expression and variation of infor-
mation in mutually interacting co-referential chains is a text-producing strategy
aimed at achieving the right balance between two contrasting poles: (1) repetition
arising from the double summary pattern in news texts and (2) the provision
of new information. All that is happening against the background of a text type
renowned for its non-linearity, since the inverted pyramid structure reverses the
natural progression from known to new information and starts with the most
important aspect of the story, which is presented repeatedly in a cyclic manner in
the news text.
9.4 Process chains
As far as the expression of temporal relations is concerned, there is one co-refer-

ential chain that plays a particularly important role in the news text: the process
chain that revolves around the verb phrase describing the central action of the
main event. Our focus on the process chain stems from the fact that the finite verb
phrase encodes temporal deixis through the verbal tense that it contains.
There are situations when one or more cohesive items in the process chain is
realized by non-verbal or non-finite means, e.g. when an event that is expressed as
a verb undergoes a grammatical transformation and comes to be re-expressed as
a noun, and vice versa. Such grammatical transformations help journalists avoid
excessive structural repetitiveness and manage the influx of new information in
different segments of the news text. In the absence of a finite verb form, temporal
deixis is not marked explicitly. The pragmatic effect of the nominalizations re-
sults in the ‘temporal persistence’ of nominalized events (cf. Downing 2000: 356,
after Chafe 1994: 68). Instead of chronological sequencing, the event becomes
cognitively reinterpreted and organized differently – in a more distant and im-
personal manner. In this connection, Downing, referencing Eggins (1994: 59),
observes that:
Discoursively, a nominalized form enables us to distance ourselves from experi-
ence and to organise our text, not in terms of chronological sequencing, but in
terms of ideas, reasons, causality. (Downing 2000: 356)
While the forms and functions of nominal, non-finite and other transformations
in process chains deserve detailed attention, this issue is briefly illustrated, with
several examples, when discussing the variations of the double tense shift pattern
in online news texts in Section 9.5.
For the purpose of the present account, however, we will concentrate on the
ideal case, when the basic process chain is realized through finite verb phrases
and opens with a verb already in the headline. Our goal is to track the transforma-
tion of the way temporality is expressed across several textual segments of news
stories.
Process chains containing finite verbs differ from other cohesive chains in
one important aspect. The finite verb form is unlike other parts of speech in that
it combines two elements: a lexical component and a temporal specification.
While the lexical component is carried by the base stem, the temporal informa-
tion is encoded through an inflectional suffix (e.g. works, worked). Where tenses
are formed analytically rather than through inflections, the duality is present in
the whole verb phrase seen as a single unit. The verb phrase is then made up of
auxiliaries that express the grammatical categories – as well as indicate the time
referred to – and the lexical component encoded through a non-finite verbal con-
struction (e.g. a present/past participle). For the purposes of cohesion analysis,
the analytical verb phrases, which express the grammatical and semantic compo-
nents separately as a finite auxiliary verb, and the (non-finite) lexical verb can be
considered as a single unit (a tie), although it also allows for a more subtle analysis
on the level of the two subchains: temporality (tense) and lexical meaning. The
verb phrase is a unit with a distinctive dual nature.
In narratives, tense has an important cohesive function (Brown and Yule
1983: 194; Halliday 1985: 313). Verbal tense establishes a temporal zone that func-
tions, in extensive stretches of texts, as a textual superstructure that provides a
unity to a sequence of actions. After a specific time is established, typically with
the help of an adverbial of time serving as a temporal anchor, several subsequent
occurrences of tensed forms will have the same temporal reference.31 Thus, the
same tense will be used to encode temporally sequenced actions, until the se-
quence is interrupted by a different tense form (or an adverbial of time) that es-
tablishes a different temporal zone.
In news texts, the role of verbal tense in articulating cohesion across exten-
sive stretches of news texts is weakened because of the non-chronological pres-
entation of the news story. What is more, however, verbal tense is non-cohesive
31. Nominalized and non-finite forms that are cohesively linked to such tensed forms in pro-
cess chains will normally be presumed to have the same temporal reference on the basis of
co-temporality.
212 News and Time
Headline Lead Body copy Source
says has spoken said Times

admits has acknowledged replied Independent
steal have been stolen were taken Times
steal have been stolen were stolen Guardian
suggests has suggested said Telegraph
launches has launched will be conducted Telegraph
apologise express remorse have apologised Guardian

moves closes in on deal has moved Guardian
sacks was sacked was dismissed Times

[is] convicted was convicted admitted Times
crashes -- have smashed Guardian
Figure 9.1 The transformation of the temporal and notional components in process
chains (sample news texts) (all examples are from 18 December 2012)
across the three most important structural segments: the headline, the lead, and
the body copy.
Figure 9.1 illustrates the non-cohesiveness of tense in the initial textual seg-
ments of several news texts. As noted, the headline present tense, when articu-
lating the authorial voice of the paper, is typically re-expressed into either the
present perfect or the simple past in the lead, even though the simple present also
occasionally recurs. The non-cohesiveness of tense is the direct consequence of
the non-deictic use of the present tense in headlines and the semi-deictic present
perfect in the lead (and the body copy), which contrasts with the deictic past tense
in the body copy (or the lead, as the case may be). It is also significant to see the
present perfect being used in the body copy, particularly where the lead contains
the present simple tense – the switch of tense to describe a past-time event is then
deferred by one structural segment.
The present tense is, of course, the result of the operation of headline con-
ventions and the pragmatic projection of deictic centres that have been described
previously. From the encoding perspective of the production of the news text, the
present perfect (or the past) tense in the lead is shifted into the non-deictic present
tense in headlines, while from the contrary perspective of the reception of news
texts, the present tense in the headline can be considered as being reevaluated into
the present perfect (or the simple past) in the lead. The latter perspective reflects
the actual linear processing of the news text from the headline to the body copy.
The figure also indicates that there is a number of variations on the basic
double tense shift pattern. Thus, one example above contains a reinterpretation
of the main event in terms of the future tense in the body copy (launches → has
launched → will be conducted), two examples contain only the shift to the present
perfect (apologise → express remorse → have apologised), and three examples a shift
to the simple past in the lead (sacks → was sacked → was dismissed). It could be
argued that it is relatively less significant whether a news text contains an instance
of the double tense shift pattern with the three different tenses used to describe
the same past-time event, even though this idealized pattern is the focus of a more
detailed interpretation below (see Section 9.5). Rather, it is the existence of a shift
of tenses (the present simple to the simple past; the present simple to the present
perfect; the present simple to the present perfect and then to the simple past) that
is most important here: the switch stands as evidence that the grammatical choice
of tense reflects the writer’s pragmatic negotiation of the discourse space involv-
ing the event, its textual representation, and the audience.
Non-cohesiveness of the present perfect tense
The verbal phrase containing the present tense in the headline and referring to
events in the past time can be re-expressed in the lead by means of the present
perfect or the simple past tense. This observation is made, for instance, by Carter
and McCarthy (2006: 625), who mention that the present simple in news head-
lines refers either to a subsequent present perfect time-frame, or to a definite past
time-frame (cf. Tidal wave hits Papua New Guinea → A tidal wave has hit Pap-
ua New Guinea vs. A tidal Wave hit Papua New Guinea late last night). However,
they do not propose an explanation for this phenomenon, even though each tense
has a particular effect that is distinctly different from the other.
The choice of tense appears to be motivated primarily by pragmatic reasons,
although the explanation of the different interpersonal effect of the two tenses
resides in their fundamental grammatical meaning. According to Quirk et al.
(1985: 183), the past tense combines two features of meaning:
a. the event/state must have taken place in the past, with a gap between its com-
pletion and the present moment; and
b. the speaker or writer must have in mind a definite time
By contrast, the present perfect relates “a past event/state to a present time orien-
tation” (Quirk et al. 1985: 192). It has the following connotations:
a. the relevant time zone leads up to the present;

b. the event is recent; and
c. the result of the action still obtains at the present time (Quirk et al. 1985: 193)
214 News and Time
Similarly, Huddleston (1969: 783–784) notes that the present perfect tense in-
volves two ‘tense selections’: one past and one non-past. In this way, the present
perfect tense forms a bridge between the past and the present time, which is a
highly desirable quality in newspaper reporting as it emphasizes the relevance of
the event and its proximity. The ‘current relevance’ of the present perfect tense is
the defining feature, contrasting it with the simple past tense.
The predictable grammatical meaning of the present perfect tense can be
seen as having pragmatic consequences. By using the simple past tense, the writer
makes the commitment to refer to the event as completed. The present perfect, by
contrast, entails the writer’s evaluation of the event as relating to, and being rele-
vant for, the reader. The encoding of the event through the simple past tense then
places an emphasis on the referential aspect of the situation, while the present
perfect tense represents a choice motivated by interpersonal regards.
This property also lends the present perfect its name in grammars as the ‘hot
news tense’. Declerck (2006: 782) identifies a particular usage type of the indefi-
nite perfect that refers to past time. In his view, the ‘hot news reading’ of the tense
is functionally justified on the following grounds: “the sentence in the present
perfect is used to ‘announce’ a bygone situation, i.e. to present the bygone actual-
ization of the situation in question as being very recent and as having high current
significance. For example: [Have you heard?] Kim Clijsters has won the US Open!”.
The occurrence of the present perfect tense in leads is a characteristic feature
of the genre of news texts. The ‘hot news’ perfect introduces the subject of the
narrative and the actual narration is then performed in the simple past tense (cf.
Walker 2008: 300). Its use as a narrative tense for events anchored in the past time
is not that exceptional; there is, for instance, a rising trend to use the tense in spo-
ken, as well as written, sports narratives (cf. Walker 2008: 296).
Clearly, in choosing a particular tense, the speaker/writer makes a selection
that is partly motivated by the time of the event that is being described, partly by
the genre requirements and partly by the writer’s own perspective when forming
a discourse space in which past events are described and related to the readers.
In their general discussion of the discourse properties of tense, McCarthy and
Carter make the following observation, which is particularly fitting in this context:
First, choice of tense and aspect can be seen to have a discourse dimension, in
that the choices are not determined purely by semantic factors relating to ‘objec-
tive’ time. Second, tense and aspect choices have become part of the conventions
of the genre. […] These genre-related choices reflect the fact that speakers and
writers convey interpersonal and textual meanings as well as the ideational ac-
count of actions and events. (McCarthy and Carter 1994: 96)
9.5 Double tense shift pattern
When approached from the point of view of the structural composition of the
news text, the present perfect tense in the lead is located in a central position
since the lead is a textual mediator that provides a bridge between the headline
and the body copy. The lead thus plays a privileged role since it not only extends
the headline but also frames the article proper. The canonical double summary
pattern means that in the idealized situation, the same information is presented
three times because the body copy is abstracted into the lead and the lead itself is
abstracted into the headline, giving rise to a very specific pattern of layered rep-
etition and parallelism of linguistic forms and structures in the three segments.
The following article indicates the typical pattern of a double tense shift that
can be identified in many news texts:
(H) Pakistan factory fires kill 261
(S) Building safety standards criticised after dual disaster strikes underwear and
shoe factories in Karachi and Lahore
(L) Factory fires in Pakistan’s two biggest cities have killed 261 people in a disaster
blamed on barred windows, shoddy building standards and the flouting of
basic safety regulations.
(BC) In the coastal megalopolis of Karachi 191 people were killed after becoming
trapped in an underwear factory that caught fire on Tuesday evening. Many
others were injured after trying to escape by jumping out of high windows of
the five-floor building, including a pregnant woman.
(The Guardian; 12 September 2012; emphasis added)
The main event, expressed through the present tense in the headline (kill), is
re-expressed by means of the present perfect tense in the lead (have killed). This
choice of tense underlines the interpretation of the news story as having a current
relevance. At the same time, the present perfect tense launches the subsequent
narration of the event in the past tense (were killed): it frames the entire narrative
consisting of past-time events.
Before dealing with the double tense shift pattern itself, let us first address the
issue of the framing use of the present perfect, which contributes towards explain-
ing the use of tense in the example above. This property of the tense is commented
on by Carter and McCarthy (2006) as follows:
The present perfect is often used initially to provide an overall frame for the re-
porting of past events, in spoken and written narratives and reports. The present
216 News and Time
perfect verb often provides a headline or statement of a newsworthy event, fol-

lowed by a series of verbs in the past tense reporting the details.
(Carter and McCarthy 2006: 618)
They give the following examples:
Poisonous black widow spiders have invaded Britain by plane. They stowed
away in crates of ammunition flown from America to RAF Welford, Berks. A
US airman at the base near Newbury captured one of the spiders in a jar after
it crawled out of a crate.
I’ve been going to the weight-watchers but, wait till you hear this, I went first
time and I’d lost three and a half pounds, and I went last week and I’d lost half a
pound, so I went down to the fish shop and got fish and chips. I was so disgusted!
(Carter and McCarthy 2006: 618)
The framing function of the present perfect differs from other possible uses of the
tense for past-time events in narration because, in the case at hand, it is used only
once, and not repeatedly, as a substitute for the simple past tense which would
otherwise be expected in more extended stretches of narrative texts (cf. Walker
2008). The first example provided by Carter and McCarthy is clearly taken from a
news text, but they fail to place the phenomenon into the full context because they
do not consider the headline as well. This means that the sequence ‘present per-
fect → simple past’, found in general narratives, is actually realized as a sequence
of three distinct tense forms in news texts, namely ‘simple present → present per-
fect → simple past’.
Once again, if the verb phrase is traced as a single co-referential process
chain, we can see that it is characterized by two shifts of tenses. Viewed from the
encoding perspective, which takes as its starting point the journalist’s choice of
tense to describe a past-time event, these shifts concern the utilization of (1) the
simple present tense in the headline and (2) the present perfect tense in the lead
for a past-time event. Viewed from the decoding point of view, the present tense
in the headline changes into (1) the present perfect to indicate the ‘current rele-
vance’ of the event and then shifts into (2) the simple past tense to locate the event
encoding
headline lead body copy

present tense present perfect tense past tense
decoding
Figure 9.2 Encoding and decoding of the double tense shift pattern
unequivocally into a particular point in time – cf. the transformation of the tense
in the example above (kill → have killed → were killed … on Tuesday evening).
In an earlier study on the internal patterns of short news stories, McCarthy
and Carter make a similar observation about the tense shift ‘present perfect →
past’:
The convention of this particular genre is that the preview [summary of the story;
broadly corresponding to the leading paragraph] is signalled by the use of pres-
ent perfect tense, while the details change to past tense. The preview therefore
stresses the ‘now-relevance’ of the events and, once again, is receiver-involving;
the details are then related as a regular, past-tense narrative of events.
(McCarthy and Carter 1994: 97)
Because of their focus on the framing function of the present perfect, McCarthy
and Carter do not contrast the tense with the headline present tense, although one
of their examples does include it. Where the co-referential process chain consists
of recurrent cohesive items (i.e., repetition without any lexical variation), we may
end up with a triad of differently tensed forms, such as the sequence kill → have
killed → were killed identified in the last example.
The triad represents a distinct pattern that needs to be recognized as one of
the fundamental grammatical conventions of modern English news texts. The
double tense shift pattern consists of the simple present tense referring to past-
time events in the headline. That form is then re-evaluated into the present per-
fect, negotiating the current relevance of the event as hot news for the reader.
Finally, the simple past tense is used in the body copy to provide a definite tem-
poral anchorage of the event in the past (often in conjunction with an adverbial
of definite time, such as yesterday) and, by doing so, begin the actual narrative ac-
count of the news story. The pattern, which embodies two shifts and three distinct
tenses, can be schematically represented as Figure 9.3. It is illustrated with the
News text segment Conventional tense Example
headline present tense kill
lead present perfect tense have killed
body copy past tense were killed
Figure 9.3 Temporal re-evaluation within a co-referential process chain (arrows indicate
the direction of the decoding process)
218 News and Time
different forms of the verb ‘kill’, which all refer to a single event that is unequivo-
cally located in the recent past (the last example contains a syntactic transforma-
tion into the passive voice).
Several observations need to be made in this place in order to make the dis-
cussion complete. First and foremost, the pattern is based on the temporal sit-
uation in a canonical hard news story that is formulated with a verbal headline
and follows the traditional format of the two summaries. Obviously, this idealized
pattern is relatively infrequent, given the diverse types of headlines as well as the
recent trends of opening the news story with conceptual elements other than the
main event (cf. the extracted topics approach to the formulation of headlines dis-
cussed in Section 10.5). Where some part of the relevant textual segment includes
an accessed voice, the heteroglossic verbalization will likewise affect the pattern
of tenses within the process chain. Last but not least, non-finite and nominalized
forms may render the relevant forms atemporal, once again disrupting the ideal
manifestation of the triple tense pattern.
Since the double tense shift pattern is carried by the verb phrase, one should
not concentrate on the issue of tense and aspect to the disregard of the lexical
component of the verb. The next section shows that when interpreting the entire
situation within the framework of the three linguistic metafunctions, the lexical
component turns out to have an important textual role because it provides much
needed cohesion within the co-referential process chain. While the temporal
component is metamorphosing in the verb phrase – for pragmatic reasons – from
one tense to another, the lexical component ensures some stability.
The data show that there are several possible ways in which the co-referential
process chain can be realized across the three basic structural segments of news
texts. The first format is characterized by a three-fold repetition of the lexical com-
ponent. The second pattern involves a repetition of the lexical component in two
of the structural segments and its replacement (reformulation) in the other struc-
tural segment. The third pattern is an extension of the second format: it involves
the dual reformulation of the lexical component in the lead and the body copy.
9.6 Cohesion and the three metafunctions
We have seen that the operation of the double tense shift pattern, which results
in the existence of three different tenses referring to the same past-time event,
impairs the usual function of tense to provide for the cohesion of a text. Due to
their variability, the tenses do not contribute towards constructing the texture of
the news text. The different tenses are locally coherent within the structural seg-
ment in which they are used: the non-deictic present simple tense in the headline,
the present perfect in the lead, and the deictic simple past tense in the body copy.
However, their correct interpretation depends on the reader’s knowledge of the
conventions of the genre.
Towards a functional model of temporality in process chains
Let us propose a functional model of temporality in process chains based on the

idealized situation in news texts, namely when the same verb is repeated in the
headline, the lead and the body copy. The verb undergoes changes as regards its
tense, but its notional component remains the same. In Figure 9.4, the compo-
nents of the verb phrases in the individual structural segments of news texts are
mapped onto the following three-fold system of language metafunctions pro-
posed by Halliday:
– the ideational function;

– the interpersonal function; and
– the textual function
Segment headline lead body copy

present tense present perfect tense past tense
(‘kill’) (‘have killed’) (‘were killed’)
Notional ideational-function- textual-function- textual-function-

component oriented oriented oriented
(lexical meaning)
(lexical introduction (cohesion realized (cohesion realized
of new information) through lexical through lexical
repetition) repetition)
Temporal interpersonal- interpersonal- ideational-function-
component function-oriented function-oriented oriented
(grammatical
meaning) (shift of tenses to (emphasis on ‘current (placement of the
enhance relevance) relevance’) event in the past
time)
ideational-function-
oriented
(indication of past
time placement
of event)
(non-deictic) (semi-deictic) (deictic)
Figure 9.4 Functions of lexical and temporal components in process chains: the case of
lexical repetition and the double tense shift
220 News and Time
The ideational function is, broadly speaking, understood as the expression of con-
tent (i.e. the provision of information of an essentially referential nature). The
interpersonal function is related to the consideration of the persons involved in
the text’s production/consumption and their potential presence or involvement
in the text (this can be manifested, for instance, by the construction of a shared
temporal context and the emphasis on the current relevance of an event). Finally,
the textual function concerns the construction of texture and is, among other as-
pects, realized through elements that help to ‘hold the text together’ and provide
structure corresponding to relevant genre requirements.
In the case of the canonical double tense shift pattern, the notional compo-
nent provides for the cohesion of the verb phrase because the cohesive function
of tense is blocked out by the conventional tense usage in the headlines until the
moment the normal narrative mode becomes instituted in the body copy. Thus,
the lexical component of the verb performs the ideational function in the head-
line, where it conveys new information. However, the repetition of the same form
in the lead and the body copy means that the lexical component of the verb is not
adding any new information in those structural segments. Hence, instead of the
ideational function, the lexical form fulfils the textual function because the repe-
tition is cohesive and supports the coherence of the text.
The temporal and modal exponents perform quite different functions. The
present tense in the headline behaves pragmatically. It satisfies the interpersonal
function on account of its non-deictic use: the tense is conventionally shifted in
order to enhance the relevance of the story. A similar situation occurs with the
present perfect tense in the lead. It likewise conveys the interpersonal function
because the use of the tense – for events located in the past time – emphasizes the
current relevance of the event, adding to it the ‘hot news’ reading. At the same
time, the present perfect already connects the event with the past, which means
that it behaves semi-deictically. Consequently, it also has to be afforded an idea-
tional orientation. The present perfect in the lead thus has a dual function. Finally,
the switch to the simple past tense in the body copy places the event indisputably
into the past and represents the most ‘objective’ of the tense choices within the
pattern. The deictic tense is then used ideationally because it conveys the notion
of pastness. It realistically represents the past event without negotiating the dis-
course space to involve the reader.
Variations of the idealized pattern
The matter of tense shift is further complicated by the frequent occurrence of the
subhead in some news texts, which adds an additional level of textual summary,
an issue that is subject to a detailed analysis in Section 9.8. Thus, in the following
news text, for instance, the headline and the subhead use the present tense, the
lead has the present perfect, and the body copy switches into the past tense:
(H) BBC and ITV apologise to Lord McAlpine for sex abuse allegations
(S) Lawyers for broadcasters express remorse and withdraw allegations after
already agreeing to pay damages
(L) The BBC and ITV have apologised to Lord McAlpine at the high court for
“disastrously” and falsely linking him to allegations of child sex abuse.
(BC) Lawyers for the two broadcasters expressed “genuine remorse” and with-
drew the allegations in a statement read by Sir Edward Garnier, counsel for
McAlpine, on Tuesday.
(The Guardian; 18 December 2012; emphasis added)
The repetition of the present tense in the subhead indicates that the double tense
shift pattern is not to be understood mechanically as a one-to-one correspond-
ence between a linguistic form and its single point of occurrence. It is a higher
order structure that organizes the representation of cyclically presented content:
while multiple or incomplete codings of the pattern are possible, what ultimately
matters is that non-deictic tense switches into deictic tense.
The example also illustrates another effect arising from the use of the present
perfect in the lead: since it places the time of the event into the ‘recency’ zone, it
leaves out the exact expression of time. Thus, no adverbial of definite time is ex-
pressed in the lead because its appearance would, in standard grammatical usage,
require the use of the deictic past tense.32 The precise temporal anchorage (on
Tuesday) then appears only in the body copy, where it duly correlates with the
past tense (expressed; withdrew). It is almost a paradox that the pragmatic use
of the present perfect in the lead runs counter to the journalistic requirement of
listing all relevant information in the lead.
As suggested in Section 9.4, a variation of the double tense shift pattern oc-
curs when only one tense shift occurs in between the segments (cf. Figure 9.5).
If we stay with the headline present tense referring to past-time events, then this
situation has two possible instances – a switch from the present simple to either
(1) the present perfect or (2) the simple past tense. In both cases, the switch of
tense can occur either between (1) the headline and the lead or (2) the lead and
the body copy. In either case, two of the structural segments will repeat the same
verbal tense, although the actual lexical verb may differ.
32. As is mentioned later in this section, however, such adverbials as today, this morning, etc.,
increasingly combine with the present perfect or even the past tense in online news.
222 News and Time
Interestingly enough, the data do not contain any example of the present per-
fect in the lead and the body copy. Thus, it appears that if the present perfect is
used in the lead to refer to a past-time event, it is further switched – almost au-
tomatically – into the simple past tense in the body copy, rendering the present
perfect a tense whose appearance is linked to the double tense shift pattern, being
a transitional form that is bound to be switched into the past tense in the next
segment. The now-relevance reading of the present perfect also entails that the
precise temporal location of the event may need to be specified later than in the
lead, i.e. only in the body copy. All this seems to support our interpretation of the
present perfect as a ‘bridging tense’, poised as a transitional element between the
non-deictic headline present tense and the deictic past tense. The present perfect
is not the ‘final stop’ as far as the textual rhetoric of news texts is concerned.
In the absence of the simple switch ‘present tense → present perfect’, it is the
latter possibility that we need to consider, namely the single tense shift pattern
from the present tense directly into the past tense. Thus, in the following example,
the present tense in the headline is already switched into the simple past in the
lead. The present perfect, as the optional element in the tense shift pattern, is not
realized here at all. This is, no doubt, the result of the inclusion of the adverbial of
time yesterday in the lead, cf.:
(H) Met sacks police officer who took on discrimination
(L) A detective who publicly challenged discrimination in the Metropolitan
Police was sacked yesterday for “discrediting the police service and under-
mining public confidence in it”.
(BC) Detective Constable Kevin Maxwell, 34, a black, gay officer in the Counter-
terrorism Command, was dismissed after a gross misconduct tribunal held
behind closed doors. (The Times; 18 December 2012; emphasis added)
An identical situation occurs in the next example, with the difference that an
additional textual level is placed between the headline and the lead. Here, the
subhead retains the headline present tense, thus continuing the distinct style of
headlinese (cf. also the incomplete sentence structure of the subhead):
(H) White House says Obama will move swiftly on gun control after
Newtown
(S) First signs that Democrats are willing to take on pro-gun lobby as even NRA-
endorsed senator Joe Manchin says ‘we need action’
(L) The White House promised a comprehensive series of measures, including
gun control legislation, on Monday to prevent a recurrence of mass shootings
such as the “horrific” attack in Connecticut that left 20 children and six teach-
ers dead. (The Guardian; 18 December 2012; emphasis added)
These two examples indicate that the non-deictic present tense found in headlines
is thus shifted directly into the deictic simple past tense in the lead, which then
establishes the narrative mode and terminates the conventional tense shift pattern
in the introductory segments of the news text. What ensues is a simple switch
‘present tense → past tense’ (sacks → was sacked; says → promised). The resulting
situation can be interpreted analogically to the double tense shift pattern: the lex-
ical component of the verb behaves ideationally in the headline and textually in
the lead, while the temporal and modal exponents fulfil the interpersonal func-
tion in the headline and the ideational function in the lead, where they provide
information about the real-time placement of the event, cf. Figure 9.5.
The single tense shift pattern, without the ‘hot news’ present perfect in the
intermediary position, is the rule whenever the lead specifies the definite time of
the event by means of an adverbial of time. The appearance of the present per-
fect tense in the lead is ruled out by the occurrence of an adverbial of definite
past time (most typically yesterday). The adverbial blocks the writer’s possibility
of constructing a shared discourse space in which the past-time event could be
construed as co-temporal with the reader’s reception of the text. In the absence
of the adverbial in the lead, the path is open to the realization of the double tense
shift pattern. The pattern is also preserved in online news, where it can, thanks to
the replication of some textual segments in the article preview, be found spread
across up to five textual segments of the online news text (see Section 10.6 below).
For the sake of completeness, it needs to be stated that some deictic adverbi-
als of time, such as today, this morning, etc., combine with both the simple past
tense (cf. the leads “Callous thieves were being hunted today…”; The Telegraph; 18
December 2012; and “David Cameron’s favourite fox hunt, and David Cameron’s
Segment headline lead

present tense past tense
(‘sacks’) (‘was sacked’)
Notional component ideational-function-oriented textual-function-oriented

(lexical meaning)
(lexical introduction of new (cohesion realized through
information) lexical repetition)
Temporal component interpersonal-function- ideational-function-oriented

(grammatical meaning) oriented
(shift of tenses to enhance (placement of the event in
relevance) the past time)
(non-deictic) (deictic)
Figure 9.5 Functions of lexical and temporal components in process chains: lexical
repetition and single tense shift
224 News and Time
favourite huntsman, were both fined for illegal hunting today…”; The Independ-
ent; 18 December 2012) and the present perfect tense in the lead because they
encode the event as co-temporal with the anticipated time of reception of the
news text. It is perhaps little surprising that we find this usage quite common in
online news, which keeps appearing on the newspaper website round the clock. A
future comparative analysis will, however, be needed to support this observation.
In connection with the variations of the double tense shift pattern discussed
above, the question will inevitably arise as to what happens in those news texts
that encode the past-time event by means of the past tense already in headlines.
Does this strategy, which is an important way of overcoming the conventions
of headlinese (cf. Chapter 7), affect the overall pattern by actually establishing
a past-time frame of reference straightaway, i.e. a temporal zone of non-recen-
cy that is cohesively referred to with the past tense in subsequent utterances? In
other words, does the past tense in headlines cancel out the possibility for a tense
shift because the event is described in a proper deictic way from the very begin-
ning? This does not seem to be the case. Let us consider the following example
that uses the past tense as a way of indicating that the most newsworthy element
is not the unfortunate death of the schoolboy but its bizarre circumstances (cf. the
analysis of the headline in Section 7.3):
Public schoolboy died inhaling laughing gas with friends
A public schoolboy has died after inhaling laughing gas with his friends, his
devasted sister has revealed.
Joseph Benett, 17, a promising art student, suffered a cardiac arrest after
taking the nitrose oxide on August 31.
The teenager, from Hampstead, North London, fell into a coma but died last
Thursday with his family at his bedside. (The Times; 5.10.2012)
Here, the tense shift pattern within the process chain is realized as a sequence of
the following forms found spread out across the three crucial textual segments:
‘past tense → present perfect → past tense’. Evidently, usage of the past tense in the
headline does not establish the need to employ the past tense reference in the lead
as well.
At the same time, the past tense in the headline acts as a substitute for the
conventional present tense, which is not used on account of the need to refocus
from the verb to the circumstances of the event. Thus, the opening unit with-
in the underlying double tense shift pattern does not, simply, have the typical
present-tense realization that we might expect on the basis of the existence of
headline conventions. In the lead, however, we see the pattern emerging in its
regular form – as the present perfect that eventually switches into the past tense in
the body copy. With reference to the sequence of co-refential forms ‘past tense →
present perfect → past tense’ (died → has died → suffered, fell, died), we have a
seeming shift from the past tense to the present perfect. This, however, is to be
read as the non-realization of the default present tense in the headline. The lead
has its own rhetoric, independent of the temporal framework established by the
headline. Because the past tense in the headline does not establish the subsequent
narration in the past tense, the lead remains free to implement its own preferred
tense, namely the present perfect. The deictic function of the past tense to provide
anchorage for the event in a specific past-time moment and to establish the sub-
sequent narration in the past tense is limited to the lead and the body copy and
does not apply to the headline.
As it transpires, the event of the ‘death’ is actually located in the zone of
non-recency (it occurred eight days before the publication of the article), yet it
becomes encoded in the lead by means of the present perfect, which stresses its
now-relevance. This is yet further evidence of the dominant position of the pres-
ent perfect in the lead: it is a form used pragmatically in order to enhance the
main event. At the same time, the action that was really ‘recent’ – in the sense of
falling within the scope of the publication frequency of the paper – was the sister’s
verbal announcement, which, likewise, receives the present perfect, although oth-
er formulations would have been possible.33
9.7 Patterns of cohesion and co-referentiality in online news texts
As discussed in Chapter 4, online news articles differ from print news by be-
ing formally articulated on two distinct levels. Each article is introduced on the
newspaper’s home page by means of a headline and, in most cases, the lead (or
the subhead). These two textual segments provide a basic preview of the article.
At the same time, the headline functions as a hypertextual link that connects the
news text on the home page to another level – the full article that is to be found
33. The alternative formulations, however, have slightly different implications. Thus, while (a)
below assigns more importance to the revelation than the death, (b) and (c) could be perceived
as actually lacking the temporal specification of the ‘time of revelation’ by means of an adverbial
of time (e.g. *yesterday) that might be expected to be present in the lead:
(a) *A public schoolboy died after inhaling laughing gas with his friends, his devasted
sister has revealed.
(b) *A public schoolboy died after inhaling laughing gas with his friends, his devasted
sister revealed.
(c) *A public schoolboy has died after inhaling laughing gas with his friends, his devasted
sister revealed.
226 News and Time
on the article web page. The full-text article has all the standard features of news
texts: headline, subhead (if any), lead, attribution, the body copy, visual elements,
additional internal and external links, etc. The existence of two independent,
though hypertextually linked, levels that contain two parallel pairs of textual seg-
ments (headlines and leads) inevitably impacts the nature of cohesion across the
segments.
Thus, the body copy of each online news article becomes abstracted in two
headlines and two leads (as well as, occasionally, a subhead on the article web
page). The headlines and leads can be identical, although it is common for the re-
spective formulations to be slightly modified so that verbatim repetition is avoid-
ed.34 When reading an online news text, a reader then has to go through up to five
textual segments (home page headline; home page lead; article headline; article
subhead; article lead). The segments serve as textual macrostructures that can, in
the extreme case, all abstract the story. A reader may then have to go through up
to five summaries before reading the body copy (or even six, in the case of another
summarizing subhead).35
Since the full text of the article is, for most readers, accessed through a hy-
perlink on the home page, the article headline is processed only after the reader
becomes familiar with the preview furnished by the headline and the lead on the
home page. The headline and the lead on the article web page can already operate
with some information that is established on the basis of the headline and the lead
on the home page.
Non-permanence of home page article previews
The article preview differs from the article itself in terms of its lack of permanence.
The textual elements on the newspaper home page are transient, as opposed to
those segments that form a part of the full-text article. The headline and the lead
provide a point of entry into the article only on the day of publication and only
34. Various online media may differ from each other: while some prefer reformulation, others
opt for repetition.
35. Many online news articles also include a photograph under the subhead. In some cases, the
caption to the photo provides yet another summary of the story. The layering of summaries is
reminiscent of the earlier practice of multiple decks of headlines that were common at the be-
ginning of the 20th century (cf. Bell 2002: 52). However, while such multiple decks would dis-
close more and more information about the event in an essentially narrative mode, the multiple
levels of summaries in online news recycle and reformulate the same information in a cyclic
manner. Although they do gradually develop the story by introducing some new information,
the gist of the story remains the same.
as long as the hyperlinks appear on the home page of the newspaper. After the
relevance of the story recedes and the news item is pushed out of the newspaper’s
home page by more current news items or the next day’s edition of the paper, the
article preview disappears from the home page and its headline and lead are no
longer available on the page. The full text of the article typically remains accessible
even thereafter because it is relegated into the newspaper’s archives, and it can
be hyperlinked through subsequent articles on similar topics. The textual form
of the headline and the lead on the home page therefore rarely survive the pub-
lication day, while the corresponding structures on the article web page become
archived and enjoy permanence. The transience of the home page reflects the fact
that the article preview functions merely as a signpost to what is deemed to be
the actual news content provided on individual article pages. In this connection,
let us also mention the comparatively bigger fragmentation of content in online
newspapers. The two-level presentation of news means that individual full-text
articles are presented independently of each other. Readers thus cannot appre-
ciate what other news items appeared alongside a particular article – each news
text becomes decontextualized from the others. The online environment, despite
its many advantages, does not enable a retrospective ‘glimpse on the whole page’.
The home page with the article previews is not archived and is available only on
the day of publication.
The structural template for online news
The article preview on the newspaper’s home page not only adds another level
to the online news text. It is also a framing structure that provides the point of
entry to the article itself – it functions as a trigger that is linked by means of a
hyperlink to the anchor, i.e. the actual news text in its complete form. While the
preview provides the only way of accessing the news item from the home page,
it is not involved in the processing of the news text when accessed through some
other means, e.g. in the paper’s archive or through any of the other possible points
of entry into the article. Figure 9.6 schematically represents the two main levels
and the five (or six) initial textual segments. The online news text is presented as
a unit consisting of the two main parts, although the preview is non-permanent
and detachable from the article itself. In my account, I approach the preview as a
component part of the news text – despite its being detachable and ephemeral –
because it is a distinct textual unit produced by the editorial office of the newspa-
per on the basis of the article itself.
When analyzing news texts within this two level approach to online news
articles, identical headlines sometimes appear on the home page and the article
228 News and Time
Online
news text
H1
Level 1
Home page home page
L1/S1
article preview
Hypertextual link
(unidirectional)
Full text article

H2
Traditional (S2) Level 2

abstracting elements article page
L2
BC
Figure 9.6 Multiple levels of textual segments in online news (H – headline, L – lead,
S – subhead, BC – body copy)
web page. The verbatim repetition reflects the editorial practice involved in the
preparation of the newspaper’s online content: the article’s headline becomes quite
literally ‘lifted’ from the article web page and included on the paper’s home page as
a hyperlinked trigger that will lead the reader to the full content of the news report.
However, there is an alternative approach to the compilation of the news content
on the papers’ home pages: the editors use the article, with its headline and lead (as
well as subhead, if any), as an inspiration for drafting an entirely new headline (and
lead) for the home page. It is in this latter case that we obtain the textual format
in which the narrative structure of abstract can appear multiple times: we may
find it in the home page headline, home page lead, article web page headline, and
article web page lead.
The reformulation of the headline and the lead for the home page has devel-
oped, arguably, from the perceived need to differentiate the text from the article
itself and to prevent verbatim repetition of information. After readers process the
headline and the lead on the home page, it becomes redundant to repeat exactly
the same propositions in the corresponding textual segments of the article itself.
While repetition is considered “a mortal sin in news writing” (Bell 1991: 184), re-
formulation is a most welcome strategy: the two pairs of headlines and leads can
retain their summarizing functions, serving as the abstracts to the news story,
yet they can present different aspects of the story and thus sustain the reader’s
interest. Reading the same set of headlines and leads would, after all, bring no
new information to the reader despite the cognitive effort needed to process the
identical textual segments. In terms of relevance, verbatim repetition would ren-
der the information as superfluous and irrelevant if repeated for a second time.
Some newspapers, such as The Independent, routinely add an aspect of au-
thorial evaluation to the home page headlines. While some of their headlines are
identical, others differ radically from the corresponding headlines on the article’s
web pages. Let us consider two examples, both retrieved on 30 January 2013, to
illustrate this phenomenon (H1 – headline on the home page, H2 – headline on
the article web page, L1 – lead on the home page, L2/S2 – lead/subhead on the
article web page):
(a) (H1) Cloud control: Thought personal information was safe with
server in sky?
(L1) Not from the US authorities it isn’t, under new legislation
(H2)
British internet users’ personal information on major
‘cloud’ storage services can be spied upon routinely by US
authorities
(L2) All personal information stored by British internet users on major
“cloud” computing services including Google Drive can be spied upon
routinely without their knowledge by US authorities under newly-
approved legislation, it can be disclosed.
The example increases the reader’s involvement in the text through synthetic per-
sonalization (cf. Fairclough 1989; Talbot 2007; Chovanec 2009) – the readers are
addressed through the question form of the headline and the gist of the news
story is being linked to their own experience. Note also the elliptical form of the
question, which adds a familiar tone to the headline. Since the article preview is
written on the basis of the article itself, we can see that H1 and L2 were drafted in
a rather creative manner: the editor took the liberty of reformulating the relatively
factual and straightforward proposition found in H2 into the question (H1) and
answer (L1) format of the preview. While the question-answer sequence is clearly
related to the news story, H1 does not abstract the story in the traditional way, i.e.
in the detached, seemingly objective manner we find in H2.
The second example makes a similar direct and quasi-personal appeal to
readers, with H1 increasing the involvement of the reader. This time, however, L1
provides a factual summary of the news story, although the quasi-personal appeal
is continued in the subhead on the article web page (S2), from where the verbali-
zation was textually lifted and entextualized in H1:
230 News and Time
(b) (H1)
And you thought it was confusing in English: James Joyce
translated into Chinese
(L1) ‘Finnegans Wake’ sells out its first print run in China
(H2)
First print run of Chinese translation of Joyce classic
Finnegans Wake sells out
(S2) And you thought it was confusing in English… (sic)
Here, a formulation is literally lifted from S2 (And you thought it was confusing
in English) and reused verbatim in H1 in the article preview, to which an aspect
of the main event (James Joyce translated into Chinese) is added. Interestingly,
however, the proposition expressing the main event is actually found in L1, not in
H1. In the article itself, we find a regular headline H2 expressing the main event:
in terms of semantic correspondence, we see, thus, that it is L1 and H2 that are
related, rather than H1 and H2, as might be expected.
What the examples above illustrate is a novel approach to the formulation of
some article previews in online newspapers: while the headline and the lead of
the article proper retain their summarizing function and are presented as factual
accounts of the main event, the headline of the article preview on the newspaper
home page offers a subjective commentary on the news report from the point
of view of the paper, which uses that textual segment to relate to the readers in
a more direct manner. It provides the possibility for the paper to further engage
with the readers, relating the gist of the news story to their own experience. In this
sense, the news (at least in the preview) becomes ‘personalized’.
9.8 The double tense shift pattern and its variations
After establishing the double tense shift pattern in Section 9.6, and the two-level
organization of online news texts in Section 9.7, let us illustrate how the tense
shift pattern operates at the interface between the home page and the article web
page. Once again, we are interested in how cohesion and coherence are artic-
ulated in the verbal process chain and how the temporal situation is expressed
through the temporal and modal exponents of verbs.
In the account that follows, four examples are documented. The first is a rel-
atively straightforward illustration of the ‘present tense → present perfect tense →
past tense’ pattern found in the three segments of the news article and of its par-
tial replication in the home page article preview. The second example documents
one of the possible variations of the pattern when the finiteness of some verb form
disappears, being replaced with a nominalization. Though a particular form may
be rendered as nominal, the underlying tense shift pattern is – despite its incom-
plete form – still observable thanks to the cyclic presentation of information in
the various textual segments. The third example documents the role of non-finite
headlines containing a past participle, where the non-finiteness arises from the
ellipsis of the operator. It also identifies the focusing function of the present per-
fect tense: it can be used in the lead to represent the main event while some other
events, which may have occurred at the same time as the main event, are rendered
instead in the past tense. Finally, the fourth example shows how the double tense
shift pattern comes into interplay with the syntactic ordering of elements in the
individual structural segments. Although we may see major shifts in transitivity
and the appearance of quite diverse verbs across the main structural segments,
the double tense shift pattern provides a clear beacon in the text, helping the read-
er navigate coherently through the text.
Although the double tense shift pattern was identified and explained in Sec-
tion 9.6 as being located within the headline, the lead, and the body copy, the
analysis provided in this section indicates that we may have to extend our inves-
tigation to cover some other elements that have an abstracting and summarizing
function and that may be placed in various parts of the news text: this particularly
concerns the subhead and the image caption, which both tend to occur before
the lead.
Example 1. The triple tense pattern
As documented above, since online news has an additional textual level in the ar-
ticle preview, there are considerably more textual segments that have a potentially
summarizing function and that may provide information in a cyclic manner (H1,
L1, H2, L2, possibly also subheads). The article preview, consisting of two seg-
ments, typically mirrors the double tense shift pattern, though it obviously does
so in an incomplete manner.
The following example illustrates an occurrence of the ideal pattern of ‘pres-
ent tense → present perfect tense → past tense’, with one of the steps being realized
in a subhead. In this and the other examples, the whole news text is presented in
two columns. The corresponding textual segments from the newspaper web page
and the article web page are aligned horizontally in order to enable an easy textual
comparison, although in actual reality they will be processed vertically column
by column, i.e. starting with the headline and the lead on the home page (left
column) and being followed with the textual segments on the article web page
(right column).
232 News and Time
Newspaper home page Article web page

Headline Cumbria rejects nuclear Cumbria rejects underground
waste dump nuclear storage dump
(H1) (H2)
Subhead -- The only local authorities in the UK
still involved in feasibility studies have
voted against the disposal facility
(S2)
Lead The only local authorities in the Government plans to undertake
UK still involved in feasibility preliminary work on an underground
studies have voted against the storage dump for nuclear waste were
disposal facility. rejected by Cumbria county council
(L1) on Wednesday.
(L2)
Body copy -- The county and its western district
councils Allerdale and Copeland
which make up the “nuclear coast”
opposite the Isle of Man were the
only local authorities in the UK still
involved in feasibility studies for the
£12bn disposal facility.
(BC2)
Figure 9.7 Multiple summaries on the newspaper home page and the article web page
Two issues are noteworthy at this point. The first concerns the headlines,
which are very similar. They share the same syntactic structure and repeat the
verb and the tense (‘rejects’). There is only a minimal degree of reformulation
(nuclear waste dump in H1 is expanded into underground nuclear storage dump in
H2). The slight reformulation helps to avoid verbatim repetition of the headline
and assures that some new information in provided in H2, however minimal it is.
As regards the mutual relation between H1 and H2, H1 is most likely the result of
the shortening of H2, in which case we see a pattern of summarization leading to
shorter and shorter textual segments.
The second noteworthy feature is the re-use of the home page lead (L1) in
exactly the same form as the subhead on the article web page (S2). The article lead
L2 is evidently formulated as another summary of the main event (Government
plans to undertake preliminary work on an underground storage dump for nuclear
waste were rejected by Cumbria county council on Wednesday). On the level of
the article itself, we thus find three summaries abstracting the story, namely the
Structural Cohesive Tense

Article preview segment item
Home page
Headline 1 rejects present

↓ ↓
Lead 1 Have voted against present perfect
↓ ↓ ↓
Headline 2 rejects present

↓ ↓
Full-text article
Subhead Have voted against present perfect

Article page
↓ ↓
Lead 2 were rejected past
(on Wednesday)
Body copy – (n/a)
Figure 9.8 The linear sequence of cohesive items within the co-referential chain.
Circles identify tense shifts in the article preview and the article itself
headline H2, the subhead S2 and the lead L2. However, if we also include the two
summaries in the article preview, then the reader may have to go through five
summaries of the main event (H1, L1, H2, S2, L2) before learning of more details
about the event in the body copy.
The pattern of cohesion in the article above is illustrated in Figure 9.8. The
elements are arranged in a vertical manner reflecting the sequence in which they
feature within the co-referential chain and the linear order in which they are pro-
cessed by the readers.
The figure illustrates how the shift of tense in the article preview (‘present →
present perfect’) mirrors the tense shift in the article itself. The sequence ‘present
tense → present perfect tense’, extracted from the headline and the lead on the
home page, is reduplicated on the article home page in the headline and the sub-
head, while the past tense appears in the lead of the article.
The article lead L2 uses the deictic past tense because it anchors the event
into a definite past-time moment through the combination of the tense and an
adverbial of definite time (were rejected … on Wednesday). The body copy does
not encode the main event through a verb phrase (avoiding what could otherwise
be its fourth reformulation). However, the absence of a corresponding verb from
234 News and Time
the body copy does not affect the triple tense pattern because the typical tense
realizations are ‘upgraded’ by one level, through the addition of the subhead. It
is there that we find the semi-deictic present perfect tense that pragmatically en-
codes the current relevance of the event, simultaneously locating the event in the
near (recent) past and articulating a link with the readers.
The example indicates that the tense shift pattern is to be found twice in on-
line news texts because of the existence of the article preview on the front page.
The replication of the headline and the lead on the newspaper home page allows
for a partial replication of the triple tense pattern found in the structural segments
of the full-text article.
The appearance of the subhead in some articles on the article web pages is
an interesting phenomenon because that segment allows for all three tenses. The
example above indicates that the ‘hot news’ present perfect is acceptable, as is the
simple past tense and the simple present tense. While the simple past tense may
be used to provide background to the main event, the simple present tense lends
the subhead more of a headline-quality. In this sense, the subhead is a transitory
element, standing midway between the pragmatically-oriented headline and the
semantically-oriented body copy, as far as the function of tense is concerned.
Example 2. Complex chain involving nominal transformation
This example sets out to illustrate an incomplete tense shift pattern. Here, the pro-
cess chain involving the verb phrase is shown to include a textual absence from
H1 and a nominal form in BC2, yet the underlying tense pattern is still noticeable.
Apart from documenting the tense shift, the discussion revolves around the lexi-
cal and syntactic transformations affecting the verb phrase.
Let us consider the co-referential process chain that runs across the whole
news text and that is realized in the individual textual segments by means of a
series of specific cohesive items. The items that participate in the chain are: sets
(H1) → makes announcement … setting (L1) → Ø [colon structure] (H2) → has
announced (L2) → announcement (BC). Our focus here is on the chain containing
all references to the action of ‘setting (the election date)’, which constitutes the
core action of the event reported. The event of ‘setting’ is unequivocally located in
the past – it is a one-off occurrence that is partly newsworthy because of its future
implications (‘election date’).
In the headline on the newspaper home page (H1), the process chain opens
with the finite verb sets. The verb is formulated in the conventional present tense;
its monosyllabism is likewise characteristic of the typical style of headlinese.

Headline Gillard sets Australian Julia Gillard: Australian
election date election will be held on 14
(H1) September
(H2)
Subhead -- Prime minister says long campaign
will allow reasoned debate on policies
of her Labor party and conservative
opposition
(S2)
Lead Labor PM makes surprise Julia Gillard, the Australian prime
announcement of 14 September minister, has announced a general
poll, setting up country’s longest election will be held on 14 September,
ever election campaign. surprising the electorate and ending
(L1) speculation about when she would
take her minority government back to
the polls.
(L2)
Body copy -- The announcement by Gillard,
Australia’s first female prime minister,
kicks off what some commentators
have described as the longest election
campaign in history. […]
(BC2)
Figure 9.9 The parallel segmentation of online news texts on the newspaper home page
and the article web page (The Guardian; 30 January 2013)
The next cohesive element within the process chain appears in the lead L1. As
part of the article preview on the newspaper home page, the lead is an additional
textual segment that provides a further summary of the story. It both abstracts
the news text and establishes the perspective from which the story is covered (cf.
the evaluative attribute surprise in the phrase makes a surprise announcement).36
The cohesive element that participates in the process chain is the verbo-nominal
phrase makes announcement, which is subsequently re-expressed in the present
participle setting up. Both forms are in fact participating in the chain. The non-fi-
nite form setting up belongs to the process chain on account of the repetition in
36. This inclusion of the evaluative item needs to be read as a strategy of achieving a maximal
relevance of the lead for the reader. As Cotter (2010: 162) notes, the maximal relevance occurs
as a result of the combination of the primary news element (the most important element of the
story) and the ‘hook’ or angle that the journalist chooses in approaching the story.
236 News and Time
L1 of the lexical form of the verb from H1. The verb phrase makes announcement,
by contrast, participates in the chain for semantic reasons: in the context of the
article, ‘making an annoucement’ is locally equivalent to ‘setting up (an election
date)’. The verb sets in H1 can thus be seen as being transformed through refor-
mulation into makes announcement in H2, with the two belonging to the same
process chain on account of semantic (the same meaning) and formal (the same
word class) reasons. The participle setting up in H2 is likewise co-referential be-
cause it semantically refers to the same process, though in a non-finite manner.
The finite verb forms in H1 and L1 are tensed in the same way (sets → makes
announcement). Both verbs use the present tense non-deictically: the announce-
ment, which gave rise to the ‘setting up of the election date’, was made in the past
but the present tense is used in order to enhance the impression of immediacy.
We can thus see that the verbal phrase significantly attends to the interpersonal
function in the headline and the lead. Interestingly, while the lead in the article
preview L1 retains the present tense found in the headline H1, the lead in the
article itself (L2) switches into the present perfect (has announced).
The situation becomes more complex when the reader follows the internal
hyperlink leading to the web page with the article itself. As noted previously, the
article headline in this case cannot be taken as a text-opening segment since it is
preceded by the home page headline and the lead. This means that the elements
within the headline on the article web page need to be considered with respect of
the preceding verbal context established by the earlier textual segments.37
In the article analyzed here, The Guardian newspaper chose to make the two
headlines different. While the home page headline H1 provides a true summary
of the event through the monoglossic voice of the paper, the article headline H2
uses a heteroglossic strategy of less explicit summarization: the colon structure
representing reported speech. The latter headline identifies the news actor first
and follows the attribution with the indirect reported speech sourced to that ex-
ternal accessed voice (Julia Gillard: Australian election will be held on
14 September). This format means that the headline does not contain any ex-
plicit mention of some lexical item that could be formally co-classified within the
process chain that starts in the home page headline and continues in the home
page lead. The action of ‘setting the election’, which stands at the centre of the pro-
cess chain, is actually constituted by the news actor’s entire utterance (election
will be held): it is performed rather than reported in the headline to the article.
The speech act of ‘announcement’, which forms a part of the news, is likewise
37. This does not apply when the article is accessed directly, e.g. through the newspaper’s on-
line archives or the hyperlinks provided in other articles/media.
expressed indirectly – through structural means (namely the combination of the

attribution and the colon structure).
Similarly, the subhead does not encode the main action either; instead, it
elaborates on the news actor’s explanation of the newsworthy aspect of the story
(Prime minister says long campaign will allow reasoned debate on policies of her
Labor party and conservative opposition). The elaboration is related to the eval-
uation of the election date as a ‘surprise’ in the lead on the home page. We can
see that this subhead, accompanying the headline on the article web page, does
not function as an additional summary. Instead, it is evidently related to the lead
found on the newspaper’s home page. The lead L1 provides the angle for the story
(makes surprise announcement) and it is the surprise effect that is taken up in the
subhead (long campaign will allow reasoned debate) rather than the event itself
which is expressed in the headlines. The absence in the subhead of any item relat-
ing to the process chain under investigation means that the structural segment of
the subhead does not participate in the cohesive chain at all, although it is – as we
have seen – discursively related to the lead on the home page.
Another verbal element within the process chain occurs in the lead of the
full-text article. This time, there is the reporting verb ‘announce’ that emphasizes
the current relevance of the event, as expressed with the present perfect tense in
L2 (has announced a general election will be held on 14 September). Here we see
that the verbo-nominal construction found in the first lead L1 (makes a surprise
announcement) is turned into a fully verbal construction (has announced). In this
part of the chain, the cohesive link is established through the formal identity of
the word stems which carry the lexical meaning (announcement in L1; announced
in L2), though the two expressions are formulated as different parts of speech.
Finally, in the first paragraph of the body copy, the element participating in
the co-referential process chain recurs again as the noun announcement (The an-
nouncement by Gillard, Australia’s first female prime minister, kicks off what some
commentators have described as the longest election campaign in history). This nom-
inalization belongs to the process chain because it is co-referential with the same
concept that is expressed verbally (or verbo-nominally) in the previous structural
segments – what matters in assigning items to the co-referential chain is their
semantic affinity (i.e., co-reference) rather than a strictly formal classification on
the basis of grammatical criteria. As a result of the nominal transformation, the
original process (‘setting the election date’) becomes conceptually redefined and
grammatically re-classified. That results in the loss of temporal indicators. The
chain ceases, at this point, to indicate time since the focus of the message shifts to
other aspects of the story (…kicks off what some commentators…).
238 News and Time
Headline SETS [ATTRIBUTION: ] ELECTION

WILL BE HELD
Subhead -- --
Lead makes announcement has announced

setting up
Body copy -- The announcement
Figure 9.10 The progression of the process chain from the home page to the article web
page (The Guardian; 30 January 2013)
Let us schematically represent the whole co-referential chain, as it winds its

way through the two textual levels of the news text from the home page to the
article web page. The connecting lines in Figure 9.8 between the individual co-
hesive items indicate the ties existing between the individual expressions. The
chart includes a circle for the indirect speech in the headline on the article page,
which does not, formally speaking, form a part of the process chain, even though
the phrase election will be held effectively realizes the action of ‘setting the date’
by virtue of a reported performative speech act. What participates – in a rather
elusive and vague manner – in the process chain is, rather, the representation of
the phrase as indirect speech, rather than the lexical content of the heteroglossic
utterance.
The figure captures the lexical as well as grammatical transformations that the
process chain undergoes during its progression through the text. What emerg-
es from this complex example is a situation that approximates to the idealized
temporal pattern for encoding past-time events through the double tense shift
(present tense → present perfect → past tense). There is the present tense in the
first headline H1 (sets), echoed in the present tense in the first lead L1 (makes
announcement) and reformulated into the present perfect in the lead of the article
proper L2 (has announced). This particular example does not go on to encode
the main event through the past tense because the focus of the story shifts and
the next cohesive item within the process chain is realized as a nominalization.
In fact, the nominal form serves as a way of avoiding the necessity to encode the
event through the more distancing past tense, which indicates a definite break
from the present time.
Example 3. Variations on the triple tense pattern
In the next example, the structural situation of the news text is further complicat-
ed by the presence of the subhead and an image caption on the article home page.
The subhead uses the past tense to describe the circumstances under which the
main event happened. The main event (‘killing’) is not mentioned explicitly in the
subhead, which means that the verbal process chain does not have any realization
in this particular structural segment. It is, however, realized in a caption under a
photograph of the victim (‘who was shot and killed’). Because of this, the image
caption is included in the figure among the other structural segments. Although
it intrudes into the basic discourse structure of the news text, it is likely to be pro-
cessed by the reader reading the article in the sequence represented in Figure 9.11.
In this case the co-referential process chain involves lexical variation – differ-
ent verbs are chosen to describe the main action in the various segments. More-
over, ellipsis of auxiliaries is involved in the headlines, making them non-finite.
The temporal situation across the relevant segments is as follows: ellipted present
tense in H1 (shot dead) → present tense in L1 (is killed) → ellipted present tense in
H2 (shot dead) → simple past tense in image caption (was shot and killed) → pres-
ent perfect in H2 (has been killed) → simple past tense in the body copy (was shot).
Since the subhead merely identifies the victim and does not encode the event, it
does not participate in the co-referential process chain.
Germane to my analysis of the news text as embodying the double tense shift
pattern is an interpretation of the non-finite verb phrase in the headline as an
instance of ellipted present tense, i.e. in harmony with the prevailing headline
conventions.38
Formally speaking, there is no tense shift in the article preview. Both H1 and
L1 use the present tense in the passive voice (with ellipted and full auxiliary, re-
spectively). In the article itself, the triple tense pattern is observable once again,
despite the presence of the non-finite headline and the interposed subhead and
38. However, the full, non-ellipted headline could be construed in the past tense as well (*US
inauguration girl was shot dead), but the meaning of the headline would, arguably, be
somewhat different. Rather than reporting the fact of the girl’s death, the past tense could imply
some prior knowledge of this fact (since the death would be contextualized as a result of the
‘non-recency’ status of the past tense; cf. Section 7.3). The usage of the past tense would actually
enable the headline to refocus on the manner of the death as the most newsworthy (and new)
element.
240 News and Time

Headline US inauguration girl Hadiya Pendleton: girl who
shot dead performed at inauguration shot dead
(H1) in Chicago
(H2)
Subhead -- Fifteen-year-old was talking with friends
when gunman opened fire on group
(S2)
Image caption -- Hadiya Pendleton, who was shot and killed
when a gunman opened fire on a group of
students.
(IC2)
Lead Fifteen-year-old Hadiya A 15-year-old girl who performed in
Pendleton, who performed Barack Obama’s inauguration festivities has
during inauguration been killed in a Chicago park as she talked
festivities in Washington, with friends by a gunman who apparently
is killed in Chicago park. was not even aiming at her.
(L1) (L2)
Body copy -- Chicago police said Hadiya Pendleton was
in a park about a mile from Obama’s home
on Tuesday afternoon when a man opened
fire on the group. Hadiya was shot in the
back as she tried to escape.
(BC2)
Figure 9.11 Non-finite ellipted headlines and the simple past in the subhead
image caption – that are both formulated in the simple past tense. Though the
pattern involves the conventional ellipsis of the auxiliary verb be from the head-
line, the gradual shift of tenses is clearly evident in the following abbreviated rep-
resentation, cf.:

[is] shot dead (H1, H2) / is killed (L1) → has been killed (L2) → was shot (BC)
Two more observations on the tense in the lead are in place here. The present per-
fect tense in the article lead is apparently so strong a conventional feature that it
is used in L2 even though the past-time frame for the article is already established
by the subhead and the image caption. Moreover, employing the present perfect
tense for the main event (has been killed) contrasts with the simple past tense
used twice in the same sentence. However, the past tense verbs in the subordinate
clauses, though describing events that are located in exactly the same time zone
(as she talked; was not even aiming at her), provide background information to
the most newsworthy event. This seems to indicate that the choice of the pres-
ent perfect tense with the verb ‘kill’ has a focusing function: the tense assigns
prominence to an action that constitutes the ‘main event’ of the news story. By
contrast, the circumstantial actions, despite being simultaneous, receive the past
tense. There is even no possibility of using the present perfect tense to refer to
those particular circumstantial actions in this lead: the resulting sentence would
not be grammatically correct.39
The second point concerns the different treatment of tense in L1 and L2. The
article preview is entirely in the present tense. While the present tense auxiliary
is ellipted in H1, it is used in L1 (is killed), where it seems to add emphasis to the
main process. In L2, by contrast, the tense is already changed into the present
perfect (has been killed), most likely in connection with the fact that L2 provides
all the supplementary information about the circumstances of the main event. L2
is, thus, much more ‘narrated’, while L1 has a distinct ‘headline feel’, thanks to the
preservation of the present tense.
Example 4. The triple tense pattern as a cohesive structure
Let us illustrate the triple tense pattern with one more news article, this time with
the subhead in the present tense. The aim is to demonstrate how the tense shift
pattern, represented here again in its ideal format, acts as a structural template,
and how other co-referential chains interact with the central process chain. (For
the sake of simplicity, with this news text only the article is analysed; the home
page segments and the image captions that identify the girl have been omitted.)
The text of the article consists of the elements as shown in Figure 9.12.
After the present tense appears in the headline (wins), the subhead uses the
present tense as well (rules), making the utterance ‘timeless’, particularly with
view to the future implications of the court’s decision. The present perfect appears
in the lead (has been granted), emphasizing the current relevance of the event, and
the past tense is used only in the body copy (ruled) that specifies the real-time an-
chor for the event (on Thursday). The tense pattern then includes a double use of
39. One of the leads analysed in Section 9.6, however, includes the following dual use of the
present perfect: A public schoolboy has died after inhaling laughing gas with his friends, his dev-
asted sister has revealed. The first occurrence of the tense (has died) corresponds to the main
event (the verb also occurs in the headline and clearly forms a part of the tense shift pattern
centring on the verbal process chain). The second form, carried by a reporting verb (has re-
vealed), is justified since it denotes an action located within the ‘recency’ zone.
242 News and Time
Segment Article
Headline Icelandic girl wins right to be called gentle breeze
Subhead Court rules Blær, which means ‘gentle breeze’, can be legally used as
15-year-old’s name despite opposition from authorities
Lead A 15-year-old Icelandic girl has been granted the right to legally use the
name given to her by her mother, despite opposition from the authorities.
Body copy A court ruled on Thursday that Blær (which means “gentle breeze”) can be
legally used as the girl’s first name.
Figure 9.12 An article with a triple tense pattern (The Guardian, 31 January 2013)
the present tense, followed by the interpersonally motivated present perfect, and
finally the deictic past tense:
wins (H) → rules (S) → has been granted (L) → ruled (BC)
In terms of cohesion analysis, it is worth noting how the lexical component of the
verb changes across the individual structural segments of the news text. The dif-
ferent verbalizations of what is taken to be cohesive items within the same co-ref-
erential chain are motivated by the different perspectives taken in the individual
segments. The perspective alternates from a focus on the girl in the headline (Ice-
landic girl wins) to the court in the subhead where it appears in the subject role
identifying the agent ultimately responsible for the outcome of the case (Court
rules). Then the focus in the lead switches back to the girl (A 15-year old Icelandic
girl has been granted) and then again to the court in the body copy (A court ruled
on Thursday).
The alternation between the passive and the active voice does not affect the
cohesive nature of the elements either. In the headline, the news actor appears in
the syntactic position of the subject and performs the semantic role of the agent of
the active verb ‘win’. The news actor enjoys the same syntactic position in the lead,
although it performs the semantic role of the patient/beneficiary of the action,
expressed by means of the passive voice of the verb ‘grant’.
Figure 9.13 illustrates how the verb phrase can stand at the centre of several
mutually interacting cohesive chains. The horizontal lines represent the syntac-
tic ordering (the actual linear sequence) of the elements from the three cohesive
chains in the individual structural segments of the news text. The vertical ar-
rangement indicates the progression of the cohesive chains, with the links be-
tween the constituents of the chains represented with arrows.
Headline girl wins right [n.a.]
Subhead court rules Blær

15-year-old
Lead girl has been granted [n.a.]

the right [implied: court]
Body copy court ruled Blær
Figure 9.13 Interaction between a process chain and two other co-referential chains
(The Guardian, 31 January 2013)
The co-referential process chain enjoys a privileged position across the main
structural segments of the news text. The process chain interacts, among other
aspects, with two lexical chains denoting the ‘girl’ and the ‘court’ as the two entities
that are connected through the central predication that assigns them the seman-
tic roles of the agent and the beneficiary (cf. ‘The court [agent] gave [action]
the girl [beneficiary] a right’). In the headline, only the ‘girl’ chain is realized
in connection with the process chain (Icelandic girl wins (right)…). The ‘court’
chain is absent from the headline because that social actor is not represented at
all; the focus is entirely on the outcome of the action benefitting the girl. The sit-
uation changes, however, in the subhead, which connects the process chain with
the ‘court’ chain (Court rules…). The ‘girl’ chain, though realized in the subhead,
is not explicitly linked with the process chain because it is subsumed within an
extensive phrase that elaborates on the nature of the right given to the girl (…
Blær, which means ‘gentle breeze’, can be legally used as 15-year-old’s name…). That
whole phrase (including the embedded realization of the ‘girl’ chain, cf. 15-year-
old) is conceptually co-referential with the expression right in the headline (which
of course forms a cohesive chain of its own). The lead is the segment where the
two crucial lexical chains that revolve around the process chain interact more
visibly, though in an implicit way – they do so through the passive construction
with the deleted semantic agent (A 15-year-old Icelandic girl has been granted the
right…). The passive, however, leaves a syntactic gap that the readers can fill on
the basis of information that is already known from the previous context (… has
244 News and Time
been granted the right [by a court] …). The repeated switch of focus from the girl
in the headline to the court in the subhead and to the girl again in the lead is final-
ly completed in the first paragraph of the body copy, which starts with the court
again and elaborates on its decision, embedding the reference to the ‘girl’ chain
within another extensive phrase conceptually co-referential with the ‘girl’s right’
chain (…can be used legally as the girl’s first name).
We can see that thanks to its central position, the process chain acts as a pivot
around which the other co-referential chains revolve and interact with one anoth-
er. While the linear order of the items that make up the chains in the individual
structural segments changes, and the items themselves are subject to various syn-
tactic and lexical transformations (including their total textual absence, recov-
erable deletion arising from passivization, and diverse forms of representation
of social actors), the central proposition expressed across the various structural
segments of the news text remains essentially the same.
9.9 Concluding remarks on tense and textuality
As shown in this chapter, tense choice in news texts has a demonstrably textual
dimension to it. The usage of tense in headlines is only the proverbial ‘tip of the
iceberg’, with other interesting phenomena lying submerged in the subsequent
textual segments. When, as far as tense choice is concerned, an investigation
into the textual rhetoric of headlines is complemented with the textual rhetoric
of news texts, news texts are found to follow a distinct macrostructural pattern
of systematic variation of tense. Although the analysis of co-referential process
chains (those connecting the repeated references to the action/state constitut-
ing the main event) reveals some variations and incomplete realizations of the
pattern, it demonstrably involves the uni-directional shift between the following
tenses: the present tense, the present perfect, and the past tense.
I have argued that the diverse tense forms found across the different struc-
tural segments serve semantic as well as pragmatic functions and do not affect
the coherence of the news text. The liberation of verbal tense to serve other than
semantic (deictic, ideational) functions is directly related to one of the fundamen-
tal constructional principles of news text, namely the inverted pyramid, that pro-
vides for the cyclic presentation of information across several textual segments
that make up news texts. The repetition and reformulation of the information
combines with the pragmatically-oriented systematic variation of verbal tense in
the introductory segments that are, in their ideal realization required by the in-
verted pyramid structure, formulated as summaries of the news story.
In online news, the pattern of tense shift is observable in additional textual

segments found in the home page article preview. Here, the tense shift pattern
essentially repeats the usage in the article itself, though only two textual segments
are available for its realization (the headline and the lead/subhead). The situation
in online news is seen as a replication of the basic double tense shift pattern rather
than its extension. This is because the textual segments that constitute the article
preview are not unique and inseparable components of the online news text: they,
in fact, provide an extra level that adds another headline and lead/subhead.
Chapter 10
Temporal structure of news reports
The discussion of temporal aspects of news texts would not be complete without
considering the internal temporal structure of news reports. After documenting,
in the previous chapters, the use of tenses in headlines and offering a pragmatic
explanation for tense variation within the main textual segments of news texts, I
will complement the debate in the present chapter with by focussing on the con-
cept of non-chronology.
While non-chronology is closely related to the coding of temporality, it is an
issue that plays a less central role in the pragmatics of tense and time and, thus,
in my interpretation of the temporal situation in the opening segments of news
texts. Non-chronology is linked to the top-bottom organization of news texts
and is manifested in the characteristic narrative structure of written news, with
the narrative move of providing an ‘abstract’ having a particularly strong role in
headlines and leads – a role that arises from the traditional summarizing function
of these textual segments. Alternative accounts – namely, the semantic-cognitive
approach to thematic structure and the conceptual structure of event frames – are
also discussed in order to shed additional light on the initial abstracting segments
(the headline and the lead).
The chapter argues that non-chronology affects the textual (surface) pres-
entation of the event and does not affect the actual underlying order of the com-
ponents of the entire news event: readers are actually able to correctly infer those
from the text. An analysis of the temporal structure of a sample news story traces
how the individual subevents are ordered with respect to each other and how
they are presented in the news text. It is shown that the non-linearity is, in fact,
partly cyclical, with the different subevents making repeated incursions into the
past and the future. In this way, they revolve around the main event that becomes
re-expressed several times in the article. The chapter concludes by explaining
this cyclicity in connection with the orbital organization of news texts and inter-
preting news time as having an interpersonal dimension. In other words, since
non-chronology is a non-iconic representation of the event, it is interpreted as
arising from other than purely ideational concerns.
248 News and Time
10.1 Non-chronology and the narrative structure of news stories
The internal temporal organization of news stories is marked by its substantial

complexity. News time has been studied, most prominently, in connection with
the narrative structure of news stories and the conceptual frame of the main
news event. The top-down organization of news texts affects the chronology of
the event reported because it brings about the reorganization of the individual
narrative elements of the story. Let us outline the narratological and cognitive
perspectives on news stories in order to understand the systematic nature of ex-
pressing temporal deixis in news texts from a different perspective than the one
applied in the previous chapters. Despite the different theoretical underpinnings,
the various approaches discussed here agree that the conceptual category of time
is organized in news stories in a non-chronological manner. Such sequencing of
the elements of the story is associated with the particular textual organization of
news texts in terms of the inverted pyramid structure. The non-chronological
textual structure is ultimately related to the operation of news values, which can,
by giving prominence to certain aspects of the story, be used for focusing the
headlines and opening the news articles.
Some very relevant findings into the organization of news time have resulted
from analyses considering news stories as narratives. This issue has been system-
atically addressed in several studies by Bell (1991, reprinted as Bell 1994 and fur-
ther extended in Bell 1995 and 1998) and critically developed by Schokkenbroek
(1999). Bell bases his approach on the observation that “[j]ournalists do not write
articles. They write stories” (1991: 147). By consistently referring to the news ar-
ticle as a ‘story’, he treats the hard news text within the conceptual framework
common in narratology, emphasising that the news story is more than merely a
text filled with facts. Instead, it has some of the crucial attributes found in stories:
a structure, direction and viewpoint.
News stories can be contrasted with face-to-face personal narratives within
the framework proposed by Labov and Waletzky (1967) and Labov (1972), along
the following cognitive and structural categories:
1. The abstract – provides a short summary of the event(s)

2. The orientation – establishes the time, place and persons involved
3. The complicating action – narrates the central event that forms the basis for
the telling of the story
4. The evaluation – specifies why the story is told
5. The resolution – establishes what happened at the end
6. The coda – signals the end of the story and a return from the time of the nar-
rative to the present.
Chapter 10. Temporal structure of news reports 249
While personal narratives typically follow the above order of the six structural
elements (with evaluation often expressed across the whole text), news stories are
rather different. The difference consists not only in the different linear arrange-
ment of the individual narrative components but also their frequent recurrence,
because the components may be expressed several times in various parts of the
news article. This is the result of the recycling of the major aspects of a news story
in different textual segments and their presentation through different voices.
Beyond the headline, a news text opens – almost without exception – with
the lead, which performs the role of the abstract. In many cases, the lead serves
as an additional abstract, conveying essentially the same information as the head-
line in case of summarizing headlines (cf. White 2003: 87). It can also serve other
functions, performing the narrative steps of the complicating action (materialized
through its focus on the main event), the evaluation, as well as the resolution.
This is partly because the category of the abstract also subsumes elements that
otherwise feature in the categories of orientation (cf. the identity of the main news
actors and the spatio-temporal setting of the event), complicating action/reso-
lution (cf. the main newsworthy event and/or its outcome), and evaluation (cf.
the subjective perspective present in the choice of particular lexical and syntactic
structures to encode the event, as well as the choice of the topic itself, which is
governed by news values). As mentioned previously, some headlines have a dis-
tinctly dual structure, in which case the initial part provides the orientation and
the second part (e.g. following the colon) describes some other narrative category,
e.g. the complicating action/resolution, or, in the case of heteroglossic headlines,
explicit evaluation through the accessed voice. Occasionally the linear sequence
of the two parts may be reversed. The lead, as the second textual segment with an
abstracting function, is similar to the headline in its capability to include aspects
of several narrative categories at the same time.
As regards the other narrative categories, evaluation concerns the focusing of
the story in a particular direction. In personal narratives, it typically occurs at the
end. If we take evaluation in the broad sense of ‘focusing the story’, then it has a
strong presence in the lead because this structural element of the news text “forms
the lens through which the remainder of the story is viewed” (Bell 1991: 152). Al-
though evaluation in news texts is present in the text-medial position of the lead,
it is also present initially in headlines because they open up the story by selective-
ly focusing on particular aspects of the event. The selection of the content and
its representation through linguistic means may impose a particular evaluative
framework (White 2003: 78). What is also involved is the articulation of certain
ideological positions (cf. the critical tradition established by Fowler 1991, such as
Richardson 2007) but also subjectivity and personal involvement (cf. Bednarek
250 News and Time
2006). As discussed in Section 6.3, the choice of the present tense in headlines
can be interpreted as one of the means that embody the cognitive category of
‘evaluation’ understood in the sense of a narrative element providing a focus on
some aspect of the story.
The complicating action is the main event reported in the story: in narratives,
it is described as a chronological sequence of the individual steps that make up
the event. In news texts, however, such linear progression is absent. Results are
presented first; causes are specified later. The non-chronology is, in fact, among
the most characteristic features of news stories:
Where chronological order defines the structure of personal narrative, a com-
pletely different force is driving the presentation of the news story. Perceived
news value overturns temporal sequence and imposes an order completely at
odds with the linear narrative point. It moves backwards and forwards in time,
picking out different actions on each cycle. […] This wilful violation of our ex-
pectations that narratives usually proceed in temporal succession is distinctive of
news stories. It may also have repercussions for how well audiences understand
news stories. (Bell 1991: 153)
Resolution concerns the manner in which stories are ended – it is their termi-
nation which makes them into rounded stories. News stories, by contrast, often
“finish in mid-air” (Bell 1991: 154). This is the result of the incremental fashion
in which they are built from segments that contain information with a decreasing
degree of importance.
The coda, as an optional element in personal narratives marking the shift
from the narrative back to the context of the narration, is absent from news sto-
ries. That is because the boundary points between the narrative of the story and
some non-narrative context need not be signalled through verbal means.
The most relevant finding that arises from the application of the structur-
al categories of personal narratives onto news texts is that the latter recount the
story in a non-chronological sequence. The atemporal organization is motivated
by the operation of news values that reorganize the content according to require-
ments other than the need to proceed in a linear manner. The process of encoding
thus does not follow the sequence of the events that make up the news story. Fig-
ure 10.1 illustrates the non-chronological temporal structure of a sample article,
following the format used by Bell to highlight the mismatch between the narrative
structure and the overall time structure.
STORY TIME
News article
STRUCTURE STRUCTURE
(headline)
Abstract BP is hit with a record $4.5bn Time 1/2
penalty over Gulf spill damage (past/present)
(subhead)
Oil giant pleads guilty on 14 counts Time 1 (past)
Orientation Managers face manslaughter charges Time 3 (future)
(lead)
BP will pay the biggest criminal Time 2 (future)
penalty in US legal history by some
Complicating distance after reaching a $4.53bn Time 1 (past)
action (£2.9bn) settlement with American
authorities over the Gulf of Mexico
spill.
Evaluation (body copy)

Two men who worked for the oil Time 4 (past)
giant at the time of the disaster are, Time 5 (future)
meanwhile, facing involuntary
manslaughter charges, while a third
Resolution was yesterday charged with lying to Time 1 (past)
federal investigators and obstructing Time 6 (past)
an enquiry into the spill by the US
congress.
Figure 10.1 Narrative analysis of a sample news story and its simplified time structure
(after Bell 1991) (The Independent; 16 November 2012)
Figure 10.1 provides the opening of a medium-sized article in a daily newspa-

per. While in short articles it is fairly easy to chart the categories of the narrative
structure and the individual times that are encoded within the event, with longer
news texts the complexity rises significantly, and it becomes clear that assigning
some parts of the news text to the individual narrative categories is increasingly
difficult, requiring substantial intuition and subjectivity of judgement. Evidently,
the narrative categories work well for the headline, the subhead and the lead, but
the situation in the body copy is much less clear-cut.
Before proceeding with the temporal structure of news stories (see Sec-
tion 10.2), let us briefly discuss some specific narrative features that we may find
in news stories. In addition to the different story structure, Bell (1991: 155) iden-
tifies several other differences between news stories and face-to-face narratives:
252 News and Time
1. Narratives are personal, news stories report on the experience of others;40

2. Narratives are provided from the narrator’s point of view, news stories cite a
range of sources;
3. News are obsessed with giving precise numbers, narratives often lack relative
precision;
4. Personal narratives have simple syntax, news stories use long and complex
sentences.
News stories are thus characterized by content that is based on:
– vicarious (non-reflexive, other-oriented) experience and the framing of the

narrative from an external perspective;
– heteroglossic composition with a multiplicity of voices involved, which re-
sults in numerous deictic shifts that increase the temporal complexity of the
news text;
– the rhetoric of numbers and the intention to document the event in as precise
(or quasi-precise) way as possible; and
– syntactic complexity.
Bell observes that non-linear structure of news is imposed on the story in the
process of news production when the journalist reassembles bits of information
in a newsworthy order (Bell 1991: 172).41
Bell’s conclusion that “In news, order is everything but chronology is noth-
ing” (1991: 172) is refined by Schlokkenbroek (1999). In her analysis of narra-
tive structures and time in news stories, she shows that some narrative elements,
particularly in short news stories, may be intertwined and hard to distinguish
from each other. That complexity arises from the journalistic need to write brief
and compact texts. However, she shows that the temporal organization of events
40. The distinction should not be seen as a strict dichotomy. What qualifies as personal narra-
tives are also accounts of fictional events, and stories narrated by fictional narrators (cf. literary
fiction). Similarly, the precise nature of narrator identity is of little technical significance at
this point (i.e., personal/fictional identity in the case of face-to-face stories vs. non-personal/
institutional identity in the case of news stories that are cast as impersonal accounts even if
bylined to individual journalists).
41. It is noted that the information is often obtained from an interview that originally cov-
ered the relevant events chronologically. To the four features mentioned in Bell, one might
add another, namely the fact that news stories are composed as a collaborative effort by several
journalists/editors, who interact with the text at various stages of its production process (e.g.
by formulating the headline). This feature, however, is not crucial since personal narratives,
though usually produced by a single speaker, can also rely on the collaborative involvement of
other interlocutors who can co-construct the narrative.
on the level of discourse structures frequently matches the structure of the event
itself. The mismatch between the two structures is not an obstacle blocking the
comprehension of news stories because readers can always infer the underlying
event structure where it deviates from the chronological description.
The non-chronology of news stories does not concern the event sequenc-
es themselves but their textual presentation. It is, thus, a surface feature of the
events reported, while the underlying order of the individual subevents can be
inferred – as in the previous example. By distinguishing time structure and narra-
tive structure, Schokkenbroek argues that news stories are characterized by tem-
poral displacement (i.e., the rendering of events in a different sequence) rather
than a non-chronological event sequence (1999: 76). The temporal displacements
are textually marked by means of several linguistic devices, such as temporal con-
nectors, anchors and verb tense.
In addition to temporal displacement, the chronological discourse structure
of the news story is complicated due to the interference of the narrative category
of evaluation. As mentioned above, evaluation concerns the ways in which the
significance of a narrative can be foregrounded through various formal means
known as ‘evaluative devices’. Following Labov, evaluation is divided into internal
and external, depending on whether it operates below the level of the clause (and
thus complicates the narrative syntax) or above the level of the clause, e.g. when
the narrator “steps outside the story and tells the listener or reader the point of
the narrative in a direct mode” (Schokkenbroek 1999: 81). Very frequently, the
external evaluation in this sense is achieved by an accessed voice of some news
actor who either provides an eyewitness account or a personal commentary on
the event, which can be realized through various discourse representations differ-
ing in the degree of directness that is assigned to the accessed voice (e.g. direct/
indirect/free direct speech):
The events of news stories are externally evaluated by interrupting and suspend-
ing the basic action by means of inserting orienting information, direct or third
person report. The events are internally evaluated through the use of foreground-
ing devices such as Intensifiers, Comparators, Correlatives and Explicatives.
(Schokkenbroek 1999: 88)
External evaluation and the description of the event through the accessed voices
of various news actors results in the alteration between a third-person (objec-
tive) and first-person (subjective) narrative. The frequent shifts of the narrative
perspective bring about further shifts in the temporal structure of the news text.
254 News and Time
10.2 Temporal structure of news stories
As to temporal structure, Bell identifies the sub-components that make up the

main event and arranges them in a natural temporal sequence in what could be
basically identified as a script of the main event. If we apply this approach to the
article provided in Figure 10.1 above, we can see that the main event consists
of the announcement of the settlement (even though another major issue is ap-
pended to it, namely the possible prosecution of the managers). The sequence of
actions that make up the main event is approximately as follows:
– The conclusion of the settlement between the parties (= unspecified past-time

action)
– The announcement of the settlement (= yesterday)
– The approval of the settlement by court (= indefinite future time)
– The payout of the sum of money (= indefinite future time)
As the next paragraph of the article indicates, the court approval is still pending,
which means that the modality attached to the main event (‘BP will pay’) needs to
be reinterpreted from certainty to future possibility:
In its settlement, which remains subject to court approval, the FTSE 100 oil
giant agreed to plead guilty to 14 counts of criminal misconduct to resolve
all federal criminal charges and claims by the Securities and Exchange
Commission (SEC) relating to the Deepwater Horizon spill in 2010, which
killed 11 workers and spilled nearly 5 million barrels of crude oil into the Gulf
of Mexico.
At the same time, there is a great deal of other information in the article that
is also presented in a non-chronological manner because it provides the back-
ground to the main event by setting up the context, outlining the history of the
case, discussing the possible future developments, etc. Thus, the following ele-
ments are also expressed, clausally and nominally, in the five textual segments of
the story provided so far. Some of the lexical clues are provided in brackets:
– Major past incident from 2010 (gulf spill)

– Comparison with previous settlements made in the past (record)
– Employment status of two workers (who worked … at the time of the disaster)
– Bringing charges in court (was yesterday charged with)
– Examination by the police (lying to federal investigators)
It would be possible to arrange all these on a temporal line, starting with the
oil spill in 2010, i.e., the original event that actually triggered all the other steps
(investigation; questioning; settlement; present and future charges; etc.). Howev-

er, the article also contains further background information that is related to the
main event in a less direct way. This concerns similar previous incidents as well as
possible future outcomes, cf.:
[…]
The fines announced yesterday far outstrip BP’s last major settlement with the
Department of Justice in 2007 when it paid out about $373m to resolve three
separate inquiries into a 2005 Texas refinery explosion, an Alaska oil pipeline
leak and fraud for conspiring to corner the US propane market.
BP’s shares fell by 0.35p to end the day at 425.4p.
[…]
A conviction would see the two men each face up to 10 years in prison on
each of the of 11 seaman’s manslaughter counts, up to 8 years on each of the
11 involuntary manslaughter counts and up a year behind bars on count of
violating the Clean Water Act.
The additional events include:
– 2007 settlement by BP
– 2005 explosion
– Alaska oil pipeline leak
– fraudulent activities in the US market
– drop in the value of shares
– future conviction
– sentence following a future conviction
– violation of the Clean Water Act
The inclusion of all this background information, combined with quotations

which shift the temporal frame of reference through direct quotes, is what signif-
icantly complicates the temporal structure of news texts beyond the non-chrono-
logical presentation of the individual components of the main event.
There are several possible representations of the temporal composition of a
news article. We can mark the time of each event, as encountered in its textual
representation in the news article, without arranging the textual presentations in
terms of their actual underlying sequence. This approach is used in identifying
the time structure of the last example in Figure 10.1 above – the events are marked
with linearly sequenced numbers as Time 1, Time 2, Time 1, Time 3, Time 2,
where the repetitions indicate the return to a time zone that has been mentioned
previously.
Delin (2000) presents a different representation of time in news articles. She
organizes the events that make up the news story according to the order in which
256 News and Time
e10 e12 e14 e16 e18 e20 e22

e1 e2 e3 e4 e5 e6 e7 e8 e9 e11 e13 e15 e17 e19 e21 e23 e24
s1
s2
s3
s4 s5
s6
s7
s8
s9
Figure 10.2 A complex timeline of a news story distinguishing between ‘events’

and ‘states’ (Delin 2000: 21)
they actually occurred in real life, distinguishing between ‘events’ (i.e. process-
es) and ‘states’. The timeline for the news story is then represented in a linear,
chronological sequence of ‘events’ that make up the story, alongside parallel lines
standing for the individual ‘states’ that show a temporal overlap with the ‘events’
and each other (cf. Figure 10.2). The temporal structure of an article is then rep-
resented by a non-sequenced arrangement of the individual events and states, as
they are mentioned from the opening of the story in the headline until the end of
the news story.
Thus, in order to create a timeline of the ‘Record BP Penalty’ story above, we
have to first extract all the events and states from the entire article and then ar-
range them in the order in which they occurred in real life. As regards the actions
identified above, the sequence of events and states may be as shown in Figure 10.3
(note that identifying the beginnings and ends of states necessarily involves a de-
gree of subjective judgement, cf. Delin 2000: 17).
Apart from the obvious difficulty of distinguishing between ‘events’ and
‘states’, there are several other issues that arise when the events that make up the
story are being extracted from the news report. Some of the events and states are
coded in nominalizations (an enquiry) or through nominal premodifications that
actually combine two temporally distinct events (manslaughter charges). Some
events may have a more complex temporal structure on account of, e.g., premod-
ification through superlatives. Thus, the phrase ‘the biggest criminal penalty in US
legal history’ is not only related to the events of ‘(past) settlement’ and ‘(future)
payment’ but also invites a comparison with the entire time zone preceding the
penalty. If all these nuances are taken into account, the timeline of the news story
may become exceedingly complicated.
Event Description Time

E1 2005 explosion 2005
E2 2007 settlement by BP 2007
E3 Alaska oil pipeline leak unspecified past time
E4 fraudulent activities in the US market unspecified past time
E5 lying to federal investigators unspecified past time
E6 obstructing an enquiry into the spill unspecified past time
E7 Deepwater Horizon spill in 2010 2010
E8 killing of 11 workers 2010
E9 spilling of 5 million barrels 2010
E10 violation of the Clean Water Act unspecified past time
E11 reaching a settlement with US authorities unspecified past time – recent past
(MAIN EVENT) (yesterday?)
E12 BP pleading guilty unspecified past time (yesterday?)
E13 third man charged yesterday
E14 announcement of fines yesterday
E15 drop in the value of shares yesterday
E16 court approval of the settlement unspecified future time
E17 payment of the biggest criminal penalty unspecified future time
in US history
E18 managers facing charges unspecified future time
E19 possible future conviction unspecified future time
E20 possible sentence following a possible unspecified future time
future conviction
S1 two men working for BP unspecified past duration
S2 enquiry into the spill by the US Congress unspecified past duration
Figure 10.3 The temporal sequence of events and states in the order of occurrence
in the ‘Record BP Penalty’ story
Another complication arises when we try to arrange the events in a sequence

because many of the events are simultaneous (or almost simultaneous), and it is
near to impossible to decide on their precise order, which is an issue that Delin
(2000) does not appear to address explicitly. While the actual times of some of the
events are provided in the text (cf. the expressions in italics in the list of events
above), the majority are left unspecified, and it is up to the reader to make sense
of the ordering of the individual events.
258 News and Time
Given those reservations, we can see that the MAIN EVENT of the news story is
only the 11th event in the sequence of all events mentioned in the article. That is
because the main event is related to a number of prior events (as a consequence),
and some previous background context is also provided in the form of the earliest
events mentioned in the story. The non-chronology of the news text is evident –
the news report opens with the main event E11 and then moves forwards and
backwards in time to encode the other events (for more details, see below).
As regards the temporal sequencing of events, the main event in E11 can be
assumed to have happened in the recent past, presumably ‘yesterday’, even though
the definite time adverbial is explicitly used only in connection with the ‘charg-
ing of the third man’ (E13), the ‘announcement of fines’ (E14), and the ‘drop in
the value of shares’ (E15), but not the actual main event E11. In terms of the
sequenced events, E11–E15 all happened within the same time zone, namely ‘yes-
terday’, with E11–E13 most likely being simultaneous, and E14 and E15 following
shortly thereafter. In a similar way, events related to the ‘oil spill’ (‘Deepwater Ho-
rizon spill’ in E7, ‘killing of 11 workers’ in E8, ‘spilling of 5 million barrels’ in E9,
and ‘violation of the Clean Water Act’ in E10) occurred at the same time.
As mentioned above, the temporal anchorage for most of the events is left un-
specified and needs to be inferred. However, an event that is initially unspecified
in terms of its time can also be anchored to a specific point in time. The initial
general placement of the event in some time zone (e.g. the past) is then comple-
mented with a more precise time. Thus, for example, the ‘oil spill in the Gulf of
Mexico’, an event encoded as ‘Gulf spill damage’ in the headline and ‘the Gulf of
Mexico spill’ in the lead, is linked to a specific point in time only in the second
paragraph of the body, cf. …the Deepwater Horizon spill in 2011, which killed 11
workers and spilled nearly 5 million barrels of crude oil into the Gulf of Mexico.
What Figure 10.3 represents well is the multitude of the different events and
states that are mentioned in an article. The timeline, however, does not capture
the non-chronological order in which the different times are actually arranged in
the news text.
For that reason, let us consider a third possible representation of the temporal
structure of news stories, one that does more justice to the cyclic pattern in which
news content is presented in news texts. As mentioned above, each news story
contains a MAIN EVENT, which is always specified in the summarizing lead and
is often present in the headline as well. If the time of the MAIN EVENT is coded as
Time 0 (in order to indicate its centrality), the other subevents will be coded with
minus and plus signs, depending on whether they precede or follow the MAIN
EVENT. In this way, they will extend in two opposite directions along the timeline.
Events identified in this way have the advantage that their marking makes it im-
mediately obvious in what direction the sequence of events narrated in the news
Time 0 (E11: BP hit with a penalty) headline

Time –1 (E7: Gulf spill)
Time 0 (E12: BP pleading guilty) subhead
Time +1 (E18: charges faced by managers)
Time +2 (E17: BP will pay) lead
Time 0 (E11: reaching a settlement)
Time –1 (E7: spill)
Time –2 (S1: men who worked) body copy
Time –1 (E7: at the time of disaster)
Time +1 (E18: facing charges)
Time 0 (E13: was yesterday charged)
Time –3 (E5: lying to investigators)
Time –1 (E7: spill )
Figure 10.4 Article timeline across the main textual segments (italics indicate the formal
realization of the relevant time)
story moves with respect to the MAIN EVENT. With canonical news stories, the
MAIN EVENT is typically located in the recent past and projected into the readers’
receiving time; in that case, all subevents with the minus sign will be located in
the past and all subevents with the plus sign will be in the future. The arrangement
of events then revolves around the central position of Time 0 of the MAIN EVENT.
There appear two possible ways of identifying the times of the events that
precede and follow the MAIN EVENT marked with Time 0. First, they can be iden-
tified with sequential numbers indicating their point-of-occurrence in the news
text, without regard to the actual ordering of the subevents in real life. The plus/
minus times would thus indicate the zigzagging pattern in which the story line
moves through the news text while pivoting around the MAIN EVENT. The article
in Figure 10.3 then has the following timeline (see Figure 10.4; the repetition of
the numerals indicates the repetition of a particular subevent).
This representation lets us see how the timeline in the story moves in cycles
from Time 0 into the past and into the future again. It is evident that the cyclicity
in the main textual segments of the story arises from the fact that the headline, the
subhead, and the lead serve as textual summaries of the main event and repeat the
basic information, while adding some minor modifications. Eventually, the body
copy repeats the information once again, while finally elaborating with additional
circumstances.
The second method of arranging the subevents around the MAIN EVENT iden-
tified as Time 0 is more complex and requires the prior extraction of all subevents
from the news story. They are all then arranged chronologically along a timeline
(as long as their actual order can be plausibly decided). Here, the MAIN EVENT
is conceptualized as Time 0 again and all previous events are identified with the
260 News and Time
minus sign. This arrangement then renders the following extracted timeline of
chronologically sequenced events:
…, Time –3, Time –2, Time –1, Time 0, Time +1, Time +2, Time +3, …
The representation of the actual linear sequences in the news story then moves in
a similar zigzag fashion as in the case of the previous method; only the numerals
indicate the relative ‘distance’ between two events in the story rather then the
order in which they occur within the news text. In this way, the linear analysis
may indicate, for instance, that the temporally most distant events (i.e., the pre-
vious cases that form the background to the news story and the possible future
outcomes), tend to occur towards the end of the article, thus being truly marginal.
Regardless of the exact method of representing the individual subevents, the
approach of placing the main event in the centre of the time axis as Time 0 has
the advantage that it can illustrate how the other actions and circumstances men-
tioned in the text variably shift to the past and the future in a non-chronological
zigzag fashion.
10.3 Modelling the internal structure of news texts
Apart from analyzing news stories as narratives, there are other approaches to the
study of the structure of news articles. I will briefly review three models which
approach the news text in terms of the thematic structure, the conceptual frame of
the event, and the orbital organization of content. My overall argument is that the
characteristic non-chronology in news can be interpreted, at least partly, as aris-
ing from the organization of content according to subjectively perceived criteria
and the assessment of what the text writer considers relevant and newsworthy. In
that sense, the non-chronology has an interpersonal dimension.
Thematic structure of news texts
Van Dijk’s (1988) semantic-cognitive approach looks at the thematic structure

of news, identifying such semantic categories as main event, consequences,
circumstances, previous event, history, verbal reaction, evaluation,
and expectation. Bell (1991) refines the categories and applies them simulta-
neously to the major textual components of news (headline, lead, body copy)
and the six elements of the narrative structure of news stories in the Labovian
paradigm presented in the previous section. In Bell’s model, a news text consists
of three major components: abstract, attribution and story. Each of those
components contains certain textual and semantic categories and is to be found
News text
Abstract Attribution Story
Headline Lead Source Place Time
Episode 1 Episode n
News Journalist’s
agency byline
Event 1 Event n
Attribution Actors Setting Action Follow-up Commentary Background
Time Place Consequences Reaction Previous History

episodes
Context
Evaluation
Expectations
Figure 10.5 The model structure of a news text (Bell 1991: 171)
in different textual segments. Thus, abstract is realized in the headline and the
lead. Attribution is realized through the categories of source (news agency; or
journalist’s byline), place and time. Story is the most complex: it can be made up
of several episodes, each consisting of several events. Each event, in turn, has its
own complex structure, consisting of attribution, actors, setting (time and place),
action, follow-up (consequences and reaction), commentary (context, evaluation
and expectations), and background (previous episodes and history). The model is
graphically represented in Figure 10.5.
Although the classification provides an attractive model of the thematic
structure of news texts, it has some drawbacks. It might be objected that it mixes
narrative, discourse-structural, and semantic categories. Likewise, the delimita-
tion of the individual components appears not to be entirely fool-proof. As noted
by Bell himself, for instance, the abstract inevitably contains components of
the story. In addition, the separation of the abstract from the story itself is
somewhat problematic since that conception is applicable to the canonical sum-
marizing headline, on which it is based. In order for the model to be tenable, the
category of the abstract would have to be conceived of in a more general sense
262 News and Time
than the way it is defined by Labov and Waletzky (1967), e.g. not as a story sum-
mary but, rather, as a news-text-opener, in order to reflect the diverse types of
headlines. Another problematic category is evaluation since that can permeate
the entire text rather than being limited to the commentary section.
Last but not least, the category of attribution in this model appears on two
levels: as (1) a high-order discourse-structural feature pertaining to the news text
as such and reflecting the preoccupation with the process of news production,
and (2) a low-order semantic component of an event that makes up the news sto-
ry, which can refer to the actual source of information or the voice that is accessed
in order to narrate the event. Because the attribution functions quite differently
in each case, it would be more suitable to make a terminologocial distinction
between the two, e.g. by using the expression ‘meta-attribution’ to describe the
source, place and time of the story, i.e. referencing of a story to a news agency or a
particular journalist. The actual attribution would then consist of identifying the
sources of specific statements, formulations, and perspectives present in the text.
The latter tend to be quite diverse, particularly in news texts incorporating multi-
ple voices. The attribution of voice can even be present in headlines where it can
attribute a whole story to such an external news source and perform the function
of hedging the propositional content of the headline.
In general, the model can be successfully applied to canonical articles but its
categories are less suitable for more unusual news texts (cf. Ungerer 2000: 178).
It is understandable that the model reflects the components of stories that help
those involved in the news production process to isolate the individual compo-
nents, but its complexity also prevents it from being used as a true analytical tool.
Conceptual structure of the news story & event frames
Some of the drawbacks have been addressed in a model of the news text offered
by cognitive linguistics. Rather than trying to account for the news text simulta-
neously from a variety of perspectives, the cognitive account is methodologically
consistent in that it starts to deal with the news text first of all on the level of
the conceptual composition of the entire event, referred to as the event frame
(cf. Ungerer 2000). The news text is built around the conceptual category of the
main event, with the other conceptual categories appearing in a non-chronolog-
ical order. That kind of presentation of the components of the event constitutes
a conceptually natural story line and forms the basis for readers’ expectations
when processing a traditional news text. The pattern is a result of the historical
development of news discourse – the sequential organization of the news text,
common in the 19th century, was replaced with the top-down (inverted pyramid)
structure in modern news stories. As a result, traditional news stories do not open
with background information that appears in the opening position of the frame
that describes the event.
The non-sequential organization of the news story has been elaborated by
Ungerer (2000), who identifies several conceptual categories in the event frame
of news stories. The categories, arranged sequentially (i.e., in the natural order in
which they exist within the event frame) are: history; circumstances; previ-
ous event; main event; (verbal) reaction and evaluation; consequence
of verbal reaction; and expectation.
Most interestingly, Ungerer observes that not many modern news stories ac-
tually start with the main event – a fact that could strike us as surprising, particu-
larly if we take into account the traditional view of articles as having the top-down
organization. Instead, the point of entry into the article is achieved through some
other conceptual category within the frame. Such alternative openings to news
stories are referred to as ‘extracted topics’. Their significance lies in the fact that
they are promoted into the most prominent position in news texts, i.e. headlines.
The reason why some other point of entry into the article than the main event
is favoured is because the print media are not the sole mediators of news content.
Radio and television broadcasts and news sources are faster in delivering hard
news than the traditional print newspapers, which are no longer able to deliver
news with the same degree of recency. In many cases they, thus, cannot afford to
open the news story with the main event because that is already known to the
audience and, hence, is relatively less important. In a situation where some prior
knowledge of the main event can be expected, the printed newspapers may need
to look for other ways of mediating the information to their readers.
A similar observation is made by Jucker (2005), who comments on the de-
layed mediation of some information in the print media, e.g. the results of match-
es in the genre of sports reporting. Thus, news reports on sports and competitions
will focus on other information than the result, because “the text assumes that
most readers will already have been informed about these facts by some faster
medium, such as television or the Internet. It concentrates on background infor-
mation” (Jucker 2005: 17).
Ungerer (2000: 184) observes that “by deviating from the natural point of en-
try, the main event, the discourse structure of the news story is no longer in
accordance with the natural flow of conceptualization, neither with sequential
ordering nor with the top-down conceptualization, which is event-based”. He
suggests that this may adversely affect the comprehension of such stories, because
the extracted topics can disrupt readers’ expectations about the way news stories
are constructed.
Nevertheless, the ‘extracted topics’ approach is connected to yet another phe-
nomenon increasingly observable in modern newspapers: it is rarely the case that
264 News and Time
major news events are covered in a single article. Instead, several shorter articles
are printed in a cluster, typically with each of them focusing on a different aspect
of the story – an aspect that typically correlates to one of the conceptual cate-
gories (e.g. history; verbal comments; consequences, etc.). Ungerer calls it
the ‘package approach’ and argues that this form of presentation of news content
may actually counterbalance the staleness of the news itself. He observes that “the
event … may no longer be newsworthy, but will look more topical and more at-
tractive when packaged in this way” (2000: 188).
The extracted topic approach is also found – as already mentioned – in re-
ports on sports matches, which otherwise preserve a significant degree of the
chronological sequence of the events reported. However, individual newspapers
may cover the events in different ways. It has been noted that the popular newspa-
pers “are particularly keen to use extracted topics to preserve topicality of sports
reports, either by re-perspectivising the sequential report of the match … or by
altogether abandoning the idea of sequentiality in favour of isolated articles with
a dominant top-down structure” (Ungerer 2000: 193). Other papers still tend to
preserve the chronological sequence of the sports event in their sports reporting.
Let us sum up at this point what the cognitive approach can add to our discus-
sion of time in news. Because of the publication frequency, daily print newspapers
inevitably lose out to other media in the race for bringing the most up-to-date
news stories, where the most newsworthy component of the story is the concep-
tual category of the main event. The papers then have to look for a way of com-
pensating for the delay; and, in order to increase the relevance of the stories for
their readers, they refocus away from the main event, which can be expected to
be already known, to the other conceptual categories. The extracted topics and the
package approach then form a sort of ‘added value’ that the print media can offer,
as opposed to radio and TV broadcasts.
An interesting situation arises with modern electronic media. Online ver-
sions of established newspapers have become some of the major sources through
which the public obtains recent and topical information, readjusting the previous
balance between the written print media and the spoken broadcast media. Since
the major news websites are constantly updated, they bring current news round
the clock. As a result, online news channels may paradoxically see a return to the
more traditional pattern of news story, i.e. one characterized by opening of the
news text with the main event in the headline, because little prior knowledge of
the event from other media may be expected to exist on the part of the audience.42
42. What we consider here are news reports on events that have already happened, i.e. those
that can be formulated in terms of a focused story with a narrative structure revolving around
some main event. For an obvious reason, we leave out the phenomenon of live reporting,
At the same time, online newspapers can very effectively utilize the extracted
topics and the package approach by covering the story in a cluster of accompa-
nying articles, providing supplementary material and including external links.
Such hypertextually linked journalistic texts, supplementary materials and other
external and internal resources are often unified by means of a single major news
item that signposts the whole cluster with a headline identifying the main event
of the entire event frame.
Regardless of whether they diminish the importance of the main event cat-
egory in print headlines or actually increase it in headlines in online newspapers,
the changes in the nature of the news story have a pragmatic motivation. The
news is not presented in a vacuum – the journalists do not pursue some ideal of
an entirely objective, decontextualized presentation of facts, but formulate the
news text with a close regard to the perceived needs of the audience.
The textual production of the news story is unlike that of other informa-
tion-oriented genres, e.g. encyclopaedia entries, in that the journalist (and/or the
headline writer) needs to negotiate the level of expected knowledge of the event
on the part of the audience at the time the news story is processed and consumed.
This implies that a degree of subjectivity is involved that may result in the decision
to open the story with an extracted topic rather than the main event. The generic
structure of news texts can then be subtly manipulated in order to achieve specific
communicative effects.
Orbital organization and the interpersonal dimension of news time
The last approach to be mentioned is grounded in the systemic-functional tradi-

tion and is related to that school’s orientation toward the interpersonal dimension,
namely the role of subjectivity in news reporting texts with regard to evaluative
and interpretative positionings (Hunston and Thompson 1999; White 2003;
Martin and White 2005; Bednarek 2006).
which describes events as they are unfolding, since such reports (typically strictly chronolog-
ically organized) verbalize unfolding reports where the category of the main event might not
be known yet. Live news reports are then headlined differently – they usually identify the topic
or the occasion on which the report is being made and not the outcome (main event) or any
other conceptual category involved in the event (history; circumstances; previous event;
main event; (verbal) reaction and evaluation; consequence of verbal reaction; and
expectation). Occasionally, however, they do resemble the headlines of live news reports,
cf. LIVE: Fitch threatens to strip America of ‘AAA’ rating (The Telegraph; 16 October
2013).
266 News and Time
Particularly useful in our context is the work by White (2003), who develops
the idea of the rhetorical effects of news articles based on their text organizational
features. Following Iedema et al. (1994), he notes that many traditional hard news
texts manifest a cyclic pattern – a pattern referred to as ‘orbital organization’. News
reports typically open with extensive evaluation and “are organized around nodes
of evaluation which recur repeatedly and regularly as the text alternates between
this explicit passing of judgement and the detailing of the events upon which this
judgement is based” (White 2003: 68–69). While the evaluation is typically ex-
pressed through reported speech, the cyclic pattern present in many news reports
makes them, perhaps surprisingly, similar to personal gossip narratives.43
The structure of the news text is not only cyclic but also discontinuous. It is
made up of textual chunks that elaborate the information contained in the seg-
ments functioning as the abstract to the story: the headline and the lead. As White
notes,
the body of the report can be broken down into unconnected chunks each of
which acts to specify some aspect of the headline/lead by, for example, repeat-
ing it, elaborating it, contextualising it or offering a challenge to it. Thus, all the
chunks which make up the body reach back or reference the headline/lead in
some way, elaborating […] or providing further information. (White 2003: 83)
The discontinuity is, of course, closely related to the non-linearity of news stories.
Since their narrative structure is not sequential or chronological, the individual
blocks of text can be relatively freely rearranged, without significantly affecting
the coherence of the text. The headline, however, has a crucial text-organizing
function: “the important text forming relationship is not one which holds be-
tween adjacent chunks but the one which operates, often at a distance, between
each chunk and the headline/lead upon which it depends and which it elaborates”
(White 2003: 83). Understandably, the orbital, discontinuous pattern of textual
organization affects chronology as well: news stories are characterized by a dis-
rupted timeline.
White makes another interesting observation that is, by extension, germane
to our overall interpretation of temporality in news texts as having a significant
interpersonal dimension, cf.:
43. The idea of the cyclic presentation of the story is, of course, tied to the inverted pyramid
structure and the abstracting patterns found in headlines and leads. It ultimately goes back
to Bell, who notes that “the story cycles round the action, returning for more detail on each
circuit, and interspersing background and other events. The technique moves like a downward
spiral through the available information” (1991: 168).
The chronological orientation, the fundamental text to time-line iconicity of all

traditional narrative sub-types is indicative of a concern with, or focus upon,
how processes unfold in real or fictional worlds and hence with the relationship
of cause and effect. There is a clear focus, then, within narrative upon what Hall-
iday terms the ‘ideational’, upon the representation of external-world entities and
events and the logical relationships which hold between them.
(White 2003: 85–86)
Thus, if we view the chronological presentation of the components of a narrated

event as being ideationally-oriented, it is easier to see how the non-chronology
may, by implication, be perceived as non-ideational, i.e. motivated by some other
factors than the authors’ attempt at a mimetic representation of the extralinguistic
world. The rearrangement of the components of the event into a non-chrono-
logical sequence, as discussed above, is guided by a number of factors that have
an interpersonal motivation, such as the drive to satisfy the socially constructed
news values, and the text author’s attempt at a subtle negotiation of the relevance
of the facts for the benefit of the anticipated audience. All that is carried out with
respect to the text’s local context, namely the expectation that the news text will
be processed at a given socio-cultural place and historical time.
What the orbital model proposed by White and others does not, however,
take explicitly into account is that the recurrent representation of the processes
that make up the event can have two different realizations, which may need to
be distinguished. Thus, we can find the orbital pattern in the body copy, where it
operates as a true orbital structure. The individual ‘orbits’, made up of a combina-
tion of several actions, stand as the relatively independent textual chunks that can
be rearranged. The orbit in this sense is understood as a textual representation of
some components of the event that make a recursion into the past, disrupting the
chronology of the text, with another orbit going back to the past once again, etc.
The second type of orbital structure is, arguably, fundamentally different in
that the orbits have to be interpreted much more closely with respect to the top-
down pattern of textual organization of hard news texts. The repeated recursions
into the past, constituting the repetitive orbits, occur not within the body copy
but across the three basic structural elements of news articles: the headline, the
lead, and only then the body. The repetitions are thus related to the fundamental
property of news text to repeat information, encapsulated in the canonical func-
tion of the lead (and often, though not always, also the headline) to abstract the
information contained in the story. Thus, in the case of articles with a summariz-
ing headline, the basic sequence of processes constituting the gist of the report is
expressed three times: in the headline, the lead and the body copy. In addition,
however, the body copy itself may show the orbital pattern as well, as long as
the same process is described repeatedly, e.g. through eye-witness accounts, news
268 News and Time
actors’ comments, etc. An orbit within the body copy (usually the initial one) can
then form part of two distinctly different orbital patterns: one articulating the
general text-constructional features of news texts (headline – lead – body copy)
and the other related to the complex non-chronological internal structure of the
body copy itself.
10.4 Final remarks
As this chapter demonstrates, the non-chronology found in news texts is not

purely an accidental arrangement of diverse elements. Since some specific organ-
izing principles are at work, there is a perceptible pattern to the way temporality is
expressed in news texts. The cyclic presentation of content results, above all, in the
non-chronological but repetitive presentation of newsworthy events. Particularly
where the presentation of the history and the consequences of the main event are
referred to, the temporal structure may show a zigzagging pattern, with the time-
line moving repeatedly into the past and the future with respect to the main event.
Thus, although the temporal structure of news articles may, due to the non-
chronological arrangement of the events reported, appear to be fragmented, the
coding of temporality in news stories shows a distinct underlying pattern. As de-
monstrated above, this situation is directly related to the top-down organization
of the text and the multiple summarizations of the main event. While the inverted
pyramid structure entails that the news story is built around the main event
or some other component selected as the most newsworthy aspect of the story,
the multiple summary pattern ensures the repetition and re-expression of infor-
mation in various component parts of the news text, giving rise to the familiar
cyclic pattern. The interpretation of many of these phenomena as being principa-
lly reader-oriented lends support to our understanding of the whole issue of the
organization of time in news texts, when viewed from the combined perspectives
of ‘textual rhetoric of news texts’ and ‘textual rhetoric of headlines’, as a partly
pragmatic phenomenon.
Chapter 11
Conclusion
One of the most fundamental choices that all language users must make when en-
coding texts concerns the expression of temporality. The reference to time perme-
ates all discourse, if only because each finite clause must include a tensed form of a
verb and each non-finite or nominalized element is marked by the absence of such
a form. The selection of an appropriate tense by the speaker/writer, as well as the
decision to encode the content in finite vs. non-finite means, is more than a ‘se-
mantic’ description of some event; it is a pragmatic act through which the speak-
er/writer positions themselves with respect to the content communicated as well
as the possible recipients of the text. In general pragmatic terms, temporal deixis
thus connects the speaker/writer, the text, and the audience. Deixis has a subjec-
tive element to it because deictic forms are always formulated from a particular
position. The speakers/writers make the linguistic choices they do also because
they are motivated by their personal, as well as interpersonal, aims and goals.
It is from this position that I have approached the expression of temporality
in news texts. Since intersubjective meanings are constructed through texts, texts
are understood to be dialogic. Through particular lexical and syntactic choices,
they create a shared discourse space that comprises the readers. It is argued that
in news texts, the choice of certain tense forms is motivated by pragmatic factors.
Temporal deixis in news articles forms a system in which a number of distinct
patterns can be observed, such as the canonical use of some tenses, tense shifts
and the effects of heteroglossia. In modern online news texts, the traditional,
well-established patterns are reused, while some novel trends appear as well, par-
ticularly in relation to the unstable temporal anchorage of online texts and their
more complex textual structure comprising the additional level of the home page
news preview.
11.1 Tense shifts
A central finding in my analysis of verbal tense in news texts consists of the im-
portance of tense shifts. These shifts require us to pay attention to deictic and
non-deictic use of tense, particularly since the latter typically have a pragmatic
270 News and Time
motivation. In news, at least three tense shifts functioning in this way can be
identified:
1. ‘past tense → present tense’. This tense shift operates in the news production
process during the formulation of headlines. The reference to past-time events
is conventionally rendered through the present tense and can be explained as
a result of a complex projection of deictic centres.
2. ‘present tense → past tense’. This tense shift, observable in the news produc-
tion process as well, operates counter to the first shift. It appears to be the rule
in heteroglossic headlines, where the encoding of a past-time event is real-
ized in the past tense. Thus, the conventional present tense in authorial voice
yields to the past tense in accessed voice, which eliminates the impersonality
of the headline style.
3. ‘present tense → present perfect → past tense’. This double tense shift is ob-
servable across the main structural segments of news texts and consists of the
actual material presence of differently tensed forms and their transformations
within the co-referential process chains. Unlike the previous shifts, this dou-
ble tense shift pattern is physically realized in its entirety, with different tenses
being used for different reasons in different textual segments.
In my socio-pragmatic approach, I propose that the issue of temporality be treat-

ed within two mutually related systems that I have labelled the ‘textual rhetoric
of headlines’ (cf. the first two tense shifts above) and the ‘textual rhetoric of news
texts’ (cf. the third shift above).
11.2 Temporality and the textual rhetoric of headlines
With respect to the former system, a significant part of this book has dealt with
the explanation and interpretation of temporal deixis in news headlines. The con-
ventional present tense, commonly used in English headlines to refer to past-time
events, is conceptualized as the result of a projection of deictic centres. The model
of deictic centre projection, which is offered as an explanation for the relationship
between tense and time in headlines, operates with several distinct temporal pe-
riods: event time, coding time, publication time, and receiving time. While the
occurrence of the present simple tense in headlines is semantically enabled by the
absence of an adverbial of definite time (which would, if used, otherwise commit
the writer to the deictic past tense – as tends to be the case in the lead or the body
copy), the conventional selection of that particular tense draws on its numerous
grammatical meanings and creates the illusion of a shared time zone in which the
Chapter 11. Conclusion 271
temporal barrier between the event and the processing of the news text is discur-
sively removed. Not only does the non-deictic simple present tense in headlines
give the semblance of temporal co-presence between the event and the reader but,
thanks to its semantic properties, it also describes the event in a timeless manner.
The event described is thus presented as being vivid and having universal validity.
The conventional existence of the simple present has numerous consequenc-
es, both as regards headlines themselves and the remaining parts of news articles.
In headlines, apart from the rather high proportion of non-finite and nominal
structures, other tenses are frequently used as well. The non-deictic present tense
contrasts with the deictic simple past tense, of which both can be used to re-
fer to past-time events. However, there is a distinct pattern in the distribution
of the tenses. The present simple tense is associated with the authorial (internal)
voice of the paper, i.e. it tends to occur whenever the event is encoded from a
quasi-objective, impersonal position. The past tense, by contrast, typically marks
heteroglossia – it appears in stretches of text that are formulated as some external
(non-authorial) voice. The external voice can be accessed in a number of ways,
such as direct/indirect/semi-direct quotation, and identified as such through in-
verted commas, the colon structure, multimodal anchorage to an accompanying
visual, etc.
The use of the past tense has a somewhat paradoxical effect. The encoding of
past events is marked with a shift of tense: from the normally deictic simple past
to the non-deictic present tense, which we find as the default tense in headlines in
the paper’s authorial voice. Where the past tense is used in non-authorial accessed
voice, we actually experience a shift of tense in the opposite direction. That is,
the paper relinquishes its privilege to encode the content in its own voice, which
would call for the use of the conventional present tense, and transfers the encod-
ing power to some other entity, whose verbal reaction or comment is captured in
a seemingly unchanged, ‘authentic’ way in the simple past tense.
The present simple and the past simple tenses, thus, both strive to create a
similar pragmatic effect, only they go about it in different ways. By using the
present simple tense in its own voice, the paper projects the deictic centre and
establishes the illusion of a shared temporal co-presence for the benefit of the
reader – itself a part of the conventions of headlinese. By using the past simple
tense in the non-authorial accessed voice, however, the paper achieves a seeming-
ly unmediated access to the frame of the event reported, e.g. through the eyes of
an eyewitness or a direct participant. The accessed voice has the advantage of con-
noting the impression of immediacy, authenticity and direct participation. Last
but not least, the past tense in the accessed voice in headlines is typically accom-
panied by the use of such grammatical words as articles, pronouns and auxiliaries
272 News and Time
that are normally omitted in headlines formulated in the paper’s authorial voice.
The co-occurrence of the deictic past tense and the other forms is indicative of the
shift of the language style in headlines from the arguably detached and imperson-
al forms characterized by the minimal syntax of headlinese, to more personal and
involved conversational language.
The proper effect of the deictic past tense in accessed voice in headlines can
be appreciated only against the background of the headline conventions, which
prefer to use the stereotypical present tense for referring to past-time events.
However, text segments belonging to an external accessed voice are permitted
to go counter to the conventions of headlinese. There is thus tension as well as
complementarity between the two tenses in headlines – they are balanced with
respect to each other.
In addition to its significance in heteroglossic headlines, we also encounter
the deictic simple past tense in those headlines that are formulated in the paper’s
own voice. In this situation, several patterns can likewise be observed. The simple
past is used for reporting events that are classified as non-recent, despite the fact
that they may have come to the attention of the media only recently. The simple
past also serves to establish a particular event as a background to some other,
more newsworthy event that has a more central position thanks to its encoding
through the simple present tense in the same headline. Another backgrounding
function of the past tense occurs in headlines of satellite articles that accompany
the most important news text within a cluster of articles on a related topic. To a
certain extent, the subordinate headlines of articles forming such a cluster have
a similar function as multiple decks of heads and subheads in early news stories.
Within an article cluster, there is a hierarchical organization of texts that applies
to headlines as well.
It appears that in order to capture the complexity with which temporality is
encoded in many such headlines, we may have to modify our perception of the
temporal zone that constitutes ‘the past’ by distinguishing two related, yet inde-
pendent past zones. The first zone is the recent past, which, in terms of its length,
correlates approximately with the publication frequency of the paper. The events
that occur within this time frame are normally encoded through the simple pres-
ent tense and are automatically assumed to be ‘recent’. The second zone is the
non-recent past, which precedes the previous issue of the periodical publication.
It is in connection with events located in the non-recent past that the simple past
tense is used as a backgrounding device. This, however, does not discount the fact
that papers occasionally enhance the recency of stale news that technically fall
within the ‘non-recent past zone’ by encoding them in the present simple tense,
even though such a practice may be perceived by some readers as somewhat ma-
nipulative when they realize that the item is in fact non-recent.
11.3 Temporality and the textual rhetoric of news texts
The headline is far from being an autonomous textual unit. Not only can it stand
in a complex hierarchical relationship to other headlines within an article cluster
but it is also the opening part of its own article. The investigation of temporality
in news texts thus may start with headlines, where some of the most characteristic
formal features are found, but it needs to be followed up with analysis of the rhet-
oric of news texts beyond the headlines. I have demonstrated that one of the clues
to understanding temporal deixis in news texts is their structural organization
into three linearly sequenced segments: the headline, the lead and the body copy.
These segments are characterized by the cyclic presentation of newsworthy infor-
mation. Since the headline and the lead typically abstract information as textual
summaries of the entire story, the same information can – in an ideal case – be
expressed three times. While the headline encodes the event in the verb phrase
by means of the conventional present tense, the lead does so either in the past
tense in combination with a definite adverbial of time or, increasingly, the present
perfect in the absence of such an adverbial. The body copy then shifts into the past
tense. This gives rise to the triple tense pattern that is observable in many news
articles despite lexical and syntactic transformations of the verb phrase across the
three textual segments.
The fact that three different tenses can be used in the three structural seg-
ments to refer to one and the same event located in the recent past has significant
consequences for our understanding of this pattern of tense use in news texts.
If we consider the triple tense pattern in its pure, idealized form, then it is only
within the third segment (the body copy) that tense is used deictically. The simple
past tense – typically accompanied by an adverbial of definite time – anchors the
event in the past time. In this sense, the past tense acts referentially since it com-
municates salient temporal information about the event.
By contrast, tense use in the headline and the lead is non-deictic. The simple
present in the headline and the semi-deictic present perfect in the lead have a
strong interpersonal orientation: they do not act referentially since they do not
provide a temporal anchorage of the event to a particular time zone. Instead, they
heighten the current relevance of the event. While the simple present is a conven-
tional tense resulting from deictic projection, the present perfect negotiates the
relationship and the relevance of the event for the benefit of the reader at the time
of the reader’s consumption of the news text. The tense in the headline has the
strongest interpersonal orientation – it establishes the conventional contact with
the reader by postulating temporal co-presence. The current relevance is then sus-
tained by means of the present perfect in the lead. Eventually, the past tense in the
body copy concludes the temporal rhetoric by finally determining the real-time
274 News and Time
placement of the event. As mentioned above, this manipulation of tenses in the

headline and the lead is enabled by the operation of the two textual segments as
summaries that abstract the narrative contained in the news story. In short, the
double tense shift pattern forms a part of the rhetoric of hard news and is a con-
ventional resource that journalists can draw on when producing their texts.
Heteroglossia can affect the nature of grammatical tense in any of the three
segments. Since headlines frequently access external voices and contain stretches
of direct/indirect reported speech, other tenses than the simple present may be
used in the headline in order to break the conventions of headlinese and give
it a semblance of more authentic conversational language.44 Heteroglossic head-
lines – as well as the other structural segments – will inevitably result in a mod-
ification of the triple tense pattern because its idealized format applies only to
monoglossic stretches of texts formulated in the internal (authorial) voice of the
paper. The occurrence of a linguistic form within an accessed voice means that
the operation of genre conventions that regulate the news text is suspended over
the entire stretch of such an external voice. Manifested through the presence of a
direct/indirect speech quote, the accessed voice itself constitutes an interpersonal
feature that modifies the temporal situation of the text in its own way: by provid-
ing the reader with a seemingly in medias res access to the narrated event.
11.4 Temporality in online news
While online news contains the same underlying pattern of tenses, it displays
several other, quite specific features. Probably the most significant is the existence
of two textual levels on which the news text is presented. The first level consists of
the article preview on the newspaper home page and the second level is the article
itself on its own web page. Since the article is most usually accessed through a
hypertextual link from the newspaper’s web page, the article preview forms an
extention prefacing the news text. The preview thus adds another headline and,
usually, also a lead. Both are processed by the audience before reading the article
itself. As regards the cyclic presentation of information, online news can thus
contain up to six textual segments that recycle the same information (home page
headline, home page lead, article headline, subhead/caption, article lead, article
body copy). While there is usually some variation between the corresponding tex-
tual segments in the preview and the article itself, the tense shift pattern actually
appears twice in online news.
44. The seeming authenticity of a quote, particularly in headlines, does not imply that it is a
verbatim account. What matters is the impression, not the actual truthfulness to the original.
The shift of the deictic past into the non-deictic present, which is the result
of the encoding process of news production in the ‘bottom’ to ‘top’ direction, and
the shift of the non-deictic present tense into the semi-deictic present perfect and
the deictic past tense in the ‘top’ to ‘bottom’ direction in which the text is read, are
two sides of the same coin: what matters, arguably, is the fact that a switch of tense
occurs, not necessarily in what direction the tense is changed. The variability of
tense within the verb phrase is evidence that it is used for other purposes that
truthfully describe some newsworthy past-time event. Verbal tense and its shifts
function as pragmatic devices for constructing a shared deictic discourse space
with the intended recipients of the news texts.
In the leads of online news, there appears to be a preference for using the pres-
ent perfect over the simple past. This phenomenon has several explanations. The
present perfect expresses the current relevance and the recency of a news item in
a situation where there is no definite cut-off point: online news is produced on a
rolling news cycle that constantly sees newer and newer stories being uploaded to
the paper’s web page. Thus, news that appears online cannot, as opposed to tradi-
tional print news, be assumed to have happened ‘yesterday’ – the news is more re-
cent than that. The present perfect, as a grammatical category, has a semi-deictic
property since it can simultaneously refer to a past-time event yet rely on a shared
temporal context. Its emphasis on current relevance, arguably, makes the choice
of this tense a particularly suitable one for reporting online news.
The interpretation of the tense shifts also squares with other features com-
monly found in online news that have an impact on the coding of temporal deixis.
One of these is the time stamp, which is present in the attribution line of each
article, anchoring it either in definite time or, by means of deictic time adverbials,
in a time that is explicitly related to the moment the web page is accessed. The hy-
perlinking of articles also allows readers to trace a topic easily through the news-
papers’ past issues placed in their archives in a complex intertextual network.
Backtracking a particular topic through internet links in online newspapers often
takes the form of a certain preferred reading path since not all articles are mutu-
ally hyperlinked to the same extent. The explicit coding of temporal deixis in such
hypertextual networks is required in order to help the readers navigate the indi-
vidual news articles which, though mutually interlinked in a complex diachronic
way, are much more autonomous from other surrounding articles or, indeed, the
online ‘issue’ of an newspaper on a given day. Hence, online news tend to rely on
the masthead, the strapline or the attribution line much more than traditional
print articles, which are temporally contextualized in an unequivocal manner by
means of their inclusion in a particular issue of a newspaper.
Hopefully, the present pragmatically-oriented analysis and interpretation of
the coding of temporal deixis in news texts has shown that the topic holds much
276 News and Time
potential for further research. It is evident that the advent of new technologies
and the online transformation of the news industry has not only led to the emer-
gence of some new genres, text types and discursive structures, but it has also had
an impact on how the relationship between the text producer, the text and the
audience are linguistically encoded. While online news has achieved what had
probably always been the ultimate goal of journalists, namely the instant delivery
of information without any time lag, some new forms have emerged for the man-
agement of the discursive space and the encoding of temporality in such news
texts. Since news texts are constantly changing, evolving and developing, further
research into how traditional genre conventions are modified and how some new
communicative forms emerge is bound to yield many interesting findings.
References
Austin, John L. 1962. How to Do Things with Words. Oxford: Oxford University Press.
Bakhtin, Mikhail M. 1981. The Dialogic Imagination. Trans. by Caryl Emerson, and Michael
Holquist. Austin: University of Texas Press.
Bednarek, Monika, and Helen Caple. 2012. News Discourse. London and New York: Continuum.
Bednarek, Monika. 2006. Evaluation in Media Discourse: Analysis of a Newspaper Corpus.
London and New York: Continuum.
Bell, Allan. 1991. The Language of News Media. Oxford UK and Cambridge MA: Blackwell.
Bell, Allan. 1994. “Telling Stories.” In Media Texts: Authors and Readers, ed. by David Graddol,
and Oliver Boyd-Barrett, 100–118. Clevedon: Multilingual Matters.
Bell, Allan. 1995. “News Time.” Time & Society 4 (3): 305–328.
DOI: 10.1177/0961463X95004003003
Bell, Allan. 1998. “The Discourse Structure of News Stories.” In Approaches to Media Discourse,
ed. by Allan Bell, and Peter Garrett, 64–104. Malden MA and Oxford: Blackwell Publishing.
Bell, Allan. 2002. “Dateline, Deadline: Journalism, Language, and the Reshaping of Time and
Place in the Millennial World.” In Georgetown University Round Table on Languages and
Linguistics. Linguistics, Language and the Professions: Education, Journalism, Law, Medi-
cine, and Technology, ed. by James E. Alatis, Heidi E. Hamilton, and Ai-Hui Tan, 46–66.
Washington, D.C.: Georgetown University Press.
Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan. 1999.
Longman Grammar of Spoken and Written English. Harlow: Longman.
Brown, Gillian, and George Yule. 1983. Discourse Analysis. Cambridge: Cambridge University
Press. DOI: 10.1017/CBO9780511805226
Bublitz, Wolfram, and Neal R. Norrick (eds). 2011. Foundations of Pragmatics. Berlin and Bos-
ton: Walter de Gruyter. DOI: 10.1515/9783110214260
Bublitz, Wolfram, and Neal R. Norrick. 2011. “Introduction: The Burgeoning Field of Prag-
matics.” In Foundations of Pragmatics, ed. by Wolfram Bublitz, and Neal R. Norrick, 1–22.
Berlin and Boston: Walter de Gruyter. DOI: 10.1515/9783110214260.1
Bucaria, Chiara. 2004. “Lexical and Syntactic Ambiguity as a Source of Humor: The Case of
Newspaper Headlines.” Humor 17: 279–309. DOI: 10.1515/humr.2004.013
Bühler, Karl. 1934[1990]. Theory of Language. The Representational Function of Language.
Trans. by Donald Fraser Goodwin. Amsterdam: John Benjamins.
Caffi, Claudia. 2007. Mitigation. Amsterdam: Elsevier.
Cap, Piotr. 2008. “Towards the Proximization Model of the Analysis of Legitimization in Polit-
ical Discourse.” Journal of Pragmatics 40 (1): 17–41. DOI: 10.1016/j.pragma.2007.10.002
Cap, Piotr. 2010. “Proximizing Objects, Proximizing Values. Towards an Axiological Contri-
bution to the Discourse of Legitimization.” In Perspectives in Politics and Discourse, ed. by
Urszula Okulska, and Piotr Cap, 119–142. Amsterdam: John Benjamins.
DOI: 10.1075/dapsac.36.11cap
278 News and Time
Cap, Piotr. 2011. “Micropragmatics and Macropragmatics.” In Foundations of Pragmatics, ed.

by Wolfram Bublitz, and Neal R. Norrick, 52–76. Berlin and Boston: Walter de Gruyter.
Carter, Ronald, and Michael McCarthy. 2006. Cambridge Grammar of English. A Comprehen-
sive Guide. Cambridge: Cambridge University Press.
Chilton, Paul. 2004. Analysing Political Discourse: Theory and Practice. London: Routledge.
Chilton, Paul. 2005. “Vectors, Viewpoints and Viewpoint Shift: Toward a Discourse of Space
Theory.” Annual Review of Cognitive Linguistics 3: 78–116. DOI: 10.1075/arcl.3.06chi
Chovanec, Jan. 2000. Interpersonal Aspects of Printed News Discourse: A Sociolinguistic Study of
the Woodward Case Reporting in the Electronic Telegraph. An unpublished Ph.D. disserta-
tion. Brno: Masaryk University.
Chovanec, Jan. 2003a. “The Mixing of Modes as a Means of Resolving the Tension between
Involvement and Detachment in News Headlines.” Brno Studies in English 29: 52–66.
Chovanec, Jan. 2003b. “The Uses of the Present Tense in Headlines.” Theory and Practice in
English Studies, Vol. 1, ed. by Jan Chovanec, 83–92. Brno: Masaryk University.
Chovanec, Jan. 2005a. “From Zeroo to Heroo: Word-play and Wayne Rooney in the British
Press.” Slovak Studies in English 1: 85–99.
Chovanec, Jan. 2005b. “The (Un)conventional Use of the Simple Past Tense in News Headlines.”
In Patterns. A Festschrift for Libuše Dušková, ed. by Jan Čermák, Aleš Klégr, Markéta Malá,
and Pavlína Šaldová, 71–81. Praha: Charles University.
Chovanec, Jan. 2008. “Focus on Form: Foregrounding Devices in Football Reporting.” Dis-
course & Communication 2 (3): 219–242. DOI: 10.1177/1750481308091908
Chovanec, Jan. 2009. “Simulation of spoken interaction in written online media texts.” Brno
Studies in English 35 (2): 109–128.
Chovanec, Jan. 2010. “Online Discussion and Interaction: The Case of Live Text Commentary.”
In Cases on Online Discussion and Interaction: Experiences and Outcomes, ed. by Leonard
Shedletsky, and Joan E. Aitken, 234–251. Hershey, PA: IGI Global.
DOI: 10.4018/978-1-61520-863-0.ch012
Chovanec, Jan. 2012. “Implicit Meanings and Racism in Political Advertising.” In Discourse In-
terpretation: Approaches and Applications, ed. by Olga Dontcheva-Navratilova, and Renata
Povolná, 55–77. Newcastle upon Tyne: Cambridge Scholars Publishing.
Conboy, Martin. 2007. The Language of the News. London and New York: Routledge.
Cotter, Colleen. 2010. News Talk: Investigating the Language of Journalism. Cambridge:
Cambridge University Press. DOI: 10.1017/CBO9780511811975
Crystal, David, and Derek Davy. 1969. Investigating English Style. London and Harlow: Longmans.
Crystal, David. 1988. The Encyclopedia of the English Language. Cambridge: Cambridge Uni-
versity Press.
Culpeper, Jonathan. 2011. “Historical Sociopragmatics: An Introduction.” In Historical Socio-
pragmatics, ed. by Jonathan Culpeper, 1–8. Amsterdam: John Benjamins.
DOI: 10.1075/bct.31.01cul
Declerck, Renaat. 1991. Tense in English. Its Structure and Use in Discourse. London and New
York: Routledge.
Declerck, Renaat. 2006. The Grammar of the English Verb Phrase. Volume 1: The Grammar of the
English Tense System. Berlin and New York: Mouton de Gruyter.
DOI: 10.1515/9783110199888
Delin, Judy. 2000. The Language of Everyday Life. London: Sage.
Dor, Daniel. 2003. “On Newspaper Headlines as Relevance Optimizers.” Journal of Pragmatics
35: 695–721. DOI: 10.1016/S0378-2166(02)00134-0
References 279
Downes, William. 1998. Language and Society. Cambridge: Cambridge University Press.
DOI: 10.1017/CBO9781139163781
Downing, Angela. 2000. “Nominalization and Topic Management in Leads and Headlines.”
In Discourse and Community: Doing Functional Linguistics, ed. by Eija Ventola, 355–378.
Tübingen: Gunter Narr Verlag.
Durant, Alan, and Marina Lambrou. 2009. Language and Media. A Resource Book for Students.
London and New York: Routledge.
Dušková, Libuše. 1988. Mluvnice současné angličtiny na pozadí češtiny. Praha: Academia.
Dynel, Marta. 2009. Humorous Garden-Paths. A Pragmatic-Cognitive Study. Newcastle upon
Tyne: Cambridge Scholars Publishing.
Fairclough, Norman. 1989. Language and Power. London and New York: Longman.
Fairclough, Norman. 1995. Critical Discourse Analysis: The Critical Study of Language. Harlow:
Longman.
Fairclough, Norman. 2001. Language and Power, 2nd ed. Harlow: Pearson Education.
Ferguson, Charles A. 1983. “Sports Announcer Talk. Syntactic Aspects of Register Variation.”
Language in Society 12: 153–172. DOI: 10.1017/S0047404500009787
Fillmore, Charles. 1997. Lectures on Deixis. Stanford: CSLI Publications.
Fillmore, Charles. 1975. Santa Cruz Lectures on Deixis. Bloomington: Indiana University Lin-
guistics Club.
Firbas, Jan. 1992. Functional Sentence Perspective in Written and Spoken Communication.
Cambridge: Cambridge University Press. DOI: 10.1017/CBO9780511597817
Fleischman, Suzanne. 1990. Tense and Narrativity: From Medieval Performance to Modern Fic-
tion. Austin: University of Texas Press.
Fludernik, Monika. 1993. The Fictions of Language and the Languages of Fiction. London and
New York: Routledge.
Fludernik, Monika. 2003. “Chronology, Time, Tense and Experientiality in Narrative.” Lan-
guage and Literature 12 (2): 117–134. DOI: 10.1177/0963947003012002295
Fludernik, Monika. 2009. An Introduction to Narratology. Abingdon and New York: Routledge.
Fowler, Roger. 1991. Language in the News. Discourse and Ideology in the Press. London and
New York: Routledge.
Fowler, Roger. 1996. Linguistic Criticism, 2nd ed. Oxford: Oxford University Press.
Green, Georgia N. 1996. Pragmatics and Natural Language Understanding. Mahwah, NJ:
Lawrence Erlbaum.
Green, Keith. 1995. “Deixis: A Revaluation of Concepts and Categories.” In New Essays in
Deixis, ed. by Keith Green, 11–25. Amsterdam: Rodopi.
Green, Keith. 2006. “Deixis and Anaphora: Pragmatic Approaches.” In Encyclopedia of Lan-
guage and Linguistics, 2nd ed., ed. by Keith Brown, 415–417. Oxford: Elsevier.
DOI: 10.1016/B0-08-044854-2/00328-X
Grice, Herbert P. 1975. “Logic and Conversation.” In Syntax and Semantics, Vol. 3, ed. by Peter
Cole, and Jerry L. Morgan, 41–58. New York: Academic Press.
Hall, Jim. 2008. “Online Editions: Newspapers and the ‘New’ News.” In Pulling Newspapers
Apart: Analysing Print Journalism, ed. by Bob Franklin, 204–211. Abingdon and New York:
Routledge.
Halliday, Michael A.K. 1978. Language as a Social Semiotic. The Social Interpretation of Lan-
guage and Meaning. London: Edward Arnold.
Halliday, Michael A.K. 1985. An Introduction to Functional Grammar. London: Edward Arnold.
280 News and Time
Halliday, Michael A.K. 1994. An Introduction to Functional Grammar, 2nd ed. London: Edward
Arnold.
Halliday, Michael A.K., and Ruqaiya Hasan. 1976. Cohesion in English. London: Longman.
Halliday, Michael A.K., and Ruqaiya Hasan. 1985. Language, Context, and Text: Aspects of Lan-
guage in a Social-semiotic Perspective. Oxford: Oxford University Press.
Halliday, Michael A.K., and Christian Matthiessen. 2004. Introduction to Functional Grammar.
3rd ed. London: Edward Arnold.
Harcup, Tony, and Deirdre O’Neill. 2001. “What is News? Galtung and Ruge Revisited.” Jour-
nalism Studies 2 (2): 261–280. DOI: 10.1080/14616700118449
Harder, Peter. 1995. Functional Semantics. A Theory of Meaning, Structure and Tense in English.
Berlin: Mouton de Gruyter.
Hasan, Ruqaiya. 1984. “Coherence and Cohesive Harmony.” In Understanding Reading Com-
prehension: Cognition, Language and the Structure of Prose, ed. by James Flood, 181–219.
Newark: International Reading Association.
Hasan, Ruqaiya. 1985. “The Texture of a Text.” In Language, Context and Text, ed. by Michael
A.K. Halliday, and Ruqaiya Hasan, 70–96. Geelong: Deaking University Press.
Hoey, Michael. 2001. Textual Interaction. An Introduction to Written Discourse Analysis.
London and New York: Routledge.
Holquist, Michael. 1990. Dialogism: Bakhtin and his World. London: Routledge.
DOI: 10.4324/9780203330340
Huang, Yan. 2007. Pragmatics. Oxford: Oxford University Press.
Huddleston, Rodney D., and Geoffrey K. Pullum. 2002. The Cambridge Grammar of the English
Language. Cambridge: Cambridge University Press.
Huddleston, Rodney. 1969. “Some Observations on Tense and Deixis in English.” Language 45
(4): 777–806. DOI: 10.2307/412335
Hunston, Susan, and Geoff Thompson (eds). 1999. Evaluation in Text. New York: OUP.
Iarovici, Edith, and Rodica Amel. 1989. “The Strategy of the Headline.” Semiotica 77 (4): 441–
459. DOI: 10.1515/semi.1989.77.4.441
Iedema, Rick, Susan Feez, and Peter White. 1994. Media Literacy. Sydney: Disadvantaged
Schools Program, NSW Department of School Education.
Ifantidou, Elly. 2009. “Newspaper Headlines and Relevance: Ad Hoc Concepts in Ad Hoc Con-
texts.” Journal of Pragmatics 41 (4): 699–720. DOI: 10.1016/j.pragma.2008.10.016
Jacobs, Geert. 1999. Preformulating the News: An Analysis of the Metapragmatics of Press Releas-
es. Amsterdam: John Benjamins. DOI: 10.1075/pbns.60
Jakobson, Roman. 1984[1932]. “The Structure of the Russian Verb.” In Russian and Slavic
Grammar Studies 1931–1981, ed. by Linda R. Waugh, and Morris Halle, 1–14. Berlin:
Mouton & de Gruyter.
Jaszczolt, Katarzyna M. 2002. Semantics and Pragmatics. Meaning in Language and Discourse.
London: Longman.
Jones, Peter. 1995. “Philosophical and Theoretical Issues in the Study of Deixis: A Critique of
the Standard Account.” In New Essays in Deixis, ed. by Keith Green, 27–48. Amsterdam:
Rodopi.
Jucker, Andreas H. 1992. Social Stylistics: Syntactic Variation in British Newspapers. Berlin and
New York: Mouton de Gruyter. DOI: 10.1515/9783110851151
Jucker, Andreas H. 2002. “Hypertextlinguistics: Textuality and Typology of Hypertext.” In Text
Types and Corpora. Studies in Honour of Udo Fries, ed. by Andreas Fischer, Gunnel Tottie,
and Hans Martin Lehmann, 29–51. Tübingen: Gunter Narr Verlag.
References 281
Jucker, Andreas H. 2005. “Mass Media Communication from the Seventeenth to the Twen-
ty-first Century.” In Opening Windows on Texts and Discourses of the Past, ed. by Janne
Skaffari, Matti Peikola, Ruth Carroll, Risto Hiltunen, and Brita Wårvik, 7–21. Amsterdam:
John Benjamins. DOI: 10.1075/pbns.134.04juc
Jucker, Andreas H. 2012. “Linguistic Levels: Pragmatics and Discourse Analysis.” In Historical
Linguistics of English, ed. by Alex Bergs, and Laurel Brinton, 197–212. Berlin and New
York: Mouton de Gruyter.
Kaltenböck, Gunther, Wiltrud Mihatsch, and Stefan Schneider (eds). 2010. New Approaches to
Hedging. Bingley: Emerald Group.
Kazunari, Miyahara. 2009. “Why Now, Why Then?: Present-tense Narration in Contemporary
British and Commonwealth Novels.” Journal of Narrative Theory 39 (2): 241–268.
DOI: 10.1353/jnt.0.0030
Knox, John S. 2007. “Visual/Verbal Communication on Online Newspaper Home Pages.”
Visual Communication 6 (1): 19–53. DOI: 10.1177/1470357207071464
Knox, John S. 2009. “Punctuating the Home Page: Image as Language in an Online Newspaper.”
Discourse and Communication 3 (2): 145–172. DOI: 10.1177/1750481309102450
Knox, John S. 2010. “Online Newspapers: Evolving Genres and Evolving Theory.” In Applied
Linguistics Methods: A Reader, ed. by Caroline Coffin, Theresa M. Lillis, and Kieran
O’Halloran, 33–51. London and New York: Routledge.
Koyama, Wataru. 2009. “Indexically Anchored onto the Deictic Center of Discourse: Grammar,
Sociocultural Interaction, and ‘Emancipatory Pragmatics’.” Journal of Pragmatics 41 (1):
79–92. DOI: 10.1016/j.pragma.2008.09.002
Kress, Gunther, and Robert Hodge. 1979. Language as Ideology. London: Routledge and Kegan
Paul.
Kress, Gunther, and Theo van Leeuwen. 2006. Reading Images: The Grammar of Visual Design,
2nd ed. Abingdon and New York: Routledge.
Kress, Gunther R. 1977. “Tense as Modality.” UEA Papers in Linguistics 5: 40–52.
Labov, William. 1972. “The transformation of experience in narrative syntax.” In Language in
the Inner City: Studies in the Black English Vernacular, ed. by William Labov, 354–396.
Philadelphia: University of Pennsylvania Press.
Labov, William, and Joshua Waletzky. 1967. “Narrative Analysis.” In Essays on the Verbal and
Visual Arts, ed. by June Helm, 12–44. Seattle: University of Washington Press.
Langacker, Ronald W. 1991. Foundations of Cognitive Grammar. Stanford: Stanford University
Press.
Le, Elisabeth. 2012. “Gateway to the News: Headlines on Le Monde’s Home Page and Front
Page.” Discourse, Context & Media 1: 32–44. DOI: 10.1016/j.dcm.2012.05.004
Leech, Geoffrey N. 1971. Meaning and the English Verb. London: Longman.
Leech, Geoffrey N. 1983. Principles of Pragmatics. London: Longman.
Lennon, Paul. 2004. Allusions in the Press: An Applied Linguistic Study. Berlin and New York:
Mouton de Gruyter. DOI: 10.1515/9783110197334
Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press.
Levinson, Stephen C. 2004. “Deixis.” In The Handbook of Pragmatics, ed. by Laurence R. Horn,
and Gregory Ward, 97–121. Malden MA and Oxford: Blackwell.
Lewis, Diana M. 2003. “Online News: A News Genre?” In New Media Language, ed. by Jean
Aitchison, and Diana M. Lewis, 95–104. London and New York: Routledge.
282 News and Time
Locher, Miriam A., and Sage L. Graham. 2010. “Introduction to Interpersonal Pragmatics.” In
Interpersonal Pragmatics, ed. by Miriam A. Locher, and Sage L. Graham, 1–13. Berlin and
New York: De Gruyter Mouton.
Lyons, John. 1977. Semantics, Vols. 1 & 2. Cambridge: Cambridge University Press.
Marmaridou, Sophia. 2000. Pragmatic Meaning and Cognition. Amsterdam: John Benjamins.
DOI: 10.1075/pbns.72
Martin, James R. 2001. “Cohesion and Texture.” In The Handbook of Discourse Analysis, ed. by
Deborah Schiffrin, Deborah Tannen, and Heidi E. Hamilton, 35–53. Malden and Oxford:
Blackwell Publishing.
Martin, James R., and David Rose. 2007. Working with Discourse: Meaning Beyond the Clause,
2nd ed. London: Continuum.
Martin, James R., and Peter. R.R. White. 2005. The Language of Evaluation: Appraisal in English.
London and New York: Palgrave Macmillan.
McCarthy, Michael, and Ronald Carter. 1994. Language as Discourse. Perspectives for Language
Teaching. London and New York: Longman.
McIntyre, Dan. 2006. Point of View in Plays: A Cognitive Stylistic Approach. Amsterdam: John
Benjamins. DOI: 10.1075/lal.3
Mey, Jacob. 2001. Pragmatics. An Introduction, 2nd ed. Oxford and Malden, MA: Blackwell.
Molek-Kozakowska, Katarzyna. 2013. “Towards a pragma-linguistic framework for the study of
sensationalism in news headlines.” Discourse & Communication 7 (2): 173–197.
DOI: 10.1177/1750481312471668
O’Keeffe, Anne, Brian Clancy, and Svenja Adophs. 2011. Introducing Pragmatics in Use. London
and New York: Routledge.
Phillips, Angela. 2010. “Old Sources, New Bottles.” In New Media, Old News. Journalism & De-
mocracy in the Digital World, ed. by Natalie Fenton, 87–101. London: Sage.
DOI: 10.4135/9781446280010.n6
Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehen-
sive Grammar of the English Language. London and New York: Longman.
Rafferty, Eamonn. 2008. “Headlines.” In Pulling Newspapers Apart: Analysing Print Journalism,
ed. by Bob Franklin, 212–219. Abingdon and New York: Routledge.
Redden, Joanna, and Tamara Witschge. 2010. “A New News Order? Online News Content Ex-
amined.” In New Media, Old News. Journalism & Democracy in the Digital World, ed. by
Natalie Fenton, 171–186. London: Sage. DOI: 10.4135/9781446280010.n11
Richardson, John E. 2007. Analysing Newspapers: An Approach from Critical Discourse Analysis.
Basingstoke: Palgrave Macmillan.
Rooney, Edmund J., and Oliver R. Witte. 2000. Copy Editing for Professionals. Champaign, Il:
Stipes Publishing.
Russial, John. 2004. Strategic Copy Editing. New York: The Guildford Press.
Sanders, José, and Gisela Redeker. 1993. “Linguistic Perspective in Short Stories.” Poetics 22:
69–87. DOI: 10.1016/0304-422X(93)90021-8
Schaffer, Deborah. 1995. “Shocking Secrets Revealed! The Language of Tabloid Headlines.”
ETC: A Review of General Semantics 52 (1): 27–46.
Schiffrin, Deborah. 1981. “Tense Variation in Narrative.” Language 57 (1): 45–62.
DOI: 10.1353/lan.1981.0011
Schneider, Kristina. 2000. “Emergence and Development of Headlines in British Newspapers.”
In English Media Texts Past and Present: Language and Textual Structure, ed. by Friedrich
Ungerer, 45–65. Amsterdam: John Benjamins. DOI: 10.1075/pbns.80.05sch
References 283
Schokkenbroek, Christina. 1999. “News Stories: Structure, Time and Evaluation.” Time & Soci-
ety 8: 59–98. DOI: 10.1177/0961463X99008001004
Segal, Erwin M. 1995. “Narrative Comprehension and the Role of Deictic Shift Theory’.” In
Deixis in Narrative. A Cognitive Science Perspective, ed. by Judith F. Duchan, Gail A. Bruder,
and Lynne E. Hewitt, 3–17. Hillsdale, NJ: Lawrence Erlbaum Associates.
Semino, Elena. 1995. “Deixis and the Dynamics of Poetic Voice.” In New Essays in Deixis: Dis-
course, Narrative, Literature, ed. by Keith Green, 145–160. Amsterdam and Atlanta, GA:
Rodopi.
Simpson, Paul. 1993. Language, Ideology and Point of View. London and New York: Routledge.
DOI: 10.4324/9780203312612
Sperber, Dan, and Deirdre Wilson. 1986. Relevance: Communication and Cognition. Oxford:
Blackwell.
Sperber, Dan, and Deirdre Wilson. 1995. Relevance: Communication and Cognition, 2nd ed.
Oxford and Cambridge: Blackwell Publishers.
Straumann, Heinrich. 1935. Newspaper Headlines: A Study in Linguistic Method. London: Allen
& Unwin.
Studer, Patrick. 2003. “Textual Structures in Early Newspapers. A Corpus-based Analysis of
Headlines.” Journal of Historical Pragmatics 4 (1): 19–44. DOI: 10.1075/jhp.4.1.03stu
Studer, Patrick. 2005. “Text-initiating Strategies in Early Newspaper Headlines.” In Opening
Windows on Texts and Discourses of the Past, ed. by Janne Skaffari, Matti Peikola, Ruth
Carroll, Risto Hiltunen, and Brita Wårvik, 65–79. Amsterdam: John Benjamins.
DOI: 10.1075/pbns.134.08stu
Studer, Patrick. 2008. Historical Corpus Stylistics. Media, Technology and Change. London:
Continuum.
Taavitsainen, Irma, and Andreas H. Jucker. 2010. “Trends and Developments in Historical
Pragmatics.” In Historical Pragmatics, ed. by Andreas H. Jucker, and Irma Taavitsainen,
3–30. Berlin and New York: De Gruyter Mouton.
Talbot, Mary. 2007. Media Discourse: Representation and Interaction. Edinburgh: Edinburgh
University Press.
Tanskanen, Sanna-Kaisa. 2006. Collaborating Towards Coherence: Lexical Cohesion in English
Discourse. Amsterdam: John Benjamins. DOI: 10.1075/pbns.146
Thibault, Paul J., and Theo van Leeuwen. 1996. “Grammar, Society and the Speech Act: Renewing
the Connections.” Journal of Pragmatics 25: 561–585. DOI: 10.1016/0378-2166(95)00058-5
Thogmartin, Clyde. 1991. “The Pragmatics of French Newspaper Headlines.” In Levels of Lin-
guistic Adaptation, ed. by Jeff Verschueren, 249–266. Amsterdam: John Benjamins.
DOI: 10.1075/pbns.6.2.16tho
Thompson, Rick. 2010. Writing for Broadcast Journalists, 2nd ed. Abingdon and New York:
Routledge.
Ungerer, Friedrich (ed.). 2000. English Media Texts Past and Present. Language and Textual
Structure. Amsterdam: John Benjamins. DOI: 10.1075/pbns.80
Ungerer, Friedrich. 1999. “Newspapers as Text Ensembles.” In English via Various Media, ed. by
Hans-Jürgen Diller, and Gert Stratman, 17–37. Heidelberg: Winter.
Ungerer, Friedrich. 2000. “News Stories and News Events. A Changing Relationship.” In English
Media Texts Past and Present: Language and Textual Structure, ed. by Friedrich Ungerer,
177–195. Amsterdam: John Benjamins. DOI: 10.1075/pbns.80.12ung
Van Dijk, Teun A. 1988. News as Discourse. Hillsdale, Hove and London: Lawrence Erlbaum
Associates.
284 News and Time
Van Hout, Tom, and Geert Jacobs. 2008. “News Production Theory and Practice: Fieldwork
Notes on Power, Interaction and Agency.” Pragmatics 18 (1): 59–85.
Van Leeuwen, Theo. 1996. “The Representation of Social Actors.” In Texts and Practices:
Readings in Critical Discourse Analysis, ed. by Carmen Caldas-Coulthard, and Malcolm
Coulthard, 32–70. London: Routledge.
Van Leeuwen, Theo. 2009. “News Genres.” In Handbook of Communication in the Public Sphere,
ed. by Ruth Wodak, and Veronika Koller, 345–362. Berlin: Mouton de Gruyter.
Vandelanotte, Lieven. 2009. Speech and Thought Representation: A Cognitive-Functional Ap-
proach. Berlin: Mouton de Gruyter. DOI: 10.1515/9783110215373
Verschueren, Jef, Jan-Ola Östman, Jan Blommaert, and Chris Bulcaen (eds). 2003. Handbook of
Pragmatics. Amsterdam: John Benjamins.
Verschueren, Jef. 1999. Understanding Pragmatics. London: Arnold
von Stutterheim, Christane, Mary Carroll, and Wolfgang Klein. 2003. “Two Ways of Construing
Complex Temporal Structures.” In Deictic Conceptualisations of Space, Time and Person,
ed. by Friedrich Lenz, 97–133. Amsterdam: John Benjamins. DOI: 10.1075/pbns.112.07stu
Walker, Jim. 2008. “The Footballer’s Perfect – Are Footballers Leading the Way?” In The Lin-
guistics of Football, ed. by Eva Lavric, Gerhard Pisek, Andrew Skinner, and Wolfgang
Stadler, 295–303. Tübingen: Gunter Narr Verlag.
Weisser, Martin. 2013. “Pragmatics in the Analysis of Discourse and Interaction.” In The En-
cyclopedia of Applied Linguistics, ed. by Carol A. Chapelle, 4335–4541. Wiley Blackwell.
DOI: 10.1002/9781405198431.wbeal0938
White, Peter R.R. 2000. “Media Objectivity and the Rhetoric of News Story Structure.” In
Discourse and Community: Doing Functional Linguistics, ed. by Eija Ventola, 379–397.
Tübingen: Gunter Narr Verlag.
White, Peter R.R. 2003. “News as History: Your Daily Gossip.” In Re/reading the Past. Critical
and Functional Perspectives on Time and Value, ed. by John R. Martin, and Ruth Wodak,
61–89. Amsterdam: John Benjamins. DOI: 10.1075/dapsac.8.05whi
White, Peter R.R. 2006. “Evaluative Semantics and Ideological Positioning in Journalistic Dis-
course: A New Framework for Analysis.” In Mediating Ideology in Text and Image: Ten
Critical Studies, ed. by Inger Lassen, Jeanne Strunck, and Torben Vestergaard, 37–67.
Amsterdam: John Benjamins. DOI: 10.1075/dapsac.18.05whi
White, Peter R.R. 2011. “Appraisal.” In Discursive Pragmatics, ed. by Jan Zienkowski, Jan-Ola
Östman, and Jef Verschueren, 14–36. Amsterdam: John Benjamins.
White, Peter R.R. 2012. “Exploring the Axiological Workings of ‘Reporter Voice’ News Stories –
Attribution and Attitudinal Positioning.” Discourse, Context & Media 1: 57–67.
DOI: 10.1016/j.dcm.2012.10.004
Wolfson, Nessa. 1978. “A Feature of Performed Narrative: The Conversational Historical Pres-
ent.” Language in Society 7: 215–237. DOI: 10.1017/S0047404500005534
Wolfson, Nessa. 1979. “The Conversational Historical Present Alternation.” Language 55 (1):
168–182. DOI: 10.2307/412521
Wright, Susan. 1987. “Tense Meaning as Styles in Fictional Narrative: Present Tense Use in J.M.
Coetzee’s in the Heart of the Country.” Poetics 16: 53–73.
DOI: 10.1016/0304-422X(87)90036-2
Yule, George. 1996. Pragmatics. Oxford: Oxford University Press.
Yus, Francisco. 2011. Cyberpragmatics: Internet-mediated Communication in Context. Amster-
dam: John Benjamins. DOI: 10.1075/pbns.213
Index
A article cluster 10, 75, 83, 86, 98, background information 80,
absolute deictics 42 101, 129, 145, 174–177, 185, 264, 107, 133, 166, 169, 175, 185, 200,
absolute time 41, 73 265, 272 241, 255, 263
abstract 62, 65, 101, 130, 142, article preview 1, 7, 59, 61, 62, backgrounding 9, 132, 133, 143,
150, 209, 226, 228, 231, 247, 66, 71, 73–77, 108, 154, 205, 153, 158, 160, 165–167, 177,
248, 249, 260, 266, 267, 274 225–227, 230, 231, 233, 239, 274 185, 272
accompanying article 74, 80, article web page 8, 61–66, 77, background knowledge 99,
86, 175, 176, 265 78, 90, 105, 108, 109, 226–233, 140, 164, 169, 183, 192
active voice 180, 183, 186, 200, 236–238 backshift of tense 21, 38
242 as 170, 199 backtracking 79, 80, 82, 92, 275
actuality modality 48, 197 assumption of recency 43–45, block language 10, 179, 180,
addressee 19, 27, 31, 34–36 51, 54, 67, 100 188, 197, 202
adjectival premodification 9, in leads 105 body copy 37, 40, 61, 63, 89, 91,
107, 109 asynchronous communication 120, 137, 205, 206, 212, 220
adverbial of time 32, 33 breaking news 62, 76
see time adverbial atemporality 125, 174, 218 broadcast 35, 76, 124, 129, 130,
advertisement 88, 92, 180 atemporal organization 250 139, 141, 142, 263, 264
agency copy 57 attention-getting device 130 broadcast media 165
alignment 19, 20 attribution 1, 66, 68, 70, 77, 78, broadsheet 7, 76, 97
allusion 169 90, 104, 106, 159, 160, 171, 176, Bühler 29
ambiguity 141, 179, 182, 183, 191, 192, 226, 236, 237, 260, byline 1, 67, 78, 86, 252, 261
185–187, 189, 200 262, 275
anchor 211, 227, 241, 253 audience 1, 6, 7, 23, 46, 67, 88, C
anchorage 67, 68, 163 97, 164, 187, 250, 263, 265, 267, calendar time adverbials 56
absence of explicit 58 269, 274 calendrical expression 49, 50,
home page 71 audience participation 92 73, 74, 78
implicit 39 authenticity 172, 190, 271, 274 calendrical time expression 86
missing 46 auxiliary 115, 147, 151, 179, 182, capitalization 194
visual 101, 192 184, 187, 195, 198 caption 90, 102, 121, 134, 226,
anchorage point 30, 51, 61, 67, absence 188 231, 239, 240
68, 70, 86, 163 ellipsis 115, 190, 240 categorization 187
angle 235, 237 see also ellipsis causative meaning 186, 199
anteriority 109, 161–163, 166, modal 146 causativity 200
169, 178 non-ellipted 115, 147, 179, central article 175–177, 185
appraisal theory 19, 22 189, 190, 200, 239 chain
archived article 9, 80, 82, 83, presence 188, 193–195, 197 cohesive 137, 208, 211, 237,
90, 227 242
archived news 78, 82 B co-referential 207, 209, 210,
archived page 80, 90, 91 background 58, 98, 168, 170, 216, 233, 234, 238, 239,
article 260, 261 242–244, 270
history 69, 70 background event 176 process 210, 211, 216, 230,
satellite 9, 98, 129, 158, 234, 235, 237, 239, 241,
174–177, 272 244, 270
286 News and Time
chain interaction 241–243 colloquialization 157 continuous coverage 4

chat 32, 33, 165 colon structure 101, 119, 151, convention 7, 17, 22, 24, 26, 28,
chronological organization 265 159, 169, 170, 173, 195, 234, 66, 117, 118, 121, 122, 124,
chronology 23, 248, 250, 252, 236, 237, 249, 271 127, 132, 134, 136, 139, 141,
264, 266, 267 comment 154, 156, 172, 174, 182, 183,
circumstances 194, 198–200, metalinguistic 48, 133 195, 197, 198, 201, 202, 212,
224, 239, 241, 259, 260, 263 commentary 60, 106, 130, 157, 219, 220, 224, 239, 270,
circumstantial information 253, 261, 262 271, 272
106, 200 live 33 discontinuation 145, 158,
clause sport 128, 129 179, 190, 195
condensed 148, 163, 183, 194 subjective 230 genre 18, 20
finite 113, 114, 147 common knowledge 136 suspension 121, 152, 274
main 133, 166–168 communication violation 137, 145, 188, 250
non-finite 112–115, 147, 154, displaced 32–39, 131 co-presence 27, 32–34, 271
168, 198, 200 electronic 33 fictitious 130
object 104, 148 face-to-face 32, 34, 50, 248, physical 32
relative 107, 166–170, 194 250 temporal 9, 32, 33, 271, 273
reported 104, 116, 133, 151 synchronous 39 copula 111, 114, 119, 153, 154,
reporting 104, 113, 116, 117, synchronous online 33 180, 189, 196
133, 159, 195 written 39 co-reference 193, 207, 209, 237
subordinate 104, 113, 133, communicative dyad 136 co-referential chain 207–210,
166, 167, 170, 198, 200, 240 compactness 155, 180 216–218, 233, 234, 238–244
temporal 170, 199 complex sentence 115, 146, co-referentiality 170, 225, 243
clock time 49, 50 168, 252 cost-effectiveness 185
cluster see article cluster complicating action 248–251 co-temporality 34, 44, 73, 130,
news 129 complicating event 136 211, 223, 224
co-construction of narrative comprehension of news 49, creativity 118, 120, 229
252 253, 263 critical linguistics 27, 183, 249
coda 248, 250 compression cross-reference 84
coding time 33–38, 41, 43, time 52, 90 current relevance 3, 9, 11, 28,
46–49, 52–55, 68, 105, 109, 131, compression of temporal space 46, 47, 57, 59, 73, 83, 126, 156,
137, 138, 270 52 157, 161, 164, 197, 214–217, 220,
cognitive computer-mediated 222, 234, 237, 241, 273, 275
approach 247, 260, 262, 264 communication 32, 165 customization of news content
category 248, 250 conceptual 89, 91
displacement 29 category 262–265 cyclicity 247, 259
effort 229 frame 248, 260 cyclic organization 209
frame 31, 185 structure 176, 247, 262 cyclic presentation 3, 11, 23,
processing 185, 210 condensation 111, 119, 149, 168, 210, 226, 231, 244, 247, 258,
reinterpretation 210 183, 194, 200 259, 266, 268, 274
structure 101, 168 conjunction 170, 199, 207
coherence 18, 25, 58, 136, 205, consequence 207, 258, 260, 261, D
207, 208, 220, 230, 244, 266 263, 264, 268 dated news 49
cohesion 18, 25, 58, 137, 181, content time 34 dateline 51, 67, 68, 70, 71, 74,
207–209, 211, 217, 218, 220, context 75, 77
226, 230, 233, 234, 242 situational 16, 20, 184 date stamp 88, 90
cohesion analysis 207–209, social 16, 25 deadline 49, 54, 59, 67, 68, 70
211, 242 sociocultural 17, 30 decks see headline decks
cohesive temporal see shared decoding 4, 35, 36, 38, 216, 217
chain 137, 142, 208, 237, 242 temporal context decontextualization 16, 183,
harmony analysis 208 contextual effects 44, 54 227, 265
tie 207, 238 continued coverage 189, 196 default tense 123, 201, 225, 271
Index 287
deictic discourse space 27, 31, 39, 110, entextualization 229

anchorage 33–36, 39, 42, 44, 143, 213, 214, 220, 223 evaluation 19, 22, 123, 135, 142,
49, 51, 56, 67, 138 compression 153, 201 172, 190, 191, 195, 229, 235,
anchoring 33 discursive gap 121 248–250, 253, 260, 262,
centre 27, 29–31, 34, 36, 37, dislocation 263, 266
52, 79, 86, 105, 162 physical 33 external 135, 253
centre projection 4, 25, 46, temporal 33 internal 130, 132, 136, 253
52, 58, 68, 110, 132, 137, 138, displaced communication 32 evaluative device 253
212, 270 displaced context 33, 35, 36, event frame 11, 247, 262, 263,
centre shift 32 38, 39, 131 265
centre switch 35 displacement event time 37, 46, 47, 49, 55, 68,
non-simultaneity 36 cognitive 29 129, 137, 270
projection 8, 24, 34, 36, 37, spatial 33 evidentiality 131, 142, 191
51, 52, 273 temporal 32, 33, 38, 253 expectation 260, 263
reference dormant article 85 external news source 150, 262
negotiation of 27 double headline pattern 64 external voice see voice,
shift 22, 27, 48, 53, 252 double summary 207, 210, 215 external
shift theory 29 double tense shift 215, 238 extracted topic 218, 263–265
simultaneity 30, 34–38, 55 double tense shift pattern 205,
tense 4, 6, 102, 123, 125, 215, 216, 220–222, 224, 230, F
220, 221 231, 239, 274 Facebook 69
time 34–36 variation 212 face-to-face conversation 32
time adverbial 33, 42, 43, dramaticality 130, 132–136, 138 face-to-face narrative 251
48, 49, 51, 55, 75, 102, 142, facticity 182
275 E factuality 140
time expression 41 earlier today 105 fictional narrator 252
deixis 26 early news discourse 75, 272 fictitious co-presence 130
hearer-centred 30 echoic utterance 159, 181 finiteness 179
personal 32 economy of expression 154 first-person narrative 253
speaker-centred 30 editorial fluidity of online news 68
temporal 34 content 89 focalization 29, 30
delayed lead 206 office 67, 68, 227 foregrounding 132–134, 136,
detachment 172, 191 practice 7, 182, 228 165, 167, 195, 253
determiner 119, 157, 173 egocentricity 30, 31 fragmentation 61
developing story 141 electronic media 264 of body copy 88, 90
diachronic analysis 80, 83–86, ellipsis 111, 114, 115, 119, 124, of news text 91
139, 196, 262 146, 148, 154, 168, 180, 183, of online content 227
dialogic 19, 20, 157 188, 190, 194, 208, 239, 240 frame of reference 35, 39
dialogicality 21, 24, 269 email 33, 69 frame of temporal reference
dialogue 21, 24 emerging news 76 28, 39, 42, 43, 53–55, 61, 69,
diegesis 135 encoding 4, 20, 22, 32, 35, 36, 71, 75, 255
direct speech 121, 151, 152, 159, 38, 110, 146, 161, 164, 191, 197, framing article 75
171, 172, 184, 189–191, 193, 201 198, 216, 238, 251, 269, 271, 275 free direct speech 171, 176
disalignment 19, 20 encoding time 163 front page 47, 61, 62, 234
disambiguation 123, 182, 183, engagement 122, 230 front page preview 62
186, 197 enhancement 149, 171, 225 function
discontinuity 266 enhancement of news value 162 interpersonal 136
discourse organization 25 enhancement of recency 143, functionalism 17
discourse representation 21, 272 futurity 104, 110, 140, 145, 146,
29, 38, 253 enhancement of relevance 220 151, 153
288 News and Time
G semantic compactness 155 I

garden-path effect 141, 183 semantic function 122, 139, ideal reader 19
gatekeeping 92 144, 149, 188 ideational function 18, 26,
genre 19, 91 structural complexity 160 220, 223
gossip 266 summarizing 206, 249, ideology 249
grammar of little texts 121 261, 267 illusion of a shared time 1, 130,
grammatical metaphor 208 syntax 107, 119 137, 270, 271
tautological 7, 170, 173 illusion of authenticity 156,
H headlinese 7, 10, 64, 66, 100, 190, 274
habitual present 127, 128, 142 107, 115, 118, 122, 124, 127, 141, image 92
headline 61, 66, 88, 99, 118, 145, 146, 152, 154, 156, 172, 173, image caption 231, 239, 240
121, 130 177, 179, 198, 201, 222, 224, immediacy 73, 99, 122, 132,
19th century 139, 140, 262 234, 271, 272, 274 136–138, 142, 152, 172, 201,
broadcast 124 hedging 19, 119, 120, 144, 236, 271
canonical 124 149–151, 191, 262 imperative 117
clausal 101, 110, 112–114, heterogeneity 7, 21 impersonality 10, 190, 270
157, 179 heteroglossia 4–6, 20–22, 24, indirect speech 121, 151, 152,
conventions 39, 109 116, 121, 133, 145, 147, 150–152, 159, 171, 189, 193, 195, 201
decks 140, 188, 226, 272 156, 158, 167, 171, 173, 178, 189, infinitive 146–149, 152, 180
dialogic structure 157 190, 192, 193, 195, 202, 218, information
dual 157, 159, 169, 201, 249 252, 269, 271, 274 known 169, 209, 210, 243
dual structure 151, 170 hierarchical organization 272 new 169, 209, 210, 220, 232
equational 151 hierarchical structure 207 statistical 103
fictional 183 historical news discourse information density 120
finite 110, 179 analysis 139 information flow 9, 158, 165,
formulation 161 historic present 9, 102, 131, 132, 166, 168–171, 177, 187
function 99 134–136, 139, 165 information value 179, 180,
heteroglossic 137, 172, 189, history 60, 68–70, 79, 260, 261, 188, 197
192, 249, 270, 274 263, 264, 268 instantaneous present 127–130,
hierarchy 98, 175, 177 home page 7, 61–63, 65–68, 71, 137, 138
home page 105, 226, 228, 73–75, 77, 78, 105, 108, 109, interaction 16, 17, 23, 27, 30, 31,
229, 236 154, 205, 225, 227, 228, 230, 33, 92, 135
identical 190, 227 232, 236, 238, 274 patterns 15
minimal 101 homogenization of news content textual site of 23
mini-narrative 199 59 interactiveness 23, 28, 182
monoglossic 160, 171, 173, homonymy 120 interactivity 23, 188
176, 187, 189, 196 hook 235 interlocutor 32, 33
monologic 160 hot news tense 131, 157, 214, internal evaluation device 130,
narrative form 156, 157 220, 223, 234 132
nominal 110–112, 148 hub article 84, 85 interpersonal
non-finite 110, 124, 179, 239 humour 99, 157, 183 function 18, 19, 26, 27, 136,
online 105 hyperlink 60–62, 79, 80, 82, 219, 220, 223, 236
popular press 101, 120 84–86, 88, 90, 226, 227, 236, intimacy 136
pragmatic function 122, 275 intersubjective meaning 269
144, 149, 188 hypermedia 82 intersubjectivity 18, 19, 27
processing 187 hypertext 60, 61, 80, 82, 91, 92 intertextuality 18, 92, 121, 140,
question form 198, 229 hypertextuality 1, 4, 8, 18, 171, 185
quotation 206 59–61, 82, 89, 92, 226, 265 inverted commas 150, 159, 167,
rhetoric 99, 102, 136, 144, hypertextual network 79, 80, 171, 172, 184, 190–192, 271
149, 150, 167, 181 82, 83, 85, 91, 275
Index 289
inverted pyramid 3, 5, 11, 25, 39, live modality 48, 120, 143–145, 149,
205, 210, 244, 248, 262, 266 broadcast 32 181, 188, 193
involvement 55, 157, 188, 220, content 8, 80, 82, 90, 91, 131 deontic 150
229, 249 coverage 8, 129 epistemic 143, 149, 182, 196
news 76, 78, 129, 265 modal verb 110, 149, 150, 156,
J online text commentary 32, 189, 191, 193, 195, 220
joke 132 33 mode 18, 21
joke telling 132, 136 reporting 68, 264 spoken 23, 32
journalism 7, 143, 252 text reporting 32, 52 written 32
modern 59 ticker 62, 76 monosyllabism 120, 121, 152,
online 32, 52, 59, 88 local coherence 205 181, 234
print 49, 59 location 9, 29, 32, 42, 106–109 multimodal anchorage 192, 271
journalistic norms 7 multimodality 60, 89, 90, 92,
M 93, 101, 120, 182
K macrostructure 99, 130, 139, multimodality in news 20,
key article 84, 85 205, 207, 226, 244 90, 101
main event 1, 65, 155, 157, 165, multiple summary pattern 65,
L 167, 168, 176, 185, 198, 199, 226, 232, 268
landing page 63 205, 206, 210, 213, 218, 225,
latest news 73, 76, 78, 165 230–232, 234, 247, 250, 254, N
lead 2, 28, 37, 39, 61, 63, 66, 101, 255, 258–260, 262–265 narration 21, 29, 130, 134,
104, 110, 120, 137, 142, 157, marked utterance 118 214–216, 225, 250
205, 206, 212, 220, 267 masthead 48, 51, 71, 73, 75, narrative 123, 131, 136, 211, 248,
online 105 100, 275 250–253, 267, 274
lead rhetoric 225 maximal relevance 235 face-to-face 248, 251, 252
lexical maxim of quantity 50, 55 fictional 134, 140, 252
ambiguity 183 meaning first-person 29, 253
choice 190 construction of 17 personal 134, 248–250,
density 107, 169 intersubjective 19 252, 266
features 120 meaning potential 17 space 136
transformation 65, 244, 273 mental space 31, 136 third-person 29, 253
lexico-grammar 17–19, 25, 117 meta-attribution 262 narrative category 136, 249, 253
linearity 83, 233, 242, 249, 250, metadata 68–70, 75, 78 narrative element 248, 250, 260
256, 266, 273 metafunction 6, 17, 97, 166, narrative structure 65, 228,
linear progression 5, 137, 250 205, 218, 219, 223 247, 248, 250, 252, 253, 264
link metaphor narratology 20, 29, 135, 248
bi-directional (two-way) 82 conduit 23 narrator 29, 38, 135, 252, 253
diachronic 82, 83, 85 metaphorical use 129–131, 157 negative prosody 186
external 60, 88, 90, 92, 265 metonymy 107 negotiation
hypertextual 60, 69, 92, 225, micropragmatics 15 of spatial context 33
228, 274 mimesis 135 network see hypertextual
internal 60, 62, 82, 83, 86, mimetic representation 267 network
88, 89–91 minimal information chunk new information 68, 169, 181,
intertextual 60, 61, 83 140 209, 210, 220, 226, 232
intratextual 60 minimalist data chunk 60, 62 news
synchronic 82, 83, 85 minute-by-minute reporting 24-hour 59, 105
uni-directional (one-way) 76 archived 82
82–84 mitigation 19 bulletin 165
literary studies 21, 29, 135 modal auxiliary cycle 44, 165
literary texts 20, 38, 134, 139, avoidance 146 daily 83
140, 252 edition 48, 88, 227
290 News and Time
hard 218 version 68–70, 75, 88, 91, non-simultaneity 33, 36

hot 44, 131, 157, 214, 217, 223 143, 264, 275 non-textual element 60, 62, 82
live 76, 78, 129, 265 newsbite 61, 74 non-validity 174, 177, 192
macro-structure 62, 63, newsdesk 88, 165 normativity 118
71, 77 newspaper notional component 205, 206,
non-recent 76, 80, 161 archive 1, 63, 76, 78, 80, 86, 219, 220, 223
old 43, 46 89, 91, 165, 227, 236, 275 nucleus 176
online 59, 60, 79, 80, 83, daily 54 nucleus-satellite structure 176
88, 105, 108, 109, 164, 205, electronic 51 number 252
221, 223–227, 231, 234, 245, historical 75 nut graf 206
269, 274 online 59, 68, 70, 75, 129,
popular 131, 264 164, 227, 230, 275 O
print 57, 61, 83, 105 popular 76, 162, 264 objective time 75, 78, 79, 121
printed 68, 130 print 70, 80, 163, 164, 263, objectivity 143, 182, 190
recent 76 264 online
research article 46, 131 printed 51, 67 article 60, 67, 68, 85
rolling 59, 264, 275 section 198 chat 32, 165
second-day 55 weekly 44, 52 content 227, 228
short 30, 217, 252 newsworthiness 9, 46, 76, 103, edition 88
soft 98, 206 106, 107, 131, 142, 153, 160–162, forum 89
stale 43, 54, 76, 86, 143, 165–168, 170, 176, 182, 194, journalism 32, 52, 59, 88
264, 272 199, 206, 209, 239, 241, 252, media 8, 59, 68, 80, 86, 90,
traditional 137, 262 260, 264, 272 164, 165, 226, 264
news newsworthy content 92 news 83
actor 101, 108, 117, 119, 142, newsworthy event 37, 40 news site 61, 70, 98
173, 176, 190, 194, 236, nominalization 11, 111, 120, 174, news stories 64
242, 253 183, 192, 200, 208–211, 218, publication 60, 61, 74
agency 164, 261, 262 230, 237, 238, 256, 269 version 40
cluster 1, 9, 60, 74, 75, 80, nominal phrase 107, 112, 113, online news 4, 50, 51, 70, 76,
82, 83, 92, 98, 174–176, 264 119, 120, 148, 163, 166, 169, 91, 92
comprehension 49, 253, 263 183, 193 addictiveness of 70
cycle 44, 105 nominal premodification 256 optimal relevance 54, 55, 185
flow 68, 76 nomination 187 orbit 267, 268
package 63, 175, 176 non-calendrical time adverbial orbital organization 247, 260,
processing 253, 263, 267, 274 26, 43, 45 266, 267
production process 48, 57, non-chronological structure of orbital structure 11, 267
70, 83, 137, 138, 252, 262, news 4 ordinal numeral 103
270, 275 non-chronology 23, 25, 58, 99, orientation 248, 249
site 4, 61, 70, 71, 76, 79, 82, 162, 206, 210, 247, 250, 253– orientation time 34
85, 98 256, 258, 260, 262, 267, 268 origo 29, 30
structure 23 non-cohesiveness 212, 213
text non-deictic tense 4, 6, 102, P
construction of 18 125, 221 package approach 60, 175, 176,
formulation 5 non-deictic time expression 41 264, 265
linear progression 5, non-factive presupposition 173, passive
137, 250 174 construction 10, 119, 168,
update 4, 59, 61, 69, 70, 73, non-linearity 210, 247, 252, 266 194, 243
76, 78, 90, 141, 164, 165, non-recency 9, 86, 145, 158, transformation 148, 186,
196, 264 160–162, 165, 166, 168, 169, 209, 218
value 1, 57, 99, 131, 148, 162, 177, 178, 224, 225, 239, 272 voice 180, 183– 187, 195, 200,
194, 209, 248–250, 267 non-sequentiality 263 208, 242
Index 291
passivization 244 premodification 107, 108, 120, publication frequency 47, 53,
past issue 83, 84, 275 121, 168, 256 54, 75, 83, 131, 138, 161, 163,
past participle 148, 149, 162, adjectival 107, 109 164, 225, 264, 272
183, 184, 186, 187, 194, 231 pre-present zone 124, 156, 164 publication time 46, 47, 49, 51,
past tense 45, 161 pre-recorded programme 32, 55, 74, 163, 270
defocusing function 160, 35
167, 168 present perfect 57, 110, 154, 214, Q
deictic 233 220, 222, 234, 275 quotation 29, 55, 119, 159, 171,
in headlines 224 framing use 215 189–191, 193, 255, 271
semantic function 167 hot news 157 quote 156, 206, 255
patterns quasi-narrative use 157
of interaction 17 semi-deictic 212, 220, 273 R
pendant article 84, 85 present tense 124, 157 radio 139, 263, 264
periodical 53 modal certainty 182 radio phone-ins 32
periodical press 43, 44, 52, 138 non-deictic 181, 212 reader 143, 149, 226, 229, 247,
periodical publication 43, 47, universal reference 126 259, 263, 273
52, 75, 164, 272 presumption of recency 9, 54, reader participation 88
periodicity 54 73, 86, 99, 131 reading path 63, 80, 83, 88,
personalization 229, 230 preview 217 120, 275
personal narrative 134, 248– see article preview reality status 143, 149, 160, 194
250, 252 previous event 260, 263 real life 256, 259
personal opinion 190 primary tense 35, 109, 125 real time 32, 39, 68, 129, 142,
persuasive discourse 31 printed newspaper 41, 44, 61, 241
photo gallery 90 68, 80, 83, 263 re-anchorage 86
photograph 90, 92, 101, 134, printing time 48 receiving time 26, 34–39, 41,
182, 226 print journalism 49, 59 43, 44, 46, 47, 49, 51, 53, 55,
photograph caption 102, 121, print media 51, 54, 165, 263, 57, 68, 79, 105, 109, 129, 137,
134 264 162, 163, 259, 270
point of entry 62, 175, 181, 226, print version 88 anticiated 37
227, 263 prior knowledge 183, 185, 192, recency 110, 122, 126, 143, 147,
point of view 252 239, 263, 264 163, 164, 165, 181, 263, 275
political discourse 31 process chain 205, 206, 211 reception 4, 32, 33, 45, 47, 48,
polyphony 21 processing 210, 212, 227, 229, 54, 55, 70, 78, 127, 201, 212,
popular press 98, 101, 120 233 223, 224, 267, 273
positioning 6, 18, 19, 23, 24, 31, processing effort 44, 54 anticipated 5, 9, 27, 38, 68,
36, 120, 191, 206, 265 producer 17, 23, 36, 52 71, 137
dialogistic 20 production 33 assumed 45
ideological 191 of news texts 17 asynchronous 33
possessive construction 101 production deadline 49, 54, 59, non-simultaneous 33
pragmatic function 265, 269, 67, 68, 70 of news texts 17
275 production schedule 51 recipient 17–19, 23, 36, 42–44,
pragmatics projection 36 69, 206, 275
European 15 pronominalization 192 recycled news 46
interactional 19, 23 proximity 26, 214 redundancy 100, 101, 196, 228
interpersonal 16 proximization 31 reformulation 46, 166, 173, 174,
socio-interactional 15, 18 pseudo-quote 159, 171, 172 193, 209, 218, 226, 228, 232,
prediction 9, 104, 146, 152, 200 publication cycle 44, 83 236, 244
pre-emptiveness 26, 50, 55 publication date 28, 43, 48, 49, register 17, 18, 117, 118, 121,
pre-emptiveness of time 52, 54, 61, 73, 137, 163, 227 129, 136
adverbials 42, 48, 50 relative deictics 42
292 News and Time
relevance 57, 122, 134, 143, 149, temporal context 33, 38, 49, story-telling 135
185, 214, 227, 235 54, 68, 143, 147, 220 strapline 275
relevance theory 44, 54, 122 time frame 41, 69 structure
repetition 209, 217, 218, 226, virtual space 33 cognitive 101, 168
228, 232, 234, 235, 244, 268 shelf life 43, 47, 51, 52, 100 dialogic 157
lexical 219 shift in perspective 132, 135 grammatical 101, 110, 177
repetitiveness shift of point of view 22, 36 narrative 135
structural 109, 210 shift of temporal perspective style 23
reported discourse 20, 21 128 as a marker of identity 7
reported speech 38, 113, 171, shift of tense 26, 37, 99, 110, condensed 100
266, 274 119, 127, 129, 137, 153, 181, 201 conventional 184
reporting verb 132, 133, 151, 237 sidebar 176 conversational 6, 23, 172,
resolution 248, 249, 250 simultaneity 126, 128, 129, 272
rhematic information 30, 169, 199, 257 detached 121
170 see also deictic simultaneity dramatic 130, 133, 134
rhetoric 31 situation time 34, 124, 156 familiar 229
rhetoric of numbers 252 social headline 9, 118, 121, 147, 180,
rolling news 59, 105, 224, 275 actor 106, 168, 170, 183, 222, 234, 270
243, 244 impersonal 270
S networking site 69 interactive 6, 28
satellite article 9, 98, 129, 158, semiotic 17, 18 neutral 181
174–177, 272 socio-cultural context 187 newspaper 7
scare quotes 159, 172, 191 sociolinguistics 16 of oral narrative 102
schema 58 sociopragmatics 16, 17 of reading 44
script 58, 254 source 261, 262 online news 60
second-day lead 55, 206 spatio-temporal setting 32, 249 personal 6, 98, 197, 198
semantic speech-act verb 133, 195 telegraphic 100, 118, 121, 180
agent 148, 183, 184, 186, speech representation 21, 29, stylistics 20, 21, 135
208, 243 121, 151, 173, 184, 271, 274 subevent 165, 168, 171, 247, 253,
compactness 155 split context 32, 33, 35 258, 259
correspondence 5, 230 spoken subhead 3, 61, 63–66, 98, 171,
incompleteness 192 broadcast 32, 129, 130 220, 222, 225, 231, 234, 237,
incongruity 194 broadcast media 264 239, 241
patient 183, 184, 186 communication 32, 39 subjective time 79
role 170, 183, 184, 186, 242 narrative 134, 214, 215 subjectivity 150, 230, 249, 265
semi-deictic tense 6, 212, 220, sponsored content 89 subordinator 166, 199–201
272 sports 129, 197, 214, 264 summarization 99, 103, 130,
semi-modal ‘be to’ 125, 141, announcer talk 129 139, 142, 206, 209, 226, 228,
147, 180 commentary 128, 129 230, 232, 236, 247, 258
sentence pattern 101, 111 reporting 263, 264 summary 63, 65, 99, 100, 116,
service button 69 stale news 43, 54, 76, 86, 143, 139, 140, 182, 205, 206, 226,
setting 32, 101, 106, 261 272 232, 244, 259, 274
shared stance 19, 172, 191 superlative 103, 256
context 5, 25, 27, 34, 40, 42, epistemic 19 superstructure 100, 211
49, 56, 99, 101, 130 stand-alone unit 61, 80, 92, 98, switch of tense 5, 135, 136
discourse space 1, 223, 269, 121, 142, 175, 181, 192, 196 synchronous online chat 32, 33
275 state present 126–128 syntactic
frame of reference 39 story 260 ambiguity 183
frame of temporal reference line 259, 262 complexity 111, 112, 182,
28, 43, 53, 55 time 136 194, 252
space 33 update 169 ordering 231, 242
Index 293
structure 101, 106, 183, 232, text colony 18, 175 today 40, 50, 52, 55, 68, 105,
249 text messaging 33 143, 221, 223
subordination 104, 107 textual tomorrow 42, 49, 50, 52, 53, 55
transformation 209, 218, absence 208, 234, 244 tonight 40, 41, 141
234, 273 chunk 266, 267 top-down organization 23, 130,
synthetic personalization 229 function 11, 18, 26, 220, 223 248, 263, 264, 267
systemic-functional grammar lifting 59, 228, 229, 230 top-down structure 39, 60,
6, 18, 265 non-permanence 226, 227 247, 262
rhetoric 244, 270 topicality 57, 147, 264
T transformation 62 transformation 11, 108, 155, 174,
tabloid 76 textuality 18, 25, 92, 205, 244 205, 208, 210, 217, 234, 238,
tabloid rhetoric 162 texture 218, 220 244, 270
telegraph 75 theatricality 129, 130 transience
telephone conversation 32 thematic information 30, 169, of article preview 63, 226,
temporal 170, 193, 194 227
anchorage 1, 4, 5, 59, 60, 68, thematic structure 247, 260, transitivity 184, 231
71, 74, 75, 77, 78, 86, 88, 261 trigger 227, 228
103, 109, 110, 142, 143, 162, third-person narrative 29, 253 triple tense pattern 97, 218, 231,
178, 181, 233, 258, 269, 273 this morning 221, 223 234, 239, 241, 242, 273, 274
axis 39 thumbnail image 61, 88 turn taking 30, 32
co-presence 9, 32, 33, 271, ticker 62, 76 Twitter 69
273 live 76 typesetting 198
currency of events 163 time typography 64, 76, 88, 120, 181,
displacement 32, 33, 131, 253 calendrical 41, 49, 68, 71, 189, 193
mapping 79, 80, 82, 92 79, 86
metacomment 48 compression 1, 52, 90, 153 U
simultaneity 48 time adverbial 5, 39, 99, 134, unambiguousness 193
temporality 142, 162, 223 universal validity 271
functional model 219 absence 104, 138, 221, 270 update see news update
tense absence of 57, 100 updateability of news 76
atemporal 9, 123, 125 deictic 33, 49, 58, 223 utterance-pragmatics 15
default 123, 125, 126, 145, non-calendrical 28, 43, 45,
158, 201, 271 49, 58 V
deictic 125, 135, 142, 145, non-deictic 49, 50 variability 6, 109, 201, 275
221, 269 time axis 46, 51, 73, 84, 86, 260 verbal
non-cohesive 211 time caption 74 comment 132, 142, 176, 264
non-deictic 102, 109, 123, time frame 38, 49, 51, 53, 161, context 171, 182, 183, 185,
125, 130, 139, 142, 221, 236, 224, 272 187, 236
269, 271 timelessness 112, 125, 127, 131, reaction 156, 176, 207, 260,
non-shifted 156 134, 142, 241, 271 263, 265, 271
pragmatic function 123, 145, time line 34, 35, 56, 73, 77, 92, verbatim repetition 159, 171,
170, 178, 213, 220, 244 163, 254, 256, 258, 259, 266 184, 190, 209, 226, 228–230,
pragmatic properties 110 time of orientation 34–36 232, 274
semantic function 145, 170 time of publication 73, 75, verbo-nominal phrase 111,
semantic properties 109, 77, 78 235, 237
123, 125 time of situation 34, 47, 162 verbo-visual 20, 88, 120
tense selection time of utterance 34, 35, 37, 44, verb phrase 3, 5, 10, 114, 115,
horizontal 137 47, 125, 162 180, 182, 183, 189, 210, 218,
vertical 137 time-shifted programme 131 220, 234, 275
tense shift 5, 66, 212, 223, 224, time stamp 73, 86, 275 analytical 154
233, 234, 241, 269, 274 time structure 250, 253 transformation 234
294 News and Time
verbs of communication 131, 177, 179, 187, 189, 191, 199, W

132, 133, 141 218, 249, 252, 253, 262, weather forecast 152
version 270, 271 web page 61, 64, 65, 71, 73, 77,
online 40, 51, 80, 165, 264 attribution 20, 22 79, 165
print 52, 88, 137, 164 authorial 21, 22, 116, see also article web page
video 60, 90, 92 150–152, 159, 168, 171, 173, word play 99, 101, 120, 157
video conference 32 176, 189–191, 195, 196, written
viewpoint 30, 248 270, 271 communication 32
visual 78, 88, 90, 101, 120, 181, external 6, 20, 21, 116, 133, media 1, 32, 264
192, 198, 226, 271 150–152, 156–159, 172, 173,
indication of identity 171 184, 189, 191, 192, 271, 274 Y
resolution of ambiguity 182 institutional 20, 159, 171 yesterday 33, 40–42, 45, 49–52,
visual anchorage 101, 181, 192 internal 198 54, 55, 86, 105, 137, 142, 222,
visual grammar 120 juxtaposition 22 223, 225, 258, 275
vividness 9, 129–134, 152, 271 monoglossic 236
voice 20, 21 multiplicity of 21, 22 Z
accessed 21, 22, 144, 145, non-authorial 6, 152, 156, zigzagging pattern 259, 260,
158, 160, 164, 167, 171–173, 159 268

Chovanec. Pragmatics of Tense and Time in News

Uploaded by

Document Information

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

Chovanec. Pragmatics of Tense and Time in News

Uploaded by

Copyright:

Available Formats

Pragmatics of Tense and Time in News

Pragmatics & Beyond New Series (P&BNS)

Editor Associate Editor

John Benjamins Publishing Company

the American National Standard for Information Sciences – Permanence

Library of Congress Cataloging-in-Publication Data

© 2014 – John Benjamins B.V.

Part I. Temporal deixis in print and online news

Part II. Textual rhetoric of headlines

6.6 Potential ambiguity of the simple present tense in headlines 141

Part III. Textual rhetoric of news texts

9.3 Information chaining 208

 Brno, June 2014

1.1 Goals and objectives

One of the most striking and noticeable linguistic phenomena characterizing

A powerful 8.2-magnitude earthquake hit Chile’s Pacific coast on Tuesday,

1.4 Overview of chapters

appended to archived news texts and is frequently personalized to specific readers.

information within clusters of articles on related topics, and elaborates on the

complicated (e.g. on account of the presence of subheads, the existence of nomi-

Temporal deixis in print and online news

2.2 Functionalism and Halliday’s metafunctions

Although the present work is pragmatically-oriented, it is also influenced by

2.4 News discourse analysis

by processing and inferring information (ideational dimension) and by being

Temporal deixis and news discourse

The pragmatic category of deixis significantly contributes towards the construc-

3.1 Deixis and interaction

Deixis is a means of connecting language, the communicative situation, the ex-

reference […] in possible contexts to consideration of the cognitive methods that

is achieved in a number of ways. In the present account, we will pay attention to

its successful interpretation on the presupposition of a temporal context shared

3.2 Deictic centre

From a cognitive pragmatic perspective, Marmaridou (2000) defines a deictic

If we adopt a general pragmatic perspective on the elements that are related

3.3 Deictic projection

1. correspond to the encoder (speaker/writer); or

You are now leaving West Berlin. [a written notice]

Noting that the non-projected time of orientation is the default interpretation in

What these examples illustrate is the well-documented semantic-pragmatic no-

3.4 Deictic projection in news texts

Simultaneous communication Deictic centre overlap CT = RT

Displaced communication a) No projection CT ≠ RT

1. by means of an explicit mention of a definite calendrical point that serves to

3.5 Time adverbials and shared temporal context

The successful interpretation of all the underlined expressions is possible thanks

3.6 Deictic and non-deictic time expressions

3.7 Time expressions in news texts

moment of utterance – “could, under the circumstances described…, be indexed

3.8 Deictic centres in print newspapers

News News News

3.9 Pre-emptiveness of deictic time adverbials

Let us now propose a model of deictic projection as it operates in periodic news

It is the deictic expressions that tend to be preferred, at least in most face-to-

The following example illustrates a similar situation of pre-emptiveness, with the

The phrase on Thursday appeared in a lead to a news article published, inciden-

3.10 Modelling deictic projection in news texts

(b) from RT point of view yesterday today

Figure 3.4 Temporal reference of the time adverbial yesterday

time directly derive from the medium’s periodicity (day/week/month).6 While a

As a result, there may be no explicit temporal indicator anchoring the original

3.11 Temporal deixis and tenses

Temporal deixis in online newspapers

The discursive management of temporality in traditional print journalism dif-

Brno, June 2014