Professional Documents
Culture Documents
PDF A Multidisciplinary Framework of Information Propagation Online Susannah B F Paletz Ebook Full Chapter
PDF A Multidisciplinary Framework of Information Propagation Online Susannah B F Paletz Ebook Full Chapter
https://textbookfull.com/product/multidisciplinary-management-of-
gastroesophageal-reflux-disease-sebastian-f-schoppmann/
https://textbookfull.com/product/sarcoma-a-practical-guide-to-
multidisciplinary-management-peter-f-m-choong/
https://textbookfull.com/product/information-asymmetry-in-online-
advertising-1st-edition-wiktor/
https://textbookfull.com/product/the-information-literacy-
framework-case-studies-of-successful-implementation-association-
for-library-and-information-science-education-heidi-julien/
1001 Ways to Stay Young Naturally Susannah Marriott
https://textbookfull.com/product/1001-ways-to-stay-young-
naturally-susannah-marriott/
https://textbookfull.com/product/information-governance-for-
healthcare-professionals-a-practical-approach-1st-edition-robert-
f-smallwood-author/
https://textbookfull.com/product/sound-propagation-through-the-
stochastic-ocean-1st-edition-john-a-colosi/
https://textbookfull.com/product/ethics-of-digital-well-being-a-
multidisciplinary-approach-christopher-burr/
https://textbookfull.com/product/digital-detectives-solving-
information-dilemmas-in-an-online-world-1st-edition-fulton/
SPRINGER BRIEFS IN COMPLEXIT Y
Susannah B. F. Paletz
Brooke E. Auxier
Ewa M. Golonka
A Multidisciplinary
Framework of
Information
Propagation
Online
123
SpringerBriefs in Complexity
Series Editors:
Henry D. I. Abarbanel, University of California, Institute for Nonlinear Science,
La Jolla, CA, USA
Dan Braha, New England Complex Systems Institute,
University of Massachusetts, North Dartmouth, MA, USA
Péter Érdi, Department of Physics, Center for Complex Systems Studies,
Kalamazoo College, Kalamazoo, MI, USA
Karl J Friston, University College London, Institute of Cognitive Neuroscience,
London, UK
Hermann Haken, University of Stuttgart, Center of Synergetics,
Stuttgart, Germany
Viktor Jirsa, Université de la Méditerranée, Centre National de la Recherche
Scientifique (CNRS), Marseille, France
Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences,
Warsaw, Poland
Kunihiko Kaneko, Research Center for Complex Systems Biology,
The University of Tokyo, Tokyo, Japan
Scott Kelso, Florida Atlantic University, Center for Complex Systems and Brain
Sciences, Boca Raton, FL, USA
Markus Kirkilionis, Mathematics Institute and Centre for Complex Systems,
University of Warwick, Coventry, UK
Jürgen Kurths, University of Potsdam, Nonlinear Dynamics Group,
Potsdam, Brandenburg, Germany
Ronaldo Menezes, Department of Computer Science, University of Exeter,
Exeter, UK
Andrzej Nowak, Department of Psychology, Warsaw University,
Warszawa, Poland
Hassan Qudrat-Ullah, King Fahd University of Petroleum and Minerals,
Dhahran, Saudi Arabia
Peter Schuster, University of Vienna, Vienna, Austria
Frank Schweitzer, ETH Zurich, System Design, Zürich, Switzerland
Didier Sornette, ETH Zurich, Entrepreneurial Risk, Zürich, Switzerland
Stefan Thurner, Section for Science of Complex System,
Medical University of Vienna, Vienna, Austria
Linda Reichl, University of Texas, Center for Complex Quantum Systems,
Austin, TX, USA
SpringerBriefs in Complexity are a series of slim high-quality publications
encompassing the entire spectrum of complex systems science and technology.
Featuring compact volumes of 50 to 125 pages (approximately 20,000–45,000
words), Briefs are shorter than a conventional book but longer than a journal article.
Thus Briefs serve as timely, concise tools for students, researchers, and professionals.
Typical texts for publication might include:
• A snapshot review of the current state of a hot or emerging field
• A concise introduction to core concepts that students must understand in order to
make independent contributions
• An extended research report giving more details and discussion than is possible
in a conventional journal article,
• A manual describing underlying principles and best practices for an experimen-
tal or computational technique
• An essay exploring new ideas broader topics such as science and society
Briefs allow authors to present their ideas and readers to absorb them with
minimal time investment. Briefs are published as part of Springer’s eBook collection,
with millions of users worldwide. In addition, Briefs are available, just like books,
for individual print and electronic purchase. Briefs are characterized by fast, global
electronic dissemination, straightforward publishing agreements, easy-to-use
manuscript preparation and formatting guidelines, and expedited production
schedules. We aim for publication 8-12 weeks after acceptance.
SpringerBriefs in Complexity are an integral part of the Springer Complexity
publishing program. Proposals should be sent to the responsible Springer editors or
to a member of the Springer Complexity editorial and program advisory board
(springer.com/complexity).
A Multidisciplinary
Framework of Information
Propagation Online
Susannah B. F. Paletz Brooke E. Auxier
Center for Advanced Study of Language Philip Merrill College of Journalism
University of Maryland University of Maryland
College Park, MD, USA College Park, MD, USA
Ewa M. Golonka
Center for Advanced Study of Language
University of Maryland
College Park, MD, USA
This Springer imprint is published by the registered company Springer Nature Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To David Leon Paletz, who has conducted
multidisciplinary research in this area for
decades and who inspires us all to cross
boundaries with rigor and insight.
Acknowledgments
The authors are grateful to Erica Michael, Michael Maxwell, and Nikki Adams for
comments on a prior draft, to the CASL Researcher’s Forum attendees for helpful
comments during our presentation, to our families for their support and patience,
and to Rebecca Goolsby for her guidance on and funding of this project.
Funding/Support This material is based upon work supported, in whole or in part,
with funding from the United States Government Office of Naval Research grant
12398640. Any opinions, findings, and conclusions or recommendations expressed
in this material are those of the author(s) and do not necessarily reflect the views of
the University of Maryland, College Park, and/or any agency or entity of the United
States Government.
vii
Contents
1 Introduction���������������������������������������������������������������������������������������������� 1
1.1 Background �������������������������������������������������������������������������������������� 3
1.2 Overview of the Framework ������������������������������������������������������������ 5
2 Sources of Messages �������������������������������������������������������������������������������� 9
2.1 Originating Content�������������������������������������������������������������������������� 10
2.2 Receiving Content���������������������������������������������������������������������������� 10
2.3 Gathering Content���������������������������������������������������������������������������� 11
2.4 Summary of Sources of Messages���������������������������������������������������� 13
3 Reactions to the Message and Messenger���������������������������������������������� 15
3.1 Affect and Engagement�������������������������������������������������������������������� 16
3.1.1 High Arousal-Specific Emotions������������������������������������������ 17
3.1.2 Other Affective Engagement������������������������������������������������ 19
3.2 Cognition������������������������������������������������������������������������������������������ 20
3.2.1 Belief in the Original Message �������������������������������������������� 21
3.2.2 Other Cognitive Reactions Rendering Belief
Unnecessary�������������������������������������������������������������������������� 32
3.3 Both Affect and Cognition���������������������������������������������������������������� 33
3.4 Summary of Reactions by Genuine Users���������������������������������������� 35
4 Motivation to Share �������������������������������������������������������������������������������� 37
4.1 Needs������������������������������������������������������������������������������������������������ 38
4.1.1 Impression Management and Self-Enhancement����������������� 38
4.1.2 Self-Consistency Motives and Social Identity���������������������� 39
4.1.3 Accuracy ������������������������������������������������������������������������������ 40
4.1.4 Affiliation������������������������������������������������������������������������������ 41
4.2 Sociopolitical and Economic Motivations���������������������������������������� 43
4.3 Summary of Motivations������������������������������������������������������������������ 44
ix
x Contents
Index������������������������������������������������������������������������������������������������������������������ 103
Chapter 1
Introduction
Abstract As the use of social media platforms has increased, so have they become
a new domain in information warfare. Before tackling the roots or spread of misin-
formation or disinformation, it is important to understand why people share any
information on social media at all. This book presents a broad, multidisciplinary
review and creation of a theoretical framework of the factors that have been shown
to, or might, influence sharing information on social media, regardless of its verac-
ity. The act of sharing information online is made up of several categories of factors:
sources of messages, reactions to the original message and messenger, the motiva-
tion to share, the ability to share (and perception of the ability to share), and then, of
course, actual sharing behavior. In addition, while genuine actors may have reac-
tions to the original message and messenger, there also exist non-genuine actors that
have pre-programmed or pre-planned reactions. We also qualitatively examined 20
fake news stories in two different languages as they appeared in social media in
order to illustrate factors affecting information propagation and identify potential
gaps in the literature.
Social media are a relatively new channel by which people not only acquire, but
also share information. Social media networks have also been used to spread mis-
information, since at least 2010 on Twitter about the Democratic candidate Martha
Coakley to the 2016 U.S. presidential election (Mustafaraj & Metaxas, 2017). The
revelations of Russian disinformation campaigns on social media against the U.S.
population during the 2016 election have identified a stark vulnerability in the secu-
rity of the United States (e.g., Sydell, 2017; United States of America v. Internet
Research Agency LLC, 2018; Waltzman, 2017; Woolley & Howard, 2017). Fake
news, as it is currently popularized,1 is not new, nor is propaganda (e.g., Allcott &
Gentzkow, 2017; Jowett & O’Donnell, 2015; Lazer et al., 2018; McKernon, 1928;
Pratkanis & Aronson, 2001). However, there has been an increase in this kind of
‘information war’ by Russia against the United States and European countries at
least since 2014 (e.g., Paul & Matthews, 2016; Prier, 2017; Woolley & Howard,
2017). Information warfare does not simply occur between nations: Both jihadi
Islamic groups and right-wing extremists also recruit and spread propaganda online
(e.g., Benigni, Joseph, & Carley, 2017; Bowman-Grieve, 2013; Caiani &
Wagemann, 2009; Derrick, Sporer, Church, & Ligon, 2016; Prier, 2017; Vidino &
Hughes, 2015).
Social media-based information conflict has sparked the interest of government,
academia, and industry alike. Key research topics underlying the flood of online
communication are why and how people share narratives and information online.
Narratives, in this context, refer to coherent stories that are shared with multiple
people rather than isolated pieces of information (Green & Brock, 2005; Hinck,
Kluver, & Cooley, 2017). A narrative might describe an activity or conflict consist-
ing of a storyline with a beginning, middle, and end, rather than a single fact, and it
may imply or state a context, how, and why (van Krieken & Sanders, 2016). Readers
of news stories can be more involved in narrative reporting, for instance, compared
to more neutral, ‘hard news’ reporting (van Krieken & Sanders, 2016). This review
examines studies of information propagation, and often focuses on narratives. Given
its conceptual breadth, we also use the term “message” to include narratives.
Although narrative propagation is not new, social media has made spreading stories,
including false ones or ones with false elements, easier.
Before tackling the roots or spread of misinformation (incomplete, vague,
ambiguous information) or disinformation (intentionally untrue information;
Cooke, 2017), it is important to understand why people share any information on
social media at all. The propagation of information has many antecedents, causes,
and moderating factors, including amplifiers and suppressors. Information propaga-
tion has been studied for decades across a range of disciplines: psychology,
marketing, sociology and social network analysis, political science and political
communication, human-computer interaction (HCI), journalism, and information
sciences. This book leverages existing knowledge about the spread of information
in general, why people are convinced by information they see, the effects of differ-
ent kinds of messages on human affect and cognition, and what might make some-
one go from interest to sharing, as well as some possible cross-cultural differences.
Most individual articles focus on a small number of factors that might be useful,
1
Fake news historically has included yellow journalism and other information in news media that
is deliberately inaccurate or misleading. For most of this book, we use the term ‘fake news’ to refer
to “news articles that are intentionally and verifiably false, and could mislead readers” (Allcott &
Gentzkow, 2017, p. 213). Note that intent is hard to prove, and propaganda can include a mix of
falsehood and truth. Although we will refer to ’fake news,’ particularly in reference to the corpus
we analyze, we will often also refer to the more technical terms of misinformation and disinforma-
tion (Cooke, 2017). Fake news may include one or both types. While disinformation is intention-
ally untrue, misinformation may have elements of truth in it (Cooke, 2017).
1.1 Background 3
with few works attempting a comprehensive view (see Hermida, 2014, for one such
attempt). In addition to surveying the literature, we conducted a bottom-up, qualita-
tive analysis of 20 fake news stories shared via social media in English and Russian,
in terms of both language and social media users and platforms (Methodology in
Appendix A). We use insights from these stories to contextualize and illustrate the
findings from the literature, as well as to add to our theoretical framework. By
assembling many factors from across a wide literature, future research can develop
measures to quantify those factors, with the goal of discovering which ones have
predictive power for encouraging sharing behavior in conjunction with each other in
real-world social media and within different cultural and language contexts.
Thus, this work presents a broad, multidisciplinary review of the factors that
have been shown to, or might, influence sharing information on social media,
regardless of its veracity. This book begins with an introduction to the problem, then
covers the background and an overview of a high-level framework of information
sharing. The framework flows through the different ways information is acquired or
viewed from a source, to reactions by the target sharer and motivations to share, to
the ability (and perceptions of the ability) to share before leading to sharing behav-
ior. Finally, this framework distinguishes between genuine and non-genuine (inau-
thentic) actors: Non-genuine actors are individuals who are pretending to be
someone they are not, and can include bots, which are automated and driven by
algorithms designed to interact or share information.
1.1 Background
In the U.S., social media use has become synonymous with digital and mobile life.
A Pew Research Center study found that in November 2018, at least 69% of U.S.
adults used at least one social media site (Pew Research Center, Internet, &
Technology, 2018). This finding is a significant jump from November 2010, when
only 45% of American adults said the same. Among popular social media sites in
the U.S., including Facebook, Twitter, Instagram, LinkedIn and Pinterest, Facebook
remains the most popular, with 68% of U.S. adults using it and 74% of them visiting
the site daily (Pew Research Center, Internet, & Technology, 2018).
Social media platforms are proving popular destinations for news consumption,
specifically. In August 2017, 67% of Americans reported getting at least some of
their news on social media, with, again, Facebook leading the way (Shearer &
Gottfried, 2017). Forty-five percent of Facebook users say they get news on the site,
followed by YouTube (18%) and Twitter (11%; Shearer & Gottfried, 2017), though
about a quarter of Americans report getting news from two or more social media
sites.
Many users openly struggle with misinformation and disinformation. A Pew
study using 2016 survey data found that 23% of respondents said they had shared a
made-up news story—some knowingly, some unknowingly (N = 1002; Barthel,
Mitchell, & Holcomb, 2016). A larger percentage of respondents in this study, 64%,
4 1 Introduction
stated that fabricated news stories “cause a great deal of confusion about the basic
facts of current issues and events.” A different study of 1200 respondents found that
only 14% of American adults viewed social media as the most important source of
election news, but that all the respondents were exposed to at least one, or perhaps
several, fake news articles (Allcott & Gentzkow, 2017). Fake news websites relied
more heavily on social media traffic than true news, which relied on social media
traffic relatively less (42% versus 10%). Of concern, in a study of disconfirmed
rumors on Twitter, most misinformed users (86–91%) spread the false rumors rather
than expressing doubt about them or seeking confirmation of the rumor, and most
(78–96%) of the debunked-rumor-spreading users neither deleted nor clarified their
original posts (Wang & Zhuang, 2018).
Social media have made the cost of spreading information low, making it easy to
influence public opinion (Allcott & Gentzkow, 2017; Shao, Ciampaglia, Varol,
Flammini, & Menczer, 2017; Shifman, 2014). Opinions are expressed attitudes
(Glynn, Herbst, O’Keefe, Shapiro, & Linderman, 2004), whereas attitudes are a
mix of behavior, affect (feelings), and cognition (beliefs; Breckler, 1984; Glynn
et al., 2004). Attitudes may be overtly or implicitly contained within narratives,
including in how narratives are presented and framed. Social media are, of course,
not only used for sharing individual political opinions. Social media are used by
large companies to advertise; by smaller-scale creators and businesses to promote
themselves and cultivate an audience and fan base; by politicians to share directly
their thoughts with constituents and others; and by individuals to share news,
memes, and their lives with each other. On the one hand, social media serve as a
new ‘public sphere’ that enables minority and oppressed voices to gain an audience
by circumventing gatekeepers, thus enhancing free speech (Debatin, 2008; Shifman,
2014). On the other hand, social media platforms are designed to encourage shar-
ing, regardless of the authenticity or benevolence of the content, and thus support
the best and worst of human psychology: the needs for attention, affiliation, and
status; the desire to control and dominate narratives; and the attraction of novelty
(e.g., Hermida, 2014; Tufekci, 2018). Malicious actors, be they state-sponsored or
unaffiliated with a state, also use social media platforms to organize, spread their
narratives, recruit, disrupt, undermine, and outright harm (see Goolsby, 2013;
Goolsby, Galeano, & Agarwal, 2014; O’Sullivan, 2018; Paul & Matthews, 2016;
Prier, 2017; Sindelar, 2014; Tufekci, 2018; Vidino & Hughes, 2015; Waltzman,
2017; Woolley & Howard, 2017).
Social media therefore can suffer from a series of problems, including deliber-
ately false information, too much information, and malicious or hostile actions
(Goolsby, 2013; Tufekci, 2018). This hostility can affect individuals outside of the
digital environment, such as when individuals’ personal information is deliberately
posted online (called doxing) and malicious actors falsely report to law enforcement
an emergency, resulting in the deployment of a SWAT (Special Weapons and
Tactics) team to an unsuspecting person’s home (swatting; Tufekci, 2018). Even
beyond deliberate malicious actions, compared to traditional media, new media
encourage audiences to follow content they already agree with (Paletz, Owen, &
Cook, 2018). Citizen journalists (and those trying to be journalists) lack the
1.2 Overview of the Framework 5
resources to truly uncover the truth and are primarily reactive (Paletz et al., 2018).
In addition, most new media platforms are controlled by a few large corporations
(e.g., Facebook), with smaller social media organizations increasingly acquired by
larger ones (Arsenault & Castells, 2008; Paletz et al., 2018). Further, social media
raise a range of privacy concerns (Ellison, Vitak, Steinfield, Gray, & Lampe, 2011;
Paletz et al., 2018; Trepte & Reinecke, 2011). Not only are people on social media
publically sharing what used to be private word-of-mouth and rumor (Hermida,
2014), but social media are used to express and promote hatred, extremism, and
fanaticism, and “are rife with muddle and nonsense, distortion and error” (Paletz
et al., 2018, p. 27). With such downsides, one wonders why anyone remains engaged
in social media, let alone shares information. Understanding why people share
information online is key to sorting through the muddle (Paletz et al., 2018).
The virality, or widespread sharing, of messages themselves has been studied exten-
sively. One conceptual model of sharing in social media, based on successful viral
marketing campaigns, suggests four success factors with the acronym ‘SPIN’:
“spreadability of content based on personal factors, the propagativity of content
based on media type, the integration of multiple media platforms, and the succes-
sive reinforcement of messaging” (Mills, 2012, p. 166). Within this SPIN frame-
work, the spreadability refers to likeability of the content of the message and
whether the sharer feels others in the social network will have a similar reaction;
propagativity refers to the ease with which consumers can continue to distribute or
redistribute the content based on both qualities of the content itself and the initial
sharer’s social network; integration refers to the strategic use of multiple social
networks simultaneously; and although most content does not achieve this final
stage, nexus refers to the “successive reinforcement of the campaign by virtue of
sequentially releasing units of viral content” (Mills, 2012, p. 168). Our framework
incorporates the factors behind both spreadability and propagativity and goes
beyond Mills’ model to include insights from a range of disciplines. Related
research utilizes network analysis to examine the relationship between different
kinds of memes based on their content, form, and stance (Shifman, 2013, 2014),
rather than the social network of individuals sharing that information (e.g., Segev,
Nissenbaum, Stolero, & Shifman, 2015). This research suggests that when a meme
spreads and changes, it retains some of its essential and unique ‘hooks’ and generic
attributes (Segev et al., 2015).
The higher-level constructs in this framework draw from a range of disciplinary
theories. For example, from political communication, we take the different ways in
which media collect narratives (Paletz, 2002). From psychology, we draw on the
interplay and distinctions among cognition, affect, and behavior; the impact of cul-
ture on the three of those; the distinctions among different types of cognitive
processing; persuasion; and the importance of both individual differences and the
6 1 Introduction
social situation (see below, e.g., Ajzen, 1991; Breckler, 1984; Brock & Green, 2005;
Cialdini & Goldstein, 2004). This framework does not presume sharing, but instead
identifies factors that may encourage or discourage sharing and the potential inter-
actions among those factors.
Most of the components of this framework relate to the process of evaluation that
a person goes through, explicitly or implicitly, before making a decision to share
information. At the most abstract level, this evaluation involves information that
originates from somewhere, a psychological reaction on the part of a real person
engaged with that information (or a pre-programmed reaction by an entity that is not
genuine), a motivation to share, and then an assessment as to what kind of sharing
is possible and/or desirable. Thus, in our framework, the act of sharing information
online is made up of several categories of factors: sources of messages, reactions to
the original message and messenger, the motivation to share, the ability to share
(and perception of the ability to share), and then, of course, actual sharing behavior
(Fig. 1.1). Dotted lines indicate the general temporal flow of a process which may,
but need not, indicate direct causal relationships. Solid lines indicate potential
directional influence. For example, the sources of messages do not cause the same
reaction in all people, but the messages elicit some kind of psychological reaction
by genuine actors. That reaction is influenced by their dynamic motivations. This
high-level framework has much in common with fundamental psychological theory
that suggests that attitudes and norms feed into behavioral intentions that then feed
into behavior, dependent on the perceived ability to perform the behavior (e.g.,
Madden, Ellen, & Ajzen, 1992). In addition, we recognize that an increasing amount
of online activity is conducted by non-genuine actors (e.g., Arnsdorf, 2017;
O’Sullivan, 2018; Woolley & Howard, 2017). These include bots, or algorithms
designed to share (Lokot & Diakopoulous, 2016; Shao et al., 2017; Varol, Ferrara,
Genuine Actors
Reactions to the
Message and
Messenger
Sources of Ability to Sharing
Messages Share Behavior
Motivation to
Share Context and match
between context and
other model
components
Non-Genuine
Actors
Davis, Menczer, & Flammini, 2017) and/or sockpuppets, or multiple fake identities
through which individuals create the illusion of support (or disdain) by pretending
to be different people (Bu, Xia, & Wang, 2013). Sockpuppets can be bots or can be
humans who are hiding behind a false identity. As defined here, non-genuine actors
on social media do not react to original messages with authentic or real psychologi-
cal responses (i.e., with affective and/or cognitive processes). Instead, they are pro-
grammed to propagate messages based on preexisting and adaptive plans. That
noted, genuine human actors are behind the creation or goals of bots and sockpup-
pets, and they have specific motivations of their own (e.g., economic, political),
such that genuine actors’ motivations impact non-genuine actors. Finally, the
broader context and match between that context and the component pieces of this
framework, including culture and the sharer’s placement within their social net-
works, influence and touch all the other components. The highest level framework,
represented in Fig. 1.1, includes all of these components (see also Fig. B1 in
Appendix B for a detailed version of the framework).
Chapter 2
Sources of Messages
Abstract Messages originate from a variety of sources. Social media users may
create content, observe a message or narrative, or seek one out. What people view is
influenced by what they search for and what is being shared already in their social
networks.
commonly, social media users see information and narratives that those they follow
share, and then may decide to (or not to) share in turn. Finally, social media users
may actively seek out others’ sources, such as online news sources or blogs (web-
based public journal logs), and then gather them to share on Facebook, Twitter, or
some other platform.
Social media users can also receive content, viewing it on their social media pas-
sively, actively, and/or via the social media platform’s algorithms. In the study of
tweets during Typhoon Haiyan, the most common type of tweet (43.4%) was sec-
ondhand reporting of the disaster, or tweeting information that arose from some-
where else (Takahashi et al., 2015). As noted previously, two-thirds of Americans
2.3 Gathering Content 11
reported getting their news via social media (Shearer & Gottfried, 2017). Social
media users’ digital networks have a significant impact on what content they are
exposed to and what they consume. As they follow specific accounts of friends,
family members, celebrities, businesses, and media organizations on social media,
what those groups and individuals share dictates the content appearing in their
feeds. Homophily is a basic principle such that individuals who are similar tend to
connect with each other (McPherson, Smith-Lovin, & Cook, 2001). Individuals
who are similar to each other socioculturally, demographically, professionally, ideo-
logically, and so on will tend to join together into networks.
For some, this clustering may create an echo chamber of information. In echo
chambers, individuals are largely exposed to conforming opinions (Flaxman, Goel,
& Rao, 2016). A study of 10.1 million Facebook users found strong liberal versus
conservative homophily, particularly for liberals (Bakshy, Messing, & Adamic,
2015). In fact, individuals chose to view less cross-cutting content than what was
available to them. However, the same study also found cross-ideological friend-
ships: of those who reported their ideology, over 20% of an individual’s Facebook
friends were from an opposing ideological affiliation. This large study suggests that
most peoples’ echo chambers are porous. That noted, in a self-report study of 103
participants, individuals engaged less on Facebook if they perceived more diversity
in their networks compared to those who perceived more similarities (Grevet,
Terveen, & Gilbert, 2014). In other words, sharing was more likely to occur in the
context of homophily, when friends exposed the user to like-minded content and the
user was more assured to have a positive audience.
Researchers examining homophily need to take into account the dynamic
(changeable) nature of friendship online (Noel & Nyhan, 2011). Two processes can
make social networks more homogeneous: Users can become more like those in the
existing network, and users can unfriend (drop from their social media network)
individuals who are dissimilar (Noel & Nyhan, 2011). These dynamic processes
increase homophily over time, making cross-sectional studies of the effects of
homophily on social influence potentially biased (Noel & Nyhan, 2011). Even with
that caveat, homophily is an established phenomenon that impacts the information
that social media users observe (Bakshy et al., 2015; Flaxman et al., 2016).
In addition, Mills’s (2012) SPIN theory of sharing identified size of network as a
potential factor of propagativity, or the ease with which users can share information.
This theory proposed that the bigger the network, the more viral content may show
up on a user’s social networking sites because of an increased chance of exposure.
Thus, individual exposure to a greater array of, and more viral, content, is in part a
function of the user’s network size.
Individuals, of course, are not simply passive receivers of information: They also
seek it out (Fisher & Julien, 2009; Kuhlthau, Heinström, & Todd, 2008). Social
media users may actively seek out news sources, science articles, or other content
12 2 Sources of Messages
directly and share those links on social media (analogous to journalists’ gathering
of information, Paletz, 2002). Information search can be a multi-stage process in its
own right, including affective and cognitive responses (Kuhlthau et al., 2008).
Conducting information search for a class project, for example, is a knowledge
construction task that entails an initial increase and then subsequent decrease in
uncertainty and anxiety (Kuhlthau et al., 2008). Less structured information search
processes could also occur in the context of sharing information on social media.1
The phenomenon of homophily can also affect what individuals seek out, as indi-
viduals gather and check news from sources they trust and follow like-minded friends
and family on social media (termed selective exposure; Paletz, Koon, Whitehead, &
Hagens, 1972; see also Chap. 3.2.1). Guess, Nyhan, and Reifler (2018), using a
nationally representative survey matched with web traffic data (N = 2,525), found
selective exposure mainly by Trump supporters and/or older adults (60+) of fake
news during the October to November period in the 2016 U.S. presidential cam-
paign. This seeking out of fake news sites seemed to occur via Facebook as a signifi-
cant conduit, and predominantly involved Trump supporters seeking out pro-Trump
fake news sites.
Flaxman et al. (2016) defined four channels through which news stories can be
discovered: direct, aggregator, social, and search. Direct discovery means a user has
gone directly to a news domain (such as nytimes.com); the aggregator channel
refers to platforms like Google News or Apple News where users are presented with
a set of links of related news topics hosted on other sites; social involves the use of
a social media platform, like Facebook, Twitter or an e-mail service; and search
involves the use of a web query on a search engine like Google, Bing or Yahoo. In a
study that analyzed web-browsing records using a data set of 2.3 billion distinct
page views, Flaxman et al. (2016) found that much of news consumption comes
from individuals simply visiting the homepage for their preferred news outlet,
which tended to be mainstream media sources. Their findings echoed other studies
(e.g., Bakshy et al., 2015) that the use of social networks and search engines are
associated with the highest levels of segregation by ideology. Thus, although social
media users often obtain their news directly from news sites rather than from their
social contacts, the users remain ideologically isolated (Flaxman et al., 2016). Given
that news sources themselves tend to link to similar perspectives in the stories they
post (Turetsky & Riddle, 2017), this finding means that active gathering does not
remove the possibility of echo chambers.
That noted, selective exposure and homophily do not always lead to blind accep-
tance of information. In an interview study of 58 people who had just chosen to
watch a film on the Vietnam War in a movie theater, 28 disliked it, with 9 consider-
ing it propaganda and 7 saying they had become more sympathetic to America’s war
in Vietnam because they had viewed the film (Paletz et al., 1972). The researchers
determined qualitatively that the American audience was annoyed and frustrated by
1
There is a significant difference between our model and the Information Search Process model
(Kuhlthau et al., 2008): Our cognitive and affective factors are reactions to information rather than
reactions to the process of going through information search.
2.4 Summary of Sources of Messages 13
the foreign film’s unconventional structure and style, as well as its hectoring tone.
This study illustrates that even with homophily, reactions to source material cannot
be taken for granted.
Social media users may create content, observe a message or narrative, or seek one
out. What people view is influenced by what they search for and what is being shared
already in their social networks. Once they receive, create, or find information, gen-
uine actors then have some kind of reaction, be it affective, cognitive, or both.
Chapter 3
Reactions to the Message and Messenger
Abstract Genuine actors will react to the message they read and the messenger
they encounter. We categorized these reactions into those related to affect and
engagement (e.g., high arousal-specific emotions such as surprise or disgust) and
those related to cognitive factors that influence belief, including factors that prompt
individuals to engage in heuristic thinking. These affective and cognitive factors
often interact in complex ways. We also categorized entertainment, humor, and
intellectual engagement as inherently both related to affect and cognition. Believing
in the content of the message is not necessary for individuals to share it online.
Genuine actors react to the messages and messengers they encounter. Psychology,
particularly the fields of social influence, persuasion, and decision making, has gen-
erated many findings regarding how human beings process and judge messages
(Brock & Green, 2005). We divide this section into primarily affective and cognitive
reactions with the caveat that for at least two decades, researchers have increasingly
understood how these two can impact each other and are interconnected (e.g.,
Loewenstein & Lerner, 2003; Sharot, Delgado, & Phelps, 2004). The factors
described in the affect and cognitive sections will influence each other, often in
subtle ways, even if they were often studied in isolation. In particular, affect is
related to ideology and belief in complex ways (Papacharissi, 2017). For instance,
beliefs are more likely to be changed if the piece of news is positive rather than
negative (Sharot & Garrett, 2016). In both the United States and the Netherlands,
sensitivity to feelings of disgust is associated with conservativism (e.g., Brenner &
Inbar, 2015; Inbar, Pizarro, Iyer, & Haidt, 2012). Experiencing a threat may make
liberals shift to conservativism, and a recent study suggests that inducing feelings of
safety and security may shift some conservatives to more socially progressive
stances (Napier, Huang, Vonasch, & Bargh, 2017). Given that political ideology can
High Arousal
Belief in Original
Emotions
Message
Other Affective
Belief Unnecessary
Engagement
influence what types of arguments may be persuasive (Jost & Krochik, 2014), affect
and cognition are not truly separate in this domain. We therefore also discuss factors
that are inherently and conceptually both cognitive and affective (e.g., intellectual
engagement).
The affect section is further divided into high arousal emotions and other affec-
tive engagement, and the cognition section is divided into factors that encourage
belief and attitudes, and those cognitive reactions unrelated to belief (or such that
belief is unnecessary; see Fig. 3.1).
Past research has suggested that emotion is a factor in sharing behavior (e.g., Berger
& Milkman, 2012; Hasell & Weeks, 2016; Peters, Kashima, & Clark, 2009). High
arousal emotions are those that feel more intense and may entail greater physiologi-
cal arousal. A high-arousal, positive emotion might be intense joy; fury is a high-
arousal, negative emotion. Contentment is a low-arousal, positive emotion, and
light sadness is a low-arousal, negative emotion. In general, stories and articles that
elicit emotions are more likely to prompt an intention to share than stories and
articles that do not elicit emotions. Theory and research on rumors, a type of actively
shared narrative, suggest that they tend to fall into four types: those that express fear
of a negative outcome, those that express hope for a positive outcome, those that
express hostility toward a group of people, and those that express curiosity about
intellectually puzzling rumors (Bordia & DiFonzo, 2007; Silverman, 2015). Three
of these types involve strong, arousing emotions, and the fourth is an intellectually
engaging state. Rumors are thought to spread due to a combination of threat man-
agement and sensemaking, and high arousal rumors are more likely to be shared
(Silverman, 2015).
In researching the valence of hundreds of New York Times articles that made the
most emailed list, Berger and Milkman (2012) found that positive content was
shared more than non-positive content. The researchers also found that content with
high arousal emotions like anger and anxiety boosted the likelihood that something
might be shared, regardless of a positive or negative valence.
Research on non-digital information-sharing behavior suggests that emotions
play a significant role in sharing and participants’ willingness to pass along infor-
mation offline, as well (Berger, 2011; Peters et al., 2009). In particular, the level of
arousal of an emotion may make a difference. In a non-digital study of intention to
share information, emotionality had a significant effect on participants’ willingness
to pass along anecdotes (N = 160; Peters et al., 2009). Anecdotes that had high or
medium emotionality were more likely to be shared than anecdotes with low emo-
tionality, a finding that the authors replicated with high versus low emotionality in a
smaller sample that actually shared the anecdotes (N = 40). In addition, anecdotes
that aroused interest, surprise, disgust and happiness were rated as being likely to be
shared, whereas anecdotes with fear, contempt, or sadness were not significantly
likely to be shared. In their second, smaller study, Peters et al. (2009) did not find
any differences for the specific emotions of happiness, sadness, or disgust, control-
ling for emotionality. In a separate study, arousal, but not positive or negative
valence, had a significant effect on intention to share news stories (N = 93; Berger,
2011). In that study, low arousal emotions were contentment or sadness, whereas
high arousal emotions were amusement or anxiety.
In an example of how content triggering high-arousal emotions may impact shar-
ing, a false story from the English corpus about Australia becoming the “most
microchipped nation” was shared multiple times on Facebook (Appendix A, Table
A1, story #433). In a public post on the user’s personal profile in 2017, one Facebook
user shared a link to a version of the story on YouTube. The user added the
18 3 Reactions to the Message and Messenger
Fig. 3.2 Facebook posts about fake story of Australia microchipping citizens
c ommentary, “Do these people realise what they are actually doing to there body’s
[sic].this.is soooooo worrying …” (see Fig. 3.2, post on left).
Another Facebook user posted a link to a version of the story on a site called sur-
vivaldan101.com. The Facebook user added commentary by writing, “And Apocalypse
is already upon them ♥ Excellent job!! Everybody will see what will happen to them
for accepting to be slaves of satan!! Stupids [sic] miserable animals!! They are not
people any more!!” (see Fig. 3.2, post on right). These comments illustrate the find-
ings by Peters et al. (2009): The information presented in this article could arguably
arouse interest, surprise and disgust in the user. By evaluating the commentary added
by Facebook users, we observed that this story invoked worry and anger.
There has also been extensive research on online sharing behavior and inten-
tions. Students were more likely to email an article if they had been jogging lightly
in place, a task that increases physiological arousal (Berger, 2011). Although the
sample was small (N = 40), the effect was large, with 75% of those who jogged
versus 35% of those who did not sharing the article. This study suggests that it is the
interpretation of the feelings of physiological arousal that encourages sharing,
rather than the specific emotions that may accompany the physiological arousal. A
separate study examined the motivations and behaviors of sharing by using semi-
structured interviews with 40 consumers born between 1978 and 1994 (Generation
Y, and mostly university students) who forwarded video content (viral messages) in
their online networks (Botha & Reyneke, 2013). The researchers showed both
content-specific (applicable to a certain group) and general videos to participants in
order to understand their emotional responses and to gauge whether or not the par-
ticipants would share the content. The general video, which was described with the
words “funny” or “laugh” was more likely to be shared; however, participants who
did not find it funny did not feel compelled to pass it along. In addition, the more
familiar the participants were with the content, the more likely they were to have an
emotional reaction and to pass it on. Valence in this study also proved important: If
participants had a positive emotional reaction to the video, they were more likely to
pass it on, but if they had a negative (or no) emotional reaction to the video, they
were not likely to share the video. However, the negative emotions captured in this
study were disinterest or boredom, rather than high-arousal negative emotions like
anger. Thus, although this study highlights the potential importance of positive
affect, it also confounded arousal and affect valence.
3.1 Affect and Engagement 19
Intense emotions are not the only type of affect related to sharing information and
narratives. Another affective reaction involves positive valence and may entail mod-
erate physiological arousal: the feeling one has when presented with something
cute. Information, narratives, and pictures may get shared on the internet simply
because they are cute (Abad-Santos, 2017; Nittono, Fukushima, Yano, Moriya, &
Paterson, 2012). Some scholars suggest that cute things are popular because they
produce positive feelings (Nittono et al., 2012). Here, we are making a distinction
between cute as adorable and cute as sexy or pretty, with the focus on the former.
Creatures are identified as cute when they have infantile (babyish) features such as
large eyes and large foreheads, heads large relative to bodies, and round cheeks
(Alley, 1981; Glocker et al., 2009). Humans are thought to instinctively respond
with caregiving desires to cute creatures (Glocker et al., 2009), even if those crea-
tures aren’t baby humans (Golle, Lisibach, Mast, & Lobmaier, 2013). In an attempt
20 3 Reactions to the Message and Messenger
3.2 Cognition
and Hagens (1972) study had a negative emotional reaction to the movie’s tone and
style, which influenced how persuasive they found its message. This section focuses
on summarizing the sizable literature on factors that influence belief in an original
message, which has been assumed to promote sharing online (e.g., Broniatowski,
Hilyard, & Dredze, 2016). However, belief is not always necessary for why indi-
viduals share narratives online. In a second subsection, we touch on those situations,
which we observed in our qualitative research and anecdotally.
There are several factors that influence readers to believe the content they read (see
Table 3.1). These factors involve source and message credibility; confirmation and
related biases; message availability, accessibility, and fluency; framing; deliberate
persuasive techniques; and individual differences of the target audience, or in this
case, potential sharer. Two theories of persuasion—the elaboration likelihood model
and the heuristic-systematic model—converged in their identification of two sepa-
rate modes of persuasion and processing social information: heuristic (or periph-
eral) and systematic modes (or central; e.g., Chaiken & Maheswaran, 1994; Chen &
Chaiken, 1999; Petty & Cacioppo, 1986; Petty, Cacioppo, Strathman, & Priester,
2005). Whereas systematic or central processing involves logically, attentively, con-
sciously, and effortfully weighing the pros and cons of information, heuristic (or
peripheral) processing entails using fast, simple heuristics and cues. Heuristic pro-
cessing can only occur if the relevant heuristics are both available in memory and
relevant (accessible) to the situation, with frequent use likely resulting in chronic
readiness of the heuristic to be used (Chen & Chaiken, 1999). The scientific litera-
ture on cognition identifies these two modes as System 1 (heuristic) and System 2
(systematic). These modes have also been argued to be fundamentally different
types of reasoning and memory in general, even beyond the processing of social
information, and to have biological bases (e.g., Evans, 2003; Smith & DeCoster,
2000). Many, but not all, of the factors listed below involve activating one of these
types of information processing. Although many of these factors involve heuristic
processing, it would be inaccurate to assume that relying solely on systematic pro-
cessing would inure people from spreading fake news. Many of the factors below
entail both types of processing, and systematic processing is less helpful when the
information available is incorrect or misrepresented.
Source and Message Credibility
The first set of factors, which generally but not exclusively involves heuristic think-
ing, entails credibility and attractiveness of the source, as well as message credibil-
ity (e.g., Heesacker, Petty, & Cacioppo, 1983; Pornpitakpan, 2004; Schwarz, Sanna,
Skurnik, & Yoon, 2007; Swire, Berinsky, Lewandowsky, & Ecker, 2017). For exam-
ple, one of the inaccurate news stories in our qualitative analysis was titled “NPR:
25 Million Votes For Clinton ‘Completely Fake’ – She Lost Popular Vote” (see
Appendix A, Table A1, story #106 for more detail). National Public Radio (NPR) is
commonly considered a high-standard and generally nonpartisan news source,
though the American right considered it leftist and biased in favor of Clinton, or at
least non-independent (Langlois, 2016). In this context, NPR may have been cited
to either identify the source as credible or to imply that the story must be true
because it conflicted with the perceived liberal bias of NPR. Facebook users repost-
ing this story explicitly raised the issue of NPR as source credibility for liberals,
such as “NPR? Wow! I think we all knew this to be true” and “My liberal friends
will break their butts saying this is ‘fake News.’”
Both source attractiveness and credibility have consistent effects on persuasive-
ness, making a message more appealing (Petty et al., 2005; Pornpitakpan, 2004).
Physical attractiveness seems to have a positive effect when perceived expertise of
a source is low (Pornpitakpan, 2004). Source credibility is similarly often linked to
message credibility. In a study of 220 students at a journalism school, individuals
judged online information, but not advertisements, based on the perceived credibil-
ity of the web source (New York Times versus a personal home page; Greer, 2003).
Quick cues and heuristics may indicate credibility even when the cues are not indic-
ative of accuracy. For example, non-native speakers of English were rated as less
credible than native speakers when stating trivia facts (Lev-Ari & Keysar, 2010).
Although the participants were able to correct this bias when the accents were mild,
they were unable to overcome this bias when the accents were heavy. In another
study, claims were more likely to be believed when the individuals to whom they
were attributed had easily pronounced names, even controlling for region of origin
of those names (Newman et al., 2014). Source credibility can be judged by per-
ceived authority (power), trustworthiness, and/or expertise (Pornpitakpan, 2004).
For example, in our Russian corpus, the story on foreign nationals collecting bio-
logical material in Russia had a huge resonance in social media, possibly due to the
fact that it was originated by President Putin and aired on national TV, and thus
considered highly credible (see Appendix A, Table A2, story #1).
Expertise entails whether the source is perceived as making accurate statements
(Pornpitakpan, 2004). A series of studies suggested a small but consistent effect of
having citations in a text on judgments of the truth of claims (Putnam & Phelps,
2017). The citations may serve as a heuristic for perceived correct expertise.
Credibility can also be attributed because of a contrast to nearby information. In a
large study (N = 877), a news story embedded in an impolite partisan blog appeared
Another random document with
no related content on Scribd:
Emperor’s frown, his, viii. 309.
empty praise or solid pudding, iii. 171.
empurpling all the ground, x. 187.
emulation, the native hue of, etc., xii. 201.
enameller of the moon, the, v. 300.
enchantments drear, x. 41.
encroachment, the figure of, iii. 75.
ends of verse and sayings of philosophers, i. 394; xi. 489.
endure to the end for liberty’s sake, ix. 162.
enemy had written a book, O that mine, vi. 205.
enemy of the human race, viii. 284; ix. 321.
enfeebles all internal strength of thought, vi. 71.
enforc’d to seek some covert nigh at hand, etc., xi. 503.
England had made Bonaparte, etc., iii. 99.
English nation, universal, vii. 167.
enlarge the conceptions or warm the heart of the spectator, to, vi.
134.
enriched, ix. 211.
Enter Sessami, vii. 86; xii. 120.
Entire affection scorneth nicer hands, viii. 455; ix. 22; xi. 524; xii.
238, 259.
envy, malice, etc., xii. 381.
Epicuri de grege porcus, iii. 42.
Epithalamia were thrown into his coffin, x. 214.
equal want of books and men, viii. 29.
equally great on a ribbon or a Raphael, ix. 352.
Erasmus aut Diabolus, Aut, ix. 34 n.
Ere the sun through heaven, etc., x. 271.
Eremites and friars, etc., xii. 337.
error of the time, the very, xi. 251.
escap’d from Pyrrho’s maze, etc., iii. 258.
essence of genius is concentration, x. 279.
Et ego in Arcadia vixi, vi. 172.
Eternal City, a part of the, ix. 232.
ethereal braid, sky-woven, xii. 203.
etherial braid, thought woven, iv. 216.
Ethiopian change his skin, Can the, etc., vii. 240, 370.
Et quum conabar scribere, versus erat, v. 79.
Et toi, guerrier infortuné, etc., xi. 282.
Et vous êtes Yorick! vii. 33.
eulogy to kill, Oh! for a, xii. 285.
European, when he has cut off his beard, If an, etc., vi. 157 n.
Even from the tomb, etc., vi. 120; xii. 159.
Even then (admire, John Bell! my simple ways), iv. 305 n.
even to o’erflowing, ix. 382.
even when he killed a calf, xii. 336.
ever charming, ever new, viii. 352.
ever lifted leg, viii. 11.
ever strong, upon the stronger side, etc., xii. 459.
every good work reprobate, to, vii. 135; x. 235.
Every moment brings, etc., iii. 207.
everything by starts, and nothing long, i. 104.
everything by turns and nothing long, xi. 515.
every variety of untried being, i. 23.
every time we called for bread, and, xii. 142.
evidence of things unseen, the, x. 86.
Evident to any one who takes a survey, it is, etc., xi. 101.
Ex uno omnes, vii. 51; viii. 366.
exact scale, according to an, viii. 93.
exaggerated evils, iii. 209.
Examines his own mind and finds nothing there, etc., vi. 124.
excellencies bear to be united, Some, etc., vi. 143.
Excellent Brutus, viii. 59.
Exchange the shepherd’s frock of native grey, etc., i. 113.
Excise, monster, iii. 465.
exhalation, Like an, etc., xii. 261, 292.
expatiates freely there, v. 102.
exploded author, that, xi. 287.
extravagant and erring spirit, vii. 16; x. 145.
Extremes meet, This is the only way of, etc., i. 97–8.
exuberant strength of my argument, iv. 21.
eye to look at, not to look with, ix. 34; xii. 354.
eye offend thee, If thine, etc., xii. 305.
eye, with lack-lustre, xii. 31, 59.
eye-judging sex, an, xii. 436.
eyelids many graces sat, Upon her, etc., x. 83, 348.
eye-pleasing flowers, v. 323.
eyes and see them, have, vi. 159.
eyes, in their arms, in their, etc., i. 45; xi. 273.
eyes of youth, x. 391.
eyes shall see me, All, ix. 69; viii. 148; x. 191.
eyes, with sparkling, etc., xii. 43.
F.
Fables for the Holy Alliance, iv. 360.
face to face, etc., xii. 43.
face was as a book, his, etc., xii. 271.
facilis descensus Averni, iii. 161.
fade by degrees into the light of common day, they, i. 250.
faded to the light of common day, ix. 62.
fænum in cornu, ix. 244.
Fain would I to be what our Dante was, etc., ix. 394; xi. 202.
faint shadow of uncertain light, Like a, vi. 113.
Fair, and of all beloved, I was not fearful, etc., v. 213.
fair clime, the lonely herdsman stretch’d, In that, etc., i. 114.
Fair moon, who with thy cold and silver shine, etc., v. 299.
Fair Semira, viii. 248.
Fair variety of things, the, ix. 332.
fairest of the fair, xii. 61.
fairest princess under sky, vi. 238; x. 242.
Fairfax and the starry Vere, vii. 232.
Fairy elves beyond the Indian Mount, etc., v. 274.
faithful remembrancers of his high endeavour, etc., vii. 430; xii.
116.
Fall blunted from the indurated breast, iv. 274.
fall degrades, But ’tis the, etc., iii. 46; vii. 368; xi. 475.
fall into misfortune, xi. 349.
fallacy, In terms a, etc., xii. 113.
Fall’n was Glenartny’s stately tree, etc., xii. 324.
false, sophistical, unfounded, etc., iii. 370.
famous for the keeping of it up, v. 131.
famous poet’s page, iv. 346; ix. 178; x. 243.
famous poet’s pen, ix. 178.
famous poet’s verse, x. 243.
famous poet’s wit, i. 23.
Fancy was a truant ever, Th’ enthusiast, vi. 72.
fancies and good-nights, xii. 224, 285.
fanciful chimeras, such, etc., iv. 282.
far darting eye, viii. 180.
far from the madding strife, vi. 100.
far from the sun and summer gale, iv. 266.
farce is over, now let us go to supper, The, vi. 150.
fared sumptuously every day, iv. 150.
farthest from them is best, iv. 261.
fashion of an hour mocks the wearer, The, etc., xi. 438.
fat and fair a bird, and how, etc., vii. 303.
fate and metaphysical aid, viii. 378.
Fate, I follow, etc., xii. 3.
father of lies, the, x. 327.
fault, it was ever the, etc., iii. 55.
faultless monsters which the world ne’er saw, Those, i. 434; ii. 129;
iv. 224; vi. 263; viii. 429; ix. 129; xii. 60.
Faunus, this Granuffo is a right wise good lord, etc., v. 226.
favours secret, sweet and precious, i. 372; viii. 14.
Fear God, and honour the King, iii. 282.
Fear God, my dear Abner, etc., ix. 116.
fear no discipline of human wit, iii. 63; xii. 378.
fear of being silent strikes us dumb, The, etc., vii. 32.
feast of reason, the, and the flow of soul, ii. 10; xii. 42, 153.
feathered, two-legged things, vii. 5.
fee-grief, due to the poet’s breast, some, vi. 174.
feel is to judge, to, xi. 85.
feel what others are and know myself a man, vii. 55.
felicity, the throne of, xii. 121.
felicity can fall to creature? What more, etc., vii. 181; xii. 2, 200.
fell of hair is likely to rouse, at which our, etc., viii. 127.
fell opposite the, viii. 356.
fell stillborn from the press, vi. 65.
fellow Burke were here now, he would kill me, If that, viii. 103.
felt a stain like a wound, v. 267; viii. 289.
Ferrara! in thy wide and grass-grown streets, xi. 424.
Few (of the University) pen plays well, etc., v. 282.
Fiat justitia, ruat cœlum, viii. 440.
Fideliter didicisse ingenuas artes, etc., vii. 235.
Fie, Sir! O fie! ’tis fulsome, xi. 419.
fields his study, nature was his book, the, vi. 181.
fierce with dark keeping, vii. 182, 278; xi. 27, 164.
fiery ordeal, x. 370.
Fiery soul that working out its way, viii. 344, 378; ix. 363; x. 393;
xi. 351.
fight, The, the fight’s the thing, etc., xii. 1.
figures nor no fantasies, They have no, xii. 5, 263, 299, 379.
finds an apple, A man, etc., vii. 176.
fine by degrees, and beautifully less, v. 359; ix. 42; xi. 386.
fine fretwork he makes of their double and single entries, iv. 364.
fine oleaginous touches of Claude, ix. 35.
fine summer evenings, when in the, they saw the frank, noble-
minded enthusiast, etc., v. 363.
fine word Legitimate, iii. 284, 293.
finical speech, a, iv. 281.
fire hot from Hell, xii. 281.
fire in the room, there was a, vi. 382.
First-born of Chaos who so fair did come, etc., viii. 58.
First come, first served, i. 53.
first garden of my innocence, that, vi. 257.
first it may be demanded, etc., But, viii. 26.
first of these is the extreme affection of two extremities, etc., The,
v. 331.
first sprightly runnings, The, i. 8; viii. 97.
first was Fancy, like a lovely boy, The, etc., v. 40.
fishing rod was a stick with a hook, a, etc., vii. 161.
fishy fume, ix. 214.
fitter for heaven, he is the, viii. 269.
Fix your eye here, etc., vii. 53.
flames in the forehead, etc., xii. 169.
flat as the palm of one’s hand, as, xi. 283.
flattery that soothes the dull cold ear, the, etc., vii. 206.
Flavia the least and slightest toy, etc., ix. 147.
fleecy fools, vi. 7.
flesh and fortune shall serve, as the, xii. 304.
flies of a summer, as the, iii. 284; vii. 234.
flocci-nauci-pili-nihili-fication, iii. 33, 231, 313; xii. 169.
Flushed with a purple grace, etc., iv. 276.
fluttering the proud Salopians, etc., xii. 259.
fly high, do we not, v. 240.
fly that sips treacle, The, is lost in the sweets, v. 129, 301; vi. 96; xii.
121.
followed in the chase, etc., xii. 272.
following things are all essential to it, the, etc., xi. 68.
Follows so the ever-running sun, etc., xii. 5.
fond deceit, And let us nurse the, etc., vi. 251.
food for the critics, viii. 223.
food whereon it lives, the very, xii. 374.
Foolish daughters of Pelias, etc., xi. 46.
fools aspiring to be knaves, iii. 67.
fools rush in where angels fear to tread, ii. 366; v. 346; ix. 480; xii.
70.
foot, an hand, an eye from Nature drawn, a, etc., v. 215.
foot of fire, with the, vi. 161.
foot mercurial, His, etc., xii. 277.
for a song, xi. 435.
For after I had from my first years, etc., v. 57.
For alas! long before I was born, etc., vi. 417.
For as much as nature hath done her part in making you a
handsome, likely man, etc., v. 284.
For her dear sake, That loves the rivers’ brinks, etc., v. 255.
For how should the soul of Socrates, etc., vii. 72.
For I am nothing if not critical, viii. 170.
For that other loss, etc., i. 118.
For this medicine, etc., v. 278.
For ’tis my outward soul, etc., viii. 52.
For true no-meaning puzzles more than wit, i. 139; viii. 552.
For wit is like a rest held up at tennis, etc., vii. 42.
For whom the merry bells had rung, v. 88.
For women, born to be controll’d, etc., vii. 203.
forehead, Her ivory, full of bounty brave, i. 69.
forerunner of the dawn, a, vi. 169.
forget the things that are behind, etc., vii. 167.
Forgive me, Now I turn to thee, thou shadow Of my contracted
lord, etc., v. 272.
form and motion so express, in, etc., xii. 248.
Fortune’s fools, vi. 460.
fortune swells him, His, etc., viii. 274.
fortune, Who shall go about to cozen, etc., xii. 297.
Forum wait for us, Let the, etc., viii. 456.
found him poor, etc., iii. 217.
fountain of blood, iii. 6.
foxes have holes, and the birds of the air, The, etc., vi. 120.
frailty, very name is, x. 397.
France, restored and shaking off her chain, iii. 51.
Franciscan think to pass, And in, etc., iii. 267.
fraught with potential infidelity, x. 127.
free born Roman maid, the, viii. 457.
Free from the Sirian star, etc., vi. 211.
French have a fault, If the, etc., vi. 307; ix. 113.
Frenchman’s darling, ix. 159.
friend in my retreat, a, etc., vi. 181.
friend in your retreat, A, etc., xii. 321.
friendly man will show himself friendly, A, etc., vii. 238.
friendship of the good, The, etc., iii. 110.
From discontent grows treason, And on the stalk of treason, death,
v. 208.
from grave to gay, from lively to severe, v. 32.
from her fair head for ever and for ever, v. 73.
From injuries to arms, and from arms to liberty, iii. 424.
From that abstraction I was roused, and how, etc., i. 117.
From that hour that Disciple took her to his own home, v. 184.
From the sublime to the ridiculous, there is but one step, viii. 23,
159.
From Windsor’s heights the expanse below, vii. 13.
From worldly care himself he did esloine, etc., xi. 333.
frozen winter and the pleasant spring, the, etc., xii. 124.
full eyes and fair cheeks of childhood, the, viii. 405.
full of matter, vi. 52.
full solemne man, a, iii. 311; xi. 413.
full to overflowing, x. 286.
full volly home, viii. 302.
fuller’s earth that takes out all stains, the true, xi. 547.
fumbling for their limbs, v. 359.
Fundamental principle of the modern philosophy is the opinion,
etc., xi. 100.
furnishing matter for innocent mirth, and, viii. 36.
fury in that Gut, there is some, viii. 304.
G.
gain but glory, iii. 259.
gain new vigour, etc., xii. 156.
Gallaspy was the tallest and strongest, etc., i. 55.
garlanded with flowers, ix. 145.
Garrit aniles ex re fabellas, iii. 419; iv. 237.
gather grapes of thorns nor figs of thistles, i. 249; vii. 200.
gaudy-days, xi. 360.
gauger of ale-firkins, a, v. 131.
Gay, sprightly land of mirth and social ease, etc., ix. 93.
gayest, happiest attitude, the, etc., viii. 41; ix. 426.
generation of actors binds another, no one, viii. 384.
generations, the, were prepared, the pangs, etc., v. 67, 235.
generous friendship no cold medium knows, A, etc., iv. 263; vi.
253.
Genius is naturally a truant, etc., vii. 59.
Genius was the child of the imitation of others, etc., vi. 127.
Genius, you must have no dependence on your own, xi. 213.
gentle craft, the, v. 302.
gentle husher, vanity by name, a, etc., vi. 289; ix. 196; x. 121; xi.
555.
gentleman and man of honour, iii. 178, 181.
Gentlemen, I can present, etc., viii. 275 n.
germain to the matter, more, xii. 239.
Gertrude’s eyes, Till now, in, etc., iv. 346.
ghost of one of the old kings of Ormus, v. 231.
Giace l’alta Cartago, etc., x. 71.
giant form roll before him in the dust, seeing his, etc., viii. 344.
giddy raptures, with all its, vii. 227.
Give a dog a bad name and hang him, iv. 1; ix. 245.
give a reason for the faith that was in me, v. 302; xii. 396.
Give me the thing and I will readily give up the name, xi. 65.
give his own little Senate laws, vii. 272.
give sorrow words, the grief that does not speak, etc., vi. 39.
give to any man without compulsion, to, xi. 419.
give up a kingdom for a mass, x. 363.
give us reason with his rhyme, vii. 371.
given in the furnace of our palace, v. 279.
gives a body to opinion, it, etc., vii. 266.
gives evidence of it, viii. 424.
gladdened life, and whose deaths eclipsed the gaiety of nations, i.
157; viii. 387, 526.
glades mild-opening, etc., xii. 202.
gladiatorship, in intellectual, viii. 84.
gladly would he learn, and gladly teach, etc., iv. 285.
glares round his soul, and mocks his closing eyelids, vii. 76; xii.
204.
glass darkly, as in a, vi. 9; xii. 152.
Glorious John, xi. 535.
glimmer, and now in gloom, now in, vii. 368; xi. 424.
glimpses that make him less forlorn, iii. 275.
Gli occhi di ch’io parlai, x. 65 n.
glittered green with sunny showers, vi. 186.
glittering bride, becomes his, etc., iii. 160; vii. 279.
glory hereafter to be revealed, the, vii. 261.
glory, the, the intuition, the amenity, vii. 120.
Glory to God, etc., iii. 266; xi. 413.
gnarled oak, the, xi. 508.
gnawed too much on the bridle, iv. 279.
gnawing the skull of his adversary, etc., ix. 401.
Go, go, you’re a censorious ill man, i. 392.
go seek some other play-fellows, v. 42.
Go thou and do likewise, vi. 164; xi. 410.
Go thy ways, old world, etc., vi. 328.
Go! you’re a censorious ill woman, viii. 78.
goes sounding on his way, iv. 214; xii. 265.
goes to church in a coranto, etc., xii. 57.
going into the wastes of time, ii. 350.
God Almighty’s gentlemen, vii. 219; viii. 85.
God knew Adam in the elements of his chaos, xi. 572.
God made the country, etc., iv. 226.
God save the King, viii. 298; ix. 93.
God the Father turns a school-divine, v. 63.
Gods have eyes but they see not, Your, etc., xii. 244.
Gods of his idolatry, the, xii. 72.
Gods partial, changeful, etc., xii. 245.
God’s image carved in ebony, xii. 392.
God’s viceregent upon earth, i. 130; x. 363.
Gog’s crosse, Gammer, etc., v. 287.
golden age, in the, v. 297.
golden mean, iv. 253.
Goldsmith of the stars, the, v. 300.
good, they did it for his, vii. 208.
good clever lad, etc., iii. 68.
good haters, i. 103, 374 n.; vii. 180; viii. 269; ix. 122.
good, he means, bad fortune, xi. 387.
good-humoured fellow, Now I think I am a, viii. 103.
good king, A, should be ... a mere cypher, etc., xii. 243.
Good lord, that there are no fairies, etc., vi. 167 n.
good-nature is a fool, mere, vii. 78.
good of the country, for the, vii. 375.
good old times, iv. 249; xi. 197.
good picture and a true, a, xi. 245.
goodly sight, It was a, to sally out from his castle, etc., i. 87.
goose pie, In form resembling a, ix. 71; xi. 200.
gorge the little fame, they get all raw, They, ix. 356.
gorge rises, our very, xii. 126.
gospel is preached to the poor, iv. 295.
gossamer that idles in the wanton summer air, the, x. 44.
Gothic cathedral ... like a petrified religion, a, vi. 369.
grace above, All is, etc., viii. 402.
graceful ornament to the civil order, etc., viii. 70.
graceful ornaments to the columns, the, etc., vii. 205.
Gracious and sweet was all he saw in her, vi. 346.
grand caterers and wet-nurses of the State, etc., ix. 24.
grandeur in it, there was a, vii. 303.
Grant I was tempted: Condemn you me, for that the Duke did love
me, etc., v. 241.
grant me judgement, you, xii. 360.
grapes of thorns, You cannot gather, etc., i. 249; vii. 200.
great book is a great evil, A, v. 114; xi. 244.
great discoverers obtain, How, shall our, i. 115.
Great Divan, the nation’s, xi. 336.
great grandmother without grey hairs, a, viii. 160.
Great is Diana of the Ephesians, xi. 603; xii. 244.
great lords and ladies do not like to have their mouths stopped,
Because, vi. 301.
great man’s memory may outlive him half a year, i. 146.
great princes have great playthings, etc., iii. 243.
Great Vulgar and the Small, i. 324; ii. 18; v. 56; vi. 157; viii. 463,
518; ix. 391, 428; xi. 437.
Great wits to madness nearly are allied, x. 231.
Greater love than this hath no man, etc., xii. 99.
greater the sinner, The, etc., xii. 330.
greatest happiness to the greatest numbers, the, vii. 180, 182, 184,
185, 193.
green-eyed, spring-nailed, etc., xi. 530.
See demure.
green upland swells that echo to the bleat of flocks, vi. 186.
Grieve not for me, etc., vi. 327.
grim-visaged comfortless despair, vii. 260.
grinding law of necessity, iv. 66, 295; vii. 193, 374.
grinding the faces of the poor, iv. 2.
grinned horrible a ghastly smile, etc., xii. 11.
grinning scorn a sacrifice, To, etc., xi. 525.
grotesque ornament to the civil order, i. 46 n.
ground, however unsafe, On this, etc., vi. 128.
grove, The, Grew dense with shadows, etc., x. 264.
Grove nods to prove each alley has a brother, etc., xi. 472.
grows with our growth, etc., vii. 60; x. 336.
guide, the anchor, the, etc., iii. 211.
guide, the stay, the, etc., iv. 205.
Guido from a daub, a, ix. 480.
Guido, from want of choice, etc., vi. 139.
Guido Reni from a prince-like affluence of fortune, etc., vi. 20.
guinea and the gallows, xi. 288, 472.
guns, drums, trumpets, viii. 403; xi. 532.
H.
habit; there is nothing so true as, vi. 33; viii. 124; x. 42; xii. 398.
Had I foreknown his death as you suggest, etc., v. 241.
Had I a heart for falsehood framed, viii. 165.
Had Petrarch gained his Laura for a wife, etc., vii. 112.
Had we but world enough and time, This coyness, Lady, were no
crime, etc., v. 314.
Hæ nugæ in seria ducunt, xi. 442.
Hæret lateri lethalis arundo, i. 135; viii. 22.
Hail, adamantine Steel! etc., xi. 505.
hail-fellow well met, v. 294.
hair-breadth ’scapes, xii. 17.
hair on end, at his own wonders, with his, etc., vi. 295.
Half thy malice youth could bear, viii. 166.
halfpenny head, having a, etc., vi. 431.
haloo an anthem, xii. 349.
hand, an ear, an eye, a, xi. 484.
hand had done, whatever the, etc., ix. 420.
hands that the rod of empire had swayed, etc., vi. 14.
handsome as you, I was never so, etc., viii. 114.
hand-writing on the wall, the, viii. 144; ix. 129.
Hang both your greedy ears upon my lips, etc., v. 208.
hang upon the beatings of my heart, vi. 257; ix. 107.
hanging locks, Like to those, etc., viii. 159; ix. 47.
Hanover rats, vi. 221 n.
happy alchemy of mind, i. 65; v. 107; viii. 408.
Happy insect, what can be, etc., viii. 59.
happy things in marriage are allowed, Two, etc., i. 68.
happy warrior, xi. 327.
hardest stone, the, etc., iii. 261.
See melancholy.
hard to say if greater want of skill, ’Tis, etc., viii. 401 n.
Hark! ’tis the twanging horn, etc., xii. 240.
Harlot old, that, etc., iii. 36, 177.
hart panteth for the waterbrooks, as the, vii. 226, 307.
hashed mutton, Amelia’s, xii. 141, 327.
has just come into this breathing world, xii. 162.
Has she not gone, trowest now thou, and lost her neele? etc., v.
287.
Hast oft been chased, etc., xi. 132, 186.
Hast thou seen the down in the air? etc., viii. 56.
hate, all we, ix. 340.
hate to fill a book with things, I, etc., vii. 399.
hated, not to be, viii. 332.
hated, needs but to be seen, which to be, etc., viii. 288.
hates conchology, he, etc., iv. 277.
hath a devil, ix. 59.
haut et puissant prince, agé d’un jour, un, viii. 176.
Have I not seen the household where love was not? xii. 88 n.
have proved a monument, i. 125.
have their hands full of truths, iv. 310.
Have ye not seen sometime a pale face, etc., v. 21.
Have you felt the wool of the beaver, etc., v. 322.
He could not read them in his old age, viii. 14.
He finds himself possessed of no other qualifications ... than what
mere common observation, etc., vi. 124.
He had received it from his grandmother, etc., viii. 228.
He hath a demon, v. 153.
He instanced it too in Lord Peterborough, vii. 209.
He is indeed a person, iii. 67.
he is one that cannot make a good leg, etc., vii. 25.
He is owner of all he surveys, vii. 68.
He is ten times handsomer, etc., viii. 442.
He looks up with awe to kings, xi. 515.
He might if he had pleased have married, i. 55.
he must rank, as a universal genius, above Dryden, etc., v. 123.
He never is—but always to be wise, iii. 139; vi. 148; ix. 249.
He openeth his hands, etc., vi. 392.
He prized black eyes, v. 189; vii. 207 n.
he saw nature in the elements of its chaos, etc., v. 341 n.
He sent a shaggy, tattered, staring slave, etc., v. 210 n.
He so teased me, viii. 323.
He takes most ease, and grows ambitious Thro’ his own wanton
fire and pride delicious, v. 254.
He that is but able to express, etc., vi. 207.
He that of such a height, hath built his mind, etc., v. 309.
he was a fine fellow once, xii. 145.
he was a fine old mouser, vi. 347.
He went up into the mountain to pray, Himself, alone, and, iii.
152.
he who knows of these delights to taste, etc., vi. 173.
he’s but his half brother, viii. 74.
head to the East, Nay, nay, lay my, iv. 248; viii. 146 n.
heaping coals of fire, etc., x. 360.
hear a sound so fine, there’s nothing lives ’twixt it and silence, etc.,
vii. 40.
hear the loud stag speak, xii. 269.
heard it, but he heeded not—his eyes, ix. 165 n.
hears it not, his thoughts are far away, He, etc., ix. 234.
hears the tumult, and is still, He, i. 338; v. 90; vi. 91.
heart of hearts, yea, into our, xii. 177.
heart of man is deceitful, the, etc., xii. 304.
hearts unkind, I’ve heard of, iii. 172; xi. 515.
heaven and all its host, he shall not perish, By, etc., viii. 307.
Heaven lies about us in our infancy, i. 250; x. 358.
Heaven, nigh-sphered in, v. 51; xii. 33.
Heaven of Invention, vi. 219.
heaven-born genius, x. 178.
Heav’n’s chancel-vault is blind with sleet, while, vi. 90.
heaves no sigh and sheds no tear, i. 135; v. 30.
he! jam satis est! iv. 305 n.
Hebrew roots, although they’re found, For, etc., viii. 64.
held on their way, etc., xii. 45.
hell of waters, A, xi. 424.
Hell was paved with infants’ skulls, vi. 76, 364; vii. 243.
hem was then heard, consequential and snapping, A, etc., i. 377.
Hence, all you vain delights, v. 295.
Her armes small, her back both straight and soft, i. 227.
Her eyes are fierce, etc., viii. 448.
Her finger was so small, the ring, etc., viii. 56.
Her full dark eyes are ever before me like a sea, like a precipice, i.
70.
Her heroes have no character at all, xii. 64.
Her voice, the music of the spheres, etc., viii. 63.
her whose foot was never off the stair, vii. 319.
Her’s is the afflicted, vi. 363.
herb that would cure him, The, xi. 328.
Here and hereafter, if the last may be? xii. 115.
Here are all that ever reigned, xi. 234.
Here be truths dashed and brewed with lies, vii. 140; x. 235.
Here be woods as green As any, air likewise as fresh and sweet,
etc., v. 254; vi. 183.