Вы находитесь на странице: 1из 23

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/283213691

Unger J, Wodak R, KhosraviNik M. Critical


Discourse Studies and Social Media Data. In:
David Silverman, ed. Qualitative Research (4th
edition). London: SAGE, 2016. 277-293 .
Chapter January 2016

READS

161

3 authors:
Majid Khosravinik

Johann Wolfgang Unger

Newcastle University

Lancaster University

20 PUBLICATIONS 301 CITATIONS

14 PUBLICATIONS 57 CITATIONS

SEE PROFILE

SEE PROFILE

Ruth Wodak
Lancaster University
230 PUBLICATIONS 4,987 CITATIONS
SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,


letting you access and read them immediately.

Available from: Majid Khosravinik


Retrieved on: 31 July 2016

UngerJ,WodakR,KhosraviNikM.CriticalDiscourseStudiesandSocialMediaData.In:David
Silverman,ed.QualitativeResearch(4thedition).London:SAGE,2016.

Preproofscopy

Criticaldiscoursestudiesandsocialmedia
data1
Johann W. Ungera, Ruth Wodakab and Majid KhosraviNikc
a. Department of Linguistics and English Language, Lancaster University
b. Institut fr Sprachwissenschaft, Universitt Wien
c. Media and Cultural Studies, Newcastle University

Abstract
This chapter outlines key theoretical and methodological aspects of critical discourse
studies (CDS) and suggests how they may be applied to social media data. After a brief
overview of the main features of CDS, there is a discussion of what makes data in
digitally mediated contexts different and interesting for researchers who adopt a critical
perspective to social phenomena. Eight methodological steps from one approach to CDS,
the discoursehistorical approach, are presented with specific reference to how they can
be applied to social media data. In the final part of the chapter, a case study showing
how a Facebook focus group was set up to study political resistance is used by way of
illustration.

Keywords
Social media, critical discourse studies, digital discourse, context, argumentation, focus
groups, Facebook, genre, discoursehistorical approach


1Sections

of this chapter have previously been published in an extended form in


Khosravinik and Unger 2016
1

1. Introducing Critical Discourse Studies (CDS)


Since the late 1980s, Critical Discourse Analysis (CDA) or as it has more recently come to
be called Critical Discourse Studies (CDS) have become a wellestablished field in the
social sciences. CDS can be defined as a problemoriented interdisciplinary research
program, subsuming a variety of approaches, each with different theoretical models,
research methods and agenda. What unites all approaches is a shared interest in the
semiotic dimensions of power, injustice, and politicaleconomic, social or cultural
change in society. Semiotic here refers to systems of signs, such as words and images
that are used in meaningful ways by motivated individuals and groups in society. The
many roots of CDS lie in rhetoric, text linguistics, anthropology, media and cultural
studies, social theory, philosophy, social psychology, cognitive science, literary studies,
and sociolinguistics, as well as in applied linguistics and pragmatics.
CDS scholars do not study a specific linguistic unit per se but rather social phenomena
which are necessarily complex and thus require an approach that draws on different
disciplines and makes use of different methods. The objects under investigation do not
have to be related to only negative or exceptionally serious social or political
experiences or events; this is a frequent misunderstanding of the aims and goals of CDS
and of the term critical which, of course, does not mean negative as it does in common
usage (see Chilton et al. 2010). Any social phenomenon lends itself to critical
investigation, to be challenged and not taken for granted.
All CDS approaches are problemoriented, and thus necessarily interdisciplinary and
eclectic. CDS is characterized by a common interest in demystifying ideologies and
power through the systematic and transparent investigation of semiotic data (written,
spoken, visual or other meaningful forms and practices). CDS researchers attempt to
make their own positions and interests explicit while retaining their respective scholarly
methodologies and while remaining selfreflective of their own research process.
CDS scholars see discourse language use in speech and writing as a form of social
practice. Describing discourse as social practice implies a twoway relationship
between a particular discursive event and the situation(s), institution(s) and social
structure(s), which frame it: The discursive event is shaped by them, but it also shapes
them. Since discourse is so socially consequential, it gives rise to important issues of
power. Discursive practices may have major ideological effects that is, they can help

produce and reproduce unequal power relations between (for instance) social classes,
genders, and ethnic/cultural majorities and minorities through the ways in which they
represent things and position people (Fairclough and Wodak, 1997). As a broad
research programme, most critical studies of discourse are based on the analysis of a
topicrelated body of linguistic data positioned and explained in relation to a socio
political context with a critical angle. Within this broad framework, texts are analysed
against genrespecific (institutional, media) backgrounds to address the processes of
distribution. This includes providing some background to the nature of the data, the
range and quality of the audience, the possibilities provided by the genre of
communication, and the semiotic features of the language used (see e.g., Wodak 2011,
2014 on the fourlevel context model of the discoursehistorical approach; and
KhosraviNik 2010 on the interaction of context levels).

Three important concepts: critique, ideology and power


In various European languages, the term critical (or its translation equivalents) has a
rather complex history; it is clear, however, that proponents of CDS use discourse
analysis to challenge what they regard as undesirable social and political practices (see
Chilton et al, 2010 for an extensive discussion). The term critical is associated with
currents of thought whose roots are in the eighteenthcentury European Enlightenment
and go back to ancient Greek philosophy. Work within CDS encompasses varied
understandings of the terms critical, criticism and critique: First, critical analysis of
discourse can imply that researchers make the implicit explicit. More specifically, it
means making explicit the implicit relationship between discourse, power and ideology,
challenging surface meanings, and not taking anything for granted. Moreover, critical
discourse analysts do not stop after having deconstructed textual meanings; they also
aim at practical application of research results. Secondly, being critical in CDA includes
being selfreflective and selfcritical. In this sense, CDA does not only mean to criticize
others. It also means to critique the critical itself. Thirdly, critical analysis itself is a
practice that may contribute to social change. Reisigl and Wodak (2001: 32ff.)
distinguish between textimmanent critique, sociodiagnostic critique, and
prospective (retrospective) critique. The first of these involves careful and transparent
text analysis, while the second is based on bringing the sociopolitical and structural
context into the analysis and interpretation of textual meanings. At this level, the aim is

to reveal multiple interests and contradictions among the text producers, based on the
evidence of the text and its contexts. The final one, prospective critique, builds on these
two levels in order to identify areas of social concern that can be taken up with
practitioners and wider audiences (Reisigl and Wodak 2001: 34). Naming oneself
critical thus implies explicit ethical standards: an intention to make the researchers
position, interests, and values explicit and their criteria as transparent as possible,
without feeling the need to apologize for the critical stance of their work (van Leeuwen
2006: 293; Wodak and Meyer 2015; Angermuller et al. 2014; Wodak 2013).
Although the core definition of ideology as a coherent and relatively stable set of beliefs or
values has remained the same in political science over time, the connotations associated
with this concept have undergone many transformations. Clearly it is not easy to capture
ideology as a belief system and simultaneously to free the concept from negative
connotations (Knight 2006: 625). It is the functioning of ideologies in everyday life that
intrigues CDS researchers. Moreover, we have to distinguish between ideology (or other
frequently used terms such as stance/ beliefs/ opinions/Weltanschauung/positioning)
and discourse (Purvis and Hunt 1993: 474ff).
Quite rightly, Purvis and Hunt state that these concepts do not stand alone but are
associated not only with other concepts but with different theoretical traditions. Thus,
ideology is usually (more or less) closely associated with the Marxist tradition, whereas
discourse has gained much significance in the linguistic turn in modern social theory
by providing a term with which to grasp the way in which language and other forms of
social semiotics not merely convey social experience but play some major part in
constituting social objects (the subjectivities and their associated identities), their
relations, and the field in which they exist' (Purvis and Hunt 1993: 474).
Power is the third concept which is central for CDS: CDS researchers are interested in
the way discourse produces and reproduces social domination, that is, power abuse of
one group over others, and how dominated groups may discursively resist such abuse.
This raises the question of how CDS researchers define power and what moral standards
allow them to differentiate between power use and abuse a question which has so far
had to remain unanswered (Billig, 2008). Holzscheiter (2005) points out that much CDS
research is concerned with differentiating the modes of exercising power in discourse
and over discourse in the field of politics Power in discourse here means actors struggles
over different interpretations of meaning. This struggle for power in discursive
4

interaction relates to the selection of specific linguistic codes, rules for interaction,
rules for access to the meaningmaking forum, rules for decisionmaking, turntaking,
opening of sessions, making contributions and interventions. (ibid, 69) Power over
discourse is defined as processes of inclusion and exclusion (Wodak 2007, 2009a, b).
Finally, power of discourse relates to the influence of historically grown macro
structures of meaning, of the conventions of the language game in which actors find
themselves (Holzscheiter 2005: 61). In texts, discursive differences are negotiated; they
are governed by differences in power that is in part encoded in, and determined by,
discourse and by genre. Therefore texts are often sites of struggle in that they show
traces of differing discourses and ideologies contending for dominance.

2. CDS and Social Media


When considering how the framework can be applied to social media, we are careful to
acknowledge differences in data types and new affordances that account for the overall
qualities of texts before engaging in more detailed analysis. However, the separation of
the online world as a strikingly different discursive arena, as advocated by early studies
in computermediated communication (CMC), does not sit well theoretically with the
sociallycritical aspirations of CDS research. Thus, just as CDS scholars would not
endorse an analytical approach that strictly separates the data from their immediate or
broader context, they should also not treat offline and online as separate and
independent of one another, a world view that Jurgenson (2012) terms digital dualism.
Any text may be seen as existing within multiple contexts (see e.g. Reisigl and Wodak
2009 or van Dijk 2005 for two different approaches to context in CDS). For social media
data specifically, it is worthwhile paying some attention to the media practices and the
affordances of the technologies that allow social media data to be produced and shared,
just as for newspaper texts, journalistic and editorial practices may be a legitimate focus
of attention. Although interactivity and the possibility of users changing roles, from text
consumers to text producers, is an overall characteristic of the participatory web, this
dynamic is not always present in various subgenres of the web. Communicative
practices in the participatory web often involve a bundle of relatively static
organizationally controlled texts, e.g. adverts and about pages, as well as what is more
usually seen as social media data, namely interactive and dynamically evolving user
5

generated texts, such as comment threads on newspaper sites or under Facebook posts.
For a CDS study involving social media data, these institutional texts should be viewed
and analysed within their new interactive context, while bearing in mind that the social
nature of communication is the core quality of textual practice in the participatory web.
While distinctions between usergenerated content and official content are still
discernible, there is a general ethos of crowdsourcing in terms of what can become
more important or attract attention, and in this sense usergenerated content and formal
texts compete with each other.
In addition to positioning social media within broader sociopolitical and media
contexts, we follow Herring (e.g. 2007) in suggesting that paying attention to specific
aspects of the medium and situation allows us to classify socialmedia data more
effectively, and thus conduct a more nuanced analysis. The table below (Table 1) shows
some of the factors that can be considered in linguistic and semiotic analyses of social
media data in general, and some of these are then taken up in our case study to illustrate
how they can be integrated into a CDS approach:

Table 1 Medium and Situation Factors (from Herring 2007, quoted in Page et al.
2014)
Medium factors

Synchronicity

Asychronoussynchronous

Message transmission

Onetoone; onetomany; many


tomany

Persistence of the transcript

Ephemeral archived

Size of the message

Amount of text conveyed

Channels of communication

Words, images, sounds, videos

Privacy settings

Public, semipublic, semiprivate,


private contexts

Anonymous

Extent to which the participants


identities are represented within
a site

Message format

Architecture for displaying


interactions

Situation factors

Participation structure

Number of participants involved

Participant characteristics

Stated or assumed demographic

and ideological characteristics


Purpose

Goals of interaction (at either


individual or group level)

Topic

Subject matter

Tone

Formal or informal

Norms

Accepted practices established


by the group

Code

Language variety and choice of


script

3. How to conduct a critical discursive analysis of social


media data
There are many different applications of critical discourse studies, and indeed one of the
core principles is that it is methodologically eclectic, and that methods and theories are
applied to the research context and data at hand. Although many of the same methods of
analysis that are typically applied to traditionally powerful texts can be applied to social
media data, there are some confounding logistical issues regarding harnessing and
defining analytical materials in social media, for instance:
1. How to collect and select data from the vast amounts available on some social
media platforms;
2. How to deal with the inherent nonlinearity of textproduction and consumption
processes;
3. How to define context visvis social media;
4. How to deal with the fleeting nature of data and the constant changes in format
and functions of platforms;
5. How to incorporate systematic observations to account for media and genre
specific contexts of communication;
6. How to decide on an ethical framework that respects individuals rights and their
understanding of how public their data should be.
The nonlinear forms of communication on the participatory web (see Markham, this
volume) indicates the usefulness of more observational research approaches, in other
words attending to communicative events and practices rather than treating

communicative data just as texts to be studied in isolation. Participantobserver


methods (at least in the initial stages of a study) seem a good way to study how users
communicate, what affordances platforms permit, their interconnectivity, the features of
different genres and the broader media context. Hence, a casestudy approach to data
and analysis seems to suit a CDS approach. We suggest three broad orientations:
a. As discourse analysts we consider the social context of the users and their
communication;
b. As critical discourse analysts we are not satisfied with the mere description of
genre, content and communication;
c. As socialmedia scholars we view the participatory Web as part of a media
apparatus which is used by individuals in society, hence we do not treat digitally
mediated texts as part of a virtual world that is separate from the physical
world and reality, despite acknowledging that digitallymediated contexts have
specific features that may affect our analyses.
For the purposes of this chapter we will outline the eight methodological steps identified
within the discoursehistorical approach to CDS (Reisigl and Wodak, 2015), and give
some indications for how these can be applied in the context of social media research:
1. Activation and consultation of preceding theoretical knowledge (i.e. recollection,
reading and discussion of previous research).
Many of the explanatory theories from traditional applications of CDS (e.g.
concerned with politics, the traditional media, institutions) may continue to be
relevant, because social media data is still a part of society and its uses often
relate to political, media or other institutions. However, there is a growing critical
body of work on social media itself (Fuchs 2014), and in related areas such as
participatory media (Bruns 2008), citizen journalism (Bruns 2005), digitally
mediated activism (Potts et al. 2014) that should be consulted depending on the
nature of the research.
2. Systematic collection of data and context information (depending on the research
question, various discourses and discursive events, social fields as well as actors,
semiotic media, genres and texts are focused on).
This can form part of the initial ethnographic/participant observations, or can
involve reading more broadly about the history and traditions of a particular
social media platform or digitally mediated community. Where information is
scarce, interviews or focus groups with users of a platform or members of a
community can provide context but could also be part of the data.
3. Selection and preparation of data for specific analyses (selection and downsizing

of data according to relevant criteria, transcription of tape recordings, etc.).


Many social media platforms make it relatively easy to acquire huge volumes of
related information, because they are often searchable, and they have builtin
systems to link different groups or texts thematically (e.g. through (hash)tags,
personalised suggestions or algorithmic ordering of search results that are most
relevant for you). Of course, ethnographic/observational or participantoriented
work can help making such difficult decisions.
4. Specification of the research question and formulation of assumptions (on the
basis of a literature review and a first skimming of the data).
As suggested above, consulting previous work on digitally mediated discourse
can help researchers identify the most salient categories of analysis, hopefully
avoiding the rather unfruitful exercise of just pointing out differences between
social media data and nonsocial media data.
5. Qualitative pilot analysis (allows testing categories and first assumptions as well as
the further specification of assumptions).
Building in an additional participant stage to confront the creators/typical
audiences of the social media data being used with the initial analysis, and make
adjustments where appropriate, can be very valuable here. However, qualitative
analysis may also comprise the communicative events or contexts directly, the
results of which can then be juxtaposed with previous theoretical insights (Stage
1) or contextual information (Stage 2)
6. Detailed case studies (of a whole range of data, primarily qualitative, but in part also
quantitative).
As in point 3 above, the availability of data is not usually the issue. However,
gathering this data ethically (see Markham, this volume) and also narrowing
down the volume of data to a manageable amount can be a challenge.
7. Formulation of critique (interpretation of results, taking into account the relevant
context knowledge and referring to the three dimensions of critique).
While not every CDS project has to necessarily be critical of the technologies and
commercial priorities underlying many social media platforms, this should
nevertheless be something that researchers are aware of when formulating their
critique.
8. Application of the detailed analytical results (if possible, the results might be
applied or proposed for application).
It is almost trivially easy to make the results of research available publicly to
anyone who is interested, and this is becoming more commonplace in academia
in general. When the object of investigation is also social media, it seems
appropriate to share research results via social media when possible, for example
using accessible blog posts or easily digestible infographics.


It is beyond the scope of this chapter to give a detailed list of linguistic and semiotic
phenomena that a CDS researcher who is interested in social media might analyse, but
many of the classic analytical categories used in various approaches to CDS (see Wodak
and Meyer 2015), such as modality, presupposition, syntax, nominalisation or
metaphors, often still can be relevant, and can be combined with the factors listed by
Herring (see above). It is important to note that the choice of which categories to analyse
is highly context and genredependent. Herring (2013) distinguishes between different
kinds of genres in the participatory web: familiar (e.g. email or textbased chat),
reconfigured (e.g. review websites, which now tend to include primarily usergenerated
rather than expertgenerated content) and emergent (e.g. microblogs like Twitter). It is
important to keep these genre differences in mind when establishing analytical
categories, as phenomena that are highly salient in one genre (for instance, jokes and
irony used in a formal speech by a politician) may be less so in socialmedia contexts, or
vice versa.

4. Case study: the role of technologies and multilingualism


in digitally mediated protest
In this case study we ask in what ways do digitally mediated technologies cause or facilitate,
complicate or frustrate communication by, to and between protestors? The data is a collection
of images taken by protestors and journalists of signs and placards at physical locations
associated with particular protest movements. These texts are retrieved from Google image
search using specific search terms. We follow an abductive approach in moving between
textual features and theories related to protests, social movements, and globalisation.
[end text box]
Since the protests and uprisings in the Middle East and North Africa region, the so-called
Arab Spring, the various manifestations of the Occupy movement and the protests again
austerity measures in London and various other affluent cities, scholars in politics, media

10

studies and various other socially-engaged disciplines have been trying to unpack the role that
social media have played in these contexts.
Much of the media coverage of the Arab Spring and other protest movements such as Occupy
has involved photographs and videos of events happening in public spaces in large cities,
many of them provided by protestors using their own cameras or smartphones. But even for
people not holding a device in their hands while on Tahrir Square or on the steps of the
London Stock Exchange, their physical realities will have been mediated in some way by
social media. Tufekci & Wilson (2012) maintain that Social media alter the key tenets of
collective action [] and, in doing so, create new vulnerabilities for even the most durable of
authoritarian regimes. Their overall finding is that social media accelerated, but were of
course not wholly responsible for, those vulnerabilities (as some mainstream Western media
would have had us believe at the time). The changes in key tenants of collective action could
be seen in the content, form, distribution, and consumption of images. For example, the image
below (Fig. X.2) is of signs laid out in Tahrir Square protesting against the Egyptian
President, Hosni Mubarak, who was ousted in 2011 after massive protests across Egypt, and
particularly in Tahrir Square in one of the key events of the Arab Spring. These signs will
have been seen by thousands of people in their physical location but they may also have been
seen by thousands if not hundreds of thousands more people as images circulated on the web,
via social media and on websites associated with more traditional news media. While the
analysis of the image itself (the various texts, languages and linguistic varieties used, the
arrangement of semiotic elements, for instance drawing on Kress and Van Leeuwen 2006 and
their grammar of visual semiosis) is an important theme for CDS, in this example we are
focusing on the bigger picture: the relation of images such as these to broader media ecologies
and global socio-political changes. The image appears on the first results page of a Google
image search for Tahrir Square signs (see below), suggesting Googles algorithms rate it
highly. This may be because it is widely linked to, because it is frequently clicked on, or for
11

various other reasons that Google algorithms have decided indicate an image is worthy of
attention. The metadata available with the image also give some clues. At the time of writing,
it has been viewed approximately 2,500 times on its source page, Al Jazeeras English Flickr
page, which is not a huge number, but in addition to being on Flickr, it is also visible on the
Wikimedia Foundation page from which it was retrieved for this article. While there are
currently no Wikipedia pages linking to it, a Google image-match search shows that this
version of the photo has been used on a number of online news sites (see
http://tinyurl.com/tahrirsquaresigns). The Flickr page also gives a considerable amount of
information about the image, for instance that it was taken on a particular type of camera (an
entry-level single-lens reflex camera) that might be used by a keen hobby photographer or
perhaps a freelance journalist, rather than a mobile device or a professional-level camera.
However, the fact that the image is found on Al Jazeeras official Flickr page suggests that it
was taken either by a journalist associated with the media corporation, or that they bought the
image from someone who was not previously affiliated. All this is valuable information for an
analysis of the media/institutional context, and it gives us some clues as to the production and
consumption of this digitally mediated image, though it does not of course give all the
information that we might wish for as CDS analysts, or that might be revealed through further
ethnographic work or contact with the image producer(s) (as advocated by Androutsopoulos,
see above).

[insert figure 9.2]


Figure 9.2 Anti-Mubarak signs in Tahrir Square (Source: Al Jazeera English under a
CC:BY:SA licence)

There is of course value in doing ethnographic work in the physical spaces where events
related to a protest are taking place. Indeed, to learn more about linguistic and political
12

behaviours, and depending on the scholarly traditions we draw on, this kind of work may be
essential. Through on-the-ground ethnographic work we can get access to local (or glocal?)
representations (or sometimes misrepresentations?) of global movements, but to grasp fully
how global media (mis)represent local voices, we of course need to include texts drawn from
these media in our analysis.
Moreover, many of these movements are not about global problems but about local (or at least
national) ones, or local manifestations of global problems. Many of the most powerful actions
and images from various protests since 2011 have been oriented towards achieving local
objectives though they have then been photographed, videoed, described in blogs, put on
Twitter and then reproduced and recontextualised globally. What many activists have been
successful in doing is using intertextual references to highlight the links and similarities
between different movements where these exist and in using previous successes (e.g.
Tunisia) to present visual and verbal arguments pertaining to their present plight. Thus, we
would argue that many of these movements are at best quasi-global. We might ask why these
local protests with their quasi-global orientations happen in particular places and at particular
times. We draw on the notion of political opportunity structures that Kitschelt (1986: 58) finds
are comprised of specific configurations of resources, institutional arrangements and
historical precedents for social mobilization, which facilitate the development of protest
movements in some instances and constrain them in others. This notion seems highly
compatible with the context-sensitive analysis of institutions as well as accounting for sociopolitical contexts of communication supported by most critical discourse scholars (Figure
X.1). In the present case, the conditions that led to the Arab Spring long preceded the advent
of social media: repressive political or military regimes, restrictions on free communication,
poverty and violence, heavy-handed policing and many other factors contributed. However, as
Tufekci and Wilson (2012) argue, the advent of social media helped to facilitate and

13

accelerate different kinds of communication and we argue that SNS should thus be regarded
as one element of the political opportunity structures.
As aforementioned, in this example, we are particularly interested in specific technologies of
protest. These are not just technologies in the sense of gadgets running on electricity. Rather,
any deliberate act of semiosis makes use of a technology-or rather strategies, even if it is the
human voice. The argument is that protestors in any context will make use of the technologies
that are available to them, and they will utilize these to the maximum extent possible within
their affordances. One technology we were particularly struck by was the use of banners,
posters or placards often handwritten or hand-painted and the subsequent appearance of
these signs in images shared via social media. The use of signs of this kind is of course not
new, and their appearance in mainstream press photographs has a similarly long pedigree.
What is relatively new, however, is the speed at which images of these signs can be spread to
viewers not physically co-present and shared and recontextualised on various social media
platforms. This thus provides us with a compelling reason for studying digitally mediated
texts in this context: the strategic (and perhaps sometimes accidental) distribution of specific
images from a physical protest location via social and traditional media has become a vital
link within and between social movements. However, at the same time that protesters are
trying to get their message out, garner more support, seek help from abroad etc., they are
working unpaid for social media corporations and generating advertising revenue.
The strong intertextual and interdiscursive links between signs in various contexts also
warranted further examination. For instance, one image of a handheld sign in Tahrir Square,
photographed in February 2011, bears the slogan Egypt Supports Wisconsin Workers; One
World, One Pain, thus expressing support for workers in the US State of Wisconsin, who
were at the time protesting against proposed laws attacking their employment rights and
pensions (see http://crooksandliars.com/scarce/sign-tahrir-square). The full extent of this
intertextuality becomes apparent when digitally mediated texts are brought into the physical
14

environment of a protest, e.g. a poster showing a Facebook Like icon or Twitter hashtag,
which is then remediated by being photographed and posted and shared via social media.
Furthermore, various ways of challenging dominant ideologies, e.g. drawing on anticapitalism or calls for political freedom, as well as connections between events, e.g.
references to the Tunisian revolution in the Egyptian uprising, are intertextual bridges
between different, but related, more-or-less mediated contexts.
We might ask whether it is the emergence of digital and social media that has made these
sorts of strong intertextual links between different protest movements possible in the first
place. This is not to suggest that the technologies literally cause the protests, but it is evident
that they form part of the political opportunity structures. However, new structures around
protests, such as flat hierarchies, shared decision-making and the absence of leaders that have
characterised movements like Occupy, may have been made possible by changes in
communication structures; in other words, they may have fundamentally changed the
relationships between affected individuals and groups, as predicted by Graham (2006, see
above). This is supported by Gonzlez-Bailn et al.s (2011: 5-6) analysis of the use of
Twitter in the Spanish May protests. They found that communication via Twitter involves a:
trade-off between global bridges (controlled by well connected users) and
local networks: the former are efficient at transmitting information, the later at
transmitting behavior. This is one reason why Twitter has played a prominent
role in so many recent protests and mobilizations: it combines the global reach
of broadcasters with local, personalized relations.
Our key finding, however, is that the relationships between protest movements across the
globe are not unlike those between linguistic communities, whereby global languages
(particularly English) often dominate. Furthermore, the hegemonic power structures found in
the global economic system are partially reproduced in the way technologies and texts flow
between protests, much as in the case of academic publishing, where Western publications
15

in the English language are seen as more prestigious and desirable in many disciplines (see for
instance Merilinen, Tienari, Thomas and Davies 2008). We make these claims for two
reasons: First, while some research points to the many multilingual and intertextual signs
found in Tahrir Square (e.g. Aboelezz 2014), we have found very little evidence that the
protests around Occupy Wall Street or in London were similarly multilingual. They were
strongly intertextual but what was notably absent from these contexts was much
engagement with other global-resistance movements. This is easily supported by systematic
content and structural analysis of the most popular image search results for Tahrir square
signs vs Occupy Wall Street signs, just a few of which are depicted in Figures X.3 and X.4,
below, and image blogs such as wearethe99percent on tumblr. Again, the details of the
analysis go beyond the scope of this case study, but these involved categorising each image
according to which language(s) and language variety(ies) were used in the signs, and specific
categories that arose from the data themselves related to the messages on the signs (e.g.
regime-critical, ironic, intertextual references etc.).

[insert figure 9.3]


Figure 9.3 Google Image search results for Tahrir square signs on 8 October 2014, limited in
this chapter to images labelled for reuse with modification.

[insert figure 9.4]


Figure 9.4 Google Image search results for Occupy Wall Street signs on 8 October 2014,
limited in this chapter to images labelled for reuse with modification.

The disparity in languages used and between references to and from each protest location may
of course be just because the pictures and descriptions of the protests that we saw during our
period of data collection were already mediated by news organisations, bloggers and tweeters,
16

which/who will naturally focus on data that are globally accessible, i.e. in English, if we are
talking about Western reporting. But as suggested earlier on, it is mainly these globallyaccessible images and reports that have the potential to forge links between protest
movements. Iconic images emerge, and by being shared and linked to repeatedly, they
become more prominent in search results, which then makes them more likely to be reused. It
is the most popular images the ones linked to, reproduced, recontextualised that constitute
the global discourse on the protest movements. Thus, we argue that these are an appropriate
site for research in themselves and carry the same, if not higher, levels of authenticity as
data sources when considering global flows of media production, consumption and
prosumption. There are of course references to the Arab Spring by the protestors in Occupy
protests and even in interviews conducted with UK rioters. In fact, Tahrir was central to the
formation of Occupy, if we take the Adbusters call to occupy Wall Street as the starting
point: Are you ready for a Tahrir moment? On 17 September, flood into lower Manhattan, set
up tents, kitchens, peaceful barricades and occupy Wall Street, we found, however, that in
the signs and texts produced by the occupiers, these references were mainly instrumentalised
in two ways: First, in arguments about the right to protest, and in attempts to highlight the
hypocrisy of governments that support protests abroad while suppressing them at home; and
second, as a symbolic claim about of the global nature of the movement. However, they did
not necessarily indicate any direct engagement with or links between these protestors and
those in other global resistance movements.
The second reason for our claims about global hegemonic flows is that it is significant that
much of the infrastructure for the social media platforms and digital devices used to augment
the physical protests is based in the United States or other Western countries, and most of
this infrastructure is run by companies that are owned by shareholders who are expecting a
return on their investment. This feeds to the narratives of the protests presented in the
Western media, particularly with the labels Twitter revolution or Facebook revolution.
17

While the mainstream English-language media have often presented a compelling narrative of
liberation, democratization and social change caused by Western technology, the reality is
of course often much more complex.
With regard to linguistic practices, we see particularly notable changes in the channels and
modes of communication, rather than in the linguistic forms being used. These are not,
however, restricted to social struggles but are common to many forms of digital
communication in the public sphere. Social and digital media have undoubtedly played a role
in organizing these protests, in making the views of protestors public and in holding public
figures to account (as suggested by Tufekci and Wilson 2012). Their particular value to
protesters has been in drawing global attention to local issues and in circumventing traditional
media outlets that are restricted by state control or commercial interests. Nevertheless, they
are themselves also susceptible to state or commercial control, and thus we should not be too
utopian about their role. With regard to multilingualism, the picture is complex: much as in
international business and politics, English occupies a hegemonic position in the global
communication of issues being protested about. Multilingualism appears at times to be
symbolic and to be heavily instrumentalised, or even commodified, as part of local protest
goals.

Future prospects
CDS is constantly developing to take account of new theoretical and methodological
developments, and being adapted to new sites and contexts of research. In one sense,
social media is just another such set of sites/contexts. However, given the sweeping
changes to communication practices across the world, particularly in more affluent
countries, that have accompanied the growth of digital communication technologies, we
feel that it is necessary for CDS to take account of not only the affordances of these
technologies. This applies to how discourse is organised, shaped and disseminated e.g.
through automated algorithms, manipulation of news feeds and the interconnectedness
18

of various platforms, but also the specific methods that can be used to study data, e.g. in
accounting for new forms of communication such as intext annotation, tagging, likes,
and sharing. It also applies to the growing body of theoretical work on social media, for
example explaining new forms of media power, that will provide new concepts and
frameworks to integrate into CDS.

Summary
In this chapter we have presented some relevant considerations when applying critical
discourse studies to social media data. We have outlined some of the key theoretical and
analytical categories from traditional applications of CDS and discussed some additional
ones that are useful in digitally mediated contexts. Eight methodological steps from the
methodology of the discoursehistorical approach are described in terms of how they
can be applied to social media, and finally we demonstrate the application of some of
these concepts and methods in a specific case study.

Useful websites
http://www.discourseanalysis.net/ Research portal for discourse studies
http://cadaad.net/ Critical Approaches to Discourse Analysis across Disciplines
website, open access journal, conference announcements and other useful content
http://snurb.info/ Website of Axel Bruns on participatory media, citizen journalism
and useful research methods and tools for large volumes of digital data
http://nsmnss.blogspot.co.uk/ New Media, New Social Science blog

Suggested Reading
For an accessible introduction to linguistic research into social media, including a
chapter on ethics, one on qualitative research, and two on ethnographic research, see
Page et al. (2014). Fuchs (2014) gives a useful and thoughtprovoking introduction to
social media from a critical (Marxist) perspective. KhosraviNik and Unger (2015), in
Wodak and Meyers new edition of Methods for Critical Discourse Analysis (2015),

19

present an extended discussion of the theoretical aspects underlying critical discourse


studies of social media.
To provide a very brief selection of empirical studies, Wodak and Wright (2006) is one
of the earliest applications of critical discourse studies to social media data (in this case
a multilingual discussion forum). In two more recent studies, KhosraviNik and Zia
analyse Facebook discussions on Iranian national identity, while Angouri and Wodak
(2014) present an analysis of discourse on the Greek exit from the Eurozone in
newspaper website comments.

References
Angermuller, J., Maingueneau, D., & Wodak, R. (Eds.). (2014). The discourse studies
reader: Main currents in theory and analysis. Amsterdam: John Benjamins Publishing
Company.
Billig, M. (2008). The language of critical discourse analysis: The case of nominalization.
Discourse & Society, 19(6), 783800.
Bruns, A. (2005). Gatewatching: Collaborative online news production. New York: Peter
Lang.
Bruns, A. (2008). Blogs, Wikipedia, Second Life, and beyond: From production to
produsage. New York: Peter Lang.
Chilton, P., Tian, H., & Wodak, R. (2010). Reflections on discourse and critique in China
and the West. Journal of Language and Politics, 9(4), 489507.
Darics, E. (2014). The blurring boundaries between synchronicity and asynchronicity
new communicative situations in workrelated instant messaging. International
Journal of Business Communication, 51(4), 337358.
doi:10.1177/2329488414525440
Delanty, G., Wodak, R., & Jones, P. (Eds.). (2008). Identity, migration and belonging.
Liverpool: Liverpool University Press.
Fuchs, C. (2014). Social media : A critical introduction. London: Sage.
Goffman, E. (1979). Footing. Semiotica, 25(12), 130. doi:10.1515/semi.1979.25.12.1
Herring, S. C. (2004). Computermediated discourse analysis: An approach to
researching online behavior. In S. Barab, R. Kling, & J. Gray (Eds.), Designing for
virtual communities in the service of learning (pp. 338376). Cambridge: Cambridge
University Press.
Herring, S. C. (2013). Discourse in web 2.0: Familiar, reconfigured, and emergent. In D.
Tannen & A. M. Trester (Eds.), Discourse 2.0: Language and new media (pp. 125).
Georgetown: Georgetown University Press.
Holzscheiter, A. (2005). Discourse as capability: Nonstate actors' capital in global
governance. MillenniumJournal of International Studies, 33(3), 723746.
Jurgenson, N. (2012). When atoms meet bits: Social media, the mobile web and
augmented revolution. Future Internet, 4(1), 8391. doi:10.3390/fi4010083
KhosraviNik, M. (2010). Actor descriptions, action attributions, and argumentation:
Towards a systematization of CDA analytical categories in the representation of
20

social groups. Critical Discourse Studies, 7(1), 5572.


KhosraviNik, M., & Unger, J. W. (2015). Critical discourse studies and social media:
Power, resistance and critique in changing media ecologies. In R. Wodak & M. Meyer
(Eds.), Methods for critical discourse analysis (3rd ed.). London: Sage.
KhosraviNik, M., & Zia, M. (2014). Persian nationalism, identity and antiArab sentiments
in Iranian Facebook discourses: Critical discourse analysis and social media
communication. Journal of Language and Politics, 13(4), 755780.
Knight, K. (2006). Transformations of the concept of ideology in the twentieth century.
American Political Science Review, 100(4), 619.
Kovcs, Andrs, & Wodak, R. (2003). NATO, neutrality and national identity : The case of
Austria and Hungary. Vienna: Bhlau.
Krzyanowski, M. (2008). Analyzing focus group discussions. In R. Wodak & M.
Krzyanowski (Eds.), Qualitative discourse analysis in the social sciences (pp. 16281).
Basingstoke: Palgrave Macmillan.
Martin, R., & Wodak, R. (2001). Discourse and discrimination: Rhetorics of racism and
antisemitism. London, Routledge.
Page, R., Barton, D., Unger, J. W., & Zappavigna, M. (2014). Researching language in social
media. In Researching language and social media. London: Routledge.
Pariser, E. (2011). The filter bubble : What the internet is hiding from you. New York:
Penguin Press.
Potts, A., Simm, W., Whittle, J., & Unger, J. (2014). Exploring success in digitally
augmented activism: A triangulated approach to analyzing UK activist twitter use.
Discourse, Context and Media, 6, 6576. doi:10.1016/j.dcm.2014.08.008
Purvis, T., & Hunt, A. (1993). Discourse, ideology, discourse, ideology, discourse,
ideology. British Journal of Sociology, 473499.
Reisigl, M., & Wodak, R. (2005). Discourse and discrimination: Rhetorics of racism and
antisemitism. London: Routledge.
Reisigl, M., & Wodak, R. (2009). The discoursehistorical approach (DHA). In R. Wodak &
M. Meyer (Eds.), Methods of critical discourse analysis (2nd ed.). London: Sage.
Reisigl, M., & Wodak, R. (2015). The discoursehistorical approach (DHA). In R. Wodak &
M. Meyer (Eds.), Methods of critical discourse analysis (3rd ed.). London: Sage.
Unger, J. W. (2013). The discursive construction of the Scots language: Education, politics
and everyday life. Amsterdam: John Benjamins Publishing.
van Dijk, T. A. (2005). Contextual knowledge management in discourse production. In R.
Wodak & P. Chilton (Eds.), A new agenda in (critical) discourse analysis (pp. 71100).
Amsterdam: John Benjamins.
van Leeuwen, T. (2006). Critical discourse analysis. In K. Brown (Ed.), Encyclopedia of
language and linguistics (2nd ed., Vol. 3, pp. 290294). Oxford: Elsevier.
Wodak, R. (2007). Discourses in European Union organizations: Aspects of access,
participation, and exclusion. Text & TalkAn Interdisciplinary Journal of Language,
Discourse Communication Studies, 27(56), 655680.
Wodak, R. (2011). Complex texts: Analysing, understanding, explaining and interpreting
meanings. Discourse Studies, 13(5), 623633.
Wodak, R. (2013). Editor's introduction: Critical discourse analysis. In R. Wodak (Ed.),
Critical discourse analysis (Vol. 1, pp. xxxLiii). London: Sage.
Wodak, R. (2014). Discourse and politics. In J. Flowerdew (Ed.), Discourse in context (pp.
321346). London: Bloomsbury.
Wodak, R., & Fairclough, N. (1997). Critical discourse analysis. In T. A. van Dijk (Ed.),
Discourse as social interaction (pp. 258284). London: Sage.
Wodak, R., & Meyer, M. (Eds.). (2015). Methods for critical discourse analysis (3rd ed.).

21

London: Sage.
Wodak, R., & Wright, S. (2006). The European Union in cyberspace: Multilingual
democratic participation in a virtual public sphere? Journal of Language and Politics,
5(2), 251275.
Wodak, R., de Cillia, R., Reisigl, M., & Liebhart, K. (2009). The discursive construction of
national identity. Edinburgh: Edinburgh University Press.
Wodak, R., De Cillia, R., Reisigl, M., Liebhart, K., Hofsttter, K., & Kargl, M. (1998). Zur
diskursiven Konstruktion nationaler Identitt. Berlin: Suhrkamp.

22

Вам также может понравиться