Zappavigna, M. (2019). Language and social media: Enacting identity through ambient affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.

Full text

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
Chapter 27
Language and Social Media:
Enacting Identity through Ambient Affiliation
Michele Zappavigna
1 Introduction
Social media services are web-based technologies that support online networking amongst their
users. This chapter explores how SFL can illuminate the types of ‘ambient affiliation’ (Zappavigna
2011; 2012; in press) that are central to this social networking in terms of the new forms of
sociality that are being enacted. Social networking sites (SNS), one of the main forms of social
media, allow relationships to be established between user accounts in the form of ‘friendship’ or
‘following’ connections (Boyd and Ellison 2007). Different kinds of semiotic associations can also
be generated in the discourse produced by social media users through, for example, the
affordances of ‘metadata’ (data about data, e.g. location information), sometimes referred to as
‘conversational tagging’ (Huang et al. 2010) or ‘social tagging’. Common examples of SNS are
Facebook, Twitter, Weibo, and Instagram. While platforms constantly change with both social
and technological imperatives, what is interesting to the social semiotician is the types of
multimodal resources that are evolving as we find new ways to establish ‘communion of feeling’
(Firth 1957) with other language users in these online environments.
This chapter will begin by reviewing some of the main communicative features of social media
discourse, before considering research into social media identities across different disciplines. It
will then explore the ‘user in uses’ perspective for exploring identity, inspired by Firth (1957),
that has arisen out of SFL research in the areas of ‘individuation’ and ‘affiliation’ (Martin 2010).
This research has been focused on understanding how identities are enacted and how social
bonds are forged interactively in discourse. The chapter concludes by surveying issues of
collecting and analyzing social media corpora as linguistic data sources for both qualitative and
quantitative research projects.
2 Social Media Discourse
The main type of texts shared with ambient audiences using social networking services are
‘posts’: multimedia content typically arranged into chronological ‘streams’ or ‘feeds’. In the case
of microblogging services such as Twitter, these posts are short character-constrained messages

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
that originally functioned predominately as status updates relating to users’ activities. These
have now taken on a broad range of communicative functions, from broadcasting ideational
‘content’, to sharing feelings. An example of a post, in this case a tweet, with a prominent
interpersonal function is the following comment on the 2014 Australian federal budget in (1):
(1) Cruel callous cronyism #ThreeWordBudget #Budget2014
This is also an example of a Twitter meme, a form of linguistic and multimodal play whereby users
are invited to contribute iterations on a theme, often as a phrasal template (see further
discussion of memes in Section 4.2 below). Social media posts of this kind are usually presented
as a temporally unfolding ‘social stream’, incorporating various types of multimedia such as
images (see Figure 27.1) and video.
Figure 27.1: An example of an Instagram feed (left) and an individual post (right)
2.1 Searchable Talk: Social Media Metadata
The drive to make our discourse searchable by using metadata such as #Budget2014 has become
a prominent social impetus, realized by a range of online communicative practices. With the
advent of social media services this relation has intensified in a movement toward ‘searchable
talk’ (Zappavigna 2012; 2015; in press), communication featuring collaborative metadata
embedded in social media texts and visible within the main content of a post. Metadata is
information that describes dimensions of a data source, for instance, information about the
author of a social media post or the location where it was created. While metadata has a long
history in the domain of information management, this is the first historical period where we see
it so closely tied to enacting social relations, having extended its semiotic reach as an information-
organizing tool to a social resource for building relationships and communities.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
A popular form of ‘social metadata’, or ‘social tagging’, is the hashtag, indicated with a # symbol
followed by a keyword or concatenated phrase or clause (for an analysis of the linguistic functions
of hashtags, see Zappavigna 2015; in press). The tweet presented in the previous section contains
two instances of social tagging in the form of hashtags (#ThreeWordBudget and #Budget2014)
used to indicate both the semantic domain of the post, and to designate the post as part of a
larger Twitter meme. Hashtags of this kind are also able to realize a range of complex
interpersonal and textual functions beyond such topic-marking. Hashtags as a form of
conversational tagging enable individuals to search social media discourse to find out what
people are saying about particular domains, or to share feelings and opinions with like-minded
users (or argue with those who do not share your worldview). In this way, social metadata
supports forms of ambient communion that arise out of the ability to search for and engage with
other people’s posts in ‘real-time’ within the social stream.
What makes social metadata particularly interesting to linguists is its capacity to infiltrate the
linguistic structure of the texts that it seeks to annotate. While traditional metadata is typically
hidden from the view of users of an information system, or separated from the main body of a
text in some systematic way, social metadata is incorporated into social media communication,
and can perform a wide range functional roles in the discourse itself (Zappavigna 2015). While
social media services collect many forms of traditional metadata about the networks, users, and
content that they manage, most of this information is not presented within the main content of
a post and certainly does not form part of the linguistic function and structure of a post. In stark
contrast, social metadata is user-generated and typically acts as a kind of ‘in-text’ tagging with a
range of novel communicative functions. For example, Zappavigna (2014b) has explored how
hashtags support attitudinal alignment around iconized dimensions of experience, such as a
positive attitude regarding coffee:
(2) I do love #morningcoffee #coffee #coffeelover #blackcoffee
The tweet above presents a coupling of positive affect with #coffee, as well as using related
coffee tags to identify the social media user as a ‘coffeelover’. It enacts both alignment around
coffee as a ‘bonding icon’ (Stenglin 2008) and foregrounds coffee as part of this user’s identity
performance, concepts that we will explore in the second half of this chapter.
2.2 Social Streams and ‘Real-Time’: Streaming Data and Temporality
Time is an important dimension in social media, and feeds of social media content are sometimes
referred to as ‘social streams’. For example, Twitter describes itself as ‘a real-time information
network’ (Twitter 2013) and presents instant engagement with its content as a critical
affordance. ‘Real-time’ as a concept ascribes value to an apparent increase in the pace at which
we engage with Internet-mediated resources, and with members of our social networks.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
‘Streaming data’ is information that unfolds in sequence, typically over chronological time. For
example, blogs and microblogs are displayed in reverse chronological order and display
timestamp information indicating when the post was published. It is now common for feeds of
‘live tweets’ to appear in news broadcasts, or during programmes involving audience
commentary, such as chat shows and panels. Because of the continuous unfolding of social
streams, they are associated with listening practices where users ‘tune in and out’, consuming
content in an ad hoc manner, rather than tracking a feed exhaustively (Crawford 2009).
The term ‘real-time’ is often used uncritically as synonymous to ‘clock-time’ or ‘synchronous
time’, where social media users are posting about events almost ‘as they happen’ via mobile
media. The classifier ‘real’, however, also has a legitimating function: web-based time is
positioned as being the equal to the ‘offline’ time of face-to-face social interaction. At the same
time, it is ‘hyper-real’, potentially affording superior access to what is happening in our social
networks in any given moment, though this position is clearly available for critique.
2.3 Interdisciplinary Communication Research into Social Media
Research into social media is a broad interdisciplinary arena that has arisen out of wider interest
in using ‘big data’ as a lens through which to examine the kinds social practices that are
developing alongside digital technologies (Boyd and Crawford 2012). Social media make available
vast quantities of naturally occurring discourse, together with complex metadata that can be
collected, stored, and analyzed. Communication-focused research into social media forms part
of a broader field known as Computer-Mediated Communication (CMC) that brings together a
range of communication theories focused on understanding electronic discourse. However,
because of the rapid development of social media technologies, they can be somewhat of a
‘moving target’ for scholars (Hogan and Quan-Haase 2010). Linguistic perspectives on social
media have largely been developed within sociolinguistics, pragmatics, and computational
linguistics. Outside linguistics, work has spanned disciplines such as sociology, informatics, and
computer science. Research employing SFL is an emergent area, with early approaches focused
on issues of identity and affiliation of the kind surveyed in the following sections of this chapter.
3 Individuation: Enacting Identity with Social Media
3.1 Interdisciplinary Research into Social Media Identities
It is unsurprising that issues of identity are the focus of much social media research, given the
important function these channels have accrued in terms of digital self-presentation. Identity is
a difficult concept to pin down as enacting identity, or doing ‘identity work’ (Benwell and Stokoe
2007), means different things across disciplines. Aside from research into personal uses of social
media, there has also been great interest in professional identities, branding, and micro-celebrity

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
(Gilpin 2010; Marwick 2011; Page 2012a). One obvious way in which social media explicitly
construes representations of the self is through profile descriptions, which often include the
social media user’s ‘demographic information, interests and relationship status along with a
photograph and a self-description’ (Boyd and Heer 2006:2). More nuanced dimensions of identity
that have been considered within the broad interdisciplinary field of social media research
include ethno-racial representation (Grasmuck et al. 2009); sexual identity (Duguay 2014);
gender, in particular, in relation to social media photographs (Albury 2015; Rose et al. 2012), and
also lexical variation (Bamman et al. 2014); age, in particular, the teenage years (Boyd 2014; Davis
2011); and location variables (Schwartz and Halegoua 2014). An important theme is the issue of
privacy in terms of how much information about the self is revealed when using an SNS (Madden
2012).
Identifying and predicating demographic variation from social media sources is seen as valuable
in areas from marketing to legal inquiry. Dimensions that have been identified as important
include contextual variables such as location (e.g. location-based prediction using geotagged
corpora) and time (using the temporal metadata afforded by social media services). In addition,
variables relating to different orders of group membership have been viewed as relevant. For
example, studies have considered the effect of age and gender on blogging (Argamon et al. 2007;
Lu et al. 2010; Rustagi et al. 2009; Schler et al. 2006).
3.2 Sociolinguistic Approaches
A body of sociolinguistic studies has begun to develop considering how identity is construed in
social media texts in terms of language variation. Seargeant and Taggs’s (2014) edited volume
brings together key work in this area, including perspectives on the performance of the ‘ludic’
self (Deumert 2014), humour and impersonation (Page 2014), user-generated online reviews
(Vásquez 2014), self-presentation and social positioning (Lee 2014), and linguistic and discursive
heterogeneity in terms of entextualization and resemiotization (Leppänen et al. 2014). Some of
this type of research has focused on variables such as gender and stylistic variation in emoticon
use (Schnoebelen 2012), building on earlier approaches outside the social media domain (e.g.
Campoy and Espinosa 2012; Coupland 2007; Eckert 2000). Other approaches adopt frameworks
from related linguistic areas such as Page’s (2012b) work applying narrative theory to social
media story-telling, considering dimensions such as the use of hashtags.
3.3 The ‘Users in Uses’ Approach to Identity
Linguistics is still coming to terms as a discipline with how to account for what Firth referred to
as the language of ‘persons’ and ‘personalities’ (Firth 1950). Hodge (2014:35) undertakes an
interesting thought experiment, contemplating what would have happened ‘if Whorf and
Halliday’s work already published by 1956 were taken as paradigm-forming works that created

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
identities in linguistics’. He suggests this would have initiated a more productive engagement
with the concept of identity in linguistics: an engagement that factors in how identities are
enacted through meaning-making. It would have allowed for what Halliday (1978) refers to as a
‘social semiotic’ perspective capable of accounting not only for the ways in which identities are
construed in discourse, but also for how such construal affords different forms of sociality.
SFL has a history of exploring semantic variation that has informed current work on how identities
are enacted in discourse (Martin, this volume; Hasan and Webster 2009; Martin 2010). Following
Firth (1957), Martin et al. (2013) conceptualize social semiotic concern with identity as a ‘users
in uses’ approach, that is, an approach which considers how particular ‘uses’ of language are
performed by particular language ‘users’:
Users of language perform their identity within uses of language. Identity, in other
words, is always already conditioned by register and genre, so that who we are
depends on the roles we play in a given situation. The identities we enact with language
at a particular point in time are influenced by the particular stage of the particular
genre in which we happen to be involved. (Martin et al. 2013:468)
Martin (2009) is careful to distinguish this type of social semiotic approach from more common-
sense perspectives that tend to characterize or describe identity in terms of individual people
rather than semiotic personae, of the kind envisaged by Firth:
One thing we have to guard against here as functional linguists is a neuro/biological
interpretation of individuals and communities instead of a social semiotic one. As Firth
warns, it is not psycho-biological entities we are exploring, but rather the bundles of
personae embodied in such entities and how these personae engender speech
fellowships. We’re not, in other words, looking at individuals interacting in groups but
rather at persons and personalities communing in discourse. (Martin 2009:563)
Martin (2009:576) argues that we have exhausted the synoptic modelling potential afforded by
the realization hierarchy, which ‘crystalises snapshots of semiotic valeur at particular points in
time’. For instance, he suggests that the traditional distinction, grounded in the realization
hierarchy, between ‘dialects as different ways of saying the same thing, and coding orientation
as meaning different things’ does not accord with the general principles of SFL, since this kind of
modelling assumes that dialects do not in themselves make meaning (Martin 2009:575–6).
However, this problem can be avoided if we acknowledge that meaning is made at all levels of
stratification, and if we develop ‘a third hierarchy alongside realization and instantiation, focusing
on the allocation of the meaning potential of culture and its deployment for affiliation’ (Martin
2009:575–6).

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
Martin (2009; 2010) refers to this new hierarchy for fostering ‘users in uses’ research as the
‘individuation’ hierarchy. The aim is to account for the ways in which the meaning potential of
culture is allocated amongst personae, in other words, the particular distribution and patterns of
meaning instantiated by different personae. This hierarchy is geared towards taking account of
‘logogenesis - i.e. unfolding discourse at the instance end of the instantiation cline, ontogenesis
- i.e. individual development at the repertoire end of the individuation cline, and phylogenesis -
i.e. the evolutionary consequences of variation according to users (individuation) and uses
(instantiation)’ (Martin 2009:576). Figure 27.2 represents the individuation hierarchy from the
perspective of both allocation, that is, how semiotic resources are distributed amongst users, and
affiliation, that is, how semiotic resources are deployed to commune (Martin et al. 2013:490).
The intent is to model the relationship between personae, sub-cultures, master identities, and,
at the most generalized end of the cline, culture as a system of meaning potential. However, work
on the exact nature of this hierarchy, and what constitutes meaningful units of analysis, is still in
its infancy (e.g. see Zappavigna and Martin 2014).
Figure 27.2: The individuation cline (adapted from Martin et al. (2013:490))

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
Martin grounds his perspective on individuation in the Bernsteinian notion of culture as a
‘reservoir’ of meanings from which an individual can mobilize a particular ‘repertoire’. This
repertoire arises out of the registers and genres to which they have been exposed. He quotes the
following passage from Bernstein in order to illustrate this point:
I shall use the term repertoire to refer to the set of strategies and their analogic
potential possessed by any one individual and the term reservoir to refer to the total
of sets and its potential of the community as a whole. Thus the repertoire of each
member of the community will have both a common nucleus but there will be
differences between the repertoires. There will be differences between the repertoires
because of the differences between members arising out of differences in members’
context and activities and their associated issues (Bernstein 2000:158)
According to this kind of approach, the interpersonal dimension of meaning inflects the kind of
personae that we can take up in social life, just as the kinds of roles that these personae can
adopt is modulated by the genres into which we have been socialized (Martin and Rose 2008).
A collection of work on identity influenced by the ‘users in uses’ perspective by mostly Sydney-
based researchers was published as a volume edited by Bednarek and Martin (2010). Dimensions
that have been considered in this volume and in related work, include gender (Bakar 2014; 2015;
Hamid and Bakar 2010; Tian 2008), nationality (Tann 2010; 2013), emotionality (Bednarek 2010;
2013; in press), multimodality (Caldwell 2010), and, most relevant to this chapter, ambience in
relation to social media identities (Zappavigna 2014a). These approaches vary from close
qualitative approaches using multimodal discourse analysis, to corpus-based and quantitative
approaches (O’Donnell 2014). They also draw extensively on the Appraisal framework in order to
explore the expression of attitude and the ‘syndromes of evaluation which characterize an
individual – their appraisal signature’ manifest as ‘the idiolectal reconfigurations of meaning-
making potential by which individual authors achieve a recognisable personal style’ (Martin and
White 2005:208).
3.4 Current SFL Work on Individuation in Social Media
Work from an SFL perspective on identity performance in social media is still very much in its
early stages. Some multimodal work on representation of the self has been undertaken by
Zappavigna and Zhao (2017) and Zhao and Zappavigna (2017), and by Zappavigna (2016) in
relation to the representation of subjectivity in Instagram images and Tumblr posts. This work
has considered the different options for visually representing the self in social media images,
from the choice to explicitly represent the self in the form of a self-portrait, or ‘selfie’, or the
choice to imply or infer the presence of the self through compositional choices (e.g. a cup of
coffee near the front of an image indicating the photographer outside the frame), or through

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
inclusion of part of the photographer’s body in the image (e.g. a hand holding a coffee cup).
Zappavigna (2016) models these semiotic choices as a system of ‘subjectification’. The major
choice in this system is between ‘as photographer’ perspective, where the photographer’s
subjectivity is represented, inferred, or implied through various choices in visual structure, and
‘with photographer’ subjectivity, where the photographer’s perspective is unmarked (Figure
27.3).
Figure 27.3: Subjectification (adapted from Zappavigna 2016)
Drawing on this initial modelling, Zhao and Zappavigna’s work aims to show that the function of
the selfie as a multimodal genre is not solely to represent ‘the self’ but rather to enact
intersubjectivity, that is, to generate various possibilities of relations between perspectives on a
particular topic, issue, or experience and hence to open up potential for negotiating different
points of view.
Selfies, as performances of identity, have been widely criticized in news and entertainment media
as an inherently narcissistic practice. One of the assumptions at the core of this criticism is the
position that the naturalized reading of a selfie is ‘Look at me’. It is a proposition difficult to

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
refute, as the self/subject is the primary representational object in selfies as a visual media (Pham
2015). However, if we adopt a social semiotic approach that factors in the importance of
interpersonal meaning, we can propose a complementary reading: that the visual structure seen
in selfies foregrounds the perspective of the photographer on a particular object, phenomenon,
or issue. For instance, in the domain controversially known as ‘mommyblogging’, selfies taken by
mothers can be read as ‘Look, it is my perspective on motherhood’ or ‘Let’s look at motherhood
through my perspective’ (Zappavigna and Zhao 2017). This subtle shift of focus from ‘me’ to ‘my
perspective’ is a very important consideration, as it affords a shift in analytical focus from the
ideational to the interpersonal.
Zappavigna (2014b) has explored microblogging identities in relation to the types of bonds that
are enacted within different communities of users. For example, this work focused on a cluster
of bonds that occur in microblogging discourse about motherhood, but which also appear to be
generalizable across different types of personae and communities. The first bond is a ‘self-
deprecation bond’, where the social media user can be read as laughing off (Knight 2010), or
lightly and irreverently mocking, a stereotype such as the concept of the ‘perfect mother’ (Lopez
2009), an example of which is given in (3):
(3) Just split wine on son’s homework diary. #badmother
A co-occurring bond is the ‘addiction bond’, as in example (4), in which the user rallies around
everyday bonding icons such as coffee, wine, or technology:
(4) Need.Wine.
And finally a frazzle bond (example (5)), whereby the user relates fatigue or exasperation
resulting from engaging in the core activity of a particular community of fellowship, for instance:
(5) What a spectacularly horrendous night. Max awake for 3 hours, wanting ‘iPad! IPAD!!’ I
think NOT baby. Then twins snuck into bed @ 4am. Ugh
This interplay or ‘complex’ of bonds is also seen in personae enacted in other domains, such as
the world of computer coding: for instance, example (6) is a parallel to the above frazzle bond:
(6) Getting ready to go to bed after a long night of coding up some custom classes to handle
xml parsing and database interaction.
Section 4 details how we may identify these types of bonds by the way in which they are realized
as particular couplings of attitude and experience in discourse.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
4 Affiliation: Communing with Social Media
4.1 Developing the Cline of Affiliation
As Figure 27.2 suggests, complementary to individuation is ‘affiliation’, that is, how community
is negotiated in discourse. Language is replete with resources for enacting social relations and
forming networks of semiotic ‘bonds’ (Knight 2010; 2013; Martin 2010). We can think of the
formation of community as involving ‘bonds’ realized in discourse as patterns of values. These
bonds are negotiated in interactions, producing different types of membership or fellowship. As
we will see in the next section, these connections may be ambient in the sense that they do not
necessarily require direct contact between users and may draw on mass forms of communion of
feeling.
The discursive patterns of values associated with different bonds and bond complexes may be
instantiated across multiple dimensions of meaning ‘across metafunctions (ideational resources
such as technical and specialized lexis, interpersonal resources such as naming and vocatives)
and across strata (accents in phonology, grammatical variation and discourse semantic style)’
(Martin 2010:25). They may also span different semiotic modes (Dreyfus et al. 2011). For
instance, Martin et al. (2013) have considered the role that gesture plays in the interactive
bonding process of face-to-face encounters. This work analyzed the way that particular gestures
support the proposal and response to different kinds of attitudinal alignments by realizing
particular social bonds in the discourse. In addition Hood has considered the role of gesture in
forging alignments in classroom discourse (Hood 2011).
A key technical concept deployed in this type of work on affiliation is the notion of ‘coupling’
Martin (2000), a concept that has been used to model the kinds of connections that can be forged
across different kinds of linguistic and multimodal systems. In its more general sense, coupling
refers to textual relations that involve ‘the temporal relation of ‘with’: variable x comes with
variable y’ (Zhao 2011:144), and these co-selections may be made ‘across ranks, metafunctions,
strata and modalities which are not specified by system/structure cycles’ (Martin et al 2013:469).
Most studies applying coupling have focused on evaluative meaning, since evaluation of
proposals and propositions is a particularly important resource for construing solidarity:
Feelings are meanings we commune with, since we do not say what we feel unless we
expect the person we are talking with to sympathize or empathize with us. We express
feelings in order to share them … to build relationships; where we misjudge the
situation and get rebuffed, then a sense of alienation sets in. (Martin 2002:196)
We do not commune, however, around feelings disjointed from their ideational targets, but
around the connection of these feelings to dimensions of our experience, as ‘couplings’ of

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
attitude and ideation (Martin 2000). For example, the following tweet in (7) instantiates a
coupling of experiential meaning with negative judgement (attitude in bold, ideation underlined):
(7) She is such a hypocrite
Martin et al. (2013) represent this type of coupling with the following notation:
1
[ideation: she /
evaluation: negative judgement]. In any given encounter we are always proposing and reacting
to bonds as we negotiate couplings in discourse. It is in this sense that we can view ‘persons and
personalities as active participators in the creation and maintenance of cultural values’ (Firth
1957:186). While the term negotiation characterizes this as a very deliberate activity, the
practices at work always involve a tacit dimension: just as we do not usually consciously attend
to our linguistic choices (Zappavigna 2013), we are rarely directly aware of the patterns of bonds
that we propose and react to in an interaction.
Current work on affiliation is focused on how bonds pattern into higher orders complexes, that
is, how they ‘cluster as belongings of different orders (including relatively ‘local’ familial, collegial,
professional and leisure/recreational affiliations and more ‘general’ fellowships reflecting
‘master identities’ including social class, gender, generation, ethnicity, and dis/ability)’ (Martin et
al. 2013:490).
2
The aim is to use semiotic rather than common-sense criteria to account for the
kinds of memberships that emerge when we study coupling patterns. According to this
perspective, identities are patterns of meaning inflected by membership in networks of
fellowship. In other words, they can be thought of as the disposition to enact particular
configurations of couplings that realize particular configurations of bonds. This disposition is
informed by a persona’s particular semiotic ‘repertoire’ that arises out of the potential semiotic
‘reservoir’ available via their membership in a given community (Bernstein 2000).
4.2 Ambient Affiliation in Online Media
Affiliation has largely been developed with a view to explaining dialogic discourse, in the sense
of interactants forging dynamic, interactive alignments through rallying around, condemning, or
laughing off couplings (Knight 2010). In terms of the modalities related to face-to-face
interaction, paralinguistic clues such as gesture and posture can assist in analyzing how bonds
are orchestrated.
3
However, in communication via online media, where we do not have some of
these types of paralinguistic clues (depending on the nature of the digital channel), markers of
solidarity such as emoticons and hashtags play a pivotal role in supporting bonding (Zappavigna
1
Other notation systems and ways of visually representing how couplings function in
interactions are currently being developed.
2
For a related approach which focuses instead on the tenor dimension, see Don (2012).
3
For gesture analysis see Martin et al. (2013); for image analysis see Caple (2010).

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
2012). In addition, some forms of social media discourse are not always directly interactive: for
example, a study of a corpus of thirty-eight billion tweets in existence in 2014 found that only
twenty-five per cent of the tweets in the sample received replies (Liu et al. 2014). To add to the
complexity, ‘communion of feeling’ in social media discourse can also incorporate mass-
communicative practices, such as social tagging and producing iterations of Internet memes
(media such as catchy phrases, image macros, or videos that are widely shared through social
networks). These practices do not require direct interaction between particular users in order to
forge attitudinal alignments.
Zappavigna (2011) has employed the concept of ‘ambient affiliation’ to explore how semiotic
resources are deployed to commune, in particular, when such communion operates beyond the
negotiation of couplings within explicit conversational exchange structures. For example, mass
performances of hashtagging are ambient in this sense. Consider the following post, expressing
negative judgement regarding the 2014 Australian federal budget with the meme,
‘ThreeWordBudget’. This meme involved users posting three words, typically opinion or
sentiment, targeted at the budget, the government, the treasurer and related content in this
semantic domain. For example, the main function of the following post in (8) taken from the
#ThreeWordBudget corpus is not ideational (e.g. about specific details of the budget) but
interpersonal (sharing negative evaluation of the budget):
(8) #ThreeWordBudget Cruel and Mean
The meme involved mass expression of negative judgement and negative affect coupled with the
budget, as is shown in examples (9) to (12):
(9) #ThreeWordBudget Lie, cheat, steal!
(10) Evil, malevolent shitheads. #threewordbudget
(11) I’m Absolutely Terrified #threewordbudget
(12) #ThreeWordBudget Unjust Unfair Unwanted
This kind of data pushes our social semiotic modelling of affiliation into considering how bonds
can function outside exchange structures. Appending a hashtag to a post presupposes that there
is an ambient audience who may share or contest the values construed.
Adopting a Firthian perspective on microblogging, we might think of posts such as the above as
proposing ‘bonds’ to an ambient audience in order to invite communion of feeling. These bonds
are realized as the coupling of negative judgement/affect with the ideational label realized in the
hashtag also coordinating the meme. This function is related to the notion of ‘phatic communion’
(Malinowski 1972), where the main communicative function is interpersonal. We might think of
posts as always having dual interpersonal affordances: construing a value at the same time as

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
entering the personae construed in the text into a relationship with other personae in the social
media stream, who in turn manifest repeated coupling patterns.
4
The repeated coupling coordinated by the #threewordbudget hashtag is related to Knight’s
(2010) notion of ‘rallying affiliation’ in interactive discourse. The above posts are examples of
users ‘communing’ around shared negative assessment of the budget. We can also think of the
tag as functioning like a vocative, a call out to a putative ambient fellowship or ‘college’ of
potential aligners who share this value. This is a process that Zappavigna and Martin (in
preparation) term ‘convocation’. Understanding how communion and convocation function in
social media is a current critical challenge that is stretching our existent social semiotic tools for
understanding bonding to their limits.
5 Collecting and Analyzing Social Media Texts
5.1 Methodological and technical issues
The sections above have considered the theoretical details of how to begin exploring identity and
affiliation in social media texts. In this section we are focused on data collection and analysis
issues that inform how we can use social media texts to address these theoretical interests. What
is exciting about high-volume social media data is not simply its size,
5
but how particular
affordances of this type of data can illuminate different kinds of social relations. For instance, the
availability of metadata, allowing search, aggregation, and cross-referencing, enables different
viewpoints on a data set (Boyd and Crawford 2012). These affordances also impact how we can
apply our SFL tools to these texts.
We can group the main issues that may arise during the data collection phase of a social media
research project into two areas: methodological issues regarding the nature of the data to be
collected to answer the particular research question at hand, and technical issues about the
practical mechanics of gathering the data, which in the case of social media services, usually
involves some proficiency in using APIs (Application Programming Interfaces) for ‘scraping’ data
from the service’s databases (see Zappavigna 2012 for an overview of the key issues relating to
API use). An important methodological concern is sampling, that is, determining the scope and
criteria for collecting texts. Many of the kinds of difficulties that may be encountered when
4
For an example of a Twitter study using the concept of ambient affiliation to explore attitudinal
alignments during the 2011 Japan nuclear crisis, see Inako (2013).
5
Indeed Boyd and Crawford (2012:663) point out that ‘some of the data encompassed by Big
Data (e.g. all Twitter messages about a particular topic) are not nearly as large as earlier data sets
that were not considered Big Data (e.g. census data)’.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
sampling and collecting social media texts are not unique to electronic discourse (think, for
example, of the multitude of considerations that must be accounted for when dealing with
spoken discourse that is to be transcribed); however, the relative novelty of social media
platforms means that standards for sampling are still being developed, and we are yet to even
clearly establish what the important issues and pitfalls are. One popular form of sampling that
has received some criticism is hashtag studies where a tag is used to collect some form of
experience (Bruns 2013). However, as Crawford (2013) has noted in relation to the general field
of big data analysis, this type of perspective only captures a slice of the potential voices involved
in a dimension of social life at any point in time. It may factor out people who are not using social
media technologies, possibly the most disenfranchised within a community. As such it might
create what she terms ‘algorithmic illusions’ that pollute our data analysis.
A prerequisite for usefully accumulating social media data for linguistic analysis is being able to
specify both the unit of analysis (what we are drawing conclusions about) and the unit of
observation (the kinds of textual data that need to be sampled in order to draw these
conclusions). This is the case for both qualitative and quantitative studies. For example, a study
might investigate the social practices involved in Facebook posts (unit of analysis), realized in
specific patterns of linguistic features in these posts (unit of observation), identified by applying
any of the multifaceted types of analysis made available by SFL theory. Social media texts, and
electronic discourse more broadly, however, problematize some of the practices linguists may
have used in the past for denoting what constitutes a text to be analyzed. While all texts enter
into heteroglossic relations with other texts (Bakhtin 1981), and can be approached from both
dynamic (text as process) and synoptic (text as artefact) perspectives, determining the ‘bounds’
of a networked, electronic text is particularly challenging. For example, Highfield and Leaver
(2014) have noted the instability of Instagram posts due to the dynamic nature of this media,
which allows users to go back to posts and add or delete comments on the image. In this case
the text, is, from the perspective of mode, never complete as a communicative unit since it can
always be modified and added to.
In addition, in terms of multimodal dimensions such as presentation (e.g. font, colour choice,
layout, etc.), the text is also not static since it may be syndicated to different kinds of devices (e.g.
a smart phone) where it may appear visually different from its instantiation in a web browser
running on a desktop computer. Thus, a concern, although not unique to social media texts, is
how to collect, store, and manipulate multimodal semiotic resources. In other words, how do we
collect and track relations between elements of texts that involve multimedia? Again, we need
to think carefully about what matters to a study. For instance, do we need to collect information
about the presentational layout of a blog post, or are we concerned with the verbiage of the blog
post alone? Do we need to track relationships between the verbiage and the layout? The
seemingly simple task of establishing the dimensions of a semiotic resource is often more difficult
than it might at first appear, and opens up a range of interesting data collection issues that are

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
intimately tied to the kinds of analysis that will be able to be undertaken in a study. A
fundamental question in most social media studies will be ‘Do I need to use a database (and/or
a script for processing the data)?’
6
or ‘Can I simply collect data in its native format or a simple
format such a plain text?’ Answering these questions means knowing which variables matter to
the study and whether or not there are multiple relationships between variables that need to be
tracked.
If we are interested in tracking contextual variables alongside the ‘content’ of a social media text,
the volume of data can quickly become extremely large and unwieldy. For example, if we wanted
to track social relations based on metadata alongside the linguistic patterns in microblogging, the
data can rapidly expand in complexity. One early study, outside linguistics, claiming to analyze
the entire Twittersphere at the time, generated a collection of ‘41.7 million user profiles, 1.47
billion social relations, 4,262 trending topics, and 106 million tweets’ (Kwak et al. 2010:600). If
detailed metafunctional linguistic analyses were also to be added, the number and kind of
relationships to be quantitatively tracked would be immense. We can partially overcome this
type of problem if we ‘shunt’ between qualitative approaches and quantitative approaches to
sampling and analysis. For example, we might adopt what Bednarek (2009), developing Baker
(2006), refers to as a ‘three-pronged’ method for corpus-based discourse analysis. This approach
incorporates close manual analysis of single texts with manual, or partially automated, small-
scale corpus-based analysis, complementing quantitative work, often highly automated, using
large-scale million-word corpora. However, it needs to be noted again that, because of the kinds
of non-standard orthography and other features of social media, many of the standard tools,
such as POS taggers, available for supporting automatic analysis, may not work – for an example
of a POS tagger that has been modified to account for Twitter data, see the GATE Twitter POS
Tagger (Derczynski et al. 2013).
5.1 How Time Affects Data Collection and Analysis
The blend of asynchronous and synchronous time involved both in the production and
consumption of social media texts and streams can make data collection and analysis
problematic. On the one hand, social media texts that involve streaming data (chronologically
organized feeds of data) might have a ‘real-time’ reference point traceable via a timestamp in
the metadata.
7
On the other hand, it may also be asynchronously modified, replied to,
syndicated, and rebroadcast. When viewed as a ‘conversation-like’ interaction, a social media
post may be part of an interactional unit evolving long after the original text was published.
6
Scripts are small programs using a scripting language to automate text-processing tasks that
would otherwise have to be done manually by a linguist.
7
A timestamp is a record, usually encoded in a standard format, marking when an event
occurred.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
In addition, the social processes involved in producing a social media text, and the ‘scope’ of a
text (e.g. the bounds of a conversation-like interaction), will also look different depending on the
‘timescale’ (Lemke 2000) across which they are viewed (e.g. a week, a day, a decade etc.).
Different time scales will allow the researcher to observe different relations between the
semiotic resources that they are studying (Zhao 2010). In addition, the rapid speed at which
technology changes means that our most refined tools for social research, which have been
developed for analyzing older semiotic modes, are often unable to be rigorously applied:
Online behavior at Time X only predicts online behavior at Time X + 1 if (1) the
underlying population from which we are sampling remains the same and (2) the
medium itself remains the same. A changing media environment, under early adopter
conditions, violates both (1) and (2). Ceteris is not paribus. All else cannot be assumed
equal. One obvious consequence is that research findings are rendered obsolete by
the time they have been published. The terrain we can explore with traditional social
science techniques has narrowed in scope. (Karpf 2012:642)
In addition, time is important for working out the difficult issue of how we know where a social
media interaction begins and ends. Some of the parallel parameters we find in other modes, such
as ‘turn-taking’ (Sacks et al. 1974), that might be used to characterize an exchange are not
necessarily present in social media interactions. An interaction will look different depending on
the ‘node’ in the social network from which it is observed.
8
Some ability to track which user is
replying to, or rebroadcasting, a particular user is typically available in social media metadata.
However, representing the discourse structure of non-linear, multicast (many-to-many
broadcast) interactions is a problem that has yet to be solved. Indeed, linguists are still working
on how to represent dynamic exchanges in face-to-face interaction.
6 Future Directions
As network capabilities advance, social media is becoming more image- and video-intensive. Thus
comprehensive descriptions of social media texts will need to account for meanings made across
multiple modes, from written to spoken discourse. This will include meanings made via
paralinguistic systems such as gesture and image, for which we have a limited but emerging
multimodal SFL description. Social media research, particularly in the areas of individuation and
affiliation, will need to contribute to the development of a multimodal metalanguage for
theorizing and describing these resources.
8
Study of the properties of social media networks is undertaken in an area known as Social
Network Analysis (SNA).

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
References
Albury, K. 2015. Selfies, Sexts and Sneaky Hats: Young People’s Understandings of Gendered
Practices of Self-Representation. International Journal of Communication 9: 1734–45.
Argamon, S., M. Koppel, J. Pennebaker, and J. Schler. 2007. Mining the Blogosphere: Age, Gender,
and the Varieties of Self-expression. First Monday 12(9).
http://firstmonday.org/ojs/index.php/fm/article/view/2003/1878 last accessed, 15/8/18
Bakar, K. A. 2014. ‘A (Sensitive New Age Guy) with Difference’: Gendered Performances in Online
Personal Advertisements. Linguistics and the Human Sciences 9(1): 5–34.
Bakar, K. A. 2015. Identity in Online Personal Ads: A Multimodal Investigation. Asian Social
Science 11(15): 313.
Baker, P. 2006. Using Corpora in Discourse Analysis. London: Continuum.
Bakhtin, M. 1981. The Dialogic Imagination: Four Essays. Edited by M. Holquist. Austin: University
of Texas Press.
Bamman, D., J. Eisenstein, and T. Schnoebelen. 2014. Gender Identity and Lexical Variation in
Social Media. Journal of Sociolinguistics 18(2): 135–60.
Bednarek, M. 2009. Corpora and Discourse: A Three-Pronged Approach to Analyzing Linguistic
Data. In M. Haugh, K. Burridge, J. Mulder, and P. Peters, eds., Selected Proceedings of the 2008
HCSNet Workshop on Designing the Australian National Corpus. Somerville: Cascadilla
Proceedings Project.
Bednarek, M. 2010. The Language of Fictional Television: Drama and Identity. London:
Bloomsbury.
Bednarek, M. 2013. ‘There’s No Harm, Is There, in Letting Your Emotions Out’: A Multimodal
Perspective on Language, Emotion and Identity in MasterChef Australia. In N. Lorenzo-Dus and
P. Blitvi, eds., Real Talk: Reality Television and Discourse Analysis in Action. London: Palgrave
Macmillan. 88–114.
Bednarek, M. 2015. ‘What We Contrarians Already Know’: Individual and Communal Aspects of
Attitudinal Identity. In M. Charles, N. Groom, and S. John, eds., Corpora, Grammar, Text and
Discourse. Amsterdam: John Benjamins. 257-82

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
Bednarek, M. and J. R. Martin, eds. 2010. New Discourse on Language: Functional Perspectives
on Multimodality, Identity, and Affiliation. London: Continuum.
Benwell, B. and E. Stokoe. 2007. Discourse and Identity. Edinburgh: Edinburgh University Press.
Bernstein, B. 2000. Pedagogy, Symbolic Control and Identity: Theory, Research, Critique. Revised
edition. London: Taylor & Francis.
Boyd, D. 2014. It’s Complicated: The Social Lives of Networked Teens. New Haven: Yale University
Press.
Boyd, D. and K. Crawford. 2012. Critical Questions for Big Data. Information, Communication &
Society 15(5): 662–79.
Boyd, D. and N. B. Ellison. 2007. Social Network Sites: Definition, History, and Scholarship.
Computer-Mediated Communication 13(1): 210–13.
Boyd, D. and J. Heer. 2006. Profiles as Conversation: Networked Identity Performance on
Friendster. Paper presented at the 39th Annual Hawaii International Conference on System
Sciences.
Bruns, A. 2013. Faster Than the Speed of Print: Reconciling ‘Big Data’ Social Media Analysis and
Academic Scholarship. First Monday 18(10).
http://firstmonday.org/ojs/index.php/fm/article/view/4879 last accessed 18/8/18
Caldwell, D. 2010. Making Metre Mean: Rhythm and Identity in the Music of Kanye West. In M.
Bednarek and J. R. Martin, eds., New Discourse on Language: Functional Perspectives on
Multimodality, Identity and Affiliation. London: Continuum. 59–80.
Campoy, J. M. H. and J. A. C. Espinosa. 2012. Style-Shifting in Public: New Perspectives on Stylistic
Variation: Amsterdam: John Benjamins.
Caple, H. 2010. Doubling-Up: Allusion and Bonding in Multisemiotic News Stories. In M. Bednarek
and J. R. Martin, eds., New Discourse on Language: Functional Perspectives on Multimodality,
Identity and Affiliation. London: Continuum. 111–34.
Coupland, N. 2007. Style: Language Variation and Identity. London: Cambridge University Press.
Crawford, K. 2009. Following You: Disciplines of Listening in Social Media. Continuum: Journal of
Media & Cultural Studies 23(4): 525–35.
Crawford, K. 2013. Algorithmic Illusions: Hidden Biases of Big Data. Paper presented at the
O’Reilly Strata Conference, Santa Clara, California.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
Davis, K. 2011. Tensions of Identity in a Networked Era: Young People’s Perspectives on the Risks
and Rewards of Online Self-expression. New Media & Society 14(4): 634–51.
Derczynski, L., A. Ritter, S. Clark, and K. Bontcheva. 2013. Twitter Part-of-Speech Tagging for All:
Overcoming Sparse and Noisy Data. Paper presented at the Proceedings of the International
Conference on Recent Advances in Natural Language Processing.
Deumert, A. 2014. The Performance of a Ludic Self on Social Network(ing) Sites. In P. Seargeant
and C. Tagg, eds., The Language of Social Media: Identity and Community on the Internet.
London: Palgrave. 23–45.
Don, A. C. 2012. Legitimating Tenor Relationships: Affiliation and Alignment in Written
Interaction. Linguistics and the Human Sciences 5(3): 303–27.
Dreyfus, S., S. Hood, and M. Stenglin. 2011. Semiotic Margins: Meaning in Multimodalities.
London: Continuum.
Duguay, S. 2014. ‘He Has a Way Gayer Facebook Than I Do’: Investigating Sexual Identity
Disclosure and Context Collapse on a Social Networking Site. New Media & Society 18(6): 891–
907.
Eckert, P. 2000. Language Variation as Social Practice: The Linguistic Construction of Identity in
Belten High. Malden: Blackwell.
Firth, J. R. 1950. Personality and Language in Society. The Sociological Review 42(1): 37–52.
Firth, J. R. 1957. Papers in Linguistics, 1934–1951. London: Oxford University Press.
Gilpin, D. R. 2010. Working the Twittersphere: Microblogging as Professional Identity
Construction. In Z. Papacharissi, ed., The Networked Self: Identity, Community and Culture on
Social Network Sites. New York: Routledge. 232–50.
Grasmuck, S., J. Martin, and S. Zhao. 2009. Ethno-Racial Identity Displays on Facebook. Journal of
Computer-Mediated Communication 15(1): 158–88.
Halliday, M. A. K. 1978. Language as Social Semiotic: The Social Interpretation of Language and
Meaning. Baltimore: University Park Press.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
Hamid, B. D. A. and K. A. Bakar. 2010. Articulating Male and Female Adolescent Identities via the
Language of Personal Advertisements: A Malaysian Perspective. In A. Duszak and U. Okulska,
eds., Language, Culture and the Dynamics of Age. Berlin: Walter de Gruyter. 191–223.
Hasan, R. and J. Webster. 2009. Semantic Variation: Meaning in Society and in Sociolinguistics,
Volume 2. Sheffield: Equinox.
Highfield, T. and T. Leaver. 2014. A Methodology for Mapping Instagram Hashtags. First Monday
20(1). http://firstmonday.org/article/view/5563/4195 last accessed 18/8/18
Hodge, B. 2014. What if the Revolution Never Happened? A Thought Experiment in Linguistics
and Identity. Linguistics and the Human Sciences 9(1): 35–58.
Hogan, B. and A. Quan-Haase. 2010. Persistence and Change in Social Media. Bulletin of Science,
Technology & Society 30(5): 309–15.
Hood, S. 2011. Body Language in Face-to-face Teaching: A Focus on Textual and Interpersonal
Meaning. In S. Dreyfus, S. Hood, and M. Stenglin, eds., Semiotic Margins: Meaning in
Multimodalities. London: Continuum. 31–52.
Huang, J., K. M. Hornton, and E. N. Efthimiadis. 2010. Conversational Tagging in Twitter. Paper
presented at the Proceedings of the 21st ACM Conference on Hypertext and Hypermedia,
Toronto, Ontario, Canada.
Inako, A. 2013. Affiliation over Crisis: Physicists’ Use of Twitter Mode on Fukushima Daiichi NPP
Accident. Proceedings of JASFL 7: 1–13.
Karpf, D. 2012. Social Science Research Methods in Internet Time. Information, Communication
& Society 15(5): 639–61.
Knight, N. K. 2010. Wrinkling Complexity: Concepts of Identity and Affiliation in Humour. In M.
Bednarek and J. R. Martin, eds., New Discourse on Language: Functional Perspectives on
Multimodality, Identity, and Affiliation. London: Continuum. 35–58.
Knight, N. K. 2013. Evaluating Experience in Funny Ways: How Friends Bond through
Conversational Hum. Text & Talk 33(4–5): 553–74.
Kwak, H., C. Lee, H. Park, and S. Moon. 2010. What is Twitter, a Social Network or a News Media?
Paper presented at the 19th World-Wide Web (WWW) Conference, Raleigh, North Carolina.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
Lee, C. 2014. Language Choice and Self-presentation in Social Media: The Case of University
Students in Hong Kong. In P. Seargeant and C. Tagg, eds., The Language of Social Media: Identity
and Community on the Internet. London: Palgrave. 91–111.
Lemke, J. L. 2000. Across the Scales of Time: Artifacts, Activities, and Meanings in Ecosocial
Systems. Mind, Culture, and Activity 7(4): 273–90.
Leppänen, S., S. Kytölä, H. Jousmäki, S. Peuronen, and E. Westinen. 2014. Entextualization and
Resemiotization as Resources for Identification in Social Media. In P. Seargeant and C. Tagg, eds.,
The Language of Social Media: Identity and Community on the Internet. London: Palgrave
Macmillan. 112–36.
Liu, Y., C. Kliman-Silver, and A. Mislove. 2014. The Tweets They Are A-changin’: Evolution of
Twitter Users and Behavior. Paper presented at the International AAAI Conference on Weblogs
and Social Media (ICWSM), Ann Arbor, June 2014.
Lopez, L. K. 2009. The Radical Act of ‘Mommy Blogging’: Redefining Motherhood through the
Blogosphere. New Media & Society 11(5): 729–47.
Lu, H., J. C. Lin, K. Hsiao, and L. Cheng. 2010. Information Sharing Behaviour on Blogs in Taiwan:
Effects of Interactivities and Gender Differences. Journal of Information Science 36(3): 401–16.
Madden, M. 2012. Privacy Management on Social Media Sites. Pew Internet Report. Available
online at: www.pewinternet.org/2012/02/24/privacy-management-on-social-media-sites. (Last
accessed 16/09/2017).
Malinowski, B. 1972. Phatic Communion. In J. Laver and S. Hutcheson, eds., Communication in
Face-to-face Interaction. Harmondsworth: Penguin. 146–52.
Martin, J. R. 2000. Beyond Exchange: APPRAISAL Systems in English. In S. Hunston and G.
Thompson, eds., Evaluation in Text: Authorial Stance and the Construction of Discourse. Oxford:
Oxford University Press. 142–75.
Martin, J. R. 2002. Blessed are the Peacemakers: Reconciliation and Evaluation. In C. Candlin, ed.,
Research and Practice in Professional Discourse. Hong Kong: City University of Hong Kong Press.
187–227.
Martin, J. R. 2009. Realisation, Instantiation and Individuation: Some Thoughts on Identity in
Youth Justice Conferencing. DELTA: Documentação de Estudos em Lingüística Teórica e Aplicada
25(SPE): 549–83.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
Martin, J. R. 2010. Semantic Variation: Modelling Realisation, Instantiation and Individuation in
Social Semiosis. In M. Bednarek and J. R. Martin, eds., New Discourse on Language: Functional
Perspectives on Multimodality, Identity, and Affiliation. London: Continuum. 1–34.
Martin, J. R. and D. Rose. 2008. Genre Relations: Mapping Culture. Sheffield: Equinox.
Martin, J. R. and P. R. R. White. 2005. The Language of Evaluation. New York: Palgrave Macmillan.
Martin, J. R., M. Zappavigna, P. Dwyer, and C. Cléirigh. 2013. Users in Uses of Language: Embodied
Identity in Youth Justice Conferencing. Text & Talk 33(4–5): 467–96.
Marwick, A. 2011. To See and Be Seen: Celebrity Practice on Twitter. Convergence: The
International Journal of Research into New Media Technologies 17(2): 139–58.
O’Donnell, M. 2014. Exploring Identity Through Appraisal Analysis: A Corpus Annotation
Methodology. Linguistics and the Human Sciences 9(1): 95–116.
Page, R. 2012a. The Linguistics of Self-branding and Micro-celebrity in Twitter: The Role of
Hashtags. Discourse & Communication 6(2): 181–201.
Page, R. 2012b. Stories and Social Media: Identities and Interaction. London: Routledge.
Page, R. 2014. Hoaxes, Hacking and Humour: Analysing Impersonated Identity on Social Network
Sites. In P. Seargeant and C. Tagg, eds., The Language of Social Media: Identity and Community
on the Internet. London: Palgrave Macmillan. 46–64.
Pham, M.-H. T. 2015. ‘I Click and Post and Breathe, Waiting for Others to See What I See’: On
#FeministSelfies, Outfit Photos, and Networked Vanity. Fashion Theory 19(2): 221–41.
Rose, J., S. Mackey-Kallis, L. Shyles, K. Barry, D. Biagini, C. Hart, and L. Jack. 2012. Face It: The
Impact of Gender on Social Media Images. Communication Quarterly 60(5): 588–607.
Rustagi, M., R. Prasath, S. Goswami, and S. Sarkar. 2009. Learning Age and Gender of Blogger
from Stylistic Variation. In S. Chaudhury, S. Mitra, C. A. Murthy, P. S. Sastry, and S. K. Pal, eds.,
Pattern Recognition and Machine Intelligence. Berlin: Springer: 205–12.
Sacks, H., E. A. Schegloff, and G. Jefferson. 1974. A Simplest Systematics for the Organization of
Turn-taking for Conversation. Language 50(4): 696–735.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
Schler, J., M. Koppel, S. Argamon, and J. Pennebaker. 2006. Effects of Age and Gender on
Blogging. Paper presented at the Proceedings the Spring Symposium on Computational
Approaches for Analyzing Weblogs.
Schnoebelen, T. 2012. Do You Smile with Your Nose? Stylistic Variation in Twitter Emoticons.
University of Pennsylvania Working Papers in Linguistics 18(2): 117–25.
Schwartz, R. and G. R. Halegoua. 2014. The Spatial Self: Location-based Identity Performance on
Social Media. New Media & Society 17(10): 1643–60.
Seargeant, P. and C. Tagg, eds., 2014. Language and Social Media: Communication and
Community Online. London: Palgrave.
Stenglin, M. 2008. Olympism: How a Bonding Icon Gets Its ‘Charge’. In L. Unsworth, ed.,
Multimodal Semiotics: Functional Analysis in the Contexts of Education. London: Continuum. 50–
66.
Tann, K. 2010. Imagining Communities: A Multifunctional Approach to Identity Management in
Text. In M. Bednarek and J. R. Martin, eds., New Discourse on Language: Functional Perspectives
on Multimodality, Identity, and Affiliation. London: Continuum. 163–94.
Tann, K. 2013. The Language of Identity Discourse: Introducing a Systemic Functional Framework
for Iconography. Linguistics and the Human Sciences 8(3): 361–91.
Tian, P. 2008. Playing with ‘Femininity’: Multi-modal Discourse Analysis of Bilingual Children’s
Picture Book. The Ballad of Mulan. In Wu, C., Herke, M. and Matthiessen, C.M.I.M., eds.,
Proceedings of ISFC 35: Voices Around the World. Sydney: The 35th ISFC Organizing Committee.
308–15.
Twitter. 2013. Twitter About Page. Available online at: https://about.twitter.com. (Last accessed
16/09/2017).
Vásquez, C. 2014. ‘Usually Not One to Complain But …’: Constructing Identities in User-generated
Online Reviews. In P. Seargeant and C. Tagg, eds., The Language of Social Media: Identity and
Community Online. London: Palgrave Macmillan. 65–90.
Zappavigna, M. 2011. Ambient Affiliation: A Linguistic Perspective on Twitter. New Media &
Society 13(5): 788–806.
Zappavigna, M. 2012. Discourse of Twitter and Social Media. London: Bloomsbury.

Pre-print version only. Please do not quote this version. To appear as:
Zappavigna, M. (2019). Language and social media: Enacting identity through ambient
affiliation. In G. Thompson, W. Bowcher, L. Fontaine & J. Y. Liang [Eds.] The Cambridge
handbook of Systemic Function Linguistics. London: Cambridge University Press.pp 714-737.
Zappavigna, M. 2013. Tacit Knowledge and Spoken Discourse. London: Bloomsbury.
Zappavigna, M. 2014a. Enacting Identity in Microblogging. Discourse and Communication 8(2):
209–28.
Zappavigna, M. 2014b. Enjoy Your Snags Australia … Oh and the Voting Thing too #ausvotes
#auspol: Iconisation and Affiliation in Electoral Microblogging. Global Media Journal: Australian
Edition 8(2). https://www.hca.westernsydney.edu.au/gmjau/?p=1139 last accessed 18/8/18
Zappavigna, M. 2015. Searchable Talk: The Linguistic Functions of Hashtags. Social Semiotics
25(3): 274–91.
Zappavigna, M. 2016. Social Media Photography: Construing Subjectivity in Instagram Images.
Visual Communication 15(3): 271–92.
Zappavigna, M. 2018. Searchable Talk: Hashtags and Social Media Metadiscourse. London:
Bloomsbury.
Zappavigna, M. and J. R. Martin. 2014. Mater Dolorosa: Negotiating Support in NSW Youth Justice
Conferencing. International Journal for the Semiotics of Law / Revue internationale de
Sémiotique juridique 27(2): 263–75.
Zappavigna, M. and S. Zhao. 2017. Selfies in ‘Mommyblogging’: An Emerging Visual Genre.
Discourse, Context & Media. 20: 239-247
Zhao, S. 2010. Intersemiotic Relations as Logogenetic Patterns: Towards the Restoration of the
Time Dimension in Hypertext Description. In M. Bednarek and J. R. Martin, ed., New Discourse
on Language: Functional Perspectives on Multimodality, Identity, and Affiliation. London:
Continuum. 195–218.
Zhao, S. 2011. Learning through Multimedia Interaction: The Construal of Primary Social Science
Knowledge in Web-based Digital Learning Materials. PhD Thesis, University of Sydney.
Zhao, S. and M. Zappavigna. 2017. Beyond the Self: Intersubjectivity and the Social Semiotic
Interpretation of the Selfie. New Media & Society. 20: 1735-1754.