Category: Social theory

Paper for a special issue of the International Journal of Multilingualism entitled “Translinguistics: Negotiating Innovation & Ordinariness”

(eds. Jerry Won Lee & Sender Dovchin)

Jan Blommaert

1.Translingualism in the online-offline nexus

Three substantive claims underlie the argument in this paper.[1]One: in considering contemporary forms of translingualism one can neither avoid online sites of scripted interaction as loci of research, nor the online-offline nexus as an area of phenomenal innovation. Two: approaching such online forms of translingual interaction can benefit substantially from a radically action-centered approach, rather than from an approach privileging participants and their identity features, or privileging the linguistic/semiotic resources deployed in translingual events. And three: addressing online forms of translingual interaction from this perspective can reveal core features of contemporary social life and serve as a sound basis for constructing innovative social theory.

Of the three claims, the first one is by now widely shared (see e.g. Li Wei & Zhu Hua, this volume). There is an increasing awareness amongst students of language in society that the online social world has by now become an integrated part of the sociolinguistic economies of societies worldwide, and that the zone in which we situate our investigations should now best be defined as the online-offline nexus, with phenomena from the online world interacting with those of the offline world and vice versa. There are the specific rescaling and chronotopic features of online communication, where interaction is, as a rule not an exception, no longer tied to physical co-presence and effectively shared timespace; and where interactions as a rule not an exception include translocal and transtemporal rhizomatic uptake (cf. Tagg, Seargeant & Brown 2017; boyd 2014). And there are the outspokenly multimodal default characteristics of online communication. Taken together, it is evident that online communication must be the locus of intense translingualism. My first claim gestures towards the theme of this collection: the online-offline nexus must turn translingualism into the rule, the normal, ordinary and unremarkable sociolinguistic state of affairs.

The two other claims might demand somewhat more attention. The second claim – an action-centered perspective on online interaction – is grounded in (but transcends) a serious methodological problem complicating research: the indeterminacy of participant identities online. Given the widespread use of aliases and avatars on, for instance, social media platforms, nothing can be taken for granted regarding who exactly is involved in interactions. Whether we are interacting with a man or woman, a young or an old person, a local or nonlocal one, someone communicating in his/her ‘native’ or ‘first’ language: none of this can be conclusively established. This straightforward feature of online interactions destabilizes much of what we grew accustomed to in social studies, including sociolinguistic research. It makes us aware that our sociological imagination strongly hinged on the self-evident transparency of who people are, the communities they are members of, the languages that characterize them ethnolinguistically and sociolinguistically. The sociological sample – one of these key inventions of 20^th century social science – cannot be reliably drawn from online data.

Thus we find ourselves in a research situation in which little can be said a priori about participants and resources involved in social action. The action itself, however, can be observed and examined, and my second claim is to put the analysis of actions central in online-offline nexus research as a firm empirical basis for theory construction (cf. Szabla & Blommaert 2018). My third claim tags onto that: it is by looking at actions, and at how such actions effectively produce participants and resources, that we can get a glimpse of elementary patterns of social behavior through interaction – an opportunity for retheorizing our field. The target of this paper is to empirically demonstrate that.

I shall do so by looking at a common feature of online interaction: the use of hashtags, in this case on Twitter. The point I am seeking to make is that hashtags, as an entirely new feature in interaction interfering with established ones into a translingual whole, can be shown to be subject to rather clear and strict functions and norms of deployment. In Garfinkel’s (2002) terms, they can be shown to involve formatted actions with a high degree of normative recognizability, turning them into transparent framing devices in Twitter interactions.

2. Hashtags and translingualism

If we see translingualism (pace the editors of this collection) as the fluid movement between and across languages or – more broadly – semiotic systems, hashtags definitely can serve as prime instances of translingualism. As a feature of social media scripted discourse, the construction “# + word(s)” is a 21^st century innovation. Surely the sign “#” itself was used before the advent of social media: it was, for instance, a symbol on dial phones and was widely used elsewhere as a graphic symbol indicating numbers or, in old-school proofreading practices, indicating a blank space to be inserted in the text. But as we shall see, the social media use of hashtags cannot be seen as an extension of those previous forms of usage. When social media emerged, the hashtag was a free-floating resource that could be functionally redetermined and redeployed in a renewed sociolinguistic system. The fact that the symbol was not tied to a particular language or graphic system such as English or Cyrillic script made it, like the “@” sign, a polyvalent and user-friendly resource, capable of becoming part of global social media discursive repertoires – a process I called ‘supervernacularization’, (Blommaert 2012).[2] This means that such symbols can be incorporated – by translanguaging actions – in a nearly unlimited range of language-specific expressions while retaining similar or identical functions.

While the use of hashtags has by now become a standard feature of several social media applications (think of Facebook and Instagram) its usage is most strongly embedded in Twitter. Hashtags there tie together and construct topical units: within the strict confines of message length on Twitter, Hashtags enable users to connect their individual tweets to large thematically linked bodies of tweets. In that sense – but I shall qualify this in a moment – their function, broadly taken, is contextualization: individual tweets can be offered to audiences as understandable within the topical universe specified by the hashtag. Thus, the “#MeToo” hashtag (one of the most trending hashtags since the 2017 Harvey Weinstein scandal) ties together millions of individual tweets, produced in a variety of languages around the world, within the topical universe of gender-related sexual misconduct and abuse. As a consequence, within Twitter analytics, hashtags are used to define what is “trending” or “viral”, and other forms of big data mining on social media likewise use hashtags as analytical tools for modeling topics and tracking participant engagement and involvement (e.g. Wang et al. 2016; Blaszka 2012).

There is some work on what is called hashtag activism (e.g. Tremayne 2014; Bonilla & Rosa 2015; Jackson 2016; Mendes, Ringrose & Keller 2018) but qualitative sociolinguistic or discourse-analytic work focused on hashtags remains quite rare (but see e.g. Zappavigna 2012).In a recent study, De Cock & Pedraza (2018) show how the hashtag “#jesuis + X” (as in “#jesuisCharlie”) functionally shifts from expressing solidarity with the victims of the terror attack on the Charlie Hebdo editorial offices in Paris, 2015, to expressing cynicism and critique about hypocrisy when such forms of solidarity are being withheld from the victims of similar attacks elsewhere (as in “#jesuisIstanbul, anyone?”), or jocular and nonsensical uses as in “#jesuisCafard” (“I am a hangover”). Observe that the corpus used in De Cock & Pedraza’s study was multilingual, and that the “French” origins of “#jesuis + X” did not impede fluency of usage across language boundaries – the hashtag operates translingually.

We can draw a simple but fundamental insight from De Cock & Pedraza’s study: the functions of hashtags are unstable, changeable and dynamically productive. The same hashtag can be functionally reordered and redeployed whenever the topical field of the hashtag changes (or can be seen to be changing). In the analysis of De Cock & Pedraza, “#jesuis + X” shifts from an emblematic sign of (emotional and political) alignment to one of disalignment and even distancing. This shift in function instantiates mature enregisterment in that it offers different but related interactional stances to users; the hashtag “#jesuis + X” has become a lexicalized but elastic signifier enabling and marking a variety of forms of footing within a connected thematic domain (cf; Agha 2005). It is, to adopt Goffman’s (1975) terms now, a framing device, enregistered as such within a globally circulating and, of course, translingual, social media supervernacular. De Cock & Pedraza call the functions they described for the #jesuis + X hashtag “pragmatic”. As framing devices, however, hashtags are metapragmatic as well, they are interactionally established elements of voicing (Agha 2005). And the latter takes us to the core of my argument.

Functions of hashtags are interactionally established and should not be seen as simply the activation of latent and stable meaning potential. Seen from an action perspective, the different forms of footing enabled by a hashtag such as “#jesuis + X” represent different forms of communicative action within what Goffman called a “realm” – a “meaningful universe sustained by the activity” (1975: 46). At first glance, the difference between this formulation and the prior ones centering on contextualization, (dis)alignment and enregisterment seems minimal; in actual fact, the shift is quite substantial. We now move away from an analytical perspective focused on participants and resources (as in De Cock & Pedraza’s analysis) to one in which concrete actions are central and seen as the points from which both the participants’ roles and the values of the resources used in interaction emerge (cf. also Cicourel 1973; Garfinkel 2002; Goodwin & Goodwin 1992, 2004; Szabla & Blommaert 2018). Enregisterment, from this action perspective, does not only stand for the formation of registers-as-resources but also as the emerging of formats for communicative action, in which such formats also include the ratification of participants and the concrete mode of effective deployment of semiotic resources. Formats are framed patterns of social action, and I believe I stay very close to what Goffman suggested when I define framing as exactly that: the ordering of interactional conduct in ways that valuate both the roles of participants and the actual resources deployed in interaction between them.

3. #justsaying as action: basics

I will illustrate this by means of examples of the interactional deployment of the hashtag #justsaying. This hashtag – manifestly English in origin – is widely used on Twitter (also in variants such as #JustSayin, #justsayingg), also in non-English messages.[3] And contrary to most other hashtags, it is not a topical marker but an explicitly metapragmatic one. The expression “just saying”, in offline vernacular interaction, often indexes consistency in viewpoint and factual certainty in the face of counterargument (Craig & Sanusi 2000). Let us take a look at what can be done with it on Twitter, and concentrate on the types of action it can contribute to. In what follows, I shall use examples of #justsaying deployed in Dutch-language tweets from Belgium and The Netherlands, followed by approximate English translations. Note that there is no Dutch equivalent to #justsaying used on Twitter: it is a fully enregistered element in “Dutch” Twitter discourse.

I must first identify some basic actions performed and performable by means of #justsaying.

3.1. Standalone act

A first observation is that #justsaying is very often used for a standalone communicative act: a tweet which is not part of a Twitter “thread” (a series of interactionally connected tweets) but which appears as an individual statement, as in example 1.

Example 1: After weeks of only pictures about the heat, all media are now swamped with pictures and videos with rain, thunder and lightning. #justsaying

Those are standalone communicative acts, but evidently they are not without contextualization cues. In this tweet from early August 2018, the timing is the cue, as the author refers to the end of the heat wave that swept over Western Europe in that period. Contextualization can also take a more explicit shape, as when authors use topical hashtags tying their standalone statement into larger thematic lines (example 2).

Example 2: suggestion for #fgov … reinstate national service to enable our children to defend themselves against the aggressive #islam in our #europe. Matter of time before our #democracy has to be defended #manumilitari[4] #justsaying

In example 2, we saw that the standalone statement has an indirectly called-out and identified addressee, the Belgian Government, hashtagged as #fgov. Specific addressees can of course be directly called out through the use of the standard symbol “@”, and tweets by default have the author’s followers as audiences. Thus, a standalone communicative act does not equal a decontextualized act nor an act that doesn’t invite uptake from addressees. On social media, standalone communicative acts are interactional by definition, for the congregation of one’s Twitter followers (or a section thereof) will see the tweet on their timelines anyway, and they respond by means of “likes”, “retweets” or “comments”, as we can see in examples 1 and 2. I shall return to this point of addressee responses in greater detail below and underscore its importance.

The main point here is: such standalone tweets are, thus, framed in Goffman’s sense. They engage with existing “realms” and select participants. And what they do within such meaningful units and in relation to ratified participants is to signal a particular footing: a detached and self-initiated, sometimes implicitly offensive statement not directly prompted by the statements of others and often proposed as the start of a series of responsive acts by addressees. They trigger and flag from within a recognizable universe of meaningful acts (the registers we use on Twitter and the communities we use them with) a specific format of action involving particular forms of “congregational work”, the work we do in order to make sense of social actions and establish them as social facts (Garfinkel 2002: 245). We can paraphrase the format as:

“here I am with my opinion, which I state in a sober and detached way unprompted by others, and which I offer to you for interactional uptake”.

Let me stress this point once more: standalone acts such as those are not isolated or non-interactional, they are fully social acts performed in a collective of participants who know how to make sense of #justsaying action formats and their concrete contextualized instances. They merely initiate such action formats and, in that sense, provide an initial definition of their main ordering parameters.

3.2. Sidetracking and reframing

When #justsaying is interactionally deployed in a thread, we see partly different things. What remains stable is the sober and detached footing we encountered in the standalone instances. But very different formats of action are triggered and flagged by it. And before we engage with these formats of action, I must return to a particularly important feature of the examples that will follow: the duality of addressees. In a thread, an author responds directly to previous tweets and to those identifiable participants involved in those previous tweets. But the individual response tweet also attracts responses from other addressees: the likes and (sometimes) retweets and comments from participants not directly operating within that specific thread. Consider example 3.

Example 3: (response to @X and @Y): I’m not saying that something is wrong with large farms. Just pointing out that 200 cows are peanuts compared to the numbers in Canada. No attack. No judgment. #JustSaying[5]

While the author directly responds to two other participants (@X and @Y), her tweet receives a retweet and two likes from different Twitter users. This is important, for we see two separate lines of congregational work here: one line performed between the author and her two called-out and identified interlocutors, the authors of previous tweets; another line performed between the author and addressees not involved in the thread but responding, very much in the way described for standalone acts, to the author’s specific tweet. Two frames co-occur here, and this is important for our understanding of what follows.

A format of action frequently triggered and flagged by #justsaying in Twitter threads is “sidetracking”, or more precisely, opening a second line of framing. The thematic universe of the thread is disrupted by the introduction of another one, initiated on the same detached and sober footing as the standalone cases I discussed above (example 4):

Example 4:

(participant 1) Can anyone ask @X whether she can unblock me?

(participant 2, responding to participant 1) Me too … I don’t think I ever reacted against her … strange bitch

(participant 3, responding to participants 1, 2) Calling women ‘bitch’ seems to me to be cause for blocking. #justsaying

(participant 2, responding to participant 3) strange madam ok then?

The topic launched by participant 1 is not uncommon among active Twitter users: a complaint about being blocked by someone, @X, articulated here as an appeal to others to help being unblocked by @X. The direct response to this comes from participant 2, who endorses what participant 1 says by expanding the case: he, too, was blocked by @X, apparently for no good reason. In this response, participant 2 uses the term ‘bitch’ (‘wijf’), and this leads to the #justsaying reframing action by participant 3. From the actual case proposed by participant 1 as the topic of the thread, participant 3 shifts to an entirely different one related to the use of derogatory and sexist terminology within the moral framework of ‘proper’ Twitter usage. The shift, thus, is more than just topical: it reorders the entire normative pattern of interaction. Participant 2 immediately responds defensively by offering an alternative, only slightly less derogatory term. A new frame has been introduced and a new format of action – from collaborative work on one topic to oppositional work on another – has been started.

In opening a second line of framing, the participation framework is also redefined. In example 4, participant 1 is sidelined as soon as the #justsaying remark is made, and the direct interaction in the thread is reordered: it becomes a direct engagement of participant 3 with participant 2, and what started as a one-to-all thread becomes a one-on-one thread. A new line of action is generated by the #justsaying statement.

4. #justsaying as complex reframing

We have come to understand some of the basic actions in which #justsaying is used. Now look at example 5, an interaction started by the Mayor of Antwerp (participant 1 in the transcript) tweeting from his holiday site in Poland about the Gay Pride held in his town that day.[6] His tweet is meant as a public, one-to-all statement, and it has the expected effects: it goes viral with hundreds of “likes” and a large number of retweets. Apart from these forms of response, the tweet also develops into a thread: the Mayor gets several “comments” from participants addressed by his tweet.

Example 5

(Participant 1) I’m still in Poland but I wish all the participants in Antwerp a great Pride. [icon]Being yourself safely and freely, that’s what matters today. [icon]

(participant 2) I find the cultural promotion of extra-natural behavior not suited for a conservative party.

I have nothing agains LGBTs, have something against their bashers, but also against publicity.

(participant 2) I grant everyone their freedom, but I find the promotion of counternatural acts entirely unacceptable.

(participant 3) Let’s also prohibit publicity for traveling by plane then. People flying is a counternatural thing as well. To give just 1 example. But I’ll happily provide more examples if you wish. #justsaying #WearWithPride #antwerppride #NarrowmindedPeople

The Mayor’s public salute to the Antwerp Pride is critically commented on in two turns by participant 2, someone who clearly aligns himself with the right-wing conservative forces opposing the Pride. Observe that participant 2 addresses the Mayor in his responses and stays within the frame of the initial activity, and that his comments receive a number of likes as well as comments. The #justsaying comment by participant 3 – someone who identifies strongly with the Antwerp Pride through the use of a sequence of hashtags – is of particular interest, for it opens a new line of framing and reorders the participation framework. The Mayor is eliminated as a relevant direct addressee and the frame he started is dismissed, as the #justsaying statement by participant 3 is targeting the anti-LGBT turns made by participant 2. Thus, and very much like what we encountered in example 4 above, participant 3 gets a reply from participant 2 after his #justsaying statement:

Example 6

(participant 2) There are less people throwing up when they see a plane, than people feeling sick when they see homosexual acts.

(participant 3) Because it suits them well. The reason ensures that a message can be shared. Now that is zum kotsen (sic). Tells a lot about people. But feel free to move to Russia if it annoys you that much.

A new format of action has been started: an escalating, one-on-one fight between both participants, on the issue of what constitutes or doesn’t constitute “counternatural” conduct.

But there is more. Do note the different lines of congregational work here: while participant 3 enters into an argument with participant 2, his #justsaying statement gets eight “likes” and a retweet from Twitter users not otherwise active in this thread. So, parallel to the one-on-one thread developing within a one-to-all interaction started by the Mayor, another one-to-all thread emerges, inviting very different forms of response.

We see the full complexity here of the actions involved in reframing, and we can represent them graphically (Figure 1). On Twitter, what we see is a thread opened by the Mayor’s one-to-all tweet which triggers collective as well as individual responses, all of it within the frame initiated by the tweet (Frame 1 in figure 1). The thread, therefore, is a unit of action, but a composite and unstable one.[7] Because the #justsaying comment by participant 3 shapes, within the thread, a different frame (Frame 2 in figure 1). In Frame 2, we also see collective as well as individual responses – we see the same genres of action, in other words – but they are performed in a frame shaped by the #justsaying statement by participant 3. This frame is only indirectly related to Frame 1, and it draws participant 2 – who reacted initially within Frame 1 to the Mayor’s tweet – into a different role and position, with a different interlocutor and with (partly) different audiences, on a different topic. The reframing of the actions means that they are thoroughly reformatted: while, formally, the participants in Frames 1 and 2 appear to do very similar things, the difference in frame turns their actions into very different kinds of normatively judged congregational work, creating different social facts.

Figure 1: complex reframing actions in examples 5-6

What we see in this examples is how the hashtag #justsaying appears to “open up” a seemingly unified and straightforward activity to different forms of social action invoking, and thus proleptically scripting, different modes of participation and different modes of uptake, appraisal and evaluation. It interjects, so to speak, entirely different formats of action into a Goffmanian “realm”, enabling the shaping of very different “meaningful universes sustained by the activity”. As a framing device, #justsaying is thus more than a pragmatic-and-metapragmatic tool. It is something that proleptically signals various allowable modes of conduct and various forms of ratified participation and congregational work in social activities that appear, from a distance, simple and unified.

5. Hashtags and translingualism revisited

The latter remark takes us to fundamental issues in methodology. Many years ago, Goodwin & Goodwin (1992: 96) told us that “there are great analytical gains to be made by looking very closely at how particular activities are organized”. They made that point in a paper that demonstrated that what is usually perceived as one activity – a “conversation”, for instance – actually contains and is constructed out of a dense and complex web of distinct smaller actions, all of which have important contextualizing dimensions and many of which reorder the patterns of roles and normative scripts assumed by the participants. About participants, the Goodwins (2004) later also observed that the frequent use of generalizing category labels such as “speaker” and “addressee” again obscure important differences and shifts in the actual actions performed by participants in social interaction. One is not always an “addressee” in the same way during a speech by a “speaker”, for instance: sometimes one is a distant addressee, at other moments an involved one; one’s response behavior can be cool and detached at times and deeply engaged and emotional at others, positively sanctioning specific parts of the talk and negatively sanctioning others. The appeal launched (and continuously reiterated) by the Goodwins was for precision in analyzing social action as a key methodological requirement for discourse analysis, something they shared with the likes of Garfinkel and Goffman, and something that motivated my efforts in this paper. I tried to demonstrate that the interactional deployment of the hashtag #justsaying involved multiple and complexly related forms of social action, including the profound reframing of activities in such ways that morphologically similar actions (e.g. “likes” or comments) are formatted differently – they are part of different modes of making sense of what goes on.

The complexity of such discursive work, performed by means of a hashtag productive across the boundaries of conventionally established languages, to me demonstrates advanced forms of enregisterment and, by extension, of communicative competence (cf Agha 2005, 2007). This implies – it always implies – advanced forms of socialization, for enregisterment rests on the indexical recognizability of specific semiotic forms within a community of users who have acquired sufficient knowledge of the normative codes that provide what Goffman called “a foundation for form” (1975: 41). Translated into the discourse of translingualism, the complexity of discursive work performed by means of #justsaying demonstrates how translingual forms of this type have acquired a “foundation”, in Goffman’s terms, and operate as enregistered, “normal” features of semiotic repertoires within a community of users. Such users are able to recognize #justsaying (even across language boundaries) as indexing a shift in interactional conduct, introducing a different frame and allowing different forms of footing in what might follow. Translingual practice of this kind is an established social fact.

But recall the compelling appeal by the Goodwins: we must be precise here. The rules for such translingual practices as were reported here are not generic, they are specific to concrete chronotopically configured situations of social media communication: interactions on Twitter. The community of users, likewise, is ratified as competent in the use of such forms of discursive practice only within that area of social life – the valuation of their competence cannot be generalized or extrapolated without elaborate empirical argument. And so the translingual practice I have described here is a niched social fact, part (but only part) of the communicative economies of large numbers of people occasionally entering that niche.

The niche is new: at the outset of this paper I insisted that the use of hashtags in the way described here is a 21^st century innovation, an expansion and complication of existing communicative economies. Which is why I find it exceedingly interesting, for novelty means that people have to learn rules that are not explicitly codified yet; they have to actually engage in the practices and perform the congregational work required for an emerging code of adequate performance, in order to acquire a sense of what works and what doesn’t. They cannot draw on existing sets of norms of usage. My analysis of #justsaying has, I believe, shown that the use of hashtags cannot be seen as an extension and continuation of prior forms of usage of the symbol “#” – the symbol is used in ways that are specific to the social media niche that emerged in the last couple of decades, and the rules for its deployment are, thus, developed through congregational work performed by people who had no pre-existing script for its usage. As mentioned before, the value of semiotic resources (such as the hashtag) and the identities of its users (as competent members of a community of users) emerge out of the actions performed.

In that sense and from that methodological perspective, the use of hashtags directs our attention to fundamental aspects of the organization of social life, of meaning making, of interaction, and of language. There is room now for a theorization of translingualism in which, rather than to the creative bricolage of cross-linguistic resources, we focus on complex and niched social actions in which participants try to observe social structure through their involvement in situations requiring normatively ratified practice – I’m paraphrasing Cicourel (1973) here – in emerging and flexible communities populating these niches of the online-offline nexus.

References

Agha, Asif (2005) Voice, footing, enregisterment. Journal of Linguistic Anthropology 15/1: 38-59.

Agha, Asif (2007) Language and Social Relations. Cambridge: Cambridge University Press.

Blaszka, Matthew, Lauren Burch, Evan Frederick, Galen Clavio & Patrick Walsh (2012) #Worldseries: An empirical examination of a Twitter hashtag during a major sporting event. International Journal of Sport Communication 5:4: 435-453.

Blommaert, Jan (2012)Supervernaculars and their dialects. Dutch Journal of Applied Linguistics 1/1: 1-14.

Blommaert, Jan (2018) Durkheim and the Internet: On Sociolinguistics and the Sociological Imagination. London: Bloomsbury.

Bonilla, Yarimar & Jonathan Rosa (2015) #Ferguson: Digital protest, hashtag ethnographyand the racial politics of social media in the United States. American Ethnologist 42/1: 4-17.

boyd, dana (2014) It’s Complicated: The Social Life of Networked Teens. New Haven: Yale University Press

Cicourel, Aaron (1973), Cognitive Sociology: Language and Meaning in Social Interaction, Harmondsworth: Penguin Education

Craig, Robert & Alena Sanusi (2000) ‘I’m just saying…’: Discourse markers as standpoint continuity. Argumentation 14/4: 425-445.

De Cock, Barbara & Andrea Pizarro Pedraza (2018) From expressing solidarity to mocking on Twitter: Pragmatic functions of hashtags starting with #jesuis across languages. Language in Society: 47/2: 197-217.

Garfinkel, Harold (2002) Ethnomethodology’s Program: Working Out Durkheim’s Aphorism. Lanham: Rowman & Littlefield.

Goffman, Erving (1975 [1974]) Frame Analysis: An Essay on the Organization of Experience. Harmondsworth: Penguin.

Goodwin, Charles & Marjorie Harness Goodwin (1992) Context, activity and participation. In Peter Auer & Aldo DiLuzio (eds.) The Contextualization of Language: 77-99. Amsterdam: John Benjamins.

Goodwin, Charles and Marjorie Harness Goodwin (2004), Participation. In Alessandro Duranti (ed.), A Companion to Linguistic Anthropology, 222–44, Malden: Blackwell

Hymes, Dell (1996) Ethnography, Linguistics, Narrative Inequality: Toward an Understanding of Voice. London: Taylor & Francis.

Jackson, Sarah J. (2016) (Re)imagining intersectional democracy from black feminis to hashtag activism. Women’s Studies in Communication 39/4: 375-379.

Mendes, Kaitlinn, Jessica Ringrose & Jessalyn Keller (2018) . #MeToo and the promise and pitfalls of challenging rape culture through digital activism. European Journal of Women’s Studies 25/2:236-246.

Szabla, Malgorzata& Jan Blommaert (2018) Does context really collapse in social media interaction? Applied Linguistics Review(in press).

Tagg, Caroline, Philip Seargeant, Philip & Amy Brown (2017). Taking Offence on Social Media: Conviviality and Communication on Facebook. London: Palgrave Pivot

Tremayne, Mark (2014) Anatomy of protest in the digital era: A network analysis of Twitter and Occupy Wall Street. Social Movement Studies 13/1: 110-126.

Wang, Yuan, Jie Liu, Yalou Huang & Xia Feng (2016) Using hashtag graph-based topic model to connect semantically related words without co-occurrence in microblogs. IEEE Transactions on Knowledge and Data Engineering 28/7: 1919-1933.

Zappavigna, Michele (2012) Discourse of Twitter and social media: How we use language to create affiliation on the Web. London: Continuum.

Notes

[1]I dedicate this paper to the memory of Charles Goodwin, a source of inspiration and an engaging interlocutor for several decades, who sadly passed away while I was developing the analysis reported here. This paper is part of a project I call “Online with Garfinkel”, in which I explore the potential of action-centered analyses of online-offline communication. A precursor of the project is Blommaert (2018). ACKNOWLEDGMENTS.

[2] The point that the widespread availability of online technologies has reshaped the sociolinguistic system is missed by some critics of notions such as translanguaging, who point to the prior existence of formally similar or identical forms of language and/or script to argue that there is nothing ‘new’ happening. In such critiques, Hymes’ (1996) important warning is disregarded: that the study of language is not merely a study of the linguistic system – the formal aspects of language, say – but also and even more importantly the study of the sociolinguistic system in which language forms are being distributed, functionally allocated and deployed in concrete social circumstances. The arrival of the internet has caused a worldwide change in the sociolinguistic system, provoking enormous amounts of sociolinguistically new phenomena. And even if such phenomena have linguistic precursors, they do not have any sociolinguistic ones. See Blommaert (2018) for a discussion.

[3] I collected a small corpus of #justsaying examples from my own Twitter account between March and August 2018 (N=186), and found the hashtag incorporated into English, Dutch, Danish, Spanish, Hindi, Bulgarian and Arabic tweets. Hashtags are also (and increasingly) used offline in marches and other forms of public demonstrations, and in advertisements.

[4] “fgov” is the Twitter name of the Belgian Federal Government; “manu military” means “by the use of military force”. The author of this tweet is a former MP for a Flemish extreme right-wing party.

[5] One can note the explicit description of the footing for #justsaying statements here: “No attack. No judgment. #JustSaying”.

[6]The Mayor is a controversial, very outspoken right-wing politician. The “victory” icon he posts at the end of his tweet is a campaign emblem of his party, and the phrase “being yourself safely” is a direct reference to the Mayor’s re-election program.

[7] In Szabla & Blommaert (2018) we analyzed a long discussion on Facebook and called the entire discussion (composed of the update, comments and subcomments) the “main action”. In a more traditional sociolinguistic vocabulary, one can also see the overall unit of action the “event”.

Jan Blommaert, Laura Smits, Noura Yacoubi

Draft chapter, Handbook of Discourse Studies (eds. A. De Fina & A. Georgakopoulou). London: Routledge 2019.

Abstract: The point of departure for this chapter is the contention that the online-offline communicative economy in which we now live compels us to rethink some of the core vocabulary and assumptions underlying our thinking about ‘context’ and ‘contextualization’ in discourse studies. We formulate a set of proposals grounded in the interactionist tradition and focused on (inter)action rather than on participants and communities. Next, we propose to see contextualization as a process that takes us from chronotopes over frames to formats of action, situationally deployed.

Introduction: Online-offline action

In his classic Cognitive Sociology, Aaron Cicourel made the following general observation:

The problem of meaning for the anthropologist-sociologist can be stated as how members of a society or culture acquire a sense of social structure to enable them to negotiate everyday activities. (Cicourel 1974: PAGE)

This statement can serve as an extraordinarily accurate description of what was later called and methodologically developed as “contextualization” (Gumperz 1982, 1992; also Auer & DiLuzio 1992; Duranti & Goodwin 1992). Yet two components of the statement demand closer attention: “social structure” and “everyday activities”, for since the beginning of the 21^st century the realities of social structure and the range and modes of everyday activities have been profoundly affected by the generalized introduction of a layer of online social life, complicating the offline social world on which these earlier formulations of contextualization were based. In this chapter, we intend to sketch the complications emerging from discourse produced interactionally in such an online-offline environment now serving as the backdrop for what Ron Scollon (2001) called “the nexus of practice”.

We must pay closer attention to the aspects of contextualization that have changed, we said, but that does not mean that we must do so from within a methodological tabula rasa. We believe the effort can be profitably made by means of some central insights and principles from within the interactional tradition of discourse studies. In fact, all the scholars already mentioned here belonged to that stream of ethnographically grounded studies of actual situated discursive practice, which has been the richest source of fundamental reflections on the notion of context and its role in social interaction. It is from that source that we can draw the general principles directing our discussion:

1. Context should not be seen as an abstract, stable or latent presence; it is a resource deployed in concrete socially situated meaning-making action: context is always contextualization.In that sense, it is highly unpredictable, evolving, dynamic and unstable. Also, while contexts operate at various scale-levels and structures a multitude of concrete interactions, the analytical point of departure is their situated effects on making sense. To quote Herbert Blumer in this respect: “People (…) do not act toward culture, social structure or the like, they act toward situations” (Blumer 1969: PAGE)
2. Contextualization is the key to making sense, because it consists of interactionally constructed indexical connections between actual discursive features and relevant chunks of sociocultural knowledge (Silverstein 1992; Hymes 1996; Gumperz 2003; Agha 2007).
3. Such indexically deployed and invoked knowledge is never neutral but always evaluative and in that sense moral, and by extension identity-related (e.g. Goodwin 2007). Making sense is a moral judgment grounded in socioculturally available normative-behavioral scripts situationally projected onto persons. Goffman (1974) called such moralized scripts “frames”; the ways we implement them have been variously called (with distinctions not overly relevant here) “indexical order” (Silverstein 2003) and “orders of indexicality” (Blommaert 2005). The concepts are joined by their emphasis on (Bakhtinian) evaluative uptake and on the dimension of social order as part of meaning-making practices – recall Cicourel’s statement quoted above.
4. The contextual resources that people draw upon in interaction have to be recognizable, but not necessarily shared (Garfinkel 2002; also Blommaert & Rampton 2016: 28-31). Sharedness is evolving as the interaction proceeds but can also evolve as a shared sense of misunderstanding, i.e. a shared sense that very little of substance is shared in the interaction. What needs to be recognizable is the broad outline of a format of interaction, a general script for social action.

We can see that these principles favor action over content and participants, and situated and evolving effects over a priori categories (such as speech acts, conversational maxims, “meaning” and “understanding”). The reason for these preferences is that, due to the changes described above, very little can be taken for granted with respect to what is “ordinary” and “normal” in communication. To name just a few of the widely used assumptions that need to be qualified: the assumption that communication is self-evidently a human-to-human activity has been challenged by human-machine interactions, and has thus become a variable rather than a stable feature. This, of course, has numerous knock-on effects on widely used criteria in theories of meaning: intentionality, agency, (human) rationality. Even more widespread is the assumption that the most “normal” or primitive form of communication – in the sense of: the kind of communication on which we base our fundamental theoretical imagination – is unmediated, spoken dyadic face-to-face interaction in shared physical timespace and between persons sharing massive amounts of knowledge, experience and sociocultural norms within a sedentary community (an offline conversation between similar people, in short). The online world has critically destabilized that assumption by inserting scripted, multimodal, non-simultaneous, translocally mobile, multiparty and technologically heavily mediated forms of everyday communication into the communicative economies of very large numbers of people, not as peripheral modes of interaction but as important, inevitable ones. We now communicate intensely with interlocutors with whom we do not share much (not even acquaintance as a natural person and a human subject), across space and time, and through complex modes of non-acoustic semiotic work.

Our core vocabulary and assumptions derive from an implicit sociological imagination of which we assume that it reflects the true state of things. Changes in the state of things often take some time before they translate into an alternative sociological imagination (cf. Mills 1959; also Blommaert 2018a). In the meantime, however, they render some of our core vocabulary for talking about language, interaction and meaning-making less salient and applicable, and invite a focus on the phenomena we can identify as constants. The constant feature, we would argue, is social action – a synonym, as Anselm Strauss (1993) among others emphasized, for interaction. Even if we now communicate with machines, with unknown mass audiences (as in mass online gaming), by means of delayed, asynchronous messages scripted in new forms of graphic visualization and design – we are still performing interactions in attempts to make sense of our world. Taking social action, defined in this sense, as our ontological point of departure enables us to start describing and understanding old and new patterns of interaction, how they intersect and how they structure our social lives.

With these principles established, we shall now engage with four different sets of issues, all of them inspired by the changes we noted earlier: the transition from an offline world of communication to an online-offline one. Some of these issues are not new – they have been constant features of debates on context and contextualization – but demand a renewed engagement in view of changes in the world of communication. In reviewing them, we will make proposals for reimagining aspects of them and for adopting another vocabulary in our descriptions of them.

Beyond the macro and micro: recognizability and formatting

A persistent feature of discussions of context and its uses in scholarship is the use of the “micro-macro” dichotomy (occasionally turned into a triad by inserting “meso” in between). “Micro”-contexts would then be the factors affecting and informing local, situated events: the timespace frame, the participants, the immediately and directly relevant social roles, the topic, and so forth. “Macro”, in turn, would stand for the nonlocal, broader factors in which the event can be situated and by which it is indirectly affected: the wider historical, sociocultural and political parts of the picture making (at least part of) the event understandable (see the discussion in e.g. Goffman 1964; Silverstein 1992;Cicourel 1992; Duranti 1997; Blommaert 2015a).

While such distinctions might be discursively and heuristically helpful, they are methodologically unhelpful from the perspective we formulated above. They do point to a fundamental fact: the non-unified and complex nature of context – any context – but they do so in an inaccurate way. Certainly when we become aware of the ways in which they rest on a particular sociological imagination, and of the ways in which and structure an epistemological field. The sociological imagination on which the dichotomy between “micro” and “macro” rests is the one sketched earlier: a world in which we can separate and isolate specific aspects of social life as being the direct conditions for conduct – the local, sedentary, individual, variable and mundane aspects – while other aspects appear to only indirectly inflect such conditions for conduct, due to their remoteness and their stable, collective character. The first set of facts we could call “processual” factors, and they would always be unique, while they others would be “procedural”, and they would be general. The first set would index “community” – a specific small-scale group involved in shared practices, but diverse and changeable – while the latter would index “society” – the organized, stable, enduring, systemic large-scale group characterized by common institutional characteristics. Obviously, this imagination of the social world is far removed from what Castells (1996), in a visionary text, called the online-offline “network society” (cf. also Blommaert 2018a).

The dichotomy between “micro” and “macro” also structures an epistemological field in which “micro” would stand for the anecdotal, the concrete, the singular, the possible exception, the empirical and the “token”, while “macro” would point to the systemic, the abstract, the generalizable, the norm, the theoretical and the “type”. Thus, so-called “micro-sociologists” and ethnographers would be dismissed as scholars whose attention to the uniquely situated features of cases precludes any attempt towards valid generalization, because generalization can only be made at a “macro” level of analysis where analytical detail has to be surrendered to abstraction (see the discussions in e.g. Mills 1959, Blumer 1969, Giddens 1984).

From a viewpoint privileging social action, all of this is highly unproductive, and the acuteness of the problem was repeatedly emphasized by Pierre Bourdieu and others. Bourdieu – often seen as a “macro”-sociologist whose work speaks to society at large – would emphasize that concepts such as “habitus” (a general concept) could only emerge through ethnographic attention to actual situated practice, not by statistical surveys. It was by observing the struggles of Algerian farmers to come to terms with a new market economy that Bourdieu saw the actual working of capitalism as propelled into socioculturally inhabited modes of practice (Bourdieu 2000; cf Blommaert 2015b). The big things reside in the small things, and the most inconspicuous and uniquely situated social action is, in that sense, “systemic” and “typical”, as well as the source for theoretical generalization. Evidently, the same insight animated Goffman’s work on interactional ritual and frames (1967, 1974): even if all instances of human interaction are unique, they display general characteristics and patterns sufficient to lift them from “micro” to “macro” relevance (cf. also Rampton 2016).

This is the point where we can start formulating a proposal in line with the principles outlined above. And we can draw for inspiration on the authors just mentioned, as well as on Garfinkel’s (2002) uncompromising formulations of the issue. Garfinkel saw recognizability as the key to understanding the social nature of interaction, and as already mentioned above, recognizability should not be equated with sharedness of norms, assumptions and worldviews. It is a recognition of the joint potential for specific modes of action that gives such action the character of “congregational work”, he argued – work performed collectively because we are jointly involved in it. We enter jointly into an action of which we know very little outside its possible general features, and we jointly construct such actions as forms of social order. This order can be entirely ad hoc, temporary and ephemeral. But while it lasts, it is a firm order that generates roles and identities along with a range of moral codes controlling (mis)behavior.

It is this aspect of recognizability, generating congregational work and its social outcomes, that renders distinctions between the factors discriminating “micro” and “macro” aspects of the act meaningless. Since acts are social, they will draw on available and accessible social resources – from the different social positions from which we enter the action, the kinds of language and discourse we use, over the topic, to the actual things we say, hear, write or read (cf. also Briggs 2005). And even if we see that such resources are unevenly distributed, a degree of order will emerge from the action itself. The latter was exemplified in a magnificent study by Charles Goodwin (2004), in which a man who,following a stroke, had lost almost all of his linguistic capabilities was shown to engage in lengthy and complex interactions with his friends and relatives. Evidently, the absence of shared linguistic resources imposes constraints on what can happen in such forms of interaction – resources are crucial contexts for interaction (Blommaert 2005: 58-62); but when we intend to understand what is happening, recognizability is the key.

Recognizability, however, is not an empty and random container. We recognize particular social situations and their features as something specific – a quarrel, a lecture or a Facebook update – on the basis of perceived properties of the situation (what Garfinkel  called ”autochthonous order properties”, 2002: 245) associated with what Goffman called “frames”: the ways in which we organize our experience. Recognizing a situation means framing it along what we could call a general indexical vector, i.e. entering that situation as one that imposes and enables specific forms of interaction, one or different orders of indexicality. When we recognize something as a Facebook update, we recognize that it enables (among other actions) different forms of response, and that it imposes keyboard writing and a specific set of symbols (e.g. emoticons) as techniques for responding to it. When we recognize the particular update as an instance of trolling, we recognize it as enabling an unfriendly response, and so forth. This we can call, following Garfinkel, formatting: shaping the particular situated interaction in “typical” (i.e. generic, non-unique) ways and bringing the “sense of social structure” mentioned by Cicourel into the particular action we are engaged in with others.

A lot of what we do in the work of contextualization is moving from recognition through framing to formatting. We do so dialogically in congregational work with others, and we do so by drawing upon socioculturally marked – indexicalized – resources that acquire a general direction in such activities. This, we propose, is the cornerstone of the argument here. We can now proceed to elaborate it further.

Chronotopes, scales and synchronization

In every moment of interaction, contextualization draws upon specific and non-unified resources (cf. Cicourel 1967, 1974; Silverstein 1992). Both dimensions are crucial if we wish to avoid undue simplifications such as “the context for this utterance is X”. The contextual resources drawn upon in contextualizing concrete interactions are inevitably multiple and layered (cf. Blommaert 2005). But they are not infinite, and not without structure and pattern. If we draw upon Goffman’s frames, we see that social experience is organized into such structures and patterns, in which particular forms of interaction – with attributes to be discussed in a moment – are attached to specific social situations in forms comparable to what Bakhtin called “chronotopes”.

Bakhtin developed the notion of chronotope (literally “timespace”) as a way to describe the sociohistorical layering in novels, more precisely the ways in which invocations of particular sociohistorical frames structured “voices” in specific situations, infusing them with identity scripts, moral orientations, participation frameworks (Goodwin & Goodwin 1992), expected and unexpected normative modes of conduct and roles within the situation – in short, the full sociocultural value of otherwise random forms of action (see the discussion in Blommaert 2015a and Blommaert & De Fina 2016). Thus in a fairy tale, the Big Bad Wolf is exactly that: male, big and bad, a threat to the others, and someone to be defeated by the others. Chronotopes, seen from this rather orthodox Bakhtinian perspective, provide moralized behavioral scripts in specific social situations (we called them formats above), and the recognition of social situations as specific (e.g. as a formal meeting) will prompt such scripts: as soon as the chair announces the beginning of the meeting, we will all reorganize our conduct, assume a different set of body poses, discursive patterns and relations with the other participants (e.g. respecting the chair’s formal leadership and the differential allocation of speaking rights), and align with the congregational work performed by the others. As soon as the meeting is over, we can shift back into another register of conduct, and the opponent during the meeting can turn into an ally in the pub during the post-meeting drink. Chronotopes impose formats on those inhabiting them, and this means that from the potentially infinite aspects of context animating events, a specific subset will be invoked and deployed as the normative script for conduct within that chronotopic situation, as the specific bit of social order to be followed by all those involved. Violating or disrupting that order – Goffman called them frame breaking – comes with moral judgments: everyday notions such as inappropriateness, rudeness, insolence, being off-topic, or trolling come to mind (cf. Blommaert & De Fina 2016; see Tagg, Seargeant & Brown 2017 for social media examples).

Chronotopes are, we believe, a useful gloss to address the specific nature of context and contextualization, one that forces us to examine with utmost precision what is elsewhere simply called “the context” of actual interactions. The notion also offers us a view of context as active, something that structures action and makes it socially recognizable and, thence, socially valued. The demand for precision will almost inevitably lead to outcomes in which particular chronotopes are

• composed of several different actions and types of actions, as when someone checks his/her email or takes orders for sandwiches during a formal meeting – where each of these will have to proceed along the specific formats for such actions. Goffman (1974: 561) clearly pointed to that (using the term “realm statuses” for what we call formats here), and see e.g. Goodwin (2013) for excellent discussions.
• Connected to other chronotopes, as when the relations between participants in a formal meeting are affected by already existing interpersonal relationships specific to other areas of social life or when the history of a particular issue is invoked as a frame for discussing its present status, or even when quoted or indirect speech is introduced into interaction embedding one chronotope and its actual voices into another one (e.g. Voloshinov 1973; Goodwin 2003).

Both outcomes are particularly interesting, for they take us to the issue of the non-unified nature of context and bring issues of scale into view (cf. Blommaert 2015a). Scale can best be understood as reflections and expressions of how social beings experience dimensions of sociocultural reality as indexical vectors, as informing the general normative patterns that shape formats of action (cf. Das  2016; Carr & Lempert 2016). Scales, thus, are interpretive and normative-evaluative, suggesting distinctions between what is general and what is specific, what is important and what is not, what is widely known and what isn’t, what is valid and what is not, what can be widely communicated and what cannot, what can be widely recognized and understood and what cannot. There is nothing stable, absolute or a priori about scales – we can obliquely recall our discussion of the “micro-macro” distinction here – for what we see in actual discursive work are scalar effects. To give a simple example: when the history of a particular issue is invoked as a decisive argument in discussing its present status, then that history is presented as a way of upscaling the current issue to normative levels immune to contemporary petty or personal concerns (“We already discussed and decided this point in January, there is no point in returning to it now!”) Conversely, when someone raises a point which is not seen by others as belonging to the most general normative layer of what goes on, it can be downscaled (“This is a detail” or “This is just your personal opinion”). In their actual deployment, scalar effects are indexically ordered degrees of moralization in social actions.

The presence of such non-unified (plural and scaled) contexts in concrete situations brings us to a third notion: synchronization (cf. Blommaert 2005: 131-137, 2018b). The scalar effects we just mentioned occur in real-time and on-the-spot moments of interaction, in a sort of evolving “synchrony” which hides layers of non-synchronous resources and folds them together into momentary and situated instances of making sense. We call this process synchronization because the highly diverse resources that are deployed as context are focused, so to speak, onto one single point in social action. In other branches of scholarship this process would be called “decision making”, with strong undertones of individual rational calibration. From an action-centered perspective, synchronization is a collaborative social act in which the format, not the rational calculation of its actors, is predominant (cf. Goodwin 2013).

Within such formats, synchronization ensures the degree of coherence we expect to find in interactions as an essential component in making sense of situations.

Formatting and nonlinear outcomes

Coherence, however, must not be imagined as a straight line from premises to conclusion. Neither can formats be imagined as closed boxes with extraordinarily transparent orders of indexicality, generally known to all participant. As said earlier, order is evolving and contingent upon the congregational work performed by participants. Recognizing a situation, we explained, proceeds through perceived order properties of such situations that can be framed into formats, then guiding the actions of participants. But outcomes cannot be linearly predicted from the starting conditions, because multiple forms of action can emerge within the same format, and be coherent to the participants. In other words, different kinds of actions can be ratified as properly within the format; formats allow nonlinear actions, and when it comes to normativity in connection to formats, we see a relatively open and relaxed form of normativity there.

This violates several older assumptions about communication. In speech act theory, J.L. Austin famously distinguished clear “felicity” conditions for smooth and “correct” interaction, while deviations of them (even a violation of one of them) would make the interaction “unhappy”, or “infelicitous”.  Equally famous are Grice’s (1975) “maxims” for conversation – conditions for maintaining a well-ordered mode of interaction with any other interlocutor. Both (and many others) grounded their theories into widely shared folk views of the strong normative order required for interaction. Another set of assumptions that is violated by the nonlinearity-within-the-format we mentioned is that underlying the kind of naïve survey methodology devastatingly criticized by Cicourel (1964) and others. In such survey enterprises, the stability of the format is used as an argument for the stability of its outcomes. Concretely, it is assumed that as long as we ask the same questions in the same format to large numbers of respondents, the answers will be commensurable because each respondent was addressed identically. Converted into the terms we are using here, stable formats will generate linear actions, since every action will be an identical response to an identical prompt. Cicourel’s penetrating critique targeted the impressive amount of ignorance about actual forms of communication buried inside this methodological assumption, leading to the incredible suggestion that hundreds of different people would all have identical understandings of a question (and its meanings for the analyst) and that the actual (and highly diverse) conditions of the question-and-answer events would not have any effects on the respondents.

The fallacies of such assumptions can be shown through the following example, involving the present authors. In late 2017 Jan Blommaert set up a small practical exercise in research interviewing for MA students including Laura Smits and Noura Yacoubi. The instructions were clear: pairs needed to be formed and the roles of interviewer versus interviewee needed to be assigned; the interview was to proceed in English and (unbeknownst to the interviewees), the interview had to contain some potentially frame-disturbing elements. One of these elements was the opening question: “who are you really?” The format, we can see, was entirely scripted and uniform for all the teams.

Laura and Noura were both interviewees and were interviewed by classmates with whom both had a history of friendly personal encounters and lengthy conversations – in Dutch. All of them – interviewers and interviewees – were also students in the same year of the same program track at Tilburg University. Thus we can suppose other elements of potential stability to be there: shared membership of a clearly defined community, a shared history of interaction making all participants familiar with each other’s speech habits and idiosyncrasies, and also enabling all to know quite well who the other “really” was. Laura and Noura, however, responded to this question in radically different ways. Let us look at the sequences following the question; in the transcript “I” stands for “interviewer” and “R” for “respondent”.

Laura’s answer

I:          SO Laura*, who are you REALLY?

R:        Who are I (am) really.. Eu::hm. What do you want to know of me. What is–what is really?

I:          TELL me something about yourself

R:        Okay. I’m Laura .. Laura Smits .. I a::m twenty-three years old .. eu::hm.. I study Global Communication here at Tilburg university I play volleyba::ll I have a little sister, I have a boyfriend, and I live in Tilburg eu::hm furthermore<1> I think<1> I am very happy at the mome::nt in the situation I live in .. eu::hm ja* enjoying life/ …

I:          Okay.

Noura’s answer

I:          Uhm .. who are you really?

R:        Who I am?

I:          Yes

R:        Well.. what do you mean? What do you want to know?//

I:          Yeahh who are you?//

R:        That is a.. difficult question [Laughing]/

I:          Why is it difficult?//

R:        Because you are asking *a lot* at the same time. Do you want to know my characteristics, my name, my birth, my hobbies, do you want to know my study?

I:         Tell me what *you think* who you really are//

R:        *Dude* [Laughing] well I am a… Dutch, well Moroccan-Dutch girl, born here, I’m uuhh 22 years old. Uuhm who I am? <2> Well I am a student that is part of my identity, I *feel* as a student, I am.. living the life of a student. Uhmm.. I am studying global communication/

I:         Ohh

R:        What a coincidence [Laughing]

I:         Me too [Laughing]

R:        Can you ask.. can you ask the question more specific?//

I:         Is this really who you really are?

R:        Well it’s uhm.. it is quite a lot who I am I mean.. also history comes into pla::y, also family comes into pla::y uuh who I am yeah I am a human being//

I:         Okay but/

R:        Punt

We see that Laura and Noura are both initially looking for the right frame, as both ask for clarification of their interviewers’ actions (“what do you want to know”?). Both, consequently, receive a reiteration of the question (part of the instructions given by Jan to the interviewers). But what follows are two entirely different courses of action. Laura instantly aligns with the perceived frame and gives what we could call a “profile answer” – the kind of clearly organized factual and affective information offered on social media profiles and in short introductory “pitches” to unknown people. She “neutralizes”, so to speak, the interviewer whom she considered to be a close friend, and addresses her in her role as an interviewer performing an unusual kind of interaction, which in the same move is “normalized”: this is an interview, it’s strange, but we’ll do it the way it should be done. The synchronization towards the format is complete in Laura’s case. Noura, by contrast, does not exit the interpersonal and intertextual frame, but engages in several turns of metapragmatic negotiation with the interviewer (also someone with whom she maintained a very friendly personal relationship), expressing discomfort and resistance to align with the format in utterances such as “dude” and “punt” (meaning “period”, “that’s it”). And while she does offer a kind of “profile answer” at some point, the answer is followed by a repeated request for clarification of what goes on. The chronotope of interpersonal friendship sits uncomfortably with that of the training interview, and synchronization is a process that demands quite a bit of construction work here. Note, however, that later in the interview Noura offers long and detailed autobiographical-narrative answers; the synchronization demands more work but happens eventually.

If Austin’s felicity conditions would be rigorously applied here, Noura’s initial response would perhaps be called “unhappy”, a “misfire”. Laura’s response would, from a similar perspective, be “correct” and “happy”, as it articulates the linear uptake of the interviewer’s action. From the viewpoint of making sense of the particular situation, however, Noura’s actions and those of Laura are equivalent and fit the format in spite of their substantial differences. What we can take from this is that uniformity in format does not guarantee uniformity in actions – a confirmation of Cicourel’s critique of assumptions to the contrary – and that diverse lines of action can occur within the same format, even if some actions are not linear responses to what preceded. Formats are not one-size-fits-all and linear–normative units.

Context collapse versus expansion

At this point, our action-centered proposal is complete: we see contextualization as the recognition of a situation through perceived order properties of such situations, that can be framed into formats, then guiding the actions of participants. We submit that it is applicable to interaction online and offline, since it avoids many of the core assumptions (and vocabulary) that are challenged by features of online interaction.

In studies of online interaction, “there are great analytical gains to be made by looking very closely at how particular activities are organized” (Goodwin & Goodwin 1992: 96). The advantages of that tactic can be illustrated by looking at an issue widely debated in the world of social media research: “context collapse”, i.e.

“the flattening out of multiple distinct audiences in one’s social network, such that people from different contexts become part of a singular group of message recipients”. (Vitak 2012: 541)

The theoretical and empirical validity of the concept of context collapse has been criticized by several discourse analysts (Georgakopoulou 2017a, 2017b; Tagg, Seargeant & Brown 2017; Szabla & Blommaert 2018). Indeed, online technology “complicates our metaphors of space and place, including the belief that audiences are separate from each other” (Marwick & boyd 2010: 115) and has taken us from a world of relatively transparent audiences to that of far less transparent “networked publics” (boyd 2011). But such complications cannot be solved by drawing on the sociological imagination we sketched earlier: that of “normal” dyadic face-to-face communication with well-known similar people in a tight community – which is what happens in the literature on context collapse. Such an anachronistic imagination spawns an abstract conceptualization of context as something which is only transparent when we situate humans in transparent situations in transparent communities, where “audiences” are known and trusted and people have full control over what they do in social action. When we move into the online world of online audiences and inconspicuous overhearers, of lurkers, aliases and bots, and of algorithms regulating the traffic and distribution of messages, such theoretical and analytic instruments obviously cease to be useful and have to be replaced by more flexible and precise ones.

In a case study of a long and highly complex discussion on a large Facebook group for Polish people living in The Netherlands, we used the action-centered perspective described here (Szabla & Blommaert 2018). At first glance, the case would be eminently qualified for context collapse: we had an enormous community of effective and potential participants, large enough to speak of a “networked audience” consisting of people who did not know each other. The lengthy nature of the online discussion may have disturbed our “metaphors of space and place” and the particular rudimentary platform affordances of Facebook may have complicated our expectations of coherence and sequentiality in dialogue, as responses to a prompt may not appear in adjacency but be separated by several intervening responses from others – a practical problem of synchronization, in fact. Facebook formats interactions in a curious way, and people may lose their bearings in such formats.

Our first empirical observation obviously complicated things further: the general activity of a “discussion” was, in actual fact, a mosaic of different actions, some linear and connected to the initial action (a request from a Polish-origin journalist for assistance in the making of a documentary on the labor conditions of Polish workers in The Netherlands) and many nonlinear, embedded and parallel to the initial action. People would indeed respond to the journalist’s request (and be redirected to the private messaging section of Facebook) but would also attack the orthographic errors in het Polish writing, discuss linguistic correctness in relation to Polish identity; they would accuse and scold each other on specific statements they had made, venture conspiracy theories about journalists and Polish émigrés, offer general observations about the work ethos of Polish and Dutch workers, and so forth. Each of these different lines of actions was normatively recognizable as a different chronotopic unit of participants, topics, orders of indexicality and moral codes, and was formatted accordingly.

The second observation, however, was that people found their way around this terrifically complex web of actions. The non-sequentiality of scripted Facebook interaction, the meandering of topics and participants and the generally confusing character of what went on did not appear as an obstacle for participants to participate in the specific parts of the event in which they got involved. We saw participation frameworks shift along with topic shifts, in such a way that just handfuls of people would be involved in an action, and know quite well who their actual addressees were and how they should proceed, and how they could migrate to another participation framework or exit the discussion when lines actions were closed. In other words: we saw plenty of congregational work shaping formats and subformats and connecting or disconnecting parts of the discussion from other parts. Participants made sense of the specific actions in which they were involved – they performed adequate contextualization work throughout, even if that included self- and other-correction and rectification, necessitated by the awkward Facebook discussion affordances. They recognized the specific situations, framed and formatted them into indexically ordered discursive actions. No contexts appeared to collapse; instead we saw an amazing density and intensity of contextualization work – context expansion, if you wish.

Conclusion

The example of context collapse versus context expansion brings us back to our point of departure: the need to rethink our commonly used notions of context and contextualization so as to make them useful and accurate for addressing a world of communication in which ordinary dyadic face-to-face conversation is no longer the Archimedian point and foundation for theory. Contemporary discourse analysts must be aware that the sociological imagination balancing on this Archimedian point is anachronistic, and that we cannot accurately address the phenomenology of contemporary communication without sacrificing that imagination. Doing that does not mean that we are left empty-handed to the task of analysis. We can fall back on reasonably robust tools and approaches that do not carry that bias of anachronism or can be refashioned so as to be free of it. In this chapter, we have made some proposals in that direction. Let they be a prompt for others to think along.

References

Agha, Asif (2007) Language and Social relations. Cambridge: Cambridge University Press.

Auer, Peter & Aldo DiLuzio (eds.) (1992) The Contextualization of Language. Amsterdam: John Benjamins

Austin, John L. (1962) How to Do Things with Words. Oxford: Clarendon Press.

Blommaert, Jan (2005) Discourse: A Critical Introduction. Cambridge: Cambridge University Press.

Blommaert, Jan (2015a) Chronotopes, scales and complexity in the study of language in society. Annual Review of Anthropology 44: 105-116

Blommaert, Jan (2015b) Pierre Bourdieu: Perspectives on language in society. In Jan-Ola Östman & Jef Verschueren (eds.) Handbook of Pragmatics (2015): 1-16. Amsterdam: John Benjamins.

Blommaert, Jan (2018a) Durkheim and the Internet: Sociolinguistics and the Sociological Imagination. London: Bloomsbury.

Blommaert, Jan (2018b) Chronotopes, synchronization and formats. Tilburg papers in Culture Studies paper  207. URL

Blommaert, Jan & Anna De Fina (2016) Chronotopic identities: On the spacetime organization of who we are. In Anna De Fina, Didem Ikizoglu & Jeremy Wegner (eds.) Diversity and Superdiversity: Sociocultural Linguistic Perspectives (GURT Series): 1-15 Washington: Georgetown University Press.

Blommaert, Jan & Ben Rampton (2016) Language and superdiversity. In Karel Arnaut, Jan Blommaert, Ben Rampton & Massimiliano Spotti (eds.) Language and Superdiversity: 21-48. New York: Routledge.

Blumer, Herbert (1969) Symbolic Interactionism: Program and Method. Berkeley: University of California Press.

Bourdieu, Pierre (2000) “Making the economic habitus: Algerian workers revisited.” Ethnography 1(1): 17–41.

boyd, dana (2011), ‘White Flight in Networked Publics? How Race and Class Shaped American Teen Engagement with MySpace and Facebook’, in Lisa Nakamura and Peter Chow-White (eds), Race after the Internet, 203–22, New York: Routledge.

Briggs, Charles (2005) Communicability, racial discourse and disease. Annual Review of Anthropology 34: 269-291.

Carr, E. Sumerson & Michael Lempert (2016) Introduction: the pragmatics of scale. In E. Sumerson Carr & Michael Lempert (eds.) Scale: Discourse and Dimensions of Social Life: 1-23. Oakland: University of California Press.

Castells, Manuel (1996) The Rise of the Network Society. London: Blackwell.

Cicourel, Aaron (1964) Method and Measurement in Sociology. New York: The Free Press.

Cicourel, Aaron (1967) The Social Organization of Juvenile Justice. New York: Wiley.

Cicourel, Aaron (1974) Cognitive Sociology: Language and Meaning in Social Interaction. Harmondsworth: Penguin Education.

Cicourel, Aaron(1992) The interpenetration of communicative contexts: Examples from medical encounters. In Alessandro Duranti & Charles Goodwin (eds.) Rethinking Context: 291-310. Cambridge: Cambridge University Press.

Das, Sonia (2016) Linguistic Rivalries: Tamil Migrants and Anglo-Franco Conflicts. New TYork: Oxford University Press.

Duranti, Alessandro (1997) Linguistic Anthropology. Cambridge: Cambridge University Press.

Duranti, Alessandro & Charles Goodwin (eds.) (1992) Rethinking Context. Cambridge: Cambridge University Press.

Garfinkel, Harold (2002) Ethnomethodology’s Program: Working Out Durkheim’s Aphorism. Lanham: Rowman & Littlefield.

Georgakopoulou, Alexandra (2017a) ‘Whose context collapse?’ Ethical clashes in the study of language and social media in context. Applied Linguistics Review 8/2-3: 1-32.

Georgakopoulou, Alexandra (2017b) Small stories research: A narrative paradigm for the analysis of social media. In Anabel Quan-Haase & Luke Sloan (eds.) The Sage Handbook of Social Media Research Methods: 266-281. London: Sage.

Giddens, Anthony (1984) The Constitution of Society. Berkeley: University of California Press.

Goodwin, Charles (2003) Embedded context. Research on Language and Social Interaction 36/4: 323-350.

Goodwin, Charles (2004) A competent speaker who can’t speak: The social life of aphasia. Journal of Linguistic Anthropology 14/2: 151-170.

Goodwin, Charles (2007), Participation, Stance and Affect in the Organization of Practice, Discourse and Society, 18 (1): 53–73.

Goodwin, Charles (2013) The co-operative, transformative organization of human action and knowledge. Journal of Pragmatics 46/1: 8-23.

Goodwin, Charles & Marjorie Harness Goodwin (1992) Context, activity and participation. In Peter Auer & Aldo DiLuzio (eds.) The Contextualization of Language: 77-99. Amsterdam: John Benjamins.

Goffman, Erving (1964)The neglected situation. American Anthropologist 66/6 (part 2): 133-136.

Goffman, Erving (1967 [1982]) Interactional Ritual. New York: Pantheon Books.

Goffman, Erving (1974 [1975]) Frame Analysis: An Essay on the Organization of Experience. Harmondsworth: Penguin.

Grice, H. Paul (1975) Logic and conversation. In Peter Cole & Jerry Morgan (eds.) Syntax and semantics. 3: Speech acts: 41-58. New York: Academic Press.

Gumperz, John (1982) Discourse Strategies. Cambridge: Cambridge University Press.

Gumperz, John (1992) Contextualization revisited. In Peter Auer & Aldo DiLuzio (eds.) The Contextualization of Language: 39-53. Amsterdam: John Benjamins

Gumperz, John (2003) Response essay. In Susan Eerdmans, Carlo Previgniano & Paul Thibault (eds.) Language and Interaction: Discussions with John J. Gumperz: 105-126. Amsterdam: John Benjamins

Marwick, Alice &danah boyd (2010) I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New Media and Society 13/1: 114-133.

Mills, C. Wright (1959 [2000])The Sociological Imagination. New York: Oxford University Press.

Rampton, Ben (2016) Foucault, Gumperz and Governmentality: Interaction, power and subjectivity in the 21^st century. In Nikolas Coupland (ed.) Sociolinguistics: Theoretical Debates: 303-328. Cambridge: Cambridge University Press.

Scollon, Ron (2001) Mediated Discourse: The nexus of Practice. London: Routledge

Silverstein, Michael (1992). The indeterminacy of contextualization: When is enough enough? In Peter Auer & Aldo Di Luzio (eds.) The Contextualization of Language: 55-76. Amsterdam: John Benjamins.

Silverstein, M. 2003.Indexical order and the dialectics of sociolinguistic life.Language & Communication 23: 193-229

Strauss, Anselm (1993) Continual Permutations of Action. New York: Aldine de Gruyter.

Szabla, Malgorzata & jan Blommaert (2018) Does context really collapse in social media interaction? Applied Linguistics Review 9/2, https://doi.org/10.1515/applirev-2017-0119.

Tagg, Caroline, Philip Seargeant, Philip & Amy Brown (2017). Taking Offence on Social Media: Conviviality and Communication on Facebook. London: Palgrave Pivot.

Vitak, Jessica (2012) The impact of context collapse and privacy on social network site disclosures. Journal of Broadcasting and Electronic Media 56/4: 451-470.

Voloshinov, Valentin (1973) Marxism and the Philosophy of Language. Cambridge MA: Harvard University Press.