21.1 Introduction: Social Media
Social media, together with messaging apps, are nowadays the most active and dynamic areas of internet-mediated communication. As described in Reference Yus, Barron, Gu and SteenYus (2017a: 555), the generation of content by users, the availability of areas for its publication and sharing, and the options for commenting on other users’ uploaded information crosscut many popular sites such as Facebook, Twitter, Instagram, YouTube, and so forth (see Reference boyd and Papacharissiboyd 2010). Social networking sites are often equated to social media, but the latter is a broader umbrella term (also covering blogs, for instance). Besides, these media have become highly imbricated and hybridized with the users’ offline interactions, to the extent that a certain level of online–offline congruence is nowadays expected on these sites (Reference BolanderBolander 2019; Reference YusYus 2022).
An important aspect of social media that pertains to any analysis of context on these sites is the fact that social media interactions take place through interfaces, whose design offers users different options for contextualization, ranging from cues-filtered text-based messaging to richer media such as video-mediated interactions. Furthermore, through these interfaces different types of discourse are exchanged, from plain text to multimodal discourses (e.g., memes; see Reference Eisenlauer, Hoffmann and BublitzEisenlauer 2017: 229).
This chapter addresses key issues regarding context (and contextualization) on the Internet and specially on social media. The chapter starts with general issues regarding context and the static/dynamic pragmatic approaches to explaining context. Next, an account is provided of the important online–offline dichotomy and how contextualization involves both settings. The main analysis of the chapter is provided in the next section, which includes a preliminary account of online discourses and so-called contextual constraints influencing the users’ options for contextualization, and then moves on to three basic areas of context on social media: personal, interactive, and social.
21.2 Context: From Typologies to a Set of Assumptions
There is general agreement within pragmatics that context (or contextual information) plays a major role during interpretation. However, there has been an important evolution regarding how context is pictured in pragmatics and its role in comprehension, from an initial stage that can be labeled static view to today’s dynamic view. In the first – static – stage, context was somehow taken for granted and the main interest lay in proposing types or varieties of context, assuming, as Reference Jones, LeVine and ScollonJones (2004) correctly remarks, that context is given beforehand, a sort of static “theatre-stage backdrop.” By contrast, today’s dynamic view entails a picture of context as not given, but “sought” as part of the inferential strategies devoted to obtaining an adequate interpretation of the utterance (or text, or multimodal discourse, among other possibilities). At this stage, all kinds of context are, simply, information (a set of assumptions in relevance-theoretic terminology, see Reference Sperber and WilsonSperber and Wilson 1995), mentally accessed and represented, and whose accessibility is both predicted by the addresser and carried out by the addressee in order to turn the schematic string of words into a relevant proposition. Which information from context is retrieved and processed in parallel to the interpretation of the utterance depends on the demands of the particular utterance, and accessibility to that information directs the addressee in one inferential direction or another (see Reference Blommaert, Smits, Yacoubi, De Fina and GeorgakopoulouBlommaert et al. 2020: 53).
Crucially, this pairing of contextual information and the utterance being processed takes place when inferring any stimulus, regardless of whether the utterance being processed is produced in a face-to-face environment or through an interface. In this sense, one of the claims of cyberpragmatics (Reference YusYus 2011) is that the qualities of the medium for internet-mediated communication (app, platform, interface, website, etc.) impact the options for contextualization on the Internet. Social media are a good example of how evolutions in interface design constrain or favor strategies of contextualization (see Section 21.4.2 below).
Within the dynamic view of context, relevance theory proposes the term manifestness in relation to context. In a nutshell, as summarized in Reference YusYus (2016a), people construct different concepts and representations of the world, just as their personal experiences are different. This array of information is called cognitive environment, which is made up of manifest assumptions, the ones that a person is capable of representing mentally on a particular occasion. The sum of all the assumptions that are manifest to a person makes up their cognitive environment. Crucially for contextualization, when a speaker produces an utterance, he/she expects that certain information will be manifest to the interlocutor so as to interpret the utterance correctly (this information then becoming mutually manifest to both interlocutors, a term that entails a rewriting of more traditional notions such as shared knowledge).
Manifest information may be retrieved from different sources, most importantly from encyclopedic background knowledge, but also from the physical surroundings where the interlocutors are located, among others. Needless to say, the same applies to utterances produced and interpreted on social media. In this sense, Reference Zappavigna, Hoffmann and BublitzZappavigna (2017: 450) correctly remarks that social media research has to dismiss the idea that context is stable and assume its dynamism instead: Any public or mass media communication integrates a range of meanings from different contexts to which any particular audience may not necessarily have unmediated access. It is nevertheless possible for a media text to target specific groups through modulating particular communicative choices.
21.3 Context Accessibility across the Online–Offline Divide
If context is information that is accessed in parallel to the interpretation of an utterance, then the capacity of the channel (website, interface, app, portal, etc.) to convey (i.e., provide interlocutors accessibility to) this contextual information is crucial to guarantee a successful interpretive outcome. In this sense, some research claims that the internet is unable to convey all the contextual information that is sometimes important to arrive at the intended interpretation, especially in cues-filtered text-based communication such as the one carried out through messaging apps, compared to the contextually rich environment of a face-to-face scenario. Besides this critical position on internet communication, in Reference YusYus (2011) other areas in this debate were highlighted:
(a) Some studies claim that online and offline communication are simply different, one major quality in this respect being the “mutual monitoring possibilities” that these technologies favor, “the different ways in which they allow us to be present to one another and to be aware of other peoples’ presence” (Reference Jones, LeVine and ScollonJones 2004: 23). (b) Other studies propose a picture of complementarity or extension regarding online and offline scenarios (Reference Jones, LeVine and ScollonJones 2004: 24; Reference Bolander and LocherBolander and Locher 2020). (c) Finally, in Reference YusYus (2011) the picture is that of hybridization between online and offline scenarios, the user turning nowadays into a node of intersecting interactions of a physical, virtual, and mostly mixed online–offline quality. Furthermore, it should be added that users often make the most of an apparently limited channel of virtual communication and manage to access (and make accessible) contextual information. As remarked in Reference YusYus (2022: 35) regarding the communication of nonpropositional information attached to a text (important as part of contextualization strategies), users actually manage to convey as many feelings and emotions online as in face-to-face communication, and with a similar (if not higher) intensity. The key lies in how users detect and exploit the linguistic and expressive resources (i.e., the affordances) that the interface provides for the communication of these affective attitudes and effects. A good example is to be found in messaging apps, whose users rely on text alteration and the use of stickers, GIFs, and emojis to transfer their attitudes, feelings, and emotions more accurately (and hence to aid interlocutors in a more fine-grained contextualization of their communicative intentions).
In sum, the picture that should be concluded regarding the online–offline divide in terms of contextualization is that, obviously, a scale of contextualization may be arranged depending on how rich the interface is in its capacity to convey and allow access to contextual information, but internet users make the most of these possibilities and often obtain as much contextual information as in a richer face-to-face scenario. Users even act upon the affordances of the interface so as to obtain greater richness and options to manage their interactions. As stressed in Reference YusYus (2022: 57), users adapt and reshape technologies to their own communicative needs. This is what happened with Twitter users, who were dissatisfied with its interface options and started to label tweets by using hashtags, as well as creating the retweet nomenclature. These end-user innovations were eventually incorporated into the Twitter interface by the company.
Figure 21.1 represents some of the issues that interests us regarding context and the online–offline divide and which will be developed in the next sections. In the center of Figure 21.1, “discourse” is located. In offline environments, discourse is typically conveyed in an oral, written, or multimodal form, often taking advantage of parallel vocal and visual nonverbal behavior that normally aids interlocutors in the right inferential direction. Offline discourse may also exhibit community-bound features whose proper contextualization builds up barriers of discursive specificity. Finally, individuals also exhibit their own linguistic features as unique idiolects. These may also constrain contextualization in offline scenarios. These offline features are correlated with those found online. Discourse on the Internet is often typed, but users rely on emojis, GIFs, stickers, and text alteration (repetition of letters, creative use of punctuation, etc.) to achieve their communicative needs. Richer means of interaction include recorded audio files and phone/video calls through messaging apps. Multimodal discourses (e.g., memes) also abound in online communication, and the aforementioned community- and individual-based linguistic features are also found online (Reference Yus and WrightYus 2015).
Figure 21.1 Discourse and contexts across the online–offline divide.
Online and offline discourses favor or limit the kind of contextualization that may be performed during interpretation in both settings. In this chapter, context will be studied by preliminarily distinguishing between the labels of personal, interactive, and social.
Offline personal context includes the person’s name, physical appearance, and their body as anchoring sources of identity. Online personal context retains these attributes but adds the possibility of interacting via nicks and avatars and neutralizing the user’s body, a clear anchorage of identity offline. Most importantly, in both offline and online personal contexts, there is a certain amount of information that is manifest to the person/user when interpreting discourses. This information may vary radically from person to person (or user to user) and affect inferential strategies. On other occasions, though, much of this information is of a common-sense, encyclopedic quality, and its accessibility is expected by the addresser and almost unconsciously retrieved by the addressee when turning coded inputs (oral utterances, typed messages, audio files, etc.) into meaningful interpretations. This is why contextual information is not only predicted between interlocutors but also revealed as the interaction unfolds successfully. An often-cited example is (1) below:
(1) Max: How was the party? Did it go well?
Amy: There wasn’t enough drink, and everyone left early.
(2) Parties are not successful if people leave early and they run out of drink.
(3) The party was awful and did not go well.
It is clear that Amy’s main intended interpretation of her utterance is the implicature in (3). In this case, Max’s access to common-sense background information such as (2) will easily lead to that implicature. It should also be noted that the explicit interpretation of Amy’s utterance also demands much contextualization, again retrieving general background knowledge from context. Firstly, drink is narrowed into specifically “alcoholic drinks.” Secondly, everyone is also narrowed to “everyone at the party.” Finally, the specific time frame of early depends on what is commonly assumed to be “early” for parties in the age group to which Max and Amy belong.
Continuing with Figure 21.1, interactive context entails the default form of contextualization typically addressed in pragmatics, namely the one in which two interlocutors exchange utterances and hold expectations of context accessibility (and assume this accessibility, sometimes wrongly) so that utterances are correctly inferred, and the conversation runs smoothly into an optimal outcome. In relevance-theoretic terms, the speaker/user in an interaction not only makes manifest information to the interlocutor but makes it mutually manifest to both interlocutors, the intersecting area of manifest information of the interlocutors, which becomes larger as a consequence of this mutual manifestness. An example is the irony in (4) (Reference YusYus 2016a: 210–211):
(4)
[Tom and John on a cold, wet, windy English day in London]. Tom: [Smiling, with a distinctively ironic tone of voice] When a man is tired of London, he is tired of life.
In this example, a great deal of contextual information invalidates any chance that the speaker intended to communicate the explicit interpretation of that utterance, and Tom expects his interlocutor to access specific information from context that will allow him to get the irony. Clearly, Tom is confident that John will be aware of the miserable weather, that he will also draw further contextual information including general encyclopedic information (nobody could possibly like to live in such a place), mutuality of information with the speaker (perhaps Tom has repeatedly complained about how miserable life in London is), and nonverbal communication (probably Tom has produced his utterance with an ironic tone of voice and a smile). This contextual information becomes mutually manifest as a result of the successful communication of the irony.
Contextualization via interactions takes place across the online–offline divide and is only differentiated by the interface constraints that may favor or limit the accessibility to contextual information between interlocutors. In other words, there is only one way in which the human cognitive system pairs information from utterances with contextual information to yield adequate or meaningful interpretations, but the act of contextualization itself may be constrained, to a greater or lesser extent, by the interfaces used for communication (in online scenarios) and other negative features such as ambient noise or bad pronunciation (in offline scenarios). This will be addressed in Section 21.4.2 below.
Finally, social context in Figure 21.1 refers to information that is taken for granted within a community of individuals/users in both online and offline scenarios. This contextual information is enacted and successfully managed by those who belong to the group or community, fostering feelings of group membership and creating barriers of discursive specificity, as already mentioned. Needless to say, the aforementioned context categories (personal, interactive, social) are not mutually exclusive and in fact they overlap in most of today’s interactions.
21.4 The Contexts of Social Media Discourse
Social media are a good example of the contexts briefly mentioned above. On these sites and platforms, thousands of textual, visual, and multimodal discourses are exchanged among users, all of them demanding contextualization. Similarly, users hold a personal array of information being manifest to them and making up their personal cognitive environment (personal context). This manifest information becomes mutually manifest through the whole array of interactions that can take place on social media, ranging from synchronous messaging chats to asynchronous commenting–replying on users’ posts (interactive context). Finally, as proposed in Reference YusYus (2007), information of social or communal assumptions is typically intended but may also leak from these interactions, especially if these entail particular uses of language, jargons, vocabulary, or registers, making up a store of group-specific information (social context).
A good example is the social networking site Facebook. As depicted in Figure 21.2, its desktop interface differentiates areas for these types of context. The list of Facebook groups indicates the kind of social gathering and sources of information that the user is keen on, with an offset of communal bonding and feelings of group membership. The Messenger area is devoted to (mainly synchronous) chats with other users. Finally, the user’s profile includes the main picture of the user, the profile photo, and the content made manifest by the user as entries or ephemeral stories. The profile also includes personal information about the user (studies, current job, relatives, etc.). All of this information exchanged on a unique site is one of the most outstanding features of social networking sites (see Reference Spilioti, Tracy, Ilie and SandelSpilioti 2015). Reference Bolander and LocherBolander and Locher (2020) illustrate this with the inclusion of text comments on photo-sharing sites; text (and video) responses to YouTube videos; text (and voice) chat during multiplayer online games; and text messages from mobile phones posted to interactive TV programs. They add the trend to engage with these sites and technologies and with other users via a growing range of mobile devices (see Reference YusYus 2022), adding a layer of information regarding whether the users are physically immobile or on the move in different physical, material spaces while using these devices.
Figure 21.2 Personal, interactive, and social contexts on a Facebook profile.
21.4.1 Online Discourse
As any type of discourse, online discourse is coded and has to be contextualized inferentially, with a lesser or greater gap between what is coded and what is meant (and eventually interpreted). How this task is accomplished depends on the interface affordances, as already pointed out. Twitter, with its 280-character limit, tends to provoke greater gaps between “coded” and “meant.” Texts on messaging apps are also cues-filtered, but users manage to convey their intentions more neatly thanks to “accompanying props” such as text alteration, emojis, GIFs, and stickers, among others. Finally, video-mediated interactions online exhibit, on paper, the richest medium in terms of options for contextualization, even if the medium itself is still limited compared to face-to-face communication. As argued in Reference YusYus (2022: 112–113), authentic expressions of emotion (mostly in face-to-face scenarios) are essential to our understanding. But those nonverbal hints disappear on pixelated video or are even frozen, smoothed over, or delayed. Indeed, to recognize emotion, we have to embody it, which makes mirroring essential to empathy and connection, and “when we can’t do it seamlessly, as happens during a video chat, we feel unsettled because it’s hard to read people’s reactions and, thus, predict what they will do” (Reference MurphyMurphy 2020: n.p.).
Besides, online discourses on social media are often not delimited, with a clear structure and expected reading patterns. Instead, they are often link-mediated and the user has to apply cohesion and coherence to these discourses, often taking the responsibility for which path is taken and what eventual interpretation is obtained, all that through dynamic contextualizing tasks.
Crucially, as already mentioned in passing, interface affordances (with parallel options for contextualization) are imposed on the social media users, but users can act on the interface in order to achieve their communicative needs, also when aiding interlocutors in an adequate contextualization (Reference GruberGruber 2019: 71). Reference Eisenlauer, Hoffmann and BublitzEisenlauer (2017: 56) points in the same direction when remarking that the function of a particular social medium “rests on the interplay of the platform’s communicative exigencies and the ways in which individual users adapt to these affordances in different ways. As shown, the functions that users associate with a particular service may or may not correspond with the initial functions of the developers.” Regarding contextualization, this interplay of site affordances and user innovation impacts the eventual options for contextualization that users are granted and, hence, also impacts the final interpretation obtained from discourses on social media.
A similar interplay occurs in the case of messaging areas within social media interfaces. Initially, these were mainly text-based, but both users and companies increased the options for contextualization through these interfaces. Companies gradually incorporated emojis, GIFs, stickers, audio files, text type (bold, italics, and so on) and images into their interfaces, again increasing the options for contextualization (see Reference YusYus 2011; Reference Yus, Barron, Gu and Steen2017a: 556; Reference Yus2022; Reference Pano Alamán and Mancera RuedaPano Alamán and Mancera Rueda 2014: 245).
21.4.2 Contextual Constraints/Affordances
As already highlighted in passing, online communication in general and strategies of contextualization in particular are constrained by the qualities of the interface through which online exchanges take place, that is, its affordances, defined in Reference EisenlauerEisenlauer (2014: 73) as the possibilities “afforded by an object or an environment in relation to social actors and their individual capabilities.” Contextual constraints may be defined as aspects which underlie acts of communication and users’ interactions (i.e., they exist prior to the act of communication) and determine their eventual (un)successful outcome. They frame, as it were, communication and have an impact both on the quality of interpretation and on the willingness to engage in future interactions.
Specifically, regarding social media interfaces, as described in Reference YusYus (2022), the profile provides users with different affordances that restrict or allow for the kinds of discourse that may be used therein. By making the interface more user-friendly, designers expect to provide an effortless experience while using their social media interfaces, what Reference CostaCosta (2018: 3651) labels affordances-in-practice, a term that highlights the idea that “social media affordances are the results of the repeated interactions between humans and platforms. They are the results of social and material enablement. As such, they are not fixed and stable properties but are implicated in different ongoing processes of constitution, which may radically vary across social and cultural contexts” (see also Reference Eisenlauer, Hoffmann and BublitzEisenlauer 2017: 232–234). To achieve user satisfaction, the information presented on the interface has to be clearly arranged, with suitable links that allow users to find, create, and process social information easily. Other researchers, not within pragmatics, have also addressed this layering of relevance in multimodal discourses (see Reference Kress and van LeeuwenKress and van Leeuwen 1996; Reference Hassan, van Leeuwen, Habil and HassanHassan and van Leeuwen 2008; Reference van Leeuwen, Stöckl, Caple and Pflaegingvan Leeuwen 2020). A term proposed in this direction is salience, decided by viewers’ instinctive judgment of the importance of various elements in a composition. The greater the importance of an element, the greater is its salience. This salience cannot be measured objectively and stems from complex interaction among many factors including size, sharpness of focus, color contrast, perspective, placement of visual elements, and specific cultural factors. The degree of salience results in a hierarchy of attention-degree among the visual elements of images. Another useful term is framing, this time related to whether there are devices that separate or connect visual elements in an image and impact their eventual interpretation (see Reference Chen and HeChen and He (2015).
In any case, what matters most from the (cyber)pragmatics point of view is whether users’ interactions are favored or limited depending on site design. Constraints become crucial especially regarding smartphones, where users show less willingness to expend effort in managing their interactions on small touchscreens. Desktop profiles and their smartphone counterparts may be similar, but some features inevitably need to be discarded or rearranged from the main desktop interface to the smartphone app in order to simplify the items displayed on the screen and make app management as effortless as possible. The most radical example of this variation between desktop and app interface is Facebook (Reference YusYus 2022), whose desktop site provides a visual display of several relevant items which become simultaneously available to the user (see Figure 21.2). By contrast, using the Facebook app entails two separate apps for the kind of activity that users perform on a single desktop screen. This may seem like a trivial alteration to some readers, but having to switch between apps to manage communication and interaction on Facebook discourages many users, who bitterly complain about the effort involved.
Interface constraints are further constrained by the site’s algorithmic procedures, which select who appears on the user’s feed or which content is presented to them, among other filtering activities (supposedly predicting the eventual relevance of these automatic choices for the end user; see my proposal of algorithmic manifestness in Section 21.4.3 below). Unlike many studies of social media that conceptualize its interface affordances as technological issues separated from its users (Reference CostaCosta 2018: 3650), now these are viewed as intertwined with users with implications for interaction, contextualization, and sociability on these sites.
In any case, and despite their limitations compared to contextually rich face-to-face interaction, social media allow for meaningful and rewarding interactions whose contextualization need not necessarily be thwarted by interface design. As summarized in Reference QuinnQuinn (2016: 593), social media offer many options for rewarding relationship maintenance, including their broadcast capabilities (dissemination of news and relevant information), and their messaging and chat functions (for one-to-one or group communication). Social media also preserve information that can subsequently be retrieved and reenacted in the future. Persistence of publications and interactions also translates into a stable and dedicated social address, “which gives others the ability to locate and contact their connections over time and across geographic, temporal, and social distance, further extending the ability to tap resources” (p. 593). Below, two examples are provided of interface constraints. The first one of a negative quality (messaging areas embedded in social networking sites such as Facebook or Instagram), and the other of a positive quality (the use of hashtags on Twitter to allow for better contextualization). In this sense, it should be noted that although it might sound as if constraints necessarily have to be negative (in the sense of “limiting” choices and options), in reality, within cyberpragmatics these are also regarded as possibly exhibiting positive qualities. For example, in the just-mentioned example, hashtags are useful in constraining the wide variety of interpretations that the accompanying text might generate, thus reducing mental effort and increasing eventual relevance.
Some qualities of messaging areas embedded in social networking sites are a burden for optimal contextualization of the utterances typed therein. One of them is disrupted turn allocation. This refers to the situation in which the system mixes up the conversational turns that are produced in the messaging conversation (see Reference YusYus 2017c). For example, in Reference Yus and Romero-TrilloYus (2016c) the following WhatsApp conversation was reproduced. It took place between a female (A) and a male (B) user:
(5)
A: La voy a facturar en el aeropuerto [I’m going to check her in at the airport]. B: Eso [That’s it]. A: Egipto, que allí los idolatran [To Egypt, since they idolise them there]. A: O a Marruecos pa q aprenda lo que vale un peine [Or to Morocco, so that she learns the tough way]. Q está muy mimadita [because she’s too spoiled]. B: Yeah A: Yastan aqui mis padres [My parents are here already]. B: Que vea que la vida no es solo hacer trastadas [She has to realise that life is not all about playing tricks around]. A: [emoji of anguish]. B: Ohhhh. Planazo Ohhhh. Great plan]. A: Total [Totally].
In this conversation, the initial topic is how angry A is with her naughty cat. Halfway through this dialogue, A informs B that her parents have just arrived (Yastan aquí mis padres), but B’s next message is still related to the naughty cat, since the system has reproduced messages in strict order of arrival. Similarly, A’s next message, an emoji of anguish which codes a whole proposition (roughly “my parents’ visit depresses me”), does not refer to the cat either, although it follows B’s cat-related message. These mixed-up threads in messaging conversations may be a potential source of misunderstanding or increased processing effort.
On the other hand, an example of positive interface constraint on social media is the Twitter hashtag and its role in aiding contextualization. Its communicative role was a discursive innovation by Twitter users and not initially integrated into the interface, as already mentioned. When attached to the tweet, the hashtag may perform a number of interesting functions aiding the audience in contextualization and in the right inferential direction. For example, Reference PuschmannPuschmann’s (2015: 30) and Reference DayterDayter’s (2016) study uses of hashtags for varied identity-related pragmatic strategies. Another example can be found in Reference ScottScott (2015), for whom the hashtag performs several roles when it comes to obtaining the intended interpretation of the tweet next to which they are typed. For instance, the hashtag in (6a) allows the reader to locate the referent for “it” and obtain the derivation of an optimal interpretation like (6b). And the hashtag in (7a) allows for the inference of the user’s propositional attitude toward the tweet, interpreted as (7b) (see also Reference WikströmWikström 2014; Reference Zappavigna, Hoffmann and BublitzZappavigna 2017: 445–446):
a. She’s done it! An amazing amazing effort. Please txt FIVE to 70510 #davina #windermere.
b. British television presenter Davina McCall has finished her charity swim across Lake Windermere.
a. One week from today I can start throwing again. #finally.
b. The user is relieved that she can start throwing one week from today.
21.4.3 Personal Context on Social Media
The user’s personal context is made up of all the information (the set of assumptions in relevance-theoretic terminology) that is manifest to the user, making up their cognitive environment. This information is activated and updated through repeated interactions and associated contextualization. In this sense, an important part of this manifest information refers to the user’s identity, both offline and online, and how identity-centered strategies (e.g., self-presentation) allow for a more specific self-concept and self-awareness. In fact, in Reference YusYus (2011, Reference Yus2014, Reference Yus and Carrió-Pastor2016b, Reference Yus2022) it is claimed that most of the users’ communicative activity on social media is related to identity claims and identity shaping, with an expectation of online–offline congruence, as already stated (see Reference Davidson and JoinsonDavidson and Joinson 2021). Therefore, the user’s personal context includes not only the information that is manifest to them but also the information that they decide to make manifest to other users, often with an identity-related communicative strategy in mind.
Manifest information is relevant to the user in three main cases, according to relevance theory: (a) when it leads to previously held assumptions being corroborated; (b) when these assumptions are contradicted and eventually erased; and (c) when it combines with existing assumptions to yield relevant conclusions. In the case of personal information strategically made manifest to other users on social media, audience validation is essential to reinforce the validity of the user’s personal assumptions regarding their identity.
Besides this information strategically made manifest to others, the user also has access to a lot of information from their peers, friends, and acquaintances, some of which combines (i.e., is contextualized), in a relevant way, with the user’s previously held assumptions to yield a more accurate and updated personal cognitive environment. In this sense, it should be noted that much of the information that is manifest to the user is not as a consequence of the user’s decision to find it but is presented to the user through the system’s algorithms that decide which information is likely to be relevant to the user by judging from the user’s previous activity online. This tracking activity may be labeled algorithmic manifestness. As described in Reference Yus, Piskorska and WałaszewskaYus (2017b) regarding Facebook, the site automatically filters and makes manifest information to the user in three different ways: (1) By affinity: the more one contacts somebody and the more someone visits the profile and engages in conversations with him/her, the more likely that Facebook will show updates from that user. (2) By the relative value of content: for example, updates on user status, e.g., on the user being no longer married etc., are more valuable and the system emphasizes them, but other types of content are also underlined by the system if it detects the user’s tendency to see that kind of content. (3) By time: obviously, the most recently published entries have prominence over the oldest. In short, content is also monitored and managed by the site’s software and algorithms (see Reference LomborgLomborg 2014: 2).
21.4.4 Interactive Context on Social Media
If information manifest to the user (or made manifest by the user to other users) is at the heart of personal context, mutual manifestness of information is at the heart of the interactive context on social media, since they “facilitate a distinctive form of social interaction online, creating a constantly expanding network of social relationships characterized by varying degrees of familiarity and tenuousness and by the exchange of symbolic content in multiple formats and modalities that is made available to others with varying degrees of openness and restrictiveness” (Reference ThompsonThompson 2020: 6–7).
During the various types of interactions that users may carry out on these sites (synchronous text-based, asynchronous multimodal, etc.), users expect their audience to have access to the information that will allow them to understand their utterances correctly (i.e., that this information will be manifest to that audience) and, as a consequence of successful contextualization, that information becomes part of their mutual cognitive environment, made up of mutually manifest assumptions (as vividly revealed by successful comprehension). An example is the audience’s background store of information regarding the user, which is crucial for communicative strategies such as irony. Indeed, ironists usually rely on information supposedly shared (mutually manifest) between the interlocutors or a specific group of friends or acquaintances. Such a mutuality is not only assumed or predicted but also revealed by the irony itself. This desire to foreground mutuality via irony is also present in online communication. In Yus (forthcoming), the Facebook post in (8a) below is proposed, which foregrounds information on marriage and sex roles. It assumes a mutuality (mutual manifestness) of information with other female Facebook users regarding marriage and their husbands’ lack of cooperation in taking care of their children. The male user’s reply to the post in (8b) holds similar expectations of mutuality with certain male users and ironically counters that prior criticism. He echoes wives’ habit of pulling the blanket to their side of the bed, leaving husbands uncovered and cold. Both posts rely on the mutual manifestness of this specific mutuality of information, likewise enhanced and revealed through such ironic remarks.
a. Facebook post: Voy a acostarme esta noche en el lado de la cama de mi marido. Por lo visto en ese lado no oyes a los niños cuando se despiertan por la noche.
[I am going to sleep on my husband’s side of the bed tonight. Apparently, on that side you can’t hear children when they wake up in the middle of the night].
b. Reply to post: Yo en el de mi mujer. Al parecer allí sí hay manta.
[I will sleep on my wife’s. Apparently, on that side there is a blanket].
Again, social media interactions are influenced by interface-related contextual constraints. On sites such as Facebook, interactions normally comprise chains of successive messages related to an initial post which are sequentially uploaded by the system. This attribute does not necessarily mean that these interaction messages are disorganized or lack cohesion and coherence. In fact, users tend to build on one another’s contributions and these turns are thematically organized by the system, that is, the system arranges the messages in such a way that the threads can be visually identified and do not get mixed up. For example, those which may be labeled as comment-centered interactions on Facebook are visually distinguished from entry-centered interactions, the former appearing with a smaller letter size and being placed by the system immediately below the comment that triggered the post (see Yus, forthcoming).
Furthermore, users typically post information on social media (as entries, ephemeral video stories, etc.) with three types of expectation in mind: (a) an expectation that this information will be relevant enough to arouse the audience’s attention with an expectation of relevance; (b) an expectation of mutuality of information allowing the audience to contextualize this information correctly; and (c) an expectation that this information will provoke the audience’s reaction to it and prospective interactions around it in terms of “likes,” comments, etc., what in Reference YusYus (2014) was called interactivity trigger. The problem is that the user’s audience is made up of a very heterogeneous mixture of intimate friends, acquaintances, colleagues, former students, etc. who have different accessibility to background contextual information regarding the user and the information posted on the site, and therefore will react differently to this information, ranging from cases of “joy of mutual manifestness” revealed from correct interpretation and subsequent dialogues based on the information posted, to cases of inability to contextualize the information properly, leading to misunderstanding and puzzlement.
The heterogeneous quality of users making up the audience on social media was called context collapse by Reference Marwick and boydMarwick and Boyd (2011). Therefore, as highlighted in Yus (forthcoming), despite the possibility of interacting with hundreds of users on the site in question, users tend to interact with only a handful of users on the friends list, others being left in a permanent state of “to be contacted if ever required or desired.” Within this environment of context collapse, with varied users and heterogeneous audiences, many users do not really know how many people make up their audience on the site, and many can only identify a smaller percentage of their network correctly, so this “interactive narrowing” comes as no surprise (Reference Szabla and BlommaertSzabla and Blommaert 2020: 252). In this environment, audiences are, rather, imagined audiences of which the user has no clear picture (see Reference Beam, Child, Hutchens and HmielowskiBeam et al. 2018: 2299) or invisible audiences that escape user control (Reference VitakVitak 2012: 453). Besides, as was mentioned above, both friends and their entries on the site are managed by the site’s algorithm that decides which users are visible and which of their content can be accessed by other users.
Context collapse is an interesting notion that affects interactions not only on social media but also on other types of online interaction (Reference AndroutsopoulosAndroutsopoulos 2014: 63), and has an impact on contextualization and (un)successful inferential outcomes too. As Reference Marwick and boydMarwick and boyd (2011: 122) comment regarding Twitter, there is a requirement to present a verifiable, singular identity to the user’s audience, but this makes it impossible to differentiate self-presentation strategies depending on group variables such as degree of closeness or job relatedness, thus creating tension as diverse groups of people flock to social network sites under the same unifying label of “audience.” In other words, users must hold expectations of relevance and expect that their entries will interact fruitfully with the cognitive environments of groups of users that would normally not be brought together as a unique virtual group.
In order to navigate these tensions arising from reactions of (ir)relevance toward the same entry addressed to this myriad of users, a variety of tactics may be adopted, such as using multiple accounts, pseudonyms, and nicknames, among others, to obscure their real identities. Of course, these tactics depend on the purposes and agreed uses of the site itself. In certain networking sites such as Facebook or LinkedIn, there is an expectation of what in Reference YusYus (2022) is called offline–online congruence, that is, the expectation that the user remains more or less the same across these environments (see Reference van Dijckvan Dijck 2013).
In this sense, one of the most frequent techniques is so-called lowest-common denominator effect, when individuals only post information that they believe their broadest group of acquaintances will find relevant and non-offensive (see Reference DuguayDuguay 2016: 894). In relevance-theoretic terms, users tend to make manifest information that will probably be found relevant (will interact positively with the heterogeneous audience of the user) via contextualization with the audience’s background store of contextual information against which this new piece of information is inferred. Other strategies include users negotiating “multiple, overlapping audiences by strategically concealing information, targeting tweets to different audiences and attempting to portray both an authentic self and an interesting personality” (p. 894; see also Reference Gil-Lopez, Shen, Benefield, Palomares, Kosinski and StillwellGil-Lopez et al. 2018: 128).
It should be noted that context collapse may occasionally be strategically used. This is the foundation of Reference Davis and JurgensonDavis and Jurgenson’s (2014) differentiation between context collisions and context collusions. The former occur when people’s manifest information and overall online behavior accidentally jumps to non-predicted audiences, with different social environments unintentionally and unexpectedly crashing into each other (p. 480). This process could produce negative communicative outcomes, such as disagreement (for instance, manifest information contradicting the background beliefs of some users in that audience) or that audience’s inability to contextualize the information properly. On the other hand, collusions occur when people strategically use various contexts of their social network in exchange for social utility, that is, “the process whereby social actors intentionally collapse, blur, and flatten contexts, especially using various social media” (p. 480). Potential benefits include increased bonding through engagement.
21.4.5 Social Context on Social Media
Social context is made up of assumptions that are manifest to a whole collectivity within a virtual social media group, community, or gathering. The users in these groups are aware of the specificity that these assumptions entail and also of the discursive barriers generated out of this communal access to specific sets of assumptions. Very often, the users’ (in)ability to access these background assumptions during contextualization signals whether these users belong to the social group or are outside, as already mentioned above in passing. Normally, background social assumptions are assumed and hence not coded, which makes inference more difficult or impossible by outsiders. Clearly, contextualization acquires importance not only at a purely discursive level but also socially. It helps interlocutors frame utterances within their social and collective environment, and aids users in shaping their identities, their position in groups, and the state of their networks of friends and acquaintances.
Group specificity is identified when contextualizing and inferring discourses exchanged within the bounded virtual group, but it may also be signaled when producing discourses therein. For example, certain genres are essential to understand how to participate in the collective actions of a community within the trend of participatory culture. Whether textual or social, genres are important framing devices, especially as generic conventions set up expectations (Reference Erickson and SpragueErickson 1997; Reference Bou-Franch and Garces-Conejos BlitvichBou-Franch and Garces-Conejos Blitvich 2018; Reference YusYus 2018: 119). An example is the LOLCat genre. Reference MiltnerMiltner (2014) uses this family of memes to illustrate how the form and structure of the LOLCat are distinct, the proper execution of the generic conventions being essential to its appeal. Participants repeatedly refer to font, text placement, image subject, and animal characterization as integral to the enjoyment of a LOLCat. In other words, adherence to specific design elements is considered paramount to LOLCats enjoyment and also reinforces collective bonding and awareness of group membership.
21.5 Concluding Remarks and Future Projections
Context is information that is accessed in order to interpret utterances correctly, that is, to turn schematic coded discourses into meaningful interpretations. Since context is information stored, expected, and retrieved, it is also the array of assumptions making up the individual’s personal context, the assumptions made mutually manifest between interlocutors during interactions and the assumptions crosscutting the members of certain groups or communities. In the case of internet-mediated communication and specifically social media, this varied contextual information is also at work, but interfaces play a major role in how this information is presented, activated, processed, and combined with online discourses in the accomplishment of ordinary virtual interactions. In this case, as hinted in Yus (forthcoming), when applying pragmatic parameters such as context and contextualization, we are faced with two apparently contradictory statements. On the one hand, internet makes no difference, in the sense that online users also interpret other users’ utterances with the aid of contextual information, just like interlocutors in offline scenarios. Therefore, applications of pragmatics to this virtual environment should be straightforward. However, on the other hand internet makes all the difference, since virtual communication takes place through interfaces, often in a cues-filtered environment, typically text based (even nowadays), and with fewer options and resources for contextualization (e.g., lack of nonverbal communication, of physical co-presence, etc.).
In future research regarding context online, some trends are bound to remain (or become) important, some of which were sketched in Reference YusYus (2019):
1. What is coded versus what is meant. Social media discourses exhibit a great variability of possibilities for contextualization, but the inferential “filling up” of coded content to be turned into propositional interpretations will remain an essential object of future research. Users will continue to use text-enriching strategies to get a more reliable interpretation of their discourses as well as to convey the feelings and emotions associated with them.
2. The pervasiveness of multimodal discourses. (Cyber)pragmatics will be particularly interested in those instances of multimodal discourses whose contextualization entails the combination of the partial meanings of the modes making up the discourse to yield relevant implicated conclusions (see Reference Kress and van LeeuwenKress and van Leeuwen 1996; Reference Jewitt, LeVine and ScollonJewitt 2004; Reference van Leeuwen, Norris and Maiervan Leeuwen 2014, Reference van Leeuwen, Tannen, Hamilton and Schiffrin2015).
3. Typed discourses with additional support. Many users still prefer to resort to cues-filtered discourses such as typed texts, and will continue to “color” these texts with emojis, stickers, text alteration, and GIFs. Contextualization may be aided by these enriching strategies applied to plain text, an interesting research area.
4. Varying sources of mutuality in the physical–virtual interface. As claimed in Yus (forthcoming), smartphones are used continuously by people in physical scenarios, and even in offline social gatherings where oral communication should be the norm. In this sense, there are prospects of research for (cyber)pragmatics regarding sources of mutuality of information and how online and offline sources of information overlap, complement, or contradict each other (Reference Yus, Xie, Yus and HaberlandYus 2021). An example is the mixed face-to-face and virtual interactions, for example the situation of interlocutors who chat face to face and at the same time text other physically absent interlocutors with their smartphones.
5. The importance of contextual constraints. These have been highlighted as impacting the quality of contextualization strategies in this chapter. These “framing qualities” will remain an important issue in future (cyber)pragmatic research.
6. Big data and their impact on users’ interactions. As also remarked in this chapter, algorithms decide which content is offered to the user and which user may interact with whom. As such, it is an interesting area for future (cyber)pragmatic research on context and contextualization. This automatic assessment of sources of interest by algorithms does play a part in what content and users are presented to the user, and therefore also on contextualization strategies.
7. Pragmatics of virtual agents. Nowadays, computer assistants, bots, and agents (e.g., Cortana, Siri, Alexa) are increasingly entering our daily lives. These have an amazing capacity to access information on the Internet and come up with supposedly relevant answers to the user’s queries. In the near future, researchers in computer science will be interested in how to “train” these devices to become more “pragmatic” and move closer to a human way of processing information, and also closer to the way humans combine in-coming information with context to yield relevant interpretations. This is especially noticeable in the case of online bots, which can interact through oral or typed dialogues with users (especially customers in shopping websites) and access a lot of background information available online in order to tailor the conversation to the user’s needs.
8. Online polylogues. Nowadays, interactions are increasingly devoid of the traditional dyadic quality, that is, utterances are no longer exchanged between one single sender and one single receiver with a single intended interpretation. Interactions are now more frequently multiparty and form different layers or polylogues, either text-based or with the aid of visual and multimodal content. This constitutes an interesting area of (cyber)pragmatic research regarding how these partial contributions are contextualized and contrasted with the overall interpretation of the thread or interaction as a whole (see Reference Bou-Franch, Lorenzo-Dus and Garces-Conejos BlitvichBou-Franch et al. 2012; Reference 474Bou-Franch and Garces-Conejos BlitvichBou-Franch and Garcés-Conejos Blitvich 2014).
9. Media convergence and multiple simultaneous areas of interaction. New interfaces for interaction are becoming popular, for example on smartphones in the shape of either dedicated apps or interfaces that “migrate” from the desktop environment to the new screen-size realm of the smartphone ecosystem, with changes in the presentation and management of the content made available through these apps having an impact on how information is contextualized.
