Embodiment and simulated interaction in online stance expression

The expression and exchange of stance drives much social media discourse, including internet memes. We demonstrate how, even in the absence of actual face-to-face communication, online discourse and memes rely on the dynamics of embodiment and dialogue in comparable ways, while also developing specific constructional forms for this with no direct face-to-face equivalent. We introduce the notion of simulated interaction to refer to the combinations of embodied expression, images, and the structures of (apparent) quotation and dialogue allowing online communicators to vividly represent experience and signal stance.

1 Introduction and approach taken

Online discourse as it unfolds on various social media platforms is rife with expressions and exchanges of stance. It is no exaggeration to say stance exchange is the lifeblood of a platform such as X/Twitter: expressing viewpoints which can then be ‘liked’, shared, commented on, or quoted with or without further comment is what continually fills up the so-called ‘timeline’ people read, refresh and respond to (e.g., Wikström, 2019; Vandelanotte, 2020). Similarly, internet memes are driven by internet users’ need to make light of, critique or ironize all sorts of daily frustrations, large or small, and their popularity increasingly molds discourses online as well as offline (Dancygier and Vandelanotte, 2017b; Dancygier and Vandelanotte, 2025).

Internet memes in particular have been the focus of much popular interest as well as scholarly research into how they can be defined, how they emerge and how they contribute to public discourse (e.g., Jenkins, 2014; Shifman, 2014; Wiggins and Bowers, 2015; Milner, 2016; Wiggins, 2019). The foundation laid by this line of research has clearly demonstrated the degree to which the notion of memes itself has evolved since the term’s initial coining by Dawkins (1976), where memes were defined as units of cultural replication, on the model of ‘genes’ as ‘selfish’ and virulent units of genetic replication; examples include, e.g., catchphrases, tunes, fashions, architectural styles and the like. Internet memes are quite different in clearly involving the active agency and creativity of meme communicators, who are not more or less passive ‘hosts’ but who deliberately devise textual and visual modifications and remixes of established meme patterns in order to target further areas of experience they want to express stance toward.

Internet memes and other forms of online discourse have generated considerable study of political meanings and social ramifications (e.g., Milner, 2016; Ross and Rivers, 2017; Denisova, 2019; Paz et al., 2021; Zappavigna and Logi, 2024). Also, studies of pragmatic aspects of such discourse types have touched on questions such as common ground, intertextuality and humorous incongruity (e.g., Yus, 2018; Wikström, 2019; Xie, 2022; Attardo, 2023). The strand of research we have broadly approached these discourse types from, and on which we build here, is one inspired by the tenets of construction grammar and cognitive linguistics, as represented in, e.g., Dancygier and Vandelanotte, (2017b); Dancygier (2017); Lou (2017); Zenner and Geeraerts (2018); Bülow et al. (2018); Piata (2020); Vandelanotte (2021); Kang et al. (2023). This line of work treats internet memes as multimodal constructions, i.e., pairings of form and meaning. Specific memetic forms are governed by rules which can be described in a ‘grammar’ of memes, while the meaning of a meme emerges from integrating the meaning of the memetic template with the situations and frames evoked, with important roles in this meaning emergence played by figuration and viewpoint. Dancygier and Vandelanotte (2025) propose an overview of the patterns of meaning typically involved, across image and text.

The role of stance in internet memes and social media discourse more generally has been highlighted (e.g., Shifman, 2014; Dancygier and Vandelanotte, 2017b; Droz-dit-Busset, 2022), but we propose to analyze specific patterns of interactions used in internet memes effectively for the purposes of stance exchange. Our understanding of stance is informed by the work of Du Bois (2007), who defines stance as “a public act by a social actor, achieved dialogically through overt communicative means (language, gesture, and other symbolic forms), of simultaneously evaluating objects, positioning subjects (self and others), and aligning with other subjects, with respect to any salient dimension of the sociocultural field.” Among the aspects of this definition which we want to highlight are its multimodal orientation, not restricting itself to the verbal stream in communication (cf. Andries et al., 2023), and its posit of stance as a ‘tri-act’: “I evaluate something, and thereby position myself, and thereby align with you.” Du Bois also rightly stresses the frequent resonance of “the current stance act … with a stance taken in prior discourse”—a scenario of which, we would argue, much online discourse exchanges present a heightened case—and introduces the terms of “stance lead” and “stance follow” to refer to an initial stance and a response to it. Dancygier (2012), in her analysis of negation of stance verbs, has introduced the notion of stance-stacking to characterize constructions which accommodate multiple stances being stacked one upon the other. We believe many examples of internet memes incorporate such complex stance configurations. More specifically, they often involve forms, such as facial expressions, body postures and scenes of interaction and dialogue between agents, that lend themselves to emotional interpretation, and are often used to vent about life’s daily smaller or larger frustrations, or about people’s amusing or annoying foibles. Our focus is thus on the expression of emotional or affective stances (Ochs, 2006; Goodwin et al., 2012; Couper-Kuhlen and Selting, 2017), where we concur with Goodwin et al. who argue that emotion is a “contextualized, multiparty, multimodal process” (2012: 18) which “emerges in unfolding interaction” (2012: 24).¹

2 Materials and research questions

The present article is an offshoot of the larger project reported on in Dancygier and Vandelanotte (2025), which focuses on proposing a ‘grammar of memes’, demonstrating the emergence of constructions in which images become structural components, while making language forms adjust to the emergence of multimodal rules. We base this project, and also the current article, on a large and growing collection of internet memes, relying on a combination of meme collecting sites (such as Knowyourmeme.com), internet searches, and our own observations on social media platforms such as X/Twitter, Facebook and Reddit.² Our aim in data collecting is to discover the range and classes of internet meme types that can usefully be distinguished along major parameters, such as the involvement of ad hoc vs. more entrenched images, various kinds of grid-like arrangements of the space of the meme, use of fictively quoted clauses and lines of dialogue, and the different uses and arrangements of text (e.g., using lines of top text and bottom text within the space taken up by the image, vs. using text in a blank space above the image, vs. using single-word or phrasal labels on top of individual components within the image). The aim is thus not to capture the minutiae of pop culture references in memes (which existing online resources do very well), nor each and every new minor variation in meaning expression, but rather in identifying conventionalizing or conventional form-meaning patterns in the data.

In this article, we re-examine some of our data, and include new cases, to elucidate two main research questions. Both of these questions find their origin in the specific type of ‘multimodality’ involved in the kinds of image-text combinations we will consider. Where the special issue’s main focus is on embodied interaction in face-to-face settings, or in mediated, virtual forms of such settings, we look at online discourse that is not face-to-face. Relatedly, our interest here lies with combinations of image and text—different semiotic resources accessed visually—rather than with combinations of language and embodied behavior such as gesture, gaze and posture. While we recognize these frameworks are different and are thus sometimes referred to using different terminology (such as polysemioticity vs. multimodality), we have always stressed the potential for fruitful crossfertilization between the two approaches (Dancygier and Vandelanotte, 2017a, Vandelanotte and Dancygier, 2017). In this article specifically, we therefore turn to two questions which a reconsideration of our recent data sets (Dancygier and Vandelanotte, 2025) have thrown up: (i) what role do visual depictions of forms of embodied behavior (including facial expressions, gestures, postures) play in internet memes, and how; and (ii) what role do ‘pretend-conversations’, sometimes extending over multiple lines of exchange, and often including images as ‘turns’ in an exchange, play in memes and other social media discourse? While the artifacts we study are very different from face-to-face settings (‘real’ or virtual) in which discourse develops over time, our concluding section reflects on the relevance of including such forms of multimodality in investigations of the embodied interactions they so succinctly reflect back at online communicators. In addition, we contribute to the theorizing, in cognitive linguistics, of Pascual’s key notion of fictive interaction (Pascual, 2002, 2014; Pascual and Sandler, 2016), by proposing the term ‘simulated interaction’ to capture the more complex forms our examples bring to light. ‘Simulated interaction’ is, in the approach we propose, a communicative mode which builds on familiar conversational patterns and mechanisms of ‘multimodality in interaction’ (Feyaerts et al., 2017a) to give a salient form of expression to evaluative and emotional stances. We rely on Pascual’s idea of the ‘conversation frame’, while showing how ‘simulated interaction’ develops its own formal parameters and constructional forms, and how these emergent mechanisms serve the purposes of stance expression in online discourse. We rely on internet memes as our primary online discourse genre, because their form and meaning provide a succinct but compelling demonstration of the forms and functions of simulated dialogic patterns.

3 Analysis and discussion

3.1 Depicted bodies in internet memes constructing emotional stance

Internet memes are typically combinations of images and text. A more elaborate analysis of the role these two semiotic modes play in memes is beyond the scope of this paper, and so we will focus on some limited observations regarding the role memetic images play in constructing stance.

Memetic images are essentially of two kinds—entrenched or non-entrenched, but most (if not all) of them use representations of body postures and facial expressions. The non-entrenched images are often random selections made by the Meme Maker, guided by the easily recognized expressivity of the posture or face represented, while the entrenched ones play a unique role of simulating a personality or event type.

Classic examples of so-called image macro memes, where images are entrenched, combine a recurring image, emblematic of the meme pattern, with text neatly divided across top and bottom areas of the space taken up by this recurrent image. Two very basic examples which we discussed in previous work (Dancygier and Vandelanotte, 2017b) can be relied on here to illustrate the case; both reflect sexist stereotypes which exist around women and arguments. The example of the One Does Not Simply meme, shown in Figure 1, suggests it is futile, to the point of impossibility, to try to win an argument with a woman. In one example of the so-called Good Girl Gina meme (Figure 2), the image of the meme’s stock character (a smiling woman looking very happily and confidently into the camera) is accompanied by the lines of text “gets mad of you” (as top text, i.e., shown near the upper edge of the image) and “tells you why” (as bottom text, shown near the lower edge). The example suggests that it’s only a truly exceptionally “good girl” like “Gina” who will tell you why she got mad at you (implying that most women, when they are angry, stubbornly expect you to just know or guess why).

Figure 1

Figure 1. One Does Not Simply meme. Reproduced from Imgflip.com.

Figure 2

Figure 2. Good Girl Gina meme. Reproduced from Imgflip.com.

Both these examples depict faces of Meme Characters—Boromir, a character from the film The Lord of the Rings in the One Does Not Simply example, and the stock character of Good Girl Gina, taken to represent women who display highly considerate and virtuous behavior. The memes are not ‘about’ these characters as actual ‘referents’; rather, they are used to communicate generally recognizable scenarios and the Meme Maker’s stances about these, within an intersubjective context where it is expected that the meme communicator’s primary online audience can relate to and perhaps share the attitudes expressed. As we have argued in other work (Dancygier and Vandelanotte, 2017b; Dancygier and Vandelanotte, 2025), image macro memes perform two related tasks—the meme template establishes a category of stance (e.g., the belief that some goals cannot be achieved, in the One Does Not Simply case), while every next meme building on the template represents a new instance of the specific stance-evaluated behavior (e.g., every next example of what Good Girl Gina would do in a situation is evaluated as welcome and generally positive).

While the embodiment on display is necessarily limited and stilted—with only a still image available—even in such a simple case, the facial expressions, postures and gestures are very important in construing the overall meanings. In the One Does Not Simply example, the combination of the still from the film and the signature line “one does not simply”—originally used to talk about the dangerous and near-impossible task of ‘walking into Mordor’—together build the category of unachievable endeavors, to which meme communicators add new examples with each new bottom line of text. The strained look of concentration on Boromir’s face, and the ‘point-making’, focused accompanying hand gesture, are part of what lends the original film scene its intensity and significance. Part of the effect of the meme then lies in the almost anticlimactic application of this portentous context to everyday gripes and frustrations which are hardly of the life-threatening variety, resulting in a mildly mocking, ironic overall viewpoint of venting about life’s smaller difficulties, perhaps even hinting at an awareness of the (sexist) stereotypes involved.

In the Good Girl Gina example, the smiling face of Gina, resting on the right hand and looking happily and confidently into the lens, is meant to call up the personality type of the stock character of the ‘good girl’ (part of a cast of Meme Characters, including Scumbag Steve and Good Guy Greg, among others). Proficient meme communicators know, when they see the image macro of Good Girl Gina, that a generic “when/if X-then Y” ‘predictive’ reasoning (Dancygier, 1998) will be presented, giving another instance of considerate female behavior. In the example here, we could reconstruct the kind of meaning prompted for, with the implied sexist nod to the meme audience, as something like this: ‘when a considerate woman gets mad at you, then (unlike what you and I usually experience at the hands of women) she will explain the reasons for her anger so you can respond appropriately’. The integration of image and text is so tight in this meme construction that it allows suppression of the subject argument, normally required in English grammar, given that in this genre, it is provided visually, via the image macro: when (Good Girl Gina) gets mad at you. This can be seen as a multimodal application of Ruppenhofer and Michaelis’ (2010) notion of genre-based argument omission, which they discuss in relation to the language of, for instance, diaries or recipes, where the subject or the object is routinely left out thanks to the conventions of the specific genre (e.g., Went to the cinema for a diary entry; place on the stove and bring to a simmer for a recipe).

More generally, we would argue that for meme proficient communicators—people who are familiar with how memes construct meaning—the meanings of memes are as tightly constrained as those of ordinary linguistic constructions: for firmly entrenched meme constructions such as One Does Not Simply or Good Girl Gina, the memetic conventions and the form-meaning patterning very strongly cue the intended meanings. Of course, there may be occasional Meme Viewers who are not familiar with memetic grammar in general, or with a particular (perhaps only just emerging) memetic template, especially when it relies on culturally rich frames which the viewer may not be familiar with. For entrenched cases, though, knowledge of such frames (for instance, that of Boromir and The Lord of the Rings) is arguably less important: a communicator who is ‘very online’ will learn to understand and use the meme even without having seen the film scene, even if perhaps they may draw less aesthetic or humoristic enjoyment from the experience compared to Tolkien aficionados.

It is important to add at this point that the still images provided in internet memes are not always unambiguously interpretable in emotional terms. We may think we can easily ‘read faces’, and body language more broadly, but usually we have very rich contextual cues, missing from the still image. The more general point still, made by Barrett (2017), is that facial expressions do not automatically and unambiguously represent emotions. Thus in a case like Good Girl Gina, we construe our understanding of the Meme Character’s facial expression to fit the meaning of the meme. This is perhaps more striking still in some of the other stock characters that are part of the series: Scumbag Steve, for instance, does not look particularly awful or unpleasant (his expression could just as well be construed as nonplussed, clumsy or even shy), or the (less commonly used) character of Good Guy Greg does not show a face which we automatically take to be that of someone you can definitely trust and rely on to do the right thing. In the broader history of internet memes, one well-known example has come to be known as ‘Sudden Clarity Clarence’, described on Knowyourmeme.com as featuring a “young man at a party staring into the distance as if he is experiencing an epiphany.” The clue is in the “as if,” which illustrates Barrett’s point about interpretation; we do not actually know if there was indeed ‘sudden clarity’ or an epiphany taking place, but this is the meaning construed in the resulting internet meme.

Given these observations, it stands to reason that image, text, and memetic grammar cooperate in determining our interpretations of embodied features in the images featured in internet memes. As well, we assume the degree to which a given image is open to varying interpretations and emotional alignments can vary. Let us first briefly highlight a few examples—across different types of meme patterns—that are quite strongly premised on cueing for clear categories of emotional stance—essentially likes and dislikes—and where the main embodied features contributing to this stance construal are in Meme Characters’ bodies. Two that feature image macros—where the images are a constant recurring feature across the instances of the meme—are the Drake meme and the Two Guys on a Bus meme. The former features two contrasting stills from a music video by hiphop artist Drake, arranged in a grid with the two images on the left, and matching words (or sometimes pictures again) on the right (for an example, see Figure 3). The embodied postures in the two Drake images involve a very pronounced opposition, one showing dismissal and dislike (where Drake turns away from what is presented to him and displays an open-palm-out ‘blocking’ or ‘stopping’ hand gesture), the other suggesting acceptance and appreciation (Drake standing physically close to what is presented to him, sporting a satisfied grin and pointing the index finger toward the presented text, in apparent approval). The simultaneous presence of the two recurrent images helps Meme Viewers confirm that construal of contrasting stances is involved—firm rejection and blissful acceptance. An example where a single image macro incorporates two opposing stances is that of the Two Guys on a Bus meme, exemplified in Figure 4. This is a so-called labeling meme (e.g., Vandelanotte, 2021; Dancygier and Vandelanotte, 2025), in which parts of a single scene presented in an image are overlaid with words or phrases that do not identify those parts in the depicted scene, but rather call up an entirely different frame (the best-known example being the Distracted Boyfriend meme, applying the shift in romantic attention to other shifts in attention, preference or allegiance; see, e.g., Walker, 2023). In the Two Guys on a Bus meme, a desperately sad looking man on the left hand side of a bus is looking at a dark rock wall, whereas a happy looking man on the right hand side enjoys (and apparently photographs) a sunny view of a mountainous landscape; the labeling then applies this, typically, to contrasting stance-marked situations, e.g., ‘does not smoke cigarettes’ vs. ‘smokes cigarettes’. Importantly, the ‘good’ versus ‘bad’ option is represented in a rather complex way. There is the view outside (beautiful or boring), there is the idea of smoking (bad for your health) versus non-smoking (good), and the Meme Characters’ faces (happy/sad). But the way these stance contrasts align is complex: nice-view-and-good-mood are aligned with smoking (which should be marked as ‘bad’), while boring-view-and-bad-mood are aligned with non-smoking. It is this inconsistency that leads the Meme Viewer to the conclusion that the Meme Maker considers smoking to be good (something that brings pleasure)—and that way both sets of stances are coherently aligned. In other words, the generic stance toward smoking is overridden via meme-internal stance alignments, which are a more important determinant of stance than independent knowledge.

Figure 3

Figure 3. Drake meme. Reproduced from Imgflip.com.

Figure 4

Figure 4. Two Guys on a Bus meme. Reproduced from Imgflip.com.

Our examples so far have featured entrenched images—so-called image macros that are reused again and again, but with altered text. Non-entrenched images, that are not themselves fixed, can likewise serve as strong prompts for the construal of emotional stance. So-called when-memes, analyzed extensively in Lou (2017, 2021) as a case of multimodal simile, often present such cases. In a when-meme, there is a single when-clause, not completed textually, but instead followed by a suitable ‘ad hoc’ image to fill the missing ‘slot’ (the then-part of a predictive construction). The image is unrelated to the scenario described in the when-clause, except in one important respect: it shows a response or situation which feels the same (hence Lou’s proposal to treat when-memes as cases of ‘simile’). One example is given in Figure 5, where the recognizable situation of a boss asking an employee to do one final (possibly complex) bit of work just before work time is scheduled to end, makes the employee feel the same kind of annoyance or disappointment as could be seen in the facial expression of the actor Leonardo DiCaprio (not, in fact, being told to work late, but appearing at some red carpet event). Even if the disappointed face may seem to be just an expression of an individual’s emotion, we agree with Goodwin et al. (2012, p. 17) that “the scope of an emotion is not restricted to the individual who displays it,” and that “emotions constitute public forms of action.” Even though we only see one person depicted, that person’s emotion still communicates something in an interactive setting. The second person pronoun you typically featured in when-memes is generic, rather than deictic: there is no specific deictic ground with a specific addres(see Dancygier, 2021), and this evokes the assumption that the Meme Viewer can share the Meme Maker’s stance. The meaning of the meme overall is paraphrasable as a similative statement such as, When you are about to leave work and the boss says “before you go…,” you feel like Di Caprio felt when not being awarded an Oscar yet again.

Figure 5

Figure 5. When-meme depicting a disappointed face. Reproduced from Boredpanda.com.

Interestingly, several of Lou’s (2017, 2021) examples, and many other when-memes beyond, involve not humans, but animals—something also seen in purely textual simile (e.g., as proud as a peacock, cf. Veale, 2012). For instance, a photo of a koala attaching itself firmly to a person’s ankle completes the when-clause when you are at a party full of people you do not know so you u stay with ur friend the whole time (Lou, 2017, p. 121). Similarly, many when-memes use existing paintings (cf. Piata, 2020), as with the example of Degas’ famous absinthe drinker, preceded in a when-meme by the clause when you are on your lunch break and consider not going back. Thus in when-memes in general the images represent salient postures and facial expressions which match the experience described in the when-clause. The focus on embodied representations of emotions and experiences makes when-memes interesting cases of exploitation of visual representations of embodied behavior, especially because the person actually feeling something (disappointment or social awkwardness), and using the meme to express this feeling (i.e., the Meme Maker), is not represented in the image at all.

Such when-memes are thus quite directly about the similarity of stance evoked by two different situations: one described textually, the other, unrelated, prompted for in the image. More generally, the examples we have seen in this section show how images of faces, gestures and postures help prompt Meme Viewers to simulate interactional stance meanings, applying them to the situations and events described in the meme text. They achieve this despite the fact that in these examples, the meme does not show interaction between people, leaving a fuller interaction involving other participants to be inferred. In the next section, we turn to embodied interactions between multiple depicted bodies, multiplying the opportunities for stance expression.

3.2 Multiple bodies, multiple stances

We now turn to a different form of stance complexity: as soon as more than one body is represented in the image featured in a meme, an interpersonal dynamic can be activated, expressed in embodiment terms but also in concrete actions and interactions between the people depicted. The emergence of interpersonal stance configurations is not automatically evoked, as we could see in the Two Guys on a Bus meme, where the stances represented are relevant to the Meme Maker, not to a scenario in which the Meme Characters interact.

The popular Distracted Boyfriend meme, however, is a different case. Each individual Meme Character in the image macro has relevant facial features: suggesting admiration and heightened interest on the part of the central male character, confidence to the point of self-satisfaction in the passing girl in red, and annoyance and anger in the newly ignored girl in blue, holding hands with the male character. All of these interpretations only really make sense combined with the movements and postures shown, in particular the man’s turning to keep looking, over his shoulder, at the admired girl in red passing by. These emotional dynamics, played out in human relationships we can easily recognize, then become applied to an unrelated frame through the application of textual labels, as with the dieting frame in the example in Figure 6, where the lure of pizza proves too strong for someone supposedly on a diet. Overall, the meme provides an ironic comment on all too human foibles—preferring the new attractive thing over something that we already have and that is good for us. Another labeling meme that shows an interaction—in this case with very strong force dynamic impact (Talmy, 1988) resulting from a physical altercation—is the Will Smith Slapping Chris Rock meme, based on a still from an incident during the 2022 Oscars ceremony in which Smith slapped Rock in response to a joke the latter made about the former’s wife. In one example, for example, the attacker (Smith) is labeled as “Monday,” and the undergoer (Rock) as “me trying to enjoy the weekend,” applying the strongly negative, punishment-exacting stance to the effect the start of the work week has on people coming out of a weekend.

Figure 6

Figure 6. Distracted Boyfriend meme. Original photo by Antonio Guillem, licensed via Shutterstock.

In addition to depictions of people (or, as we discussed in Section 3.1, animals), memes may also rely on cartoon depictions. An interesting example which shows important embodied emotional stance is that of the Drowning High Five meme, based on a cartoon by artist Gudim. The event structure here is particularly rich, as four scenes develop across the four panels of the grid (see Figure 7): (1) a hand outstretched from a large expanse of water (suggesting “not waving, but drowning,” to quote the famous Stevie Smith poem); (2) another hand approaching from the upper left hand corner, suggesting an approaching offer of help, and thus a positive, hopeful emotional stance; (3) a high five gesture—itself expressive of very strong positive stance, usually congratulatory—being performed by the approaching hand; (4) back to the original, single hand sticking out of the water, but more deeply submerged than in the first frame, surrounded by bubbles suggesting further immersion into the deep waters—signifying that the hope felt earlier turned out to be but false hope, and no actual help was offered. A very large portion of existing Drowning High Five memes, including the example in Figure 7, apply the original scene of physical distress (drowning) to the domain of mental distress, as a way of commenting on well-meaning but ultimately useless advice around depression (represented by snippets of mental health advice being quoted, such as “hang in there,” or “you have such a good life compared to some, just be happy”).

Figure 7

Figure 7. Drowning High Five meme. Reproduced from Reddit.com user BasicSadBish, 27 March 2019.

The simulated expressivity in the examples above is closely correlated with some naturally recognizable aspects of facial expression and/or body posture. Throughout the discussion above we focused on aspects of embodied expression or behavior which are read as expressions of stance. There are differences in pathways of stance simulation across the examples we presented (and, naturally, many other examples), such that when-memes (Figure 5) may signal stance through easily recognizable facial expressions, while image macro memes like Good Girl Gina rely on the entire template to present the behavior described as pleasant and praiseworthy. Specific cases may vary, but it seems justified to postulate a formal dimension present in many memes—facial expression and body posture as tokens of stance.

However, while memes are in general a stance-focused and image-based genre of internet discourse, many of them also include token snippets of what looks like conversation. In Figure 5, for one, the disappointed facial expression is triggered by a quoted phrase “before you go,” which forces the Meme Character to abandon the plan to leave work for the day. The Drowning High Five grid meme shows quoted discourse as well, while also creating a story—a sequence of events which starts with a plea for help and ends in drowning. These two features—apparent discourse snippets prompting a narrative through correlation with an image or images—are a very salient aspect of memetic form and meaning. In Section 3.3 we will set the agenda for exploring memetic narratives and the role of apparent quotations in them, also showing how such discourse structures depend crucially on memetic images. Sections 3.4 and 3.5 will then offer a broader discussion of relevant examples and memetic meaning-making strategies, focusing on verbal interactions between speakers which may be explicitly depicted (Section 3.4) or, across a variety of formats, not depicted (Section 3.5).

3.3 Narratives, discourse snippets and memetic simulation

Our first example is the It Will Be Fun, They Said meme, in Figure 8. In general, this meme type condenses a story sequence across two lines of text and a picture: ‘advice – acting on the advice – distressing outcomes that contradict the original advice’. The ‘advice’ part is given in the piece of discourse presented in the top line of text: “Go to grad school, they said,” complemented by the positive stance expressed by the same unidentified “them” in the bottom line of text: “It will be fun, they said.” Combining this with the negative emotional stance prompted by the picture of the inconsolably crying man leads viewers to pragmatically infer a scenario where the advice was followed, but led to bitter disappointment. Note that the ‘lower’ positive and negative emotional stances of promised fun and actual distress feed into an overall Discourse Viewpoint (Dancygier and Vandelanotte, 2016), which takes an ironic stance toward the whole narrative ‘reversal of stance’ sequence.

Figure 8

Figure 8. It Will Be Fun, They Said meme. Reproduced from Quickmeme.com.

The meme in Figure 8 is a good example of the many ways in which memes construct stance: by telling a brief but emotionally loaded story, by using representations of conversational discourse as tokens of narrative events, and by profiling the image as a representation of the final event in the story. In Figure 8 the language of the meme is a more or less standard form of Direct Speech, where the utterances are made to look as if they were quoted verbatim, and the reporting clause (they said) completes the construction. We are referring to this example here as a predictable pattern of a narrative which includes dialogic pieces. In our further discussion, we will show how memes construct stories which build the dialogic structure by inserting images in their place. In this case, though, the image represents the concluding event in the sequence and supports the simulation of emotional meanings which constitute the core of the meme.

In what follows we will highlight the important role of simulation in the emergence of emotional meaning in memes such as the ones discussed here (cf. Sweetser, 2012; Bergen, 2012; Feyaerts et al., 2017b). In understanding linguistic and visual inputs, our brains “run embodied simulations” (Bergen, 2012, p. 195); in Barrett’s succinct summation, “[s]imulations are your brain’s guesses of what’s happening in the world” (Barrett, 2017: 27). Our capacity to recognize and relate to depictions of embodied experiences allows us to ‘fill in’ and also simulate the impact of the partial visual inputs the memes provide us with: we ‘feel’ (at least in an attenuated sense, by simulation), and can empathize with, the frustration felt by the girl in blue in the Distracted Boyfriend meme, for instance, or what it feels like to cry uncontrollably as in the It Will Be Fun, They Said meme; likewise, we can simulate, perhaps even with some sense of dread or of flinching respectively, what drowning feels like (in the Drowning High Five meme), or what a slap to the face is like (in the Will Smith Slapping Chris Rock meme). Even if animals, or cartoon characters, or film characters, rather than ‘ordinary’ human characters are depicted, as soon as bodies are depicted, we are primed and ready to be attuned to feelings felt and emotions experienced.

The simulations are further supported by the use of memetic quotation—i.e. not the most standard type of quotation taking place in fully detailed deictic grounds, with fully identified speaker, addressee, time and place coordinates (cf. Dancygier, 2021), but rather snippets of discourse being used to quickly evoke frames and attitudes. We zoom in on discourse exchanges of such a ‘fictive’ kind (Pascual, 2002, 2014; Pascual and Sandler, 2016) in the remainder of this paper, first in examples of internet memes showing Meme Characters engaging in verbal exchanges (section 3.4) and then in memes and X/Twittter discourse making broader innovative use of longer exchanges in a kind of faux dialogue format identifying speech participants not visually, but by means of noun phrases (“me,” “partisan Twitter,” etc.) (section 3.5). We will also discuss the role of images in these dialogic formats and argue that the complexity of such usage calls for a more specific investigation of how the unusual discourse and visual forms yield the representation of stance. We will refer to such instances as ‘simulated interaction’.

3.4 Depicted bodies interacting verbally in dialogue memes

In research on quoted speech generally, the view that quotations are demonstrations or ‘depictions’ (Clark and Gerrig 1990; Clark, 2016) has been highly influential—and suits many memetic quotation examples well: rather than having ‘actually’ been said by some identifiable speaker to a specific addressee, memetic quotes are used to quickly point up attitudes and stances, which can then be responded to by other components of the same meme. A typical, older example is that of the Said No One Ever meme (Dancygier and Vandelanotte, 2016), for instance in an example like “I love your Crocs, said no one ever”: the postposed speech clause “said no one ever” effectively reverses the viewpoint initially expressed. Apparently, in the Meme Maker’s view, the idea that anyone could love anyone’s Crocs is so laughable that it could not be expressed by anyone ever. Our earlier It Will Be Fun, They Said example (Figure 8) shows another typical kind of stance reversal, as do many examples of Be Like memes (Vandelanotte, 2019; Dancygier and Vandelanotte, 2025).

Particularly where more speakers are involved in a non-actual exchange, it is helpful to refer to the notion, developed in cognitive linguistics, that is related in spirit to that of quotations as demonstrations, namely that of “fictive interaction” (Pascual, 2002, 2014). The central idea is that in fictive interaction, the frame of a face-to-face conversation is used in language to structure meanings that in fact involve no actual conversation. Examples include fictive speech acts as in Call me old-fashioned, but… (which are not ‘actually’ asking an interlocutor to call the speaker old-fashioned), and uses of apparent Direct Speech snippets at lower structural levels, like at the phrasal level, as in a “yes we can” attitude. What happens in such types of cases, then, is that something we know from real interactions—having a speaker, an addressee, a speech event in which they participate—is borrowed in order to represent mental construals such as emotional stances.

Our examples below, and also in section 3.5, illustrate several types of memes which rely on the structure of a dialogue to construct stance, while not always using fully profiled dialogic structures; also, such memes do not always represent viable conversational discourse. We consider such uses to belong to the category of ‘simulated interaction’, where memetic conversational-looking discourse structures do not just rely on evocation of interactions where specific stances are involved (which would apply broadly to fictive interaction examples). In the cases we consider below, the evocation of the dialogue form as such (rather than a specific discourse turn with accessible meaning such as “yes we can”) is used to construct the representation of stance. Additionally, the construction of stance is supported via the very visual organization of the meme panels. In this section, we will consider three patterns which illustrate different strategies Meme Makers rely on.

We start by referring back to the Two Guys on a Bus meme. The contrasting stances there were allocated to two sides of the image macro, while there was no dialogic discourse added. The distinction between the stances was marked by different views outside the window of the bus, the facial expressions of the two men, and the labels identifying the target of positive/negative evaluation. For comparison, the Mad Men meme, illustrated in Figure 9, combines two stills from the drama series Mad Men, where a junior employee, with raised eyebrows, is shown saying “I feel bad for you” in the top image, and a more senior colleague in the bottom image, sharing an elevator ride after a pitch meeting, says “I do not think about you at all” while looking serious and frowning slightly. There is already text in the image macros: by convention, we read these lines of text (“I feel bad for you,” “I do not think about you at all”) as subtitles, and apportion them to the depicted speakers. On top of that, however, the meme adds labeling, thereby reusing the two-part image macro to apply the sequence of stance lead (pity/disappointment) and stance follow (rejection of the previous stance) to a range of topics—for instance pitting Europe against the US, or ‘the 49 other states’ against ‘New Jersey’. In the example we include here (Figure 9), Android users are identified as expressing pity, whereas iPhone users are arrogantly dismissive.

Figure 9

Figure 9. Mad Men meme. Reproduced from Imgur.com user RoyBattysDove, 8 September 2016.

Importantly, the top/bottom organization of the meme, and the not highly expressive faces and body postures weaken the assumed semblance of a conversational context (in the original scene, the characters are in fact standing next to each other in an elevator, both facing the door of the elevator, which explains the viewer-facing arrangement). In fact, the only indication of a discourse connection between the top and bottom panels is the pronoun you—the deictic pronoun identifying the addressee. The Meme Viewer has to create the interactive context needed solely on the basis of the content of the conversational lines and the use of deictic pronouns. It is worth noting, though, that deictic forms are a sufficient component to prompt the understanding of the entire meme in terms of simulated interaction.

A more complex, quite popular example is the Anakin and Padmé dialogue meme, using a four-cell grid to represent exchanges between two Star Wars characters (Figure 10). The basic grid structures a much more natural interactive pattern, with Anakin on the left-hand side, and Padmé on the right. The characters’ eye-gaze also suggests interaction. Additionally, the changing stances are clearly signaled via facial expression: the shot of Anakin’s face changes from the top left cell (where his line of text is “I’m going to change the world”) to the bottom to become more close-up and with an intense expression on his face. There is no text in this bottom left cell. Meanwhile the image of the Padmé character changes drastically, from smiling and unconcerned in the top right cell, to troubled and sombre in the bottom cell, suggesting a very different intonation being used for the dialogue line, repeated in both of her cells, “For the better, right?.” In this way, the left-to-right, top-to-bottom sequential ordering structures changing and evolving stances, in which the good intentions of the ‘opening gambit’ (in the top left cell) end up being called into question. Labeling and/or altering the text subsequently shifts the meaning of the original dialogue—which is more extensive in the actual film dialogue, and discusses forms of government (basically, democracy vs. dictatorship)—to the frame evoked by the textual and visual additions and alterations. In a simple case, as in Figure 10, the basic grid is not changed but labels are added in the four cells, here labeling Anakin as “Elon Musk” and Padmé as “The Internet” – presumably in a critique of Musk’s influence on cryptocurrency markets by tweeting out conflicting information. Many varieties of the basic meme exist—for instance changing the dialogue lines, changing the pictures of people (with other pictures or even paintings, e.g., of Napoléon and Josephine; see Dancygier and Vandelanotte, 2025) or using visual labels (e.g., superimposing a flag to identify “Anakin” with a specific country). All these varieties typically respect some important essentials: the zoom-in for the Anakin character, and lack of a second dialogue line; the very stark change in stance facially expressed by the Padmé character, and the nature of the grid supporting the idea of conversational interaction. Textually, there is an added effect of the (near-)repetition of Padmé’s dialogue line, the second iteration of it being uttered with strong tints of worry and disbelief, suggested by the embodied features. In this type of dialogic grid many elements combine to create stance—facial expressions represented, the nature of the default dialogic pattern, and the additional framing prompted by labeling. But, perhaps most importantly, by applying the dialogic pattern to a different frame (as with ‘Elon Musk/the Internet’), the meme is used to express a stance or an opinion on the content of that frame, e.g., Musk’s role in the changed perception of the Internet, and not to assume any spoken (fictive or otherwise) interaction between the entities profiled (Musk and the Internet). The attitudes and stances signaled via facial expressions and conversational quotes are applied to entities and situations in the frames, attributing stance to culturally rich concepts such as ‘the Internet’. Discourse snippets and facial expressions in the grid are thus used to simulate attitudes in entities not represented in the memes as such.

Figure 10

Figure 10. Anakin and Padmé meme. Reproduced from Reddit.com user XipingVonHozzendorf, 22 May 2021.

A more recent example showing speech participants visually, and ‘relabeling’ their dialogue, is the David Beckham “Be Honest” meme. The meme presents a three-by-two grid showing stills from a documentary about former footballer David Beckham and his wife Victoria in which she (depicted on the left) talks about them both being “very working class,” and he (shown on the right, listening in on his wife’s interview from behind a door) insists repeatedly she should “be honest” until she relents and answers his question about what car her dad drove her to school in (a Rolls Royce, as it happens—not “very working class,” then). On the “David” side of the grid, the bottom right image shows him, eyes closed and eyebrows raised, responding with “Thank you,” in recognition of his wife finally properly conceding (as he sees it) his point. Figure 11 shows an example of the “Be Honest” meme, which maintains the text on the “David” side, but modifies the “Victoria” side to make it about something else entirely.

Figure 11

Figure 11. David Beckham “Be Honest” meme. Reproduced from X.com user AWrites116, 6 June 2024.

The form of the grid, with three tiers organized top-to-bottom, and gradually changing the stance represented (more insistence on the David side prompting more toning down on the Victoria side), additionally relies on the left/right allocation of conversational lines and facial expressions to the two participants in the conversation. It is enough to imagine the effect of any changes in the organization of the grid (such as, putting Victoria on the right in the middle tier, or moving David’s bottom tier image up top) to note that the conversation which viewers are asked to simulate is structured to represent the specific stance-shift happening throughout the conversation. The facial expression, the conversational turns and the nature of the grid jointly organize the simulated interaction involved.

The original exchange in the Beckham documentary is a fascinating piece of dialogue in its own right with the husband, listening in on his wife’s interview from behind a door, intervening to force her to be more truthful. But it is perhaps even more fascinating to see how this was given a second lease of life as a meme in which people are forced to recognize that the things they say (e.g., “I wrote for 3 hours”) often do not accurately reflect what is actually happening (e.g., hardly having written anything at all—a feeling all too recognizable to many writers). The meme is presumably not used for straight-faced, hard-hitting critique, as suggested by the smiling expression on David Beckham’s face, or indeed the light tone of examples circulating on the Internet. Rather, it installs a mildly ironic, sympathetic distance (Tobin and Israel, 2012) from which to consider the human weakness in overstating one’s credentials or achievements. Thus, proficient meme communicators who know what this particular artifact is used for, will have no difficulty identifying the emerging meaning and getting the meme’s general point, regardless of the specific topics it is applied to: ‘sometimes we all like to present ourselves as just that little bit better or more virtuous than we can actually truthfully claim to be.’

The examples we turn to in the next section add further dimensions of the concept of simulated interaction—including, for instance, the use of non-text lines (enclosed within asterisks) describing behavior, or indeed images, as turns in a dialogic exchange. These seem to us surprising and quite complex forms used in stance evocation, quite distinct from the use of existing conversational patterns (like in fictive interaction) if only because some of the forms centrally relied on do not actually appear in conversation at all.

3.5 Simulated interaction formats not depicting speakers in memes and online discourse

We have seen the role that representations of embodied states and patterns of conversation play in simulated interaction. In this section we look more closely at the structure and functions of dialogic sequences in memes and other forms of online discourse which do not as such depict apparent ‘interlocutors’ in the way the Mad Men, Anakin and Padmé, and David Beckham “Be Honest” memes discussed in the previous section do. We will consider the role of pictures and emoji in different dialogue formats (using quotation marks in alternating lines, or using noun phrases to introduce different speakers). We will also discuss the use of the Me pronoun, referring to a Meme Character who represents the Meme Maker’s experience. Finally, we will also highlight the absence of speech, and the staging of ‘non-speakers,’ in stance evocation.

We start with two examples that present turns in an apparent dialogue partly as discourse enclosed in quotation marks, and partly as pictures, which are profiled as turns or moves in conversational sequences. As such examples show, memetic discourse is not naturally interpretable in terms of familiar discourse categories such as ‘narrative’ or ‘conversational’. In many cases, for example, fictive dialogic lines are used as representations of narrative events, and narrative sequence is represented by the sequence of lines in a faux dialogue used in the meme. As many of our examples show, such faux-dialogic lines can further be substituted with images, which have a higher humorous and stance-forming impact. Our contention is that the memetic re-construal of discourse types has two goals—(1) creating humor, while (2) simulating stance-loaded responses to the situations described.

Both dialogic and image-based construal can be seen in Figure 12 (@Stephenlough95 on X/Twitter, 24 December 2021), where the process presumably starts from the picture of a COVID self-test being forged by adding in, in red pencil, the all-important second red line indicative of a positive test result. This picture was, for some time, popular on X/Twitter, and led several online communicators to imagine suitable fictive utterances, inviting someone to perform a duty or chore they would rather get out of, to which the picture provided the desired solution—getting out of something one does not want to do. The image in Figure 12 fictively responds to the fictive request to set the table by using an image as a token metonymically representing a verbal response such as “I cannot, I have COVID-19.” Overall, the piece of simulated interaction communicates a stance about the task being avoided (dislike), but arguably also comments on the pandemic, making light of its new-found rituals, in this case regular self-testing.

Figure 12

Figure 12. Faked Covid Test meme. Reproduced from X.com user Stephenlough95, 24 December 2021.

Figure 13 (@McJesse on X/Twitter, 15 January 2018) shows a longer virtual exchange, with three lines of dialogue, which we are lexically prompted (by items such as “bae,” for “babe,” and talk of “parents” and “coming over”) to understand as being between young lovers. The impossibility of coming over safely (“getting my brakes fixed”) turns out to be swiftly ignored when the exceptional circumstance of absent parents is mentioned—the promise this holds appears to warrant extreme risk taking, but the picture shows the unlucky outcome in which “bae” crashed the car into a house along the way. The lines of fictive dialogue here are presented using quotation marks, without any accompanying reporting clauses to identify the speakers (whose generic identity we glean from context, as suggested above). This further underscores the fictive nature of the simulated interaction: no specific referents are involved at all, but a type of situation is pithily illustrated, including its disastrous outcome, exaggerated to the point of absurdity. The switch from a typical and realistic nature of the first three lines of the dialogue, to the highly evocative image representing a disastrous result creates humor through exaggeration. In a sense, such instances, similarly to when-memes, evoke a simile—a trope known for creating vivid and exaggerated imagery—so that the point of the joke is a suggestion that a young lover would prefer to risk certain death rather than miss the opportunity to be with someone they love.

Figure 13

Figure 13. Bae Come Over meme. Reproduced from X.com user McJesse, 15 January 2018.

Another strategy, very common on microblogging platforms such as X/Twitter, is that of adopting the type of dialogue format sometimes found in press interviews, in which different interlocutors are identified by means of noun phrases preceding a colon (cf. Vandelanotte, 2020, 2021; Dancygier and Vandelanotte, 2025). In Example (1), the rather broadly defined collective ‘speaker’ (people on Earth) changes their emotional stance completely after an intervention on the part of the starry ‘night sky’—an intervention consisting in simply existing (and, we assume, in the communicator’s view, being beautiful). The initial stance is represented by the basic sad face emoticon :( and the altered stance after consideration of the night sky by its counterpart, the basic happy face emoticon :). While the visual component here is really quite minimal (using emoji that are part of the available character set), we also have very extensive examples using only emoji. One such example (discussed in Dancygier and Vandelanotte, 2025) represents a Brexit negotiation between ‘the UK’ and ‘the EU’—the two partners in the negotiation being represented as ‘speakers’ by means of their respective flag emoji preceding the ‘speaker’-introducing colon—with the content of the exchanges being expressed solely using emoji, colons and arrows. An example such as this clearly shows how the general schematic form of a dialogue is used in quite original and unusual ways to evoke stance, and change in stance. Rather than take an existing form of possible ‘real’ dialogue and use it for another end, as in fictive interaction, here the apparent dialogue form itself is altered to allow things it does not actually allow in face-to-face conversations.

(@poetastrologers on X/Twitter, 13 April 2019)

The example in Figure 14 features use of the ‘me’ pronoun often found in memetic discourse to represent the Meme Maker’s experience —for instance, of impatience. ‘Me’ is the memetic pronoun of choice, playing a role very much like a demonstrative marker—‘what you see in the image represents my behavior/posture/facial expression/etc./ in the way that allows you (the Meme Viewer) to recognize the experiential or emotional stance I am describing’. The use of ‘me’ is a common feature of memetic discourse, as it allows the Meme Maker to represent their emotional stance in a somewhat impersonal and humorous way.

Figure 14

Figure 14. ‘Me’ meme. Reproduced from Reddit.com user dalex0001, 21 October 2017.

Another very common pattern, the Me/Also Me meme, relies on clearly describing the Meme Maker’s intentions or plans, only to show their failure to live up to them; one example we collected, for instance, shows as ‘Me’ line “I need to save money this month,” and includes a picture of a woman with lots of shopping bags for the ‘Also Me’ part. In Figure 14, though, we see a faux-dialogic formula used to display types of behavior in response to recognizable everyday situations. In Figure 14, a parcel is scheduled for delivery (metonymically evoked by means of the fictive quote “your order has been shipped”), and the apparent dialogic response on the part of the ‘me’ faux-conversationalist is not anything verbal, but a depiction of the kind of expectant, hopeful attitude—as it happens, here embodied by the Meme Character of a dog sat on a chair and peering longingly out of the window.

For comparison, Figure 15 looks as if it is quite predictably structured as dialogue, with ‘me’, then ‘person’, then ‘me’ again. But in fact, each of these three lines is quite different, and none of them introduces speech, not even fictive speech. The first me-line describes what precedes the main event in focus—the Meme Maker’s intention to leave a bench; the second line then introduces another Meme Character, “person,” whose ‘discourse’ move is to appear on the scene, and take a seat on the bench. The final part of the meme uses the me VERBing pattern, describing the Meme Character (representing the Meme Maker) as waiting a bit before leaving, so as not to awkwardly suggest they are leaving because of the arrival of “person.”³ Here, a demonstration does follow the me phrase, even if it is not a verbal quote: a depiction of a man (actor Keanu Reeves) sitting on a bench sporting a somewhat blank expression. Even if, using standard conventions of dialogue in writing, the meme example here seems very complex and unusual, the artifact works because it builds on existing conventions of meme grammar—particular uses of pronouns (me typically representing the Meme Maker’s views), roles such as Meme Maker and Meme Character, constructions such as me VERBing and its use with images. It also works because of the recognizability and relatability of the behaviors and stances described —we recognize the awkwardness of the situation, but we can also share with the Meme Maker a sense of the humor or even silliness of how we try to deal with this awkwardness.

Figure 15

Figure 15. ‘Me/Person’ Man on the Bench meme. Reproduced from Reddit.com user cej98, 19 July 2020.

We should also note here how this particular meme uses the faux-dialogic convention while using grammatical forms that would typically be considered unacceptable. First, ‘Me:’ is followed by a third person form is—which follows at least two memetic discourse conventions: using me to point to the Meme Character representing the Meme Maker’s experience in the event described, and the third person is to follow the standard memetic usage which avoids talking about Meme Characters as if they were simply the same as the Meme Maker. In the next line, the text sits next to me is framed by asterisks, now a standard memetic convention to represent *action*. It is only in the final line, after the two characters in the story being told are established, that the Meme Maker can use the first-person pronoun to elaborate on their experience. Overall, this rather complex narrative meme uses faux dialogue in a rich way in order to simulate an emotional stance which explains superficially useless behavior (staying on when wanting to leave) in order not to hurt the feelings of a complete stranger—a situation and a stance that are relatable to the Meme Viewer.

The use of ‘silent’ lines, as with the night sky merely “being starry” in Example (1), or a person “sitting next to me” in Figure 15, is striking. Earlier, we also noted the use of silence as a ‘conversational turn’ in the Anakin and Padmé memes (Figure 10)—the line (combined with Anakin’s impassive stare) which makes Padmé realize that her assumption about Anakin’s good intentions may be a mistake. A further related pattern was identified in Vandelanotte (2020), exemplified here in Example (2), in which a series of “empty” quotes, on the part of ‘non-speakers’ (“nobody,” “absolutely no one,” etc.) leads up to a culminating turn in which someone says something completely uninvited, and presented as irrelevant and unwanted. In the example, an ‘iNfLuEnCeR’—the spelling reveals the online communicator’s disdain for the mere category of people—starts a social media post as “A lot of you have asked about my skin care routine” (presumably continuing to advertise a particular brand of which they received freebies). The contrast between all the preceding lines of ‘nothing being said or asked’, and the influencer’s fictive statement about “a lot of you” having asked about skin care, further underscores the negative emotional stance toward the way influencers operate. Interestingly, silent ‘turns’ in memes and other online discourse forms suggest online language users are acutely aware of the potential significance of silence in dialogue—a topic that is very actively explored in current approaches to conversation analysis (e.g., Hoey, 2020, 2021). Overall, (2) suggests an absence of speech, made expansive by dividing it over multiple lines of ‘non-dialogue’ on the part of ‘non-speakers’, followed by a stretch of unwanted, unasked-for speech—the influencer’s intervention, presented as an annoying, self-centered and financially motivated form of engagement.

(2) Nobody:

Absolutely no one:

Not a single soul on this Earth:

Not even their mom:

iNfLuEnCeR: “A lot of you have asked about my skin care routine...”

(@cdcxpe on X/Twitter, 16 April 2019)

Note, again, in these examples how they deviate from existing conversational structure: the use of images, emoji, ‘action’ lines like *Sits next to me* presented where a dialogue line ought apparently to go, or a sequence of ‘absent’ dialogue lines from ‘non-speakers’ (nobody, absolutely no one, etc.) all do things which work in online visual-textual contexts, but not in the actual kinds of dialogic expressions the notion of fictive interaction is premised on. The more specific notion of simulated interaction is proposed to fill this gap.

4 Conclusion

We have tried to show, in this contribution, that despite their important differences, there is much that unites the ‘embodied interaction’ and ‘polysemiotic’ paradigms of multimodality research. Viewed from one perspective, the data types we have been presenting are far removed from face-to-face interaction (even ‘mediated’ online forms such as Zoom calls). At the same time, however, we see that even the snapshot depictions of embodied features—including facial expressions, gestures, (changes in) postures—however much reduced they may be compared to full face-to-face interaction, contribute much to the emergence of meaning, especially stance meanings, in internet memes. Likewise, the use of simulated dialogue formats in ways that are hard to imagine in face-to-face exchanges (presenting descriptions of behavior, for instance, or pictures, as ‘turns’ in the exchange) suggests interesting new ways of using the basic interactional frame of a dialogic exchange. Over the course of our exploration, we have suggested that online discourse is taking on specific features of embodied interaction, and developing strategies that constitute a new formula, which we have termed simulated interaction. Importantly, these communicative strategies specifically serve the needs of stance construction, the marking of contrasting stances, and also stance negotiation or shift. In other words, the wealth of stance phenomena online is developing expressive means which exploit the online reliance on visual means of expression and the need for brevity and clarity, while re-framing forms of interaction, both embodied and discourse-based. The new formulae of stance simulation call for an in-depth discussion, and this paper shows some potential directions worth exploring.

We have proposed the notion of simulated interaction as a more suitable concept, compared to fictive interaction, for the multimodal types of online discourse we have investigated. For one thing, some of our examples, such as the Distracted Boyfriend meme, concern not conversations, but purely embodied interactions (without a verbal component), which are sufficient prompts for Meme Viewers to construe stance exchanges. In addition, we have described a number of unique forms specific to online discourse which do not exist in ‘actual’ conversations and so cannot properly be analyzed as applying the existing ‘conversation frame’. These include the use of images, emoji, descriptions of behavior (such as “Sits next to me”), or indeed of ‘absent’ speech by ‘non-speakers’, as apparent discourse moves. More broadly, we have seen various forms which blur the boundaries between narrative and conversation: examples such as the Bae Come Over meme or the ‘Me/Person’ Man on the Bench meme use long ‘faux’-dialogic exchanges to in fact structure narrative sequences, with images representing either an emotionally loaded climactic event (Bae Come Over) or embodied features (Man on the Bench). Other examples, too, mix discourse snippets and narrative sequence (e.g., the It Will Be Fun, They Said meme, or the dialogic grid memes) in the pursuit of stance expression. We are thus proposing a concept which reflects the richness and complexity of online communication. It also relies in complex ways on two modes of multimodal discourse—multimodality in interaction (including gesture, body posture, facial expression, etc.) and image-text multimodality; specifically, images can be inserted in slots (such as dialogue entries) normally reserved for text, and play a broad array of stance evocation and narrative roles.

Overall, we have tried to show that much of the formal innovation observed in internet discourse is driven by a number of important discourse goals. The primary such goal is to represent experience, while not making the artifacts uniquely connected to the internet user’s life and beliefs—hence, for instance, the appeal to generic pronouns or noun phrases such as “person,” or the drive to connect experiences between unrelated frames and situations (e.g., a distracted boyfriend and a failure to follow a diet). Much of what is being communicated is opinions, reflections, humorous commentary, irony, and sometimes sheer amazement at things we say and do to maintain the social fabric, while being aware of our own flaws and quirks. Humor itself is also a goal, as is brevity or ‘linguistic economy’—if a couple of faux dialogic lines will do, why tell a long story? In this pursuit of pithiness, the use of images is especially effective, and as we have seen, images are used liberally to represent experience, and they can be naturally embedded in what looks like ordinary dialogue. These strategies of stance expression form the grounding of what we have termed ‘simulated interaction’. Further research will, we hope, throw more light on the ways in which online communicators respond to, but also co-construct, instances of stance expression through simulated interaction (see, e.g., Dancygier and Vandelanotte, 2025, Chapter 10).

Data availability statement

Publicly available datasets were analyzed in this study. The data analyzed in this study is part of an ongoing collection of examples, collected by both authors from various public online sources over the years, for purposes of qualitative (not quantitative) analysis. The same or similar examples can be found on meme collecting sites, social media platforms or using online search engines. Further inquiries can be directed to the corresponding author.

Ethics statement

Ethical approval was not required for the study involving human data in accordance with the local legislation and institutional requirements. Written informed consent was not required, for either participation in the study or for the publication of potentially/indirectly identifying information, in accordance with the local legislation and institutional requirements. The social media data was accessed and analyzed in accordance with the platform’s terms of use and all relevant institutional/national regulations.

Author contributions

BD: Writing – original draft, Writing – review & editing, Conceptualization. LV: Writing – original draft, Writing – review & editing, Conceptualization.

Funding

The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. Funding allowing open access publication was provided by The University of British Columbia, Canada.

Acknowledgments

Our thanks to the editors of the Research Topic on Stance-Taking in Embodied and Virtual Interaction for inviting us to contribute, and to reviewers for valuable feedback.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

1. ^While we recognize the distinctions drawn in psychology between affect and emotion (as reported, for instance, in Barrett (2017): Chapter 4), we accept that the term ‘affective stance’ has gained wider currency in linguistics-oriented work, and note that the term is used interchangeably with ‘emotion as stance’ in Goodwin et al. (2012). We think our examples of stance expression are more complex and fine-grained, and more situation-bound, than the kind of basic, ever-present sense of affect (in its dimensions of valence and arousal) described by Barrett (2017), and so seem to have more to do with emotion.

2. ^An internet meme generally has many creative mothers and fathers. Even in a case where one might know specifically who made one particular meme object, it will rely on pre-existing and pre-circulating visual material as well as on the pattern of variation communally established; more often, it is simply not, or no longer, possible to verify which internet account (let alone which ‘biographical’ person) posted a particular iteration or innovation. For a thoughtful reflection on this, we refer to Milner’s (2016, pp. 221–232) appendix on methods and ethics. Wherever possible, in Section 3.5 in particular, where we can source an example to a specific post on X/Twitter, we have provided the source information.

3. ^For reasons of space, we do not elaborate our approach here in terms of mental spaces and blending (e.g., Fauconnier and Turner, 2002). For a treatment of many aspects of meaning-making in memes along these lines, we refer to Dancygier and Vandelanotte (2025).

References

Andries, F., Meissl, K., de Vries, C., Feyaerts, K., Oben, B., and Sambre, P. (2023). Multimodal stance-taking in interaction – a systematic literature review. Front. Commun. 8:1187977. doi: 10.3389/fcomm.2023.1187977

Crossref Full Text | Google Scholar

Attardo, S. (2023). Humor 2.0: How the internet changed humor. London: Anthem Press.

Google Scholar

Barrett, L. F. (2017). How emotions are made: The secret life of the brain. Boston & New York: Houghton Mifflin Harcourt.

Google Scholar

Bergen, B. (2012). Louder than words: the new science of how the mind makes meaning. New York: Basic Books.

Google Scholar

Bülow, L., Merten, M.-L., and Johann, M. (2018). Internet-Memes als Zugang zu multimodalen Konstruktionen. Zeitschrift Angewandte Linguistik 2018, 1–32. doi: 10.1515/zfal-2018-0015

Crossref Full Text | Google Scholar

Clark, H. H. (2016). Depicting as a method of communication. Psychol. Rev. 123, 324–347. doi: 10.1037/rev0000026

PubMed Abstract | Crossref Full Text | Google Scholar

Clark, H. H., and Gerrig, R. J. (1990). Quotations as demonstrations. Language 66, 764–805.

Google Scholar

Couper-Kuhlen, E., and Selting, M. (2017). Stance and footing. Supplementary online chapter “C” Of interactional linguistics: Studying language in social interaction. Cambridge: Cambridge University Press.

Google Scholar

Dancygier, B. (1998). Conditionals and prediction: Time, knowledge and causation in conditional constructions. Cambridge: Cambridge University Press.

Google Scholar

Dancygier, B. (2012). “Negation, stance verbs, and intersubjectivity” in Viewpoint in language: A multimodal perspective. eds. B. Dancygier and E. Sweetser (Cambridge: Cambridge University Press), 69–93.

Google Scholar

Dancygier, B. (2017). Viewpoint phenomena in constructions and discourse. Glossa 2:37. doi: 10.5334/gjgl.253

Crossref Full Text | Google Scholar

Dancygier, B. (2021). Fictive deixis, direct discourse, and viewpoint networks. Front. Commun. 6:624334. doi: 10.3389/fcomm.2021.624334

Crossref Full Text | Google Scholar

Dancygier, B., and Vandelanotte, L. (2016). “Discourse viewpoint as network” in Viewpoint and the fabric of meaning: Form and use of viewpoint tools across language and modalities. eds. B. Dancygier, L. Wei-lun, and A. Verhagen (Berlin & Boston: De Gruyter Mouton), 13–40.

Google Scholar

Dancygier, B., and Vandelanotte, L. (2017a). Viewpoint phenomena in multimodal communication. Cogn. Linguist. 28, 371–380. doi: 10.1515/cog-2017-0075

Crossref Full Text | Google Scholar

Dancygier, B., and Vandelanotte, L. (2017b). Internet memes as multimodal constructions. Cognit. Linguist. 28, 565–598. doi: 10.1515/cog-2017-0074

Crossref Full Text | Google Scholar

Dancygier, B., and Vandelanotte, L. (2025). The language of memes: patterns of meaning across image and text. Cambridge: Cambridge University Press.

Google Scholar

Dawkins, R. (1976). The selfish gene. Oxford: Oxford University Press.

Google Scholar

Denisova, A. (2019). Internet memes and society: Social, cultural, and political contexts. New York: Routledge.

Google Scholar

Droz-dit-Busset, O. (2022). “So-called influencers”: Stancetaking and (de)legitimation in mediatized discourse about social media influencers. Discourse Context Media 49:100629. doi: 10.1016/j.dcm.2022.100629

Crossref Full Text | Google Scholar

Du Bois, J. W. (2007). Stancetaking in discourse: Subjectivity, evaluation, interaction, 139–182. Amsterdam and Philadelphia: John Benjamins.

Google Scholar

Fauconnier, G., and Turner, M. (2002). The way we think: Conceptual blending and the Mind’s hidden complexities. New York: Basic Books.

Google Scholar

Feyaerts, K., Brône, G., and Oben, B. (2017a). “Multimodality in interaction” in The Cambridge handbook of cognitive linguistics. ed. B. Dancygier (Cambridge: Cambridge University Press), 135–156.

Google Scholar

Feyaerts, K., Oben, B., Lackner, H. K., and Papousek, I. (2017b). Alignment and empathy as viewpoint phenomena: the case of amplifiers and comical hypotheticals. Cognit. Linguist. 28, 485–509. doi: 10.1515/cog-2016-0109

Crossref Full Text | Google Scholar

Goodwin, M. H., Cekaite, A., and Goodwin, C. (2012). “Emotion as stance” in Emotion in interaction. eds. A. Peräkylä and M.-L. Sorjonen (Oxford: Oxford University Press), 16–41.

Google Scholar

Hoey, E. M. (2020). When conversation lapses: The public accountability of silent Copresence. Oxford: Oxford University Press.

Google Scholar

Hoey, E. (2021). “Sacks, silence, and self-(de)selection” in On sacks: methodology, materials, and inspirations. eds. R. J. Smith, R. Fitzgerald, and W. Housley (London: Routledge), 130–142.

Google Scholar

Jenkins, E. S. (2014). The modes of visual rhetoric: circulating memes as expressions. Q. J. Speech 100, 442–466. doi: 10.1080/00335630.2014.989258

Crossref Full Text | Google Scholar

Kang, J.-i., Hanbeom Jung, A., Kwon, Y., and Kwon, I. (2023). Meaning extensions of internet memes: a case study of the ‘if 2020 was a(n) X’ meme. Rev. Cognit. Linguist. 21, 178–209. doi: 10.1075/rcl.00131.kan

Crossref Full Text | Google Scholar

Lou, A. (2017). Multimodal simile: the “when” meme in social media discourse. English Text Construct. 10, 106–131. doi: 10.1075/etc.10.1.06lou

Crossref Full Text | Google Scholar

Lou, A. (2021). Multimodal similes: Toward a cognitive understanding of Similative meaning. Vancouver: PhD dissertation, The University of British Columbia.

Google Scholar

Milner, R. M. (2016). The world made meme: Public conversations and participatory media. Cambridge, Mass: The MIT Press.

Google Scholar

Ochs, E. (2006). “Linguistic resources for socializing humanity” in Rethinking linguistic relativity. eds. J. J. Gumperz and S. C. Levinson (Cambridge: Cambridge University Press), 407–437.

Google Scholar

Pascual, E. (2002). Imaginary trialogues: Conceptual blending and fictive interaction in criminal courts. Utrecht: LOT.

Google Scholar

Pascual, E. (2014). Fictive interaction: The conversation frame in thought, language, and discourse. Amsterdam & Philadelphia: John Benjamins.

Google Scholar

Pascual, E., and Sandler, S. (Eds.) (2016). The conversation frame: Forms and functions of fictive interaction. John Benjamins: Amsterdam & Philadelphia.

Google Scholar

Paz, M. A., Mayagoitia-Soria, A., and González-Aguilar, J.-M. (2021). From polarization to hate: portrait of the Spanish political meme. Social Media Society 7. doi: 10.1177/20563051211062920

Crossref Full Text | Google Scholar

Piata, A. (2020). Stylistic humor across modalities: the case of classical art memes. Internet Pragm. 3, 174–201. doi: 10.1075/ip.00031.pia

Crossref Full Text | Google Scholar

Ross, A. S., and Rivers, D. J. (2017). “Internet memes as polyvocal political participation” in The presidency and social media: Discourse, disruption and digital democracy in the 2016 presidential election. eds. D. Schill and J. A. Hendricks (New York: Routledge), 285–308.

Google Scholar

Ruppenhofer, J., and Michaelis, L. A. (2010). A constructional account of genre-based argument omissions. Construct. Frames 2, 158–184. doi: 10.1075/cf.2.2.02rup

Crossref Full Text | Google Scholar

Shifman, L. (2014). Memes in Digital Culture. Cambridge, MA: The MIT Press.

Google Scholar

Sweetser, E. (2012). “Introduction: viewpoint and perspective in language and gesture, from the ground down” in Viewpoint in language: A multimodal perspective. eds. B. Dancygier and E. Sweetser (Cambridge: Cambridge University Press), 1–22.

Google Scholar

Talmy, L. (1988). Force dynamics in language and cognition. Cogn. Sci. 12, 49–100. doi: 10.1207/s15516709cog1201_2

Crossref Full Text | Google Scholar

Tobin, V., and Israel, M. (2012). “Irony as a viewpoint phenomenon” in Viewpoint in language: A multimodal perspective. eds. B. Dancygier and E. Sweetser (Cambridge: Cambridge University Press), 25–46.

Google Scholar

Vandelanotte, L. (2019). Changing perspectives: something old, something new. Pragmatics 29, 170–197. doi: 10.1075/prag.18046.van

Crossref Full Text | Google Scholar

Vandelanotte, L. (2020). (Non-)quoting and subjectivity in online discourse. E-rea 17.2. doi: 10.4000/erea.9782

Crossref Full Text | Google Scholar

Vandelanotte, L. (2021). Creative constructs, constructions, and frames in internet discourse. Construct. Frames 13, 160–191. doi: 10.1075/cf.00050.van

Crossref Full Text | Google Scholar

Vandelanotte, L., and Dancygier, B. (2017). Multimodal artefacts and the texture of viewpoint. J. Pragmatics 122, 1–9. doi: 10.1016/j.pragma.2017.10.011

Crossref Full Text | Google Scholar

Veale, T. (2012). Exploding the creativity myth: The computational foundations of linguistic creativity. London: Bloomsbury.

Google Scholar

Walker, Aidan. (2023). How to read a meme: Context, agency and knowledge in distracted boyfriend. New York: MA thesis, Columbia University.

Google Scholar

Wiggins, B. E. (2019). The Discurse power of memes in digital culture: Ideology, semiotics, and intertextuality. London: Routledge.

Google Scholar

Wiggins, B. E., and Bower, G. B. (2015). Memes as genre: a structurational analysis of the memescape. New Media Soc. 17, 1886–1906. doi: 10.1177/1461444814535194

Crossref Full Text | Google Scholar

Wikström, P. (2019). Acting out on twitter: affordances for animating reported speech in written computer-mediated communication. Text Talk 39, 121–145. doi: 10.1515/text-2018-2021

Crossref Full Text | Google Scholar

Xie, C. (2022). The pragmatics of internet memes. Amsterdam: John Benjamins.

Google Scholar

Yus, F. (2018). Identity-related issues in meme communication. Internet Pragmatics 1, 113–133. doi: 10.1075/ip.00006.yus

Crossref Full Text | Google Scholar

Zappavigna, M., and Logi, L. (2024). Emoji and social media paralanguage. Cambridge: Cambridge University Press.

Google Scholar

Zenner, E., and Geeraerts, D. (2018). “One does not simply process memes: image macros as multimodal constructions” in Cultures and traditions of wordplay and wordplay research. eds. E. Winter-Froemel and V. Thaler (Berlin and Boston: De Gruyter Mouton), 167–193.

Google Scholar