Exigence at the Dawn of Recommendation Media: Dramatizing Salience in Audio Memes

ABSTRACT

This article looks at how exigence is made publicly observable in user-based media operating on recommendation algorithms. Messaging in these rhetorical environments often takes the form of imitative behaviors rather than statements inviting a direct response. Examined in the article are two audio memes from TikTok representing two modes of imitation: one a physical imitation meme associated with the Woman, Life, Freedom protests in Iran, and the other a narrative imitation meme where participants objectify endemic social problems. The findings suggest that the responsorial imperative of audio memes can either intensify the speed and urgency with which an exigence is experienced, or it can bring urgency to endemic problems. The studies also find that the formal qualities of a given audio meme constrain both how an exigence is communicated as well as what kinds of exigences the meme can be taken up for in the first place.

KEYWORDS:

• Audio memes
• exigence
• genre studies
• mimesis
• the rhetorical situation
• TikTok

The Twilight of Social Media

It’s possible that, in years to come, the third decade of the twenty first century will be seen as the twilight of social media. It won’t be because giants like Facebook and X (formerly Twitter) are mired in scandal and plummeting levels of credibility (though they are). It won’t even necessarily mean that those giants will not remain giants; but they will not look the same as they did in the days of social media. That’s because something more appealing has come along, something that delivers novelty-packed dopamine spikes even more efficiently than the endless scrolling function ever could. Reacting to Meta’s announcement in July 2022 that Facebook would be converting its newsfeed to a recommendation algorithm similar to that of its other property, Instagram, in order to compete with the astronomical success of TikTok, tech financier Michael Mignano coined the term recommendation media, predicting that it would become the hegemonic mode of user-based media production for the foreseeable future (Mignano, “End of Social Media”). The difference, Mignano explains, between social media and recommendation media is that the former relies on the construction of social graphs to distribute content.¹ Under the social media model, a user sees content either from those whom they follow or that with which another user in their network has engaged, and advertising as well as the prioritization of content in a user’s feed is based upon the strengths of those networks. While a user may follow others and be followed in a recommendation medium, the algorithm also recommends content based upon predictions about the user’s evolving interests as well as content that may be trending at the moment, exposing the user to content they would not have otherwise seen from inside of a social graph. Depending on engagement – which can simply mean the user watches a video all the way through – the thematic trajectory of the user’s feed can change in situ. Rolling, in other words, displaces scrolling. This produces an unlikely combination of hyper-personalized interfaces and delivery of more diverse content, diminishing the influence of a user’s social network on the mediated world around them.

Given the enormous place of user-based media in twenty-first century social and political lives, any shift in the logic and formal interfaces of such media ought to be of great interest to rhetoricians. Dependent largely on advertising revenue, the main activity of these media is the transformation of attention into capital. The intensive mining for attention, in turn, affects the ways in which users attend to the mediated objects and events before them. Thus as we attempt to grasp the relevance for rhetoric of a shifting logic in user-based media, it’s not a bad idea to take attention as our point of departure. Who and what are worth paying attention to at a given moment? How is attention paid? Implicated in such questions are some fundamental areas of rhetorical concern, including perceptions of reliability and trust. For instance, the relationship between epistemic siloing in social media bubbles and political polarization is well known. One’s social horizon comes to converge upon their epistemic horizon. So how does an actor evaluate the reliability and relevance of information when that social horizon has been disrupted? Similarly, the turn to recommendation media has meant that the imprimaturs of authority and celebrity have less currency than they would have had on social media. Those with a lot of followers on social media had, if not a standing reserve of trust, then an ability to steer the conversation in a certain way, to point out to others what’s worthy of attention. In recommendation media, the work of indexing relevance has been outsourced to the algorithm, a non-human actor thrust into the constellation of actors that make up a rhetorical event. No doubt a fruitful area of inquiry, then, would be a reevaluation of what it means to be a rhetorical actor in these media environments. The concern of this article, however, is even more basic: the rhetorical event itself. Given the new constraints on an actor’s ability to call attention to themselves and issues they consider important, it becomes important to understand how a moment comes to be imbued with salience (becoming, in other words, an exigent moment) so that a public may be motivated to seize the moment and act upon it.

We know that a major constraint upon capturing a market’s share of attention in a recommendation medium depends upon an actor’s ability to “please the algorithm” (Abidin & Kaye, 60). The algorithm behind TikTok, for example, is a closely guarded corporate secret; but it is at least clear that pleasing this algorithm often means imitating trending sounds, gestures, and visual compositions that are circulating at a given moment. Thus, a particular trend (often a meme) becomes an event with its own imperative to respond. This, then, poses an interesting problem for anyone looking into the structure of rhetorical situations in recommendation media, particularly the problem of exigence. How, in environments where algorithms are effective social actors, are we to understand the distinction between the responsorial imperative of a social exigence from that of a social trend? This is not a trivial or dismissive question; nor is it a radical suggestion that there is no real distinction between the two. What I am suggesting is that the two may be linked. Specifically, what the research presented in this article shows is that the practical imperative to imitate a current trend means that the demonstration of an exigence is formally constrained (topoi, syntax, etc) by the range of repeating forms at a given moment.

Although the arguments in this article are situated within the wider context of a shift from social media to recommendation media, I have chosen to draw my examples from TikTok. Although TikTok is predated by other recommendation media, such as YouTube and Vine, its widespread adoption in recent years has put it at the vanguard of the new hegemony in user-based media. Imitating TikTok’s success means not only adopting a recommendation algorithm but a “pivot to video,” specifically short video formats, such as Instagram/Facebook Reels and YouTube Shorts (Moore). These formats share several key affordances, including easy uptake of audio from other videos, which means that algorithm-pleasing behaviors are similarly oriented. Previous research on TikTok (Zulli & Zulli; Abidin & Kaye) shows that the affordances for dialogic interactions are weak relative to those in social media, which may have the effect of destabilizing the relationship between speaker and public, as well as putting limitations on the discursive and metadiscursive moves used to demonstrate exigence. Instead, socialization more often happens through various forms of mimesis, which usually involves the dramatization of moods, experiences, and events. The most ubiquitous form of imitative performance is the audio meme, in which creators overlay their own visual content with audio from existing content. Using existing audio not only signals to audiences that the performance is responsive and timely, but making use of a trending sound, again, increases the likelihood that the recommendation algorithm will make the content visible to larger audiences. Though audio meme videos reuse audio segments and replicate other aesthetic features, each memetic video is an individual communicative action with its own intention. The obvious constraint of an audio meme video is that the creator does not speak in their own words. Audio meme videos may include text on top of other visual content, but this is necessarily limited. The demonstration of exigence as well as enthymematic elements must therefore be dramatized rather than stated explicitly. The dramatization, for instance, of exigence must then be constrained by the aesthetic features already present in the meme, which includes both formal features of the original video as well as features and semantic coding that have accreted in the meme’s repetition. Such repeating features can have appreciable effects on anything from the emotional valence of an exigence to the topical and syntactic structures in its dramatization, which can, in turn, constrain the type of exigence an audio meme can take up in the first place.

I examine in this article the ways in which the formal features of audio memes constrain the communication of exigence through observations of two TikTok audio memes. The first audio meme, a response to the killing of Mahsa Amini in September 2022 by Iranian morality police, consists almost entirely of physical (gestural) imitation.² There, I observe that the meme takes on a sort of naming function, offering little or no descriptive and prescriptive information about the exigence but instead simply actualizing it for a public, as well as actualizing the public itself, through a chain of intentions. The second audio meme examined—the Mando meme—originates in a comedy bit where the uses of topos and syntax have been coopted to dramatize other exigences. My observations of the Mando meme suggest that multiple exigences can hitch a ride, so to speak, with a single audio meme. In such cases, the meme takes on what I will call a meta-exigent function, in which the intense circulation of a rhetorical object creates a responsorial imperative that runs orthogonally to the social imperative of any particular exigence associated that that rhetorical object. No matter which direction the meme-exigence relation takes, however, the ride is not free. The formal features of a meme act as constraints on the way the exigence is communicated, shaping anything from the topical structure of an issue to the emotional valences that organize a public’s relationship to the exigence.

Exigence: Ontology and Understanding

Ever since New Rhetoric approaches like Lloyd Bitzer’s theory of the rhetorical situation complicated the steady sender-receiver-message triangle, much of the debate about the rhetorical situation has been about where (or in whom) an exigence originates. Positions in that debate lie on a spectrum between realist and anti-realist ontologies. The arch realist is Bitzer himself. He explicitly opposes “real situations” to “sophistic ones,” claiming that the former are “objective” and “publicly observable,” whereas the latter must be “asserted as real” (11, emphasis added). Although Bitzer is confident that “rhetoric is a mode of altering reality […] by the creation of discourse which changes reality through the mediation of thought and action,” an exigence is part of that reality; it is—with greater or lesser clarity—perceptible to both speaker and audience prior to its mediation in discourse (4). Early on, Richard E. Vatz staked out the anti-realist position, arguing that the reality of an exigence is a matter of selection rather than reflection. Vatz, for instance, responds to Bitzer’s characterization of the Kennedy assassination as an exigence by asserting that the reports of the shooting, suffused as they were with “salience,” were what subsequent rhetorical actions responded to rather than the shooting itself (160). Later entries into the debate maintained an anti-realist position while calling into question the notion of stable, autonomous subjectivities that had been assumed on both sides of the issue (Smith & Lybarger; Biesecker; Phelps), with Barbara Biesecker observing, “Either speaker or situation is posited as logically and temporally prior, one or the other is taken as origin” (114). Instead of existing between stable identities of speaker and audience, the rhetorical situation, in this view, is a site in which those identities are articulated rather than merely being merely influenced by the mediation of an exigence.

But there is also the epistemological question. Deciding whether an exigence exists prior to speaker and public or if it emerges in interaction is different (though not unrelated) to the question of how rhetorical actors collectively achieve categorical knowledge of what kind of exigence exists. The epistemological question was of particular interest to genre theorists. Side-stepping the ontological question, Kathleen Hall Jamieson argues that the understanding of “an unprecedented rhetorical situation grows not merely from the situation but also from antecedent rhetorical forms,” (163). Thus, what Vatz calls “salience”—the perspicuity and relevance of an event—is neither evident in the event nor does it originate solely in the rhetorical choices of individual actors. “Exigence,” explains Carolyn R. Miller, “is a particular set of social patterns and expectations that provides a socially objectified motive for addressing danger, ignorance, separateness” (“Genre as Social Action” 158). “Social patterns” and “socially objectified” are the operative terms here. The singular event can only speak to us as a recurrence. Miller’s contribution to the ontological question is to contend that material events are singular; individuals may perceive a material event, but in order for an event to become a motive for social action, it must exist at the level of meaning. And in order to exist at the level of meaning, it must be recurrent; it must be understood as a type. This is where genres play an important epistemic role because they too are recurrent phenomena, and taking up a particular genre in response to an exigence creates an association with previous exigences with which the genre was taken up, fostering a social understanding of type, or definition, of the present exigence.

The potential weakness of grounding the social epistemology of exigence in recurrence is that we are in danger of seeing exigences as a succession of discrete events that are nevertheless known by types. And, as Jenny (Edbauer) Rice is keen to say, “The elements of rhetorical situation simply bleed” (9). This pronouncement comes as part of Rice’s reframing of the rhetorical situation in which the situation is no longer seen as a discrete place in time but rather a distributed phenomenon, “a mixture of processes and encounters” that “operate within a viral economy” (13). As an ontological argument, Rice’s is a realist position, though a new materialist one that would be unrecognizable to Bitzer’s own realism. Rice’s picture of the rhetorical situation fits comfortably with a wave of reconsiderations of rhetorical time emerging in 21st scholarship (Trapani & Maldonado; Graham; Gries; Rickert; Syverson). Inspired by Bergsonian and Deleuzean thought, these newer approaches replace the intuition of time as an extensional container of discrete segments with a sense of events as intensities. The experienced speed of an event is heavily influenced by the circulations of textual and visual objects, and the relative intensity of such circulations can imbue a given exigence with a sense of reality and urgency to audiences. But as Rice points out, the circulation of a rhetorical object can operate orthogonally to the exigence with which it was originally associated, thus taking on a meta-exigent function. She illustrates this phenomenon with the Keep Austin Weird campaign that began in the early 2000s. Originally created as a sticker campaign by local businesses to protest tax-breaks being given to big box stores and, in general, to counter corporate homogenization of the city, Keep Austin Weird was soon coopted by other actors, for other audiences and purposes – these included, of all things, the Cingular Wireless corporation. The fact that a rhetorical action such as the Keep Austin Weird campaign can serve as a response to multiple, simultaneous, sometimes conflicting exigences, suggests that the exigence itself (cf. exigence type in genre theory) does not alone determine the form of the action that is taken up as a response.

In Rice’s example, the circulation of Keep Austin Weird itself became an event that encouraged rhetorical action, demonstrating that even the urgency to act can emerge, at least in part, from an object saturating a rhetorical ecosystem, without needing to wait on an individual rhetorical actor to imbue it with salience. Though it is nowhere named as such, it is telling that Rice’s main example for her distributed model of the rhetorical situation—Keep Austin Weird—is essentially a meme, an idea or communicative action that repeats itself through imitation across various actors, media, and events. As we see in genre theory, it is recurrence that allows for a social understanding of what kind of situation an action is responding to. But as a repeating action like a meme drags itself across a topological web of events, it does not do so with perfect fidelity. It’s a sticky path. Clinging to it are the intentions of actors in those events; recorded onto it are formal alterations, impressions of the material contexts of its mediation. Instead of analytical correspondence with an exigence or exigence type, a circulating meme creates a social understanding of an exigence by analogical association with previous iterations through the recognition of emotions, aesthetic qualities, and even implicit doxai (e.g., “weirdness is better than conformity”). The imperfect fidelity of a meme lends it a degree of semantic openness that allows it to contribute to a social understanding of multiple exigences.

Objectifying Exigence

Although I hold with Biesecker that the binary Bitzer makes between “objective” and “sophistic” exigences is unnecessary, I still think that “objective” is a useful term here. We need not subscribe to a strong form of objectivity in which there is genuine inter-subjective knowledge about a thing; we can instead take “objectivity” to mean an epistemic and affective relationship to an object (in this case, an exigence) that tends toward inter-subjective understanding. That is to say that there is a confluence of understanding—if not agreement—about the category of exigence that exists (health, economy, security, etc.) and that the exigence has perceptible affective qualities, such as urgency, relative speed, and other valences. In simpler terms, we can say that an objective exigence is one that is becoming publicly observable, tending toward consensus on its salience. Granting this weak form of objectivity, it is possible to describe the attempt by an actor or actors to make an exigence publicly observable as a rhetorical act: to objectify exigence. One of the things that the research in this article will make clear, however, is that, contra Vatz, qualities like salience are not wholly supplied by individual rhetorical actors. The saturation of a rhetorical ecosystem, for instance, by a meme can have a greater effect on perceptions of urgency and speed than the work of any individual actor.

Following Biesecker’s argument that exigence and public co-emerge, we can say that exigence objectification (a confluence of categorical understanding and perception of affective qualities) is achieved through rhetorical identification. That is to say that any such confluence of understanding and perception emerges alongside a sense of “acting together” (Burke, 21). Identifiable and durable communities can be sustained by typified ways of “acting together,” such as genres, which prescribe rhetorical roles and even ways of responding to an exigence (Bazerman; Orlikowski & Yates; Devitt; Freadman). There are identifiable communities on recommendation medium like TikTok, such as “2SLBGTQ+ TikTok” (Hiebert & Kortes-Miller), that transcend any particular hashtag identification, but it is not clear how durable and immediate those networks are in comparison to those in social media; and it is therefore unclear how immediately observable any given exigence would be for such a public. Alexa Hiebert and Kathy Kortes-Miller, for instance, observe that while the community feel is real for 2SLBGTQ+ members, much of the direct interaction takes place on comment streams where commenters sometimes invite the video creator to communicate on a different platform altogether. This suggests that the affordances for dialogue between two or more interlocutors are comparatively limited on the app. Otherwise, identity formation in the community comes from being able to relate to or sympathize with the experiences that creators share (Hiebert & Kortes-Miller). It is fair to say, then, that community formation and the information sharing that comes with it relies upon, to an even greater extent than with social media, shared aesthetic features such as narrative style. Dianne Zulli and James Zulli argue that instead of networks based upon “interpersonal connections,” TikTok, by cultural and structural means, positions “mimesis as the basis of sociality on the site” (1873). The limited means of interpersonal communication encourages users to participate in a community by mirroring behaviors and imitating aesthetic features. As a form of sociality, mimesis need not be considered the impoverished cousin of interpersonal dialogue. Indeed, as Kenneth Burke maintains, the “aesthetic congruity” in imitation creates “consubstantiality by community of ways (‘identification’)” (131). Furthermore, mimesis is not a flat phenomenon. Just as there are identifiable registers in interpersonal communication, there are different ways of imitating that correspond to constraints such as the material affordances of a medium, cultural norms, as well as individual and community intentions. In the following section, I will discuss both the functions of mimesis, as well as the modes of its enactment within the specific medium of TikTok.

Functions and Modes of Mimesis

Burke counts imitation as “an essentially dramatistic concept,” and indeed drama is a good starting place for talking about mimesis in recommendation media (131). Drama is where we find the first function of mimesis, which is to bring an audience in closer proximity to the situation being dramatized. We can begin with Plato’s negative definition of mimesis in which he contrasts it to diegesis. Diegesis, Socrates tells Adeimantus, is “simple narration” that “avoids concealment,” where the speaker maximizes control over a text’s meaning (392d-393c).³ The specific reference here is to concealment in a character (e.g., Homer speaking through Chryses), but Plato later uses mimesis in reference to the imitation of objects and emotions through different media. Such mimetic elements may include mediated effects such as background music or filtered lighting. These effects imitate the mood of an action or event, all of which are more immediate to the audience because they are not what the audience is consciously attending to, but rather what is shaping the attentional experience itself.⁴ Again the first function of mimesis is one of proximity. In everyday interactions, we mirror behaviors to create identification and emotional proximity. Speaking through a character does create distance in interpersonal communication; however, speaking through a character also brings the audience into closer proximity with the events dramatized. The imitation of mood through the convergence of mediated effects brings the audience closer to the speeds and emotional valences of the dramatized events.

The second function of mimesis is generative. It is what Walter Benjamin calls “the mimetic force,” which, counterintuitively, is a force of change, of innovation (65). The mimetic object is a reproduction not of the object itself but of the perception of an object which is bound to a singular moment: “It slips past, can possibly be regained, but cannot really be held fast … It offers itself to the eye as fleeting and transitory as a constellation of stars” (66). What is perceived in that moment is the essence of the thing to be reproduced or projected upon our understanding of another object or context. Since this perception of essence is impressionistic—existing within a singular context—the reproduction of it in another context will necessarily create something new. This is why Aristotle considered poetics a wholly mimetic art. It might also be a reason why Aristotle severed rhetoric from poetics—not because the sophistic practice of imitating famous speeches produced unoriginality but because it was unruly in the novelty it produced. It is in this second, generative function that mimetics overlaps with memetics, in which a repeating object like a meme undergoes formal and semantic changes enabling it to adapt to and be meaningful in newer and diverse contexts.

While the functions of mimesis may be considered universal, the modes by which it is actualized are dependent upon their mediation. Diana Zulli and David James Zulli observe that TikTok is mimetic on two distinct but related levels: digital affordance and sociality. They point out that even “the set-up process and default page prompt users to engage with content conducive for imitation and for the purpose of imitation” (1877). In contrast with social media platforms, which prompt the user to begin building a network of friends in the set-up process, TikTok prompts new users to select genres whose content appeals to them. Thus, on the consumption side, users are already oriented to view content within the context of repeating types. Imitation is also encouraged on the production side. In addition to being offered a menu of visual effects templates, the three major tools of imitative production are stitching, duetting, and the Original Sound function. Briefly, stitching is when a creator splices another video onto their own; duetting is when a creator places their own content on top of another video so that the two run simultaneously; and the Original Sound function allows a creator to take the sound from another video (music, ambient noise, speech, etc.) and pair it with original visual content. All of this is possible on an older recommendation medium like YouTube, but TikTok provides these tools as part of the production template. This cultivates what Kaye et al. describe as circumscribed creativity, which “refers to creative potential being shaped or guided by platformization” that serves to push “passive users to become active users with relative ease compared to other content creation platforms” (246). In the mode of circumscribed creativity, phenomena like audio memes become templates for individual communicative actions just as much as they are sites of vernacularization for pop culture.

Again, direct social engagement between TikTok users is comparatively limited, and so social engagement often takes the form of imitative actions. One of the major pathways for this, as Zulli and Zulli observe, is through challenges where a creator solicits certain kinds of actions from other creators. Challenges often take on similar and even templatized forms, such as “Tell me you’re an X without telling me you’re an X.” Zulli and Zulli further divide types of imitative actions on TikTok into three modes, including “physical imitation – copying dance moves – reactive imitation – capitalizing and expanding on someone else’s video – and narrative imitation – describing the same types of experiences” (1881, emphasis added). Physical and narrative imitation will be most interesting for the memes analyzed in this article. We may expand the category of physical imitation to include not just the copying of dance moves but also the imitation of certain intentional gestures, such as the one examined in the Woman, Life, Freedom meme in this article. The category of narrative imitation also needs to be expanded. There are, of course, plenty of videos where the creator—whether prompted by a challenge or not—narrates a type of personal experience in a straightforward and diegetic way, but the other major mode of narrative imitation on TikTok is mimetic in its very form. These are the videos in which the creator acts out a type of experience as if the event is ongoing or existing in repetition, drawing the viewer into a sense of temporal and emotional immediacy. This mimetic type of narrative imitation is not merely an offshoot of diegetic narrative imitation but is indeed built into the DNA of the app.

TikTok originated as a lip-synching platform called Musical.ly (Kaye, Xu & Jing). The Original Sound function is a legacy of its lip-synching origins, and lip-synching videos remain ubiquitous. One of the oldest and most prominent genres on the still young app is the POV genre, which grew out of lip-synching videos. Specifically, it began with creators lip-synching the chorus of the Taylor Swift song, “You need to calm down,” which is “You need to calm down/You’re being too loud” (Haskins, “This Meme”). There, creators would act out recurring situations in their lives, addressing an imagined interlocutor, where lip-synching the song’s chorus would serve as either a resolution or a punchline. Eventually, videos in which a recurrent situation is acted out went beyond lip-synching to the song, and those POV videos became even more elaborate as the fifteen-second video limit expanded to sixty seconds.⁵ As Noah Roderick notes, even the “indicative mood” of POV video titles suggests a present but recurrent type of situation (407). One example comes from user nicoleciravolo whose POV video is titled “you get sent to the principles office for dress code but the secretary does! not! f*ck! around!” In it, the creator acts as a high school secretary and the viewer-as-interlocutor is a young woman whose male teacher has sent her to the principal’s office because the tank top she is wearing was supposedly distracting the boys in the class. It is clear that the video is making salient the pressing problem of young women being sexualized by restrictive school dress codes. No doubt there is a deliberative claim implicit in the video (i.e., school dress codes should be changed), but taken on its own terms, the video is no more or no less than a dramatization of an ongoing event-as-problem. It acts rhetorically as an objectification of exigence. Furthermore, this conceit of placing the viewer as an imaginary interlocutor inside the dramatization of a situation type creates the conditions of temporal and emotional immediacy that would not be present in a diegetic narrative of a single experience.

Audio Memes and the Moment

We can also talk about effects of what Abidin and Kaye call “the aural turn in memes” in terms of rhetorical time (61). First, it is worth reviewing the relationship between exigence and visual memes in social media. Memes in general, as Ryan Milner argues, “function as a ‘media lingua franca,’ where individuals can express themselves in an understood vernacular” (2). “Getting the ‘nationwide inside joke’ of memes,” he adds, “means adhering to broadly accepted aesthetic practices and touching upon resonant cultural moments” (2). There is thus an element of kairos to the use of memes. In other words, effective expression through a meme is not just about understanding the range of contexts in which the meme makes sense but about understanding which meme captures the cultural mood and social exigences of the moment.

Even so, I would argue that visual memes in social media (be they single images, panels or GIFs) can have relatively longer lifespans, capable of being called upon at different moments. The circulation of the “This is fine” meme on social media serves as an illustrative example. During the first year of the COVID-19 pandemic, Sweden, led in part by its state epidemiologist Anders Tegnell, adopted a markedly different strategy than most other countries in Europe by eschewing lockdowns and downplaying the protective powers of face masks against viral transmission. Tegnell, the public face of this minimal intervention strategy, came under heavy criticism, particularly by European and North American expats on social media. A meme commonly used to criticize Tegnell was the “This is fine” meme, which was a panel image taken from the comic strip Gunshow in 2013, originally titled “The Pills are Working.” In it, a cartoon dog happily sits at a table drinking coffee while the room he is in is on fire, with the speech balloon reading “This is fine.” In time, the meme was even adapted to depict Tegnell himself in place of the dog at the table. By 2020, the “This is fine” meme had already been circulating around various social media platforms for six years (a lifetime in internet years), frequently being used to criticize political leaders for inaction in the face of different crises. A visual meme like “This is fine” thus becomes one of a stock of enthymemes that can be called upon at the opportune moment and context.

The temporal relation between audio memes and exigence in recommendation media, I argue, is much closer. In general, audio memes in recommendation media are more evanescent than visual memes in social media. This, again, has to do with the imperative to “please the algorithm.” It is never possible to say just when a trend is going to peak (although there exist social listening apps used to monitor user, hashtag, and sound trends). But in order to get maximum visibility for one’s content, it is obviously advantageous to join the trend before it has peaked. The creators who have “[internalized] recommendation logics” will continue to seek out other trends before they peak. While the selection of an audio meme for a rhetorical purpose is just as deliberate as it is for the selection of a visual meme, the creator using an audio meme will be more constrained in their choice if they wish to reach a larger audience.

But a meme’s relationship to a moment is not enough to explain the rhetorical work it is capable of doing. Indeed, Eric Jenkins and Monica Huzinec contend that Milner overestimates the power of a moment to launch a meme, arguing “the question of meme emergence demands consideration of formal cause,” including the meme’s affective intensity (e.g., surprise) against an information-saturated background (403). Abidin and Kaye hold that the audio meme is more than just a legacy feature or a unique selling point for TikTok; aurality brings with it an affective intensity that pictorial and textual experiences do not have: the “earworm” effect (59). “Earworm” of course implies a sound—a musical or verbal phrase or else an ambient noise—burrowing into one’s mind, repeating itself without any external stimuli, becoming instantly recognizable when the listener is re-exposed to it. Moreover, as Abidin and Kaye point out, it is not just the content of the sound that repeats itself in an audio meme but also an ambient mood that accrues with its repetition. Thus, the affective power of the audio meme is not exhausted by the original content of sound itself or by the intentions of its sender, but rather lies in what Thomas Rickert calls rhetorical “attunement,” which shapes “how one finds oneself embedded in a situation” (9, emphasis added). According to Rickert,

Many elements, human and nonhuman, contribute to the manifold ways affect emerges, takes shape, and finds its trajectories (affect here being understood as a worldly disposition and not a subjective emotional state). […] We cannot posit a simple cause-effect relationship wherein a sonic element directly achieves the desired result in accordance with intention (145).

The accreted mood of an audio meme can be said to emerge out of the interface of the sound as a singular object and as an object-in-iteration. The sound as a singular object possesses its own extensional (syntax, running time, etc.) and aesthetic qualities, while the object-in-iteration brings with it the intentions of those who would repeat it. It is, as Jenkins and Huzinec put it, “the shared form of the meme, across all the iterations, that allow different versions to be, and to be recognizable as, one instance of the larger meme” (403). Rickert further asserts that “attunement is nothing static,” meaning that as the ambient mood accompanying an iterative action changes, so do rhetorical intentions toward it (8). Repeated imitations of aesthetic form can, over time, become an engine of change, altering users’ understanding of the meme’s relationship to an exigence type (and thus the motives of rhetorical actors), or even allowing a meme to be taken up for different exigence types altogether.

As I will attempt to illustrate in the following sections, rhetorical actors can not only use an audio meme to call attention to an exigence, but through the accretion of mood as well as the coopting of formal elements like topoi, a meme can affect how an exigence is attended to as well. Furthermore, depending upon the intensity of repetition within a given moment, an audio meme can both effect a sense of urgency and draw the audience temporally closer to the event.

Physical Imitation: Woman, Life, Freedom

The phrase, “Woman, Life, Freedom” (WLF) comes from the Iranian Kurdish slogan, Zan, Zendegi, Azadi, with both versions frequently appearing as hashtags on social and recommendation media to express solidarity with young protesters in Iran. The protests began in September 2022, following the death of a 22-year-old Iranian Kurdish woman, Mahsa Amini, after being in police custody for supposedly violating Iran’s sumptuary laws for women. In addition to the massive and persistent protests on the streets of Iranian cities, members of the Iranian and Kurdish diasporas, along with others in solidarity with them, have protested Amini’s death in cities all over the world, with many inside and outside of Iran calling for the removal of the current regime (Dagres).

As with many social movements of the past decade, the WLF movement has effectively used social media and recommendation media to inform, network, and mobilize. In addition to text and still image content on various platforms, those involved in the movement also frequently share video clips taken from Iranian streets as well as from international protests. One such video, posted to TikTok by user jkk.lomis on 18 September 2022—just two days after Amini’s death—features a series of four stitched videos in which three women and one man, looking directly at the camera, are cutting or shaving off portions of their hair. The video includes the text, “Iranian people are cutting their hair short the killing of a 22-year-old girl Mahsa Amini by the police in Iran bc of hijab.” The video’s use of sound is mimetic: there is no sound from the inside of the video; instead the video is overlaid with the chorus from a live version Tom Odell’s 2012 song, “Another Love.” The featured lyrics from the song are:

And if somebody hurts you I wanna fight/But my hand’s been broken one too many times/So I’ll use my voice, I’ll be so fucking rude/Words they always win, but I know I’ll loose.

And I’d sing a song, that’d be just ours/But I sang ‘em all to another heart.

This quickly became the movement’s signature audio meme, with young women and men showing solidarity with the movement by creating videos of themselves cutting their hair using the same audio clip (Dagres).

Cutting one’s hair as a sign of protest draws upon a Persian mourning custom that dates back at least as far as the tenth century epic poem, The Shahnameh (Wright). It might therefore be understood, in part, as an assertion of Iranian cultural identity over the Islamic Republic of Iran’s theocratic regime. The other source of the meme is the song itself. Odell’s original 2012 single is slow, melancholic love song featuring vocals and piano. It underwent a resurgence in popularity after it was appropriated by those protesting Russia’s invasion of Ukraine in February 2022, where, isolated from the rest of the song, the theme of the chorus was recontextualized from one of personal heartbreak to one of collective defiance. Thus, when it was taken up as an audio meme for the WLF movement, it did so from an existing context.

“Another Love” is not the only song associated with this exigent moment. Live protest events are, in fact, more likely to feature demonstrators singing Shervin Hajipour’s “Baraye,” which was written for the movement, and includes tweets from young Iranians in its lyrics (Dagres). Furthermore, the use of the song for videos in the Ukrainian context is varied and might only loosely constitute and audio meme. There, several different versions of the song are used, and they often feature more than just the chorus. The visual subjects of those videos are also extremely varied, including images from military action, cities in the aftermath of destruction, street protests, as well as video creators themselves. By contrast, the use of Odell’s song for WLF is largely uniform, using the same portion of the song from the same version. Importantly, the version used for the WLF meme comes from a live performance, with the audience singing the chorus in unison rather than Odell singing it alone. Some videos using Odell’s song include images of protests, but, overwhelmingly, they feature young women looking into the camera and cutting their hair so that the latter can be considered a meme onto its own. Some examples of the meme, like the early video from jkk.lomls, include textual information in various languages, such as “I stand with the women of Iran” (mona.noroozi) or “Je me lève pour Mahsa Amini” (dana_leclerc), while many others have no text at all.

Using Zulli and Zulli’s categories of imitation, this meme would be classified as physical imitation, since the one visual constant is the imitation of gesture. As a physical imitation meme, it distinguishes itself from earlier Ukraine protest videos in aesthetic effect as well as in type of content depicted. First, while the use of “Another Love” serves the proximity function of mimesis in both contexts, the Ukraine videos that show things like military actions and civilians next to the shelled remnants of their homes also convey quite a bit of diegetic information. Such depictions are individual pieces of a narrative, and the mimetic element of music serves as an affective thread similar to a leitmotif in film. Those videos are not only multimodal but the interplay of diegesis and mimesis makes them multidimensional. The WLF meme, on the other hand, reveals very little about the exigent moment of the Iranian protests. Instead of informing and persuading, the WLF meme functions as premediation, which is the affective structuring of “events that are about to happen or are in the process of happening,” occurring prior to the formation of more durable narratives of the event (Papacharissi 309). The relative intensity of premediation around an anticipated or ongoing event can affect the speed at which the event is perceived as happening, thus actualizing it in its immediacy.

In its premediating role, the meme further effects something between metonymy and proper naming. On the one hand, the relationship between the meme and the exigence is not arbitrary since the mood the meme effects is meant to be a part of the felt essence of the WLF movement. Thus the meme becomes, by analogical association, a token of the movement. On the other hand, the meme acts as a referential chain similar to what Saul Kripke describes with respect to names: “When the name is ‘passed from link to link,’ the receiver of the name must, I think, intend when he learns to use it with the same reference as the man from whom he heard it” (96, emphasis added). To intended reference, the iterative chain of the meme adds affective intensity. Such chains, as Robert Hariman argues, are capable of providing “frameworks with which some events become more likely or legible or legitimate than others” (321). In this case, it is all three. The chain makes the event legible and legitimate because, while there are various hashtags associated with it on TikTok, the meme has no name for itself (I’ve been calling it the “WLF meme” for convenience). And yet it serves as a recognizable token for the exigence that it is objectifying, with a public emerging as the ones who would recognize the token. The meme also makes the event seem more likely because of the intensity of its circulation and iteration. Because it is a meme and not a heterogeneous series of videos corresponding to trending topics or hashtags, it brings with it the urgency to participate that comes with the recommendation algorithm. The urgency of the recommendation algorithm increases the meme’s iterative speed, which in turn creates a sense of immediacy with an ongoing event. In addition to the public being brought into closer temporal proximity with the event, a sense of anticipation is also created. The imitation of sound and gesture within a small window of time creates a feeling of spontaneous synchronization, as if to signal a coming phase change, or perhaps a revolution.

Narrative Imitation: Mando (Tell Me Something That’ll Piss Me Off)

As we saw in with the WLF audio meme, coupling the same portion of a particular version of a song with physical imitation allowed for the public identification of a single exigence (the struggle for civil rights in Iran). Besides the cultural context of the physical imitation (cutting or shaving hair), the emotional attunement afforded by the song itself probably also played a role in objectifying exigence. As is the case in theater and cinema, music is easily taken up for the mimetic role of telegraphing or exaggerating the emotional content of a scene. Non-musical sounds such as speech, on the other hand, do not take to this mimetic role quite so readily because speech is more often diegetic, or part of the material of the scene itself. However, many of the most popular audio memes at a given time on TikTok use speech for their sound. Some of those memes fall under Zulli and Zulli’s category of reactive imitation, where audio clips of, for instance, motivational speeches or political arguments are used with the visual subject of the video telegraphing agreement, disdain, or rapt attention. In other cases, a speech clip may perform a proximal mimetic role, such as overlaying The Barber’s (Charlie Chaplin) final speech from The Great Dictator with videos presenting images of Putin and Ukrainian resistance, the pathos of the speech echoing the emotional content of the images (visegrad24).

More commonly, speech audio gets memetized for ironic humor. That is, a speech clip gets taken out of its original context and placed into unexpected but resonant visual contexts. The source of this irony is, therefore, situational incongruity. Incongruity is already a key concept in visual rhetoric (Tseronis; Mohanty & Ratneshwar; McQuarrie & Mick). As Edward F. McQuarrie and David Glen Mick put it, incongruity as a rhetorical figure “provides the means for making the familiar strange” (“Figures of Rhetoric in Advertising Language” 426). Incongruity in images consists of one recognizable image being merged with the form of another, for example, in a March 2018 cover of The Economist depicting Donald Trump’s face as a big, orange grenade (Tseronis, 388). Here, incongruity is not only a source of humor and/or pleasure, but “it prompts cognitive processing for resolution of the perceived incongruity” (Mohanty & Ratneshwar, 444). It suggests categorical information about a visual subject without spelling out a predicate (i.e., Trump is a volatile leader). I argue that incongruity in speech-based audio memes prompts similar “cognitive processing for resolution,” but here, the cognitive processing is distributed and iterative, which means that there is not a stable coupling of subject and category to be resolved. Instead what gets resolved and conserved through imitation are formal characteristics that constrain how an exigence gets objectified by a meme, if not the kind of exigence the meme takes up in the first place.

To illustrate what the conservation of form looks like in an audio meme, I will discuss the Mando (or Tell me something that’ll piss me off) meme, whose sound went viral in summer 2022. The original video from user artimuswolz was posted to TikTok on 10 June 2022 and is just six seconds long. The creator (and visual subject) is holding camera below eyeline, tilted in such a way that it resembles a low Dutch angle.⁶ The video could be placed in the POV genre, except that instead of facing an imagined interlocutor, the subject is addressing someone (“Harry”) who is present but off camera (perhaps in the next room). The subject is telegraphing ennui. The sequence of dialogue follows as:

1. [Subject to himself]: “Ugh … I need to feel something.”

2. [Subject shouts away from camera]: “Harry, can you tell me something that’ll piss me off?”

3. [Harry]: “Boba Fett’s better than Mando!”

4. [Subject springs from his chair]: “Yup. That’ll do it!”

Here, “Mando” refers to The Mandalorian which, along with The Book of Boba Fett, are series on Disney+ that exist in the Star Wars fictional universe. The implication is that for Star Wars cognoscente, the very idea that there are people who don’t recognize The Mandalorian as the superior series is noxious, even painful. The subject’s emotional commitment to the bit is what makes it so absurdly funny. Since it is a humorous video, there is no expectation that the audience will take the Mandalorian-Boba Fett controversy as a salient issue. If there is an exigence for artimuswolz’s original video, it is that the video itself is waiting for engagement or even to be made into a meme.

Although historical data about when the Mando meme reached its height is unavailable, the video with the highest level of engagement using the sound was posted on 17 July 2022. By that time, the sound had already been used in over 90,000 videos (Cavender). Some early examples of the meme can be categorized as physical imitation, as they feature a visual subject who is merely lip synching the dialogue of the original sound while imitating the ennui of the original video. The earliest observable example of the sound being used for something besides physical imitation was posted on 16 June 2022 by user yoursupportivedad. Using a camera angle similar to the original, the visual subject lip synchs the lines 1, 2, and 4 from the original video and puts his hand to his ear during line 3. The video is overlaid with text of the original subject’s lines, but during line 3, the text reads “PEPPA PIG IS BETTER THAN BLUEY!” Here, the syntactical construction of the original is conserved under the translation of content. Other early videos in the meme stick to the “X is better than Y” structure, with X and Y being competing pop culture products. The inclusion of text would become a lasting feature of the meme, although it also became common for text to appear only during line 3.

By the height of its virality in mid to late July 2022, however, the “X is better than Y” structure had largely disappeared from the meme. In its place were objectionable statements that the viewer is led to understand are frequently heard by the video’s creator. For example, one video, titled “Its almost like I hear that a million times a day,” features a woman lip synching the audio from lines 1, 2, and 4, with the text of line 3 reading “I feel like I know you from somewhere,” obviously expressing exhaustion from a constant barrage of bad faith pickup lines (athenafarisx). It’s here, when the meme jettisons the “X is better than Y” structure, that it undergoes a shift to narrative imitation. The audience is brought into a repeating situation—almost in media res—where the line 3 speaker embodies an amalgamation of voices that, in turn, represent a “defect, an obstacle, […] a thing which is other than it should be” (Bitzer 6). Other line 3 examples include: “Everyone’s a bit Autistic,” “Bisexual isn’t inclusive of trans and nb people,” and “Man bashing is a trend.” The exigences dramatized in the Mando meme often reflect concerns about social identity (gender, race, disability, etc.) but beyond that, there is little in the way of coalescence around a particular issue. Examples can be found of line 3 texts that refer to issues contemporary with the meme, such as the Depp v. Heard defamation trial (which concluded in June 2022), but these are nowhere as common as references to more endemic social issues.

When it comes to the relationship between an audio meme and an exigence, then, we see a marked difference between a physical imitation meme like the WLF meme and a narrative imitation meme, like Mando. Mind you, these are just two examples, and it would be a mistake to overgeneralize the distinction. Nevertheless, it is clear that audio memes are capable of relating to exigence in at least two ways: 1) a 1:1 relationship between meme and exigence (WLF) and 2) a meme taken up for a variety of exigences (Mando). I would argue that in the latter case, the audio meme functions as a meta-exigence. The meta-exigence function can be usefully contrasted to the memeplex, or “an interdependent collaboration of memes” that get coded for meaning as they cluster around a single issue (Johnson, 38). The meta-exigent function is instead similar to that of a genre, in that it can be taken up for more than one exigence, provided there is a perception that the various exigences are of a type. The key difference between meta-exigence and genre is that genre uptake is driven by the eruption of an event (whether singular or perennial) in which the exigence is located. For instance, in the case of a deadly wildfire, emergency officials may take up a press conference, repeating certain rhetorical moves employed in a press conference responding to an earthquake. With the kind of meta-exigence we’re seeing here, on the other hand, the audio meme is the event. It brings with it its own duration (its moment of virality) and with it an urgency to respond. In its meta-exigent function, the audio meme supplies the crucial ingredient of kairos to rhetorical actors who wish to call attention to an endemic problem.

But meta-exigence in audio memes is no free-for-all. The formal characteristics that are conserved, and then coopted, serve to constrain the ways in which exigence is objectified. In the case of the Mando meme, we see in the original video the “X is better than Y” structure that might be identified with the Aristotelian topos of disputed goods (I. §7). That topos was conserved in early iterations but was later dropped, indicating that participants in the meme perceived the topos as an accidental rather than an essential quality. Perceived as essential were the call-and-response structure and the affective quality (or pathos topos) of ennui. The conservation of ennui in particular allowed for the meme to be exapted for other rhetorical purposes, but I would argue that the dramatization of ennui—the feigned exhaustion and dissatisfaction—largely constrained the meme to objectify problems considered to be persistent rather than immediate. Had the topos of disputed goods been perceived as essential, the kinds of exigences the meme took up, let alone the way in which they were framed, might have been quite different.

Conclusion

If it is the case that an exigence does not reveal itself as a publicly observable object prior to the intervention of rhetorical actors, then attending to an exigence must not be regarded as separate from the exigence itself. Within a durable social network, there may be a reliable range of issues that are already considered salient which a rhetorical actor can draw upon to call attention to a particular exigence. However, inside of the intense attention economies of recommendation media, the rhetorical actor must either manufacture attention or—more often—coopt it. Some share of attention can always be manufactured through brute force: volume, spectacle, and sex. When it comes to affective power, it is easy to overestimate the chiaroscuro of noise and information, the toneless pleasure of studium against the felt “sting” of punctum (Barthes, 27). But as this article has shown, there is also the earworm effect to consider, the amplification of a theme or mood in consonant repetition. Before any particular exigence is identified, a public is already attuned to the situation’s speed, emotional valence, and even logic. Circulating forms have our attention before they shape our intentions toward a problem. And since those forms are circulating, they do not belong to any particular actor, public, or moment, except to the moment of their circulation.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

¹ A social graph is a representation of a user’s network that can be used to understand the nature of the user’s connection with an individual or community (personal, professional, political, etc.) as well as which users act as major network hubs within a given realm of association (Yang et al.).

² Due to the sensitive political nature of some of the videos I discuss, I have chosen not to make a permanent record of the videos, should a user decide to delete their posts. I have therefore excluded image captures of videos and the real names of users (even where available), referencing only user handles and currently available URLs.

³ Given the Greek term for “truth” as aletheia, or unconcealment, it becomes pretty clear where Plato stands on the morality of mimesis.

⁴ David Saltz names as an example the “deep red glow [pulsating] over the virtual landscape” as Caliban “writhed in pain” in an intermedial production of The Tempest (125).

⁵ At the time of writing, the time limit for TikTok videos is at ten minutes.

⁶ Without additional context, the low Dutch angle effect in the video may seem accidental. A look at the corpus of artimuswolz’s posts, however, reveals a clear preference for the technique. The Dutch angle, together with the content of his videos, contributes to an online comedic persona of someone who is constantly on the edge of anger or extreme disappointment.