Probably the first Lolcat

Anyone with a cursory familiarity with the internet will recognise the archetypal meme instantly: an image that can be humorous, but isn’t always, and that’s overlaid with header and/or footer text. It might be an image of cats, dog(e)s, frogs, or other animals; celebrities or pop culture scenes; hi def or lo brow photo-shopped or MS Paint graphics that are real, drawn, or computer generated; white Impact font in ALLCAPS or primary-colour clouds of Comic Sans text; or, at the limit of what we might consider an image, the looped movement of the GIF. This content is more or less schlocky and popular. In broader terms, the word has been extended to other forms of digital media, including easily re-shared (often inspirational) phrases; short videos, like vines; or even embeddable audio clips and sound-boards. The meme has sloughed off its neo-Darwinist associations and is now taken to be internet culture’s minimal, circulating unit.



The meme is not just an image, but it is usually consumed, produced, disseminated and manipulated through graphic user interfaces and distributed platforms. The interface is the non-visual meme’s visual frame. Perhaps, like Hito Steyerl’s “poor images”, the meme moves like an “errant idea” (Steyerl, 2012: 32), tracing the path of that which goes astray. Steyerl also says of the poor image that, “[i]n short: it is about reality” (44). The meme is internet schlock, but its errant movements also expose us to the massively distributed networks that are less a virtuality than an other elemental milieu. Like the more strictly defined internet images that Maria Olson describes, memes “circulate in excess” (Olson, 2008: 280). This excess, the internet’s, is what the meme indexes. Not only because it exploits network infrastructure to make internet culture move, but because the meme is, in a very basic way, in excess of itself.



The meme is always more than a particular instance because it has a double character. To function as a meme, media must express an organising syntax. The meme-instance is also, always, a member of a meme-series. But this quality makes it this almost-contradictory excessive unit hard to theorise. The meme resists reduction to its putative “content”, troubling typical – discrete, additive – understandings of cultural objects.


A definition of the meme like the one Limor Shifman offers is indicative. Shifman’s definition identifies three main characteristics: it encapsulates “a group of digital items sharing common characteristics of content, form, and/or stance”; it states that these items “were created with awareness of each other”; and it stipulates that these items “were circulated, imitated, and/or transformed via the internet by many users”(Shifman, 2013: 41). Or: the meme is serial; produced collectively; and it circulates. Whilst Shifman’s definition is useful, it is also limited.


Probably the earliest Doge meme

Take the Doge meme, one of the most popular (Anglophone) memes of the past few years. The popular Know Your Meme (KYM) platform, a user-generated meme encyclopaedia, constructs a taxonomy of the Doge meme by identifying its common characteristics. These include the ur-content of the now-iconic Shiba Inus breed of dog and the Comic Sans MS font overlay in basic, bright colours; the twisted misuse of the formal grammatical modifiers, much, many, so, very and such (as in typical phrases like “very Doge, much cool”); and the related ironic “contemplative” stance that, it is argued, these features engender. KYM historicises these features by relating them to key platforms, like the “Shiba Confessions” Tumblr, or media, like the seminal clip from the early comic web series, Homestar Runner, that is credited for spelling out and so initiating writing/saying D-O-G-E.


How are we to use these materials to understand what a meme might be? Taken alongside Shifman’s framework, the data that KYM archives can be stitched into a history by enumerating instances and bringing them in to a more or less adequate genealogy. The second-person plural past tense Shifman uses – memes “were” created, “were” shared – defines the meme as a nominalistic collection of instances with a trackable provenance. This approach allows us to understand the meme as an accretion of instances that merge, like tributaries, into its contemporary instantiation. But this reduction to the collection of instances elides the meme’s collective, excessive character, because the “group” it describes is not the same as the “series”.


This approach fails to escape a basic theoretical aporia confronting any theory of the meme: that “Doge” designates both a Doge, an instance, and the Doge, the series. The series can neither be reduced to nor transcend the collection of instances that comprises it. Or: it is in excess of its apprehensible content. This claim can be clarified by Georges Bataille’s distinction between a “restrictive” and a “general” economy. In Shifman’s model, making memes is the prerogative of participatory users and amounts to a kind of collective activity. It presents what Bataille calls a “conventional set of human activities”, or a restricted, enumeratable group (Bataille, 1988: 68). In the face of this model we might ask, what’s the return on a meme? Can the meme be limited to the sum of its constituents, or does the meme put something more into play? How does a meme, like the Smaug meme below, emerge out of, mutate, or deconstruct the Doge?


Smaug-Doge (Smoge?)


The tendency to reduce the meme-series to its content is symptomatic of a broader problem in media studies. A meme is only a meme, or a meme-series, when it enters circulation. Yet media studies so often fails to account for what the mechanism that underpins distributed cultural production, the circulation of media, actually is. When the meme ceases to circulate, it ceases to exist. The “irreducible movement” that Bataille ascribes to his “general economy” (68) is integral to both its mediation and its basic continuation. To understand why the meme is in excess of its content, we need to ask what it means for a meme to return.




Circulation could be described as fundamental to large-scale media, but it’s not very well theorised in media studies. Our usual understanding of circulation is problematic, because it’s tautological. The etymology of circulation, which denotes or has denoted something spreading or moving through something else, is inscribed with its own media theory. Circulation is usually understood as the circulation of content through a medium. Sianne Ngai, for instance, defines circulation as “the technologically-mediated movement and dissemination” of things like “information, discourse and commodities” (Ngai, 2012: f.n.1, p. 246). For memetic media to be reduced to content, then, they need a carrier: the networks that content passes through. But this division between content and network, or matter and form, generates a problem.


The problem plays out like this. When we ask the question, “what is circulated”, we get the following answer: content. But if we follow this question with another, “what is circulation?”, we get a tautological answer: circulation is the circulation of content. If content is already understood as, by its very nature, having already been put in circulation by its technological substrate, the circular nature of our conception of circulation becomes clear. Content circulates; technology circulates content. The concept of the circulation of content is without content.


I’m just going to leave this here


“Content” is the death mask of circulation. Content is content because content is content, which is to say because it isn’t anything in particular. If the meme is analogous to content, it has no utility as a concept. It would signify a contradictory impulse: to understand the meme as a unit that is both informational, and so seemingly free to circulate; but also phenomenal, and so understood through and for us. The impulse to reduce the meme to its instances does violence to the concept because it assumes that the meme’s singular, phenomenal form is primary. It elides the meme’s irreducible, nonhuman movement because it assumes that the economy of the meme produces returns that can be cashed in by an aggregate of human actors.


That’s probably not realistic, Keanu.


We can apprehend this movement’s velocity when we ask, instead, what it means for a meme to return. Karl Marx’s concept of circulation helps us to frame this question. Marx argues that the circulation of capital must be “grasped as a movement, and not as a static thing” (Marx, 1993: 185). Capital’s movements intersect with, but are not equivalent to, internet culture’s. But, we can nevertheless begin by focusing on circulation understood in excess of its infrastructural carriers. Circulation is the process through which something returns – but not necessarily to us. Circulation forms a circle, “autonomising” a process that is abstracted from its contents (Marx, 185). Sheared from the dynamics of capitalist value creation, this process – Verselbstständigung, achieving an independent existence – autonomises the meme as a processual series. The meme that enters circulation returns to itself.


If the meme is neither content nor instance, what kind of media is it? The concept itself exceeds the mere image, encapsulating GIFs, vines, bits of text and short audio clips. Memetic mediation is irreducible to the instance, but it shouldn’t be confused with its network infrastructure. The meme in excess of itself reorders what we think of as media, because its mode of mediation exploits the processual, constantly mutating distribution of the network itself – and makes it available for manipulation.


(In simple terms)



Sybille Krämer argues that “”mediation” must be understood as radically extrinsic, as the back-and-forth of circulation itself” (Krämer, 2015: 54). What we would otherwise call the meme’s “content”, its intrinsic qualities, are secondary to its extrinsic quality, that it has entered circulation. The instance or its content is a retroactive epiphenomenon of the process of the meme returning to itself. Steyerl says of the poor image that it “is no longer about the real thing – the originary original” (Steyerl, 2012: 44). The Doge’s genealogy is secondary. What it’s about are “its own real conditions of existence: [] swarm circulation, digital dispersion, fractured and flexible temporalities” (44). To return to what I said she said before, “in short: it is about reality” (44).


The Doge meme has also morphed into a Cyrptocurrency with a real world value, mixing alternate modes of circulation – and the modes of reality they produce.


The meme is a kind of media, only not the kind we’re used to dealing with. Bernhard Siegert has critiqued media studies’ tendency to reduce media to usually screen-based devices and interfaces (Siegert, 2015: 87) – to that which is discrete. But the meme’s circulation mediates, and can be used to manipulate, a specific relationship to reality. The meme can only return to itself, or exceed itself, because it can collect a certain kind of return made possible by another kind of abstraction.


What we call the internet is not an internet, but a series of networks governed by protocols, technologies and distance (Dourish, 2015). These networks trace a very terrestrial material infrastructure: the letters labelling Google’s search interface transmitted to us, along fibre-optic cable routes first laid for telegraphic communication (Easterling, 2014), from massively-polluting data centres whose cooling pipes echo this logo’s hues. The internet is not an abstraction or a frictionless space through which data flows, because we can point to where it quite literally is. But circulation nevertheless generates a kind of abstraction that subsists in the joints of its massive, global distribution.



Benjamin Bratton’s fiction of “the black stack”, which describes this global distribution, has a certain kind of utility for thinking memetic mediation. Global data movements might not flow freely – or at all (Sutherland, 2013) – but they are in excess of our ability to either comprehend or to calculate them meaningfully. As John Durham Peters quips, “ontology, whatever else it is, is usually just forgotten infrastructure” (Peters, 2015: 35). The internet is not reducible to a re-presenting network map. Rather, this kind of mapping is a “cultural technique” that, given Siegert’s argument that ontology emerges out of “concrete ontic operations” (Siegert, 2015: 87; 2013), provides us with one way of dividing and mediating the internet’s distribution. The production and circulation of memes is another kind of technique.


If the meme has any utility as a concept that is non-medium-specific, it is because it emerges from a cultural technique that is specific to a massively distributed network real: the meme that returns, in excess of itself, is first made as a kind of offering.



The fiction of the stack or of massive distribution has a limited kind of phenomenological truth: the internet is in excess of us. The eventual heat-death of our solar system undoes any empirical efficacy that Bataille’s concept of a “solar economy” might have had as energetic apeiron. But his idea of squander is still useful for thinking the excessive dynamics of memetic circulation. Making memes is a kind of expiation, a kind of offering made to what we construe as excess itself. Any return that we might get from this meme is secondary to the originary return it must first make to itself.


The limits of the solar real


Memes aren’t for us, but they participate in what Alexander R. Galloway calls the “furious” network real (Galloway, 2013). Memes are non-media-specific because the basic technique that presupposes them is the mediatic mode of making an offering to the network. This is, in the end, why the meme’s content is epiphenomenal. The “content” of the meme, or what we consume, is retroactively condensed as a set of associated, and, contingent techniques: the Shiba Inu, Comic Sans, etcetera. It’s real “content” that we consume is not, originally, what the meme contains, but that it is able to circulate.


How a meme doesn’t work


This expiatory technique is processual and recursive: we – where this term is understood, always, in terms that are singular and collective (see Stiegler, 2009) – try to manipulate the abstract reality of massive distribution by making memes. Olson has described internet images as “the vertebrae of a body that we otherwise seek to theorise as amorphous” (Olson, 2008: 281). Memetic mediation gives the network a furious, mutating, threatening spine.


But if this expiatory technique traces a mode of existence, it is one in which what is shared is only shared indirectly and may never return to us, as either profit or as experience. By exploiting the furious reality of distributed networks, this technique traces the outlines of alien processes that we can only ever access obliquely. The mode of existence of technical beings (see, i.e., Simondon 1980) introduces scale into experience. What is huge may be entirely indifferent to what isn’t. Though our sun radiates huge amounts of energy, it will one day die. We rely on it, but it is radically indifferent to us.


Like magnets, the fractal notion of scalability is rather mystical


Or: we may make our offerings to the internet, but the internet will remain unknowable in its entirety. Contra Bruno Latour, networks are not necessarily available at all of their points. This is an abstraction, but it’s also real. The relationship between the act of expiation and subjectivity is oblique at best. Its effect on us is less like the sun’s excessive rays and more like an enveloping atmosphere, a milieu that can only partially be perceived (Peters, 2015: 46). What does it mean for experience if the act of giving to an abstraction is not only not reciprocated, but that the non-reciprocity of what is massively larger than us can be exploited as a technique? The meme errs because it doesn’t rely on us. The furious “non-human” swarms.




Bataille, Georges. The Accursed Share: An Essay on General Economy. New York: Zones Books, 1988.

Bratton, Benjamin. “The Black Stack”. E-Flux No. 53 (2013). Online: http://www.e-flux.com/journal/the-black-stack/

Dourish, Paul. “Protocols, Packets and Proximity: The Materiality of Internet Routing. In Signal Traffic: Critical Studies of Infrastructure, eds. Lisa Parks and Nicole Starosielski. Urbana: University of Illinois Press, 2015.

Easterling, Keller. Extrastatecraft: The Power of Infrastructure Space. London: Verso, 2014.

Galloway. “Love of the Middle”, in Galloway, Thacker and Wark, eds. Excommunication: Three Inquiries into Media and Mediation. Chicago: Chicago University Press, 2014. 25-75.

Krämer, Sybille. Medium, Messenger, Transmission: An Approach to Media Philosophy. Trans. Anthony Enns. Amsterdam: Amsterdam University Press, 2015.

Marx, Karl. Capital Vol 2: A Critique of Political Economy. Trans. David Fernbach. London: Penguin, 1993.

Ngai, Sianne. Our Aesthetic Categories: Zany, Cute, Interesting. Harvard: Harvard University Press, 2012.

Olson, Maria. “Lost Not Found: The Circulation of Images in Digital Culture”. In Words Without Pictures, 18 September 2008, online: http://wordswithoutpictures.org/main.html?id=276

Peters, John Durham. The Marvellous Clouds: Toward a Philosophy of Elemental Media. Chicago: Chicago University Press, 2015.

Shifman, Limor. Memes in Digital Culture. Mass.: MIT Press, 2013.

Siegert, Bernhard. “Cultural Techniques: Or the End of the Intellectual Postwar Era in Germany”. Online: Theory, Culture & Society, 2013. 1-18.

—         “Media After Media”. In Media After Kittler, eds. Ileni Ikoniadou and Scott Wilson. London: Rowman & Littlefield, 2014.

Stiegler, Bernard. Acting out. Trans. David Barrison, Daniel Ross and Patrick Crogan. Stanford: Stanford University Press, 2009.

Steyerl, Hito. “In Defense of the Poor Image”. In The Wretched of the Screen. Berlin: Steinberg Press, 2012. 31-44.