Part 3/7 of the essay series “Slang of Content Nerds to Art and Back“
Deniability has no political direction. The same ironic frame that protected the dissident normalizes prejudice regardless of the flag it is flown under. And the cost of the joke is often carried by someone quite different from its target. This part of the series attempts to distinguish constructive transgression from corrosive transgression, and proposes four tests for telling them apart. None of them asks about the author’s intentions; they examine what ideology and attitude a meme actually carries, spreads and normalizes.
I. The symmetry problem that normalises hate speech
The earlier essays in this series have all, in one way or another, defended deniability. Dada’s negation needed room to move without final accountability; dissident art used double-coding as a shield for the weaker party against totalitarian power; the remix techniques of meme culture deliberately build ever-narrowing circles of initiates. In all of these, deniability has been something defended through the position of its user: the censored, the underdog, the creator working at the margins. And that is easy to agree with.
But deniability has no political direction of its own. It is a technique, not a stance. That is precisely why the field also contains phenomena that force us to set sympathy aside for a moment and examine the whole machinery from a more critical angle.
It is well known that extremist movements, fascist and other reactionary accelerationist circles in particular, have used “ironic” transgression systematically to shift the Overton window. The method has countless variations, but it can work, for example, like this: First, something extreme is said, seasoned with humour. When people don’t laugh, it is explained to have been a joke, and the listeners may even be needled for lacking a sense of humour. Once enough people have heard the claim enough times within the frame of “a joke,” the claim has become familiar. And familiarity resembles acceptability.
This mechanism was documented precisely in the mid-2010s research on online radicalisation. “Ironic” racism or misogyny works as a gateway in which the frame of humour enables gradual exposure without anyone ever having to admit believing in any extremist claim. The audience merely accepts that in certain circles “people joke about somewhat rough things” or favour “dark humour,” and out of the repeated experience a quiet, “boys will be boys” style of acceptance slowly develops. That is socially and intellectually far easier than setting about to analyse and condemn every single “joke” with a furrowed brow. A sense of humour is socially valuable currency. Few want to be branded a killjoy.
The only problem is that every single instance of presenting such a questionable claim habituates the audience to regarding extremist statements — and sometimes outright hate speech — as somehow a normal part of public conversation. The mechanism is pure statistics. The statistical link between repetition and acceptance has been demonstrated over and over.
So how should we regard it when the same tactic is used by someone who feels they are on the side of the good? Take, for example, a “woke”-leaning meme sharer who decides to share a meme produced by fascist accelerationists “ironically,” or perhaps to point out what the other side of the fence is up to.
Is the outcome structurally different? How much does the context of publication matter? Certainly the sharer’s inner circle knows their intentions. But does even that really matter? The meme is shown to people who cannot be as sure of the sharer’s intentions, unless the motive is explained. And explaining is exactly what the makers and sharers of memes hate. The outcome is that a questionable meme was shown to a large number of people, and every one of them was exposed to the questionable claim one more time.
The mechanism itself contains no value-assessing or regulating system. It is only a mechanism. So should this “woke” meme sharer be granted a moral pardon, because they feel they are “on the side of the good” while merely “playing around with slightly rougher stuff” under the cover of humour and exposing the other side?
I think not. Showing a racist or misogynistic claim in public is an act of normalising it, whatever the motive and whatever the sharer’s intention.
II. The Overton window doesn’t check your political views
Shifting the Overton window is a process, not a position. The window moves according to which claims are repeated in public space often enough and within an acceptable enough frame — not according to who presents the claim or with what justification. This is the core of the window’s entire mechanics: it is a statistical, not a moral, process. When a stereotype — whether its target is an ethnic group, a sexual minority or a linguistic minority — circulates in public space within the frame of humour, its circulation normalises the stereotype itself, regardless of the political purpose for which it was originally set in motion.
Here the disintegration of interpretive keys, discussed in the previous part of the series, becomes decisive in a new way. When an in-group joke circulates widely enough, its original key, “this is irony, this is a parody of power, this is aimed upward”, does not travel with the image. What arrives is only the image and the content within it, without explanation, without context, and above all without the sharer’s intentions.
We use a real case as our example here. It is a political caricature of the Finnish prime minister in which he is mocked by associating him with being Swedish-speaking and with homosexuality. So we are operating close to the lowest level of exploiting prejudice. And for the reasons set out in this essay, we do not reproduce the original image.
A recipient who does not know the publisher’s political self-understanding, and who was not present in the community in which the “correct” reading was self-evident, gets hold of only the surface: an associative chain that repeats an old, demeaning juxtaposition with new energy. The deniability that was supposed to protect the maker does not protect the recipient. And least of all does it protect the group whose traits were used as ammunition.
This is exactly the mechanism that research has documented specifically in analyses of “trolling” and “for the lulz” culture. The ironic frame does not neutralise the material; it only makes the material circulable and harder to challenge, because the challenger always looks like someone who just “doesn’t get the joke.” The researchers who have followed this phenomenon the longest have arrived at the conclusion that the mechanism is an instrument, not an ideology. It works with equal force from either direction, because its effect does not depend on the user’s inner conviction but on what actually circulates, and in what form.
For the same reason, hoisting the other side’s communications up the flagpole and loudly wringing one’s hands over them only gains them more visibility — and exposes more people to the harmful content. This mechanism is at least, to some degree, already familiar to a broad audience. “Don’t feed the troll.”
III. Two accelerationisms
For the mechanism described above to settle into its political frame, we must unpack a concept that haunts the background of this entire field: accelerationism. The word circulates in online discussion so carelessly that its two main branches are constantly confused with one another. And this very confusion is at the heart of the case at hand.
The shared starting point of both branches is the idea that the crises of the present system are solved not by braking but by accelerating. The system’s internal contradictions must be driven to their point of culmination so that something new can be born. But that is where the commonality ends, and the difference is far sharper than the kinship of the names suggests.
Left accelerationism — whose best-known formulation is probably Srnicek and Williams’s accelerationist manifesto from the early 2010s — is at bottom a programme of construction. Its thesis is that the technology, infrastructure and organisational capacity produced by capitalism are not enemies but an unfinished inheritance. They must be taken over and redirected toward goals at which market logic would never arrive on its own: toward time freed by automation, toward common planning, toward shared prosperity. Whatever one thinks of the programme, its structural character is indisputable. It requires institutions, long-term planning, broad coalitions and — this is decisive — trust. None of these can be built by corroding the space of public conversation. It is, in a way, a continuation of the Nordic social-democratic process and consensus politics. Not very sexy. Mundane grind and grinding compromise are hard to sell to anarchists enamoured of revolutionary romance.
Right accelerationism is a different animal. At its philosophical end sits the late work of Nick Land and neoreaction, for whom democracy itself is an irreparable error. At its practical and most violent end sits collapse-accelerationism, whose slogan “there is no political solution” sums up the entire programme. Since the system cannot be repaired, its collapse must be hastened. The goal is not a better system but collapse itself, upon whose ruins the strongest of the survivors will then build whatever they build. And here it is essential to note what such a programme requires. No construction of any kind before the collapse, only corrosion: the fastest possible eating-away of trust, of the capacity for conversation, of institutions, and of shared reality.
From this follows a structural asymmetry that is decisive for this essay. Transgressive shock content is by its nature corrosion. It does not build coalitions, does not grow trust and does not produce institutions; it corrodes the ground of conversation regardless of who spreads it and why. The mechanism is symmetrical. It consumes shared space in exactly the same way in the hands of either side, but the benefit is not symmetrical. Chaos is not a neutral currency. It benefits the party whose goal is chaos. For the builder, a corroded conversational space is a defeat in every case. For the corroder it is always a victory, including when it is carried out by a political opponent on their own initiative and at their own expense. One might almost say especially then.
And now back to our case. By every indication, the publisher of the meme under examination conceives of themselves as a left accelerationist. The provocation is meant to serve some great awakening, the shock effect is meant to expose the rot of the system, and the uglier the conversation becomes, the closer the rupture is. Their chosen instrument simply cannot produce a single one of the things a left programme would require. It does not build a coalition; it drives away allies. It does not grow trust; it proves that their camp cannot be trusted. It does not shift the Overton window in the direction they hope, but in the direction toward which a repeated stereotype always shifts it. They are not a left accelerationist using a questionable means. They are a voluntary subcontractor of right accelerationism, who has filed their work under the wrong flag.
IV. Edgelord: a role, not a developmental stage
How is such a category error possible in the first place? The answer is found in the figure named in this essay’s title: the edgelord.
The edgelord is an established role-expectation of online culture, with roots in the forum culture of the 2000s and its “for the lulz” economy. The logic of the role is simple. Social currency is earned through norm-violation, and status within the in-group is measured by how far one dares to go. The reaction is the reward, regardless of the content. Shock, disapproval and “ratios” are all equally valid currency, because they prove that the line was crossed and the crosser noticed. In this economy, the “rough joke” is not an instrument of communication but a performance, whose value is measured by the force of the audience’s flinch.
It is important to understand that this is specifically a role and not a developmental stage. The role is often associated with the traits of the teenage rebel position: testing limits, irritating authorities, seeking attention by any means available. But because it is a cultural register and not a biological age, it is available to anyone, at any stage of life. A fifteen-year-old can take up the role, and a fifty-year-old can take up the role. The latter case is sadder to watch, because the status the role offers no longer converts into anything outside itself.
The role also carries its chronic occupational disease, which online-culture research has called irony poisoning. A long-maintained ironic position, in which one never has to be accountable for anything, because it’s all “banter”, eventually erodes the maker’s own capacity to tell what they actually believe. Irony that began as distance from one’s own claims imperceptibly becomes distance from one’s own judgment. This is the bridge that connects the edgelord to the accelerationism analysis of the previous chapter. Irony that no longer knows its own side inevitably drifts to the side that the mechanism it uses serves.
And it is precisely here that the party which knows its own side extremely well enters the picture: the professional reactionary propaganda machine. From its point of view the edgelord, regardless of the level of their political self-understanding, is a free distribution layer. The machine’s artefacts are designed so that their circulation does not depend on the sharer’s conviction: the meme carries its associative chain all the way, whether it was shared by a loyal supporter, an “ironic” edgelord, or a hand-wringing opponent.
The goals are known, and they are also well documented. Normalising claims through repetition, because familiarity resembles acceptability. Exhausting moderate voices with endless “joke or not” interpretive labour (and shifting the burden of proof onto the recipient more generally). Blurring the line between sincerity and parody so that no claim can any longer be bindingly read as anyone’s position. And, as the finest achievement, recruiting opponents as one’s own amplifiers.
The old observation of online culture known as Poe’s law is this machine’s bedrock. Without an unambiguous marker, parody and sincere extremism cannot be told apart. Hence it is worthwhile for a machine bent on corrosion to produce content that lives precisely on that zone of indistinguishability. It is for exactly this reason that far-right, corrosive “humour-sludge” is so dominant in the content stream. And it is exactly why there is a great temptation on the opposing side to join in, because it looks like the spot where the current runs strongest and the attention foams highest.
From the accelerationist canon one might borrow one more concept that describes all this more precisely than any term of propaganda analysis: hyperstition, the self-fulfilling fiction. The “ironic” extremist claim is hyperstition in miniature. It is set in motion as a fiction, with the unspoken caveat “no one really means this.” Every repetition produces more believers for it, until the fiction has gathered reality around itself. Those who shared it “as banter” find themselves, at some point, standing in the middle of a movement they helped speak into existence.
V. That boring killjoy question again: who bears the cost?
Here it is worth separating two different questions that are easily conflated: who is the target of the joke and who bears the cost of the joke. They are not the same thing, and it is precisely their difference that reveals the mechanism’s partiality most sharply.
Let us return to that recent real example. A meme in which the sitting prime minister is mocked by associating him with a stereotyped image of homosexuality and of being Swedish-speaking. The target of the joke is a politician in power: a legitimate target of mockery in any political satire, because power both deserves its criticism and can withstand it.
The cost of the joke, however, is borne by entirely others: gay people and Swedish-speaking people, who are not part of this power struggle at all, but whose (imagined) identity traits are borrowed as demeaning ammunition. The prime minister does not suffer from being associated with homosexuality or with being Swedish-speaking. He was in power before the joke and remains in power after it. But every time such a juxtaposition circulates within the frame of humour, it reinforces once again that old, stale associative chain in which refinement, cultivation or a certain way of speaking signify weakness and ridiculousness precisely because they deviate from the traditional, blockheaded norm of masculinity. This is a blow to the side disguised as a blow upward: the target is the power-holder, but the splatter flies into the eyes of bystanders.
Incidentally, the joke is so tasteless and so detached from reality that one would be tempted to suspect it of being a far-right false-flag operation, were the source not known. Regardless of intention, the joke succeeds in reviving already-overdue prejudices; in generating sympathy for the prime minister (which is an achievement in itself!) because he is being attacked so uglily and so tastelessly; and, on top of all that, in driving away allies by demonstrating that the maker or makers who count themselves on the left evidently have a wretchedly racist and homophobic taste — which, by propaganda’s iron laws, casts doubt on the whole left: “we always knew they were an untrustworthy bunch of trash”. So polarisation between groups grows and corrosive acceleration advances. No small propagandistic achievement for one ill-considered meme. In the previous chapter’s terms: a single such artefact does the work of three campaigns, for free, for the very machine its publisher imagines they are opposing.
This is not a unique case; it merely happens to be the cleanest example at hand, and precisely for that reason it is analytically valuable. It reveals the mechanism clearly without any need to assess an individual publisher’s inner motivation. It is neither necessary, nor even possible from the outside, to know whether the publisher felt they were being brilliantly sarcastic in resorting to already rather dated prejudices, or whether they were venting some unintegrated aggression of their own. Neither answer changes the meme’s operation in the world. The joke circulates in the same way, carries the same associative chain and causes the same cost, regardless of what the publisher’s inner experience was.
This is in fact a liberating way to approach the whole question. Instead of trying to assess the state of the maker’s soul, which is a near-impossible and, in this context, ultimately irrelevant task, one can assess the artefact and its circulation, which is both more precise and fairer. For the same reason, the previous chapter’s edgelord analysis is to be read as a description of a cultural register, not as anyone’s individual diagnosis. The role exists and is recognisable from the artefacts it produces, regardless of who has taken it up on any given occasion, and why.
VI. Counter-argument: doesn’t irony have a direction after all?
For fairness’s sake, it is worth taking seriously the strongest counter-argument that a defender of “ironic” transgression would present. And it is not a non-existent one.
Cultural studies has a long and well-founded tradition that distinguishes two structurally different phenomena: a group’s internal self-irony and parody (queer camp, drag, a marginalised group’s own mockery of its own stereotypes) on the one hand, and the borrowing of a group’s traits from outside as a weapon against some third party on the other.
The first has historically functioned as empowering. When the target itself takes over and exaggerates its own stereotype, the gesture disarms the stereotype, because control remains with the one it concerns. An “ironic” homophobic joke told by a gay person about themselves, and the same joke told by a straight person against a third party, are not the same act even when the wording is identical. Context and power-position are part of the meaning. In popular culture the phenomenon has perhaps been explained most clearly to a broad audience by African-American culture and by the discussion around gender and sexual minorities, with the core question of who may say what, and in what circumstances.
This distinction is valid and important, but it does not reach far enough to protect the case at hand; in fact it sharpens the problem. In the case of the prime minister meme, the maker evidently belongs to neither of the borrowed groups (they were not presenting their own stereotype), and the target (Prime Minister Orpo) likewise belongs to neither of the groups whose traits were used. So this is not internal self-irony, nor even a direct attack on the target’s own identity. It is the borrowing of two bystander groups’ identity traits to smear a third. This is something for which the cultural-studies tradition has no defence as good as the one it has for internal parody. It empowers no one, because none of the joke’s parties is the one the borrowed traits actually concern.
VII. Back to the interpretive keys
In the previous essay we sought to show that deniability can protect a small individual against totalitarian power, when the keys of interpretation are held by a bounded community. This essay shows that when the same technique is used in a broad, fragmented public sphere — where no one any longer controls who reads which layer — deniability ceases to be a shield and becomes an instrument for evading responsibility, regardless of the political flag under which it moves. The difference is not in who uses the technique, but in whether the keys are controlled. And in broad public distribution they can never be fully controlled.
Poe’s law gives this observation its final, almost mathematical form. In sufficiently broad distribution, the shrinking of unmarked irony into mere surface is not a risk but a wholly certain and predictable factor. The question is not whether a recipient farther from the core can read the artefact wrongly, but how many intermediary hands are needed before the misreading becomes statistically dominant. And the answer is: fewer than any publisher would like to believe. Close friends may still understand, but already at the level of acquaintances significant scatter begins to appear.
From this follows a practical and uncomfortable principle, one more honest than any defence that appeals to political direction. A transgressive gesture is responsible for what it actually sets in circulation, not for what its maker hoped it would mean. If irony is to be kept as a shield and not a weapon, it must be marked so clearly that its shrinking into surface is impossible even for the third-degree recipient. Otherwise it is not irony but merely a prejudice that has made itself a temporary safe harbour beneath the laughter. And that safe harbour is available to anyone, regardless of which side’s flag they imagine themselves to be sailing under.
VIII. The craftsman’s ethics of transgression
The conclusion of this essay would be easy to read as a prohibition: don’t transgress, don’t ironise, stay on the clean side of the line. That would be a wrong reading, and against the spirit of the whole series. The earlier parts have shown that transgression, double-coding and deniability have produced some of the most significant art of the last hundred years. The problem is not transgression but bad transgression — and the bad is told from the good in exactly the same way as in any craft: does the maker control their instrument, or the instrument its maker?
I therefore propose, instead of a prohibition, a craftsman’s ethics that condenses into four tests. They follow directly from everything said above, and not one of them appeals to the maker’s inner conviction, only to the artefact and its circulation.
The first test: who controls the keys? Is there a community that holds the “correct” reading, and does the artefact stay within that community’s reach, or has it been released into free circulation, where the keys disintegrate at the very first share?
The second test: who bears the cost? Not who is the target, but from whose hide the joke’s energy is actually torn. If the cost is borne by someone other than the target — and especially if it is borne by a group that is not a party to the power struggle at all — the transgression has failed, regardless of how clever it felt to its maker.
The third test: does the irony survive the degrees of distancing? Does the frame survive the journey without the maker? If the artefact’s “correct” reading requires knowing the maker’s intention, Poe’s law guarantees that it will not survive. The risk of an effect contrary to the intention rises quickly.
The fourth test: does it build or corrode? Does the gesture leave behind more or less trust, more or less capacity for shared conversation? This is the acceleration test: corrosion always serves the corroder, under whichever flag the eroder operates.
And there exists an entire canon that passes all four. Camp and drag, in which the owner of the stereotype exaggerates their own into harmlessness. Self-parody, in which the cost is borne by the maker themselves, belongs to the oldest and surest ways of earning the right to laugh. Satire that strikes power with power’s own traits — with its decisions, deeds, words, hypocrisy — and not with identities borrowed from bystanders, is for exactly that reason more devastating. It gives its target no opportunity to cast itself as a victim, and gives its audience no reason for pity. The same transgressive energy that, wrongly aimed, works as subcontracting for the opposing side, produces — rightly aimed — what transgression at its best has always produced: the moment in which power looks ridiculous, without the bill falling on any bystander.
For the logic of hyperstition, after all, works in both directions. If a repeated fiction makes itself true, then the transgressive maker’s real power lies not in what they dare to say, but in which fiction’s speaking-into-truth they take part. And this question is more acute right now than ever, because the very environment in which the choice is made is itself only being born.
IX. The terminus of nihilism is the starting point of a new adventure
For everything this essay series has observed so far — the layering of meanings, the control and disintegration of interpretive keys, the pervasiveness of irony, the inflation of deniability, the divergence of cost from target — is not merely a diagnosis of past and present online culture. It is inevitably also a requirements specification for what is to come.
Whatever those digital environments turn out to be in which people meet one another over the coming decades, they will collide with these problems. I do not believe any environment can succeed without offering strong answers to them. How does context travel with the artefact? How can the in-group reading and the out-group reading live side by side without tearing either apart? How can meaning ripen gradually, instead of either staying hermetic or collapsing into surface all at once?
The present algorithmic feed is the worst possible architecture for these questions. It tears every artefact from its frame and distributes it to precisely the audience whose reaction is strongest — meaning that the loss of the keys is scripted into the system itself. But nothing compels us to continue this way. It is entirely possible to imagine environments in which meaning has a topology: nearness and distance, degrees of development, zones in which a joke is still a joke, and boundaries beyond which something of the frame travels with it. In such an environment, this essay’s edgelord problem would not disappear, but it would change from a force of nature into a design question. And that is an enormous difference.
Here it is worth making one restoration of value that may come as a surprise. The so-called brainrot culture of the youngest online generation, deliberately meaningless, self-consuming, appearing to the adult eye as pure noise, is tempting to write off as a symptom of decay. That is a mistake (one which, incidentally, older generations always seem to repeat). Looked at more closely, the phenomenon is something quite different. For it is the first intuitive response to a culture in which every meaning is already weaponised. When you grow up in an environment where every image is the carrier of someone’s agenda, every irony possibly sincere and every sincerity possibly irony, it is entirely consistent to withdraw to a zone in which nothing means anything — because it is the only zone in which no one can lie and there is no risk of misinterpretation. Brainrot is the zero point of meaning in the same sense that Dada’s gibberish poetry was the zero point of language. Not stupidity and not terror, but an immune reaction. An exact replica of nonsense, made from an environment that has become nonsensical. It is not to be idealised, but neither is it to be despised. It is the starting point of something new, not a place to settle down and live.
Nihilism has a terminus, and it is always the same. Dada shouted its negation to the end and discovered that afterward it was Monday and something still had to be done. The Dadaists became Surrealists, Constructivists: builders of something, at least, even if it seemed strange at first. Punk roared its own voice hoarse — “no fun, you’ve been had” — and out of it sprouted the decade of post-punk, DIY culture and independent institutions. It built more than any grown-up propriety that preceded it. When the last negation has been shouted, only two options remain: to build, or to surrender the ruins to the one who has been waiting for them all along. And as chapter III of this essay showed, there are already creeps queuing for the ruins, with rather chilling plans at the ready.
The return of values, whose first signs are already in the air, the cautious new course of sincerity, “post-irony” that is really just honesty after irony, is therefore not a nostalgic regression but the next phase of the same movement. Building is not something to leave to the fascists. They are bad at it, but they are the only ones who want to do it alone and on their own terms.
Building begins with taking the material seriously. A culture begins to exist the moment one can distinguish genres within it. When forms emerge from the noise, conventions from the forms, and from the conventions that quiet web of agreements against which a single act can be skilful or clumsy, faithful or rebellious. Meme culture already has these genres, though it itself rarely consents to admit anything so bourgeois.
Ville Kiiski
The next essay deals with the genres of meme culture and their connections to earlier cultural layers, and seeks to sketch possible bridges to what is to come.
Key background materials for the interested reader
Whitney Phillips, This Is Why We Can’t Have Nice Things: Mapping the Relationship between Online Trolling and Mainstream Culture (MIT Press, 2015). A foundational study of trolling and “for the lulz” culture; shows how the ironic frame makes material circulable and hard to challenge.
Alice Marwick & Rebecca Lewis, Media Manipulation and Disinformation Online (Data & Society Research Institute, 2017). Explicitly documents the gateway mechanism in which milder memes serve as a bridge to more extreme material. Freely available: datasociety.net
Ashley Feinberg, “This Is The Daily Stormer’s Playbook” (HuffPost, 2017). A leaked style guide from a far-right publication that instructs writers to keep content deliberately ambiguous — a primary source, in the machine’s own voice, for what this essay says about Poe’s law.
Further: Angela Nagle, Kill All Normies (2017), a more journalistic and more contested overview of the same period’s online culture, but frequently cited.
Leave a Reply