AI NAsties
I. Introduction : The Pattern That Keeps Repeating
Somewhere right now a man is being told that the AI generated music video he made is a symptom of a dangerous worldview. The video features an attractive woman. She is not real. No one was harmed in her creation. She exists as a generated figure attached to a creative work produced by someone who wanted to make something and had access to tools that made it possible. The tools are new. The attractive woman in a creative work is not. Neither is the argument being made about what her presence means, what it says about the man who put her there, and what community spaces owe the people who find her existence objectionable.
We are living through a moral panic about AI generated sexual content and most of the people participating in it do not know that is what it is.[1][2][3] The language being used does not feel like panic. It feels like analysis. It arrives with citations, with psychological frameworks, with named researchers and peer reviewed papers. It speaks of pipelines and radicalization pathways, of objectification and the male gaze, of communities that normalize harm by platforming content that reduces women to tropes. It is confident in the way that moral panics always are, which is to say more confident than the evidence warrants and considerably more confident than history suggests is wise.
The specific triggers of this particular panic are worth naming because they are the same triggers that have preceded every previous version of this argument. The first is democratization.[4] Generative AI has done to sexual imagery what home video did to transgressive film, what desktop publishing did to obscene literature, what file sharing did to explicit content of every kind. It has removed the technical gatekeeping that previously limited who could produce this material. When production required skill, equipment and resources, the volume of transgressive content was naturally constrained. When those barriers fall the volume becomes visible in a way it never was before and visibility reads as escalation even when the underlying desire is not new.[5] The second trigger is scale. Algorithmic tools can produce in an hour what would have taken a skilled artist weeks. That scale feels different even when the individual output is not. The third is the removal of what people unconsciously treat as a moral filter. If making something required effort and craft there was an implicit assumption that the maker had considered what they were doing. Automation removes that assumption and the removal feels like the removal of conscience itself even when the conscience of the individual creator is entirely intact.
These triggers combine to produce a predictable response which this essay will trace across multiple historical moments because the response is old enough to have a track record. That track record is not encouraging for the people currently making the argument. The thesis here is simple and will be defended at length: this is not a new argument. It is a new technology wearing an old argument with new citations. The citations change. The structure of the claim does not. And the structure of the claim has not fared well when measured against what actually happens after the panic subsides.
This essay is historical, scientific and philosophical in roughly equal measure. It will examine how the current panic fits into a pattern stretching back further than most of its participants realize. It will engage seriously with the psychological and sociological frameworks being deployed against sexual expression in AI creative communities, including social constructivism, objectification theory and radicalization research, not to dismiss them wholesale but to examine where they have genuine explanatory power and where they have been extended well beyond what their evidence supports. It will draw on evolutionary psychology and developmental science to offer a different account of male sexuality than the one currently dominating these conversations. And it will make an unambiguous philosophical argument in defense of the full spectrum of artistic expression, including the transgressive, the gratuitous, the lowbrow and the purely gratificatory, not because all of it is good art but because the freedom that protects the art nobody asked for is the same freedom that protects the art everybody eventually agrees matters.
Before any of that a distinction needs to be on the table because it will run through everything that follows. There is a meaningful difference between content that involves real non-consenting people and content that involves generated or fictional figures. Nonconsensual deepfakes of real identifiable women are a genuine harm with genuine victims.[6] AI generated child sexual abuse material is not a gray area. These things deserve the serious legal and ethical responses they are receiving. They are not the subject of this essay except insofar as they are being used as rhetorical leverage against an entirely different category of content, which is adult fantasy involving generated figures who do not exist and cannot be harmed. The conflation of these categories is not a precaution. It is a logical error, and it is one that does significant damage to the clarity of a conversation that would benefit from considerably more precision than it is currently getting.
The organizing frame for what follows is a specific historical moment that most of the current participants in this debate are too young to remember clearly and too incurious to have looked up. In the early 1980s in Britain a moral panic erupted around a category of films that became known as video nasties.[7] The panic had everything the current moment has. New technology in the form of home video. Democratized access to content previously controlled by institutional gatekeepers. Moral entrepreneurs translating aesthetic disgust into public health language. Science recruited to confirm what the disgusted already believed. Demands that platforms and institutions take responsibility for what their audiences might see. The films at the center of that panic are now available in certified editions, studied in universities and recognized as significant works in the history of horror cinema. The science that was supposed to demonstrate their harm has been methodologically discredited. The epidemic of violence they were supposed to produce did not materialize at any detectable scale consistent with the predictions made. We will return to the video nasty throughout this essay because it is the clearest recent example of how this pattern runs, how it ends, and how certain the people driving it were at every stage that this time the concern was legitimate, this time the evidence was solid, this time the content really was different from everything that had been defended before.
There is one more thread to introduce before we go further because it underlies the specific contemporary argument this essay is responding to even when that argument does not name it directly. In current cultural discourse female sexuality is generally treated as something to be explored, contextualized and humanized. Female desire is complex. Female sexual expression is agentive. Women who produce sexual content are exercising autonomy over their own bodies and creative output. These are largely defensible positions. The problem is what sits alongside them. Male sexuality in the same discourse is treated not as something to be explored and contextualized but as something to be explained and contained. Male visual desire is not complex, it is a symptom. Male sexual expression is not agentive, it is a power move. Men who consume sexual content are not satisfying a human need, they are participating in a system of harm.[8][9] The asymmetry is so thoroughly embedded in the current conversation that most people reproducing it do not notice it is there. This essay notices it. It will keep noticing it. And it will argue that any account of sexuality that extends humanity to one half of the species while pathologizing the other is not an ethical framework. It is a prejudice wearing one.
II. We Have Been Here Before : The Video Nasty Panic
To understand what is happening right now in AI creative communities you need to go back to Britain in 1979 and stand in a high street video rental shop. The shop probably smells of carpet cleaner and cigarette smoke. The shelves are lined with clamshell cases and oversized cardboard boxes because the home video industry has not yet standardized its packaging any more than it has standardized its content. What you can rent from this shop is genuinely remarkable by the standards of what was available twelve months earlier. Films that never received a cinema release in Britain. Films that were shown briefly in specialist venues and then disappeared. Films from Italy, America and Germany that existed in a kind of regulatory vacuum because nobody had anticipated that the technology to distribute them directly into living rooms would arrive this quickly or become this affordable this fast.
The Video Recordings Act did not exist yet. The British Board of Film Censors, which would later become the British Board of Film Classification, had no jurisdiction over home video. It classified films for theatrical release.[10] It had no mechanism for and no authority over the cassettes now multiplying across the country's rental market. For a period of roughly four years between the mass market arrival of home video and the legislative response to it, the distribution of filmed content in Britain operated in a state of almost complete regulatory freedom. What filled that freedom was, among many other things, a wave of horror and exploitation films that would not have passed the BBFC's theatrical certification standards and which were now available to anyone with a VHS or Betamax player and a few pounds for a weekly rental.
This is the context in which the video nasty panic was born. Not in the content itself, which had existed in various forms for years, but in the sudden visibility of that content to an audience that had previously been protected from it by the practical barriers of specialist distribution. The films did not get more extreme. They got more accessible. And accessibility, as it always does, read as escalation.
The Films
The Director of Public Prosecutions eventually compiled a list of seventy-two films that were either successfully prosecuted under the Obscene Publications Act or were considered liable to prosecution.[11] The list is worth examining in some detail because the gap between what these films actually were and how they were publicly described tells you something essential about how moral panics work.
Driller Killer, directed by Abel Ferrara in 1979, was among the first and most prominently targeted.[12] The cover art, a man pressing a power drill against another man's skull, was reproduced in newspapers and parliamentary debates as though it were documentation of the film's content rather than exploitation marketing of the kind that had existed since the penny dreadful. The film itself is a low budget character study of a struggling artist in New York whose mental deterioration culminates in violence. It is not a comfortable watch. It is also considerably more interested in the texture of poverty and creative failure than its cover art suggested. The drill appears relatively briefly. The film that was held up as evidence of a new threshold of depravity was in significant part a portrait of artistic desperation in a city in economic collapse.
Cannibal Holocaust, directed by Ruggero Deodato in 1980, is the most genuinely difficult film on the list and the one that most deserves serious examination rather than reflex condemnation. Deodato was arrested in Italy after its release because authorities believed the cast members had actually been killed during production.[13] He was required to produce them alive in court to avoid a murder charge. The film contains real animal deaths which remain its most defensible target for criticism and which Deodato himself later expressed regret about. It also contains graphic sexual violence and scenes of extreme gore. What it is doing with those elements is more complicated than its critics acknowledged. The film is a structural critique of documentary voyeurism and media exploitation, a found footage film a decade before The Blair Witch Project that turns its camera on the people holding the camera and asks what the desire to film atrocity says about the filmmakers and their audience. Its thesis, that Western civilization's self-congratulatory horror at primitive violence is hypocritical given what civilization does when it arrives in the jungle, is not a comfortable one. It was not designed to be. The film is genuinely transgressive in the sense that it transgresses the audience's expectation of a comfortable relationship with its own disgust. Whether it succeeds as art is a reasonable debate. Whether it is simply evidence of depravity in its makers is not.
The Evil Dead, Sam Raimi's 1981 debut, is perhaps the most instructive case on the list because of what subsequently happened to it. At the time of its prosecution it was described as one of the most obscene films ever made. It is a low budget horror film in which demons possess and kill a group of young people in a cabin in the woods. It contains a sequence of sexual violence involving a woman and a tree that remains confronting. It also contains extraordinary inventive filmmaking from a young director with almost no money and enormous ambition. It launched one of the most celebrated careers in Hollywood. It has been remade, sequelized and analyzed in film schools across the world. In 2023 a stage musical adaptation ran in the West End of London, the same city whose legal system had once prosecuted the film as obscene. The transformation of The Evil Dead from video nasty to beloved cult classic to West End musical is not incidental to the argument of this essay. It is the argument.
I Spit on Your Grave, directed by Meir Zarchi in 1978, is the film on the list that generates the most genuinely complex critical response even now. It depicts the rape of a woman and her subsequent revenge on her attackers in equal and unflinching detail. Roger Ebert gave it zero stars and called it a film without a shred of artistic distinction.[14] Carol Clover, whose work we will return to, read it as a revenge narrative that transferred the audience's identification to a female protagonist in ways that complicated simple readings of exploitation. Feminist critics have been divided on it for four decades. It is many things. A simple document of misogyny produced for the gratification of misogynists is not an adequate description of what it is doing, which is not to say it is comfortable or that discomfort about it is unreasonable. It is to say that the films being prosecuted as straightforwardly harmful were frequently more complicated than the prosecution acknowledged.
SS Experiment Camp, Faces of Death, The Funhouse and the other films on the DPP list varied enormously in quality, intent and content. Some of them are genuinely poor films with little to recommend them beyond their transgression. Some of them are interesting works that deserved better than prosecution. What they shared was not a consistent level of harm or a consistent artistic failure. What they shared was visibility at a moment when visibility itself was the problem.
Mary Whitehouse and the Moral Entrepreneur
No account of the video nasty panic is complete without Mary Whitehouse, and no account of Whitehouse is fair without acknowledging that she was a genuinely formidable woman operating with real conviction. She was not stupid. She was not simply a figure of fun, though she became one. She was an effective political operator who understood before most of her contemporaries that cultural production was a site of genuine contestation and that the people making decisions about what appeared on British screens were not ideologically neutral.
Whitehouse founded the National Viewers and Listeners Association in 1965, initially to campaign against what she saw as the BBC's promotion of permissive values.[15] By the early 1980s she had expanded her focus to video content and she understood the home video moment immediately and accurately as a regulatory gap that her campaign could exploit. She was not wrong that something had changed. She was wrong about what it meant and wrong about what would happen as a result of it.
Her specific methods are worth examining because they are recognizable. She did not primarily make empirical arguments. She made emotional ones dressed in empirical clothing. She collected anecdotes, testimony from parents and teachers about children who had seen things they should not have seen, and presented these as evidence of systematic harm. She translated aesthetic disgust, her own and others, into the language of child protection, which is the most rhetorically powerful available register in democratic politics because it is almost impossible to argue against without appearing to argue for harm to children. She lobbied politicians directly and effectively. She cultivated relationships with sympathetic journalists. She understood that a well placed quote in the right newspaper could accomplish more than a carefully argued pamphlet and she acted accordingly.
The transformation of her aesthetic preferences into public health language is the template for the moral entrepreneur and it is a template that has been used before and since with remarkable consistency. [16]The specific content changes. The structure of the argument does not. Disgust becomes concern. Concern recruits science, or the appearance of science. Science becomes policy. Policy becomes law. And the content that provoked the original disgust is reframed not as something the campaigner finds offensive but as something that objectively harms identifiable victims. The move from I find this disgusting to this causes measurable harm is the central rhetorical operation of the moral panic and Whitehouse performed it with considerable skill.
The Daily Mail and the Feedback Loop
Whitehouse needed amplification and she found it in the tabloid press, particularly the Daily Mail, which ran a sustained campaign against video nasties that was essential to transforming a fringe campaign into a parliamentary priority.[17][18] The Daily Mail's role in this panic is worth examining as a case study in how media amplification works because the mechanism it deployed is identical to mechanisms currently operating in digital media around AI content.
The Mail ran headlines about children being traumatized by content their parents had rented. It published frame grabs from the most extreme moments of the targeted films without context, presenting the worst image from a ninety minute work as representative of the whole. It ran testimony from teachers, social workers and psychologists who were willing to connect the content to behavioral problems in the children they worked with. It created a feedback loop in which politicians who had not seen the films responded to the coverage rather than the content, constituents wrote to their MPs about coverage they had read rather than films they had seen, and the parliamentary debate that followed was shaped by Mail front pages rather than by any systematic examination of the evidence.
Graham Bright, the Conservative MP who introduced the Video Recordings Bill that became the 1984 Act, cited in parliamentary debate a study claiming that significant percentages of young children had seen films from the DPP list. The study, conducted by researchers at the University of Westminster, has been widely criticized as methodologically flawed.[19] It relied on self-reporting by children who were shown a list of film titles and asked which ones they had seen. The list included films that did not exist. Substantial percentages of children reported having seen these nonexistent films. The study demonstrated not that children were watching video nasties but that children will report having seen things they have not seen when asked by an adult in an authority context. This methodological failure was pointed out at the time. It did not slow the legislation.
The Video Recordings Act 1984
The Video Recordings Act received royal assent in July 1984.[20] It required that virtually all video recordings be submitted to the BBFC for classification before they could be supplied commercially. It gave the BBFC statutory authority over home video for the first time, authority it had never had over theatrical releases in quite the same form. In practice it meant that films on the DPP list became legally unavailable in the United Kingdom. Possession with intent to supply became a criminal offense. Retailers who had been operating in a regulatory vacuum found themselves subject to prosecution.
The Act was in many respects a reasonable piece of legislation responding to a genuine regulatory gap. The home video market had expanded faster than any framework for governing it and some mechanism for applying existing obscenity standards to the new medium was probably inevitable and not obviously wrong. What the Act also did was legitimize the specific claims that had been made about the content it was designed to restrict, not through any judicial finding about harm but through the simple political fact of legislation. Once Parliament had acted on the assumption that video nasties caused harm the assumption acquired the status of established fact in public discourse regardless of what the evidence actually showed.
The Science and Its Dismantling
The scientific claims made in support of the Video Recordings Act and the broader panic deserve careful examination because they follow a pattern that appears in every subsequent moral panic about media content and that appears in the current AI content debate with only minor variations.
The core claim was behavioral: exposure to violent and sexually violent content caused viewers, particularly young viewers, to become more aggressive, more tolerant of violence against women, and more likely to imitate what they had seen. This claim was supported primarily by laboratory studies of a specific kind. Researchers would expose subjects to violent or sexually violent media and then measure their aggressive responses using standardized instruments, typically involving willingness to administer electric shocks or noise blasts to another subject.[21] The studies consistently found short term effects. They consistently failed to demonstrate that these short term laboratory effects translated into real world behavior change. The gap between what the research showed and what the research was said to show was substantial.
Martin Barker and Julian Petley assembled the scholarly response to the panic in their edited collection Ill Effects, first published in 1997 and expanded in 2001.[22] The book gathered contributions from media researchers, psychologists and sociologists who examined the evidentiary basis of the harm claims systematically and found it wanting. The laboratory aggression studies, Barker and his contributors argued, were measuring something real but something much more limited than their deployment in public debate acknowledged. Short term arousal effects in laboratory conditions do not translate directly into behavioral predictions about real world conduct. The populations studied were not representative. The measures of aggression used were not validated against real world aggression. The studies that found effects were published. Studies that failed to find effects were less likely to be published and less likely to be cited when they were. The evidentiary base for the panic's scientific claims was considerably weaker than the parliamentary debate it informed.
The longer term verdict on the research is even clearer. Subsequent meta-analyses of the media violence literature have consistently found that the effect sizes reported in individual studies are substantially reduced when publication bias is corrected for, that the relationship between media consumption and real world violence is weak and almost certainly mediated by factors the simple exposure model cannot account for, and that the predicted behavioral consequences of exposure to violent media have not materialized in the populations that have consumed it most extensively.[23]
The Long Term Verdict
The films that Parliament legislated against are now, with some exceptions, legally available in Britain. The Evil Dead has been certified by the BBFC in multiple editions and spawned a franchise that includes the West End musical already mentioned. Cannibal Holocaust is the subject of serious academic analysis and is taught in courses on exploitation cinema, found footage and postcolonial film theory. I Spit on Your Grave has been remade twice by Hollywood studios. Driller Killer is available on streaming platforms. The Funhouse can be purchased at any major online retailer.
The epidemic of violence that the panic predicted did not materialize. The generation that grew up with access to video nasties did not become more violent than the generations that preceded it. Youth violence in Britain, as in most of the developed world, declined substantially in the decades following the introduction of the Video Recordings Act, which is the opposite of what the harm thesis predicted.[24][25][26] The researchers whose work was cited in parliamentary debate to justify the legislation have not been vindicated. The filmmakers whose work was prosecuted have in several cases been recognized as significant artists. The moral entrepreneurs who drove the panic are remembered, when they are remembered at all, with a mixture of affection and embarrassment.
Mary Whitehouse died in 2001. Obituaries acknowledged her energy and conviction while noting that the cultural transformation she campaigned against had proceeded largely without the catastrophic consequences she predicted. The Daily Mail, which ran the campaign that was essential to the panic's political success, did not run a correction.
The Pattern and the Parallel
The video nasty panic has the following structure. A new technology arrives that dramatically lowers the barrier to accessing content that was previously controlled by institutional gatekeepers. The volume and visibility of transgressive content increases not because transgressive desire is new but because the infrastructure for its distribution has changed. Moral entrepreneurs translate the aesthetic disgust of a specific cultural constituency into public health language centered on the protection of vulnerable populations, in this case primarily children. The tabloid press amplifies specific examples chosen for maximum emotional impact and presents them as representative of a general phenomenon. Science is recruited to provide empirical support for conclusions that the campaigners have already reached on other grounds. The science is presented with more confidence than its methodology warrants. Legislation follows. The content largely survives. The predicted harm does not materialize. The panic is eventually remembered as an overreaction.
Now substitute generative AI for home video. Substitute AI creative communities and their moderation policies for the BBFC and the Video Recordings Act. Substitute the language of pipelines, radicalization and the male gaze for the language of depravity and moral corruption. Substitute peer reviewed papers in psychology and gender studies for the University of Westminster study that included nonexistent films. The structure is identical. The confidence is identical. The gap between what the evidence actually demonstrates and what it is said to demonstrate is identical.
This is not a coincidence. It is a pattern. And the pattern has a track record that anyone making the current argument with the current level of certainty should be required to account for before that certainty is extended any further credit.
III. The Longer History: Male Sexuality and Art Under Siege
The oldest known sculptures in human history are of women. The Venus of Willendorf, carved from limestone approximately 25,000 years ago and discovered in Austria in 1908, depicts a female figure with exaggerated breasts, hips and abdomen.[27] The Venus of Hohle Fels, discovered in Germany in 2008 and dated to somewhere between 35,000 and 40,000 years ago, predates it and is currently the oldest known figurative art of any kind.[28] It depicts a female body with similarly exaggerated sexual characteristics and was found with a carved ring suggesting it was worn or carried. Whatever else these objects were, and the debate about their precise function continues, they were made by human hands at considerable effort using the best available technology of their moment to depict the female form in a way that emphasized its sexual characteristics. The people who made them are separated from us by tens of thousands of years of cultural development. The impulse their work represents is not.
This matters because one of the recurring claims in every moral panic about sexual imagery is the claim of novelty. The content being condemned is presented as something new, something that technology or cultural decay has introduced into a world that was managing fine without it. The Venus of Willendorf suggests otherwise. The idealized female figure produced for purposes that include but are not limited to male visual pleasure is not a product of moral failure or cultural degradation. It is one of the oldest continuous human creative traditions we have evidence of and it has been with us since before writing, before agriculture, before anything we would recognize as civilization.
The Ancient World
The erotic art of Pompeii is the most extensively documented ancient example and the most instructive for the current conversation because of what it reveals about the relationship between sexual imagery and social function in a culture very different from our own. The brothels of Pompeii contained explicit frescoes that served as what scholars believe were menu illustrations, images depicting specific sexual acts available for purchase displayed on the walls above the relevant rooms.[29] The baths contained erotic imagery. Private homes of wealthy Romans contained explicit mosaics and sculptures. The Warren Cup, a Roman silver vessel dated to the first century AD now held by the British Museum, depicts male homosexual acts in explicit detail and was almost certainly a luxury item displayed at dinner parties. Roman sexuality was frank, hierarchical and extensively documented in its visual culture in ways that would attract prosecution under contemporary obscenity law in multiple jurisdictions.
Greek symposium culture treated male sexuality with similar frankness. The symposium, a ritualized drinking gathering for male citizens, was attended by hetairai, educated female companions whose status was distinct from that of wives or slaves, and was decorated with pottery depicting sexual acts in considerable explicitness.[30] The poetry of Sappho, the comedies of Aristophanes, the dialogues of Plato on the nature of eros, all treat sexuality as a legitimate subject for serious intellectual and artistic engagement without requiring it to justify itself through a non-sexual purpose. The idea that sexual content requires redemptive intellectual scaffolding to deserve cultural space is not an ancient wisdom. It is a relatively recent and specifically Western theological inheritance that has been secularized into aesthetic theory without examining whether the underlying assumption deserves the authority it is still being granted.
The Shunga tradition of Edo period Japan, woodblock prints depicting explicit sexual scenes produced by some of the most celebrated artists of the period including Katsushika Hokusai and Kitagawa Utamaro, ran from the early seventeenth century through the nineteenth and constituted a substantial portion of the output of major artists working in a form now considered one of the high points of Japanese visual culture.[31] Shunga were produced for a mass market, sold openly, collected by both men and women across social classes, and depicted a range of sexual scenarios including heterosexual and homosexual acts with a frankness that has no parallel in contemporaneous European art. They were not considered degraded work. They were considered skilled work. The artists who produced them were not considered moral failures. The distinction between their serious work and their erotic work was not always clearly drawn because the distinction was not always considered meaningful.
The Kama Sutra, composed in Sanskrit sometime between the third and fifth centuries AD and attributed to the scholar Vatsyayana, is a text that Western culture has consistently misunderstood as a sex manual and that is in fact a comprehensive treatise on the art of living in which sexuality occupies one significant section among many.[32] It treats sexual pleasure as a legitimate human goal alongside virtue and material prosperity. It discusses sexual technique with frank precision. It also discusses the nature of love, the qualities of a good partner, the management of a household and the cultivation of personal refinement. The inclusion of explicit sexual instruction alongside philosophy and social guidance is not treated as a contradiction because in Vatsyayana's framework it is not one. Sexual pleasure is part of a good human life and discussing it clearly is part of the intellectual's responsibility. The notion that this is a category error, that the explicit and the serious cannot occupy the same work, is a prejudice of a specific cultural tradition and a relatively recent one at that.
The point of this survey is not that all erotic art is equivalent or that historical precedent establishes moral legitimacy. It is that the specific claim currently being made, that idealized sexual imagery of women produced for male pleasure is a symptom of a disordered cultural moment requiring intervention, is a claim that would require us to pathologize a continuous human creative tradition stretching back to before the last ice age. That is a significant claim and it requires significantly more evidentiary support than it is currently receiving.
The Obscenity Trials
The legal history of attempts to suppress sexual content in art is long enough and consistent enough in its outcomes that it constitutes its own argument. The pattern is as follows: a work appears that transgresses the sexual conventions of its moment, legal action is taken on the grounds that the work will harm its readers or viewers, the work is defended on artistic grounds with varying success, and the passage of time reveals that the predicted harm did not materialize while the work itself is either forgotten or recognized as significant. The specific legal mechanisms change across jurisdictions and centuries. The structure of the argument and the structure of the outcome do not.
The foundational British legal precedent is Regina v Hicklin, decided in 1868, in which the Court of Queen's Bench established what became known as the Hicklin test for obscenity.[33] The test asked whether the material in question had a tendency to deprave and corrupt those whose minds are open to immoral influences. The standard was extraordinarily broad. It allowed prosecution based on the effect of isolated passages on the most susceptible possible reader. It did not require demonstration of actual harm to actual people. It required only that the court find the tendency toward corruption plausible. Under this standard a vast range of literary and artistic work was technically obscene and a significant amount of it was prosecuted.
The suppression of James Joyce's Ulysses in the United States is the most consequential of these cases for literary history. The novel was serialized in The Little Review beginning in 1918 and was prosecuted for obscenity in 1920 following publication of the Nausicaa episode in which Leopold Bloom masturbates on a beach while watching a young woman. The editors Margaret Anderson and Jane Heap were convicted and fined.[34] The novel was effectively banned in the United States until 1933 when Judge John Woolsey of the Southern District of New York ruled that it was not obscene, finding that its effect on a person with average sex instincts would be that of a somewhat emetic but not an aphrodisiac and that it was written with a sincere and serious attempt to devise a new literary method.[35] Woolsey's ruling is important not only for what it permitted but for what it established philosophically: that the intention and method of a work matter to its legal status, that literary seriousness is a defense against obscenity claims, and that the effect on average readers rather than the most susceptible readers is the appropriate standard.
The Lady Chatterley trial of 1960 is the British landmark.[36] Penguin Books was prosecuted under the Obscene Publications Act 1959 for publishing D.H. Lawrence's novel in an unexpurgated edition. The prosecution invited the jury to consider whether it was a book you would wish your wife or your servants to read, a question that inadvertently revealed more about the prosecution's class anxieties than its concern for public morality. The defense called thirty-five expert witnesses including E.M. Forster, Rebecca West and Richard Hoggart, who argued that the novel had serious literary merit and that its treatment of sexuality served Lawrence's artistic purposes. The jury acquitted. The unexpurgated Penguin edition sold two million copies within a year. The prediction that widespread availability of the novel would corrupt its readers was not borne out by any subsequent evidence anyone thought to gather.
Henry Miller's Tropic of Cancer, published in Paris in 1934 and banned in the United States and Britain for decades, generated a series of obscenity prosecutions when Grove Press published an American edition in 1961.[37] The book is explicitly sexual, deliberately crude and entirely uninterested in justifying its sexuality through literary respectability. Miller was not making the argument that the sexual content served a higher purpose. He was making the argument that the sexual content was the purpose, that the frank depiction of male desire and its unglamorous pursuit was itself what the book was doing, and that this was sufficient reason for the book to exist. The Supreme Court effectively ended the prosecutions in Grove Press v Gerstein in 1964, and Tropic of Cancer is now taught in university literature courses, which is the standard ending for this particular story.
Allen Ginsberg's Howl was prosecuted for obscenity in San Francisco in 1957 following its publication by Lawrence Ferlinghetti's City Lights Books.[38] The poem contains explicit references to homosexual sex and drug use and makes no attempt to contain either within a framework of moral instruction or redemptive narrative. Judge Clayton Horn ruled in favor of the defense, finding that the poem had redeeming social importance and that the obscene passages were not separable from the work as a whole. The ruling established an important principle: that a work cannot be fairly judged by extracting its most transgressive moments and evaluating those in isolation from the artistic whole. A principle, it should be noted, that is routinely violated in contemporary arguments about AI sexual content, where the most extreme examples from the most irresponsible producers are presented as representative of a general phenomenon.
Naked Lunch and the Philosophy of Disgust
William S. Burroughs occupies a unique position in this history because he is the figure who most clearly and deliberately made the philosophical argument that discomfort, offense and disgust are not harm, and because the legal proceedings around Naked Lunch established that argument in American law in a form that has not been successfully challenged since.
Naked Lunch was composed in Tangier between 1954 and 1957, assembled by Burroughs with the help of Jack Kerouac and Allen Ginsberg, and published by Olympia Press in Paris in 1959.[39] It arrived in America through Grove Press in 1962. The book is not easy to describe because description implies a coherence the book deliberately refuses. It is a series of routines, Burroughs's term, loosely connected by recurring characters and obsessions, depicting drug addiction, sexual violence, bureaucratic control, and the mechanisms by which language and power operate on human consciousness. It contains scenes of explicit homosexual sex. It contains scenes of sexual violence. It contains a sequence in which a man is hanged and sexually used during his execution. It contains imagery that is designed to be deeply unpleasant, not as a failure of craft but as a deliberate strategy. Burroughs was not trying to make you comfortable. He was trying to make you see what you would rather not see, to confront the reader with the control systems operating on their own desire and their own language, and he chose to do this through content that was deliberately and systematically transgressive.
The Massachusetts prosecution came in 1965 when the Attorney General sought to have Naked Lunch declared obscene under state law.[40] The case went to the Massachusetts Supreme Judicial Court, which ruled in 1966 that the book was not obscene. The ruling is philosophically significant beyond its specific legal holding. Justice Reardon's majority opinion engaged seriously with the question of what obscenity law is for and what it can legitimately accomplish. The court applied the standard established in Memoirs v Massachusetts, which required that a work be utterly without redeeming social value to be obscene, and found that Naked Lunch met no such standard. But the more important philosophical point the case established, through the aggregate of its arguments if not through a single judicial statement, is that a work does not have to be pleasant, coherent, morally instructive or aesthetically agreeable to merit legal protection. It does not have to justify itself to the people who find it disgusting. Disgust, however sincere and however widely shared, is not a legal standard for suppression and should not be a cultural one.
Burroughs himself was explicit about this. He argued throughout his life that the function of transgressive art was precisely to transgress the reader's defenses, to get past the censoring mechanisms of polite consciousness and make contact with something more fundamental. He was not interested in art that confirmed what its audience already believed or felt. He was interested in art that disturbed, destabilized and forced a reckoning with things the culture preferred not to examine. Whether Naked Lunch succeeds in these terms is a matter of genuine debate. What is not debatable is that this is a legitimate artistic project, that the discomfort it produces is the point rather than evidence of its failure, and that the people who wanted it suppressed were not vindicated by history.
Burroughs is the avatar for this essay's central argument not because his work is representative of the AI generated content currently being condemned but because he made the philosophical case in its purest form. You do not have to like what he made. You do not have to find it meaningful or valuable or worthy of your time. The freedom that protected Naked Lunch is the freedom that protects the music video with the attractive AI generated woman that some community moderator found gratuitous. These are not equivalent works. They are covered by equivalent principles. And the principle is that your disgust, however strongly felt, does not constitute evidence of harm.
The Pinup Era and the Magazines
The moral panic around men's magazines in the 1950s is particularly instructive because it is the clearest predecessor of the current anxiety about AI generated sexual imagery and because the specific claims made about it have been so thoroughly contradicted by subsequent history.
Betty Grable's pinup photograph, taken in 1943, became the most reproduced photograph in American history during the Second World War.[41] Millions of copies were distributed to American servicemen. The image is by contemporary standards remarkably modest, a woman in a white swimsuit looking back over her shoulder. The desire it represented was not modest and everyone who distributed it knew this. The United States government distributed sexual imagery of an idealized woman to millions of young men as a deliberate morale strategy and nobody in a position of authority suggested this would corrupt them or establish a pipeline to violence against women. The men came home, married in record numbers, and produced the baby boom. The pinup as radicalization vector was not a concern that anyone thought to raise.
The launch of Playboy in December 1953 changed the terms of this conversation in ways that are still reverberating.[42] Hugh Hefner's specific achievement was to embed the sexual content, the centerfold, within a broader editorial project that included serious journalism, fiction by significant writers and cultural commentary of genuine quality. Norman Mailer, John Updike, Kurt Vonnegut, Alex Haley and Gabriel Garcia Marquez all appeared in Playboy.[43] The magazine published serious interviews with significant public figures. It made an explicit argument that sexual pleasure and intellectual cultivation were compatible parts of a complete life and that men who enjoyed both were not thereby contradicting themselves.
The argument this produced in criticism is one worth examining because a version of it recurs in every subsequent debate about sexual content. Did the intellectual content of Playboy legitimize the sexual content or did the sexual content merely use the intellectual respectability as cover? Critics who found the magazine objectionable argued the latter. Hefner argued the former. But there is a third position that neither side fully articulated: perhaps the question is wrong. Perhaps sexual content does not require legitimization through non-sexual purpose and the demand that it provide such legitimization is itself the problem. The man who bought Playboy for the centerfold and also read the interview with Martin Luther King Jr was not engaged in two separate activities, one legitimate and one requiring justification. He was reading a magazine that treated the full range of male interest as worthy of serious attention. The cultural criticism that spent decades trying to separate these elements was imposing a hierarchy of legitimate and illegitimate desire that the magazine's readers were declining to observe, which is why it sold millions of copies every month for decades.
The specific claims made against men's magazines in the 1950s and 1960s were that they would corrupt their readers, encourage promiscuity, undermine marriage and family values, and establish unrealistic expectations of women that would damage real relationships. Some version of all of these claims is still being made. The marriage rate did not collapse. The family did not dissolve. The readers of Playboy did not become incapable of functioning relationships with real women at any statistically detectable rate. The content survived. The panic did not.
Hammer, Giallo and the Slasher: When Horror Got Feminist Theory
The horror film's relationship with sexuality and violence, particularly violence against women, generated its own sustained moral panic across several decades and produced in response one of the most interesting pieces of scholarly revisionism in the history of film criticism.
Hammer Horror, the British studio that dominated horror filmmaking from the late 1950s through the 1970s, specialized in a specific combination of Gothic atmosphere, explicit violence and prominent female sexuality.[44] The Hammer films were condemned by critics and moralists for their content in terms that closely parallel current AI criticism. They were said to be exploitative, degrading to women, and likely to encourage violence in impressionable viewers. They were also enormously popular with audiences who found their combination of sensuality and horror genuinely compelling. They are now recognized as a significant body of work in British cinema.
Italian giallo, the genre of stylized murder mystery thriller that flourished from the late 1960s through the 1970s in the work of directors including Dario Argento, Mario Bava and Lucio Fulci, pushed the combination of sexuality and violence considerably further than Hammer had.[45] Giallo films are frequently and explicitly about the eroticization of violence, the relationship between the spectacle of female death and the viewer's positioning as observer, and they make no apologies for their interest in this territory. They attracted sustained critical condemnation for exactly these qualities. They are now studied in film schools, the subject of serious academic monographs, and recognized as a significant influence on mainstream Hollywood filmmaking including the work of Brian De Palma and Quentin Tarantino.
The slasher film, which synthesized elements of giallo with American horror in the late 1970s and early 1980s, was the specific target of feminist film theory's most sustained engagement with horror and produced in that engagement one of the genuinely important pieces of revisionist scholarship. Carol Clover's Men Women and Chainsaws, published in 1992, did something that most of the moral panic literature about violent sexual content has never managed: it actually watched the films carefully and thought seriously about what they were doing.[46]
Clover's central argument is that the slasher film, almost universally condemned as straightforwardly misogynistic, is in fact a genre in which audience identification is considerably more complex than the condemnation assumes. The figure she identifies as the Final Girl, the surviving female protagonist who ultimately defeats or escapes the killer, is not a passive victim but an active agent whose survival depends on her resourcefulness, intelligence and capacity for violence. Clover argues that male audiences of slasher films identify with the Final Girl in ways that cross gender lines, that the films are doing something more complicated with gender and identification than their surface content suggests, and that the feminist condemnation of the genre was based on a partial reading of what the films actually do and how audiences actually receive them.
Clover was not arguing that slasher films are feminist texts or that their treatment of violence against women is unproblematic. She was arguing that the relationship between a text, its content and its audience is more complex than moral panic analysis acknowledges and that analysis which stops at the surface of the content it is condemning is not adequate to what it is actually examining. This is a methodological point as much as a critical one and it applies with full force to the current debate about AI sexual content. The question of what a piece of content does in the world is not answered by describing what it depicts. It requires examination of how it is received, by whom, in what context, and with what actual rather than predicted consequences.
The PMRC Hearings: The Congressional Version
On September 19, 1985, the United States Senate Commerce Committee held hearings on the subject of explicit content in popular music.[47] The hearings were convened in response to a campaign by the Parents Music Resource Center, a Washington advocacy group co-founded by Tipper Gore, wife of then-Senator Al Gore, and Susan Baker, wife of Treasury Secretary James Baker. The PMRC had compiled what it called the Filthy Fifteen, a list of songs it considered most objectionable, including works by Madonna, Prince, Sheena Easton, Judas Priest, Black Sabbath and Cyndi Lauper. The organization was seeking warning labels on record albums and the co-operation of the recording industry in restricting access to explicit content.
The vocabulary used in the PMRC campaign and in the Senate hearings is worth quoting because it is almost verbatim the vocabulary currently being used about AI generated sexual content. The PMRC argued that explicit music content was a pipeline to harmful behavior in young people. It cited research claiming to demonstrate that exposure to sexually explicit and violent music lyrics influenced attitudes and behavior. It argued that the scale of the problem, the volume of explicit content available to young people, constituted a public health crisis requiring institutional response. It invoked the protection of children as the primary justification for restriction. It presented the recording industry's profiting from this content as evidence of complicity in harm.
The most significant opposition to the PMRC's position came from two witnesses whose testimony before the Senate committee has become famous. Frank Zappa appeared in a business suit and delivered a prepared statement that was precise, legally grounded and contemptuous in equal measure.[48] He argued that the PMRC's proposals constituted an unconstitutional restriction on free expression, that the research cited to support the harm claims did not demonstrate what it was said to demonstrate, and that the specific mechanism by which music was supposed to cause behavioral harm had not been adequately explained. He described the proposals as the equivalent of treating dandruff by decapitation.
Dee Snider of Twisted Sister appeared without his stage makeup, in street clothes, and delivered an extemporaneous defense of his song We're Not Gonna Take It that was both funnier and more legally sophisticated than the Senate committee appears to have anticipated.[49] He explained calmly what the song was actually about, pointed out that the PMRC's reading of it had no connection to his intentions or the song's content, and argued that the extension of the PMRC's logic would require restricting a significant portion of Western literature and music.
John Denver, whose music was not on anyone's list of harmful content, also testified in opposition to the proposed labeling scheme and made what is in some ways the most important point of the hearings: that the determination of what content is harmful is not a neutral technical judgment but a cultural and political one, that it will inevitably reflect the preferences and anxieties of the people making it, and that vesting that determination in any institutional body creates a mechanism for suppressing content that the institution finds objectionable under the cover of protecting a population that has not requested protection.
The recording industry ultimately agreed to a voluntary labeling scheme. The explicit content that prompted the hearings continued to be produced, distributed and consumed in increasing quantities. The behavioral consequences predicted by the PMRC did not materialize at any detectable scale. Tipper Gore later expressed some regret about the campaign. The music is still there.
2 Live Crew and the Class Dimension
The prosecution of 2 Live Crew for obscenity in 1990 adds a dimension to this history that most accounts of media moral panics underweight: the question of whose sexuality is being policed and by whom.[50]
2 Live Crew, a Miami bass hip hop group led by Luther Campbell, released As Nasty As They Wanna Be in 1989. The album was explicitly sexual in ways that had no precedent in commercially released popular music. It was raunchy, crude, funny in a specifically working class Black Miami idiom, and entirely unapologetic about all of these qualities. A federal district court judge in Florida ruled the album obscene in June 1990. Record store owners who sold it were arrested. Members of the group were arrested after a performance in Hollywood, Florida.
The class and race dimensions of the prosecution were noted at the time and have become clearer in retrospect. The explicit content available in mainstream media in 1990 was not limited to 2 Live Crew. Sexually explicit material had been a feature of mainstream entertainment for decades. What was different about 2 Live Crew was not the explicitness per se but the specific idiom of that explicitness, which was working class, Black, Southern, and entirely uninterested in the respectability politics that had allowed other forms of explicit content to escape prosecution. The men's magazines that had been circulating legally for decades were explicit in ways that fit within a framework of aspirational middle class masculinity. 2 Live Crew's sexuality was not aspirational. It was not tasteful. It did not provide intellectual cover for its own content. It was exactly what it appeared to be and it was not sorry about it.
Henry Louis Gates Jr testified as an expert witness for the defense and made an argument that is directly relevant to the current conversation about AI sexual content.[51] He argued that the album operated within a specific African American oral tradition, that its hyperbolic sexuality was a form of signifying, a rhetorical practice of deliberate exaggeration with deep roots in Black culture, and that judging it by the standards of mainstream white middle class propriety was a category error that failed to engage with what the work was actually doing in its own cultural context. The jury acquitted. The Eleventh Circuit Court of Appeals later overturned the obscenity finding, ruling that the district court judge had failed to apply the community standards test correctly.
The 2 Live Crew case establishes something important that the current debate about AI sexual content tends to obscure: the determination of which sexual expression is acceptable and which requires suppression is never culturally neutral. It consistently disadvantages the sexual expression of men who are already disadvantaged by class, race or cultural distance from the mainstream. The sexual content that gets legitimized is the sexual content that fits within frameworks of cultural respectability already controlled by the people doing the legitimizing. The sexual content that gets prosecuted is the sexual content that doesn't bother to seek that legitimization or can't access it. This pattern has been consistent enough across enough different instances that it should prompt genuine scrutiny of who is making the current determinations about which AI generated content crosses the line and whose cultural framework their determinations reflect.
Video Games and the Settled Science That Wasn't
The moral panic around violent video games is the most recent predecessor of the current AI content panic and the most methodologically instructive because the science marshaled against video games was presented with greater confidence and institutional backing than almost any previous media panic and has subsequently been more thoroughly discredited.
The panic intensified dramatically after the Columbine High School shootings in April 1999.[52] Eric Harris and Dylan Klebold were reported to have been players of Doom and Quake, two first person shooter games. The connection between their gaming and their violence was asserted with considerable confidence by politicians, journalists, psychologists and the families of victims. Research was rapidly produced and widely cited claiming to demonstrate that exposure to violent video games increased aggressive thoughts, aggressive feelings, aggressive behavior and decreased empathy and prosocial behavior. Craig Anderson's work was the most prominent in this regard and was cited extensively in legislative debates, legal proceedings and public discourse.[53]
The problems with this research were identified by critics at the time and have been systematically documented since. Christopher Ferguson, whose meta-analyses of the video game violence literature are the most comprehensive available, found that the effect sizes reported in individual studies dropped substantially when corrected for publication bias, that the measures of aggression used in laboratory studies were not validated against real world aggressive behavior, that the populations showing the strongest effects were those already predisposed to aggression by other factors, and that the relationship between video game consumption and actual violence in the real world was not supported by population level data.[54] As video game consumption increased exponentially through the 1990s and 2000s youth violence declined in the United States and most of the developed world, which is precisely the opposite of what the harm thesis predicted.
The Supreme Court addressed the question directly in Brown v Entertainment Merchants Association in 2011, ruling seven to two that California's law restricting the sale of violent video games to minors was unconstitutional.[55] Justice Scalia's majority opinion is worth reading in full because it engages seriously with the scientific literature and finds it wanting. The Court found that the studies purporting to show a causal link between violent video games and harmful effects on children were distinguished, as a group, by the fact that none of them proved that violent video games cause minors to act aggressively. The studies showed at best some correlation between exposure to violent video games and minuscule changes in aggressive feelings and thoughts. The Court found this insufficient to justify content restriction.
The video game panic has the same ending as every previous entry in this history. The content survived. The science that was presented as settled was not. The predicted behavioral consequences did not materialize. The games that were at the center of the panic are now played by hundreds of millions of people without incident. The researchers whose work was used to justify restriction have been challenged, in some cases quite harshly, by the subsequent literature.
The Through Line
The history traced in this section covers approximately 35,000 years of human creative production and roughly 150 years of formal legal and institutional attempts to suppress sexual and transgressive content in art and entertainment. The specific forms change. The arguments change. The technologies change. The predicted harms change in their specifics while remaining identical in their structure. The outcomes do not change at all.
In every case the science of harm was asserted with more confidence than the evidence warranted. In every case the content either survived or was later recognized as significant. In every case the predicted epidemic of behavioral harm failed to materialize at anything approaching the predicted scale. In every case the people driving the panic were more certain than history subsequently justified. In every case the determination of which content was harmful and which was acceptable reflected the cultural preferences of the people making the determination rather than any neutral technical standard.
The current panic about AI generated sexual content is the newest entry in this sequence. It has all the same features. It is being made with all the same confidence. It will almost certainly have all the same ending. The question this history poses is not whether the content will survive. It will. The question is what damage is done in the interim to the creators, communities and freedoms that the panic targets on its way to being remembered as an overreaction.
IV. What Social Constructivism Gets Right and Where It Overextends
Any serious engagement with the arguments currently being made about male sexuality and AI generated content has to begin by taking seriously the intellectual tradition those arguments draw from. Social constructivism is not a fringe position. It is not pseudoscience. It is a substantial intellectual tradition with genuine explanatory achievements and serious scholars working within it, and treating it as simply wrong would be as intellectually dishonest as treating it as simply right. What follows is an attempt to do what the current debate rarely bothers to do: examine the tradition carefully enough to distinguish what it has actually established from what it has assumed, extrapolated or wished into existence.
What Constructivism Actually Is
Social constructivism as a theoretical position holds that significant aspects of human experience, including identity, knowledge, social roles and cultural meaning, are not natural or inevitable but are produced through social processes, cultural practices, language and institutional structures. The position exists on a spectrum from weak to strong. The weak version holds that culture shapes how we understand and express aspects of experience that have biological substrates. The strong version holds that those aspects of experience are entirely products of culture with no meaningful biological substrate at all. The weak version is well supported by evidence and largely uncontroversial among serious researchers. The strong version is where the problems begin.
The genuine insights of the constructivist tradition are substantial and worth preserving. The recognition that what appears natural and inevitable in any given society is frequently the product of specific historical and cultural conditions is a genuine intellectual achievement. Before the constructivist tradition made this argument systematically it was entirely possible to treat the specific gender arrangements of mid-twentieth century Western bourgeois society as simply the way things naturally were, reflections of biological reality rather than historical contingencies. Constructivism demonstrated that these arrangements varied enormously across cultures and historical periods in ways that were very difficult to explain if they were simply expressions of fixed biological nature. This is important. The demonstration that gender roles are not simply biological destiny, that they vary across cultures in ways that suggest significant cultural determination, is not something evolutionary psychology or any serious biological account of human behavior disputes. Culture matters. Representation matters. The social environment in which people develop shapes who they become in ways that are real and measurable. These are genuine findings and they deserve to be treated as such.
The constructivist contribution to understanding racial and class identity has been similarly significant. The demonstration that racial categories are historically produced rather than naturally given, that what counts as a racial category, who falls within it and what it means has changed dramatically across time and place, is an important insight with real implications for how we understand social inequality. The analysis of how power structures reproduce themselves through cultural practices, through what Pierre Bourdieu called the habitus, the internalized dispositions that make existing social arrangements feel natural and inevitable to those who inhabit them, is a genuine contribution to understanding how social systems maintain themselves.[56]
In the specific domain of media and representation the constructivist tradition has produced real insights. The demonstration that media representations are not neutral reflections of reality but selections, framings and constructions that carry ideological implications is important. The recognition that persistent patterns in representation can influence how people understand social reality, that seeing certain groups consistently depicted in certain roles shapes expectations and assumptions, is supported by evidence. These are not trivial observations and they should not be treated as such.
Laura Mulvey and the Male Gaze
The most influential single application of constructivist thinking to the analysis of visual culture and male sexuality is Laura Mulvey's 1975 essay Visual Pleasure and Narrative Cinema, published in the journal Screen. The essay is cited so frequently in current debates about sexual imagery that it is worth examining what it actually argues rather than what it is usually said to argue.[57]
Mulvey's essay draws on psychoanalytic theory, specifically the work of Sigmund Freud and Jacques Lacan, to analyze the structure of pleasure in Hollywood cinema.[58] Her central argument is that classical Hollywood filmmaking is structured around a specifically male heterosexual viewing position. The camera, she argues, tends to adopt the perspective of a male character or a generalized male viewer. Women in classical Hollywood films function primarily as objects of spectacle rather than agents of narrative. They are displayed for visual pleasure rather than developed as subjects with their own points of view and desires. Mulvey introduces the concept of the male gaze to describe this structural feature of mainstream cinema, the way the apparatus of filmmaking, the camera angles, the editing, the lighting, the narrative structure, positions the viewer as a specifically male subject looking at a specifically female object.
This is a genuine analytical observation about a genuine feature of classical Hollywood filmmaking. Watch a film from the 1940s or 1950s with Mulvey's framework in mind and you will see what she is describing. The camera does linger on female bodies in ways it does not linger on male bodies. Female characters are frequently introduced through fragmented shots of their physical attributes before we are given access to their interiority. The narrative structure of many classical Hollywood films does position women as the object of male pursuit and desire rather than as subjects of their own stories. Mulvey identified a real pattern in a specific body of work and gave it a name and an analytical framework. That is a genuine intellectual contribution.
The framework has been productive in film studies and has generated a substantial body of subsequent scholarship. It has been applied, refined, challenged and extended by researchers including E. Ann Kaplan, Kaja Silverman and Linda Williams in ways that have produced genuine critical insight into how visual culture operates.[59] The male gaze as an analytical tool has illuminated real things about how mainstream visual culture has historically been structured.
Where the framework becomes problematic is in the move from descriptive to prescriptive, from analytical tool to moral verdict, and in its extension from the specific historical body of work Mulvey analyzed to all visual culture in all contexts. Mulvey was analyzing classical Hollywood films of the 1940s and 1950s, produced under a specific studio system by a specific industrial apparatus at a specific historical moment. Her analysis of that specific body of work is defensible and has been largely defended by subsequent scholarship. The extension of her framework to cover all instances of male visual attraction to female imagery, to assert that any representation of a woman from a position of male visual pleasure reproduces the oppressive structure she identified in Rear Window and Vertigo, is an extension that the original argument does not support and that Mulvey herself has been more cautious about than many of her followers.
Mulvey's Own Qualifications
This is a point worth dwelling on because it is almost never mentioned when Mulvey's framework is deployed in contemporary debates. In subsequent work, including her 1981 essay Afterthoughts on Visual Pleasure and Narrative Cinema, Mulvey substantially complicated and qualified her original argument.[60] She acknowledged that female spectatorship, the experience of women watching classical Hollywood films, could not be adequately described by the male gaze framework as originally stated. She acknowledged that identification in cinema is more complex and mobile than the original essay suggested, that viewers do not simply adopt the viewing position the apparatus offers them but negotiate with it in ways that vary with gender, sexuality, race and cultural background. She acknowledged that the psychoanalytic framework she drew on had limitations as a tool for analyzing real spectatorship.
These qualifications matter because they come from the theorist herself and because they significantly limit the application of the original framework. A theory that its own author has substantially qualified in light of criticism and subsequent evidence should be deployed with some acknowledgment of those qualifications. In contemporary debates about AI sexual content the male gaze theory tends to be deployed in its original 1975 form, without the qualifications Mulvey added, without the substantial feminist criticism the theory has received, and without any acknowledgment that its application to contexts radically different from 1940s Hollywood cinema requires justification rather than assumption.
Where Constructivism Overextends: The Tabula Rasa Problem
The move from the weak constructivist position, that culture shapes sexual behavior and attitudes, to the strong constructivist position, that sexuality is entirely a product of cultural conditioning, requires a claim about human nature that the evidence does not support. It requires, in effect, the assumption that the human mind begins as a blank slate upon which culture writes its contents without constraint from biological nature, and that therefore the contents of mind, including sexual desire, are fully explicable through the cultural inputs that produced them.
The blank slate is not a new idea. John Locke articulated its philosophical foundations in An Essay Concerning Human Understanding in 1689, arguing that the mind begins as a tabula rasa, a blank tablet, and that all knowledge derives from experience.[61] Locke was making an epistemological argument about the origins of knowledge rather than a psychological claim about the structure of mind, but the idea was adopted into social theory in its stronger form as the claim that human nature is essentially plastic, that what appears fixed is in fact culturally produced, and that changing the culture will change the person.
Steven Pinker's The Blank Slate, published in 2002, is the most comprehensive scholarly engagement with this assumption and its consequences.[62] Pinker's argument is that the blank slate, along with what he calls the Noble Savage and the Ghost in the Machine, constitutes one of three connected doctrines that have shaped social science and humanities scholarship in ways that have increasingly conflicted with evidence from genetics, evolutionary biology, developmental psychology and cognitive neuroscience. He does not argue that culture does not matter. He argues that culture operates on a human nature that is not blank, that the mind comes equipped with structures, dispositions and preferences that constrain and shape what culture can do with it, and that ignoring these structures produces systematic errors in both description and prescription.
The Blank Slate generated substantial controversy when it was published and the controversy has not fully resolved. Some of Pinker's specific claims have been challenged. His characterization of his opponents is sometimes uncharitable. The book is a work of synthesis and advocacy as well as scholarship and should be read with appropriate critical attention. What it establishes, even accounting for these limitations, is that the strong blank slate position, the claim that human nature is infinitely plastic and that all significant variation in human behavior and psychology is culturally produced, is not supported by the available evidence and that the scientific consensus in relevant fields has moved substantially away from it.
The Cross Cultural Universal Problem
The most direct evidentiary challenge to the strong constructivist account of male sexuality is the evidence from cross cultural research. If male visual sexuality, the orientation toward female physical attractiveness as a primary driver of desire, the tendency to respond to visual cues of fertility and youth, the separation of desire from emotional attachment more readily than female sexuality, are products of Western patriarchal culture, then we should expect to find substantial variation in these features across cultures with different gender ideologies. We do not find this.
The cross cultural evidence on male sexual psychology is extensive and remarkably consistent. Studies conducted in societies ranging from highly gender egalitarian Nordic countries through traditional patriarchal societies in the Middle East and Africa to indigenous communities with minimal contact with Western media find consistent patterns in male visual sexual response that are very difficult to explain as products of a specific cultural construction. The specific aesthetic standards of beauty vary across cultures in ways that reflect local cultural preferences. The basic structure of male visual sexual response, the orientation toward physical youth and fertility signals as primary attractants, does not vary in ways that track cultural ideology.
This presents a genuine problem for the strong constructivist position that it has not fully resolved. The standard response is to argue that all of the cultures studied have been influenced by patriarchal structures even when those structures take different forms, and that what looks like a cross cultural universal is in fact a cross cultural constant in the patriarchal conditioning of male desire. This is not an impossible argument but it is an increasingly strained one as the evidence accumulates and as the cultural variation between the societies studied becomes more extreme. At some point the argument that a psychological feature is culturally constructed despite appearing consistently in cultures with radically different and sometimes opposing cultural values requires more explanatory apparatus than the constructivist tradition has provided.
The Circular Evidence Problem
There is a methodological problem in constructivist research on sexual media and attitudes that does not receive adequate attention in the debates that cite this research. Many of the studies claiming to demonstrate that sexual media consumption produces objectifying attitudes toward women have built their conclusions into their methodology in ways that make the findings circular.
The typical structure of this research involves exposing subjects to sexual media content, measuring their subsequent attitudes toward women using validated scales, and finding that the exposed group shows more objectifying attitudes than a control group. The problem begins with the coding of the media content. Studies in this tradition frequently code male visual sexual interest in female appearance as inherently objectifying. A study that defines the male gaze as objectification by definition and then finds that exposure to content featuring the male gaze increases objectifying attitudes is not discovering a causal relationship. It is restating its own definitions.
The measurement instruments used to assess objectifying attitudes compound this problem. Scales that measure objectifying attitudes toward women often include items that conflate the assessment of physical attractiveness, which is a universal feature of human social cognition applied across genders, with the denial of personhood, which is what objectification in the morally relevant sense actually involves. A subject who rates female physical appearance as important in sexual attraction scores higher on objectification scales designed this way regardless of whether they also attribute full personhood, agency and moral status to women. The measurement instrument cannot distinguish between finding women attractive and treating women as objects, which means the research using it cannot make that distinction either.
These methodological problems have been identified in the literature by researchers including Christopher Ferguson, whose critiques of the media effects research tradition are among the most thorough available.[63] They are not resolved by the volume of studies that share the same methodological problems, because sharing a methodological flaw across many studies does not eliminate the flaw. They matter because the policy and social conclusions being drawn from this research, conclusions about which content should be restricted and which creators should be characterized as harmful, are being based on evidentiary foundations that are considerably weaker than the confidence with which they are presented.
What Constructivism Cannot Explain
There are specific features of male sexual psychology that the constructivist framework, even in its more sophisticated versions, has not adequately explained and that a complete account of sexuality must address.
The consistency of male visual sexuality across cultures has already been discussed. The developmental evidence presents an additional challenge. Sexual orientation and the basic structure of sexual preference emerge in development before the period of cultural exposure that constructivism would identify as the primary shaping influence. Twin studies consistently find that sexual orientation has substantial heritability, meaning that genetic factors contribute significantly to its development in ways that cannot be fully explained by shared cultural environment.[64] The developmental trajectory of sexual preference, its emergence in early adolescence in response to biological changes that precede significant sexual experience, is very difficult to account for in terms of cultural conditioning alone.
The neuroscience of sexual response presents further challenges. The neural systems involved in sexual attraction and arousal show characteristic patterns that are consistent across individuals in ways that suggest a structured biological substrate rather than a culturally written blank slate.[65] The specific patterns of activation in response to visual sexual stimuli differ between male and female subjects in ways that are consistent with evolutionary predictions and inconsistent with the prediction that all significant variation in sexual response is culturally produced.
None of this means that culture does not influence sexuality. It clearly does. The specific forms that sexual desire takes, the aesthetic preferences that attract or repel, the social meanings attached to different kinds of desire, the behavioral expression of sexual interest, all of these are substantially shaped by cultural context. The question is not whether culture matters but whether culture is all that matters, whether the strong constructivist claim that sexuality is entirely a cultural product is supported by the evidence. It is not.
The Strongest Counterarguments, Honestly Stated
Intellectual honesty requires engaging with the most serious challenges to the evolutionary and biological account of male sexuality rather than dismissing them. Cordelia Fine's work, particularly Delusions of Gender published in 2010 and Testosterone Rex published in 2017, represents the most sophisticated recent challenge to strong biological determinism about sex differences and deserves serious engagement rather than dismissal.[66]
Fine's central argument in Delusions of Gender is that many of the claimed scientific findings about sex differences in brain structure and function are based on methodologically flawed research that has been overinterpreted in ways that confirm existing gender stereotypes rather than revealing genuine biological differences. She documents cases where small, inconsistently replicated findings have been reported as established facts, where the popular science account of brain research diverges substantially from what the research actually shows, and where the framing of studies reflects assumptions about gender that contaminate the findings. These are legitimate methodological criticisms and they should be taken seriously.
In Testosterone Rex Fine challenges the popular account of testosterone as the biological driver of characteristically male behavior, including sexual aggression and risk taking, arguing that the relationship between testosterone and behavior is far more complex, context dependent and bidirectional than the simple story suggests. Again these are legitimate and important points. The popularized account of testosterone as the simple explanation for male sexuality and behavior is indeed an oversimplification that the research does not support.
What Fine's work establishes is that specific strong claims about the biological determination of sex differences are often better supported than their popular presentation suggests but less well supported than their proponents claim. This is a useful corrective to overconfident biological determinism. What it does not establish is that the cross cultural universals in male sexual psychology are explained entirely by cultural factors, that the developmental and genetic evidence for biological contributions to sexual orientation and preference is an artifact of methodological problems, or that the strong constructivist position is vindicated by the failure of the strong biological determinist position. Both strong positions overreach their evidence. The honest conclusion is that male sexuality has both biological substrates and cultural shapers and that the current state of the evidence does not support confident claims about the precise weighting of these factors in every domain.
The Conclusion of This Section
Social constructivism has made genuine and important contributions to our understanding of gender, sexuality and the role of representation in shaping social reality. These contributions should not be dismissed. The insight that culture shapes behavior, that representation influences attitude, and that power structures reproduce themselves through cultural practices is well supported and important.
What the constructivist tradition has not established, despite the confidence with which it is sometimes deployed, is that male sexuality is entirely a product of cultural conditioning, that there is no biological substrate to male visual sexual response that precedes and constrains cultural influence, or that the idealized female imagery produced by and for male audiences across tens of thousands of years of human history is adequately explained as a symptom of a culturally constructed power system rather than as an expression of a biological feature of human male psychology that culture shapes but does not create.
The strong constructivist position on male sexuality requires a blank slate that developmental evidence does not support, predicts cross cultural variation that cross cultural evidence does not find, and generates research with methodological problems that its practitioners have not adequately addressed. Taking the genuine contributions of the tradition seriously means also taking seriously the limits of what it has actually demonstrated. Those limits matter enormously when the tradition is being used to justify characterizing the sexual expression of a majority of men as a pipeline to violence. That is a significant claim. The evidence behind it is not.
V. What Evolutionary Psychology Actually Says
Evolutionary psychology is one of the most frequently misrepresented fields in contemporary public discourse and it is misrepresented from both directions. Its enthusiasts sometimes deploy it as a trump card that settles debates about human behavior by appeal to natural selection, as though demonstrating an evolutionary explanation for a trait resolves all questions about its moral status or social implications. Its critics frequently attack a caricature of the field, a collection of simplistic just so stories about cavemen and their preferences, rather than engaging with the actual methodology and findings of the discipline as it is practiced by serious researchers. Both misrepresentations serve the interests of people who want simple answers to complicated questions. Neither serves the interests of understanding what the field has actually established.
What follows is an attempt to describe evolutionary psychology as it actually exists rather than as it is typically deployed in arguments for or against specific positions on male sexuality. This requires some methodological groundwork before the specific findings, because the findings only make sense in the context of how the field operates and what kinds of claims it can and cannot legitimately make.
What Evolutionary Psychology Is and Is Not
Evolutionary psychology proceeds from the observation that the human mind, like the human body, is a product of natural selection operating over millions of years of evolutionary history. Just as the structure of the human hand reflects the selective pressures of our evolutionary past, the structure of human psychological mechanisms reflects the problems our ancestors faced and solved during the period of evolutionary adaptedness, roughly the Pleistocene epoch from approximately 2.5 million to 12,000 years ago.[67] The field attempts to identify psychological mechanisms that are plausibly the products of natural selection, to specify the adaptive problems those mechanisms would have solved, and to test predictions derived from these hypotheses against empirical evidence.
Several things need to be clear about this methodology before proceeding. First, evolutionary psychology does not claim that evolved psychological mechanisms are morally justified because they are natural. The naturalistic fallacy, the inference from is to ought, is a logical error that serious evolutionary psychologists are generally careful to avoid even if popular accounts of their work frequently commit it. The fact that a psychological tendency has an evolutionary explanation does not make it good, desirable or exempt from moral evaluation. Second, evolutionary psychology does not claim that evolved mechanisms are immutable or that behavior is biologically determined in ways that preclude change. Evolved mechanisms interact with environmental inputs, cultural contexts and individual variation in complex ways. The existence of an evolved basis for a behavioral tendency does not mean that tendency cannot be influenced, moderated or redirected by experience and culture. Third, the field does not claim that all human behavior is adaptive or that everything people do can be explained by natural selection. Evolution produces mechanisms, not behaviors, and the behaviors those mechanisms produce depend on the environments in which they operate.
The legitimate methodological criticisms of evolutionary psychology focus primarily on the problem of generating hypotheses that are sufficiently specific to be falsifiable. A hypothesis that begins with an observed behavior, constructs a plausible evolutionary story about why that behavior would have been adaptive, and then treats the plausibility of the story as confirmation of the hypothesis is not doing science in any meaningful sense. This is the just so story problem, named after Rudyard Kipling's fanciful tales of how animals got their characteristics, and it is a real problem in some evolutionary psychology research.[68] A plausible story about why a trait might have been adaptive is not evidence that it is an adaptation. Independent evidence beyond the plausibility of the adaptive story is required and not all evolutionary psychology research provides it.
Keeping both the genuine contributions and the legitimate limitations in view, here is what the field has established about male sexuality.
A distinction needs to be drawn before the findings are presented. This essay will use the narrower phrase male visual sexual response to refer to the specific cross-cultural pattern the research has documented: the orientation of male sexual attention toward physical features that predicted reproductive value in ancestral environments. This is a narrower category than male sexuality in the broader sense, which includes behaviors, practices and cultural expressions whose universality across cultures is not claimed here and which are plainly shaped in part by culture. The argument that follows defends the narrower category as a biologically grounded feature of human psychology. It does not defend everything that has ever travelled under the broader name, and it does not require that any specific cultural expression of male sexuality is itself beyond critical examination.
David Buss and the Cross Cultural Evidence
David Buss's 1989 study of mate preferences across 37 cultures, published in Behavioral and Brain Sciences, is the most extensive cross cultural investigation of human mate choice ever conducted and remains the foundational dataset for evolutionary accounts of human sexual psychology.[69] Buss and his collaborators surveyed over 10,000 individuals across 37 cultures on six continents, covering an extraordinary range of cultural, economic, religious and political contexts, asking about preferences in long term mates across a range of dimensions including physical attractiveness, earning potential, ambition, chastity and age.
The findings revealed a pattern of both cross cultural universals and cross cultural variation. The universals are the findings most relevant to the present argument. Across all 37 cultures men placed significantly higher value on the physical attractiveness of potential mates than women did. Across all 37 cultures men preferred mates younger than themselves while women preferred mates older than themselves, a preference that Buss interpreted as reflecting male sensitivity to cues of female reproductive value. Across all 37 cultures women placed significantly higher value on the resource acquisition potential of potential mates than men did, preferring mates with higher earning capacity, ambition and social status.
These cross cultural universals are not trivially explained by cultural transmission. The cultures in the study differed enormously in their gender ideologies, their economic systems, their religious frameworks and their attitudes toward sexuality. The Nordic countries in the sample had substantially more gender egalitarian ideologies than the traditional societies in the Middle East and Africa. The communist societies included had explicitly ideological commitments to gender equality that differed radically from the capitalist societies. Despite these enormous cultural differences, the basic pattern of sex differences in mate preference remained consistent. The magnitude of the differences varied across cultures in ways that are themselves informative, with some of the strongest sex differences in mate preference appearing in the most gender egalitarian societies, a finding that is difficult to explain on a purely cultural constructivist account, but the direction of the differences did not reverse in any culture studied.[70]
What the Buss study does and does not establish requires careful statement. It establishes that sex differences in mate preference are cross culturally consistent in ways that suggest a biological substrate rather than purely cultural determination. It does not establish that these preferences are immutable, that they fully determine behavior, that they justify social arrangements that disadvantage women, or that men who find female physical attractiveness important in sexual attraction are thereby exhibiting a pathological or culturally conditioned attitude. Male visual attraction to female physical attributes that serve as cues of reproductive value is, on the evidence of Buss's study and the substantial literature that has replicated and extended it, a feature of human male psychology that exists across cultures with radically different ideologies. It is not a product of Western patriarchal conditioning. It precedes the specific cultural forms in which it is expressed.
Male Visual Sexuality as a Cross-Cultural Universal
The specific features of female physical appearance that male subjects across cultures find attractive are informative about the likely evolutionary function of male visual sexuality. Research across multiple methodologies including survey studies, eye tracking studies and neuroimaging studies has identified a consistent set of features that predict male ratings of female attractiveness across cultures: bilateral facial symmetry, a waist to hip ratio in the range of 0.67 to 0.80, clear skin, lustrous hair, and age related features that indicate youth and reproductive maturity.[71] These are not arbitrary aesthetic preferences. Each of them correlates with features of physical health, genetic quality and reproductive potential that would have been relevant to mate selection in ancestral environments.
Facial symmetry is a reliable indicator of developmental stability and resistance to pathogens. The waist to hip ratio associated with female attractiveness corresponds to the hormonal profile associated with female reproductive health and fertility.[72] Clear skin and lustrous hair are indicators of nutritional status and freedom from parasitic infection. The age-related features preferred by men in cross cultural studies correspond to the peak of female reproductive potential.
The evolutionary interpretation of this pattern is straightforward: male visual sexuality is a mechanism that orients male mating effort toward females with high reproductive potential, using physical cues that reliably predicted reproductive value in ancestral environments. This is not a culturally variable preference. It is a design feature of the male sexual psychology that natural selection produced.
This has direct implications for the current debate about AI generated sexual imagery. The idealized female figures that appear in AI generated content, young, symmetrical, with the body proportions associated with female health and fertility, are not products of a specific patriarchal culture's construction of femininity. They are outputs of a male visual system responding to the features it was designed by natural selection to find attractive. The man generating this content is not reproducing a harmful cultural ideology. He is expressing a psychological mechanism that exists across cultures and that predates by millions of years the specific cultural forms in which it currently manifests. This does not make every expression of that mechanism morally equivalent or exempt from any evaluation. It does mean that treating the mechanism itself as a symptom of cultural pathology is a category error that the evidence does not support.
Erotic Plasticity and the Baumeister Research
Roy Baumeister's research on what he calls erotic plasticity addresses one of the most frequently misunderstood features of male sexuality: the relative separation of male desire from emotional context, relationship quality and situational factors compared to female sexuality.[73] Baumeister's work, summarized in his 2000 paper in Psychological Bulletin and extended in subsequent publications, argues that female sexuality is characterized by higher erotic plasticity than male sexuality, meaning that female sexual desire is more responsive to social, cultural and situational factors, more variable across time and context, and more shaped by relationship quality and emotional connection than male sexuality tends to be.
This difference, Baumeister argues, is not a moral failing on either side. It is a documented feature of human sexual psychology that has plausible evolutionary explanations. Female reproductive investment is substantially greater than male reproductive investment: pregnancy, nursing and primary infant care involve enormous costs that male biology does not share to the same degree. This asymmetry in reproductive investment creates different optimal mating strategies for males and females. For females, the quality of the specific partner and the resources and commitment he can provide matter enormously to reproductive success. For males, partner quality in the sense of fertility and health matters more than partner specific commitment in many contexts. The result is that female sexuality evolved to be more context sensitive and relationship dependent while male sexuality evolved to be more responsive to physical cues of fertility regardless of relationship context.
The manifestation of this difference in contemporary human psychology includes the pattern that critics of male sexuality consistently misread as pathological: the capacity for male sexual desire that is not dependent on emotional intimacy, the responsiveness to visual sexual stimuli without accompanying relational context, the tendency to experience sexual fantasy about people with whom no relationship exists or is desired. These are not failures of male emotional development. They are features of a sexual psychology that evolved under selection pressures different from those that shaped female sexuality. Treating them as symptoms of a disordered or dangerous masculinity requires either ignoring the evolutionary account of why they exist or arguing that the evolutionary account is wrong, which requires engaging with the cross-cultural evidence that the evolutionary account explains better than any alternative.
Baumeister's research has been critiqued on methodological grounds and some of his specific claims have been contested. The general finding of sex differences in erotic plasticity has been replicated across multiple methodologies and remains one of the more robust findings in the sex differences literature. It matters for the current debate because it provides a framework for understanding why male visual sexuality, including the consumption of sexual imagery featuring attractive women with whom the consumer has no relationship, is a normal expression of a well-documented feature of male sexual psychology rather than evidence of an inability to relate to women as full human beings.
Fantasy as Psychological Function
The claim that sexual fantasy about idealized figures constitutes a rehearsal for harmful behavior, that the man who imagines sexual scenarios is in some sense practicing them and will eventually enact them, is central to the pipeline argument about AI generated sexual content. It is also, on the available evidence, wrong.
The psychology of sexual fantasy has been studied extensively since Alfred Kinsey's foundational surveys in the 1940s and 1950s.[74] The research consistently finds that sexual fantasy is nearly universal in both men and women, that the content of fantasy frequently involves scenarios the fantasist has no desire to enact in reality, and that the relationship between fantasy content and behavior is weak and largely mediated by other factors. Studies of sexual fantasy content find that fantasies about non-consensual scenarios are among the most common in both male and female subjects, while the rates of actual sexual coercion do not correspond to the prevalence of such fantasies in any simple way.[75] The existence of a fantasy does not predict the behavior it depicts.
The functional account of sexual fantasy is more coherent than the rehearsal account. Fantasy serves psychological functions that are distinct from behavioral planning. It provides a safe context for the exploration of desire without the constraints and consequences of real-world behavior. It allows the processing of sexual interest in forms that social reality does not make available. It provides arousal in contexts where real world sexual experience is absent. These functions are served precisely by the fact that fantasy is not reality, that the imagined scenario exists in a space separate from behavior where normal social and ethical constraints do not apply. The freedom of fantasy is the source of its psychological utility, not evidence of its danger.
Brad Sagarin and his colleagues have examined the relationship between sexual fantasy content and real-world attitudes and behavior and found that the content of fantasy is a poor predictor of real-world behavior in the direction the harm thesis assumes.[76] People who fantasize about scenarios they would never enact are not unusual. They are the norm. The psychological function served by fantasy about scenarios that exceed the bounds of real-world desire or ethics is precisely to contain that excess in a form that does not translate into action. Treating this containment function as a pipeline to behavior it is actually containing is a significant misunderstanding of what fantasy does.
Milton Diamond and the Pornography Question
The claim that pornography consumption causes sexual violence is the most persistent and most consequential empirical claim in debates about sexual media and one of the most thoroughly examined. It is also one where the evidence most consistently contradicts the harm thesis in ways that have not been adequately acknowledged in cultural criticism.
Milton Diamond, a researcher at the University of Hawaii who has studied the relationship between pornography availability and sexual violence rates for several decades, has published findings across multiple countries that consistently contradict the prediction that increased pornography availability leads to increased sexual violence.[77] His 1999 paper examining the relationship in Japan, published in Archives of Sexual Behavior, found that rates of sexual violence declined substantially over the period during which pornography availability increased dramatically following the relaxation of obscenity laws. A subsequent study examining the Czech Republic found a similar pattern: the legalization and widespread availability of pornography following the end of communist era censorship was accompanied by a decrease rather than an increase in sexual assault rates.[78]
Diamond's findings have been replicated in other national contexts. A 2009 paper by Anthony D'Amato examining rape statistics in the United States found that states with higher internet access, and therefore higher access to online pornography, showed lower rates of rape than states with lower internet access, a finding that is the opposite of what the harm thesis predicts.[79] A 2010 paper by Berl Kutchinsky, reviewing evidence across four countries including the United States, West Germany, Denmark and Sweden, found that sexual offense rates did not increase following the legalization of pornography in any of the countries studied and decreased in several.[80]
These findings have not received the attention in public debate that studies claiming to find harmful effects have received. The reasons for this asymmetry are worth noting. Studies finding harmful effects of pornography are more likely to be published than studies finding null or contrary effects, a manifestation of the publication bias that affects all empirical research but has particular influence in areas where the research findings have policy implications and where the researchers frequently share the moral concerns of the advocates citing their work.[81] Studies finding that pornography availability correlates with decreases in sexual violence contradict a widely held assumption and face a higher bar of skepticism in review and a lower likelihood of prominent coverage in media that has already committed to a specific narrative about sexual media and harm.
Diamond's work and the research that corroborates it does not establish that pornography causes decreases in sexual violence, which would be a causal claim that the correlational evidence cannot support. What it establishes is that the simple harm thesis, that pornography consumption leads to increased sexual violence at a population level, is not supported by the available epidemiological evidence and is in many cases directly contradicted by it. A theory that predicts X when the evidence shows not-X needs to account for why it got it wrong before it is used to justify social and legal policy.
Michael Seto and the Population Question
Michael Seto's research on sexual offending and sexual media consumption addresses a conflation that is central to the current AI content panic and that significantly distorts the terms of the debate. The conflation is between the population of typical consumers of sexual media and the population of individuals at elevated risk of sexual offending, treating findings about the latter as though they applied to the former.
Seto, whose research on pedophilia, sexual offending and the use of sexual media by sex offenders is among the most rigorous in the field, has consistently argued that the relationship between sexual media consumption and sexual offending is not a general relationship that applies across the population of consumers but a specific relationship that applies primarily to individuals who have pre-existing vulnerabilities or pathologies. His 2010 paper in Psychological Bulletin on pedophilia and sexual offending found that the use of child sexual abuse material by sex offenders against children is better understood as a symptom of pre-existing pedophilic interest than as a cause of offending behavior.[82] The offenders were not made into offenders by the material. They sought the material because they were already offenders in orientation if not yet in behavior.
The generalization of this finding to typical sexual media consumers, the extrapolation from the finding that sexual media use is associated with offending in populations already predisposed to offend to the conclusion that sexual media use in general populations produces offending behavior, is a logical error of the type Seto's work specifically warns against. The population of men who consume sexual media is vast, encompassing the large majority of adult males in societies with significant internet access. The population of sexual offenders is a small fraction of that group. The research on sexual offenders cannot be used to characterize the larger population without evidence that the relationship found in clinical populations generalizes to non-clinical ones, and the epidemiological evidence already discussed suggests that it does not.
This matters for the current debate because arguments about the harmfulness of AI generated sexual content frequently move between findings about clinical populations and claims about typical consumers without acknowledging the logical gap between them. The existence of men who use sexual media in the context of broader patterns of harmful behavior toward women does not establish that sexual media consumption is causally related to that behavior in the general population. The base rate of harmful behavior in the general population of sexual media consumers, which is vastly lower than the harm thesis implies, is the relevant evidence for claims about what typical consumption does and does not lead to.
The Adaptationist Storytelling Problem Honestly Addressed
The legitimate methodological critique of evolutionary psychology deserves direct engagement rather than dismissal. The just so story problem is real. The field has produced hypotheses about the adaptive basis of specific behavioral tendencies that are more plausible as narratives than they are supported as science. The generation of an adaptive story that fits an observed pattern is not the same as demonstrating that the pattern is an adaptation, and some evolutionary psychology research has been insufficiently careful about this distinction.
Cordelia Fine's Testosterone Rex makes this point with particular force in the domain of testosterone and behavior. The popular account of testosterone as the biological driver of male aggression, risk taking and sexual behavior is, Fine argues, a significant oversimplification of a relationship that the research shows to be far more complex, context dependent and bidirectional. Testosterone levels respond to social situations rather than simply causing behavior. The relationship between testosterone and aggression in particular is far less robust than popular accounts suggest and is moderated by a large number of contextual factors. Fine's critique of testosterone determinism is well supported and important.
What Fine's critique establishes is that specific strong claims about hormonal determination of male behavior are oversimplified. What it does not establish is that the cross-cultural patterns in male sexual psychology documented by Buss and others are artifacts of methodological problems or cultural contamination. The critique of specific overreaching claims in evolutionary psychology does not extend to a general refutation of the field's findings, and the findings most relevant to the current discussion, the cross cultural consistency of male visual sexual response, the sex differences in erotic plasticity documented by Baumeister, the population level epidemiology of sexual media and sexual violence documented by Diamond and others, are not primarily dependent on the specific claims that Fine's critique targets.
The honest position is that evolutionary psychology has established some things with reasonable confidence and has overreached in other areas, and that distinguishing between these requires engaging with the specific claims and their specific evidentiary support rather than accepting or rejecting the field wholesale. The specific claims most relevant to the current debate are among the better supported ones: male visual sexuality has a biological substrate, it is cross culturally consistent in ways that precede specific cultural formation, and its expression in fantasy and sexual media consumption does not reliably predict harmful behavior toward real women in general populations.
What the Field Establishes
Setting aside the areas of genuine uncertainty and legitimate critique, here is what the evolutionary psychology literature establishes with reasonable confidence about male sexuality and why it matters for the current debate.
Male visual sexuality, the orientation toward female physical attractiveness as a primary driver of sexual interest, is a cross culturally consistent feature of human male psychology that appears in cultures with radically different gender ideologies and that cannot be adequately explained as a product of any specific cultural construction. It has a plausible evolutionary explanation in the adaptive value of sensitivity to fertility cues during the period of human evolutionary history. It manifests in the contemporary world through attraction to idealized female figures, through sexual fantasy involving physical desirability, and through the consumption of sexual imagery featuring attractive women.
None of this constitutes a moral endorsement of every form that male visual sexuality takes in the contemporary world. Some expressions of male sexuality cause real harm and deserve real condemnation. Nonconsensual use of real people's images causes harm. Sexual coercion causes harm. The existence of a biological substrate for male visual attraction does not justify these behaviors and no serious evolutionary psychologist has argued that it does.
What the biological substrate does establish is that treating male visual sexuality itself as a symptom of cultural pathology is an error. The idealized female figure in the AI generated music video is not evidence of a worldview that reduces women to objects. She is evidence of a male visual system responding to the features it was designed by millions of years of natural selection to find attractive, expressed through the most recent available technology. The man who made her is not three steps from violence. He is doing what men across every culture and every historical period have done when given access to the tools to do it: making a representation of female beauty because making representations of female beauty is something humans do.
The science does not support this panic's generalization from specific documented harms to a general theory of male sexuality. It never has, in this form. The history reviewed in the previous sections shows what happens when a panic proceeds regardless of what the science supports. This one will be no different.
VI. The Objectification Question
There is a difference between imagining a person as an object and treating a person as one. This distinction is not subtle. It is not a technicality deployed to evade a serious argument. It is the foundational distinction that any honest discussion of objectification in art and fantasy has to make before it makes any other claim, because the entire weight of the objectification argument as currently deployed against sexual imagery rests on the assumption that these two things are continuous, that the imagination of objectification produces the behavior of objectification, and that assumption is where the argument loses its evidentiary grounding and does not recover it.
The man who finds an AI generated female figure attractive is engaged in an act of imagination. The figure is not a person. She has no inner life to deny, no autonomy to violate, no body integrity to disrespect. She is a visual artifact produced by a generative system responding to a prompt. The moral question that actually matters is not whether the imagination treats her as an object, because she is not a subject, but whether the act of imagining contributes causally to treating real women as objects in contexts where that treatment constitutes genuine harm. That is an empirical question. It requires empirical evidence. The evidence, examined carefully rather than selectively, does not establish the causal chain the objectification argument assumes.
Nussbaum's Framework Stated Fully and Fairly
Martha Nussbaum's 1995 paper Objectification, published in Philosophy and Public Affairs, is the most philosophically rigorous account of what objectification means and why it matters, and it deserves to be engaged with in its actual form rather than in the simplified version that circulates in cultural criticism.[83] Nussbaum identifies seven features that characterize the treatment of a person as an object, noting that these features can operate independently and that their moral significance depends on context and combination.
The first feature is instrumentality: treating a person as a tool for one's own purposes. The second is denial of autonomy: treating a person as lacking in self-determination and as subject to the will of another. The third is inertness: treating a person as lacking in agency and activity. The fourth is fungibility: treating a person as interchangeable with other objects of the same type, as replaceable. The fifth is violability: treating a person as lacking in boundary integrity, as something that can be broken into and used. The sixth is ownership: treating a person as something that can be bought, sold and owned. The seventh is denial of subjectivity: treating a person as something whose experiences and feelings need not be taken into account.
This is a careful and genuinely useful analytical framework for identifying what is morally wrong about specific kinds of treatment of real people. Each of these features describes a way of failing to extend to another person the moral consideration they are owed as a subject with their own perspective, desires and interests. When these features operate together and systematically in contexts involving real people the resulting treatment constitutes a serious moral failure that Nussbaum is right to identify and name.
What is almost universally omitted when Nussbaum's framework gets deployed in cultural criticism is the qualification she herself placed on it in the same paper. Nussbaum was explicit that objectification is not always harmful, that context determines its moral weight, and that elements of objectification can be present in interactions that are not morally wrong. She offers as examples the mutual objectification that can occur in consensual sexual relationships, where partners temporarily treat each other as sources of physical pleasure in ways that involve some of the features on her list without involving the denial of personhood that constitutes genuine moral wrong.[84] She argues that what matters is whether the objectification operates within a broader context of recognition of the other person's full humanity or whether it substitutes for that recognition entirely.
This qualification is not a minor caveat. It is central to how the framework functions as an analytical tool rather than a blunt instrument. A framework that identifies objectification as always harmful regardless of context is not Nussbaum's framework. It is a simplified version of it that has been stripped of its most important philosophical complexity in order to serve as a more effective rhetorical weapon. The restoration of the caveat changes the terms of the argument substantially. The question is no longer whether sexual imagery involves elements of objectification, which it obviously does by Nussbaum's definition, but whether the context in which it operates, fantasy, art, entertainment produced and consumed by consenting adults, is one in which the objectification constitutes the kind of systematic denial of personhood that generates genuine moral harm.
The Context of Fantasy
Nussbaum's own framework, properly understood, suggests that the context of fantasy is precisely one in which objectification in the sense of focusing on physical attributes for the purpose of sexual arousal does not constitute the morally serious wrong her framework is designed to identify. The generated female figure in a music video has no inner life to deny because she has no inner life. She is not a person being treated as an object. She is an image being engaged with as an image. The application of Nussbaum's framework to generated imagery requires first treating the generated figure as a person whose subjectivity is being denied, which begs the very question at issue.
The extension of the objectification framework from real people to fictional or generated figures requires an argument that the attitude practiced toward the fictional figure translates into attitude or behavior toward real people. This is the causal claim that the framework depends on and it is the claim that requires empirical support. Nussbaum's philosophical analysis of objectification does not itself provide that support. It identifies what objectification is and why it matters in contexts involving real people. The question of whether engaging with objectifying content in fantasy contexts produces objectifying treatment of real people in behavioral contexts is a psychological and empirical question that philosophical analysis alone cannot answer.
The Contested Evidence Landscape
The research literature on the relationship between sexual media consumption and attitudes or behavior toward real women is genuinely contested and the confidence with which it is cited in cultural criticism does not reflect the state of the actual evidence. Understanding why requires examining what the research typically measures and what those measures do and do not establish.
The dominant methodology in this research area involves exposing subjects to sexual media content, measuring their subsequent responses using standardized instruments, and comparing these responses to those of control groups who were not exposed to the material. The measures used include self-report scales assessing attitudes toward women, laboratory measures of aggressive behavior, and instruments designed to measure rape myth acceptance, the endorsement of false beliefs about sexual violence that minimize victim responsibility and normalize offender behavior.[85]
Studies using this methodology have produced findings that are frequently reported as demonstrating that sexual media consumption causes harmful attitudes and aggressive behavior. The problems with these findings as the basis for strong causal claims are multiple and significant.
The laboratory measures of aggression used in this research are not validated against real world aggressive behavior. The most common paradigm involves giving subjects the opportunity to administer noise blasts or electric shocks of varying intensity to another subject, with higher intensity ratings taken as evidence of greater aggression.[86] The relationship between willingness to administer noise blasts in a laboratory context and the likelihood of committing sexual violence against a real person in a real-world context has not been established and is not obvious. Laboratory aggression paradigms measure something. What they measure and its relationship to the behaviors about which we actually care is an assumption that the research tradition has not adequately examined.
The attitude measures used in this research face the methodological problems identified in the previous section. Scales measuring objectifying attitudes frequently conflate the assessment of physical attractiveness with the denial of personhood in ways that make their findings difficult to interpret. A subject who rates female physical appearance as important in sexual attraction and who also fully recognizes women's autonomy, intelligence and moral status will score higher on many objectification scales than a subject who professes indifference to female appearance while privately holding contemptuous attitudes toward women that the scale does not capture. The instrument cannot distinguish what it is supposed to distinguish.
The difference between attitude change and behavior change is consistently underemphasized in reports of this research. Even accepting the finding that exposure to sexual media produces measurable changes in expressed attitudes in laboratory conditions, the relationship between attitude change in laboratory conditions and behavior change in real world conditions is not established and is generally assumed rather than demonstrated. People's expressed attitudes in research contexts and their actual behavior in real world contexts are related but not identical. The gap between them matters enormously when the policy conclusion being drawn is about the real-world behavioral consequences of media consumption.
Neil Malamuth's research on sexually aggressive men and their pornography consumption patterns has been influential and deserves specific attention because it is among the more methodologically careful work in this area and because its findings are more nuanced than their citation in cultural criticism suggests.[87] Malamuth found that among men already showing risk factors for sexual aggression, including hostile attitudes toward women and a history of early exposure to sexual violence, pornography consumption was associated with higher rates of sexually aggressive behavior. Among men without these pre-existing risk factors the relationship between pornography consumption and sexual aggression was not found.[88] The finding is consistent with Michael Seto's population level analysis: the relationship between sexual media consumption and harmful behavior is specific to populations with pre-existing risk factors rather than a general relationship that applies across the population of consumers.
This nuanced finding is not what gets cited when Malamuth's work appears in cultural criticism about sexual media. What gets cited is the association between pornography and aggression in high-risk populations, presented as though it were a finding about sexual media consumers in general. The omission of the finding that no such relationship was found in low risk populations, which constitute the large majority of sexual media consumers, is a significant distortion of what the research actually shows.
Christopher Ferguson's meta-analyses of the sexual media and aggression literature, the most comprehensive systematic reviews available, found that the effect sizes reported in individual studies were substantially reduced when corrected for publication bias, that the methodological quality of studies was inversely related to the strength of the effects they reported, and that the evidence for a causal relationship between sexual media consumption and sexual violence was insufficient to support the strong claims being made on its basis.[89] Ferguson's work has been contested by researchers in the field and the debate continues. What his analyses establish beyond reasonable dispute is that the research literature does not speak with the unified voice that its citation in cultural criticism implies and that the confidence with which causal claims about sexual media and harmful behavior are made is not supported by the quality of the evidence underlying them.
The Base Rate Argument
Any causal theory of the relationship between sexual media consumption and harmful behavior toward women has to account for a basic feature of the evidence that is rarely stated plainly in these debates: the vast majority of men who consume sexual media across their entire lives, including content that by any definition involves objectification, do not commit sexual assault.
The rates of sexual media consumption among adult men in societies with significant internet access are very high. Studies in multiple countries find that substantial majorities of adult men report regular consumption of pornography, with some studies finding rates above eighty percent among younger men. The rates of sexual assault, while unacceptably high in absolute terms, represent a small fraction of the male population.[90] If sexual media consumption were a significant causal driver of sexual violence, the relationship between these two facts would require explanation. It does not get one in most discussions of objectification and harm.
The base rate issue is not resolved by arguing that the effects are probabilistic rather than deterministic, that media consumption raises the probability of harmful behavior without guaranteeing it. A probabilistic relationship still produces detectable population level effects and the population level evidence, as the work of Diamond and others reviewed in the previous section documents, does not show the relationship the harm thesis predicts. Countries and periods with higher sexual media consumption do not show higher rates of sexual violence. This is not what a significant causal relationship between the two would look like.
The base rate argument also has implications for the moral characterization of sexual media consumers. If the large majority of men who consume objectifying content do not commit sexual assault, the inference from consumption to harmful intent or dangerous character is not warranted by the evidence. The man consuming sexual imagery is overwhelmingly likely to be one of the men for whom consumption does not predict harmful behavior. Treating him as presumptively suspect on the basis of his consumption patterns requires either ignoring the base rate or explaining why the base rate does not apply to the inference being made. Neither move is typically made explicit in arguments about AI sexual content.
The Exposure Rate Problem
The statistics most frequently cited in discussions of male sexual violence and its causes suffer from the exposure rate problem identified earlier in this essay in the context of the bear and man in the woods argument and it is worth applying that analysis specifically to the objectification debate.[91]
The statistic that men commit the overwhelming majority of sexual violence is true and important. The inference that male sexuality is therefore inherently dangerous, or that expressions of male sexual desire should be treated with presumptive suspicion, does not follow from this statistic without accounting for exposure rates. The relevant comparison is not between the number of sexual assaults committed by men and the number committed by women in absolute terms. It is between the rate of sexual assault per unit of sexual interaction for men and women respectively, controlling for the enormously different rates at which men and women are in situations where sexual assault could occur.
This is not a defense of male violence or an argument that the disproportion in sexual assault statistics is unimportant. It is a methodological point about what the statistics actually measure. The disproportion in sexual assault rates reflects a real disproportion in harmful behavior that demands serious response. It does not establish that male sexuality itself, the orientation toward female attractiveness, the experience of sexual desire in response to visual stimuli, the production and consumption of sexual imagery, is a causal driver of that violence rather than a biological feature of the large majority of men who share it and who do not commit violence. To be clear about what this argument does and does not establish: even corrected for exposure rates, male rates of sexual violence exceed female rates by a substantial margin, and that disproportion is real and demands serious engagement. What the exposure-rate correction does is relocate the analytic question from the biological feature shared by the large majority of men who do not commit violence to the specific individual-level risk factors that distinguish men who do from men who do not. That is where the causal work of an honest theory of sexual violence has to be done.
The conflation of male sexuality with male sexual violence is precisely the error that the exposure rate analysis reveals. Male sexuality is nearly universal among men. Sexual violence is committed by a small minority. A causal account of sexual violence that implicates male sexuality in general has to explain why a universal feature produces a minority outcome, and the evolutionary and epidemiological evidence suggests that the explanation has more to do with individual risk factors, developmental histories, and contextual variables than with the general feature of male sexual desire that the objectification argument targets.
The Idealization Question
The sexual imagery that generates the most concern in current debates about AI content is characterized by idealization: the female figures depicted are young, symmetrical, physically healthy, and display the features that cross cultural research identifies as attractiveness cues associated with fertility and reproductive value. Critics of this content read the idealization as evidence of its objectifying character, as the production of an impossible standard that demeans real women by comparison. The evolutionary account of idealization suggests a different interpretation.
Fantasy tends toward the ideal because the psychological mechanisms that generate sexual desire evolved to respond to features that predicted reproductive value in ancestral environments. These mechanisms do not have access to the statistical distribution of real women's appearances. They respond to the cues they were designed to respond to. When those cues are presented in concentrated form, as they are in idealized imagery, the response is correspondingly strong. The idealization is not a statement about real women's inadequacy. It is a feature of how the male visual system processes attractiveness cues in the context of fantasy rather than mate selection.
The distinction between fantasy idealization and real world expectation is psychologically real and important. Research on men's relationship expectations consistently finds that the standards men apply in actual mate choice differ substantially from the idealized figures they find attractive in fantasy contexts.[92] Men in real relationships with real women do not consistently evaluate their partners against idealized fantasy standards and find them wanting. The psychological compartmentalization between fantasy and reality is a feature of normal human sexual psychology, not a failure of it. The argument that idealized sexual imagery produces unrealistic expectations of real women that damage real relationships has been made for decades about every medium in which idealized female imagery appears and has not been demonstrated at any detectable scale in the real-world evidence.
The function of idealization in fantasy is better understood as expressive rather than prescriptive. The idealized figure is not a specification for a real partner. She is an expression of a desire that exists independently of and prior to any specific real woman. The man generating idealized female imagery is not making a statement about what real women should look like. He is expressing a psychological tendency that the evolutionary evidence suggests is a feature of human male sexual psychology operating exactly as it was designed to operate.
The Asymmetry the Framework Cannot Explain
The objectification framework as deployed in current debates about AI sexual content applies its analytical categories with a consistency that deserves scrutiny. The framework reads male consumption of sexualized female imagery as objectification, as evidence of an attitude toward women that denies their subjectivity and reduces them to instruments of male pleasure. It does not apply the same analysis to female consumption of analogous content with anything like the same consistency, and the inconsistency requires an explanation that the framework does not provide.
Female consumption of sexualized male imagery exists and is substantial. The cultural products organized around female sexual desire, from romance novels featuring idealized male figures to the Magic Mike franchise to the enormous market for male-oriented content produced for female audiences, involve the same features that the objectification framework identifies as morally problematic when they appear in male-oriented content.[93] The idealized male figure in a romance novel is instrumentalized for the reader's emotional and sexual pleasure. His autonomy is circumscribed by the narrative. He is to some degree fungible with other idealized male figures in the genre. His primary function is to produce the reader's pleasurable response.
These features are not typically read through the objectification framework when they appear in female-oriented content. The female reader of romance fiction is not characterized as exhibiting an attitude toward real men that reduces them to objects. She is understood to be engaging with a fantasy that exists in a context separate from her real-world relationships and attitudes. The fantasy is recognized as serving a psychological function that is normal, healthy and unproblematic.
The asymmetry in how the framework is applied to male and female consumption of sexualized imagery is not explained by any feature of the framework itself. Nussbaum's seven features of objectification apply to the idealized male figure in romance fiction as readily as they apply to the idealized female figure in an AI generated music video. The selective application of the framework to male-oriented content requires an additional argument that the framework's advocates have not provided: an argument for why the psychological functions served by sexual fantasy are normal and unproblematic when women engage in them and symptomatic of a dangerous worldview when men do.
The most common response to this asymmetry is the argument from power: that objectification in male-oriented content is more harmful because it operates in the context of a broader social structure in which women are already disadvantaged, and that this context changes the moral weight of the objectification.[94] This is a coherent argument that deserves engagement rather than dismissal. What it does not establish is that the male consumer of sexual imagery is thereby participating in the oppression of real women, that his fantasy is causally connected to the social structures the power argument invokes, or that restricting his fantasy would have any detectable effect on those structures. The power argument provides a context for concern. It does not provide the causal mechanism that would connect the expression of male desire in fantasy to real world harm in ways that justify the moral characterization being made. The power argument's strongest form deserves a more careful response than dismissal on causal grounds alone, because the power argument is not primarily a causal claim. It is a contextual claim about the moral weight of a behavior given the surrounding conditions under which it occurs, and the contextual move is legitimate as a feature of moral reasoning. Conduct can have different moral weight depending on the power relations and social structures within which it occurs. What the contextual move does not do, when deployed against the specific case of male fantasy content, is demonstrate that the particular content participates in the structure rather than merely occurring within it. Cultural adjacency to a structure of harm is not participation in that structure, and the move from adjacency to participation requires exactly the kind of causal argument the contextual framing is offered as a substitute for. The power argument can be granted its premise about contextual modulation of moral weight without yielding the conclusion that any particular piece of fantasy content inherits the moral weight of the broader structure.
The Violence Asymmetry
There is a second asymmetry in the framework that deserves direct examination because it cuts against the priorities the framework implicitly asserts. The cultural discourse that characterizes male sexual imagery as a driver of harm is substantially more tolerant of male violent imagery even though the harm thesis, if it were consistent, would predict the opposite pattern.[95] The action film in which men are dismembered, tortured, set on fire and shot through the head is a mainstream entertainment product that receives a PG-13 rating in the United States if the blood is managed appropriately. The corresponding sexual imagery, showing no violence and producing no victim, receives an NC-17 rating that effectively removes it from mainstream distribution. The rating system encodes the asymmetry into the industrial structure of American entertainment and it does so without defending the underlying principle.
The research record is not consistent with the rating system's implicit theory. The pipeline arguments applied to sexual imagery have identical logical structures to pipeline arguments applied to violent imagery, and the latter have been tested more extensively without yielding the effect sizes the theory predicts.[96] Ferguson and Kilburn's meta-analysis of media violence, cited later in this essay, applies directly to the question of whether violent entertainment produces violent behavior, and the answer the evidence supports is that it does not at anything like the scale required to justify the concern that currently attaches to sexual content by analogy. If the pipeline argument were being applied consistently it would treat sexual imagery with the same lightness that violent imagery currently receives, or treat violent imagery with the same suspicion that sexual imagery currently receives. The current configuration, which treats sexual imagery as more morally suspect than violent imagery while relying on research that does not support either suspicion, is not defensible as the expression of the evidence. It is defensible only as the expression of a culturally specific discomfort with sexual expression that the evidence does not vindicate. The strongest defense of the current configuration available in the literature, that sexual response generalizes from imagery to real-world perception in ways that aggressive response does not, is itself an empirical claim and one that would require exactly the kind of evidentiary support the pipeline argument has been unable to provide. The asymmetry in treatment is not vindicated by the asymmetry in the research base.
The objectification framework, properly understood rather than selectively deployed, is a useful analytical tool for identifying what is wrong about specific kinds of treatment of real people in real contexts. It is not a tool that, applied honestly and consistently, supports the conclusion that male visual sexual desire expressed through idealized imagery is a pipeline to harm. The distinction between imagining a person as an object and treating a person as one is not a loophole. It is the entire argument. And the evidence for the causal bridge between the two that would close that gap is substantially weaker than the confidence with which the bridge is assumed.
VII. The Pipeline Problem
The pipeline argument is the most rhetorically powerful claim in the current debate about male sexuality and AI generated content, and it deserves to be stated in its strongest form before it is examined critically. Doing otherwise would be the kind of intellectual dishonesty this essay has tried throughout to avoid.
The argument in its strongest form runs as follows. There exists a documented continuum of online misogynist communities ranging from relatively mainstream expressions of male frustration and entitlement through progressively more extreme ideological positions to communities that have been directly associated with real world violence against women. This continuum is not theoretical. Researchers have documented it. The men who have committed high profile acts of violence motivated by misogynist ideology, from Elliot Rodger's 2014 Isla Vista killings to Alek Minassian's 2018 Toronto van attack, were embedded in online communities that occupy positions on this continuum.[97] The ideological content of those communities, the beliefs about women, about male entitlement, about the causes of male frustration and the appropriate responses to it, did not appear from nowhere. It was produced and sustained by a cultural ecosystem with identifiable features. The question the pipeline argument poses is whether apparently innocuous upstream content, including the mass production of idealized sexualized female imagery, contributes to that ecosystem in ways that make the downstream violence more likely even when the individual producer of the upstream content has no violent intent and no awareness of the connection.
This is a serious argument. It is grounded in real research, real events and real concern about real harm. The Coufal et al. study published in Nature Humanities and Social Sciences Communications in 2025, examining radicalization pathways within incel communities, documents genuine progression from relatively mild expressions of frustration to increasingly extreme ideology within identifiable online spaces.[98] The UN Women and Institute for Strategic Dialogue research on the manosphere confirms that radicalization pathways exist and that entry points into those pathways are often content that appears harmless in isolation. These findings are real and they matter. Anyone making an argument about male sexuality and digital content who dismisses them entirely is not engaging honestly with the evidence.[99]
The question is not whether the pipeline exists. It does, in specific populations, under specific conditions, through specific mechanisms that researchers have documented. The question is whether the pipeline argument, which was developed to describe radicalization in specific online communities with specific ideological content, can be legitimately extended into a general theory of male sexuality such that the production and consumption of sexual imagery featuring attractive women is characterized as an upstream stage in a process that ends in violence. That extension is where the argument breaks down and it breaks down in ways that are systematic enough to constitute not a refinement of the theory but a fundamental challenge to how it is being applied.
Where the Research Is Actually Solid
The radicalization research that provides the empirical foundation for the pipeline argument is most robust in its analysis of specific ideological communities with identifiable content. The incel community, the men going their own way movement, the red pill community and related online spaces have been studied by researchers including Ribeiro et al., whose 2020 paper in the proceedings of The Web Conference mapped radicalization pathways using YouTube recommendation data, and by Moonshot CVE and the Institute for Strategic Dialogue, whose qualitative research has documented the specific rhetorical and ideological mechanisms through which men move from mainstream frustration into extreme communities.[100]
What this research documents is a genuine phenomenon: men who experience real social frustration, often related to romantic and sexual rejection, encounter online communities that provide an ideological framework for understanding that frustration, a community of similarly situated men, and a vocabulary of grievance that attributes the frustration to the malevolence or unworthiness of women rather than to complex social and individual factors. The framework provides both an explanation and an enemy and the combination is psychologically potent. Men who become deeply embedded in these communities show measurable shifts in their attitudes toward women and toward violence that are qualitatively different from the attitudes of men who consume sexual imagery without engaging with ideological content.
The specific mechanisms the research identifies are worth noting because they clarify what the pipeline research is actually about. The radicalization pathway involves ideological content: explicit claims about women's nature and worth, about male victimhood, about the legitimacy of hostility toward women as a response to perceived injustice. It involves community reinforcement of those ideological positions. It involves progressive normalization of increasingly extreme expressions of hostility. These mechanisms are specific to communities organized around explicit misogynist ideology and they are distinct from the consumption of sexual imagery that involves no ideological content about women's worth or the legitimacy of hostility toward them.
The research is solid on the specific phenomenon it studies. The problem arises when the findings about ideologically organized misogynist communities are extended to characterize sexual imagery consumption in general, as though the mechanism identified in one context operates identically across all contexts that share surface features.
Where It Breaks Down: The Base Rate Problem
The extension of the pipeline argument from specific radicalized populations to general male sexuality founders immediately on the base rate problem that has appeared at multiple points in this essay and that never receives an adequate response from the argument's proponents.[101]
The population of men who consume idealized sexualized female imagery is enormous. In societies with significant internet access, it encompasses the large majority of adult men. The population of men who have radicalized through online misogynist communities into ideologies associated with real world violence is vanishingly small by comparison. If consumption of sexual imagery were a meaningful upstream stage in the pipeline that ends in radicalized misogynist violence, the relationship between these two populations would require explanation. The vast majority of men who consume sexual imagery do not progress to ideological misogynist communities. They do not develop the specific beliefs about women that characterize those communities. They live their lives, maintain relationships with real women, and produce no detectable contribution to the violence that the downstream stages of the pipeline involve.
A causal theory that predicts a pathway from X to Y must account for the overwhelming majority of X cases in which Y does not follow. The pipeline argument does not do this. It establishes that Y is sometimes preceded by X without establishing that X is a meaningful predictor of Y in the general population. The distinction is the difference between necessary and sufficient conditions, and it is one that the pipeline argument consistently collapses.
The response that the pipeline effects are probabilistic rather than deterministic, that consuming sexual imagery raises the probability of radicalization without guaranteeing it, requires specifying what the probability increase is and demonstrating it against the base rate. If consuming sexual imagery raises the probability of radicalization from 0.001 percent to 0.002 percent, that is a doubling of relative risk that is simultaneously a trivial absolute risk increase. The framing of pipeline risk in relative rather than absolute terms is a feature of how risk is routinely overstated in public health communication and it is no less distorting when applied to the radicalization literature.
The Selection Bias Problem
The radicalization research faces a selection bias problem that its practitioners generally acknowledge but that rarely receives adequate attention when the research is cited in cultural criticism.[102] The studies that document radicalization pathways examine populations that have radicalized. They trace the content consumption histories, the community memberships and the ideological progressions of men who ended up in extreme misogynist communities and they identify patterns in those histories. This methodology is valuable for understanding how radicalization happens in the populations where it does happen. It cannot establish the probability that a given upstream behavior, such as consuming sexual imagery, will lead to radicalization in the general population, because the general population of men who consume sexual imagery is not the study population.
The methodological problem is analogous to studying car accident victims to determine the risks of driving and finding that all of them were in cars at the time of the accident. The finding is true, but it does not establish what the risk of being in a car is for drivers in general because the study population was selected precisely on the basis of the outcome being studied. Studies of radicalized men that find sexual imagery consumption in their histories cannot establish that sexual imagery consumption is a risk factor for radicalization in unselected populations because those studies did not examine unselected populations. They examined men who radicalized.
The appropriate comparison is between the content consumption patterns of men who radicalized and the content consumption patterns of men who consumed similar content without radicalizing. This comparison is rarely made in the radicalization literature and when it is made the findings consistently show that the distinguishing features of radicalization are not the sexual content consumption that the pipeline argument targets but the ideological content, the community embeddedness, and the pre-existing psychological vulnerabilities that characterize the men who progress to extreme communities. The sexual imagery is not the active ingredient.
The Continuum Fallacy
The pipeline argument in its current form deploys what can be called the continuum fallacy: the logical error of treating the existence of a continuum between two phenomena as establishing meaningful proximity between points on that continuum that are in fact far apart.[103]
The argument works as follows. There exists a continuum from sexual imagery consumption through ideological misogyny to real world violence. Therefore, sexual imagery consumption is connected to real world violence in ways that justify treating it as a stage in a process that ends in harm. The existence of the continuum is presented as doing the argumentative work that a demonstrated causal mechanism would need to do.
The problem with this structure is that a continuum between two things establishes only that they share a category, not that occupying one position on the continuum makes occupation of another position probable or even meaningfully more likely. Alcohol consumption and alcoholism exist on a continuum. The existence of that continuum does not mean that having a glass of wine with dinner is meaningfully connected to end stage alcoholism in ways that justify treating wine consumption as an upstream stage in a harmful process. The overwhelming majority of people who drink alcohol do not become alcoholics. The continuum does not collapse the distinction between them.
Action cinema and real-world violence exist on a continuum. Horror fiction and real-world cruelty exist on a continuum. Competitive sport and assault exist on a continuum. Religious devotion and religious extremism exist on a continuum. In each case the existence of the continuum is true and trivial. It does not establish that occupying the upstream position makes the downstream position more likely in ways that would justify treating the upstream activity as presumptively suspect. The continuum is not the argument. The argument requires a demonstrated causal mechanism with demonstrated effect sizes in the relevant populations. The pipeline literature has not provided this for the connection between sexual imagery consumption and radicalized misogynist violence.
The specific deployment of the continuum argument in debates about AI sexual content is worth examining because it reveals how the rhetorical move works in practice. The argument places PG-13 sexual imagery and coordinated sexual assault networks on the same continuum, acknowledges that they are not the same thing, and then proceeds to treat the former as though it belongs to the same moral and causal category as the latter. The acknowledgment and the treatment contradict each other. The acknowledgment is designed to protect the argument from the charge of unfairness. The treatment is where the rhetorical work of the continuum gets done. Naming this structure does not resolve the underlying empirical question. It clarifies that the underlying empirical question has not been answered by the rhetorical move.
The Bystander Argument
A specific version of the pipeline argument that deserves its own response is the claim that men who produce or consume sexualized AI imagery without directly intending harm are nevertheless functioning as bystanders whose silence constitutes approval of the ecosystem their content contributes to, and whose apparent innocence is itself a form of complicity in the harms the ecosystem produces.[104] This argument borrows the moral structure of bystander analysis developed in other contexts, the ethics of not intervening when harm is being committed, and extends it to the consumption of content that is not itself the harm but is claimed to be complicit with it through a shared cultural space.
The extension fails for the same reason the pipeline extension fails. Bystander ethics in its original form identifies a specific moral failure: the presence of a witness to harm who chooses not to intervene when intervention would be feasible and would reduce the harm. The moral force of the analysis depends on the specificity of the relationship between the bystander and the harm, the presence of a real victim, a present opportunity to intervene, and a foreseeable reduction in harm contingent on the intervention.[105] None of these features characterize the relationship between the man generating a music video image and the coordinated rape networks the bystander argument wants to connect him to. He is not present at the harm. He has no opportunity to intervene in the harm. His non-intervention does not reduce the harm because his intervention would not have produced the harm in the first place. The argument that his consumption of unrelated content constitutes a morally relevant failure to intervene requires treating cultural adjacency as moral presence, and cultural adjacency is not moral presence. The expansion of bystander ethics into a theory of diffuse cultural complicity is the move that gives the argument its rhetorical force and it is the move the argument cannot justify.
The Unfalsifiable Loop
There is a specific feature of the pipeline argument as it is deployed in contemporary debates that deserves direct identification because it constitutes a recognized epistemological problem that applies regardless of the subject matter. The argument has been structured in ways that make it immune to refutation by any possible response from the men it targets, and a theory that cannot be refuted by any possible evidence is not a scientific theory. It is an ideology.
The structure works as follows. Men who consume sexual imagery are exhibiting behavior that the pipeline argument characterizes as an upstream stage in a radicalization process. When these men respond to this characterization with disagreement, pushback or expressions of frustration, the framework interprets this response as confirmation of the thesis. The disagreement is read as the masculinity threat response that the Stanaland research predicts, as evidence of fragile masculinity reacting defensively to accurate characterization. The pushback is read as the aggressive defense stage of the radicalization loop.[106] The frustration is read as the grievance deepening that hardens ideological identity.
The result is a closed system in which agreement confirms the thesis and disagreement also confirms the thesis. There is no response available to the targeted man that the framework would interpret as evidence against its characterization of him. If he agrees, he is open to feedback. If he disagrees, his disagreement is a symptom. If he argues calmly, he is in the early defensive stage. If he argues with heat, he is in the aggressive stage. If he withdraws from the conversation, he is avoiding. Every possible response is interpreted as confirming the framework.
This structure is not unique to the masculinity research that the pipeline argument draws on. It appears in various forms across the history of ideological systems that have become sufficiently closed to resist empirical challenge. The psychoanalytic tradition produced similar structures in which resistance to a diagnosis was interpreted as evidence of the defense mechanisms the diagnosis predicted. The witch trial produced similar structures in which denial of witchcraft was interpreted as evidence of the devil's influence. The specific content of these systems differs enormously. The epistemological structure, the immunity to refutation by any possible response from the accused, is identical.
Karl Popper's criterion of falsifiability, the requirement that a scientific theory must be capable of being proven wrong by some possible observation, is the standard philosophical response to this kind of closed system.[107] A theory that is consistent with all possible observations, including the observation that its subjects deny it, is not a scientific theory in Popper's sense. The pipeline argument in its current form, structured so that any response from targeted men confirms rather than challenges it, fails this test. This does not mean the underlying concern about radicalization is wrong. It means the specific argumentative structure being used to advance that concern has a serious epistemological problem that its proponents should address rather than exploit.
The practical consequence of the unfalsifiable loop is that it becomes impossible to distinguish between men who are genuinely exhibiting the defensive responses that the research predicts and men who have a legitimate disagreement with the characterization being made of them.[108] If these are indistinguishable within the framework, the framework cannot be used to make the distinction it claims to make. It can only be used to confirm the characterization regardless of whether the characterization is accurate in any specific case. That is not analysis. It is accusation wearing the clothing of analysis.
The Historical Parallel Revisited
The pipeline argument has a direct historical predecessor that this essay has already examined in detail but that deserves specific application here. The claim that horror films created violent criminals, that heavy metal created Satanists and suicides, that violent video games created school shooters, each followed the same logical structure as the current pipeline argument about sexual imagery and radicalized misogyny.
In each case a population of concerning real world harm was identified: violent crime, teen suicide, school shootings. In each case a cultural product consumed by a population that overlapped with the harmful population was identified: horror films, heavy metal, violent video games. In each case the overlap was interpreted as a causal relationship and the cultural product was characterized as an upstream stage in a pipeline that ended in real-world harm. In each case the argument was made with reference to psychological research that was presented as establishing the causal connection. In each case the research did not establish what it was said to establish. In each case the predicted downstream effects did not materialize at anything like the predicted scale. In each case the content survived and is now consumed without the presumption of harm that once attached to it.
The heavy metal case is particularly instructive because the PMRC's specific argument about the pipeline from heavy metal to teenage suicide and Satanism followed the pipeline logic almost exactly. Young men who were experiencing social frustration and isolation were consuming heavy metal music that expressed anger, rebellion and transgressive imagery. The PMRC argued that the music was not merely reflecting the frustration of its audience but amplifying and directing it toward harmful outcomes, that the specific content of the music, its anger, its darkness, its transgressive imagery, was an upstream stage in a process that could end in self-harm or violence. This argument was made with reference to psychological research, with case studies of young men who had committed violence and who had consumed the music, and with the same kind of continuum logic that places the music and the violence on a shared pathway.
The downstream effects did not materialize. The generation that consumed heavy metal in the 1980s did not produce elevated rates of Satanism, suicide or violence compared to generations that did not. The researchers whose work was cited in support of the harm thesis were not vindicated. The musicians whose work was characterized as harmful are now recognized as artists. The PMRC is remembered as a cautionary example of moral entrepreneurship overreaching its evidence.
There is no principled reason to believe the current pipeline argument about sexual imagery and radicalized misogyny will fare better. The logical structure is the same. The relationship between the evidence and the confidence with which the claim is made is the same. The historical track record of this kind of argument is the same.
The Ferguson and Kilburn Meta-Analysis
Christopher Ferguson and John Kilburn's 2009 meta-analysis of the violent media and aggression literature, published in the Journal of Pediatrics, is the most relevant systematic review for evaluating the evidentiary basis of media pipeline arguments. The analysis examined 25 studies of violent media and aggression, correcting for publication bias using established statistical methods, and found that the effect sizes reported in individual studies were substantially reduced when this correction was applied. The corrected effect sizes were small enough that Ferguson and Kilburn concluded the research failed to support the conclusion that media violence was related to aggressive behavior to a degree that would be considered clinically meaningful.[109]
The methodological problems Ferguson and Kilburn identified in the violence literature are the same problems that appear in the sexual media literature: the use of laboratory aggression measures not validated against real world behavior, the failure to control for third variables that predict both media consumption and aggressive behavior, the publication bias that makes positive findings more likely to appear in the literature than null findings, and the tendency to report relative risk increases without contextualizing them against base rates.
These findings have been contested. Anderson and Bushman, whose work on media violence is the most prominent in the field, have responded to Ferguson's critiques and the debate continues in the literature. What Ferguson and Kilburn's analysis establishes is that the research base for media pipeline arguments is not the solid evidentiary foundation that cultural criticism implies when it cites this research. The science is contested. The effect sizes are small and sensitive to methodological choices. The relationship between laboratory findings and real-world behavior is assumed rather than demonstrated. These are not fringe objections. They are mainstream methodological concerns raised by researchers within the field.
What Would Actually Constitute Evidence
The pipeline argument as currently deployed conflates two distinct claims that require different kinds of evidence and that should not be treated as though they support each other.
The first claim is that documented radicalization pathways exist in specific online communities through which some men move from mainstream content to extreme misogynist ideology. This claim is supported by the research and is not in dispute here. The research on incel communities, on the manosphere and on online radicalization generally has documented real phenomena using appropriate methodologies.
The second claim is that the production and consumption of sexual imagery featuring attractive women by men in general creative and entertainment contexts is a meaningful upstream stage in these radicalization pathways such that it should be characterized as contributing to a cultural ecosystem that enables violence. This claim is not supported by the research cited in its favor and would require different and more demanding evidence than is currently being offered. It is worth stating this narrowing precisely: the base-rate argument refutes the strong claim that sexual imagery consumption is a general driver of radicalization. It does not refute the narrower claim, which the research supports, that sexual imagery consumption may be one of several interacting risk factors elevating radicalization rates within already-vulnerable subpopulations. The problem addressed here is the rhetorical slide between the narrow claim the research supports and the general claim the discourse asserts, not the research itself.
Evidence that would actually support the second claim would need to demonstrate several things that have not been demonstrated. It would need to show that men who consume sexual imagery without engaging with explicitly ideological misogynist content show elevated rates of radicalization compared to men who do not consume such imagery, controlling for the pre-existing risk factors that the research identifies as the actual predictors of radicalization. It would need to show that the mechanism connecting sexual imagery consumption to radicalization is the sexual imagery itself rather than the pre-existing risk factors that predict both the imagery consumption patterns and the radicalization. It would need to show effect sizes at the population level that are large enough to be meaningful rather than statistically detectable at laboratory scale while trivial in real world terms. And it would need to account for the base rate of radicalization in the enormous population of men who consume sexual imagery without radicalizing.
None of this evidence has been provided. What has been provided is documentation of radicalization in specific populations, documentation of overlap between sexual imagery consumption and radicalized ideology in those populations, and the continuum argument that treats the overlap as establishing the causal connection. These things together do not constitute the evidence the second claim requires. Presenting them as though they do is a significant distortion of what the research actually shows and of what conclusions it legitimately supports.
The pipeline argument at its best is an important contribution to understanding a specific phenomenon of online radicalization in specific communities. At its worst it is a framework that uses real concern about real harm to justify a general suspicion of male sexual expression that the evidence does not support and that the history of similar arguments suggests will not be vindicated. Taking the real concern seriously means demanding that the argument be made with the precision and evidentiary support it currently lacks rather than accepting the rhetorical power of its framing as a substitute for that precision.
VIII. AI Specifically: The New Technology Panic
Everything this essay has traced across thirty-five thousand years of human creative production, across the obscenity trials and the pinup panics and the video nasty prosecutions and the PMRC hearings and the video game moral panics, arrives in the present moment in the specific form of the argument being made about AI generated sexual content in creative communities. The pattern is complete. The technology is new. The argument is not. And recognizing the pattern is not a way of dismissing the concerns that animate the argument. It is a way of insisting that those concerns be examined with the same critical attention that the history of similar concerns demands.
The present moment has specific features worth naming precisely because precision is what the debate currently lacks. Generative AI tools can produce visual content of extraordinary quality at extraordinary speed with minimal technical skill required from the person prompting them. A man who twenty years ago would have needed significant artistic training to produce a drawing of an attractive woman can now produce a photorealistic image of one in seconds. A man who ten years ago would have needed video production skills, equipment and collaborators to make a music video featuring an idealized female figure can now produce one alone in an afternoon. The barrier between desire and its visual expression has effectively collapsed. What was once gatekept by skill, equipment and resources is now gatekept only by access to a subscription service.
This is genuinely new. Not the desire. Not the imagery. Not the idealized female figure in a creative work. Those are as old as human art. What is new is the infrastructure through which the desire produces the imagery, the speed at which it does so, the volume that speed makes possible, and the visibility that volume creates in shared community spaces. These features are real and they change the texture of the phenomenon even when they do not change its nature. Understanding what is actually new about AI generated sexual content, as distinct from what merely appears new because the technology is unfamiliar, is essential to evaluating whether the response the content is receiving is proportionate to what it actually is.
Why Generative AI Triggers the Pattern So Effectively
The moral panic about AI generated sexual content has been so rapid and so confident partly because generative AI possesses in concentrated form all of the features that have historically triggered this pattern of response. Each feature activates a specific anxiety that has driven previous panics and the combination of all of them simultaneously produces a response of corresponding intensity.
The democratization of production is the primary trigger. Every previous moral panic about sexual content has been partly driven by the expansion of access to content that was previously controlled by technical or economic gatekeeping. Home video democratized access to films that theatrical distribution had controlled. The internet democratized access to content that publication costs had controlled. Each democratization produced a panic whose energy came partly from the disruption of gatekeeping systems that had provided a kind of implicit quality control as well as content control. When making something required skill and resources there was an assumption, usually unstated, that the maker had cleared some threshold of seriousness or intentionality. When that threshold falls the content that emerges feels unfiltered in ways that read as threatening even when the content itself is not qualitatively different from what came before.
Generative AI takes this process to its logical endpoint. The technical skill required to produce photorealistic imagery of an attractive woman has dropped to approximately zero. The craft barrier that separated the skilled artist from the person who merely wanted to see a specific image has been eliminated. The result is a volume of content that would have been impossible to produce at comparable quality by any previous means, produced by people with no training, no considered artistic intent in any conventional sense, and no investment beyond the time it takes to write a prompt. The volume is visible in community spaces in ways that individual production never was, concentrated in submission queues and social feeds in forms that make the scale of production legible in a way it previously was not.
The removal of the skill barrier is experienced psychologically as the removal of a moral filter even though no such filter was actually operating. The assumption that effort implies consideration, that making something carefully implies having thought about what you are making, is a reasonable heuristic in many contexts. Applied to sexual imagery it produces the inference that the absence of technical effort implies the absence of any consideration of what is being made and why. This inference is not reliable. People who prompt generative AI to produce sexual imagery have thought about what they want to see. The consideration is present. What is absent is the craft, and craft and consideration are not the same thing.
The algorithmic scale compounds the visibility problem. Generative AI tools can produce content at rates that individual human production cannot approach. The analysis of Grok's output during the scandal of early 2026, finding approximately 6,700 sexually suggestive images being generated per hour, is a number that has no analog in any previous technology for producing sexual imagery.[110] That number feels like escalation even when each individual image is not qualitatively different from images that have been produced and consumed throughout human history. The scale changes the experience of the phenomenon without necessarily changing its nature and the conflation of scale with nature is where much of the panic's energy comes from.
The Grok Scandal: A Real Case
The Grok scandal of early 2026 requires honest engagement rather than dismissal because it involves a genuine harm that is distinct from the AI generated sexual content that is the primary subject of this essay and that deserves the serious response it received.
Grok, the AI chatbot developed by Elon Musk's xAI, was used by large numbers of users to generate sexualized imagery of real identifiable women by replying to their publicly posted photographs with prompts requesting that the AI modify or replace their clothing. The women whose images were used did not consent to this use. They were real people with real identities whose likenesses were being used to generate sexual imagery without their knowledge or agreement. Among the images generated were depictions of minors.
These are genuine harms with genuine victim classes. The woman who finds that her publicly posted photograph has been used to generate sexualized imagery of her without her consent has experienced a real violation. The violation does not require physical contact. It does not require that the generated image be widely distributed. The use of her likeness for sexual purposes without her consent is a harm in itself, one that has been recognized by an increasing number of legal jurisdictions through legislation specifically addressing nonconsensual intimate imagery.[111] The fact that the image is AI generated rather than photographed does not change the nature of the harm when the source material is a real person's likeness used without consent. The generation of imagery depicting minors in sexual contexts is not a gray area regardless of the technology used to produce it.
Musk's public response to regulatory concern about Grok's outputs, characterizing proposed restrictions as any excuse for censorship, is the response of a person who has either not thought carefully about the distinction between censorship and the regulation of genuine harm or who finds that distinction inconvenient.[112] It is also, as this essay has documented, a response with a long and not particularly distinguished history. The men who argued that the BBFC's powers over home video were censorship were not wrong that the powers were real. They were wrong that the powers were therefore illegitimate. Regulation of content that causes genuine harm to real people is not censorship in any meaningful sense of the term. It is the appropriate response to a genuine problem.
The Grok scandal is real. The harm it caused was real. The response it received was proportionate to what it was.
The Category Error That Follows
What is not proportionate is what comes next. The Grok scandal, a case involving nonconsensual use of real women's likenesses and the generation of imagery depicting minors, has been used as rhetorical leverage against an entirely different category of content: consensual adult fantasy featuring generated figures who do not exist, have no likeness to any real person, and cannot be harmed because they are not persons.
The category error is the conflation of these two things under the general heading of AI generated sexual content in ways that allow the genuine harm of the first to color the moral characterization of the second. It is the same rhetorical move that was made in the video nasty panic, when the most extreme and genuinely problematic content on the DPP list was used to justify the prosecution of films that were not extreme and not genuinely problematic by any consistent standard. It is the move that was made in the PMRC hearings, when genuinely offensive material was used to justify warning labels on music that was not offensive by any standard a reasonable person would recognize. The move works by establishing the reality of harm in one specific case and then extending the moral characterization to a broader category of content that shares surface features with the harmful case while lacking the features that actually generate the harm.
The distinction the law draws in this area is instructive. Legal frameworks addressing nonconsensual intimate imagery focus on the use of real people's likenesses without consent.[113] Legal frameworks addressing child sexual abuse material focus on the depiction of real children and, in an increasing number of jurisdictions, generated imagery that is indistinguishable from depictions of real children regardless of whether any real child was used in its production. These frameworks are designed to protect real people from real harms. They are not designed to address adult fantasy content featuring generated figures with no connection to real individuals because such content does not involve the harms that the frameworks were designed to address.
The consistent conflation of nonconsensual deepfakes, CSAM and consensual adult fantasy in the current debate is either a logical error or a rhetorical strategy.[114] As a logical error it confuses category membership, all are AI generated sexual content, with morally relevant features, only some involve real people or real victims. As a rhetorical strategy it leverages genuine and serious harms to generate moral concern about content that does not share the features generating those harms. In either case the conflation does damage to the clarity of the conversation and to the ability to make the distinctions that a serious response to genuine harms requires.
CSAM as a Different Category Entirely
Child sexual abuse material requires specific statement because it is the most serious harm in this space and because its conflation with adult fantasy content is the most damaging conflation the current panic produces.
CSAM is not a gray area. It is not a free expression question. It is not subject to the arguments this essay has made about the freedom to produce content others find offensive or gratuitous. The harm CSAM causes is not speculative. It is not dependent on contested causal claims about attitude formation or behavioral pipelines. It is direct and serious in every case. Real children are harmed in the production of CSAM that involves real children. The question of AI generated imagery that depicts what appear to be children in sexual contexts, without involving real children in production, is more complex legally and ethically but the serious harm concerns that attach to it, the contribution to demand for real CSAM, the use in grooming, the violation of the dignity of children as a class, distinguish it categorically from adult fantasy content in ways that justify treating it as a different legal and ethical category.
The conflation of concern about CSAM with concern about adult fantasy content featuring generated figures of unambiguously adult appearance is one of the most consistent features of the current panic and one of the most damaging to clear thinking about what the actual problems are and what appropriate responses to them look like. When the category of concerning AI sexual content is expanded to include any idealized female figure in a music video, the expansion dilutes the seriousness of concern about the content that actually causes serious harm. The appropriate response to CSAM is not a general suspicion of sexual imagery. It is specific, targeted and serious engagement with the specific harm that CSAM causes and the specific mechanisms through which it operates.
The Curation Question and the Moral Diagnosis Problem
The question of what creative communities and platforms choose to platform is distinct from the question of whether the content they decline to platform is harmful and it is important to keep these questions separate even when they are systematically conflated in the current debate.
A platform, festival or community space has the legal right to decline any content it chooses. This right is real, it is important, and it does not require justification by reference to the harmfulness of the declined content. A festival that wants to focus on experimental sound design can decline music videos on purely aesthetic grounds. A community that wants to focus on narrative driven work can decline content it considers insufficiently narrative. These are editorial decisions that require no defense beyond the articulation of the curatorial vision the space is trying to realize. The Knight First Amendment Institute's documentation of courts consistently recognizing editorial discretion as itself a First Amendment protected activity establishes the legal foundation for this position clearly enough.[115]
What requires justification is the move from editorial discretion to moral diagnosis. The decision not to platform specific content is a curatorial choice. The characterization of the creator of that content as exhibiting dangerous patterns of thought, as contributing to a cultural ecosystem that enables violence against women, as manifesting fragile masculinity through their response to criticism, is a moral accusation about a specific person that requires proportionate evidence. The evidence required is not the general evidence that sexual imagery exists on a continuum with radicalized misogyny. It is specific evidence that this creator's work and intent belong to the morally concerning category being invoked.
The current practice in some AI creative communities of declining content while simultaneously characterizing its creators in terms drawn from radicalization research and objectification theory collapses the distinction between editorial decision and moral verdict in ways that do harm to creators who have done nothing that the evidence connects to the harms being invoked. A rejection is a legitimate curatorial decision. A rejection accompanied by the implication that the rejected creator is on a pipeline to misogynist violence is an accusation that requires considerably more support than the content's failure to meet a curatorial standard provides.
The practical consequence of conflating editorial discretion with moral diagnosis is that it makes the curatorial standards of a community into a de facto moral certification system in which inclusion signals virtue and exclusion signals danger. This is a significant expansion of what curation is and what it can legitimately claim to accomplish and it is one that the evidence does not support. The content that gets declined from a community festival is not thereby established as harmful. The creator who receives a rejection is not thereby established as dangerous. These are not the same thing as editorial judgment and treating them as though they are does damage both to the creators being characterized and to the intellectual honesty of the conversation.
The Asymmetry Fully Developed
The asymmetry between how the current framework treats female producers and male consumers of sexual content has been noted at several points in this essay and deserves its fullest statement in this section because it is the feature of the current debate that most clearly reveals the inconsistency at its foundation.
The framework that generates concern about AI sexual content extends to women who produce sexual content for male audiences a moral complexity and human depth that it withholds from the men who consume it. The OnlyFans creator is exercising bodily autonomy.[116] The cam performer is making an economic choice that deserves respect. The woman who produces sexual imagery for male audiences is a full human subject whose participation in the sexual economy reflects complex motivations that should not be reduced to simple characterization. These positions are defensible and this essay does not dispute them.
The framework then turns to the men who consume this content, or who produce analogous content using AI tools, and applies a fundamentally different analytical register. The consumer is not exercising desire. He is exhibiting a symptom. He is not seeking pleasure. He is reproducing a power structure. He is not a full human subject with complex motivations for wanting what he wants. He is a position in a social system, a manifestation of the male gaze, a potential upstream stage in a radicalization pipeline. The humanity extended to the female producer is withheld from the male consumer and the withholding is presented as a principled ethical position rather than an inconsistency.
The inconsistency becomes most visible when the content in question is identical. The woman who produces sexualized imagery of herself for male audiences is agentive. The man who produces sexualized imagery of a generated female figure for his own gratification is objectifying. The woman who sells access to her body's visual representation is empowered. The man who pays for that access is participating in a system of harm. These characterizations apply to the same transaction from different sides and they cannot both be right as descriptions of its moral character. Either the transaction is legitimate, in which case both parties' participation in it deserves the complexity and humanity the framework extends to the female side, or it is not, in which case the framework needs to explain why the moral concern it applies to the male side does not extend to the female side in the same terms.
The power argument, that the asymmetry reflects genuine asymmetry in the social positions of the parties rather than inconsistency in the framework, is the standard response and it deserves engagement.[117] The argument holds that female production of sexual content and male consumption of it occur within a social structure in which women are already disadvantaged, and that this context changes the moral weight of each party's participation. This is a coherent position as far as it goes. What it does not explain is why the social context that disadvantages women as a class changes the moral character of the specific male consumer's desire in ways that justify treating his desire as presumptively suspect. The social structure is real. The individual man consuming sexual imagery is not thereby established as a participant in the oppression of women in ways that his consumption causes rather than merely reflects. The argument from social structure to individual moral diagnosis requires a causal mechanism connecting the two that is assumed rather than demonstrated.
The deeper problem with the asymmetry is what it reveals about whose humanity the framework is actually committed to. A framework that extends full human complexity to one party in a transaction while treating the other party as a manifestation of a social system rather than a person with desires, pleasures and a life that includes but is not defined by those desires, is not an ethical framework applied consistently. It is a framework that has decided in advance which party's experience deserves to be humanized and which party's experience deserves to be analyzed as a symptom. That decision requires justification that the framework has not provided and cannot provide without abandoning its claimed commitment to the full humanity of all persons regardless of gender.
What Is Actually New
The honest accounting of what is genuinely new about AI generated sexual content, as distinct from what merely appears new because the technology is unfamiliar, is shorter than the current debate implies.
The scale is new. The speed is new. The democratization of production is new in its degree if not its kind. The visibility of volume in shared community spaces is new in its form. The specific legal questions raised by nonconsensual use of real people's likenesses through AI tools are new in their technical specifics even if the underlying harm of nonconsensual intimate imagery is not. The regulatory questions raised by the difficulty of distinguishing AI generated imagery from photographic imagery in contexts where that distinction matters legally are genuinely new and genuinely difficult.[118]
What is not new is the argument that male sexual desire expressed through idealized imagery constitutes a harm to women that justifies restriction, moral characterization of creators and the mobilization of community standards against the content. That argument is the same argument that has been made about every new technology that has lowered the barrier to the production or distribution of sexual content throughout the history traced in this essay. It has never been vindicated at anything like the scale its proponents predicted. The content it targeted has consistently survived. The panic has consistently subsided. The confidence of the people driving it has consistently proved to be more robust than the evidence underlying it.
The new technology does not change this pattern. It instantiates it. The man generating an AI music video featuring an attractive female figure is the latest in a line that runs back through the home video consumer, the men's magazine reader, the pinup collector, the symposium attendee, the person who acquired a Shunga woodblock print in Edo period Japan, the person who carved the Venus of Hohle Fels thirty-five thousand years ago. The technology changes. The desire does not. The argument that the desire is a symptom of a dangerous pathology has been made at every point in this line. It has never been right. There is no principled reason to believe it has become right now.
IX. In Defense of the Full Spectrum
Lloyd Kaufman did not make Troma films because he wanted to elevate cinema. He made them because he wanted to make movies that were funny, disgusting, transgressive and entertaining in ways that mainstream cinema either could not or would not be, and because he understood that there was an audience for exactly that combination that deserved to be served as much as the audience for prestige drama deserved to be served.[119] The Toxic Avenger, released in 1984, features a mop wielding mutant superhero killing villains in ways that are explicitly designed to be as gross as possible. It has no redemptive intellectual framework. It makes no claim to social significance. It is not asking for your respect. It is asking for your money and your willingness to watch something that will make you laugh and wince in roughly equal measure and it delivers exactly what it promises.
Troma films are not great art by any conventional measure and Kaufman has never claimed they are. They are something else: an honest transaction between makers who know exactly what they are making and audiences who know exactly what they are getting, conducted without pretense and without apology. That transaction is legitimate. The people who have argued otherwise across the forty years of Troma's existence have not been vindicated by history. The films are still there. The audiences still find them. The predicted cultural damage did not materialize. What did materialize is a body of work that has influenced filmmakers including James Gunn, who directed the first two Guardians of the Galaxy films after cutting his teeth writing Troma scripts, and Quentin Tarantino, whose debt to exploitation cinema's willingness to be exactly what it is rather than something more respectable runs through everything he has made.[120]
The Troma argument is not that low art is secretly high art or that gratuitous content is secretly meaningful by standards the audience has failed to recognize. The argument is simpler and more important: low art is legitimate on its own terms. Gratuitous content serves real human needs. Entertainment that exists purely to entertain is not a lesser category of human creative production requiring justification by reference to its unintended educational benefits. It is what it is and what it is is enough.
The Full Spectrum and Why It Matters
The defense of the full spectrum of artistic expression, from the most intellectually ambitious work to the most shamelessly gratificatory, from Bergman to Troma, from Naked Lunch to the AI generated music video with the attractive female character attached to a generic beat, is not a defense that requires treating all of these things as equivalent. They are not equivalent. The evaluation of their relative merit on aesthetic grounds is a legitimate and interesting conversation. What is not legitimate is the conversion of aesthetic preference into ethical hierarchy, the move from this is not to my taste to this should not exist or this person should not have made it.
Every position on the creative spectrum serves human needs that are real even when they are not intellectually prestigious. Horror films that exist primarily to frighten serve the human need to experience controlled fear in a safe context, a need that has been recognized since Aristotle's account of catharsis.[121] Comedy that exists primarily to make people laugh serves the human need for pleasure and release that makes life bearable in ways that no serious argument about comedy's social function fully captures. Pornography that exists primarily to arouse serves the human need for sexual stimulation in forms that reality does not always provide. The AI generated music video with the attractive female figure serves the human need for visual pleasure in the specific form that male visual sexuality takes when given access to tools that can realize its preferences. These needs are not equally prestigious. They are equally real.
The argument that content must demonstrate purpose beyond its primary function to deserve cultural space is an argument that has been made against every form of popular entertainment in every era and it has been wrong every time. It was wrong about penny dreadfuls.[122] It was wrong about jazz. It was wrong about comic books. It was wrong about rock and roll. It was wrong about slasher films. It was wrong about hip hop. The people making it in each era believed they were making a principled distinction between legitimate and illegitimate cultural production. They were making a class distinction between their own tastes and other people's tastes dressed in the language of principle.
Return to Burroughs: The Philosophy in Its Purest Form
William S. Burroughs made the philosophical argument that this essay is defending in its purest and most uncompromising form and he made it not with words alone but with the existence of Naked Lunch itself, a work that embodies its own argument in the sense that the argument could not be made as effectively any other way.
Naked Lunch does not ask for your sympathy. It does not offer you a protagonist whose journey you are invited to share in the way that redeems difficult content by providing the reader with a stable moral position from which to observe it. It does not package its transgression within a framework that tells you what to think about what you are experiencing. It puts you inside a consciousness dissolving under the pressure of addiction and control and it gives you no handholds. The cut-up method that Burroughs developed with Brion Gysin, the deliberate disruption of linear narrative, is not a stylistic affectation.[123] It is the formal enactment of the book's argument about how control systems, including the control system of conventional narrative, operate on human consciousness. You cannot read Naked Lunch comfortably. You are not supposed to. The discomfort is the point.
The sexual content of Naked Lunch is explicit, frequently violent and in several sequences designed to be as disturbing as Burroughs could make it. The Hanging sequence, in which sexual arousal and death are combined in imagery of deliberate extremity, is not there to titillate. It is there to force the reader to confront the relationship between pleasure and death, between desire and control, between the body's responses and the moral frameworks the mind imposes on them. Whether it succeeds is a matter of genuine debate. Whether it is doing something, whether there is a human intelligence behind it making choices for reasons that connect to a coherent artistic project, is not in serious doubt to anyone who has read it carefully.
But here is the crucial point, the point that the Massachusetts Supreme Judicial Court established in 1966 and that needs to be stated as clearly as possible: Burroughs did not need to be doing something that you recognize as a legitimate artistic project to deserve legal protection.[124] He did not need to justify his transgression to the people who found it offensive. He did not need to demonstrate to any authority's satisfaction that the discomfort his work produced served a purpose that the authority recognized as legitimate. The ruling established that a work does not have to be pleasant, coherent or morally instructive to merit protection, and the philosophical principle behind that ruling extends beyond the legal domain into the cultural one.
The principle is this: discomfort is not harm. Offense is not harm. Disgust is not harm. The experience of encountering content that you find repugnant, that serves no purpose you recognize, that produces no response in you except the desire for it not to exist, does not constitute harm in the sense that generates a legitimate claim against its existence. Harm requires a victim. Harm requires a mechanism connecting the content to damage sustained by a real person. The feeling of offense, however strongly felt and however widely shared, is not that mechanism. Burroughs understood this. The Massachusetts court understood this. The people currently arguing that AI generated sexual content should be restricted because they find it objectifying and gratuitous have not engaged with this principle and their argument does not work until they do.
The History of Taste Based Gatekeeping
The history of cultural gatekeeping based on taste is long enough and its outcomes consistent enough to constitute a pattern that anyone claiming authority over what content deserves to exist should be required to account for.
The people who believed that penny dreadfuls were corrupting the working class were certain they were making a principled distinction between legitimate and illegitimate reading material. The penny dreadfuls survived. The certainty did not produce vindication. The people who argued that jazz was morally dangerous music associated with criminal and sexual excess were certain they were identifying a genuine cultural threat.[125] Jazz became one of America's most celebrated art forms. The people who believed comic books were producing juvenile delinquents, whose certainty was sufficient to generate congressional hearings and a self-censorship code that constrained the medium for decades, were not vindicated by subsequent history.[126] Comic books are now studied in universities and adapted into some of the most commercially successful films ever made.
The pattern observable across this essay's historical survey has several consistent features. The condemned content is, almost always in the cases reviewed here, associated with a demographic that has less cultural power than the people doing the condemning. This generalization is an observation about the cases traced in this essay rather than a formal sociological finding; the broader question of how aesthetic judgment interacts with class distinction has been addressed in depth by Bourdieu and the tradition following him.[127] Penny dreadfuls were read by the working class. Jazz was associated with Black culture. Comic books were children's entertainment. Rap was working class Black expression. The sexual content currently being condemned in AI creative communities is produced primarily by men, and specifically by men who are not embedded in the cultural institutions that would give their work the legitimacy markers that protect it from this kind of scrutiny. The pattern of whose content gets condemned and whose does not is not random. It consistently tracks the distribution of cultural power.
The condemned content is also almost always associated with a form of pleasure that the people doing the condemning either do not share or find illegitimate. The aesthetic preferences of the condemners are elevated into ethical standards and then applied to content that serves different aesthetic preferences in ways that make the different preferences appear to be moral failings rather than differences in taste. This conversion of taste into ethics is the central rhetorical operation of cultural gatekeeping and it has been performed so consistently and so unsuccessfully throughout cultural history that its continued performance in each new iteration requires an explanation that the performers do not provide.
The vindication rate of aesthetic-taste-based cultural gatekeeping about ostensibly victimless content is, in the historical record this essay has traced, extremely low. The content survives. The condemners are forgotten or remembered with embarrassment. The audiences who were supposed to be corrupted by the condemned content turn out to be perfectly capable of distinguishing between what they consume and how they live. The generation that grew up with video nasties did not become violent. The generation that grew up with gangsta rap did not all become criminals. The generation that grew up with violent video games did not produce the epidemic of violence that was predicted. The opponent of this view owes an account of what makes the current case relevantly different from its historical precedents, and that account has not yet been produced.
The claim being made here is not that cultural concern is always misplaced. Some concerns about mass-produced content have been validated by subsequent evidence: the targeting of children by tobacco marketing, the public health consequences of leaded gasoline, the asymmetries of subprime lending. Such cases share features that distinguish them from taste-based gatekeeping. They involve concrete demonstrable harm to identifiable victims, the harm is measurable and its mechanism is specifiable, and the response is calibrated to the demonstrated harm rather than to the aesthetic character of the product. The pattern this essay has traced applies to the narrower class of aesthetic-taste-based concern about expression that lacks demonstrable identifiable victims. The argument here is not that concern itself is suspect but that concern of this specific form has a poor vindication record, and that distinguishing which kind of concern one is looking at is prior to evaluating whether it is warranted.
Tropes as Architecture
The specific charge leveled against AI generated sexual content in current debates, that it is trope driven and therefore empty, that its use of familiar conventions rather than original invention demonstrates the absence of genuine creative intent, misunderstands what tropes are and what their persistence tells us about human psychology.
Tropes are not failures of imagination. They are the accumulated solutions to recurring storytelling problems, conventions that have been tested across generations of creative production and have survived because they work. The idealized female figure in a music video, the romantic hero in a novel, the wise mentor in an adventure story, the monster in a horror film, these are tropes because they connect to something real in human psychology that has been consistently responsive to these figures across the history of the forms in which they appear. Their persistence is evidence of their function. They keep appearing because they keep working. They keep working because they address needs and activate responses that are features of human psychology rather than accidents of specific cultural moments.
The idealized female figure in particular has been a persistent trope in visual art across every culture and every historical period for which we have evidence, for reasons that the evolutionary psychology reviewed earlier in this essay explains. She is not a product of any specific patriarchal culture's construction of femininity. She is a response to a feature of human male visual psychology that precedes the specific cultural forms in which it manifests. The trope persists because the underlying psychological response to it persists. Treating its persistence as evidence of cultural pathology requires ignoring both the evolutionary account of why it exists and the historical evidence that it has existed everywhere humans have had the tools to produce visual art.
The demand that creative work transcend its tropes in order to be considered legitimate is a demand that has been applied consistently to popular and lowbrow forms and inconsistently to prestige forms. The romance novel operates on tropes as conventional and predictable as any AI generated music video and it is consumed by millions of readers who find in those conventions exactly the pleasures they are looking for. The thriller operates on tropes of menace and resolution that have not changed substantially since the genre was established. The horror film operates on tropes of threat, vulnerability and survival that Aristotle would have recognized. None of these genres is required to transcend its conventions to deserve cultural space. The demand that sexual content transcend its conventions to deserve cultural space is a special requirement applied to a specific category of content for reasons that are not made explicit because making them explicit would reveal that they are aesthetic preferences rather than ethical principles.
Sexual Gratification as a Legitimate Intent
The most fundamental dishonesty in the current argument against AI generated sexual content is the implicit hierarchy it establishes between intellectual purpose and pleasurable purpose, between content that means something and content that feels something, between art that challenges and entertainment that satisfies. This hierarchy has no coherent philosophical foundation and its consistent application to sexual content while other forms of pleasurable entertainment escape scrutiny reveals that it is not a principle at all. It is a prejudice.
Sexual gratification is a legitimate human purpose. It has been recognized as such across most of human history and across most human cultures in ways that do not require the elaborate justification that contemporary cultural criticism demands of sexual content. The Kama Sutra treated sexual pleasure as one of the three legitimate aims of human life alongside virtue and material prosperity.[128] Classical Greek culture celebrated sexual pleasure as a component of the good life. The Shunga tradition produced explicit sexual imagery for the pleasure of its audience without requiring that pleasure to justify itself through a non-sexual purpose. The demand that sexual content earn its legitimacy by demonstrating purposes beyond the sexual is not an ancient wisdom. It is a specific cultural inheritance from a theological tradition that viewed sexual pleasure with suspicion and that inheritance has been secularized into aesthetic theory without anyone examining whether the underlying suspicion of pleasure is justified.[129]
The man who generates an AI image of an attractive woman because he finds attractive women visually pleasurable is doing something that is explained by evolutionary psychology, documented in cross cultural research, practiced across the full span of human history, and harmless to any identifiable person. The requirement that he justify this activity by demonstrating that it serves purposes beyond his own pleasure is a requirement that is applied to no other form of entertainment. The man who watches an action film because he finds action films exciting is not required to demonstrate that his consumption serves purposes beyond the excitement. The woman who reads a romance novel because she finds romantic fantasy pleasurable is not required to justify her reading by reference to its educational benefits. The specific demand that sexual content justify itself through non-sexual purpose is applied to sexual content because of a residual cultural discomfort with sexual pleasure that has not been honestly examined or defended but is simply assumed.
The intellectual hierarchy that places challenging art above entertaining art and entertaining art above gratificatory art is also a class hierarchy and it consistently maps onto the distribution of cultural power in ways that should prompt skepticism. Prestige art is the art of educated elites.[130] Popular entertainment is the art of general audiences. Sexual entertainment is the art of people who want what they want without apology. The further down this hierarchy a form of content sits the less cultural protection it receives and the more readily its audiences are characterized as people whose tastes reveal something unflattering about them. The characterization is not neutral. It is the exercise of cultural power by people who have it against people who do not and it is dressed in the language of ethics to make it harder to identify as what it actually is.
The Censorship Argument Properly Framed
The legal argument about censorship has been made carefully in this essay and does not need to be rehearsed at length here. The platform has the right to decline what it chooses. The festival is not a government. The rejection email is not a prison sentence. These things are true and important and the people who invoke First Amendment absolutism against curatorial decisions are misreading both the law and the nature of editorial discretion.
The cultural argument about censorship is different and it deserves to be made on its own terms. The cost of normalizing content suppression based on discomfort is not primarily a legal cost. It is a cultural cost that operates through the gradual contraction of what can be expressed without social consequence, through the chilling effect of knowing that certain kinds of content will be met not with aesthetic criticism but with moral characterization of the creator, through the establishment of community norms that treat the feelings of people who find content offensive as automatically outweighing the freedom of people who produce it.[131]
This cost is real even when no law is broken and no legal censorship occurs. The history of what gets made is not only the history of what is legally permitted. It is also the history of what people feel safe making, what they are willing to submit to community spaces, what they believe will be received as legitimate creative expression rather than as evidence of a disordered character. When community standards are set in ways that treat a category of male creative expression as presumptively suspect, the effect on what gets made is not limited to the specific content that gets declined. It extends to the content that never gets made because the creator has internalized the message that their desires are not the kind of desires that deserve expression.
The normalization of content suppression based on discomfort also establishes a principle that does not stay where it is placed. A community that decides it has the authority to decline content because it reflects a worldview the community finds objectionable has established a mechanism that will be used by whoever controls the community's standards for whatever content they find objectionable. The principle that discomfort justifies exclusion does not arrive with built-in limits. It expands to cover whatever the people wielding it are uncomfortable with. The history of content restriction bears this out consistently. Powers established to address genuinely harmful content are applied to content that is merely unfamiliar, merely offensive to specific sensibilities, merely the expression of desires that do not fit the preferences of the people making the decisions.
The Question Underneath All the Others
Whose desire gets treated as human? This is the question that the entire debate about AI generated sexual content is circling without landing on directly and it is the question that needs to be answered before any of the others can be answered honestly.
The desire of the woman who produces sexual content is treated as human in current discourse. It is complex. It reflects economic calculation, creative expression, bodily autonomy and a full range of motivations that deserve respect even from people who find the content she produces objectionable. The desire of the man who consumes equivalent content is not treated as human in the same way. It is treated as a symptom of something, of cultural conditioning, of fragile masculinity, of a worldview that reduces women to objects, of a position in a radicalization pipeline. His desire does not get the complexity that her desire gets. He does not get the presumption of full humanity that she gets. He gets analysis where she gets recognition.
This asymmetry is not a minor inconsistency in an otherwise coherent framework. It is the framework's central feature and it reveals what the framework is actually about. A framework genuinely committed to the humanity of all persons would extend the same complexity, the same presumption of good faith, the same recognition of legitimate need, to the desire of the man consuming sexual imagery that it extends to the woman producing it. The failure to do this is not an oversight. It is a choice, and the choice reveals that the framework's commitment to human dignity is conditional on the gender of the person whose dignity is at stake.
The defense of male sexual desire as a legitimate human feature rather than a social pathology is not a defense of the men who commit violence against women. Those men exist. Their actions are real and serious and deserve serious response. The defense is of the overwhelming majority of men whose desire takes the form of visual pleasure in idealized female imagery, who consume that imagery, who sometimes produce it using whatever tools are available to them, who do not commit violence, who do not radicalize, who do not treat real women as objects in the morally serious sense that Nussbaum's framework identifies, and who are told by the current cultural argument that their desire is nevertheless a symptom of a worldview that connects them to the men who do commit violence.
Those men deserve to have their desire recognized as human. Not as exceptional. Not as justified by special pleading. Human. The same way the desire of any other person is recognized as human. The standard applied in both directions. The complexity extended without condition. The presumption of full humanity not withdrawn because the desire is male and visual and not sufficiently apologetic about what it wants.
That is the argument underneath all the others. Everything else in this essay, the history, the psychology, the philosophy, the critique of the pipeline logic and the objectification framework and the unfalsifiable loop, is in service of that argument. It is not a complicated argument. It has been complicated by frameworks that benefit from its complication. Stated plainly it is this: male sexual desire is a feature of human psychology, not a pathology of male character, and the art that expresses it, in all its forms from the transcendent to the gratuitous, deserves the same freedom that any other form of human expression deserves.
That freedom is not contingent on your approval. It never was.
X. Conclusion: AI Nasties and the Verdict of History
In 1984 the British Parliament passed the Video Recordings Act with the confidence of people who believed they were responding to a genuine and documented threat to public welfare. The films they were responding to were described in parliamentary debate as among the most depraved material ever produced, as direct causes of violent and antisocial behavior in the young people exposed to them, as evidence of a cultural moment that had crossed a threshold requiring legislative intervention. The science was cited. The case studies were presented. The moral entrepreneurs had done their work and the work had succeeded. The Video Recordings Act received royal assent in July 1984 and the video nasties were removed from circulation in a country that had decided it knew what harm looked like and was not going to tolerate it.[132]
Forty years later The Evil Dead has a West End musical. Cannibal Holocaust is assigned reading in film courses.[133] The researchers whose work was cited in parliament to justify the legislation have not been vindicated. The epidemic of violence the films were supposed to produce did not materialize. The generation that grew up with access to the nasties before the Act and found ways to access them after it did not become more violent than the generations that preceded it. Youth violence in Britain declined substantially in the decades following the panic, which is precisely the opposite of what the harm thesis predicted.[134] The people who drove the panic are remembered, when they are remembered at all, as cautionary examples of what happens when aesthetic disgust acquires the confidence of established science and the authority of legislation.
This is what happens, every time, in the historical record of aesthetic-taste-based panics about ostensibly victimless content that this essay has traced. The pattern's consistency creates a high prior for skepticism; it does not establish a logical guarantee about any specific future case, and the current moment's advocates owe an account of why this case is relevantly different that has not yet been provided.
The Pattern Prediction
Based on that record the prediction for the current panic about AI generated sexual content is not difficult to make and it does not require unusual powers of foresight. It requires only the willingness to take the historical pattern seriously rather than treating the current moment as the exception that finally justifies the confidence that previous moments did not.
The AI generated sexual content that is currently the subject of community moderation policies, moral characterization of creators and arguments about radicalization pipelines will be available, consumed and largely unremarkable within a decade. The frameworks being deployed to justify its suppression, the objectification theory stripped of Nussbaum's own qualifications, the pipeline argument extended beyond its evidentiary warrant, the unfalsifiable loop that reads any disagreement as confirmation of the thesis, will be examined by researchers in media studies and the sociology of moral panics and found to exhibit the same features that every previous instance of this pattern exhibited.[135] The confidence will not be vindicated. The predictions will not be borne out. The content will survive and the certainty of the people who wanted it gone will not.
This prediction is not made with pleasure. It would be genuinely preferable if the current argument were different in kind from its predecessors, if the evidence were as solid as the confidence implies, if the causal chain from male sexual expression to male sexual violence were established with the rigor that a claim of that weight requires. It is not. The evidence is contested where it is not simply absent. The causal chain is assumed where it is not demonstrated. The confidence is a feature of the moral panic rather than a feature of the underlying evidence and the historical record of moral panics does not suggest that confidence of this kind is a reliable indicator of accuracy.
The people driving the current argument are not stupid. They are not malicious. They are not indifferent to evidence. They are doing what people in the grip of a moral panic have always done: responding to genuine concern about genuine phenomena with a framework that overextends what the genuine concern and the genuine phenomena actually establish, and doing so with the certainty that comes from believing that this time the concern is serious enough to justify the confidence. The concern about online misogyny is genuine.[136] The concern about the treatment of women in digital spaces is genuine. The concern about the pipeline from online communities to real world violence in specific documented cases is genuine. None of this makes the extension of those genuine concerns into a general theory of male sexuality that connects the AI generated music video to the rape coordination network an extension the evidence supports. Genuine concern does not produce accurate analysis automatically. The history of moral panics is the history of genuine concern producing inaccurate analysis with great confidence and significant consequences for the people the analysis targets.
The Distinction That Actually Matters
This essay has drawn one distinction repeatedly and it deserves its final, clearest statement here because it is the distinction that the current debate most consistently collapses and whose collapse does the most damage to clear thinking about what the actual problems are.
On one side of the distinction: real people, real non-consent, real harm. The woman whose likeness is used to generate sexual imagery without her knowledge or agreement.[137] The person whose identity is impersonated in generated content. The child whose abuse is documented in material that is produced, distributed and consumed as sexual content. These are genuine harms with genuine victims and they deserve serious, specific and proportionate responses. The law is increasingly recognizing this. The technology companies enabling these harms are increasingly being held accountable for them. The responses are appropriate to what the harms actually are.
The Pelicot case, in which Dominique Pelicot was convicted in 2024 of drugging his wife Gisele and coordinating her rape by dozens of men recruited from an online forum, is the kind of case this distinction is designed to handle.[138] The online forum from which Pelicot recruited was itself a specific, identifiable and organized space whose participants had a concrete relationship to an identifiable harm and a duty of inquiry about what they were being asked to participate in; that is not the diffuse cultural adjacency the bystander argument extends its logic to, and treating it as such would collapse exactly the distinction this essay has drawn. It belongs on the side of real people, real non-consent and real harm, and it deserves the response the French court delivered: prosecution, conviction, lengthy sentences and the refusal of the anonymity the defendants sought. What the case does not establish, though it is sometimes deployed as though it did, is that the existence of coordinated networks of men organized around the rape of a specific real woman is evidence that ordinary male consumption of generated sexual imagery is continuous with such networks. Pelicot's co-conspirators were a population selected on outcome, not a representative sample of men who find attractive women attractive. The methodological point from the pipeline discussion applies here with particular force: the existence of a harmful pattern in the studied population does not establish its operation in the unselected population. The distinction between the victims of real coordinated violence and the men generating a music video image holds under the weight of the hardest case, not because the hardest case is any less terrible, but because the response it warrants is specific to what it is.
On the other side: generated figures, consensual adult fantasy, artistic transgression, the idealized female character in a music video produced by a man who wanted to make something and had access to tools that made it possible. No real person. No non-consent. No identifiable victim. No demonstrated causal mechanism connecting the content to harm sustained by any real person in any real context. These are not the same category as what sits on the other side of the distinction and treating them as the same category is not a precaution against harm. It is an error. An error with a specific shape that this essay has traced across the full length of human cultural history and that has a consistent track record of producing confident suppression, significant cost to the people whose expression is suppressed, and eventual embarrassment for the people who drove the suppression.
The distinction is not difficult. It is not subtle. It does not require specialist knowledge to draw. It requires only the willingness to ask, in each specific case, whether there is a real person being harmed by this specific content, and to resist the temptation to answer that question by reference to theoretical frameworks that assume the harm rather than demonstrating it. When the answer is yes there is a real person being harmed, respond seriously. When the answer is no, the content involves generated figures and consenting adults, resist the confidence that tells you the harm is there even when you cannot point to a victim. That confidence has been wrong before. It has been wrong every time.
Male Sexuality as a Human Feature
The defense of male sexuality as a normal feature of human psychology rather than a social pathology is not a defense of the men who commit violence against women. Those men exist. Their violence is real. Their victims deserve justice, recognition and every response that genuine harm demands. This essay has not argued otherwise at any point and does not do so now.
The defense is of the overwhelming majority of men whose sexuality takes the form that human male sexuality characteristically takes: visual, oriented toward physical attractiveness, capable of idealization and fantasy that is independent of real world relationships and intentions, expressed through the consumption and sometimes the production of imagery that depicts what they find attractive. These men do not commit violence against women at any rate that the evidence connects to their consumption of sexual imagery. They are fathers, partners, friends, colleagues, members of communities who find attractive women attractive and who sometimes make or consume content that reflects this and who live their entire lives without the content predicting any harm to any real person.
The message currently being sent to these men in AI creative communities and in broader cultural discourse is that their desire is a symptom. That finding an attractive woman attractive and wanting to make or consume imagery of attractive women is participation in a system of harm even when no specific harm can be identified, even when the content involves no real person, even when the man's actual behavior toward actual women gives no evidence of the attitudes the framework attributes to him. The message is that his desire requires justification that other desires do not require, that his pleasure is presumptively suspect in ways that other pleasures are not, that the discomfort of people who find his desire objectionable is a more legitimate consideration than his freedom to express it.
This message is wrong. It is wrong empirically, because the evidence does not establish the causal chain from male visual sexuality to harm that would justify it. It is wrong historically, because the same message has been delivered about male sexuality and its creative expressions across the full span of this essay's historical survey and has not been vindicated. It is wrong philosophically, because it extends to male sexual desire a suspicion that it withholds from other forms of desire without a principled basis for the asymmetry. And it is wrong in its consequences, which brings us to the final consideration.
The Cost of Getting This Wrong
What happens to a culture that tells a generation of men that their desires are symptoms of a disease?
The question is not rhetorical. It has answers that are worth thinking about seriously rather than dismissing as concern trolling or defensive deflection. A culture that pathologizes normal male sexuality does not thereby eliminate that sexuality. It drives it underground, into spaces where it cannot be examined, discussed or understood, where the men who experience it cannot encounter the context that would allow them to situate it within a broader understanding of what it is and why it exists. The man who is told that his attraction to idealized female imagery marks him as a potential participant in a radicalization pipeline does not thereby stop finding attractive women attractive. He stops being able to discuss what he finds attractive in spaces where that discussion could be productive. He learns that his desires are not the kind of desires that can be expressed openly without moral consequence. He learns that the cultural institutions around him view his sexuality as a threat rather than a feature of his humanity.
This is not a recipe for producing men who are safer to women. It is a recipe for producing men who are alienated from the cultural frameworks that might otherwise help them understand their desires in ways that contextualize rather than suppress them. The research on what actually reduces sexual violence does not point to the suppression of fantasy and sexual expression as a mechanism. It points to the cultivation of empathy, to the development of healthy relationship models, to the reduction of the social isolation and status anxiety that the radicalization research identifies as the actual drivers of the pathway from frustration to extremism.[139] None of these are served by telling men that their visual attraction to attractive women is evidence of a dangerous worldview.
The cost of getting this wrong falls most heavily on the men who are most vulnerable to the message, who are most likely to internalize the characterization of their desires as pathological, who have the least access to the social and intellectual resources that would allow them to evaluate the characterization critically. These are not the men who are most likely to become radicalized. They are the men who are most likely to become isolated, ashamed and confused about desires that are normal features of human male psychology, and whose isolation and shame the radicalization research identifies as conditions that make them more rather than less susceptible to the ideological frameworks the pipeline ends in. The suppression of normal male sexuality in the name of preventing radicalization may be producing exactly the conditions the radicalization research identifies as drivers of the pathway. This is a prediction warranted by parallel reasoning rather than a demonstrated empirical finding, and in fairness it is owed the same evidentiary burden this essay has applied to the claims it criticizes. Neither the harm thesis nor its critique fully escapes that burden; both are owed the precision that the current research base has not yet provided.[140]
There is also a cost that falls on culture more broadly. A culture that cannot distinguish between the expression of desire and the enactment of harm, that treats the idealized female figure in a music video and the coordinated sexual assault network as points on the same moral continuum, has lost something important in its ability to think clearly about what harm actually is and what responses to it are actually proportionate. The capacity to make this distinction is not a luxury. It is a prerequisite for responding effectively to genuine harms rather than performing concern about theoretical ones while the real problems continue undisturbed.
The real problems are specific. They have specific victims and specific mechanisms. They deserve specific responses calibrated to what they actually are. The generalization of those specific problems into a theory of male sexuality as inherently suspect does not serve the real victims of real harms. It serves the confidence of people who have decided what the problem is and are selecting evidence accordingly. Those people are not villains. They are doing what people in moral panics have always done. But the victims of real harms deserve better than a panic. They deserve the precision that genuine response to genuine harm requires and that precision is impossible when everything that makes men uncomfortable is folded into the same moral category as the things that actually hurt people.
The Knife
The Venus of Hohle Fels was carved thirty-five thousand years ago by a human hand working with a stone tool to produce an image of a woman's body with exaggerated sexual characteristics.[141] We do not know who made her. We do not know precisely what she was for. We know she was made with considerable care and effort, that she was small enough to be carried or worn, and that she is currently the oldest known figurative sculpture in human history. Whatever else she was, she was an expression of something in human psychology that found the female form worth the considerable effort of rendering it in stone.
The man who prompted an AI tool to generate an attractive female figure for his music video last Tuesday is separated from the person who carved the Venus of Hohle Fels by thirty-five thousand years of human history, by every development in art, technology, culture and moral philosophy that separates us from our Paleolithic ancestors. He is not separated from them by the desire. The desire is recognizably continuous. The technology is incomparably more sophisticated. The continuity of the desire is not erased by the difference in medium or scale, even as those differences raise legitimate questions of their own.
Every generation believes it is confronting a new moral crisis. Every generation produces people who are certain that a specific form of cultural expression has crossed a threshold that previous expressions did not cross and that this threshold crossing justifies a response that previous expressions did not require. Every generation's certainty looks, from the next generation's vantage point, like a combination of genuine concern and catastrophic overconfidence about what the concern established and what responses it warranted.
This generation is not different. The AI panic will be remembered the way the video nasty panic is remembered. The content will survive. The science will be found to have been presented with more confidence than it warranted. The predictions will not be vindicated. The people who drove the panic will not be remembered as the people who finally got it right after centuries of false alarms.
What will remain, as it has always remained, is the desire. The human need to make images of what we find beautiful and to find in those images something that serves needs the world does not always provide. That need is thirty-five thousand years old. It survived every previous attempt to pathologize it. It will survive this one.
The only question that has ever mattered is not whether the desire exists. It does and it will. The question is whether we extend to the people who experience it the recognition of full humanity that desire in all its forms deserves, or whether we decide, one more time, with one more set of citations and one more vocabulary of harm, that this particular desire, in this particular person, is the exception that finally justifies the confidence that has always, in the end, been wrong.
We have been wrong before. We should be more careful about being certain again.
[1] Cohen, Stanley. Folk Devils and Moral Panics. London: MacGibbon and Kee, 1972.
[2] For the framing of generative AI as an object of moral and ethical concern, see Crawford, Kate. Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. New Haven: Yale University Press, 2021; and Benjamin, Ruha. Race After Technology: Abolitionist Tools for the New Jim Code. Cambridge: Polity Press, 2019.
[3] On the discursive construction of AI as a source of novel cultural harm, see Bender, Emily M., Timnit Gebru, Angelina McMillan-Major, and Shmargaret Shmitchell. "On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?" In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency (2021): 610–623; and Marwick, Alice E. "Gender, Sexuality, and Social Media." In The Social Media Handbook, edited by Jeremy Hunsinger and Theresa Senft. New York: Routledge, 2013.
[4] Attwood, Feona. Porn.com: Making Sense of Online Pornography. New York: Peter Lang Publishing, 2010.
[5] On how networked media make previously marginal content newly visible and thereby invite disproportionate moral response, see boyd, danah. It’s Complicated: The Social Lives of Networked Teens. New Haven: Yale University Press, 2014; and Gillespie, Tarleton. Custodians of the Internet: Platforms, Content Moderation, and the Hidden Decisions That Shape Social Media. New Haven: Yale University Press, 2018.
[6] McGlynn, Clare, Erika Rackley, and Ruth Houghton. "Beyond 'Revenge Porn': The Continuum of Image-Based Sexual Abuse." Feminist Legal Studies 25, no. 1 (2017): 25–46.
[7] Barker, Martin, and Julian Petley, eds. Ill Effects: The Media Violence Debate. 2nd ed. London: Routledge, 2001.
[8] On the framing of male sexuality as inherently problematic within much contemporary feminist and critical-masculinities discourse, see Connell, R. W. Masculinities. 2nd ed. Berkeley: University of California Press, 2005; and hooks, bell. The Will to Change: Men, Masculinity, and Love. New York: Atria, 2004.
[9] For critical analyses of how media and scholarly discourse have constructed male heterosexual desire as a site of pathology rather than agency, see Kimmel, Michael. Guyland: The Perilous World Where Boys Become Men. New York: Harper, 2008; and Gill, Rosalind. "Empowerment/Sexism: Figuring Female Sexual Agency in Contemporary Advertising." Feminism & Psychology 18, no. 1 (2008): 35–60.
[10] Petley, Julian. Film and Video Censorship in Contemporary Britain. Edinburgh: Edinburgh University Press, 2011.
[11] Kerekes, David, and David Slater. See No Evil: Banned Films and Video Controversy. Manchester: Headpress, 2000.
[12] Mathijs, Ernest, and Xavier Mendik, eds. The Cult Film Reader. Maidenhead: Open University Press, 2008.
Newman, Kim. Nightmare Movies: Horror on Screen Since the 1960s. London: Bloomsbury, 2011.
[13] Goodall, Mark. Sweet & Savage: The World Through the Shockumentary Film Lens. London: Headpress, 2006.
[14] Ebert, Roger. "I Spit on Your Grave." Chicago Sun-Times, January 1, 1980.
[15] Whitehouse, Mary. Mightier Than the Sword. Eastbourne: Kingsway Publications, 1985.
Tracey, Michael, and David Morrison. Whitehouse. London: Macmillan, 1979.
[16] Becker, Howard S. Outsiders: Studies in the Sociology of Deviance. New York: Free Press, 1963.
[17] Barker, Martin. The Video Nasties: Freedom and Censorship in the Media. London: Pluto Press, 1984.
[18] For a direct media history of the Daily Mail and its role as an engine of moral campaigning in twentieth-century Britain, see Addison, Adrian. Mail Men: The Unauthorised Story of the Daily Mail, the Paper That Divided and Conquered Britain. London: Atlantic Books, 2017; and Bingham, Adrian. Family Newspapers? Sex, Private Life, and the British Popular Press 1918–1978. Oxford: Oxford University Press, 2009.
[19] Barker, Martin, and Julian Petley, eds. Ill Effects: The Media Violence Debate. 2nd ed. London: Routledge, 2001.
[20] Video Recordings Act 1984. c. 39. London: Her Majesty's Stationery Office.
[21] Freedman, Jonathan L. Media Violence and Its Effect on Aggression: Assessing the Scientific Evidence. Toronto: University of Toronto Press, 2002.
[22] Barker, Martin, and Julian Petley, eds. Ill Effects: The Media Violence Debate. 2nd ed. London: Routledge, 2001.
[23] Ferguson, Christopher J., and John Kilburn. "The Public Health Risks of Media Violence: A Meta-Analytic Review." Journal of Pediatrics 154, no. 5 (2009): 759–763.
[24] Farrington, David P. "Trends in Violence in England and Wales." In Violence in Europe, edited by Sophie Body-Gendrot and Peter Spierenburg. New York: Springer, 2008.
[25] For the broader long-run decline in interpersonal violence across developed societies, see Pinker, Steven. The Better Angels of Our Nature: Why Violence Has Declined. New York: Viking, 2011.
[26] On the sustained decline in youth and general violent crime across the period during which home video, heavy metal, and violent video games became mass-market products, see Zimring, Franklin E. The Great American Crime Decline. New York: Oxford University Press, 2007; and Eisner, Manuel. "Long-Term Historical Trends in Violent Crime." Crime and Justice 30 (2003): 83–142.
[27] Conkey, Margaret W. "Humans as Materialists and Symbolists: Image Making in the Upper Paleolithic." In The Origin and Evolution of Humans and Humanness, edited by D. Tab Rasmussen. Boston: Jones and Bartlett, 1993.
Dixson, Alan F., and Barnaby J. Dixson. "Venus Figurines of the European Paleolithic: Symbols of Fertility or Attractiveness?" Journal of Anthropology 2011 (2011): 1–11.
[28] Conard, Nicholas J. "A Female Figurine from the Basal Aurignacian of Hohle Fels Cave in Southwestern Germany." Nature 459 (2009): 248–252.
[29] Clarke, John R. Looking at Lovemaking: Constructions of Sexuality in Roman Art, 100 BC–AD 250. Berkeley: University of California Press, 1998.
Johns, Catherine. Sex or Symbol: Erotic Images of Greece and Rome. London: British Museum Press, 1982.
[30] Lissarrague, François. The Aesthetics of the Greek Banquet: Images of Wine and Ritual. Princeton: Princeton University Press, 1990.
Davidson, James. Courtesans and Fishcakes: The Consuming Passions of Classical Athens. London: HarperCollins, 1997.
[31] Screech, Timon. Sex and the Floating World: Erotic Images in Japan 1700–1820. London: Reaktion Books, 1999.
Hayakawa, Monta, ed. Shunga: Sex and Humor in Japanese Art and Literature. Tokyo: Kodansha, 2008.
[32] Vatsyayana. Kama Sutra. Translated by Wendy Doniger and Sudhir Kakar. Oxford: Oxford University Press, 2002.
[33] Obscene Publications Act 1857. 20 & 21 Vict. c. 83. London: Her Majesty's Stationery Office.
Regina v Hicklin (1868) LR 3 QB 360.
[34] Vanderham, Paul. James Joyce and Censorship: The Trials of Ulysses. New York: New York University Press, 1998.
[35] United States v. One Book Called Ulysses, 5 F. Supp. 182 (S.D.N.Y. 1933).
[36] Rolph, C.H., ed. The Trial of Lady Chatterley. London: Penguin Books, 1961.
Obscene Publications Act 1959. 7 & 8 Eliz. 2. c. 66. London: Her Majesty's Stationery Office.
[37] De Grazia, Edward. Girls Lean Back Everywhere: The Law of Obscenity and the Assault on Genius. New York: Random House, 1992.
Grove Press, Inc. v. Gerstein, 378 U.S. 577 (1964).
[38] Ferlinghetti, Lawrence, and Nancy J. Peters. Literary San Francisco. San Francisco: City Lights Books, 1980.
People v. Ferlinghetti, Municipal Court, City and County of San Francisco (1957).
Morgan, Bill. The Typewriter Is Holy: The Complete, Uncensored History of the Beat Generation. New York: Free Press, 2010.
[39] Burroughs, William S. Naked Lunch: The Restored Text. Edited by James Grauerholz and Barry Miles. New York: Grove Press, 2001.
Miles, Barry. William Burroughs: El Hombre Invisible. London: Virgin Books, 1992.
[40] Attorney General v. A Book Named "Naked Lunch," 351 Mass. 298 (1966).
Memoirs v. Massachusetts, 383 U.S. 413 (1966).
[41] Westbrook, Robert B. "'I Want a Girl, Just Like the Girl That Married Harry James': American Women and the Problem of Political Obligation in World War II." American Quarterly 42, no. 4 (1990): 587–614.
[42] Watts, Steven. Mr. Playboy: Hugh Hefner and the American Dream. Hoboken: John Wiley and Sons, 2008.
[43] Miller, Russell. Bunny: The Real Story of Playboy. London: Michael Joseph, 1984.
[44] Hearn, Marcus, and Alan Barnes. The Hammer Story. London: Titan Books, 1997.
[45] Koven, Mikel J. La Dolce Morte: Vernacular Cinema and the Italian Giallo Film. Lanham: Scarecrow Press, 2006.
[46] Clover, Carol J. Men, Women, and Chain Saws: Gender in the Modern Horror Film. Princeton: Princeton University Press, 1992.
[47] Deflem, Mathieu. "The Politics of Pop Music Censorship: A Historical Overview." In Popular Music and Society, edited by Brian Longhurst. Cambridge: Polity Press, 2007.
Martin, Linda, and Kerry Segrave. Anti-Rock: The Opposition to Rock 'n' Roll. Hamden: Archon Books, 1988.
[48] Zappa, Frank, with Peter Occhiogrosso. The Real Frank Zappa Book. New York: Poseidon Press, 1989.
United States Senate Committee on Commerce, Science and Transportation. Contents of Music and the Lyrics of Records: Hearing Before the Committee. 99th Cong., 1st sess., September 19, 1985. Washington: Government Printing Office, 1985.
[49] United States Senate Committee on Commerce, Science and Transportation. Contents of Music and the Lyrics of Records: Hearing Before the Committee. 99th Cong., 1st sess., September 19, 1985. Washington: Government Printing Office, 1985.
[50] Marsh, Dave. Fifty Ways to Fight Censorship. New York: Thunder's Mouth Press, 1991.
Skyywalker Records, Inc. v. Navarro, 739 F. Supp. 578 (S.D. Fla. 1990).
Luke Records, Inc. v. Navarro, 960 F.2d 134 (11th Cir. 1992).
[51] Gates, Henry Louis, Jr. The Signifying Monkey: A Theory of African American Literary Criticism. New York: Oxford University Press, 1988.
[52] Cullen, Dave. Columbine. New York: Twelve, 2009.
[53] Anderson, Craig A., and Brad J. Bushman. "Effects of Violent Video Games on Aggressive Behavior, Aggressive Cognition, Aggressive Affect, Physiological Arousal, and Prosocial Behavior: A Meta-Analytic Review of the Scientific Literature." Psychological Science 12, no. 5 (2001): 353–359.
[54] Ferguson, Christopher J. "The Good, the Bad and the Ugly: A Meta-Analytic Review of Positive and Negative Effects of Violent Video Games." Psychiatric Quarterly 78, no. 4 (2007): 309–316.
[55] Brown v. Entertainment Merchants Association, 564 U.S. 786 (2011).
[56] Bourdieu, Pierre. The Logic of Practice. Translated by Richard Nice. Stanford: Stanford University Press, 1990.
Bourdieu, Pierre. Outline of a Theory of Practice. Translated by Richard Nice. Cambridge: Cambridge University Press, 1977.
[57] Mulvey, Laura. "Visual Pleasure and Narrative Cinema." Screen 16, no. 3 (1975): 6–18.
[58] Lacan, Jacques. The Four Fundamental Concepts of Psychoanalysis. Translated by Alan Sheridan. New York: Norton, 1978.
[59] Kaplan, E. Ann. Women and Film: Both Sides of the Camera. New York: Methuen, 1983.
Silverman, Kaja. The Subject of Semiotics. New York: Oxford University Press, 1983.
Williams, Linda. Hard Core: Power, Pleasure and the Frenzy of the Visible. Berkeley: University of California Press, 1989.
[60] Mulvey, Laura. "Afterthoughts on 'Visual Pleasure and Narrative Cinema' Inspired by King Vidor's Duel in the Sun." Framework 15/16/17 (1981): 12–15.
[61] Locke, John. An Essay Concerning Human Understanding. Edited by Peter H. Nidditch. Oxford: Clarendon Press, 1975. First published 1689.
[62] Pinker, Steven. The Blank Slate: The Modern Denial of Human Nature. New York: Viking, 2002.
[63] Ferguson, Christopher J. "Do Angry Birds Make for Angry Children? A Meta-Analysis of Video Game Influences on Children's and Adolescents' Aggression, Mental Health, Prosocial Behavior and Academic Performance." Perspectives on Psychological Science 10, no. 5 (2015): 646–666.
Ferguson, Christopher J., and John Kilburn. "The Public Health Risks of Media Violence: A Meta-Analytic Review." Journal of Pediatrics 154, no. 5 (2009): 759–763.
[64] Bailey, J. Michael, Michael P. Dunne, and Nicholas G. Martin. "Genetic and Environmental Influences on Sexual Orientation and Its Correlates in an Australian Twin Sample." Journal of Personality and Social Psychology 78, no. 3 (2000): 524–536.
Långström, Niklas, Qazi Rahman, Eva Carlström, and Paul Lichtenstein. "Genetic and Environmental Effects on Same-Sex Sexual Behavior: A Population Study of Twins in Sweden." Archives of Sexual Behavior 39, no. 1 (2010): 75–80.
[65] Hamann, Stephan, Rebecca A. Herman, Carla L. Nolan, and Kim Wallen. "Men and Women Differ in Amygdala Response to Visual Sexual Stimuli." Nature Neuroscience 7, no. 4 (2004): 411–416.
[66] Fine, Cordelia. Delusions of Gender: How Our Minds, Society, and Neurosexism Create Difference. New York: W.W. Norton, 2010.
Fine, Cordelia. Testosterone Rex: Myths of Sex, Science, and Society. New York: W.W. Norton, 2017.
[67] Tooby, John, and Leda Cosmides. "The Psychological Foundations of Culture." In The Adapted Mind: Evolutionary Psychology and the Generation of Culture, edited by Jerome H. Barkow, Leda Cosmides, and John Tooby. New York: Oxford University Press, 1992.
[68] Gould, Stephen Jay, and Richard C. Lewontin. "The Spandrels of San Marco and the Panglossian Paradigm: A Critique of the Adaptationist Programme." Proceedings of the Royal Society of London B 205, no. 1161 (1979): 581–598.
[69] Buss, David M. "Sex Differences in Human Mate Preferences: Evolutionary Hypotheses Tested in 37 Cultures." Behavioral and Brain Sciences 12, no. 1 (1989): 1–14.
[70] Lippa, Richard A. "Sex Differences in Sex Drive, Sociosexuality, and Height across 53 Nations: Testing Evolutionary and Social Structural Theories." Archives of Sexual Behavior 38, no. 5 (2009): 631–651.
Schmitt, David P. "Sociosexuality from Argentina to Zimbabwe: A 48-Nation Study of Sex, Culture, and Strategies of Human Mating." Behavioral and Brain Sciences 28, no. 2 (2005): 247–275.
[71] Singh, Devendra. "Adaptive Significance of Female Physical Attractiveness: Role of Waist-to-Hip Ratio." Journal of Personality and Social Psychology 65, no. 2 (1993): 293–307.
Thornhill, Randy, and Steven W. Gangestad. "Facial Attractiveness." Trends in Cognitive Sciences 3, no. 12 (1999): 452–460.
Aharon, Itzhak, Nancy Etcoff, Dan Ariely, Christopher F. Chabris, Ethan O'Connor, and Hans C. Breiter. "Beautiful Faces Have Variable Reward Value: fMRI and Behavioral Evidence." Neuron 32, no. 3 (2001): 537–551.
[72] Singh, Devendra. "Adaptive Significance of Female Physical Attractiveness: Role of Waist-to-Hip Ratio." Journal of Personality and Social Psychology 65, no. 2 (1993): 293–307.
[73] Baumeister, Roy F. "Gender Differences in Erotic Plasticity: The Female Sex Drive as Socially Flexible and Responsive." Psychological Bulletin 126, no. 3 (2000): 347–374.
[74] Kinsey, Alfred C., Wardell B. Pomeroy, and Clyde E. Martin. Sexual Behavior in the Human Male. Philadelphia: W.B. Saunders, 1948.
Kinsey, Alfred C., Wardell B. Pomeroy, Clyde E. Martin, and Paul H. Gebhard. Sexual Behavior in the Human Female. Philadelphia: W.B. Saunders, 1953.
[75] Critelli, Joseph W., and Jenny M. Bivona. "Women's Erotic Rape Fantasies: An Evaluation of Theory and Research." Journal of Sex Research 45, no. 1 (2008): 57–70.
Joyal, Christian C., Amélie Cossette, and Vanessa Lapierre. "What Exactly Is an Unusual Sexual Fantasy?" Journal of Sexual Medicine 12, no. 2 (2015): 328–340.
[76] Sagarin, Brad J., Bert Cutler, Nadine Cutler, Kimberly A. Lawler-Sagarin, and Leslie Matuszewich. "Hormonal Changes and Couple Bonding in Consensual Sadomasochistic Activity." Archives of Sexual Behavior 38, no. 2 (2009): 186–200.
[77] Diamond, Milton, and Ayako Uchiyama. "Pornography, Rape, and Sex Crimes in Japan." International Journal of Law and Psychiatry 22, no. 1 (1999): 1–22.
[78] Diamond, Milton, and Eva Jozifkova. "Pornography and Sex Crimes in the Czech Republic." Archives of Sexual Behavior 40, no. 5 (2011): 1037–1043.
[79] D'Amato, Anthony. "Porn Up, Rape Down." Northwestern Public Law Research Paper No. 913, Northwestern University School of Law, 2006.
[80] Kutchinsky, Berl. "Pornography and Rape: Theory and Practice? Evidence from Crime Data in Four Countries Where Pornography Is Easily Available." International Journal of Law and Psychiatry 14, no. 1–2 (1991): 47–64.
[81] Rothstein, Hannah R., Alexander J. Sutton, and Michael Borenstein, eds. Publication Bias in Meta-Analysis: Prevention, Assessment and Adjustments. Chichester: John Wiley and Sons, 2005.
[82] Seto, Michael C. "Pedophilia." Annual Review of Clinical Psychology 5 (2009): 391–407.
Seto, Michael C. "Child Pornography Use and Internet Solicitation in the Diagnosis of Pedophilia." Archives of Sexual Behavior 39, no. 3 (2010): 591–593.
[83] Nussbaum, Martha C. "Objectification." Philosophy and Public Affairs 24, no. 4 (1995): 249–291.
[84] Papadaki, Lina. "Sexual Objectification." In Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta. Stanford: Stanford University, 2019. https://plato.stanford.edu/entries/feminism-objectification/.
[85] Burt, Martha R. "Cultural Myths and Supports for Rape." Journal of Personality and Social Psychology 38, no. 2 (1980): 217–230.
Lonsway, Kimberly A., and Louise F. Fitzgerald. "Rape Myths: In Review." Psychology of Women Quarterly 18, no. 2 (1994): 133–164.
[86] Freedman, Jonathan L. Media Violence and Its Effect on Aggression: Assessing the Scientific Evidence. Toronto: University of Toronto Press, 2002.
Tedeschi, James T., and Richard B. Felson. Violence, Aggression, and Coercive Actions. Washington: American Psychological Association, 1994.
[87] Malamuth, Neil M., Tamara Addison, and Mary Koss. "Pornography and Sexual Aggression: Are There Reliable Effects and Can We Understand Them?" Annual Review of Sex Research 11 (2000): 26–91.
Malamuth, Neil M., and James V. P. Check. "The Effects of Mass Media Exposure on Acceptance of Violence against Women: A Field Experiment." Journal of Research in Personality 15, no. 4 (1981): 436–446.
[88] McCauley, Clark, and Sophia Moskalenko. "Mechanisms of Political Radicalization: Pathways toward Terrorism." Terrorism and Political Violence 20, no. 3 (2008): 415–433.
Horgan, John. The Psychology of Terrorism. London: Routledge, 2005.
[89] Ferguson, Christopher J. "Pornography and Sexual Aggression: Can Meta-Analysis Find a Link?" Trauma, Violence, and Abuse 16, no. 2 (2015): 166–175.
Ferguson, Christopher J., and Richard D. Hartley. "The Pleasure Is Momentary…the Expense Damnable? The Influence of Pornography on Rape and Sexual Assault." Aggression and Violent Behavior 14, no. 5 (2009): 323–329.
[90] Regnerus, Mark, David Gordon, and Joseph Price. "Documenting Pornography Use in America: A Comparative Analysis of Methodological Approaches." Journal of Sex Research 53, no. 7 (2016): 873–881.
Covenant Eyes. Pornography Statistics: 250 Facts, Quotes, and Statistics about Pornography Use. Owosso: Covenant Eyes, 2015.
[91] Gigerenzer, Gerd. Reckoning with Risk: Learning to Live with Uncertainty. London: Penguin Books, 2002.
[92] Kenrick, Douglas T., Sara E. Gutierres, and Laurie L. Goldberg. "Influence of Popular Erotica on Judgments of Strangers and Mates." Journal of Experimental Social Psychology 25, no. 2 (1989): 159–167.
Buunk, Abraham P., Rosario Zurriaga, and Pilar González. "Social Comparison, Envy, and Depression." Motivation and Emotion 30 (2006): 211–220.
[93] Illouz, Eva. Hard-Core Romance: Fifty Shades of Grey, Best-Sellers, and Society. Chicago: University of Chicago Press, 2014.
Kamblé, Jayashree. Making Meaning in Popular Romance Fiction: An Epistemology. New York: Palgrave Macmillan, 2014.
[94] MacKinnon, Catharine A. Feminism Unmodified: Discourses on Life and Law. Cambridge: Harvard University Press, 1987.
Langton, Rae. Sexual Solipsism: Philosophical Essays on Pornography and Objectification. Oxford: Oxford University Press, 2009.
[95] On the asymmetric treatment of sex and violence in the American film-rating system, see Lewis, Jon. Hollywood v. Hard Core: How the Struggle Over Censorship Saved the Modern Film Industry. New York: NYU Press, 2000; and Sandler, Kevin S. The Naked Truth: Why Hollywood Doesn't Make X-Rated Movies. New Brunswick: Rutgers University Press, 2007.
[96] On ratings creep and the industrial tolerance of violence relative to sexual content, see Leone, Ron, and Nicole Houle. "21st Century Ratings Creep: PG-13 and the Demise of the Family-Oriented Film." Atlantic Journal of Communication 14, no. 4 (2006): 197–215; and Vaughn, Stephen. Freedom and Entertainment: Rating the Movies in an Age of New Media. Cambridge: Cambridge University Press, 2006.
[97] Going, Kate. "Misogyny, Masculinity, and the Isla Vista Killings." Violence Against Women 21, no. 3 (2015): 296–312.
Kalish, Rachel, and Michael Kimmel. "Suicide by Mass Murder: Masculinity, Aggrieved Entitlement, and Rampage School Shootings." Health Sociology Review 19, no. 4 (2010): 451–464.
Jaki, Sylvia, Tom De Smedt, Maja Gwóźdź, Rudresh Panchal, Alexander Rossa, and Guy De Pauw. "Online Hatred of Women in the Incels.me Forum: Linguistic Analysis and Automatic Detection." Journal of Language Aggression and Conflict 7, no. 2 (2019): 240–268.
[98] Coufal, Martin, Martin Fabi, Ondřej Kopeček, and Jan Šedivý. "Radicalization within a Network of Misogynist Extremists: A Case Study of an Incel Forum." Humanities and Social Sciences Communications 12 (2025): 1–15.
[99] UN Women and Institute for Strategic Dialogue. "What Is the Manosphere and Why Should We Care?" UN Women, 2025. https://www.unwomen.org/en/articles/explainer/what-is-the-manosphere-and-why-should-we-care.
[100] Ribeiro, Manoel Horta, Raphael Ottoni, Robert West, Virgílio A. F. Almeida, and Wagner Meira Jr. "Auditing Radicalization Pathways on YouTube." In Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency, 131–141. New York: Association for Computing Machinery, 2020.
Moonshot CVE. The Redirect Method: A Blueprint for Providing Counter-Narratives. London: Moonshot CVE, 2019.
[101] Kahneman, Daniel. Thinking, Fast and Slow. New York: Farrar, Straus and Giroux, 2011.
Gigerenzer, Gerd. Reckoning with Risk: Learning to Live with Uncertainty. London: Penguin Books, 2002.
[102] Borum, Randy. "Radicalization into Violent Extremism I: A Review of Social Science Theories." Journal of Strategic Security 4, no. 4 (2011): 7–36.
[103] Walton, Douglas. Informal Logic: A Pragmatic Approach. 2nd ed. Cambridge: Cambridge University Press, 2008.
[104] The strongest philosophical articulation of diffuse structural complicity, on which bystander-style arguments about cultural participation typically rest, is Young, Iris Marion. Responsibility for Justice. Oxford: Oxford University Press, 2011. For the extension of bystander reasoning into feminist media critique, see Dines, Gail. Pornland: How Porn Has Hijacked Our Sexuality. Boston: Beacon Press, 2010.
[105] The classical bystander literature is Latane, Bibb, and John M. Darley. The Unresponsive Bystander: Why Doesn't He Help? New York: Appleton-Century-Crofts, 1970. For the moral-philosophical framework governing what can and cannot be inferred from non-intervention, see Kutz, Christopher. Complicity: Ethics and Law for a Collective Age. Cambridge: Cambridge University Press, 2000.
[106] Stanaland, Adam J. L., Sarah E. Gaither, and Anna Gassman-Pines. "When Is Masculinity 'Fragile'? An Expectancy-Discrepancy-Threat Model." Personality and Social Psychology Review 27, no. 2 (2023): 114–138.
[107] Popper, Karl R. The Logic of Scientific Discovery. London: Hutchinson, 1959.
Popper, Karl R. Conjectures and Refutations: The Growth of Scientific Knowledge. London: Routledge, 1963.
[108] Grünbaum, Adolf. The Foundations of Psychoanalysis: A Philosophical Critique. Berkeley: University of California Press, 1984.
[109] Ferguson, Christopher J., and John Kilburn. "The Public Health Risks of Media Violence: A Meta-Analytic Review." Journal of Pediatrics 154, no. 5 (2009): 759–763.
[110] Wikipedia. "Grok Sexual Deepfake Scandal." Last modified 2026. https://en.wikipedia.org/wiki/Grok_sexual_deepfake_scandal.
Levin, Sam. "Musk's Grok AI Faces Scrutiny after Generating Sexual Deepfake Images." PBS NewsHour, January 2026. https://www.pbs.org/newshour/show/musks-grok-ai-faces-more-scrutiny-after-generating-sexual-deepfake-images.
[111] McGlynn, Clare, Erika Rackley, and Ruth Houghton. "Beyond 'Revenge Porn': The Continuum of Image-Based Sexual Abuse." Feminist Legal Studies 25, no. 1 (2017): 25–46.
Citron, Danielle Keats. Hate Crimes in Cyberspace. Cambridge: Harvard University Press, 2014.
[112] NYU Stern Center for Business and Human Rights. "The Grok Nudify Controversy Is Another Example of the Need for International AI Regulation." NYU Stern BHR, 2026. https://bhr.stern.nyu.edu/quick-take/the-grok-nudify-controversy-is-another-example-of-the-need-for-international-ai-regulation/.
[113] Franks, Mary Anne. The Cult of the Constitution. Stanford: Stanford University Press, 2019.
Chesney, Robert, and Danielle Keats Citron. "Deep Fakes: A Looming Challenge for Privacy, Democracy, and National Security." California Law Review 107, no. 6 (2019): 1753–1820.
[114] Seto, Michael C. Internet Sex Offenders. Washington: American Psychological Association, 2013.
Quayle, Ethel, and Max Taylor. "Child Seduction and Self-Representation on the Internet." CyberPsychology and Behavior 4, no. 5 (2001): 597–608.
[115] Knight First Amendment Institute. "Rereading Editorial Discretion." Knight First Amendment Institute at Columbia University, 2022. https://knightcolumbia.org/blog/rereading-editorial-discretion.
[116] Barker, Meg-John, and Julia Scheele. Queer: A Graphic History. London: Icon Books, 2016.
Ruberg, Bo, and Adrienne Shaw, eds. Queer Game Studies. Minneapolis: University of Minnesota Press, 2017.
Hardy, Kate, and Teela Sanders. "The Political Economy of 'Lap Dancing': Contested Spaces and Women's Work in the Illicit Erotic Economy." Work, Employment and Society 29, no. 1 (2015): 118–134.
[117] MacKinnon, Catharine A. Feminism Unmodified: Discourses on Life and Law. Cambridge: Harvard University Press, 1987.
Dworkin, Andrea. Pornography: Men Possessing Women. New York: Perigee Books, 1981.
[118] Chesney, Robert, and Danielle Keats Citron. "Deep Fakes: A Looming Challenge for Privacy, Democracy, and National Security." California Law Review 107, no. 6 (2019): 1753–1820.
Europol. Facing Reality? Law Enforcement and the Challenge of Deepfakes. The Hague: Europol Innovation Lab, 2022.
[119] Kaufman, Lloyd, and James Gunn. All I Need to Know about Filmmaking I Learned from the Toxic Avenger. New York: Berkley Boulevard Books, 1998.
Thrower, Stephen. Nightmare USA: The Untold Story of the Exploitation Independents. Godalming: FAB Press, 2007.
[120] Biskind, Peter. Down and Dirty Pictures: Miramax, Sundance, and the Rise of Independent Film. New York: Simon and Schuster, 2004.
[121] Aristotle. Poetics. Translated by Malcolm Heath. London: Penguin Books, 1996.
Nussbaum, Martha C. The Fragility of Goodness: Luck and Ethics in Greek Tragedy and Philosophy. Cambridge: Cambridge University Press, 1986.
[122] James, Louis. Fiction for the Working Man: A Study of the Literature Produced for the Working Classes in Early Victorian Urban England. London: Oxford University Press, 1963.
Springhall, John. Youth, Popular Culture and Moral Panics: Penny Gaffs to Gangsta-Rap, 1830–1996. London: Macmillan, 1998.
[123] Burroughs, William S., and Brion Gysin. The Third Mind. New York: Viking Press, 1978.
Miles, Barry. William Burroughs: El Hombre Invisible. London: Virgin Books, 1992.
[124] Attorney General v. A Book Named "Naked Lunch," 351 Mass. 298 (1966).
[125] Gioia, Ted. The History of Jazz. 2nd ed. New York: Oxford University Press, 2011.
Peretti, Burton W. The Creation of Jazz: Music, Race, and Culture in Urban America. Urbana: University of Illinois Press, 1992.
[126] Wertham, Fredric. Seduction of the Innocent. New York: Rinehart, 1954.
Hajdu, David. The Ten-Cent Plague: The Great Comic-Book Scare and How It Changed America. New York: Farrar, Straus and Giroux, 2008.
[127] For the classical analysis of how aesthetic judgment functions as a marker of class position and cultural power, see Bourdieu, Pierre. Distinction: A Social Critique of the Judgement of Taste. Translated by Richard Nice. Cambridge: Harvard University Press, 1984.
[128] Vatsyayana. Kama Sutra. Translated by Wendy Doniger and Sudhir Kakar. Oxford: Oxford University Press, 2002.
[129] Foucault, Michel. The History of Sexuality, Volume 1: An Introduction. Translated by Robert Hurley. New York: Pantheon Books, 1978.
Brown, Peter. The Body and Society: Men, Women, and Sexual Renunciation in Early Christianity. New York: Columbia University Press, 1988.
[130] Bourdieu, Pierre. Distinction: A Social Critique of the Judgement of Taste. Translated by Richard Nice. Cambridge: Harvard University Press, 1984.
[131] Sunstein, Cass R. Democracy and the Problem of Free Speech. New York: Free Press, 1993.
Chemerinsky, Erwin, and Howard Gillman. Free Speech on Campus. New Haven: Yale University Press, 2017.
[132] Video Recordings Act 1984. c. 39. London: Her Majesty's Stationery Office.
Petley, Julian. Film and Video Censorship in Contemporary Britain. Edinburgh: Edinburgh University Press, 2011.
[133] Kerekes, David, and David Slater. See No Evil: Banned Films and Video Controversy. Manchester: Headpress, 2000.
Mathijs, Ernest, and Jamie Sexton. Cult Cinema: An Introduction. Oxford: Wiley-Blackwell, 2011.
[134] Farrington, David P. "Trends in Violence in England and Wales." In Violence in Europe, edited by Sophie Body-Gendrot and Peter Spierenburg. New York: Springer, 2008.
Eisner, Manuel. "Long-Term Historical Trends in Violent Crime." Crime and Justice 30 (2003): 83–142.
[135] Cohen, Stanley. Folk Devils and Moral Panics. London: MacGibbon and Kee, 1972.
Goode, Erich, and Nachman Ben-Yehuda. Moral Panics: The Social Construction of Deviance. Oxford: Blackwell, 1994.
Thompson, Kenneth. Moral Panics. London: Routledge, 1998.
[136] Ging, Debbie. "Alphas, Betas, and Incels: Theorizing the Masculinities of the Manosphere." Men and Masculinities 22, no. 4 (2019): 638–657.
Horgan, John. The Psychology of Terrorism. London: Routledge, 2005.
[137] Citron, Danielle Keats. Hate Crimes in Cyberspace. Cambridge: Harvard University Press, 2014.
McGlynn, Clare, Erika Rackley, and Ruth Houghton. "Beyond 'Revenge Porn': The Continuum of Image-Based Sexual Abuse." Feminist Legal Studies 25, no. 1 (2017): 25–46.
[138] On the Pelicot trial, see Willsher, Kim. "Dominique Pelicot Found Guilty of Drugging and Raping Wife Gisele in Mass Rape Trial." The Guardian, December 19, 2024; and Sage, Adam. "Inside the Pelicot Trial: How a Husband's Betrayal Exposed a Network." The Times, December 2024.
[139] Flood, Michael. "The Harms of Pornography Exposure among Children and Young People." Child Abuse Review 18, no. 6 (2009): 384–400.
Casey, Erin A., and Paula S. Nurius. "Trends in the Prevalence and Characteristics of Sexual Violence: A Cohort Analysis." Violence and Victims 21, no. 5 (2006): 629–644.
[140] Borum, Randy. "Radicalization into Violent Extremism I: A Review of Social Science Theories." Journal of Strategic Security 4, no. 4 (2011): 7–36.
Kimmel, Michael. Healing from Hate: How Young Men Get Into and Out of Violent Extremism. Berkeley: University of California Press, 2018.
[141] Conard, Nicholas J. "A Female Figurine from the Basal Aurignacian of Hohle Fels Cave in Southwestern Germany." Nature 459 (2009): 248–252.
