My favorite intellectual over the past year is David Pinsof who’s had an odd career in contemporary psychology. He is an evolutionary social scientist at UCLA, a co-creator of Cards Against Humanity, the author of the Substack Everything Is Bullshit, a podcaster, a musician, and the developer of increasingly ambitious theories about political alliances, status, signaling, friendship, humor, and the stories we tell about our own motives.
The commercial success of Cards Against Humanity gave Pinsof income independent of the university. In a long 2026 interview with Zach Elwood, Pinsof said the game’s success freed him from the academic imperative to publish constantly, to optimize his work for obtaining a faculty position, and to choose projects according to their status on the academic labor market. He has acquired something like intellectual tenure without holding an academic tenure line. That explains both the freedom and the oddity of his oeuvre.
Pinsof grew up in Highland Park, Illinois, on Chicago’s North Shore. The future creators of Cards Against Humanity knew one another at Highland Park High School and graduated in 2005. A Daily Iowan profile published May 12, 2009 identified Pinsof as 21 years old.
Nothing in that 2009 profile suggests a young man who had decided to become an evolutionary psychologist. Pinsof was graduating from the University of Iowa with a double major in psychology and theatre arts. He was considering acting in Chicago or Los Angeles and mentioned theatre education. His undergraduate record holds another clue to the breadth of his interests. The University of Iowa Center for New Music lists a composition by David Pinsof called “Half Step”, performed in May 2008. Pinsof told me it was played at a composition seminar because his instructor’s girlfriend was a good enough pianist to perform it.
The music runs further back. In high school Pinsof played guitar in a band called Modus Operandi with Ben Hantoot and Eli Halpern, two of the friends who went on to create Cards Against Humanity with him. Their album Higher Shapes came out in 2005, the year they graduated, and a second, The Mourning After, followed in 2007. Pinsof later wrote the theme for Evolutionary Psychology (the Podcast) and scored short films made by his wife Annah Feinberg, who works in the entertainment industry. He has a body of unreleased music made with one of the bandmates that the two of them set aside for Cards Against Humanity, and he says a solo album sits on his bucket list.
In 2007 Pinsof wrote a ten-minute comedy called “The Bond They Shared,” performed in Iowa’s undergraduate play festival and still sold by Heuer Publishing, about three men whose shared obsession with James Bond is failing to hold their friendship together. Two decades before he began using game theory to ask why friendships form, persist, and dissolve, he was writing comedy about three men trying to keep one alive. Pinsof’s later work is preoccupied with audiences, performance, impression management, social perception, hidden intentions, and the gap between a man’s presentation and the machinery underneath it. The young actor and the mature theorist were drawn to related problems years before Pinsof had the evolutionary vocabulary he would later use to describe them. He eventually said so. Looking back on two years of the blog in July 2025, he referred to “the no-bullshit character I play.” The acting student built a public character and delivered the theory through it.
The other formative institution was a group of high-school friends. Josh Dillon, Daniel Dranove, Eli Halpern, Ben Hantoot, David Munk, David Pinsof, Max Temkin, and Eliot Weinstein kept meeting during college vacations after leaving Highland Park. They developed a deliberately transgressive party game that became Cards Against Humanity. The project went onto Kickstarter in December 2010 seeking $4,000 and raised $15,570. Kickstarter later published a case study of the game’s early success.
A Chicago magazine profile from 2014 described the eight founders as equal participants who kept making decisions together while pursuing other careers. Pinsof was then in graduate school in Los Angeles. The article caught him joining the others remotely to brainstorm jokes.
The card game produced serious money while he pursued his doctorate. He stayed with Cards Against Humanity part time, including data analysis and play-testing, while conducting psychological research.
Pinsof says comedy and psychology appeal to him for the same reason. Both require getting inside other people’s heads. Writing Cards Against Humanity also taught him a prose style. It rewarded short, provocative, cynical formulations that delivered an uncomfortable thought before the audience had time to protect itself from it. Years later he imported that voice into Everything Is Bullshit.
Pinsof entered UCLA and earned an M.A. in social psychology in 2015 and a Ph.D. in psychology in 2018. His dissertation, Evolutionary Origins of Political Ideology: Mating Strategies, Intergroup Conflict, and the Nature of Political Alliances, contains a curriculum vitae documenting his Iowa degree, UCLA training, publications, and works in progress. His graduate formation combined psychology with biological anthropology. His dissertation committee brought together Martie Haselton, Aaron Lukaszewski, H. Clark Barrett, and David O. Sears. Haselton chaired it. The committee reads as a map of his subsequent thinking. Haselton and Lukaszewski brought evolutionary psychology, Barrett evolutionary anthropology, and Sears political psychology.
The dissertation title captures a transition inside the work. Pinsof began with narrow questions about sexual politics and ended by asking whether political ideology had been misunderstood.
His first substantial publication, with Haselton, appeared in Psychological Science in 2016. “The Political Divide Over Same-Sex Marriage: Mating Strategies in Conflict?” proposed that opposition to same-sex marriage could partly reflect conflict between different sexual strategies. People averse to casual sex were likely to oppose same-sex marriage when they associated homosexuality with promiscuity. UCLA publicized the study under the headline “Why people oppose same-sex marriage.”
Pinsof and Haselton followed in 2017 with “The Effect of the Promiscuity Stereotype on Opposition to Gay Rights” in PLOS ONE. This time they varied what participants were told about gay male promiscuity. Among participants disinclined toward casual sex, information challenging the stereotype increased support for gay rights.
Pinsof doubted that people’s stated moral reasons explain their politics. A man might say his position follows from principle while the disposition that makes the principle attractive comes from another social interest.
That proposition put him into an early argument with Mark Hoffarth and John Jost. Hoffarth and Jost pointed to sexual minorities who opposed same-sex marriage as evidence that ideology can override apparent self-interest. Pinsof and Haselton answered in “When Self-Interest Contradicts Ideology: A Reply to Hoffarth and Jost.” The reply contains Alliance Theory in embryo. A conservative gay man, they suggested, might oppose same-sex marriage to show solidarity with fellow conservatives or Christians, to uphold principles that serve his group, or to conform to coalition norms. Where Jost saw ideology defeating self-interest, Pinsof moved the interest up a level, from the individual to the alliance. Everything he later built follows from that move. Self-interest need not mean immediate personal material advantage. Loyalty to allies, communities, religious groups, and political coalitions supplies powerful motives of its own.
The abortion work shows how slowly Pinsof develops an idea. At the Human Behavior and Evolution Society conference in 2016, Pinsof and Haselton presented “Pro-Life or Anti-Promiscuity? The Role of Mating Strategies in Opposition to Abortion.” The HBES program describes a study of 855 participants asking whether abortion opposition ran especially strong among people pursuing more restricted mating strategies when they believed abortion restrictions would discourage casual sex.
A descendant of that project did not surface as a major paper until 2026. Elias Acevedo, Pinsof, Jaimie Krems, and Haselton released the preprint “Concerns about Sexual Promiscuity Play an Underexplored Role in Opposition to Abortion,” now based on four studies and 3,631 participants, including data from before and after Dobbs and a nationally representative panel. It remains a preprint.
By 2018 Pinsof was presenting Alliance Theory in public. At that year’s Human Behavior and Evolution Society meeting he presented “Double Standards of the Political Mind: Empirical Support for the Alliance Theory.” The conference program describes an analysis using polling data from 1,215 Americans. The data came from Pulse of the Nation. Pinsof examined cases where Democrats and Republicans appeared willing to reverse general principles depending on which political coalition benefited from applying them.
The following year he presented “Equality for All, or for My Allies?” The 2019 HBES program records the presentation, while the dissertation gives the fuller analysis using the 2016 American National Election Studies sample of 4,271 respondents. The dissertation chapter pits Alliance Theory against accounts emphasizing egalitarianism, social dominance, moral foundations, and system justification.
There is a publication-history problem here. Neither the 1,215-person double-standards analysis nor the ANES study exists as a standalone peer-reviewed journal article. They appear in conference programs and, in the second case, the dissertation. Pinsof told me that some of the double-standards results reached the record through “Strange Bedfellows,” that he has never written a paper reviewing them, that he doubts they would fill one, and that he lacks the patience to try, though a few scholars have offered to take the lead. The ANES chapter has a better prospect. He is now working with Elias Acevedo, a doctoral student he co-advises with Jaimie Krems and Martie Haselton, to replicate and extend the findings for publication. Until that appears, some of Pinsof’s most pointed empirical attempts to separate the theory from competing explanations remain outside the ordinary journal literature.
The mature statement appeared with David O. Sears and Martie Haselton in Psychological Inquiry in 2023 as “Strange Bedfellows: The Alliance Theory of Political Belief Systems.” The title Pinsof had listed among manuscripts in preparation in 2018 had taken about five more years to become the published article.
Pinsof, Sears, and Haselton argue that political belief systems are largely constructed from historically contingent alliances. Parties bring together groups whose interests do not derive from a single philosophy. Advocates then build narratives and moral principles that make the coalition’s positions look coherent. Change the alliance structure and the apparent ideological logic changes with it.
Psychological Inquiry was the right venue because it surrounds a target article with commentaries. The reactions exposed where the argument becomes controversial. Ariel Malka and Christopher Federico argued in “Political Belief Systems Are Not Singularly Rooted in Alliance Psychology” that political belief systems cannot be reduced to alliance psychology, that stable dispositions and values constrain which positions fit together, and that the theory needed a test in which alliances and dispositions compete. Nichola Raihani and Quentin Atkinson argued in “It’s More Complicated Than That: Alliances Are One of Many Factors Shaping Political Belief Systems” that alliances constitute one causal influence among several. Yoav Ganzach and colleagues went further in “Political Ideology is Not Meaningfully Explained by Alliances and is Not Inconsistent with Attitudinal Inconsistencies.” The entire issue is a symposium on the scope and limits of Pinsof’s theory.
David Pietraszewski granted the premise and asked what follows. His commentary, “Seven Grand Challenges for Evolutionary Political Psychology or: Political Ideologies as Ad-Hoc Alliances… So What?”, posed seven challenges a predictive evolutionary political psychology would still have to meet once it accepted that ideologies are coalition bundles. The question goes past whether Pinsof has identified inconsistency and asks what the theory buys with the insight. Pinsof, Sears, and Haselton wrote in reply that they found the seven challenges challenging. Peter DeScioli of Stony Brook agreed that alliances belong at the foundation of political psychology and then turned the claim around. In “The Dangers of Alliances Caused the Evolution of Moral Principles,” he argued that impartial moral rules evolved as a way to choose sides in conflicts without the hazards of choosing by alliance. Where Pinsof looks behind moral principles and finds coalitions, DeScioli looks behind them and finds machinery built to restrain coalitions.
Antoine Marie and Michael Bang Petersen were more sympathetic but argued in “Speech Repression and Outrage from Orthodox Activists as Attempts to Mobilize Allies and Gain Status” that status competition deserved a larger role.
Pinsof, Sears, and Haselton answered the critics in “The Strange Epicycles of Political Psychology.” The reply offers a falsification criterion: alliances and rivalries, measured independently of ideological labels, should strongly predict political beliefs and votes.
John Jost’s 2024 “Both-Sideology Endangers Democracy and Social Science” addresses the larger dispute. Jost attacks arguments in political psychology that emphasize psychological symmetry between political camps. He says the baddies are on the right:
Liberal-leftists in the United States and elsewhere are demonstrably more committed than conservative-rightists to humanistic-egalitarian values, deliberative reasoning, and adherence to democratic norms. In Western societies, acts of authoritarian aggression, hate crimes, and political violence are overwhelmingly more likely to come from the right than the left. As a witness to Nazi atrocities, Kurt Lewin deeply understood the role of historical, economic, and political forces in human lives and the interdependence between democracy and social science. He rejected moral relativism and what I call “both-sideology” and offered a sophisticated critique of anti-democratic tendencies. There are perfectly understandable reasons why people—especially academics and journalists—would be tempted to draw parallels between the left and right, and indeed there are many similarities as well as dissimilarities between liberal-leftists and conservative-rightists. However, the uncritical adoption of both-sideology threatens liberal democracy itself and, with it, the only social science worth having. What we—as citizens and social scientists—need now is a renewal and revitalization of Lewin’s critical-emancipatory legacy before it is too late.
Social psychologist Mark Brandt responds to my email on Sep. 3, 2026:
Generally speaking, the challenge is separating alliances from shared worldviews. It is likely the case that we develop alliances because we share worldviews and we might also develop our worldviews because we share alliances. This is the case for an individual and their psychology, but also as a society and political system. In a more immediate way, we also infer worldviews from people’s alliances and we infer alliances from people’s worldviews.
That means that a successful experiment would find ways to pull these things apart, or to identify where they conflict with each other. I think this is challenging, but is in principle possible. As a quick example, one might investigate what happens when a person or group who typically shares both alliance and worldview stops sharing one of those factors (but keeps the other). It might also be possible to create minimal groups (i.e. they share no history/information) and see if people are more positive about those they share an alliance with vs their worldviews. In both of these examples, the devil is in the details.
The network ideas find that people’s identification with political parties and labels (e.g., “liberal”) are most central on average. This might be supportive of alliance theory. At the same time, issue positions relevant to groups (e.g., immigration policy) are not necessarily the most central on average, which might speak against alliance theory. A lot hinges on what constructs alliance theory believes measure alliances.
Lastly, I’m not sure I’d say that I personally disagree or not with David and alliance theory. The theory might well be right. My points are more about what other perspectives might predict and whether they are distinguished from alliance theory. We’ll see what the data from the Toronto study says.
I email Yoav Ganzach:
Your 2023 commentary was the most categorical in the Psychological Inquiry issue. Two questions. What empirical result would persuade you that Pinsof is right about the priority of alliances over dispositions? And has anything published since 2023, including the reply to commentaries, moved you in either direction? Pinsof tells me a preregistered study pitting alliances against dispositions is underway at the University of Toronto; I would welcome your view on what such a study would need to show.
He responds:
My critique of Pinsoff theory is not that it is incorrect. It is that it is dull.
First, roughly speaking, the model underlying the theory is: interests–>alliances–>ideology, which can be abbreviated as: interests–>ideology. But what’s new in that? (see for example Marlxist theory or rational choice theory and even dissonance theory).
Second, Alliance also explain alliance by “similarity”. However, since people may be similar in many ways, in order to be meaningful, Alliance theory must make a distinction between significant and insignificant similarities, and this is something it does not do. In fact, the most natural types of similarities that may lead to alliances are similarities in values and beliefs (i.e., dispositions).
I would very much like to hear your view about this critique.
I reply:
That is sharper than the commentary title let on, and “dull rather than wrong” is the objection I most wanted someone to make, because it is the one a theory cannot answer by adding studies.
On your first point, here is where I land. I agree that interests to ideology is old. What I think Pinsof adds is the middle term, and it does work only if you keep it. Marx’s coalitions were supposed to follow from class and did not. Rational choice cannot get anyone to the polls. Pinsof’s coalitions are contingent, and the theory predicts that when a coalition realigns, its members reverse principles within a few years without any change in interests or personality. The trade, Russia, and free-speech reversals since 2016 look like that prediction coming true. Marx would not expect a principle to flip in four years, and dissonance theory has nothing to say about which behavior a coalition demands. So I think the abbreviation drops the part that predicts. Whether Pinsof has done enough with that part is a fair question, and I would put it to him.
On your second point I think you are right, and I would like to represent it as the strongest objection the theory has drawn. If allies are picked by similarity and the theory does not say which similarities count, values are the obvious candidate and dispositions come back through the side door. The only escape I can see is for Pinsof to commit to the evolutionary version: the similarities that matter are the ones that predict side-taking in a conflict, kinship, proximity, cheap and hard-to-fake markers, and a record of past loyalty, with values downstream. Pietraszewski’s work on race as a coalition cue would be his evidence. That commits him to a prediction, that alliances built on side-taking markers reshape values while alliances built on shared values fail when the markers pull the other way. If he will not commit to it, your objection stands and the theory has no independent variable.
Two questions, if you have time. Would you accept “similarities that predict side-taking” as a repair, or do you think it still smuggles values in? And Pinsof tells me a preregistered study at the University of Toronto will pit alliances against dispositions; if alliance were measured by network and demography rather than by values, would you regard the result as bearing on your critique?
Ganzach responds:
This is the type of exchange that I seldom find in the psychological academic world, and I appreciate the opportunity.
You are right that my commentary was not sharp enough. I had to do that since editors stay aways from sharp short commentaries, especially when well known academics such as Sears and Haselton are involved. If one wants his work to be published he has to dilute his critique.
Pinsof et al present a speculative theory. They don’t bring systematic evidence, just a few examples. Furthermore, most, if not all, of their examples are from the Trump presidency. And as I write in my commentary “…the Trump period in the American conservative movement may very well be an exception rather than a rule, and deriving a general theory about the structure of ideology based on this example may be misleading. Unusual political periods make bad political theories in the same way that hard legal cases make bad law.”
I can’t see any similarities that predict side-taking between the evangelists in the south, the blue-color workers in the mid-west and Trump cronies in the east-const. In my view the Trump era political alliances all have to do with short-time interests, focused primarily around the Trump personality, and are doomed to disintegrate.
Pinsof has kept trying to give the alliance idea more formal structure. At HBES in 2025 he presented “Whose Side Are You On? The Evolutionary Game Theory of Alliances.” The conference program shows him moving from verbal descriptions of partisan inconsistency toward evolutionary game theory and the formal problem of alliance formation. The talk sat in a session on theory building alongside Daniel Conroy-Beam’s “We Need to Formalize Theory in Evolutionary Psychology,” so the turn is a movement inside the field. This may decide the future of the project. A theory that generates mathematical predictions stands in a stronger scientific position.
Another institution connected Pinsof’s commercial and scholarly lives. Cards Against Humanity created Pulse of the Nation, a project conducting nationally representative surveys of American opinion. The project says twelve polls were conducted in 2017 and 2018 using scientific sampling through Survey Sampling International, with data and cross-tabs released publicly. Pinsof served as its director and wrote the questions. He says the poll’s one requirement was that the questions be funny, and since double standards are funny, he used the slot to test predictions from the Alliance Theory he was then developing. It is almost a caricature of his peculiar career: a party-game company financing public-opinion research supervised by one of its founders, who was at the same time developing a theory of political belief, and who smuggled his hypotheses into the field under cover of a joke.
The company went through a serious workplace controversy in 2020 centered on allegations about cofounder Max Temkin and the company’s internal culture. Temkin stepped down. The company acknowledged failures and announced changes. One contemporary account is available here.
The decisive year in Pinsof’s development was 2023. A preprint of what became “The Evolution of Social Paradoxes” was circulating that year. Its preprint history lets us trace the paper before journal publication. That same March Pinsof began Everything Is Bullshit. “Strange Bedfellows” was circulating in the same period before journal publication. The academic work and the public philosophy developed together.
Pinsof has said that Everything Is Bullshit grew out of emails he wrote to himself arguing against claims that seemed obviously true to everyone else. UCLA later profiled him in a piece on academics who had migrated to Substack and described the audience around him as a “global gathering of Darwinist cynics.”
The blog contained twenty posts in 2023, twelve in 2024, ten in 2025, and six through August 18, 2026. The annual archives are available for 2023, 2024, 2025, and 2026. Forty-eight essays in under three and a half years make an oeuvre, and the falling rate has a stated cause. In his second-anniversary post Pinsof wrote that the blog had been a passion project on which he had taken no money, that he had lately felt demotivated, and that it was hard to justify the time when he had more serious work to do. He opened paid subscriptions with that post. Sidestack lists the newsletter at more than seven thousand free subscribers.
The early titles establish the territory at once: “Happiness Is Bullshit,” “The Meaning of Life Is Bullshit,” “What Are Ideologies All About?”, “Psychologizing Psychologists,” “Status Is Weird,” “Opinions Are Bullshit,” and “Your Fear of Mortality Is Bullshit.” The common move is a search for function. Pinsof starts with the explanation people consciously give for a behavior and asks what evolved social function might sit underneath it.
“Status Is Weird” is a key text. Humans seek status, Pinsof argues, but seeking it in the open discredits the seeker. Effective status competition has to disguise itself. Humility, authenticity, moral conviction, independence, and professed indifference to status all become ways of gaining status. The most effective competitor may experience himself as a man who is not competing.
Pinsof suggests that successful social signaling may require the motive to be hidden from the man doing the signaling. That idea has a pedigree running through Robert Trivers on self-deception and Robert Kurzban’s Why Everyone (Else) Is a Hypocrite.
The same idea became the backbone of “The Evolution of Social Paradoxes,” published online by American Psychologist in February 2026 after several years as a preprint. Pinsof examines paradoxical signals such as humility, conspicuous disregard for norms, and anonymous generosity. His central engine is recursive social inference. A signal can lose value when observers know the sender intended it as a signal. That creates incentives to communicate desirable qualities while concealing the signaling intention.
Nick Feltovich, Rick Harbaugh, and Ted To modeled countersignaling in 2002, the rich man who dresses down and the expert who skips his credentials, and Moshe Hoffman, Christian Hilbe, and Martin Nowak modeled anonymous giving and modesty as signal burying in 2018. Pinsof cites both. His claim is recursive. Once observers can infer the intention to advertise a trait as well as the trait, the intention becomes damaging information, and signals turn into cues, cues expose motives, and motives have to go underground again. The economists’ models are stable when everyone knows the game. Pinsof’s paradoxes collapse when the game becomes common knowledge.
Pinsof’s distinctive territory lies in the recursion. We compete to demonstrate that we are not competitive. We seek approval by displaying indifference to approval. We signal that we are not signaling. A man can gain status by making the status game disappear from view.
That recursive structure explains why the theatre student, the card-game designer, and the psychologist increasingly look like the same person. Each occupation depends on understanding what one mind thinks another mind thinks about it.
Pinsof is now trying to write the recursion down as mathematics. He shared with me the abstract of a paper in progress, tentatively titled “Human Status-Seeking as 4D Chess,” that models how each level of mindreading changes the game of status competition. At the first level, caring about others gets inferred and rewarded. At the second, status-seekers anticipate the inference and pretend to care. At the third, observers see through the pretending and penalize it. At the fourth, the seekers anticipate the penalty and pretend to be unpretentious, which produces the social paradoxes of the American Psychologist paper. The result is an arms race between pretentious signalers and cynical audiences. Pinsof argues the pretenders won, because fifth-order mindreading is too expensive to run and because being charmed by a skilled mindreader who is about to rise in status pays for itself. The paper promises accounts of charisma, authenticity, humility, performativity, and sacred values, and closes by adapting Dobzhansky: nothing in status psychology makes sense except in light of recursive mindreading. The 2023 preprint already invoked recursive mentalizing and common knowledge, so the mathematics is a reconstruction of an argument three years old, and the test of the formalism is whether it forced him to narrow anything. Whether the model survives reviewers remains to be seen. Pinsof says the humor paper carried heavy formalism too and that reviewer comments are pushing much of it into an appendix or out of the paper.
By 2024 the blog was acquiring its own conceptual vocabulary. “30 Useful Concepts about Bullshit” defined ideas such as “Darwinian cynicism,” self-bullshitting, confabulation, and Alliance Theory. Pinsof was building a general interpretive framework. He later converted much of the essay into an extended X thread, preserved by Thread Reader.
The titles run more extreme than the essays. “Democracy Is Bullshit” does not advocate abolishing democracy. Pinsof argues that democratic systems generate tribalism, rent-seeking, and perverse incentives while still producing outcomes preferable to autocracy. The rhetoric stays maximalist because maximalism is part of the Cards Against Humanity inheritance. The argument underneath has grown more qualified.
“Happiness Is Bullshit” offers a case study in both Pinsof’s ingenuity and his scientific vulnerability. His later “Happiness Is Bullshit Revisited” acknowledges that the original essay was among his most controversial. He develops the idea that happiness functions as something like a positive prediction-error signal, telling an organism that circumstances are improving relative to expectation.
The difficulty is that evolutionary function and conscious motive sit at different explanatory levels. Explaining why natural selection built a capacity for happiness does not show that a man cannot desire happiness. An ultimate evolutionary explanation can coexist with a proximate psychological one. Pinsof sometimes moves too fast from “this system evolved because it performed function X” to “the reason people say they pursue Y is bullshit.”
That belongs to a larger scientific problem. Pinsof is good at generating hidden-motive explanations. A framework that can accommodate every possible observation tells us less.
A person boasts. Status seeking.
A person acts humble. Concealed status seeking.
A person follows the crowd. Coalition maintenance.
A person rebels against the crowd. Status through independence.
A person admits craving status. Strategic candor.
A person denies craving status. Successful concealment.
Constructing a plausible evolutionary story for each behavior is easy. The challenge is specifying an observation that would show the story is wrong.
“The Evolution of Social Paradoxes” counts partly because Pinsof begins doing this. Instead of declaring humility a hidden status strategy, he tries to specify the information conditions under which concealed signals should work and fail. His evolutionary game-theory work on alliances is another attempt to produce constraints.
Pinsof is alert to the reflexive problem. If intellectuals are status seekers constructing flattering rationalizations, then the man becoming famous for exposing them is an intellectual seeking status. That is why “A Big Misunderstanding,” published in December 2025, ranks with “Status Is Weird.” Pinsof argues that intellectuals exaggerate how much social dysfunction results from ignorance, misinformation, bias, and misunderstanding. That diagnosis conveniently makes the people who manufacture knowledge indispensable. Pinsof wants us to look harder at incentives, status, alliances, competition, and institutional structure.
The essay drew a serious response from Dan Williams titled “We Are Confused, Maladapted Apes Who Need Enlightenment.” Williams agreed with much of Pinsof’s incentive-centered picture but argued that evolutionary mismatch, technological novelty, and human ignorance give knowledge-producing institutions a much larger role than Pinsof allows. Modern humans confront problems of nuclear weapons, financial markets, pandemics, and artificial intelligence that ancestral intuitions were not designed to solve. Sometimes misunderstanding is the problem.
Pinsof responded in Williams’s newsletter with “What Kind of Apes Are We?” The exchange is one of the best windows into Pinsof because the swagger falls away. He treated Williams as a serious good-faith interlocutor, acknowledged substantial agreement, and pushed back mainly on how far evolutionary mismatch should be taken. He also acknowledged his own attraction to Enlightenment ideals and the desire to understand the world better.
His 2025 essay “AI Doomerism Is Bullshit” supplied another stress test. Pinsof tried to identify a chain of assumptions behind catastrophic AI arguments and attacked them using an evolutionary conception of intelligence as a collection of specialized adaptive capacities instead of one generalized substance. The piece drew heavy resistance from readers who argued that AI-risk theories do not require all of Pinsof’s assumptions. Pinsof later wrote that of everything he had published it provoked the most backlash and bad faith, which surprised him, since claims that ideologies are propaganda and that nobody cares about improving the world had drawn nods and shrugs. The episode illustrates both his strength and his weakness. He can detect assumptions hidden inside apparently rigorous arguments, and he can also construct a target narrower than the strongest opposing case.
In 2026, “Everything Is Signaling” placed Pinsof in a lineage running through Robin Hanson and Kevin Simler, whose The Elephant in the Brain made the hidden-motive essay a genre. Much signaling, Pinsof argues, is defensive. Human beings spend much of their social effort avoiding signals that would let others classify them as stupid, disloyal, desperate, immoral, or incompetent.
His public intellectual environment has expanded alongside the Substack. In 2025 he began cohosting Evolutionary Psychology (the podcast) with David Pietraszewski, the critic who had asked what Alliance Theory buys with its insight. Pinsof did not solve the problem by defining the critic out of his circle. He brought him inside it. The show brings in researchers from across evolutionary behavioral science and discusses current work and the intellectual history of the field.
Pinsof has also moved through the podcast ecosystem around evolutionary psychology, rationalist-adjacent writing, signaling theory, and heterodox social science. He appeared on Infinite Loops, where the conversation touched on books including Harry Frankfurt’s On Bullshit, Edward Bernays’s Propaganda, Robert Cialdini’s Influence, Robert Wright’s The Moral Animal, Will Storr’s The Status Game, Howard Bloom’s The Lucifer Principle, and Douglas Hofstadter’s Gödel, Escher, Bach. He has also appeared on Chris Williamson’s Modern Wisdom.
He sits at an angle to rationalism. Much of his work runs against the rationalist habit of treating articulate reasons as privileged evidence about why people believe things. His “Utilitarianism Is Bullshit” turns the same explanatory suspicion on a philosophical tradition influential in effective-altruist and rationalist circles, and opens by admitting he used to be a utilitarian. His criticisms of AI doom, political ideology, and intellectualism reflect a deeper suspicion that explicit reasoning serves social functions with little relation to discovering truth.
His current UCLA work is widening again. The Social Minds Lab says Pinsof studies political attitudes, mating strategies, status, and signaling while using mathematical models to investigate alliance formation. The lab also describes him as leading work on the psychology of humor. Pinsof told me of three further papers. One, under review, tests Alliance Theory predictions about which groups people perceive as more discriminated against. A second, under review with Acevedo, offers an evolutionary functionalist account of awe. A third is a large preregistered study with a group at the University of Toronto that pits alliances against dispositions as predictors of political views.
The discrimination paper enters contested ground. In March 2026 Jake Womick, Kurt Gray, and nine colleagues published “Liberals and Conservatives See Different Victims” in Personality and Social Psychology Bulletin, twelve studies arguing that much of the moral disagreement between left and right comes from different assumptions about who is vulnerable to harm. Liberals see vulnerability as concentrated in groups; conservatives see it as spread across individuals. Those assumptions predicted moral judgments, implicit attitudes, and charitable giving, and could be manipulated. Alliance Theory can generate a similar surface result by a different route: allies get perceived as victims and rivals as perpetrators because the perception serves the coalition. Showing that partisans disagree about who is discriminated against will therefore prove little. The test is whether independently measured alliances predict the perceptions after vulnerability beliefs are allowed to compete.
The Toronto study is the answer to Malka and Federico. They asked for a contest between alliances and dispositions; Pinsof, Sears, and Haselton wrote in reply that independently measured alliances should strongly predict beliefs or the theory is in trouble; the preregistered study is that contest, with the predictions committed before the data arrive. If alliances outpredict personality and values, Alliance Theory moves from a redescription to a rival. If they do not, Pinsof will have run the study that showed it.
A 2025 UCLA profile of friendship research describes Pinsof’s emerging conception of friendship as an alliance problem. He asks why humans form, maintain, and terminate friendships and uses game theory to think about those decisions. Humor enters as a device that can cement friendship while marking insiders off from outsiders.
His 2026 preprint “All Mixed Up: A Solution to the Evolutionary Mystery of Humor” attempts an evolutionary account. Pinsof proposes that humor helps humans handle “mix-ups” in coordination. Misunderstandings, ambiguity, mistakes, and awkward violations can threaten social coordination. Shared laughter can establish common knowledge that both parties recognized the disruption and help defuse its costs. The ability to produce humor can therefore signal value as a coordination partner. The paper went through three public versions between February and April 2026 and grew from an intuition about awkwardness into a model with three linked adaptations, humor detection, mutual laughter, and partner choice, and a formal statement of when they could be favored. The revision history shows the direction of travel. Pinsof is trying to turn the sentence that makes you laugh into the account that says when you should. The humor work is also the one paper that joins his two careers. He has said it attracted him because it let him unite the comedy side and the psychology side of his life.
Pinsof is also working with Jaimie Krems and Daniel Sznycer on a UCLA Initiative to Study Hate project called “Tests of an Engineering Perspective of Hatred Towards Jews.” The project approaches hatred in functional terms by asking what psychological inputs activate hostile responses and what conditions alter them. That fits his preference for asking what a psychological system does before judging whether its output is morally desirable.
His publication record stays odd. ResearchGate’s Pinsof profile aggregates his papers, preprints, and related outputs, but his peer-reviewed corpus runs considerably smaller than his public intellectual footprint. His projects take five, eight, or ten years to mature. Alliance Theory appeared at conferences years before its target article. The abortion hypothesis returned a decade after its first presentation. “The Evolution of Social Paradoxes” circulated for several years before American Psychologist published it. He is now trying to formalize alliances, humor, and status itself, and says he thinks all science should move in that direction. Pinsof’s natural unit of work may be the research program.
The arrangement gives Pinsof permission to pursue large questions other researchers avoid because they do not yield rapid publications. It lets him combine political psychology, evolutionary theory, game theory, humor, philosophy, and internet writing without worrying whether a hiring committee will see a coherent specialization. It also removes some of the external constraints that make scholars narrow their claims. The same freedom that allows an original theory to develop can allow a favored explanation to expand until it covers everything.
Across all of his professional interests runs one question. Do people do things for the reasons they say they do? The young Pinsof asked whether people opposed same-sex marriage for the moral reasons they supplied. Alliance Theory asked whether political ideologies came from the values their adherents invoked. “Status Is Weird” asked whether noble conduct could conceal competition. “The Evolution of Social Paradoxes” asked how that concealment could become an evolved signaling strategy. “A Big Misunderstanding” asked whether intellectuals’ explanations of society conveniently magnify the importance of intellectuals. His happiness writing asks whether people understand the goals they believe they pursue.
Pinsof has spent a decade removing the object from his original question. What began as “Do people oppose this political policy for the reasons they say they do?” has become something close to “Does anybody do anything for the reasons they say they do?” That enlargement is the attraction of the project and its greatest vulnerability.
If Alliance Theory says coalitions influence political ideology, it persuades. If it says values are largely post hoc coalition propaganda, the evidentiary burden grows much heavier. If status theory says people sometimes disguise status seeking, it names a familiar phenomenon. If every act of humility, rebellion, generosity, authenticity, and indifference can be recoded as hidden status competition, it becomes hard to say what evidence could count against the theory.
Pinsof is at his most scientifically valuable when he disciplines his own cynicism. The formal treatment of concealed signals beats declaring everything a status game. Mathematical alliance models beat collecting partisan inconsistencies. A testable coordination theory of humor beats saying jokes help people bond. The movement from aphorism to model may decide whether Pinsof is remembered as a clever public intellectual or as the architect of a durable research program.
The popular persona conceals another feature of him. He enjoys explanation. He remains invested in doing science. He runs experiments and mathematical models, revises theories, engages serious critics, produces a research podcast, and keeps working inside UCLA with the money to walk away from it. His cynicism concerns human self-description.
That makes the title Everything Is Bullshit misleading. Pinsof believes that much of what human beings say about why they think, choose, and value what they do is psychologically superficial. His project is to uncover the machinery underneath.
The biography helped make that project possible. An actor-in-training became a social psychologist. A group of high-school friends produced a vulgar card game. The game became a fortune. The fortune freed one of its creators from the ordinary career structure of academia. The game taught him to compress uncomfortable observations into memorable sentences. Evolutionary psychology taught him to ask what those uncomfortable behaviors were designed to accomplish. Political psychology supplied the first cases. The internet supplied an audience on which to test ever broader versions of the theory.
He is an experiment in what happens when an evolutionary psychologist acquires tenure-like independence without tenure and then follows one explanatory instinct for more than a decade. The experiment now has a result to wait for. The theory that began by explaining away other people’s reasons is being made to compete against another explanation under rules announced in advance.
The Set
Start with the room. Every summer the Human Behavior and Evolution Society (HBES) meets in a hotel with bad carpet, and for four days a few hundred people argue about what the mind was built to do. The 2025 program runs to talks on alliance formation, mating strategy, coalitional cognition, disgust, gossip, and hatred. David Pinsof gave one called “Whose Side Are You On? The Evolutionary Game Theory of Alliances.” In the audience sit the people who can evaluate it. There are not many of them. Randall Collins observed in The Sociology of Philosophies that intellectual fields sustain only three to six live positions at a time, because attention is the scarce good and the network cannot hold more. The Pinsof set is a field of that size. Its members can be named, and most of them appear in the citations of one journal issue and the guest list of one podcast.
The inner ring is a laboratory lineage. Martie Haselton chaired his dissertation and has coauthored with him since 2016. David O. Sears supplied the political psychology and stands as senior author on the Alliance Theory papers. Aaron Lukaszewski and H. Clark Barrett sat on the committee. Jaimie Krems runs the Social Minds Lab where he now works. Daniel Sznycer joins him and Krems on the UCLA engineering-of-hatred project. Elias Acevedo, whom Pinsof co-advises with Krems and Haselton, carries the abortion paper, the awe paper, and the replication of the equality-for-all study. Tran Dinh and Steven Gangestad appear on the ovulatory-shift work. Behind all of them stand Leda Cosmides and John Tooby, whose Santa Barbara program supplies the grammar every one of these people speaks, along with Robert Trivers on self-deception, Amotz Zahavi on costly signals, Robert Kurzban on the modular and hypocritical mind, and Dan Sperber and Hugo Mercier on reasoning as an argumentative tool.
The second ring is the ring of worthy opponents. Most of them wrote for the 2023 issue of Psychological Inquiry. David Pietraszewski made the most dangerous concession. Suppose Alliance Theory is right, he asked in a commentary subtitled “Political Ideologies as Ad-Hoc Alliances… So What?”, and then posed seven problems a predictive evolutionary political psychology would still have to solve. Two years later he became Pinsof’s podcast cohost. Peter DeScioli agreed that alliances belong at the foundation of political psychology and then turned the claim around: in “The Dangers of Alliances Caused the Evolution of Moral Principles” he argued that impartial moral rules evolved as a way to choose sides without the hazards of choosing by alliance. Where Pinsof looks behind principles and finds coalitions, DeScioli looks behind principles and finds machinery built to restrain coalitions. Michael Bang Petersen and Antoine Marie argued in their commentary that status deserved more room than Alliance Theory gave it. Nichola Raihani and Quentin Atkinson said alliances are one factor among many. Ariel Malka and Christopher Federico said belief systems have other roots and asked for a study that would let dispositions and alliances compete. Yoav Ganzach said the theory explains little. Mark Hoffarth and John Jost were the first to push back in print, in 2017, and Jost’s later attack on both-sideology, which never names Pinsof, is the strongest statement of the position the set defines itself against in the wider argument over ideological symmetry. The Toronto collaboration that pits alliances against dispositions is the study Malka and Federico asked for, and it will give this ring something to shoot at.
The third ring is the public one. Dan Williams of Conspicuous Cognition is the closest thing Pinsof has to a sparring partner of equal weight. Robin Hanson at Overcoming Bias and Kevin Simler at Melting Asphalt wrote The Elephant in the Brain and made the hidden-motive essay a genre. Will Storr wrote The Status Game. Adam Mastroianni at Experimental History runs the same operation on psychology’s own institutions. Rob Henderson at his newsletter, Steve Stewart-Williams, Coltan Scrivner, Diana Fleischman, Geoffrey Miller, and Richard Hanania occupy adjacent stalls in the same market. Zach Elwood, Chris Williamson, and Jim O’Shaughnessy supply the microphones. Behind them stand the dead: La Rochefoucauld, Mandeville, Adam Smith of the Theory of Moral Sentiments, Nietzsche, Veblen, Goffman, and Harry Frankfurt, whose On Bullshit gave the set its title word.
What do they value? Function first. The question they ask of any behavior, belief, institution, or emotion is what it was built to accomplish, and they hold that the honest answer will rarely match the answer the actor gives. They value parsimony, and they value nerve. They value a man who can hold an unflattering hypothesis in his hand without dropping it. They value wit, and they treat a good joke as evidence, since a joke that lands proves that the audience recognized the hidden thing. They have started to value formal models more than they used to, which is why Pinsof’s move toward game theory reads as a bid for standing inside the field. He now says all science should go that way.
The immortality project: A true account of human nature that other people still use in fifty years. That is why the set treats popularization with suspicion even as it lives on Substack, and why the members who have crossed over keep insisting they are still doing science.
The status games follow from the hero system. The first is the cynicism arms race. If seeing through is the virtue, then the man who sees through one layer further wins the exchange. Hanson says medicine is not about health. Pinsof says the humble man is competing. The next man says the confession of competing is a competitive move. The game has no ceiling, which is a problem the set half-recognizes and half-enjoys. Pinsof’s “4D Chess” paper is an attempt to give the ceiling a number: the fifth order costs more than it pays.
The second is the reflexivity trump card. Admitting that your own essay is a status play cannot be answered, since any objection you raise afterward can be recoded as the play. Mary Douglas would recognize the structure. The confession functions as a purification rite, and once performed it licenses the polluting act. This is how a man can write about status seeking for three years while accumulating status and never appear to have contradicted himself.
The third is the freedom display. Pinsof has money from a card game and says so. Hanson has tenure. Mastroianni left. In a set that theorizes incentives, being visibly unbought is the highest form of capital. Robert Merton called disinterestedness a norm of science and noted that scientists have to perform it. Here the performance has been specialized: independence from career pressure is displayed, and it functions as a warrant. Trust me, I have nothing riding on this.
The fourth is hospitality to critics. When Dan Williams answered Pinsof with “We Are Confused, Maladapted Apes Who Need Enlightenment,” Pinsof replied in Williams’s own newsletter, conceded ground, and kept the disagreement narrow. Taking your strongest opponent seriously in his own house is a status move of a high order, and making a former critic your podcast cohost, as Pinsof did with Pietraszewski, is a higher one.
The fifth is boundary work. The set defines itself against three neighbors it can be seen arguing with. Against mainstream social psychology, which it regards as a machine for producing flattering findings about the researchers’ own coalition, and whose position Jost states most sharply. Against the rationalists and effective altruists, whose faith in articulate reasoning Pinsof attacks in “Utilitarianism Is Bullshit” and “AI Doomerism Is Bullshit.” And against the pop-evolutionary content business, the alpha-male podcast economy, which sells the same words with none of the discipline. Each boundary is patrolled by tone as much as by argument. You signal which side you are on by what you find funny.
The test the set applies to a joke is Pinsof’s: does the laugh prove that the audience recognized the hidden thing. A joke lands in that room when its punchline is a motive the speaker was concealing, and it dies when it asks the listener to share a piety, to hate a group, or to be shocked. Tone does the sorting because the set laughs deadpan and distrusts anything that sounds like applause. Some examples:
The mission-statement joke. “Starbucks exists to inspire and nurture the human spirit, one person, one cup, and one 40 percent margin at a time.” Pinsof used the real mission statement in “A Big Misunderstanding,” and the laugh is the gap between the stated purpose and the actual one. Every joke of this form works in the room: the university that exists to pursue truth and to keep its endowment, the charity that exists to end poverty and to keep existing.
The confession joke. “I’m not a status seeker. I’m just someone who happens to be right more often than the people above me.” The speaker exposes himself, which is the set’s cardinal virtue, and the punchline is the fourth-order move from the 4D Chess paper, pretending to be unpretentious. A close cousin: “I don’t care what people think of me. I’ve written a 4,000-word essay explaining why.”
The effective-altruism joke. “He gave a tenth of his income to shrimp welfare, then spent the rest of the evening explaining the math to a woman who left.” Pinsof made the shrimp-welfare crack. The joke works because it targets a piety of the set’s nearest neighbor and because the hidden motive, mating display dressed as impartial benevolence, is one the theory names.
The therapy-speak joke. “She set a boundary. It was around the last slice.” Or Feinberg’s version, the woman who is so good at therapy she wants to be her therapist’s favorite. The laugh is at a vocabulary of self-knowledge deployed in the service of the oldest appetites, and the set enjoys it because it confirms that the sophisticated are as mammalian as the rest.
The misinformation joke. “The misinformation experts have convened a panel to determine which of your beliefs are propaganda. Theirs are research.” That one patrols the boundary with mainstream social science, and it is funny in the room because it names the coalition.
The self-implicating Darwinian joke. “Natural selection built a mind that can prove the existence of hidden motives in everyone but its owner. Which is how you know it works.” The set’s humor at its best turns on itself, and the reflexive admission is, as the Set essay says, the confession that grants absolution.
Jokes the set does not find funny.
The clapter joke. Any punchline that is an applause line for a side. “Republicans want to ban books because they’ve never read one.” The room does not laugh, because the joke asks the listener to affirm a coalition, and the set’s status comes from noticing. The same joke with the parties reversed dies the same death. That is the symmetry the theory claims and the humor enforces.
The punch-down ethnic joke. The crude joke about a group’s stupidity or criminality, told with a grin. This is the alpha-male podcast economy’s material, and the set reads it as low-status because it reveals no hidden motive; it is the surface motive said out loud. The card game trades in exactly this content, but under rules, as a liturgy, and the laugh there is at the fact that everyone present can handle it. Told sincerely in a lab or on the podcast, it marks the teller as belonging to the wrong stall in the market.
The group-differences joke. “Nobody’s allowed to say it, but the test scores say it for them.” The room does not laugh here either, and the silence has a different quality from the silence at the punch-down joke. This one is close enough to a truth some of them hold that laughing would be a badge they cannot afford. That silence is the hereditarian boundary, and it is patrolled by tone precisely because it cannot be patrolled by argument.
The doomer joke. “Enjoy your coffee, the superintelligence will have paperclipped us by Thursday.” Pinsof’s readers are adjacent to the rationalists, and after the AI essay this joke splits the room; half laugh at the doomers and half go quiet because they are the doomers. The set found out in 2025 that this was its own sacred value, and a joke that reveals a sacred value in the audience gets the reaction Pinsof described: not laughter but backlash.
The just-so-story joke. “Evolutionary psychology: the field that can explain why you did that and why you’d have done the opposite.” Outsiders love this one. The set does not laugh, because it is the one joke that makes the set the target and whose hidden motive, unfalsifiability as career strategy, is the accusation the set most needs to be untrue. Everyone believes everything is bullshit except his own thing.
The earnest joke. Any joke that is kind. The set’s grammar treats sentiment as the raw material and never the punchline, so a joke whose warmth is the point, the kind that ends a commencement address, reads as a mission statement and gets the mission-statement treatment.
So the sorting rule is that the set laughs when a joke exposes a motive, and stops laughing at exactly the point where the motive exposed is its own. Pinsof’s humor theory says laughter creates common knowledge that both parties saw the mix-up. In the room, the mix-ups that can be made common knowledge are the neighbors’ pieties. The set’s own pieties, symmetry, the species line, the exemption of the observer, produce the silence a joke gets when everyone saw it and no one can say so.
Now the claims. The set writes as if it had no normative commitments and has several. Truth beats comfort, held with something close to religious force. Incentives beat exhortation, so the way to fix an institution is to change what it pays for.
The adaptationist commitments: The set holds that a species-typical psychological architecture exists, that it has content, and that the content can be discovered. It holds that psychological systems have functions, and that a function is a property of a trait. It holds that motives are objects that exist below awareness and can be wrong about themselves. Those are strong commitments, and the set makes them at the level of the species. At the level of the group it makes almost none: Alliance Theory’s point is that coalitions are historically contingent and that no party owns a timeless ideological essence. The set is fixed about human nature and fluid about human groups. In the AI argument Pinsof pushes the second habit further and denies that intelligence is a single substance, treating it as a bundle of special-purpose capacities. The set treats as machinery what rationalists treat as a quantity, and as contingent what political psychology treats as a trait.
The moral grammar: Sanctimony is the sin. Mood affiliation is the sin. Believing your own press is the sin. Boring is a sin, which is unusual for a scholarly community and traces to the card game. The virtues are candor about one’s own motives, updating in public, engaging the strongest version of an opponent, and taking a joke at one’s own expense. There is a confession, the reflexive admission, and it grants absolution. There is an excommunication, reserved for those caught moralizing without irony or defending a sacred cow. And there is a sacrament of controlled defilement. Cards Against Humanity is a liturgy in which a group of friends touch the forbidden together and prove by laughing that no one present is fragile. Douglas would say the game handles pollution the way rituals always do, by staging it under rules. The Substack performs the same office with moral material. The essays that end in “Status Is Weird” and “A Big Misunderstanding” invite the reader to handle a thought about himself that he would rather not touch, and the reward for handling it is membership.
A hero system that rewards nerve will reward the boldest claim. A moral grammar built on confession lets the confessor buy indulgences. A set that defines itself by seeing through everything has trouble specifying what it could fail to see through. Pinsof’s answer, so far, has been to reach for formal models, since “The Evolution of Social Paradoxes” tries to say when concealed signals should fail and the Toronto study tries to say when alliances should lose to dispositions. DeScioli’s answer is older: the thing the set cannot see through may be the moral rule itself, which on his account evolved because alliances are dangerous and people needed a way to choose sides without them. Whether the set follows Pinsof into models, or takes DeScioli’s point that some principles are not coalition propaganda, will determine whether it becomes a research program or remains a style.
The Four Questions
What coalition does he depend on for status and income.
Income and status come from different coalitions. Income is Cards Against Humanity, a company owned by eight Highland Park friends whose brand is progressive-coded, whose customers skew young and left, and whose public acts, the Trump-wall land purchase, the “Saves America” campaign that funded Pulse of the Nation, are jokes aimed at the right. His money comes from a coalition his theory would place on one side of the alliance structure he describes. Status comes from three places. The evolutionary psychology set, HBES, the Santa Barbara lineage, Haselton, Krems, Pietraszewski, is the one that can evaluate his papers and the one his formal turn is a bid inside. UCLA supplies the affiliation; he is a research scientist in Krems’s lab, not faculty, so the university’s hold on him is the lab’s goodwill. The Substack supplies seven thousand readers who are heterodox, rationalist-adjacent, and male, and who reward nerve. Sears is the bridge to mainstream political psychology and the reason the paper got a target-article slot.
Who does he risk angering if he speaks plainly.
His own set, on two subjects. The first is the hereditarian question. Evolutionary psychology has spent thirty years drawing a line between species-typical claims and group-difference claims, and its standing in the academy depends on the line holding. Alliance Theory applied would ask what coalition the line serves. He has not asked. The second is the coalition he takes money from. Alliance Theory applied to the comedy and entertainment world he and his wife work in, to the progressive institutions Cards Against Humanity donates to, or to the theatre activism of the Kilroys, would say the same things it says about Republicans and the FBI. He says them about utilitarians and AI doomers, who are near his audience but not in his house. Beyond that, plain speech about which coalitions produce hatred of Jews, the subject of the UCLA grant he shares with Krems and Sznycer, would anger someone whatever the answer, and the project’s language, an engineering perspective, inputs and conditions, is the register of a man who knows it.
Who benefits if his framing wins.
Evolutionary psychology annexes political psychology, and the annexation runs through him and Haselton. Anyone accused of hypocrisy benefits, since a theory in which everyone’s principles are coalition instruments makes the accusation costless; that is Jost’s objection, and it has force whether or not Jost’s own program is symmetric. The heterodox commentariat benefits because the theory licenses its habitual move, the recoding of moral claims as position-taking. Cards Against Humanity does not benefit and may lose, since a founder whose theory says democratic moral appeals are propaganda sits oddly beside a company whose public persona is moral appeal. Pinsof benefits in the way his own theory predicts: the man who sees through everyone gains the status that seeing through confers, and the formal turn lets him keep it while signaling that he never sought it.
What truths would cost him his position.
Saying, in print and with his own theory, what alliance the academic consensus on group differences serves would cost him the evolutionary psychology set’s protection and the lab. Saying that hostile feelings to out-groups is the logical result of different groups having different interests. Saying anything specific about what he saw inside Cards Against Humanity in the years before 2020 would cost him seven friendships and a share of a company, which is why the public record contains nothing from him on it and why his theory would predict exactly that silence. The truth he has already paid for is the AI doomerism essay, which cost him the rationalist wing of his readership and which he says drew more bad faith than anything he has written.
Buffered vs Porous Identity
Pinsof is a buffered self who has spent a decade demonstrating that the buffer is a fiction.
On the metaphysics he is as buffered as they come. No spirits, no meanings in things, a disenchanted cosmos in which the mind is a set of adaptations and the sacred is a psychological effect. He writes from the far end of Charles Taylor’s immanent frame, and the confidence of the blog’s titles is the confidence of a man for whom nothing outside the natural order can get in.
On the psychology he describes a porous self, with the porosity moved from spirits to other minds. The Pinsof self is permeated by its coalition, which sets its principles without consulting it; by its audience, whose anticipated judgments run through recursive mindreading and shape what it does before it acts; and by its own hidden motives, which it cannot see and which produce the story it tells about itself. His man is possessed, in Taylor’s sense: something not himself acts through him and he experiences the result as his own free judgment. The buffered self’s proudest possessions, its sincerity, its authenticity, its independence of mind, are on Pinsof’s account the places where the possession is most complete, because those are the signals that work only when the signaler believes them. “Status Is Weird” says that displayed independence from status is the highest status move. That is a claim that the buffer is a performance staged by porous machinery.
Taylor’s porous self is vulnerable to the sacred as a power in the world. Pinsof’s sacred values are the coalition’s social paradox, the commitments a group cannot admit are strategic because the admission would end their usefulness. The person who holds a sacred value is, on his account, possessed by his coalition and defended against knowing it. Sacredness has come back into the disenchanted world as a social force that enters the person from outside and cannot be reasoned with, which is what it was in the enchanted one. Pinsof re-enchants the self by way of the group.
Taylor’s point is that the buffered self is a historical achievement, an identity people came to inhabit, and one of its features is the view from outside, the disengaged observer who sees the machinery. Pinsof’s theory needs that observer. The Darwinian cynic who sees through everyone’s possession is himself the last buffered man, standing outside the coalition with his models. His theory says no such position exists, that the cynic has his own coalition and his own status audience and is possessed like anyone else. He concedes this in the reflexive passages and then goes on observing. Pinsof holds the buffered self as a method and denies it as a description, and the tension between the two is what his critics keep pointing at when they ask what would count against him. The buffered self can be wrong about the world. The porous self, in his account, cannot know when it is wrong about itself, and he has not said which self is writing the papers.
Annah Feinberg
Annah Feinberg is a playwright turned dramaturg turned screenwriter turned cartoonist turned graphic novelist, and the through-line in her work runs so close to her husband’s that the two of them look, from a distance, like a single project pursued by two methods.
Feinberg grew up in Deerfield, Illinois, the next town over from Pinsof’s Highland Park on Chicago’s North Shore. She took a BFA in theatre studies from the University of Illinois in 2009, the year Pinsof graduated from Iowa, and then an MFA in dramaturgy from Columbia. Her HowlRound biography says that graduate school was where she had the debt-riddled realization that she did not want to be a dramaturg. She had by then written plays. The Beautiful Beautiful Sea Next Door, a dark comedy that reworks the Medusa myth as a family drama with cellphones, was produced in New York in 2012. Numismatics, about a woman who returns to her dead mother’s house to organize a coin collection as a defense against a larger chaos, was a finalist for a Clubbed Thumb commission. The Ivories concerns an elephant invasion of the Midwest and a seventy-two-year-old woman who sets out to become a YouTube pop star to avenge her daughter’s murder. Doollee lists the plays as unpublished.
She wrote for HowlRound in those years, and two essays there show the mind before the comedy. “Confessions of a Serial Intern” describes seven theatre internships that she found useful and that also trained her, she wrote, to accept exploitation, apologize constantly, and be grateful for scraps. “Antipermanence,” from 2014, argues that nonprofit theatres profess missions built on experiment and community while their budgets, buildings, and jobs push them toward one goal, which is to keep existing. She surveyed staff at dozens of theatres to measure the gap between what the institutions said they were for and what they did. That is the question her husband would make a career of, put to an institution she knew from inside, three years before he defended his dissertation.
Feinberg moved to Los Angeles in 2013. There she was one of the founding members of the Kilroys, the collective of women in theatre that began publishing an annual list of unproduced plays by women and trans writers to answer the industry’s claim that it could not find them. Her account of the problem, in a 2014 Backstage interview, was that theatres sincerely believed in equality and kept producing the same men, and that the cause was habit and structure. She then, as her own biography puts it, ditched the theatre to become a screenwriter and cartoonist.
The Hollywood years were an apprenticeship. She was an assistant on Veep, a writers’ assistant on Arrested Development, and a script coordinator on I Love Dick. In April 2019 Variety reported that Showtime was developing her half-hour animated comedy Multifarious Maris, with ShadowMachine, the studio behind BoJack Horseman, attached. The premise is a compact statement of her subject. On the night of her thirtieth birthday Maris discovers that after one glass of wine too many she can travel into the lives she would have lived had she chosen differently, and she spends the series hopping between them in search of the one where she is living her best life. The joke is on the phrase. Feinberg has since developed adult animation for AMC as well, and her website lists two live-action features and a second graphic novel in progress.
Her humor writing is where most readers meet her. She has published in The New Yorker, McSweeney’s, Bon Appétit, and The Hairpin, and the New Yorker pieces make a small anthology of one comic method. “Millennials Are the First Generation in History to Inspire Think Pieces About Millennials” has a narrator who clicks on an article because it is about her personally and then recites the article’s list of generational experiences back as proof of her individuality. “But I Can’t Complain” keeps one rule fixed, that others have it worse, while the circumstances around it escalate from backache to a toddler who kills the neighbors’ pets to nuclear war. “What I Know About You Based on How Many of Your Friends Are Becoming Therapists,” from this spring, builds a diagnostic taxonomy out of a number and closes by sending the reader to therapy. “From Napkin to Notoriety” is about the insufferable literary wunderkind, a figure she has watched from inside theatre, graduate school, writers’ rooms, and magazines. “Why Women’s Bodies Are Bad for Women,” at McSweeney’s, takes the wellness-culture separation of self from body so literally that the solution becomes leaving the body altogether.
Her debut graphic novel, Goodbye, Dolly!, came out in 2025. It is narrated from beyond the grave by Dolly the cloned sheep, whose six adult children, the nepo lambs, wander the Scottish countryside after her death trying to work out what a second-generation clone is for. The Deerfield Public Library, her hometown library, had her on its podcast, and she said there that Dolly had made her uneasy at age ten and that she was drawn to the gap between an epochal scientific breakthrough and the speed with which the world stopped caring about it. Comic Book Herald singled out her method of drawing the sheep by hand and setting them against photographs of Edinburgh and the Scottish hills. A neurotic cartoon consciousness in the middle of a licensed stock landscape is a fair picture of her whole comic world.
Feinberg typically begins with a real human predicament, a dead mother, a murdered daughter, adult disappointment, a body, and then supplies the character with an absurd system for managing it: coin collecting, YouTube fame, gratitude practice, therapy vocabulary, generational sociology, a multiverse. The character follows the system with total seriousness. Feinberg then lets the system run until it consumes the person. The coins get organized; the mother stays dead. The narrator cannot complain; the house burns down. The emotion in these pieces is never fake. The grief is grief, the anxiety is anxiety, the wanting is real. What is wrong is the script the culture hands the person for handling it. Her subject is that misfit, and her characters are ridiculous without being frauds.
Set that beside her husband. Pinsof’s blog is called Everything Is Bullshit, and his method is to take a respectable explanation and replace it with the less respectable machinery underneath: incentives, alliances, status, signaling, confabulation. He decomposes. Feinberg takes the same respectable explanation and lets it keep talking until it incriminates itself. She literalizes. Pinsof writes that your egalitarian principle may be serving a coalition. Feinberg invents a woman so egalitarian she has ranked all her friends by how psychologically evolved they are. He explains the contradiction. She stages it. Both of them are after the distance between the story a sophisticated person tells about himself and the less dignified thing visibly going on underneath, and both trained in theatre, in neighboring suburbs, in the same year, before either had a name for the subject.
David and Annah met in the summer of 2009 while working as bus drivers and camp counselors at Discovery Day Camp in the suburbs of Chicago. In 2014 Feinberg wrote and directed a short called Gretch & Tim, about idealistic twins who fail to save their town library and decide to steal it back, and the credits list the original music as David Pinsof’s. In 2018 he composed for her Balloon Room, a short musical in which Stacy Osei-Kuffour played Hillary Clinton, and he also composed the soundtrack for Feinberg’s Fetus Monster (2024). So the music Pinsof told me he finds inherently enjoyable was, for several years, the medium through which the two of them made things together, and it began a year after she arrived in Los Angeles and four years before his dissertation.
Feinberg appears in Pinsof’s writing more than he appears in hers, and when she does she tends to mark the limit of his persona. In “Money Is Bullshit” he reduces exchange to its social functions and then reaches marriage and stops reducing. His wife, he writes, is irreplaceable, because her interests have become intertwined with his own, and a man who withheld help from her because she could not repay it would be a bad husband. He uses the phrase without irony. In “Deep Bullshit” he describes leaving a restaurant with his pregnant wife when an older woman stopped them and announced that their baby would change the world. Feinberg teared up and later agreed the prophecy was bullshit. Pinsof admits he was moved too, while diagnosing why the line worked. In “Bullshit Is a Choice” he calls himself a softie who sings the Barney song to his daughter. And in this August’s “The Correct Answer to Every Philosophical Question” he says his belief that life has no inherent meaning survived fatherhood while the feeling behind it did not. The man who explains why our deepest stories are bullshit has a family life in which understanding the machinery leaves the feeling in place.
Feinberg keeps him offstage. Her Bon Appétit essay on the weeks after her daughter’s birth is personal and names no husband. The Deerfield interview ranges across childhood, parenthood, existential dread, and sheep, and never mentions him. That tells you nothing about the marriage and something about the two writers. He brings his household into his essays because it supplies cases for the theory. She writes from inside her own head and leaves the rest of the house alone.
If Pinsof’s question is what the machinery underneath us produces, Feinberg’s is what it looks like from inside. His answer is a model. Hers is a sheep in a stock photograph of Scotland, sincerely grieving, wondering who she is, and reaching for the wrong tool.
Pinsof’s Cards Against Humanity Cadre
Josh Dillon is the astrophysicist of the group. After graduating from Highland Park High School in 2005, he went to Stanford, where he earned a B.S. in physics, and then MIT, where he completed a Ph.D. in physics in 2015 under cosmologist Max Tegmark. He is now an assistant research scientist at UC Berkeley, where he works on 21-centimeter cosmology and leads analysis for the Hydrogen Epoch of Reionization Array, an effort to use radio astronomy to study the first stars and galaxies. A Berkeley profile notes that Dillon continued working on Cards Against Humanity alongside his scientific career. He also became closely involved with the company’s Science Ambassador Scholarship, which has used proceeds from the game’s Science Pack to fund women studying STEM. In a 2014 profile of the eight founders, Dillon was described as the group’s project manager, repeatedly dragging its joke-writing sessions back on track when he was not studying astrophysics at MIT.
Daniel Dranove followed another science-heavy route. He attended Northwestern University from 2005 to 2009 and earned a B.S. in Integrated Science, later working there as a researcher. His public biography is much thinner than Dillon’s or Pinsof’s, and he appears to have deliberately maintained a low profile, but professional records continue to identify him as a co-founder and partner of Cards Against Humanity. Dranove later moved to Hawaii and became involved in another deliberately irreverent consumer business. Entrepreneur David Evans identifies Dranove as a co-founder of Optimist, a cannabis brand whose advertising drew explicitly on the outrageous stunt-marketing playbook Dranove had helped develop at Cards Against Humanity. The trajectory is characteristic of the group: a technically educated Highland Park kid who unexpectedly acquired financial independence through a joke project and then used it to experiment outside a conventional professional career.
Eli Halpern was one of the group’s writers. He went to Brown University with Ben Hantoot and graduated in 2009 with a degree in Literary Arts. He was not merely a student who happened to like comedy. Brown awarded him a Weston Fine Arts Award for poetry from his honors thesis, along with the Preston Gurney Prize in Literary Criticism of Poetry. He later worked as an editor at Groupon. Halpern and Hantoot had known each other since middle school, and at Brown they became one of the principal testing stations for the embryonic Cards Against Humanity. They brought complete decks back to campus, played them with classmates, and eliminated jokes that failed. As Hantoot and Halpern later told the Brown Daily Herald, those sessions helped turn a basement party game into a viable product. Halpern’s description of how to win the game was revealingly psychological: in a 2013 interview, he joked that the secret was to dissect human psychology until nothing seemed funny anymore. By 2014, Chicago reported that he was also touring with an interpretive dance company.
Ben Hantoot was the design and manufacturing man. He graduated from Brown in 2009 with a degree in psycholinguistics and, together with Halpern, tested early versions of Cards Against Humanity on Brown students. Hantoot handled much of the game’s graphic design and manufacturing as the project moved from homemade decks to professional production. The Brown Daily Herald account is especially useful because it captures the transition while it was happening: Hantoot described starting with decks full of “duds,” using college audiences to whittle them down, then negotiating the printing of thousands of professional decks after the Kickstarter campaign took off. He subsequently became a serial design-oriented entrepreneur. He co-founded Dinosaur Coffee in Los Angeles, became a partner in Go Get Em Tiger, and co-founded Magic Puzzle Company. Hantoot’s career suggests that the commercial success of Cards Against Humanity was not simply a freak payday for him. It became capital and credibility for a succession of businesses built around design, play and consumer experience.
David Munk is the most obscure of the seven. He attended Carnegie Mellon before transferring to Northwestern, where he earned a B.A. in economics in 2010. Public professional records identify him as a co-founder and partner of Cards Against Humanity, but unlike Temkin, Dillon or Hantoot, he has built almost no independent public persona. That does not mean he was peripheral. Chicago’s 2014 account places him squarely inside the weekly writing room and suggests that his taste helped supply some of the game’s more aggressively bawdy material. He also appeared with Halpern in a 2013 discussion of the games the founders themselves liked to play. Munk is useful precisely because he resists the temptation to turn every founder into a grand biographical story. He appears to have remained primarily a partner in an enormously successful private company while keeping the rest of his life largely private.
Max Temkin became the public face of Cards Against Humanity and for years was the founder most closely associated with its political provocations, marketing stunts and institutional culture. At Goucher College he studied philosophy and became involved with Humans vs. Zombies, the campus game that spread to colleges around the country. A Washington Post feature from 2008 caught Temkin immersed in the game before Cards Against Humanity existed. He also worked in web and graphic design and on Barack Obama’s 2008 presidential campaign. A Goucher profile described him designing websites, working for political and nonprofit clients, building software for Humans vs. Zombies and developing Cards Against Humanity simultaneously. That combination of games, design, politics and internet organizing helps explain why Temkin emerged as the group’s spokesman and chief impresario. His active role ended in June 2020 after former employees made allegations about racism, sexism and the culture of the company’s Chicago office. The other founders announced an investigation, and Temkin stepped down from active involvement. He remained an owner but no longer had an operational role.
Eliot Weinstein is the economist and quantitative-social-science figure in the group. He attended the University of Chicago, graduating in 2011 with a B.A. in economics. Weinstein and Temkin had known each other since first grade, making them one of the oldest relationships inside the Highland Park circle. A revealing 2013 University of Chicago Magazine profile caught Weinstein just as the game was becoming a serious business. He was applying to economics Ph.D. programs and remarked on the strange possibility that income from Cards Against Humanity could pay his way through graduate school. Instead of disappearing into a conventional economics career, Weinstein continued to combine the company with academic interests. The University of Chicago now identifies him as AB’11, MPH’25, reflecting his later return for public-health training. He has also served as vice president and treasurer of The Storytelling Trust. Weinstein therefore sits somewhere between Dillon and Pinsof in the larger story: another member of the Cards circle who repeatedly returned to serious quantitative and academic work while retaining the unusual freedom provided by an accidental hit created with his high-school friends.
Pinsof on Pinsof
David Pinsof’s theory says that stated reasons are rarely causes, that opinions are coalition badges, that status seeking must hide itself to work, and that intellectuals exaggerate the problems that make intellectuals necessary. He has applied every one of those claims to other people and some of them to himself. “Psychologizing Psychologists” turned the method on his field. The reflexive admission runs through the blog: yes, this essay is a status play too. What he has not done is run the theory on the machine that publishes him. The machine is Substack, and it has incentives, an audience, and a coalition of its own. His theory predicts what those will do to a writer. The prediction can be checked against the writer.
Start with the titles, since the title is the unit Substack pays for. A post travels as a headline in an email, a link on X, a screenshot in a group chat. The argument arrives later, if at all. Pinsof’s titles read as a list of executions: “Happiness Is Bullshit,” “The Meaning of Life Is Bullshit,” “Utilitarianism Is Bullshit,” “Democracy Is Bullshit,” “AI Doomerism Is Bullshit.” The arguments under them are qualified. “Democracy Is Bullshit” concludes that democracy beats the alternatives. “Happiness Is Bullshit Revisited” reworks the original into a claim about prediction-error signals that most psychologists could sign. The gap between title and argument is a fact about the corpus, and Pinsof’s theory has a name for it. In “The Evolution of Social Paradoxes” he describes signals that deliver a benefit while concealing that a signal is being sent. The maximal title delivers the status benefit of nerve, the cardinal virtue of his set, and the qualified argument supplies deniability when the nerve is challenged. The title says everything is bullshit. The essay says some things are more functional than they look. A reader who objects to the title is pointed to the essay. A reader who finds the essay tame is pointed to the title. The pair works the way his theory says paradoxical signals work: it collects the reward for boldness while retaining the ability to disclaim boldness.
His own “Opinions Are Bullshit” supplies the second lens. Opinions, on that account, function as badges of coalition membership. They are held, displayed, and defended for what they signal about whose side you are on. Apply that to the blog’s targets. Pinsof attacks happiness research, the search for meaning, utilitarianism, rationalist AI doom, mainstream political psychology, and the intellectual class’s faith in its own necessity. Every one of those targets sits inside or next to the audience that reads him. UCLA’s profile of its Substack academics called his readership a global gathering of Darwinist cynics, and that gathering overlaps with the rationalist and effective-altruist worlds he keeps attacking. Alliance Theory predicts this pattern. The cheapest status inside a coalition comes from defeating the nearest rival coalition on shared ground. Attacking utilitarianism in front of people who have read too much utilitarianism wins more than attacking, say, astrology, which no one in the room defends. His targets are the badges his audience is half-ashamed of wearing. The essays let readers take them off in public and put on a better one.
The theory also predicts what he will leave alone. The set defines itself against the hereditarian right, which shares its evolutionary vocabulary. Pinsof does not write “Egalitarianism Is Bullshit” or “Race Is Not Bullshit.” On his own account of coalitional caution there is no need to ask whether he holds views on those questions. The point of Alliance Theory is that positions are set by who you can afford to lose. A Darwinist cynic who ran the method on group differences would lose the part of his audience that keeps him respectable, and his theory says the essay that costs a coalition does not get written. That is a prediction his theory makes about him, and the blog’s table of contents matches it.
Next take the freedom display. Pinsof told Zach Elwood that Cards Against Humanity freed him from the incentives of an academic career, and he has said so often enough that the fact has become part of his public identity. “Status Is Weird” argues that displayed independence from status is a status strategy, one of the best. By his own logic the repeated mention of his financial freedom is a signal that says trust me, I have nothing riding on this, and it earns him the deference that an unbought man gets in a set that theorizes incentives. The signal happens to be true. His theory does not care whether it is true. It cares what it does, and what it does is exempt him from the suspicion he directs at everyone else. The money freed him from one incentive structure and put him inside another. Substack rewards frequency, provocation, and a recognizable voice, and it rewards them in attention, which for a man who does not need the subscriptions is the currency that remains. His theory says attention is the currency that was always at stake.
Now the numbers. The blog ran twenty posts in 2023, twelve in 2024, ten in 2025, and six through August 2026. The annual archives show the decline. Over the same years he published in American Psychologist, presented evolutionary game theory at HBES, started a research podcast with David Pietraszewski, and began writing models with Greek letters in them. He told me all science should move toward formalism. That is a stated reason. His theory says stated reasons are rarely causes and that the cause is usually a move in a status game. The move here is legible in the “4D Chess” abstract he sent me. At the fourth order, status-seekers who have been caught pretending learn to pretend to be unpretentious. A man who has gained standing as a cynic by seeing through everyone eventually has to signal that he is not merely a cynic gaining standing, and in his set the surest way to do that is to become boring on purpose. Mathematics is the costly signal. It cannot be screenshotted. It does not travel by email. It says: I was never in this for the titles. The formal turn is, by Pinsof’s own account of the fourth order, the unpretentious pose that a pretentious signaler adopts once the audience has learned to see through pretension. The pose may also be science. The theory says both can be true, and that the man executing it will believe only the second.
In “A Big Misunderstanding,” Pinsof argues that intellectuals inflate the role of misunderstanding in social dysfunction because a world full of misunderstanding needs intellectuals to correct it. Invert the argument and aim it at the blog. A Darwinist cynic inflates the role of hidden motive in social life because a world full of hidden motives needs Darwinist cynics to expose it. Everything Is Bullshit is a brand, and a brand called Everything Is Bullshit has a commercial and coalitional interest in everything continuing to be bullshit. The essays that qualify the title, that concede democracy works and happiness might be a real signal and Dan Williams might have a point, are the essays that threaten the brand, and they are also the best essays. The tension I called the central one in his career, between a favored explanation and the evidence that would limit it, is on this reading a tension between the writer and the masthead. His theory predicts the masthead will win more often than the writer would like.
Dan Williams answered “A Big Misunderstanding” with “We Are Confused, Maladapted Apes Who Need Enlightenment,” and Pinsof replied in Williams’s own newsletter with “What Kind of Apes Are We?” I read that exchange in the bio as the place where the swagger falls away. Under Pinsof’s theory it reads as something else as well. Replying in the rival’s house, conceding ground, narrowing the disagreement: these are the moves “Status Is Weird” identifies as the highest-yield status plays available, the ones that gain by appearing to give. The concession was sincere. His theory says the sincerity is what makes it work, and that the man making it need not know what it is doing for him. Making Pietraszewski, a critic of Alliance Theory, his podcast cohost belongs to the same family. A former enemy on your payroll is the alliance move that looks least like one.
Run the essay’s own questions on its author.
What does he have an incentive to understand? The essay says humans understand what pays. Pinsof’s payoffs come from an audience that rewards seeing through, a scientific set that rewards function-first explanations, and a company whose money spares him the academic market. He has understood, with great skill, that social science is coalitional, that intellectuals flatter themselves, and that cynicism sells to a certain reader. He has not understood, or has not written down, the things his coalitions do not pay him to understand: what alliance the evolutionary psychology consensus on group differences serves, what his own company’s public morality is for, or what the theory says about the heterodox commentariat that reads him. The essay’s rule is that stupidity is strategic. Applied to him, the blank spots on his map are where the rule predicts they would be.
Is his stupidity strategic? Take the paragraph on bigotry. It says ethnic minorities are accurately stereotyped as allies of the Democratic Party, that hostility toward them tracks hostility toward Democrats, and that the rest comes from status competition between low-status Whites and their nearest rivals. That is Alliance Theory carried to the edge of the question evolutionary psychology has spent thirty years refusing to ask, and it stops at the edge. Everything about bigotry is explained by alliance and status, and nothing by the beliefs about groups that the same paragraph has just called accurate. A man who thinks stereotypes are savvy and then explains prejudice entirely by coalition has left a gap, and the gap sits exactly where the cost to him would be highest.
Does he understand what he is doing? Yes, and he says so. The essay lists the goals of intellectuals as climbing hierarchies, derogating rivals, and dominating people under moralistic pretexts. The essay derogates rivals for two thousand words. Its rivals are progressive social scientists, the rationality movement, effective altruists, positive psychologists, and the press, which is to say the coalition of Democrat-allied intellectuals the essay names as such, plus the rationalist wing that sits nearest his own readers. Its moralistic pretext is truth. Its hierarchy is the one where the man who sees furthest through stands highest, and the essay is a bid for the top of it. He knows this. “Psychologizing Psychologists” and the reflexive lines throughout the blog are the confession, and the confession is, by his account in “Status Is Weird,” the move that lets a status-seeker keep seeking.
Is his advice bullshit? He gives none, and that is the tell. The essay ends by saying there is nothing to be done, the world does not want to be saved, and the study of human nature is the study of the hole. An intellectual who says the world cannot be fixed has escaped the charge of pretending to fix it, and has also made himself indispensable in the mirror-image way. If everything wrong is misunderstanding, the understanders are the most important people. If nothing is misunderstood and everyone is savvy, then the savvy are self-deceived about their savviness, and the one man who sees the self-deception is the most important person. The misunderstanding myth makes intellectuals saviors. The no-misunderstanding myth makes one intellectual the only one awake. Each story ends with its teller at the top.
Are his problems bad motives? Apply the Kahneman test he applies to Kahneman. Kahneman learned about the biases and did not change; Pinsof takes that as evidence the biases were self-serving. Pinsof has learned that status seeking must disguise itself and has not stopped; he writes about status weekly, in the register that gains it, on a platform that rewards it, and he says the reader is on the blog because he does not want to be happy. By his own reasoning, the persistence shows the motive. He would agree, which is the difficulty. A theory that predicts its own author’s behavior and is confirmed by it has explained everything and forbidden nothing, and the essay’s charge against social science, that it collects what it wants to find, is the charge his critics make against him.
Then the Starbucks test. The essay says to judge an institution by what it does, not by its mission statement. The blog’s mission statement is poking holes in the stories we tell ourselves. What it does is deliver, to a coalition of readers who dislike the same people he dislikes, a weekly demonstration that their rivals are self-deceived and that the reader, by reading, is not. That is the product. It is a good product and it is not the same as the mission.
The stated motive and the actual motive can coincide. The essay says pretending to care differs from caring and that actions speak. Pinsof’s actions include a preregistered study designed to let his theory lose, a reply written in a critic’s newsletter conceding ground, a podcast with the man who asked “so what,” and a decade of unpublished patience that no status game rewards. Those are the deeds. They are the deeds of a man who wants the theory to be true more than he wants to be seen holding it, and the essay’s own standard says to weigh them over the words. So the fair verdict is the one his theory allows and his persona resists: the motives are mixed, the coalition shapes what he looks at, the status pays for the looking, and some of the looking is real. The only misunderstanding, he says, is that there has been a misunderstanding. Applied to him, the only bullshit is the claim that it is all bullshit.
The audience, for its part, is doing what Alliance Theory says audiences do. Readers who share the badges he sells become a coalition, and the coalition pulls the writer toward what it rewards. The phrase for that is audience capture, and it is an Alliance Theory story with the writer as the captured party. The evidence of capture in Pinsof’s case is mixed. The titles stayed maximal while the arguments grew careful, which suggests he gave the audience the headline and kept the essay for himself. The posting rate fell as the formal work rose, which suggests the academic coalition began outbidding the Substack one. The AI doom essay drew the heaviest resistance from his own readers, and he did not retreat, which suggests either that he is less captured than the theory predicts or that the rationalist part of his audience is the part he can afford to lose. His theory does not let him claim the first without earning it, and the second is the alliance reading.
There is one more party the theory applies to, and I am it. A long biography of a Darwinist cynic, published by a writer who has adopted the cynic’s concepts and run them on Orthodox Jewish communal life, on academics, and on public intellectuals, is a bid for standing inside the cynic’s set. The essay signals membership by fluency. It signals nerve by criticizing the subject. It signals humility by criticizing the subject gently and by ending with the concession that he might be right. Pinsof answered the draft by calling me the closest thing he has to a biographer. On his theory that sentence is the coalition admitting a member, and the warmth I felt reading it is the reward the coalition pays. On his theory I should distrust the warmth and keep writing. On his theory he should distrust the biography and keep reading. Neither of us will, and the theory explains that too, which is the thing about it that his critics find hardest to forgive.
A Robin Dunbar Frame
The paper David Pinsof described to me under the title “Human Status-Seeking as 4D Chess” ends its arms race at a specific rung. Status-seekers pretend to care, audiences see through the pretending, the seekers then pretend to be unpretentious, and there the recursion stops, because, in his words, fifth-order mindreading makes cognitive demands that most people cannot meet. The pretenders win, on this account, partly because nobody can afford to catch them at the next level up. That is a claim about a number, and the number comes from a laboratory. Whether the number has held is a question the model has to answer before the game theory can start.
The laboratory belongs to Robin Dunbar, the Oxford evolutionary psychologist better known for the number 150. Dunbar’s larger project is the social brain hypothesis: primate brains grew to manage social complexity, and the human neocortex is the size it is because human groups are the size they are. Inside that project sits a smaller literature on what Dunbar calls orders of intentionality, borrowing Daniel Dennett’s term. First-order intentionality is having a belief. Second-order is believing something about another’s belief, which is ordinary theory of mind, the thing children acquire around age four. Third-order is believing that you think that I believe. Each further embedding adds a clause. Dunbar’s question was how many clauses an adult can track before the structure falls apart.
The first answer came in 1998, in a paper by Peter Kinderman, Dunbar, and Richard Bentall that gave adults short stories and asked questions of increasing embedding. Performance stayed high through the fourth order and dropped sharply at the fifth. Dunbar has repeated the result across several studies since. In 2007 James Stiller and Dunbar reported that the mean adult ceiling sits around the fourth to fifth order, that individuals range from about three to about seven, and that the ceiling predicts the size of a person’s social network. Later imaging work from the same group tied intentionality competence to the volume of orbital prefrontal cortex. Dunbar has built a good deal on the fifth-order limit. He argues that religion requires it, since a believer must hold that God intends that we believe that He wants us to act as He intends. He argues that Shakespeare’s audiences worked at the fourth order and Shakespeare at the fifth or sixth, and that this is why the plays are hard.
Set that beside Pinsof’s ladder. His first level is caring about others being inferred and rewarded, which in Dunbar’s counting is second-order: I believe you care. His second level, the seeker anticipating the inference and pretending, is Dunbar’s third: I believe that you believe that I care. His third level, seeing through the pretending, is Dunbar’s fourth. His fourth level, pretending to be unpretentious, is Dunbar’s fifth. And the fifth level he says nobody can run, the one that would catch the unpretentious pose, is Dunbar’s sixth. If that mapping is right, the model stops one rung past where Dunbar’s data say most adults stop, and the stopping rule is better supported than Pinsof’s own phrasing suggests, since Dunbar found the wall at five and the model needs people to fail at six. If Pinsof counts the way Dunbar counts and means the fifth order in Dunbar’s sense, then his fourth level, the crucial one where humility gets faked, is already at the ceiling for the average adult, and the model has to explain how the average status-seeker executes a strategy the average brain can barely represent. The paper will need to say which count it uses. The difference is one rung, and one rung is the argument.
The larger difficulty is that Dunbar’s wall has been fought over for a decade, and the fight is unresolved. In 2015 Cathleen O’Grady, Christian Kliesch, Kenny Smith, and Thomas Scott-Phillips published “The Ease and Extent of Recursive Mindreading, Across Implicit and Explicit Tasks” in Evolution and Human Behavior. They argued that the earlier tests were flawed: the stories broke their own conceptual chains, some questions could be answered by simple substitution, and some had no correct answer. With cleaner stimuli they found adults tracking mental states to at least seven levels of embedding, both when the states were described in words and when they were shown in filmed interactions. They also found the implicit version easier than the explicit one. Watching people is easier than reading about them. Nathan Oesch and Dunbar answered in 2017 with five objections. Then in 2023 Ross Wilson, Ales Hruby, Daniel Perez-Zapata, Sanne van der Kleij, and Ian Apperly published “Is Recursive ‘Mindreading’ Really an Exception to Limitations on Recursive Thinking?” in the Journal of Experimental Psychology: General. They rebuilt the tasks again, this time to close the shortcuts they found in O’Grady’s stimuli, and fifth-order performance fell from 80 percent on the original tasks to 17 percent on the revised ones. A second study got 15 percent. Only large bonuses, unlimited time, and coaching on a strategy raised it to 45 percent. Their conclusion was that recursive mindreading is effortful and limited like recursion in every other domain. A 2024 study by Cameron Jones, Sean Trott, and Benjamin Bergen at UC San Diego then replicated O’Grady’s seven-level result with text-only stimuli and ruled out the answer-format shortcuts. So the literature in 2026 holds two results that contradict each other, and which one you get depends on how the stories are written. Pinsof’s model borrows its central parameter from the side of a dispute that has not been settled.
That matters for the arms race in four ways.
First, the implicit finding cuts against Pinsof in the setting his model cares about most. Status competition does not happen in short stories. It happens in rooms, over months, among people who watch each other. O’Grady and her coauthors found that the deep embeddings were easier when the observer watched a social scene than when he read a description. If that holds, the cynical audience Pinsof needs to be exhausted at the fifth order is operating in the condition where fifth-order tracking is cheapest. The model’s stopping rule assumes the hardest case and applies it to the easiest. Wilson and Apperly give him a way back: their subjects could reach the fifth order when paid enough and given time, which is what a status competition offers an audience that has months to watch and something at stake. The two findings together suggest the ceiling is a cost curve, and a cost curve is what a formal model needs. Pinsof should model the price of the fifth order.
Second, the variation Dunbar found is a prediction Pinsof has not yet made. If individuals range from three to seven orders, and the range tracks network size and prefrontal volume, then the arms race does not stop at the same rung for everyone. Some audiences run the fifth and sixth orders without strain. Those audiences should catch the unpretentious pose, and in front of them the pretenders should lose. That yields something testable. Pretension should thrive where audiences are large, anonymous, and cannot spend the fifth order on each performer, and it should die in small dense groups where every member has the capacity and the time to run it. The essayist with seven thousand subscribers faces a different mindreading budget per reader than the man in a twelve-person lab, and the model should say that the essayist can afford poses the lab member cannot. Dunbar’s own linkage of mindreading capacity to network size supports the inference, since the people with the highest ceilings are the ones with the most relationships to spend it on.
Third, the same variation makes Pinsof’s second reason for the pretenders’ victory do more of the work. His abstract gives two. One is the cost of the fifth order. The other is that being charmed by a talented mindreader pays, because talented mindreaders are likely to rise and it is worth being on their side. That second reason survives O’Grady and Jones. It does not depend on a ceiling. It says that even an audience capable of seeing through the pose has a reason not to, since the person executing the pose is displaying the skill that predicts future status, and the display is information. On that reading the pretenders win by recruitment. Dunbar supplies a warrant here too. If mindreading competence predicts network size, then the skilled pretender is not merely charming; he has more allies, and joining him is the better bet. If the cost argument falls, the charm argument can hold the model up on its own, and Pinsof would do well to say so before a reviewer who has read O’Grady says it for him.
Fourth, the explicit task may be the right one for part of Pinsof’s problem. Seeing through a pose in the moment is implicit. Explaining to a third party why the pose was a pose is explicit; it requires putting the embedded clauses into words, and that is where O’Grady found the going harder and where Wilson found it hardest. Much of what Pinsof calls the collapse of a status game under common knowledge happens through talk. Someone has to say, out loud, that the humble man is competing, and say it in a way that others can follow. If the explicit ceiling sits lower than the implicit one, then a room might contain many people who privately see through a pose and few who can articulate it, and the pose survives as a public fact while dying as a private one. That is a richer version of Pinsof’s thesis than the one in his abstract, and the explicit-task data point straight at it.
The question of whether the ceiling varies by culture, training, or intelligence is where the literature is thinnest and where Pinsof’s model is most exposed. Dunbar’s samples were British undergraduates and adults. Stiller and Dunbar found the competence measure only loosely related to general intelligence, and there is little cross-cultural work at the higher orders. Wilson’s coaching result says the ceiling moves with strategy, at least in the laboratory. If it can be raised by practice, and the professions that practice it include lawyers, diplomats, novelists, actors, and comedy writers, then those groups should be the cynical audiences that break the equilibrium. Pinsof trained as an actor and wrote a game about predicting what would make friends laugh. If anyone has a raised ceiling, he does, and the model predicts he should be unusually hard to charm. If training does not raise the ceiling, then the arms race is fixed by biology and the pretenders won once, at the species level, which is the stronger and stranger claim. The model as he described it to me leaves the choice open.
The 4D Chess model has put an empirical parameter at its center and borrowed the value from a literature that is split down the middle. Formalism has a cost he has now incurred: a verbal theory can say that people eventually lose track, and nobody can object, but a model that says they lose track at five has to defend the five. Dunbar, Oesch, Wilson, and Apperly would defend it. O’Grady, Scott-Phillips, Jones, Trott, and Bergen would not. The paper that settles Pinsof’s arms race may turn out to be one of theirs.
Countersignaling
In 2002 three economists published a paper whose title could serve as a chapter heading in Everything Is Bullshit. Nick Feltovich, Rick Harbaugh, and Ted To called it “Too Cool for School? Signalling and Countersignalling,” and it appeared in the RAND Journal of Economics. Its subject was the rich man who dresses down, the Nobel laureate who introduces himself by first name, the top student who skips the honors seal, the old-money family that finds display vulgar. Each of these people declines a signal that economics since Michael Spence said he should send. The paper explained why, and it explained it with a model, an equilibrium, and a laboratory experiment. Twenty-one years later David Pinsof posted “Status Is Weird” and asked why humility, indifference to status, and refusal to signal so often win the status they disclaim. The question is whether he was answering a question the economists had already closed.
He was not, and he knew the literature. The preprint of “The Evolution of Social Paradoxes” cites Feltovich and Harbaugh for the observation that we gain status by not caring about status, and it cites Moshe Hoffman, Christian Hilbe, and Martin Nowak’s 2018 paper “The Signal-Burying Game Can Explain Why We Obscure Positive Traits and Good Deeds,” which modeled anonymous charity and the modest genius in the same tradition. So the rediscovery charge fails. The better question is what a psychologist adds to a problem economists formalized, and the answer turns on a difference the citation hides.
Start with what countersignaling says. Spence’s original model has two types, high and low, and a costly signal the high type can afford and the low type cannot. Education, in Spence’s example, tells employers nothing except that the student could bear its cost. Feltovich, Harbaugh, and To added two things. They added a third type in the middle, and they gave the receiver a second, noisy source of information about the sender, something like reputation, track record, or the way a man carries himself. With that extra channel the model produces an equilibrium the two-type model cannot. Medium types signal, to separate themselves from the low types they might be confused with. High types do not signal, because the noisy channel already separates them from the low types most of the time, and because sending the signal would lump them with the striving mediums. As the authors put it, countersignaling is a signal of confidence. Their laboratory subjects learned to do it within a few rounds.
Three features of that model matter for Pinsof. First, the countersignaler knows what he is doing and so does the audience. The old-money heir who wears a frayed sweater is understood by everyone present to be sending a message, and the message works because it is understood. Nothing is concealed. The equilibrium is a common-knowledge equilibrium in the ordinary game-theoretic sense; everyone knows the strategies, everyone knows everyone knows, and the strategies remain stable. Second, the model requires the extra channel. Take away the receiver’s noisy independent information and countersignaling collapses, since a high type who declines the signal now looks like a low type. Third, the model predicts a shape. Signaling effort should rise with quality up to a point and then fall, so that the people signaling hardest sit in the middle of the distribution. That prediction is the reason the paper counts as science. It forbids a monotonic world.
Now read Pinsof against it. His definition of a social paradox is a signal designed to conceal, from both sender and receiver, that a signal is being sent. The virtue signaler does not believe he is virtue signaling, and the person who rewards him does not believe it either; if either did, the paper says, the reward would be withheld. That is a different animal from countersignaling. The countersignaler’s refusal to signal succeeds because the audience reads it. Pinsof’s humble man succeeds only if the audience does not read it. His central claim is that status games can collapse when the players acquire common knowledge that they are playing one. Feltovich, Harbaugh, and To describe a game that is stable under common knowledge. Pinsof describes one that is unstable under it. The two accounts agree on the surface behavior, the modest display, and disagree on what holds it up.
That disagreement is Pinsof’s contribution, and it is a psychologist’s contribution. The economists modeled agents with beliefs about types. Pinsof models agents with beliefs about intentions, and beliefs about others’ beliefs about their intentions, and so on up. He derives the paradox from two cognitive capacities, cue-based inference and recursive mentalizing, and argues that once a signal of a good trait can be recognized as intended, the intention itself becomes a cue of a bad trait, the trait of wanting to be seen as good. The signal then has to go underground. Countersignaling has no mechanism for that. In the economists’ model, wanting to be seen as high type carries no penalty; it is simply the game. In Pinsof’s model, wanting to be seen is the thing the audience punishes, which is why humility is rewarded and why the reward vanishes the moment humility is seen as a bid for it.
The second thing Pinsof adds is self-concealment. Feltovich, Harbaugh, and To have no place for a sender who does not know he is sending. Pinsof’s paradoxes are, by his definition, hidden from the sender as much as from the receiver. That moves the problem from economics into the territory of Robert Trivers on self-deception, and it makes an empirical claim countersignaling does not make: that the effective humble man will sincerely report no status motive, and that the sincerity is part of what makes the signal work. That is testable. The countersignaling model predicts that the old-money heir, asked in private, will say he dresses down so as not to look like a striver. Pinsof’s model predicts the effective version of that man will say he just likes the sweater and will pass a lie detector.
The third addition is scope. Countersignaling explains a few behaviors around display and credentials. Pinsof takes the same structure and runs it through intergroup conflict, moralistic pretexts, sacred values, and the stability of political belief systems. His argument is that a coalition pursuing dominance can coordinate only if its dominance seeking stays out of common knowledge, since open acknowledgment would draw the moral disapproval that breaks coordination. Sacred values, on this reading, are the coalition’s social paradox: the thing you cannot admit is a strategy because admitting it would end its usefulness. Whether that extension survives is a separate question. It is not in the 2002 paper and could not be derived from it.
Against those additions, the economists hold two advantages.
The first is the shape prediction. Countersignaling forbids a monotonic relation between quality and display and predicts a peak in the middle. The prediction has been tested in laboratories and in field data on things like the use of honorifics and the pricing of luxury goods. Pinsof’s social-paradox account, as published, makes no shape prediction of comparable sharpness. It says concealed signals should fail when concealment fails. That is correct and hard to falsify. The countersignaling result is a template for what a sharp version would look like: a prediction that humility should be rewarded most among people whose status is already established by an independent channel, least among those whose status is in doubt, and that the middle of the distribution should be where the visible striving concentrates. If Pinsof’s model predicts that too, he should say so, because it would tie his psychology to a result economists have already found. If it predicts something different, the difference is the experiment.
The second advantage is the extra channel. Countersignaling works because the receiver has other information. That condition explains why the strategy is available to some people and not others, and why the rich can dress down while the poor cannot. Pinsof’s account does not specify the equivalent condition. When can a man afford to hide his signal? The economists answer: when something else is already carrying it. That is a resource constraint, and it predicts that concealed signaling should be a strategy of the secure and overt signaling a strategy of the insecure. Pinsof’s reading of humility as concealed competition would gain discipline from adopting it. Without it, his framework says every humble man might be competing; with it, his framework says which humble men are.
There is a fourth relative in this family worth naming, since Pinsof’s paper sits between its two parents. Roland Bénabou and Jean Tirole’s 2006 “Incentives and Prosocial Behavior” in the American Economic Review showed that paying people for good deeds can reduce the deeds, because the payment spoils the signal: once there is a material reason to give, giving no longer proves you are the giving type. That is the economics of the contaminated signal, and it is the closest formal ancestor of Pinsof’s claim that recognized intent destroys the value of a display. Bénabou and Tirole had audiences who know the incentive exists. Pinsof’s move is to make the contaminating factor the audience’s knowledge of intent itself, which pushes the model one level up the recursion.
So the ledger reads like this. Countersignaling got there first on the behavior, twenty-one years before the blog and with a model and an experiment. Signal burying got there on anonymous charity in 2018. Bénabou and Tirole got there on spoiled signals in 2006. Pinsof cites the first two and builds on all three. What he adds is the intention as a cue, the self-deceived sender, the instability under common knowledge, and the extension to coalitions and sacred values. What he has not yet added is the thing the economists had from the start: a prediction that forbids a shape, and a stated condition under which the strategy is available.
The paper he described to me under the title “Human Status-Seeking as 4D Chess” is the place to look for that. Its abstract models each level of mindreading as a stage in an arms race between pretentious signalers and cynical audiences and concludes that the pretenders won because the fifth order costs too much to run. That is a stopping rule, which countersignaling never needed because its agents never climbed past the first level. If the model also says at which level the audience’s independent information enters and how much of it a signaler needs before he can afford to go quiet, Pinsof will have joined his psychology to the economists’ result and improved on it. If it stays at the level of who pretends and who sees through, the economists will still hold the one prediction in this area that has been tested and has held.
A Paul Meehl Frame
Paul Meehl spent the last third of his life explaining why soft psychology does not get anywhere. His answer bears on David Pinsof more than on most psychologists, because Pinsof has picked a fight about explanation in a field where, by Meehl’s account, almost nothing can be refuted and almost everything can be confirmed. Meehl was the psychologist who did clinical prediction, philosophy of science, and the genetics of schizophrenia, and the University of Minnesota keeps his papers online. Three of them matter here. “Theory-Testing in Psychology and Physics: A Methodological Paradox” from 1967, “Theoretical Risks and Tabular Asterisks: Sir Karl, Sir Ronald, and the Slow Progress of Soft Psychology” from 1978, and “Why Summaries of Research on Psychological Theories Are Often Uninterpretable” from 1990. Read together they supply a test for what Pinsof has done and a specification for what he is about to do.
The paradox in the 1967 paper runs like this. In physics a theory predicts a number, and better instruments make the prediction harder to confirm because they narrow the band of results that count as a hit. In psychology a theory predicts a direction, and better methods make the prediction easier to confirm because larger samples reject the null more easily, and the null is almost always false. Meehl called the reason the crud factor. In the social world everything correlates with everything. He later reported a Minnesota study of 57,000 high-school students on 45 variables, from birth order to church attendance to what they liked to read, in which 92 percent of the pairwise correlations reached significance. None of those correlations was predicted by any theory. They were the background noise of a connected world. A theory that says X will correlate with Y in a soft domain has therefore taken a risk close to zero. It had a coin-flip chance of getting the direction right by accident, and the coin was loaded, because the theorist picked the direction after looking at the world.
From that Meehl drew a rule that Wesley Salmon had named the damn strange coincidence. A theory earns credit when it predicts something that would be a strange coincidence if the theory were false. Direction is never strange. A point value is strange. A specified ordering of several groups is strange. An interaction that reverses sign under stated conditions is strange. A predicted null, where the crud factor and every rival say there should be an effect, is the strangest of all, because the crud factor is fighting on the other side. Meehl’s word for the space of outcomes a theory tolerates was the Spielraum. The narrower the Spielraum, the riskier the prediction, and only risky predictions move a field.
Now sort Pinsof.
The first published work passes. The 2016 Psychological Science paper “The Political Divide Over Same-Sex Marriage” did not say that sexual attitudes correlate with political ones. Meehl would have yawned. It predicted an interaction: opposition should rise among people with a restricted sociosexual orientation only when they believe gay men are promiscuous, and the effect of the stereotype should be absent or reversed among the unrestricted. That is a sign prediction under stated conditions, and it narrows the Spielraum. The 2017 PLOS ONE paper “The Effect of the Promiscuity Stereotype on Opposition to Gay Rights” did better still. It manipulated the stereotype and predicted that support would move among the restricted and stay put among the unrestricted. A predicted null in one cell and a predicted effect in another is Meehl’s strange coincidence in miniature. Rival accounts that say opposition comes from religiosity, authoritarianism, or disgust sensitivity had no reason to expect that pattern.
The abortion line passes the same test. The 2016 HBES abstract “Pro-Life or Anti-Promiscuity?” predicted that opposition should concentrate among the restricted only when they believe abortion access encourages casual sex. The 2026 preprint “Concerns about Sexual Promiscuity Play an Underexplored Role in Opposition to Abortion” reports the moderation across four studies. Meehl would want to know the size of the effect and whether it survives the controls the rivals demand, but he would grant that the shape of the prediction was risky.
The double-standards work is the most Meehlian thing Pinsof has done and the least available. Alliance Theory predicts a reversal: the same respondent should endorse a principle when it helps his coalition and reject it when it hurts, and the reversal should run at a similar rate in both parties. That symmetry claim is a risk. The dispositional program, and John Jost in particular, expects asymmetry. Pinsof wrote the Pulse of the Nation questions to catch the reversals before the data came in. That is prediction preceding observation, which is the only order Meehl credited. But the analysis reached the literature as citations inside “Strange Bedfellows” and as a 2018 conference abstract. Meehl’s 1990 paper lists ten reasons research summaries are uninterpretable, and two of them are selective submission and selective publication. A risky prediction that was made, tested, and never written up is uninterpretable for a reason Meehl did not list: no one can read it.
Then comes the part of Pinsof’s corpus Meehl would have marked down. The blog essays make claims with an unbounded Spielraum. “Status Is Weird” says humility, authenticity, and independence can serve as status strategies. Can is not a prediction. The essay does not say when humility should pay and when it should cost, or how much, or for whom. “Everything Is Signaling” has the title of a theory that forbids nothing. Meehl’s complaint about soft psychology was that its theories are stated so loosely that any correlation confirms them and no correlation embarrasses them, and the crud factor guarantees a steady supply of confirming correlations. The blog is that complaint in a readable form. Pinsof knows it, which is why the same years produced “The Evolution of Social Paradoxes,” a paper that tries to say when a concealed signal should work and when it should fail. That is a Spielraum getting narrower under its author’s own pressure.
The commentaries on “Strange Bedfellows” read differently through Meehl. Raihani and Atkinson said alliances are one factor among many. Malka and Federico said belief systems have other roots. Both are true and both are what Meehl predicted the field would say, because in a crud-factor world every factor is one factor among many and every theory is a little bit right. That is the slow progress he mourned. Ganzach argued that alliances do not meaningfully explain ideology, which is an effect-size claim. Meehl would say effect size alone settles nothing when the competing theories all predict a positive coefficient. The question is who predicted the pattern, and pattern here means which issues, which respondents, which direction, and which nulls.
That is the specification for the Toronto study. Pinsof told me he is joining a group at the University of Toronto on a large preregistered paper that pits alliances against dispositions as predictors of political views. Meehl gives four demands. First, the preregistration must say in advance which issues alliances should win and which dispositions should win. Alliance Theory has a natural answer: on issues that have recently changed coalitional sides, trade, Russia, vaccines, free speech, alliance measures should predict positions and dispositional measures should fail, because the disposition did not move when the issue did. On issues tied to stable threat and disgust responses, dispositions might hold their ground. If the theory says that before the data come in, it has taken a risk. Second, the study should predict nulls. The most damaging result for the dispositional program is a set of items on which personality explains nothing once alliance is in the model, and the most damaging for Pinsof is the reverse. Both should be named. Third, the study should predict an ordering, at least, and an interval if the formal work permits it. Alliances should outpredict dispositions by more on switched issues than on stable ones. That ordering has a small Spielraum. Fourth, the study must survive Meehl’s auxiliary problem. If alliance is measured by party identification and party identification is the product of disposition, the head-to-head is rigged in a way neither side can see. Measuring alliance in a way that is independent of the outcome is the hardest part, and Meehl would say the paper stands or falls on it.
If the Toronto paper reports only that alliance measures add incremental variance to political views, Meehl’s verdict is that nothing happened. The crud factor supplies incremental variance for free. If it reports the pattern, the field will have a strange coincidence to reckon with, whichever way it goes.
This is also the Meehlian reading of Pinsof’s turn to mathematics. Meehl agreed with the aim and doubted the shortcut. Equations narrow the Spielraum only when their parameters are fixed by something other than the data being explained. A model with three free parameters fitted to the same survey it predicts is a verbal theory in symbols. Meehl argued that soft psychology should aim for what he called consistency tests, where a theory predicts a numerical relation between quantities it did not use to build itself. The evolutionary game theory of alliances Pinsof presented in 2025 has that potential. So does the recursive mindreading model he described to me under the title “Human Status-Seeking as 4D Chess.” That model’s central claim, that fifth-order mindreading is too expensive and so the pretenders win, is a risky prediction, since it forbids a world where cynical audiences get the upper hand. The question Meehl would ask is whether the model fixes the cost of each order from outside, from the mindreading literature, or tunes it to fit the observed level of pretension. The first is physics. The second is crud with Greek letters.
There is a countersignaling test waiting in the same model. The economic theory of countersignaling predicts a non-monotonic relation between quality and display: the middle signals hardest, the top and bottom stay quiet. A U-shape is a Meehlian prediction, since direction cannot deliver it and the crud factor supplies straight lines. If Pinsof’s social-paradox account predicts a U and finds one, that is a strange coincidence. If it predicts a straight line and dresses it up in recursion, it is not.
One more of Meehl’s observations lands on Pinsof’s reviewers. He wrote that theories in soft psychology do not die by refutation. They fade as people lose interest. One reason is that the field’s conventions strip risk out of papers before publication. Pinsof told me the humor paper carried heavy formalism and that reviewer comments are pushing much of it into an appendix or out. Meehl would recognize the move. The formalism is where the risky predictions live. Moving it to an appendix leaves the verbal claim, which the crud factor will confirm.
Pinsof has written an essay called “Psychologizing Psychologists” that accuses his field of producing findings that serve its own coalition. Meehl’s accusation was different and older. He said the field produces findings that serve no one, because the tests are too weak to tell a true theory from a false one, and so the true ones never win. Pinsof’s early work took real risks and survived them. His public work takes none. His next work, if it does what he says, might be the first head-to-head in political psychology that Meehl would have counted as a test.
An Imre Lakatos Frame
When David Pinsof, David O. Sears, and Martie Haselton answered their critics in Psychological Inquiry, they titled the reply “The Strange Epicycles of Political Psychology.” The word is a charge. An epicycle is a circle added to a circle so that a theory of the heavens can keep the earth at the center while the planets refuse to cooperate. To call a rival’s hypotheses epicycles is to say the rival’s theory is dying and does not know it. The critics had said something close to the reverse. Yoav Ganzach argued that Alliance Theory explains little. Ariel Malka and Christopher Federico argued that it could not be the single root of political belief. Nichola Raihani and Quentin Atkinson said the world was more complicated than that. Underneath the exchange both sides were accusing each other of the same thing: a theory that can accommodate anything and therefore forbids nothing. Neither side used the vocabulary built for that accusation. Imre Lakatos built it in 1970, and it sorts Pinsof’s career better than the vocabulary of falsification he and his critics reach for.
Lakatos’s essay “Falsification and the Methodology of Scientific Research Programmes” began from an observation that embarrassed Popper. Scientists do not abandon theories when an experiment goes wrong. They blame the instrument, the sample, the auxiliary assumption, and they keep going, and often they are right to. Newton’s theory survived the anomalous orbit of Uranus by positing Neptune. The same theory later survived the anomalous orbit of Mercury for decades by positing Vulcan, a planet that did not exist. The difference between the two cases is the difference Lakatos wanted to name. So he replaced the theory with the research program as the unit of appraisal. A program has a hard core, the claims its adherents refuse to give up. It has a negative heuristic, the rule that says redirect every refutation away from the core. It has a protective belt of auxiliary hypotheses that take the hits. And it has a positive heuristic, a plan for what to build next. Programs are never refuted by a fact. They are outgrown. A program is progressive when its adjustments predict new facts that later turn up, and degenerating when its adjustments only explain away what has already gone wrong. Neptune was progressive. Vulcan was degenerating. The Stanford Encyclopedia entry sets out the apparatus in full.
The first move is to find Pinsof’s hard core. He has two programs that share a foundation. The alliance program holds that political belief systems are built from coalitions and that principles are instruments coalitions pick up and put down. The signaling program holds that status competition must disguise itself and that recursive mindreading drives the disguise. Beneath both sits a claim his whole oeuvre refuses to surrender: the reasons people give for what they do are not the causes of what they do, and the causes are functions that natural selection built. That is the core. The negative heuristic follows from it. Never accept a stated reason as an explanation. When a man says he opposes abortion on principle, look for the mating strategy. When a party says it stands for free speech, look for the coalition that benefits. When a man acts humble, look for the concealed bid. The heuristic is what makes Pinsof readable at a glance, and it is what his critics mean when they say he can explain anything.
Lakatos would say that complaint misfires. Every core can explain anything, because that is what cores are for. The dispositional program in political psychology has one too: political attitudes express stable traits, values, and needs, and John Jost’s system justification account, which Pinsof’s dissertation tested against, holds that people are motivated to defend the arrangements they live under. Confronted with a partisan who reverses his principle when the other side benefits, the dispositional program does not fold. It posits a trait that predicts the reversal, or a measurement problem, or a difference between elites and the mass public. Those are its Neptunes and its Vulcans. The question for both programs is the same. Which adjustments have predicted something new?
Sorted that way, Pinsof’s output falls into two piles.
The progressive pile starts early. The 2016 Psychological Science paper “The Political Divide Over Same-Sex Marriage: Mating Strategies in Conflict?” generated a prediction the rival program did not: that opposition would concentrate among people averse to casual sex who linked homosexuality with promiscuity. The 2017 PLOS ONE follow-up, “The Effect of the Promiscuity Stereotype on Opposition to Gay Rights,” did what Lakatos asks: it derived a novel consequence, that correcting the stereotype should shift support among the sexually restricted and only among them, and then manipulated the stereotype and found the shift. The abortion work follows the same path. The 2016 HBES presentation “Pro-Life or Anti-Promiscuity?” made the prediction. The 2026 preprint “Concerns about Sexual Promiscuity Play an Underexplored Role in Opposition to Abortion” tests it on 3,631 people with data from before and after Dobbs. A prediction made in 2016 and confirmed on data that did not exist until 2022 is a novel fact in the strictest sense.
The double-standards work belongs in the same pile, and its history shows why the publication record matters for a Lakatosian audit. Pinsof told me he wrote the questions for Pulse of the Nation to test Alliance Theory predictions he was then developing. The predictions preceded the data. That is the sequence Lakatos rewards. But the analysis reached the record only in fragments cited inside “Strange Bedfellows” and in a 2018 conference program. An outside reader cannot verify which reversals were predicted in advance and which were noticed afterward. A program can be progressive in fact and look degenerating on paper when its predictions live in an unpublished poll.
The most progressive move is the one that has not happened yet. Pinsof is joining a group at the University of Toronto on a large preregistered study that pits alliances against dispositions as predictors of political views. Lakatos was skeptical of crucial experiments in the moment; he thought they were recognized as crucial only in hindsight, once one program had absorbed the result and the other had not. But a preregistered head-to-head is as close as social science gets to forcing the issue. It commits both programs to predictions before the data arrive. Whichever program has to add a Vulcan afterward will have been caught doing it in public.
The signaling program has its own progressive entries. “The Evolution of Social Paradoxes” does the work Lakatos would want, since it specifies the information conditions under which a concealed signal should succeed and under which it should collapse. The humor preprint “All Mixed Up” predicts that laughter should track coordination failures and that the ability to produce it should read as partner value. The “Human Status-Seeking as 4D Chess” paper he described to me tries to derive charisma, authenticity, humility, and sacred values from a single recursive model. Each of these is content-increasing. Each says something that could turn out to be false about a domain the program had not yet touched. And the positive heuristic is visible. Pinsof knows what he will build next, which Lakatos took as the mark of a program still alive: mating, then alliances, then status, then humor, then friendship, then the formal model that is supposed to hold them together.
The degenerating pile is smaller, and it lives mostly on the blog. The list in my earlier essay is the standard exhibit. A man boasts, status seeking; a man acts humble, concealed status seeking; a man admits craving status, strategic candor; a man denies it, successful concealment. In Lakatos’s terms each line is an ad hoc adjustment of the worst kind, what he called ad hoc3: a hypothesis added to save the appearances that predicts nothing beyond the case it was invented for. “Happiness Is Bullshit Revisited” sits near the line. Pinsof reformulated after criticism, which is allowed; the test is whether the reformulation, happiness as a prediction-error signal, forecasts anything the original did not. “Everything Is Signaling” widens the core to cover defensive signaling, the avoidance of being classified as stupid, disloyal, or desperate. That could be progressive if it specifies which signals get avoided and by whom. As written it reads more like an extension of the belt to cover behaviors the original signaling story had missed.
“AI Doomerism Is Bullshit” raises a different Lakatosian problem. The core says intelligence is a bundle of specialized adaptations. Applied to AI, that core generated a prediction, that general intelligence of the kind doomers fear cannot arise from scaling. Readers pointed out that the strongest doom arguments do not need the assumption Pinsof attacked. That is what an anomaly looks like from inside a program. The question is how he handles it. If the answer is that the critics have misunderstood, the program has bought itself a Vulcan. If the answer narrows the claim to something testable about what scaled systems can and cannot do, the program has bought itself a Neptune.
Now the title of the reply. Pinsof, Sears, and Haselton accuse dispositional political psychology of epicycles. The charge has content. Each new inconsistency in ideology gets a new construct: authoritarianism, then need for closure, then moral foundations, then a dozen scales, each fitted to the anomaly that produced it. That is the shape of a degenerating belt, and Pinsof is right that a theory which starts from coalitions predicts the inconsistencies in advance instead of naming them afterward. But the charge cuts both ways, and Lakatos would insist on saying so. A program that treats every stable value as post hoc coalition propaganda has its own epicycle waiting whenever a value survives a realignment. When a group keeps a principle after its coalition drops it, the alliance program must explain why, and the easy answer, that the group has a different alliance in view, is an epicycle by any other name. Which side owns the epicycles will be settled by which side can go longest without adding one. That is what the Toronto study is for.
Lakatos had one more thing to say that bears on Pinsof. He denied that it is irrational to keep working on a degenerating program. Prout’s hypothesis that atomic weights were integer multiples of hydrogen looked dead for a century and then came back with isotopes. Copernicus took a long time to outpredict Ptolemy. What stops most scientists from persisting is cost, in grants, jobs, and standing. Pinsof told Zach Elwood that Cards Against Humanity removed those costs for him. In Lakatos’s frame that changes nothing about whether his program is progressive and everything about how long he can afford to find out. The same money that lets a Prout wait out the century lets a Vulcan hunter keep hunting. The difference between the two is visible only in the ledger of predictions, and Pinsof’s ledger, for the moment, runs progressive. The abortion result came in. The Dobbs data cooperated. The concealed-signal conditions were stated before they were tested. The head-to-head is preregistered. What would tip the ledger is a run of misses each followed by a new hidden motive. So far he has been reaching for a model instead.
The Great Delusion
John J. Mearsheimer says humans are social from start to finish, that individualism is secondary, that people form strong attachments to their group and sacrifice for it, and that reason is the least important of the ways we arrive at our preferences. Pinsof would sign all of that. Strange Bedfellows opens by rejecting the view that political beliefs derive from abstract values reasoned toward, and it treats the individual as a coalition member first. Both men are Darwinians about sociality: Mearsheimer says the best way to survive is to be embedded in a group, and Pinsof’s alliance psychology is the evolved machinery for doing so. Both think liberalism misdescribes the animal.
The first disagreement is over which group. Mearsheimer’s group is the society one is born into, and his mechanism is childhood socialization: by the time your reasoning matures, your family and society have already imposed a value infusion on you, and you have limited choice thereafter. That is a dispositional account. Values go in early, from a stable source, and stay. It is closer to Malka and Federico, and to Ganzach, than to Pinsof. Pinsof’s group is the political coalition, and his central claim is that values move with the coalition and move fast. A Republican did not absorb his view of the FBI or of Putin in childhood; he acquired it between 2015 and 2018 when his alliance changed. Mearsheimer explains why a man holds his society’s values. Pinsof explains why he reverses them in four years. The two accounts can both be true, one for the deep layer and one for the shallow, but Mearsheimer gives no room for the reversals and Pinsof gives little room for the infusion.
The second disagreement is about nationalism. Mearsheimer holds that the nation is the tribe that matters most in modern life and that national loyalty is the strongest political attachment. Pinsof’s paper argues that partisan alliance overrides national alliance. On Mearsheimer’s theory that should be rare; on Pinsof’s it is what happens whenever a party leader takes a side. Mearsheimer’s own book is a case for Pinsof, since he describes liberal states invoking universal rights and then pursuing them where interest allows, which is the selective application of a principle that Alliance Theory predicts.
The third is about human rights. Mearsheimer treats rights universalism as a sincere ideology that motivates liberal states into ambitious foreign policies. He thinks it is mistaken about human nature but takes it at its word as a motive. Pinsof would not. On his account “everyone has the same rights” is a banner that a coalition raises for its allies and lowers for its rivals, and the test is whether the same people who invoke it for one victim group invoke it for another. Mearsheimer’s realism is close to that in practice, since he thinks states follow interest and dress it in rights, but he locates the sincerity in the ideology and the hypocrisy in the state, while Pinsof locates both in the individual mind.
Where Mearsheimer sits nearest to someone in the Pinsof debate is DeScioli. Mearsheimer’s inborn sentiments that shape right and wrong before reasoning arrives, and his limited choice in formulating a moral code, are a claim that moral machinery exists prior to any alliance. That is DeScioli’s point against Pinsof, that impartial principles evolved to restrain alliances.
So the passage supports Pinsof against liberalism and against the rationalist picture of belief, and it supports his critics on the two questions that matter: whether values are stable dispositions laid down early, and whether some moral sentiments are prior to the coalitions that use them.
Highland Park
Highland Park is a wealthy suburbs in Illinois, on the lake north of Chicago, with a median household income well over the state’s and a high school that sends its graduates to selective colleges as a matter of course. It is heavily Democratic and it is also heavily Jewish, with Deerfield next door in the same category, and the surnames of the eight men who founded Cards Against Humanity, Temkin, Halpern, Hantoot, Dranove, Weinstein, Munk, Dillon, Pinsof, read as a roll call from that community. This is the milieu of educated, secular-leaning, prosperous, liberal Jewish Chicago, the world John Hughes filmed in the town’s driveways. On Mearsheimer’s account, that world did the value infusion.
What does it infuse?
First, a politics. The town votes Democratic by margins that make the other party a rumor, and the intellectual class Pinsof spends his time with, by his own description, is allied with the Democratic Party. Mearsheimer would predict that a child of Highland Park emerges with progressive commitments as his default and that his later reasoning works around them. Pinsof’s blog attacks that coalition’s intellectuals more than any other, and his theory says the attacks are position-taking inside the coalition. His targets are the Democrat-allied academy, the rationalists, and the effective altruists, which is to say the neighbors, and never the neighborhood. He does not write about group differences. He does not write about the moral claims of his own company. Those are the things a Highland Park child learns not to say before he learns why, and Mearsheimer’s point is that the learning comes first and the reasons are supplied later. Pinsof’s theory says the same thing about everyone else. Highland Park is where it was demonstrated on him.
Second, an idiom. The comedy of that community is irreverent, transgressive, self-aware, and aimed at pieties, and Chicago’s improv culture sits twenty miles south. Cards Against Humanity is that idiom in a box: a group of friends handling the forbidden together to prove no one present is fragile. Pinsof told me comedy and psychology attract him for the same reason. Mearsheimer would say the attraction is socialized, that a boy raised where wit is the currency of standing will seek standing through wit, and that the choice to become a Darwinian cynic was made in the register available to him. The cynicism is a Highland Park accent.
Third, the tribe. Mearsheimer says the deepest attachment is to the group one is born into, and he means the nation. In Pinsof’s life the deepest attachment visible in the record is to eight boys from one high school who kept meeting during college vacations, built a company as equals, and have held it together for fifteen years through a public crisis that cost one of them his role. Pinsof stayed. The band with Hantoot and Halpern, the unreleased music, the shared address at the company, the money that funds his freedom: every one of those is the friend group persisting. Mearsheimer’s man makes great sacrifices for his fellow members, and the silence Pinsof has kept about 2020 is the sacrifice in this case. His own theory would call it coalitional loyalty. Mearsheimer would call it what humans are.
Fourth, the layered loyalties. Mearsheimer wrote his most controversial book about the pull of Israel on American Jewish politics, and a child of Highland Park grows up inside that pull whatever he concludes about it. Pinsof’s one professional engagement with the subject is the UCLA grant on hatred of Jews, framed as engineering, inputs and conditions, the coldest register he has. Mearsheimer would read the coldness as the buffered method of a man whose socialization made the subject hot, and would predict that a product of that town cannot approach it neutrally.
Highland Park socializes its children into liberal individualism, the ideology Mearsheimer says ignores our social nature, and it does so by means of an unusually dense tribe: one high school, one community, one set of families, one idiom. The professed values are universal rights and reason; the practiced life is loyalty, status competition, and group defense. A perceptive child raised there would notice the gap between what the town says and what it does. Pinsof’s theory is the gap. Alliance Theory says political values are the propaganda of a coalition; “Status Is Weird” says professed indifference to rank is a rank strategy; “A Big Misunderstanding” says intellectuals invent the misunderstanding myth to make themselves indispensable. Every one of those is a description of an affluent liberal suburb from the inside, generalized to the species. Mearsheimer’s anthropology predicts that a Highland Park child cannot escape his socialization. Pinsof’s career suggests that he can do one thing with it, which is to describe the machinery accurately while remaining inside it, and to call the result a theory of everyone.
Chicago’s homicides concentrate in a dozen neighborhoods on the South and West Sides, Englewood, Austin, Garfield Park, North Lawndale, and the victims and perpetrators are overwhelmingly Black. Highland Park sits twenty-five miles north on the lake, across a county line, with a homicide rate near zero in most years. A child raised there in the 1990s and 2000s knew about the violence the way he knew about weather in another state. It was in the newspaper, it was true, and it never touched him. That fact, distance from a danger one knows to be real, is the condition under which most of the town’s politics were formed, and it is where Mearsheimer and Pinsof converge on the same prediction from different directions.
The Jewish community of the North Shore is largely the Jewish community of Chicago’s West Side one or two generations removed. North Lawndale was a Jewish neighborhood until the 1950s. Its synagogues are now churches. The families moved north as the neighborhood turned Black and turned poor, and the moving was a group decision made by a tribe protecting itself, which is Mearsheimer’s model of how humans behave. So the child of Highland Park is raised inside a community that professes the civil-rights alliance between Jews and Blacks, remembers Skokie and the Freedom Rides, and lives in a town its grandparents built by leaving. That is a value infusion with a contradiction folded into it: solidarity as a stated principle, separation as a practiced one, and a family memory of why. Mearsheimer would say the memory is the deeper layer and the principle the shallower, and that the child will hold both without noticing the seam until something forces it.
Pinsof’s theory says what happens next. “A Big Misunderstanding” argues that stereotypes are accurate, that psychologists suppress the finding because it makes them look insufficiently progressive, that bigotry tracks coalition because ethnic minorities are accurately perceived as Democratic allies, and that antiracism confers elite status because the elite’s nearest status rivals are low-status Whites. Read that paragraph as the view from Highland Park. A resident there knows the crime numbers, knows which neighborhoods produce them, and holds antiracist commitments that cost him nothing and raise his standing; the people for whom the same knowledge is a daily calculation are the working-class Whites and Blacks of the city, whom the resident outranks. Pinsof has written the sociology of his own town’s position and generalized it to the species. The theory predicts that liberal Jews in Highland Park will profess equality, practice distance, send their children to a high school that is a few percent Black, back police in their own town while their coalition’s rhetoric turns on police, and experience no contradiction, because the coalition’s principles are instruments and the instruments are applied selectively. The 2020 season is the test case. Alliance Theory predicts that towns like Highland Park would sign the coalition’s statements about policing and change nothing about their own, and that is what most such towns did.
The relationship with Chicago’s Black political world adds a third strain. The city that produced the civil-rights alliance also produced the Nation of Islam, headquartered on the South Side, and Jesse Jackson’s 1984 remark about New York. A Jewish child of the North Shore grows up with both the alliance narrative and the knowledge that the alliance is not returned by everyone in it. Mearsheimer would file that under layered tribal loyalty. Pinsof’s theory would predict that the strain gets managed by proxy: hostility toward the far end of the coalition is not expressed, because expressing it would cost coalition standing, and it surfaces instead as the cold clinical register in which the subject can be handled at all. His one professional engagement with hatred of Jews is a grant framed as engineering.
Then July 4, 2022. The one act of mass violence in Highland Park’s modern history came from inside, a local White man with a rifle on a parade route, seven dead. The town had banned assault weapons in 2013 and defended the ban in court. Its response to the shooting was more of the same, and the framing ran through the coalition’s channel, gun control, because that is the framing the coalition supplies.
A liberal Jewish child of Highland Park is raised to hold egalitarian principles by a community whose safety rests on distance, whose history includes flight, and whose politics are cheap in the sense Mearsheimer gives the word. He learns the crime numbers and learns not to draw the conclusions the numbers would draw for a man living inside them. If he is perceptive he notices that the principles and the practice diverge, and if he is honest he needs a theory that lets him say so without leaving the coalition. Alliance Theory is that theory. It says everyone does this, that values are coalition instruments, that stereotypes are savvy, and that the only misunderstanding is thinking there has been one. It lets the Highland Park child tell the truth about his town by telling it about the species, which keeps him inside the town while he does it. Whether that is an achievement or a limit is the question the whole essay has been asking, and the answer is both.
Pinsof writes about race constantly and about racial group differences never, and the gap is the same shape as the one in evolutionary psychology generally.
In Strange Bedfellows race is a coalition marker. Republicans feel greater allegiance to White people than to Black people and Democrats the reverse, and the paper predicts the propaganda that follows: perpetrator biases that excuse one’s own side’s group, victim biases that inflate the grievances of allies, attribution biases that explain allies’ disadvantages by circumstance and rivals’ by character. The paper’s cleanest example is reparations, where support for compensating Holocaust survivors is bipartisan and support for compensating the descendants of slaves is partisan, which the authors read as evidence that perpetrator biases attach to local political alliances. The history section treats the Civil Rights Act of 1964 as the first of four realignments and the splitting of the working class along ethnic lines as the third. The paper’s Israel and Putin examples are meant to show that conservatives are not ethnocentric or nationalistic in any general sense, only allied. Race, on this account, is one of the cheap, visible, hard-to-fake similarities that alliance psychology uses, and the evidence he leans on is the Kurzban, Tooby, and Cosmides result that the mind stops tracking race as a category the moment a cross-cutting coalition becomes visible, extended by Pietraszewski, his podcast cohost. Race is a cue and contingent as a category.
“A Big Misunderstanding” goes further. Stereotypes about religious, ethnic, occupational, and geographic groups are accurate, he says, the accuracy is one of the best-replicated findings in psychology, and psychologists suppress it out of fear of looking insufficiently progressive. Bigotry is then explained entirely by alliance and status: hostility toward ethnic minorities tracks hostility toward Democrats, because minorities are accurately perceived as Democratic allies, and the rest comes from status competition, most acute between minorities and low-status Whites, which is why antiracism confers standing on elites whose nearest rivals are other elites. “Everything Is Signaling” adds that most signaling is defensive, aimed at avoiding classification as stupid, disloyal, or immoral, and the paradigm case of an immoral classification in his readers’ world is the racist one, though he does not say so.
Pinsof has published nothing, in the papers or on the blog, on whether groups differ in cognitive or behavioral traits for genetic reasons, nothing on IQ, nothing on the hereditarian debate that runs through the heterodox world his readers inhabit. The “I Am Superhuman” post argues that “human nature” is invoked selectively as an explanation. When his theory reaches the place where group differences would enter, as in the bigotry paragraph, it explains the phenomenon by alliance and status and moves on. The stereotypes are accurate; what they are accurate about is not discussed.
Three reasons are available.
The principled one is that the question is orthogonal to his program. Alliance Theory is a theory of beliefs about groups, not of groups. Its point is that coalition membership is historically contingent and that no party or ethnic group owns a timeless political essence, and it sits inside an evolutionary psychology whose founding commitment is species-typical architecture. A man whose whole apparatus is built to show that group-level politics is downstream of shifting alliances has no theoretical use for fixed group-level traits, and would say the accuracy of a stereotype about, say, occupational groups tells you nothing about causes. That is a coherent position and he could state it in one paragraph.
The disciplinary one is that evolutionary psychology’s standing in the academy has depended for thirty years on the line between species-typical claims and group-difference claims, and everyone in his set, Haselton, Krems, Pietraszewski, Cosmides and Tooby before them, keeps the line. Pinsof is a research scientist in a lab, not a tenured professor, and the line is a condition of the affiliation.
The coalitional one is his own theory. In the four-question exercise I put it as the truth that would cost him his position, and the bigotry paragraph is the evidence that he knows where the line is, since it walks up to it and applies the theory to everything on the near side. Alliance Theory predicts that a man will not state the principle that costs him his coalition and will experience the omission as a lack of interest. He could not, on his own account, tell you which of the three reasons is operating.
Everything about race that can be explained by who is allied with whom, he explains that way, in detail and with data. Everything that would require a claim about groups themselves, he leaves to the stereotype-accuracy literature and does not open. The hatred-of-Jews grant follows the same rule: an engineering perspective, inputs and conditions, the coldest register available, applied to the one group question he cannot avoid. If you want to know where he stands, the question to ask him is the narrow one his theory permits: does he think the accuracy of ethnic stereotypes is a coalition-independent fact, and if so what does his theory say about what it is accurate about. He has invited questions. That is the one that would show whether the silence is principle or price.
The Fearless Pinsof
What would be logical extensions of his work if Pinsof had no fear about loss of status and standing?
The question assumes a figure his own theory says cannot exist, so treat it as a limit case: what the theory points at when the cost of following it is set to zero.
The first extension is the one the theory begs for and he has not run: Alliance Theory applied to the academy that houses it. He has written that psychologists suppress stereotype accuracy out of fear of looking insufficiently progressive, that social science collects misunderstandings that serve the Democratic coalition, and that intellectuals inflate their own importance. What he has not written is the paper that would follow: a map of the academic alliance structure, which findings each coalition is permitted to publish, and a prediction, from the theory, of where the consensus will be wrong. The place the theory points at is his own discipline’s line between species-typical and group-difference claims. Alliance Theory says a principle that costs a coalition nothing to hold and everything to abandon is an alliance instrument. The line is exactly that. An unbound Pinsof would ask what the line is for, in the same register he used for the FBI and the WHO, and would predict when it will move. He would not need to take a side on the empirical question. The theory’s contribution is to say that the consensus is coalitional whichever way the facts go, and he has already said that about everyone else’s consensus.
The second is the stereotype-accuracy paper he has cited and never written. He says stereotypes about ethnic, religious, occupational, and geographic groups are accurate and the finding is suppressed. The theory then owes an account of what accurate stereotypes are accurate about and why coalitions differ in which ones they may state. An unbound version would run the double-standards method on stereotype endorsement: which groups may be characterized by which coalition, with the same trait, and what flips when the alliance flips. That is a study he could preregister tomorrow. It would be the most cited thing he ever did and it would end his affiliation with UCLA and Cards Against Humanity.
The third is the Jewish case. He holds a grant on hatred of Jews framed as engineering. Alliance Theory has a natural and unsentimental account of antisemitism: Jews as a group that sits in the elite half of the class split he describes, allied with the Democratic coalition, perceived as rivals by the low-status Whites nearest them in the hierarchy and by the minority coalitions that compete with them for the progressive elite’s patronage. The theory predicts that hostility toward Jews will be voiced by whichever coalition finds them on the other side that year, that the principle “antisemitism is unacceptable” will be applied with the same double standards as every other principle, and that the campus politics since 2023 are a realignment in progress with the propaganda being rewritten as he watches. Israel is the second half: coalition turnover every few years, positions on the draft and the courts moving with the coalition, the best non-American test of the theory anywhere, and Ganzach’s country. A Pinsof without fear would write that paper and would say what it says about his own community and Highland Park. Mearsheimer wrote the version from the outside and it cost him a decade. Pinsof’s version would come from the inside and would cost more.
The fourth is the theory applied to the coalition that pays him. Cards Against Humanity is a company that makes moral gestures and funds progressive causes; the entertainment world he and his wife work in runs on stated values and practiced alliances; the Kilroys is an alliance formed to change an institution’s output by changing who counts. Every one of those is a case study for Alliance Theory and every one is off limits. “Money Is Bullshit” is where he came closest and it turned into a defense of marriage. The unbound version treats the comedy industry the way “A Big Misunderstanding” treats the academy: as a hierarchy that dresses status competition in moral language. He has the data and he has the standing. He also has seven co-owners.
The fifth extension is inward and he has half-done it. The blog’s readers are a coalition, heterodox, male, rationalist-adjacent, and hostile to the progressive academy, and his theory says his positions are badges for it. He has confessed this in general and never in particular. The particular version would name what his audience will not let him say: that the heterodox right’s principles, free inquiry, viewpoint diversity, the courage to say the unsayable, are coalition instruments too, applied when the coalition benefits and dropped when it does not, and that the audience’s cynicism about progressive moralizing is a moral performance for a rival hierarchy. He did a version of this to the rationalists and it cost him a wing of his readership. Doing it to the rest would cost the rest. It would also be the paper that answered Jost, since it would demonstrate the symmetry on the side that pays him.
The sixth is the formal one, and it is the least dangerous and the most valuable. Alliance Theory predicts realignment and cannot yet say when. An unbound Pinsof would build the model of coalition dissolution that would settle his disagreement with Ganzach: what holds a coalition of dissimilar parts together, how long it survives the loss of the leader or the rival, and which principles get rewritten first. Applied to the Trump coalition it would be a forecast with a date. Applied to Israel it would be a test. Applied to the academy it would be a prediction of when the group-differences line moves. That is the piece of work that would make him the Pareto of his generation, and it is available at no coalitional cost, because a model is not an opinion.
Where a bloke like Pinsof goes is usually one of three places. Mandeville’s, where the man writes the scandalous truth under cover of satire and the satire is taken as the point; Pinsof is halfway there and the blog’s titles are the cover. Pareto’s, where the man leaves the coalition altogether, writes the theory of elites in retirement, and is claimed after his death by people he would have despised; Pinsof has the money for the retirement and the theory to be claimed. Or Kurzban’s, his own teacher’s teacher’s line, where the man writes the theory of hypocrisy, keeps the affiliation, and lets others draw the applications. That is where he is now.
His theory says the third is where he will stay, because status seeking must disguise itself and the disguise that works best is the one the seeker believes. The evidence against the theory is the AI doomerism essay, the one time he spoke against his audience. A man who did that once might do it again. If he does, the paper to watch for is the one on his own discipline’s line, because it is the one his theory has been predicting for three years and the one he has the least to gain from writing.
Alliance Theory on Academic Psychology
What happens if we apply Alliance Theory to the institution that produced Alliance Theory?
In “Psychologizing Psychologists”, Pinsof argues that psychologists should be treated as ordinary status-seeking, self-serving human beings. He mocks the arrangement by which psychologists discover biases in everyone except psychologists, and he includes himself in the indictment. Researchers compete to produce interesting findings, want status, want to look clever, and political psychologists are political animals. His remedies are stronger theory, preregistration, replication, adversarial collaboration, and incentives that reward truth over interest.
In “A Big Misunderstanding” he says intellectuals imagine the world’s problems as misunderstandings because a world ruined by misunderstanding makes the suppliers of understanding important. Social science accumulates findings about bias, irrationality, and misinformation that give intellectuals a heroic corrective role, and mainstream social science often advances political agendas. Then the sentence that matters here: psychologists have suppressed evidence for stereotype accuracy because they fear looking insufficiently progressive.
That is an accusation about the sociology of scientific knowledge, and there is a literature behind part of it. Lee Jussim, Jarret Crawford, and Rachel Rubinstein’s “Stereotype (In)Accuracy in Perceptions of Groups and Individuals” in Current Directions in Psychological Science reviews evidence that many aggregate stereotypes correspond substantially with measured group characteristics. Their complaint is that social psychology devoted far more attention to stereotype inaccuracy than the accuracy literature warranted. Pinsof’s explanation is coalitional. Psychologists were managing reputational danger.
If that is true, the next paper would map the academic alliance structure the way “Strange Bedfellows” maps American politics. Who are the allies and rivals? Which constituencies control professional prestige? Which findings raise an investigator’s standing with them and which impose reputational cost? Which propositions can be stated without signaling anything dangerous, and which require a display of benevolent intention before anyone will discuss their evidentiary merits?
Yoel Inbar and Joris Lammers surveyed 800 social and personality psychologists for “Political Diversity in Social and Personality Psychology” in Perspectives on Psychological Science. Six percent described themselves as conservative overall. Conservative respondents reported concern about disclosing their beliefs, and some respondents acknowledged willingness to discriminate against openly conservative colleagues in decisions about papers and hiring. The authors also found more ideological diversity on economic and foreign-policy questions than the headline classification suggested.
José Duarte, Jarret Crawford, Charlotta Stern, Jonathan Haidt, Lee Jussim, and Philip Tetlock made the broader argument in “Political Diversity Will Improve Social Psychological Science” in Behavioral and Brain Sciences. Their claim went past demographics. They proposed routes by which ideological homogeneity alters the science: values enter the construction of research questions, politically unattractive hypotheses go understudied, concepts get operationalized in ways that assume one side’s reading of a dispute, and dissenting perspectives disappear before a study is run.
Alliance Theory can improve on that literature because it makes a more discriminating prediction. Homogeneity predicts a general leftward distortion. Alliance Theory predicts selective distortion. The treatment of an idea should depend on what that idea does to the alliance structure.
Consider a species-typical evolutionary claim. A psychologist says humans evolved to track alliances, detect cheaters, compete for status, choose mates, or manage reputation. These are cynical claims about human nature and they annoy people who prefer nobler descriptions. Politically they are cheap. If everyone possesses the adaptation, no contemporary group gains status at another’s expense. The claim can be depressing without destabilizing any coalition.
Now alter one feature. Instead of an evolved characteristic of humans in general, propose a persistent average difference between politically salient groups. The empirical merits have not been established, and that is the point. Before anyone knows whether the claim is true, its coalition consequences can be calculated. One possible result weakens an environmental explanation for an inequality; another strengthens it. One implies that ending discrimination would leave a disparity in place; another implies discrimination explains more than previously thought. Each supplies argumentative resources to one coalition. Alliance Theory says those differences should matter before the science is settled.
This is where the recursive application bites. A principle that costs a coalition nothing to affirm and becomes expensive when applied to a particular ally or rival is the kind of principle Alliance Theory teaches us to distrust as an independent cause. “Follow the science” is easy when the science threatens nobody in your coalition. “Evidence should determine belief” is cheap when the evidence points where your allies already want to go. The informative case begins when the same epistemic principle generates a costly conclusion.
The best evidence here comes from Cory Clark and colleagues’ “Taboos and Self-Censorship Among U.S. Psychology Professors”, published in Perspectives on Psychological Science. The researchers first interviewed scholars to identify conclusions regarded as taboo, then surveyed 470 psychology professors. In the pilot interviews, the conclusions most often nominated as taboo were genetic, evolutionary, biological, or otherwise natural explanations for socially important group differences, especially differences where women performed less well than men or Black people less well than White people. That is the boundary Alliance Theory would tell you to inspect.
The Clark study complicates the culture-war version of the story, which is why it is useful. The professors were left-skewed; more than 90 percent placed themselves at or left of the midpoint on the study’s scale. Professors who believed taboo propositions were more likely to report self-censorship. Many feared social penalties for stating their empirical beliefs, and tenured professors reported about as much self-censorship as untenured ones. Yet most professors did not favor suppressing controversial scholarship. They opposed retracting papers or punishing researchers because conclusions might be harmful, and average support for discouraging research into the taboo propositions was low. The authors warn that the response rate was a little above 10 percent and that the survey cannot establish whether any controversial proposition is true.
Those qualifications make the finding more useful. The academy needs no censorship committee for Alliance Theory to operate. If most people sincerely endorse academic freedom and a researcher still expects ostracism, reputational damage, or professional suspicion from a motivated minority, the equilibrium sustains itself. Enforcement is distributed. Nobody issues an order. Researchers anticipate the reactions of their allies and adjust. That is a better model of a taboo than a dean in an office deciding which truths may be spoken, and it is Pinsof’s own model of how a coalition’s beliefs converge without anyone commanding them.
David Buss and William von Hippel approached the problem from inside evolutionary psychology in “Psychological Barriers to Evolutionary Psychology: Ideological Bias and Coalitional Adaptations” in Archives of Scientific Psychology. Scientists, they argued, possess coalitional adaptations, status motives, and persuasion instincts that interfere with judgment, and evolutionary psychologists are no exception: they can form their own coalition, caricature rivals, and acquire incentives to find results congenial to their camp. Pinsof’s contribution would be to formalize what Buss and von Hippel described.
The paper would replace the question “Is this topic taboo?” with measurable quantities. How much professional friction does a proposition encounter? How reluctant are researchers to test it? How many qualifications surround it? How often are moral consequences discussed alongside methodological ones? How much stronger must evidence become before the conclusion counts as established? How often do reviewers demand alternative explanations? How likely is the result to draw a press release, a replication effort, a correction, a retraction demand, or a public denunciation? “Permitted to publish” is too binary. Alliance Theory should predict a friction coefficient.
There is an old experimental model for testing it. Stephen Ceci, Douglas Peters, and Jonathan Plotkin’s “Human Subjects Review, Personal Values, and the Regulation of Social Science Research” in American Psychologist sent human-subjects committees matched hypothetical proposals whose treatment of participants was constant while the sociopolitical content changed. Socially sensitive proposals were rejected more often than neutral versions, and among the sensitive ones, proposals investigating reverse discrimination had more trouble winning approval than proposals investigating conventional discrimination. The institutional environment was 1985 and the numbers should stay there. The design is what matters: change the coalition implications while holding the science constant.
That is the architecture of the Alliance Theory experiment. Give reviewers the same methods, sample size, preregistration, effect size, and statistical uncertainty. Change only the identities of the groups or the political implication of the hypothetical result. Does perceived methodological quality move? Does the demand for replication move? Does concern about confounds move? Does the recommendation to publish move? Swap male and female targets. Swap racial groups where a design permits. Swap a group associated with the left for one associated with the right. Swap evidence of discrimination against an allied group for equivalent evidence involving a rival. Swap an environmental explanation of a disparity for a biological one while matching evidentiary strength as closely as a hypothetical allows. Then preregister the Alliance Theory prediction before collecting the data.
The theory generates a prediction about null results. A null that frustrates an allied grievance should be treated differently from a null that frustrates a rival’s. One gets described as underpowered, premature, or insensitive to structural effects; the other gets accepted as evidence that the problem was exaggerated. Either reading could be right in a particular study. The question is whether the standard moves with the coalition consequence.
It generates a prediction about causal explanations. Dispositional explanations should become more attractive for the failures of rivals and less attractive for the failures of allies, with structural explanations showing the reverse. Pinsof’s theory already makes that prediction about voters. Academic psychology is a cleaner environment in which to test whether scientists do the same thing while believing they are following method.
It generates a prediction about language. Coalition-costly findings should attract the words harmful, essentialist, reductionist, stigmatizing, irresponsible, and easily misused more often than equally strong coalition-useful findings. Sometimes the words are earned. The claim concerns the asymmetry: does the danger become salient because of a philosophy of scientific responsibility, or because this result disadvantages this ally?
It generates a prediction about selection before publication. The largest distortion may sit in the studies nobody proposes. A literature can contain individually competent papers and still present a distorted picture if investigators ask only questions whose answers vary within a comfortable range. Publication bias refers to findings that disappear after data are collected. Alliance Theory points earlier, to hypothesis-selection bias.
It generates a cross-national prediction. If academic taboos arise from universal moral principles, the same propositions should meet similar resistance wherever competent scientists evaluate them. If taboos reflect alliance structures, the boundaries should shift across political cultures, and reputational cost should track the local alliance map better than the words of the proposition.
Most of all, it generates a temporal prediction, and this is where the FBI example comes back. Alliance Theory earns its keep when an institution or principle changes political meaning faster than psychology could plausibly have changed. If attitudes toward the FBI move when the FBI’s alliances move, a stable principle of respect for authority is not doing the causal work. Do the same with a research proposition whose evidence base stays stable while its political meaning changes. Does a dangerous idea become respectable when it starts helping a different coalition? Does a fashionable line of research acquire methodological defects once its implications turn inconvenient? Do scientists notice confounds that had existed all along? That is a far stronger claim than saying academics are biased, and it is the claim the Pinsof, Sears, and Haselton reply to their critics commits them to, since they wrote there that alliances measured independently of ideological labels should strongly predict beliefs or the theory is in trouble. The academy is a population in which alliances can be measured independently of the beliefs, by department, by funder, by citation network, by who reviews whom.
The theory is symmetric. If Alliance Theory is right, conservative scientists, religious universities, military institutions, corporations, and right-leaning intellectual communities do not become impartial when they control the locally relevant prestige system. A conservative-dominated discipline should develop its own protected propositions. A corporation funding research should become sensitive to methodological weakness in findings that threaten its interests. A religious institution should discover epistemological reasons to doubt evidence costly to its commitments. Without the symmetry the theory degenerates into one coalition’s propaganda about another, which is Jost’s charge against it and the charge the design must answer in advance.
That symmetry has a reflexive edge. A study showing that the psychological academy suppresses coalition-costly findings is a coalition finding, for the heterodox network that reads Pinsof and for the political right that would cite it. Alliance Theory predicts its own reception: the paper would be embraced as a badge by one side and dismissed as one by the other, and the embrace would be selective, with the paragraphs about corporate and religious institutions going unquoted. The design has to anticipate that. The friction coefficient has to be measured in at least one prestige system that leans the other way, and the preregistration has to state what result would embarrass the heterodox coalition as well as the academic one. A paper that only finds distortion on the left has found what its audience paid for.
Alliance Theory can identify where consensus is least independent of social incentive. The reliability of a scientific consensus should decline, all else equal, as the reputational asymmetry between possible findings increases. Where every answer is equally safe, ordinary error remains. Where one answer carries rewards and another carries coalition cost, a new source of selection enters the system, and it operates at every stage: who enters the field, which question gets asked, which hypothesis gets funded, which design survives review, which result gets written up, which caveats reviewers demand, which paper gets cited, which result gets publicity, which researcher becomes radioactive, and which graduate student decides that another topic would make for an easier life.
That is what makes the application Pinsofian. His theory explains Republican reversals without a secret committee ordering voters to change principles. People sensitive to alliances converge on locally useful beliefs and experience them as sincere. The academic version works the same way. Nobody telephones a psychologist to say suppress this. The psychologist needs only a normally functioning brain, an accurate model of the local prestige hierarchy, and a career.
The species-typical versus group-difference boundary matters, then, because the two kinds of claim carry different coalition prices. A universal claim about evolved psychology can be intellectually radical and socially cheap. A claim about politically salient group differences can be methodologically ordinary and socially expensive. That inversion is where Alliance Theory says to look, and the Clark study says the boundary is more precise than species versus group: the sensitive cases were natural explanations of differences with recognizable implications for arguments about inequality, sex, race, merit, and discrimination. A taboo map that follows coalition consequences is what the theory predicts.
An unbound Pinsof would ask what that boundary is for. He would ask why an evolved capacity is an insight into human nature in one paragraph and a dangerous essentialism in the next. He would ask whether the methodological objections appear before or after the coalition consequence becomes visible. He would ask whether scientists can predict one another’s political reactions to findings more accurately than the findings themselves, and which academic principles survive when the group labels are reversed. Most of all he would predict when the boundary will move.
He has told us that political principles follow alliances, that psychologists are ordinary humans with status motives, that his discipline suppressed a literature for reputational reasons, and that the people who psychologize everyone else must be psychologized. The remaining step is small, and it is the one his own theory predicts he will not take, since he holds a research scientist’s post in a lab inside the coalition the paper would map. That prediction is a test. If the paper appears, the theory has a counterexample in its author. If it does not, the theory has one more confirmation, and the only man who could have written it will have shown why nobody does.
Stereotype Accuracy
David Pinsof has already said the dangerous part. What he has not done is turn it into the experiment his own theory requires.
In “A Big Misunderstanding,” Pinsof rejects one of social psychology’s stories about stereotyping, the one in which stereotypes are cognitive mistakes waiting to be corrected by better educated people. “Our beliefs about religious, ethnic, occupational, and geographic groups are pretty accurate,” he writes. He calls stereotype accuracy one of psychology’s most robust findings, then explains why most educated people have never heard of it. Psychologists, he says, suppressed the finding because they feared appearing insufficiently progressive.
That is a large accusation and, by the standard of his own theory, an incomplete one. Alliance Theory says that invoking progressivism as the explanation explains nothing, since progressivism is itself a bundle of positions that needs explaining. If political principles are generated by alliances, then the theory has to identify the alliance structure underneath the taboo. Which groups are protected? Which are available for derogation? Which generalizations can be made about an ally and which about a rival? What happens to the rule when the alliance changes? That is the stereotype paper Pinsof has cited his way up to and never written.
There is a serious literature behind the first half of his claim. Lee Jussim, Jarret Crawford, and Rachel Rubinstein’s “Stereotype (In)Accuracy in Perceptions of Groups and Individuals” argues that stereotype accuracy is among the larger and more replicable effects in social psychology. Accuracy needs unpacking before anyone gets carried away. The researchers distinguish correspondence from discrepancy: people can correctly perceive which of two groups is higher on a characteristic while exaggerating the size of the gap. They distinguish consensual stereotypes, the average belief in a population, from the stereotypes held by particular people. Those distinctions separate an empirical claim from the slogan that stereotypes are true. In the studies Jussim and his colleagues reviewed, consensual beliefs about demographic and occupational differences often corresponded strongly with criterion data. Political stereotypes got the direction right while exaggerating the magnitude. National character stereotypes measured against personality inventories performed badly; Robert McCrae, Jussim, and a large international group found weak relationships between national character stereotypes and measured national personality.
Perhaps nobody can answer whether stereotypes are true in general. The serious problems are which stereotypes are accurate, what kind of accuracy they have, how people learn the relevant differences, where they exaggerate and where they suppress, and why some statistically defensible descriptions of groups count as ordinary observation while others count as morally contaminated thought.
Jussim has been asking versions of this for years. Social Perception and Social Reality argued that social psychology had exaggerated bias and underestimated accuracy. More useful for Pinsof, Jussim, Sean Stevens, and Nathan Honeycutt’s “Unasked Questions About Stereotype Accuracy” asked why a field had neglected certain questions, proposed cognitive, motivational, and social-incentive explanations, and suggested testing whether researchers who identify with social-justice goals treat evidence of accuracy and inaccuracy asymmetrically, and whether scientists fear sanctions for studying politically charged accuracy. That paper stops one step short of Pinsof. It has the sociology and lacks the theory of who protects whom.
Pinsof supplies that. In “What Are Ideologies All About?” he points to Democrats who reject negative stereotypes based on place of birth while describing Southerners as racist, and who insist on equality between the sexes while repeating favorable stereotypes about women’s empathy and multitasking. Whatever one makes of the examples, he had located the anomaly: the variable may be the identity of the stereotyped group. He even ran the pilot. The Southerners item came from Pulse of the Nation, the poll he wrote the questions for in 2017 and 2018, and “Strange Bedfellows” cites it. He also owns the experimental template. At the 2018 Human Behavior and Evolution Society meeting he presented “Double Standards of the Political Mind: Empirical Support for the Alliance Theory,” whose logic is simple. If positions derive from principles, people should apply the principles consistently. If principles are weapons for defending allies and attacking rivals, the rule should change when the beneficiary changes. He found double standards in both parties.
Run that method on stereotyping.
The dependent variable should be something other than stereotype endorsement. What Pinsof needs to measure is stereotype license. Give subjects a statement about a group and ask several questions. How accurate is it? How certain are you? How acceptable is it for an ordinary person to say it, for a journalist to print it, for a scientist to investigate it? If a high-quality study supported it, should a journal publish the result? Would saying it make you think less of the speaker? Would you punish an employee, professor, or public figure for saying it? Those measure different things. Factual belief differs from social permission, and scientific legitimacy from personal approval. A person can believe a proposition is probably true and regard saying it as vicious; another can doubt its accuracy and consider it legitimate to investigate. Social psychology blurs these judgments because the word stereotype already carries a moral charge. Pinsof should separate them, and then manipulate the coalition.
The first experiment should use no real ethnic groups. Invent two populations. Tell subjects that Group A and Group B differ by a specified amount on an innocuous but socially meaningful characteristic, and give everyone the same evidence. Tell one set of participants that Group A is allied with people like them and Group B with their opponents; reverse the alliances for another set. Nothing about the evidence, the size of the difference, or the trait changes. Only the coalition changes. Then ask whether it is acceptable to describe the group by the trait. If estimates of the statistical difference stay stable while willingness to state, investigate, or punish discussion of it shifts with coalition, Pinsof has separated belief from license, and he has done it in a design no argument about an existing racial or religious stereotype could match. The result to look for is an interaction: the effect of the description depends on the subject’s alliance with the target, and the same evidence acquires a different moral meaning attached to friend or foe.
People may hold a general principle against characterizing groups, in which case objections should stay stable across alliance manipulations. Their concern may be foreseeable harm, in which case judgments should track the severity and consequences of the description after coalition is controlled. It may be historical vulnerability, in which case status and history should outperform alliance; Jake Womick, Kurt Gray, and colleagues have published twelve studies arguing that assumptions about who is vulnerable to harm drive partisan moral disagreement, and their account is the natural competitor here. People may simply distrust negative claims more than positive ones, in which case valence should dominate. Alliance Theory earns its keep only if it beats those alternatives with the labels swapped. That is what separates a theory from a grievance. “Progressive psychologists suppress uncomfortable truths” is a grievance until someone says what they would suppress, under what conditions, compared with what control, with what predicted effect size.
The second study should use real groups and objective criteria. Choose characteristics with good population data. Estimate the distributions before collecting stereotype judgments. Measure correspondence and magnitude. Then measure each participant’s alliance with each group directly, through warmth, common interests, willingness to defend the group, perceived common enemies, electoral alignment, and willingness to accept costs on its behalf. Alliance Theory itself should refuse that shortcut. Two questions then run at once. Does alliance predict error in estimating the group? Does alliance predict permission to state an accurate estimate?
Those are different predictions. A strong version of the theory predicts distortion at the level of belief: people underestimate negative traits among allies and exaggerate them among rivals. A darker result would be dissociation. Subjects might know what the data show while policing who may say it. The function of the taboo would then be the regulation of speech in the service of coalition management.
Philip Tetlock got close to this a quarter century ago. In “The Psychology of the Unthinkable: Taboo Trade-Offs, Forbidden Base Rates, and Heretical Counterfactuals,” Tetlock and his colleagues showed that some reasoning becomes morally contaminated because of what entertaining it appears to reveal about the thinker. Racial egalitarians were reluctant to use race-linked base rates and angry at people who did, and merely contemplating certain calculations produced outrage and moral cleansing. Tetlock asked what turns a base rate into a forbidden one. Pinsof can ask what turns the prohibition on and off. If the same statistical reasoning becomes offensive when it disadvantages an ally and mundane when it disadvantages a rival, the rule has shown its boundary. The operative principle is closer to “do not deploy damaging descriptions against groups I am motivated to defend” than to “never reason statistically about groups.”
That reading gives political correctness a new description. Alliance Theory would treat it as a system of asymmetric speech permissions. Every coalition keeps reputational rules about which descriptions of its members are acceptable, which insults are intolerable, which generalizations are dangerous, and which attacks on rivals can be redescribed as social analysis. Conservatives should keep such rules too, and the paper depends on showing it. A weak version documents liberal hypocrisy about stereotypes and joins the culture war. A strong version predicts mirror structures without mirror images. The favored groups and traits need not match, because the coalitions do not match; the logic should. The theory makes the question of whether liberals or conservatives are more biased boring. The interesting question is conditional: biased about whom, permissive toward what, protective under which alliance structure, and does the principle move when the favored target moves.
The third study therefore needs alliance reversals. Find groups whose relationship with American coalitions has changed. Use archival surveys, newspaper corpora, elite rhetoric, and repeated measures. The prediction is stronger than the trivial one that opinion changes. It is that the moral vocabulary governing group description changes as coalition membership changes. When a group becomes an ally, descriptions that were ordinary acquire the language of stigma and harmful stereotyping, negative generalizations require more qualification, individual variation becomes salient, structural explanations become attractive, and the group’s positive traits become more discussable. When a group becomes a rival the movement partly reverses: aggregate tendencies become relevant, cultural explanations become respectable, and a population can be called authoritarian, ignorant, conspiratorial, privileged, fundamentalist, or backward without the insistence that individuals must never be reduced to averages. The words vary; the asymmetry is the prediction. Researchers can preregister a set of alliance changes independently of stereotype language, build a dictionary of individualizing and generalizing rhetoric, and test whether the language applied to the group changes after the alliance moves. The historical study is weaker causally than the fictional-group experiment and stronger ecologically. Together they would be formidable.
A fourth study should turn to scientists, and this is where the claim in “A Big Misunderstanding” becomes testable. Do psychologists suppress stereotype accuracy because they fear appearing insufficiently progressive? Do not infer the answer from a publication record or from cancellation anecdotes. Randomize the evidence. Give social scientists matched abstracts describing the same methods and numbers; in one condition the disadvantaged population is Group A, in another Group B, with the identities chosen in advance from groups whose alliance positions have been measured. Ask researchers to rate methodological quality, importance, ethical concern, likelihood of recommending publication, anticipated harm, and willingness to collaborate with the author. The methods stay fixed and the political meaning moves. The strongest version recruits reviewers across ideological camps and preregisters the interaction between reviewer alliance and target alliance. A liberal researcher should grow more skeptical when a finding threatens an allied group, and a conservative researcher should show the same selectivity toward groups inside the conservative coalition. If the asymmetry appears only among progressives, Alliance Theory has found something narrower than a species-wide coalitional psychology and should say so. The same design runs on journalists and voters, and a headline version can ask participants to choose between “members of Group A differ on average” and “Group A is more X,” holding the statistics constant while the license moves.
There is reason to expect accuracy and distortion to coexist. Jonathan Haidt, Brian Nosek, and Jesse Graham’s “The Moral Stereotypes of Liberals and Conservatives” found that Americans perceived the direction of moral differences between liberals and conservatives correctly while exaggerating their size. A belief can carry information and propaganda at once. Alliance Theory should predict the stretching. If stereotypes are partly tools for navigating alliances, you need enough accuracy to identify allies, enemies, and likely behavior, and exact calibration may not serve your interests. Exaggerating an enemy’s faults mobilizes allies; underestimating an ally’s preserves solidarity; broadcasting an accurate but damaging trait of an ally is costly even when knowing it privately is useful. That fits Pinsof’s argument better than the bare claim that stereotypes are savvy. Savvy differs from unbiased. A useful political animal may need accurate private maps and distorted public propaganda, and natural selection rewards the organism that uses information well. Sometimes that means saying the true thing, sometimes staying quiet, sometimes exaggerating, and sometimes accusing the person who said the true thing of sinister motives. That is the evolutionary hypothesis buried in his essay.
In his 2018 UCLA dissertation, Pinsof studied stereotypes and political attitudes. The missing paper would reunite the early empirical Pinsof with the later theorist.
It would force precision about suppression. Scientists can neglect an accurate finding because they think it theoretically dull, because a criterion is bad, because the group categories are unstable, because population averages predict individuals poorly, because publication could do harm, or because careers follow fashion. All of those are compatible with a biased literature, and none of them is censorship. The Pinsof version should measure anticipated reputational sanction directly. Ask scholars whether a researcher would lose collaborators, invitations, jobs, friends, or standing for publishing each result, and compare the perceived danger across swapped targets. Jussim, Stevens, and Honeycutt pointed at this; Pinsof can supply the theory predicting where the sanctions fall.
One protection is socially non-negotiable. Group-level accuracy must stay separate from individual judgment. An accurate difference between two means says nothing about a particular member of either group; distributions overlap and individual information can dominate. Jussim’s work repeats this at every turn, and the paper should say it aggressively, because otherwise critics can dismiss the project as a defense of prejudice. The question here is not whether employers should discriminate or police should profile. It is what happens when a true proposition about a group becomes expensive to utter, and what sets the price. The price itself becomes the dependent variable: how offensive is the statement, how much would you dislike its author, should the person apologize, should a university investigate, should a newspaper publish it, should the professor be invited to speak, would you share the article, read it privately, or avoid clicking because being seen reading it might reveal something about you. At that point the study has become a theory of how intellectual taboos form.
That connects Pinsof to the argument about political homogeneity in social psychology. José Duarte, Jarret Crawford, Charlotta Stern, Jonathan Haidt, Lee Jussim, and Philip Tetlock argued in “Political Diversity Will Improve Social Psychological Science” that homogeneity shapes which questions get asked, which hypotheses look plausible, and which interpretations get scrutiny. Their argument was institutional. Alliance Theory could make it predictive. A homogeneous field should have a map: safe quadrants and dangerous ones, group differences that are easy to publish because they flatter allies or damage rivals and others that require throat-clearing, results attributed quickly to culture and others that trigger a search for structural causes, populations readily discussed as aggregates and others that trigger concern about essentialism, explanatory variables welcomed when they distribute blame one way and called reductionist when they distribute it the other. Those differences can be measured, and this is where Alliance Theory becomes a sociology of knowledge or stays a clever way to call intellectuals hypocrites.
Pinsof has made the accusation. In “Psychologizing Psychologists,” he says psychologists are reluctant to apply their theories to themselves; in “A Big Misunderstanding,” he says social scientists collect misunderstandings because the picture makes intellectuals useful. So psychologize the psychologists experimentally, and then psychologize the psychologist doing it, because Pinsof is inside the game. He is a UCLA-affiliated researcher, he says on his own page that he works at Cards Against Humanity, and he publishes provocative arguments about status and self-deception while moving through institutions with coalitions and incentives of their own. Alliance Theory predicts that he has alliances, groups he finds easier to generalize about, propositions he scrutinizes harder because of their politics, and heterodox positions that earn him standing in one audience because they cost him standing in another. It predicts that the role of brave truth-teller resisting progressive conformity can itself serve a coalition. He acknowledges the danger. In the comments under “A Big Misunderstanding,” when a reader observed that intellectuals might pursue status by promoting flattering theories, Pinsof replied that another route to status is saying true and insightful things, that he is trying to take it, and that whether he succeeds is uncertain. Build that uncertainty into the design.
The paper should preregister the result that would hurt the theory as well as the one Pinsof hopes for. If subjects keep the same rules for allies and rivals once evidence, harm, status, and history are controlled, the coalition account loses ground. If reviewers evaluate identical studies identically regardless of target, the suppression claim loses ground. If alliance measures predict little once stable moral commitments are in the model, the theory loses ground. If real alliance changes do not predict changes in license, the historical prediction fails. A paper that could wound the theory that generated it is the only kind worth the cost.
The most memorable result would be a three-way separation of truth, belief, and permission. Democrats and Republicans estimate the same group difference with equal accuracy; both know it exists; Democrats punish mentioning it when the target is a Democratic ally and Republicans punish mentioning an equivalent difference when the target is theirs. Nobody misunderstands anything. That result would tie “A Big Misunderstanding” to Alliance Theory more tightly than anything Pinsof has published. The misunderstanding model says bad stereotypes persist because people have bad information, so the corrective is better information. Pinsof’s model says people can have adequate information and regulate its expression because information has social consequences. Give them better data and nothing need change. They already knew. The fight is over what may be said, by whom, about whom, in front of whom.
The full project has five layers: whether the belief corresponds with reality, whether individuals hold it, whether they grant permission to state it, what happens when alliance is manipulated, and what penalties fall on the speaker. Accuracy, belief, license, alliance, punishment. Social psychology has studied each in some form. Pinsof has a theory that puts them in one causal sequence, and its deepest prediction is neither that stereotypes are accurate, which Jussim fought for, nor that people have double standards, which Pinsof has shown, but that stereotype taboos are generated by alliance structures and move when they move.
That would explain why arguments about stereotypes resist resolution by facts. The combatants answer different questions. One asks whether the difference is real. The other hears a question about whether people like us should permit people like them to say this about people we protect. A factual answer cannot settle a coalition dispute, because accuracy is one input into the social judgment, and the person who keeps presenting data looks obtuse to the enforcer while the enforcer looks terrified of facts to the empiricist. Both may understand what is happening. It also predicts where scientific controversies explode: where evidentiary claims cross coalition boundaries, so that better measurement can intensify conflict by removing the methodological grounds for dismissal, and the argument turns moral. Why are you studying this, why that group, who will misuse it, why do you care. Those questions can be legitimate. Alliance Theory predicts their frequency will be uneven, and that unevenness is measurable too. A final experiment could manipulate the author’s stated motive, describing the researcher as concerned with truth, as an activist against prejudice, or as a critic of progressive orthodoxy, and see whether evaluations of identical evidence move with the presumed coalition of the scientist. If they do, the scientist has become part of the stereotype and the theory has swallowed the laboratory.
Then the wager. There is no evidence that UCLA or Cards Against Humanity would sever ties with Pinsof over such a paper, and saying so would be speculation dressed as reporting. My bet is that a preregistered, methodologically careful paper demonstrating coalition-dependent permission to state accurate stereotypes would become the most discussed work he had produced, and that the social costs would rise sharply once the study moved from fictional groups and parties to ethnicity, religion, and other protected identities. Whether it would end either relationship is a prediction to be stated before the outcome rather than retrofitted after. The theory tells him what to record if trouble begins: who objects, to which conditions, to the method or the target, whether they would accept the design with the labels reversed, whether they attack the result, the question, or the scientist, whether they demand more evidence for costly claims than for convenient ones, and whether people who dislike the result discover a universal principle they never applied to its mirror image. The reaction becomes a further study, made comparable by randomizing, blinding, and swapping, since treating every criticism as confirmation would make the theory unfalsifiable and many criticisms could be right.
Pinsof’s best idea is that political morality becomes intelligible when you stop listening to the announced principle and look at the alliance underneath. Stereotype discourse is a severe test because the announced principles here are among the most charged in educated life: do not essentialize, do not generalize, treat people as individuals, follow the evidence, protect the vulnerable, do not suppress inconvenient findings. All of them sound universal. Alliance Theory says to check where each appears and where it disappears, whose interests it protects and whose it injures, and then switch the names. The accuracy literature has spent decades asking whether people’s beliefs about groups correspond with the groups. Pinsof’s missing paper should ask whether our principles about stereotyping correspond with our stated principles. I suspect the correlation will be lower.
Alliance Theory and the Jews
The third paper David Pinsof has not written is sitting on his desk.
Pinsof is part of the UCLA team working on “Tests of an Engineering Perspective of Hatred Towards Jews.” The project is led by Jaimie Arona Krems with Daniel Sznycer and Pinsof, and its framing tells you the register. Hatred is treated as a potentially functional emotion, an evolved program for managing relations with people or groups whose existence is perceived as costly, including because they are seen as taking important resources. The researchers propose to derive predictions from that model and test them with representative surveys, large datasets, and controlled experiments. Ignorance, pathology, moral failure, and the residue of ancient prejudice are all set aside as starting points.
Put that project beside Pinsof, David Sears, and Martie Haselton’s “Strange Bedfellows: The Alliance Theory of Political Belief Systems,” published in Psychological Inquiry, and the missing paper becomes obvious. Alliance Theory begins from the same refusal of sentiment. Political beliefs arise from alliances; people identify allies and rivals and recruit principles to defend one and attack the other; the principles can be sincerely felt and strategically deployed at once; and the principle predicts a person’s side less well than the identity of the ally does. The antisemitism project asks what makes Jews objects of hatred. Alliance Theory asks what makes any group an object of political hostility. The two programs ought to meet, and Strange Bedfellows already reaches toward the meeting twice. It cites Israel as the case that breaks intolerance theory, since conservatives, who are supposed to dislike foreign outgroups, view Israel as an ally. And it cites reparations as the case that reveals perpetrator bias, since support for compensating Holocaust survivors is bipartisan while support for compensating the descendants of slaves is partisan, which the authors read as evidence that perpetrator biases attach to local political alliances. Jews are already in the paper as the group that shows the theory working. The next step is to make them the test.
American Jews sit awkwardly on the coalition map. They are a small religious and ethnic minority with a long history of persecution, and they are unusually educated and prosperous. Pew’s Religious Landscape Study reports that about two-thirds of Jewish adults hold at least a bachelor’s degree and roughly half report household incomes of $100,000 or more, while about 69 percent of Jewish voters identified with or leaned toward the Democratic Party in Pew’s 2024 analysis. Even that description fractures. Orthodox Jews are the conspicuous counterexample: in Pew’s 2020 survey, 75 percent identified as Republicans or Republican leaners. A group that looks Democratic from a distance contains a religious subgroup aligned the other way.
That is the sort of group Alliance Theory should want, because it defeats the story in which disadvantaged minorities join the left and privileged groups join the right. Pinsof’s theory already rejects that story. It describes an alignment in which highly educated professionals joined the Democratic coalition while less-educated and small-town Whites moved Republican, and it entertains the possibility that lower-status members of an ethnic majority turn hostile toward minorities they perceive as competitors for resources, status, or elite patronage. The Jewish case makes the framework harder. A minority can be historically vulnerable, economically successful, culturally prominent, heavily represented among professionals, and allied with a coalition full of less affluent minorities at the same time. It can be an ally along one dimension and a rival along another, and the perceptions can change faster than the demographics.
That gives Alliance Theory an account of antisemitism that owes nothing to an antisemitic personality. Hostility toward Jews should depend partly on the perceived alliance position of Jews in the conflict at hand, and the form of the hostility should change with the coalition producing it.
There is evidence pointing that way. Eitan Hersh and Laura Royden’s “Antisemitic Attitudes Across the Ideological Spectrum” surveyed roughly 3,500 American adults and separated overt antisemitic beliefs from political double standards. Overt antisemitism was substantially more common on the right, especially among young adults on the far right. Their experiments produced a second result that looks designed for Alliance Theory. Very liberal respondents were more likely to say Jewish Americans should denounce Israel before joining social-justice activism than to impose the analogous requirement on Muslim Americans regarding Muslim countries. Very conservative respondents produced the opposite and larger double standard against Muslims. The prejudices were unequal. The moral burden moved with the target group, which is what Pinsof’s theory says political cognition does: a coalition invokes a rule against its rival and discovers qualifications when the rule reaches an ally.
Alliance Theory therefore offers a method for studying antisemitism. Instead of asking whether someone dislikes Jews, swap the groups while holding the conduct constant. Give subjects matched cases involving Jewish Americans and Muslim Americans, Jewish students and another minority, pro-Israel organizations and organizations representing another foreign-policy constituency. Randomize the group and keep the alleged behavior identical. Ask about collective responsibility, offensive speech, campus exclusion, political litmus tests, foreign loyalty, discrimination, protest tactics, institutional punishment, and whether hostile generalizations count as bigotry. Then manipulate the coalition cues. Describe the Jewish target as strongly Democratic, strongly Republican, or neutral; as Zionist, anti-Zionist, Orthodox, secular, wealthy, working class, activist, or apolitical. If the theory is doing work, some principled judgments should move when the coalition information moves and the conduct does not.
The campus politics after October 7, 2023 make that experiment timely. Graham Wright, Shahar Hecht, and Leonard Saxe surveyed 3,810 non-Jewish undergraduates at 60 American universities in spring 2024 for “Antisemitism, Israel, and Political Ideology on the American College Campus,” published in Politics, Groups, and Identities, and found two distinguishable patterns. Explicitly anti-Jewish attitudes were more common among students with far-right identities. A different cluster, intense hostility toward Israel combined with propositions that formal definitions and most Jewish students in the researchers’ work regarded as antisemitic, was associated with the left. That is more useful to Alliance Theory than a horseshoe. The theory needs the two forms of hostility to be neither morally identical nor equal in size; it needs to explain why the rhetoric differs. On one side “Jew” can remain the explicit hostile category. On the other, open hostility toward Jews stays taboo while “Zionist” or “Israel supporter” carries the antagonism. Where criticism of Israel ends and treatment of Jews as a rival coalition begins is an empirical boundary, and the theory offers a way to test it. Hold the act constant and vary the actor. Does excluding a student organization from a campus coalition look different when the organization is Jewish rather than Muslim, pro-Israel rather than pro-Palestinian, Republican rather than Democratic? Does an accusation of dual loyalty become more acceptable when the foreign state changes? Does demanding that an American answer for another country’s conduct seem legitimate when the American is Jewish and the country is Israel, and illegitimate when another group is substituted? Those are experimental conditions.
The same method can test the more combustible hypothesis, competition among minorities. Hersh and Royden separately found higher antisemitic attitudes among some young Black and Hispanic respondents than among comparable White respondents and considered several explanations. That finding establishes no competition among minorities for progressive patronage. It makes competition a hypothesis that Alliance Theory should be able to test. Prime scarcity. Tell one group of subjects that university positions, political attention, philanthropic resources, institutional protections, or public recognition are scarce and another that they are abundant. Vary whether Jews are described as receiving more or less of the resource. Measure hostility. The UCLA engineering model predicts that perceived cost should matter. Alliance Theory adds that the effect of cost should grow when the costly group is also presented as allied with one’s rivals. If neither manipulation matters, the theory loses ground. If both matter and interact, the two programs fit together.
Antisemitism is old enough to resist a coalitional explanation. Anti-Jewish stereotypes have survived monarchies, socialist states, Christian and Muslim societies, capitalist and communist economies, periods when Jews were poor and periods when some communities became affluent, periods of exclusion from elite institutions and periods of conspicuous presence inside them. A theory that reduced that history to contemporary Democratic and Republican alliances would explain too little. The plausible contribution is narrower and falsifiable. Alliance Theory might explain which pieces of the inherited repertoire become politically useful at a given moment, which accusations become respectable, which become forbidden, and which coalition discovers an exception to its usual rules. It also produces predictions across time. If Jews move in public perception from ally to rival, hostility should rise in the coalition they are leaving. If they move toward a coalition, that coalition should become readier to recognize antisemitism, defend Jewish grievances, and read ambiguous conduct charitably, with the reverse among its rivals. The vocabulary should lag the realignment, leaving an interval in which everyone insists his principles are unchanged while his applications change in view.
Israel gives the theory its second laboratory. Pinsof and his coauthors argue that Alliance Theory should travel across countries and periods, and Israel is attractive because its coalitions change often and its bargaining is visible. The Knesset held five elections between April 2019 and November 2022, and parties became indispensable allies, unwanted partners, opposition rivals, and governing partners within a short span. Instead of comparing timeless Israeli left and right personalities, the theory could follow the same constituencies across coalition changes. What happens to confidence in courts when one’s coalition controls the government, and when it loses power? What happens to beliefs about judicial supremacy, executive restraint, protest, civil disobedience, military service, religious exemptions, settlements, or subsidies when a party that benefits from a rule moves from rival to indispensable partner?
The judicial fight supplies material. In February 2023 the Israel Democracy Institute found that 87 percent of opposition-party voters supported letting the Supreme Court strike down laws incompatible with the Basic Laws, against 44 percent of governing-coalition voters, and coalition voters were themselves divided, with only 39 percent of Likud voters supporting an automatic coalition majority on the Judicial Selection Committee. The messiness helps. Alliance Theory predicts pressure toward coalition-serving belief, not obedience.
The ultra-Orthodox draft may be the best test, because the perceived cost of an alliance changed sharply after October 7 while the alliance held. Ultra-Orthodox parties stayed indispensable as the burden on soldiers and reservists became salient. The Israel Democracy Institute reported in 2025 that roughly three-quarters of Likud voters and 79 percent of Religious Zionism voters supported Haredi enlistment, despite their coalition’s dependence on the Haredi parties, and other IDI surveys found large majorities of non-Haredi Jewish Israelis favoring broader conscription. That is no simple vindication of the theory, and it is more useful than one. An ally can become too expensive. Pinsof’s framework treats interdependence as one of the bonds of alliance, so Israel lets him test what happens when a partner’s value inside parliament stays high while the social cost of accommodating him rises outside it. Does ideology bend far enough to keep the alliance, or does the alliance break? A theory of alliances owes a prediction about that boundary.
Israel also supplies a contrary finding Pinsof has acknowledged. Yoav Ganzach and Yaacov Schul’s “Partisan Ideological Attitudes: Liberals Are Tolerant; the Intelligent Are Intolerant”, in the Journal of Personality and Social Psychology, found conservatives more intolerant of ideological outgroups than liberals, especially among extremists, and Pinsof, Sears, and Haselton cite it as an exception to the literature they marshal for symmetric intolerance. A serious program wants that exception. The task is to ask when Alliance Theory predicts symmetry and when the surrounding structure produces asymmetry: whether some societies produce more balanced rival coalitions, whether status, institutional control, threat, or the number of available partners changes the expression of intolerance. Israel could be the comparison that forces the theory to state its boundary conditions, and Ganzach, who has told me he regards the theory as dull rather than wrong and its evidence as confined to the Trump years, is the critic best placed to say whether the Israeli case looks like Pinsof’s theory or like something else.
Meanwhile the American alliance around Israel is moving. In April 2026, Pew found that 80 percent of Democrats and Democratic-leaning independents viewed Israel unfavorably. Republicans remained more favorable overall and had developed a generational split, with 57 percent of Republicans under 50 viewing Israel unfavorably against much warmer views among older Republicans. That is evidence that the old alignment has become unstable, and the transition is the moment Alliance Theory should care about most. Once a new coalition settles, political actors construct histories in which their current position was always entailed by their principles. The valuable moment is before the settlement, when principle and alliance have not been reconciled. If Israel becomes less securely coded as a conservative ally, Republican rhetoric about Israel should change. If American Jews stay disproportionately Democratic while Israel becomes contested inside the Democratic coalition, attitudes toward Jews, Jewish Americans, Israelis, and Zionists should differentiate, and positions that traveled together should separate. That allows preregistration. Measure the strength of the Israel-Republican association in 2026. Predict which attitudes will follow it if it weakens. Measure whether respondents distinguish American Jews from Israel more sharply as the coalition shifts, and whether definitions of antisemitism move with them. Come back in two or four years and score it. That would be Alliance Theory doing something harder than explaining yesterday.
Pinsof grew up in Highland Park, Illinois, and was one of eight childhood friends who created Cards Against Humanity. Kickstarter’s history of the game describes Max Temkin and Eliot Weinstein meeting in first grade and the eventual group of Josh Dillon, Daniel Dranove, Eli Halpern, Ben Hantoot, David Munk, David Pinsof, Max Temkin, and Eliot Weinstein graduating from Highland Park High School in 2005. A Highland Park profile of the founders described all eight as politically progressive and Temkin as seeing the game as informed by progressive concerns about race, class, privilege, and cultural norms. Highland Park is a heavily Jewish suburb; after the 2022 parade shooting the Jewish Telegraphic Agency described it that way and quoted a local rabbi estimating that roughly half its population was Jewish. None of that establishes anything about Pinsof’s private identity. It establishes something more relevant to an intellectual biography: antisemitism is no remote topic dropped into his portfolio. The politics of Jews, liberalism, status, elite institutions, and progressive coalition membership is the social world he came from.
Alliance Theory is easy to enjoy when it exposes someone else’s double standards. Conservatives support free speech until the speaker insults the flag; liberals support unions until the union represents police; conservatives oppose handouts until the beneficiaries are rural; liberals oppose inequality until the rich people in the example are movie stars. The joke is that nobody has principles. The theory gets serious when the target is close enough that every classification hurts.
Are Jews an oppressed minority or an elite? Alliance Theory’s answer is that they can be either for purposes of political cognition, depending on the comparison, the salient resource, and the active conflict. A Jewish student excluded from an organization is a vulnerable minority. A Jewish donor at an elite university is a powerful member of the establishment. An Orthodox family is religiously conservative. A secular Jewish professor is part of the progressive professional class. An Israeli civilian is a victim of terror; an Israeli official is the representative of a powerful state. None of the descriptions is false. Coalition politics selects which one is foregrounded, and the propaganda battle runs through classification. Calling Jews a persecuted minority activates one moral architecture; calling them privileged Whites activates another. Calling Israel a refuge activates one set of analogies; calling it a Western-aligned power activates another. Calling an American Jew a minority citizen activates a rule against collective responsibility; calling the same person a Zionist may suspend it. The empirical question is why political actors select one classification at one moment, and Alliance Theory has the answer ready: select the description that helps allies and hurts rivals.
That is where John Mearsheimer and Stephen Walt enter, with a different argument. Their 2006 paper “The Israel Lobby and U.S. Foreign Policy” argued that a loose coalition of organizations and individuals sympathetic to Israel exerted unusual influence over American Middle East policy, and denied that the lobby meant American Jews collectively or any conspiracy. The backlash was extraordinary. Michael Massing’s “The Storm over the Israel Lobby” in The New York Review of Books documents charges of antisemitism, conspiracy thinking, distortion, intimidation, and scholarly irresponsibility from several directions, and Mearsheimer and Walt answered with “Setting the Record Straight.” They approached the problem from outside political psychology, asking whether an organized coalition influenced policy. Pinsof’s question is a level deeper. Why do people become intensely interested in Jewish power when Jews are perceived as rivals and in Jewish vulnerability when they are perceived as allies? Why does one coalition notice a lobbying organization and another notice discrimination? Why does one side read criticism of Israel as antisemitism while the other reads the accusation of antisemitism as a silencing tactic, and under what conditions do the readings reverse? Alliance Theory cannot answer by choosing a winner. It has to turn the double-standards machinery on everyone.
So a good experiment manipulates the perpetrator’s coalition as well as the target’s. Present identical anti-Jewish remarks attributed to a far-right activist, a progressive activist, a Muslim activist, a Christian nationalist, a Black activist, or an apolitical person. Ask whether the remark is antisemitic, whether the speaker should be punished, whether context mitigates, whether the target is reasonably offended, and whether the speaker’s movement bears responsibility. The theory predicts the judgments depend partly on who said it. Then reverse the target, using comparable remarks about Muslims, Christians, Whites, Black Americans, Hispanics, men, or women. If respondents rediscover context, individual responsibility, free expression, historical oppression, or the danger of stereotyping when the identities change, the theory has something measurable.
Political alliances help determine when inherited anti-Jewish ideas become useful, which versions become permissible, which speakers get charitable readings, which victims get institutional protection, and how consistently universal rules against prejudice are applied. That follows from Alliance Theory. The UCLA hatred project provides the target. Hersh and Royden provide evidence that anti-Jewish attitudes and double standards distribute unevenly across ideology. Wright, Hecht, and Saxe provide evidence that campus hostility takes politically distinct forms. Israel provides the comparative system. Ganzach and Schul provide the result that keeps the theory from assuming symmetry. Current polling provides an alignment that is moving. Highland Park provides the proximity.
A strong Pinsof paper would map when Jews are perceived as allies and when as rivals. It would distinguish Jews from Israel, Israel from Zionism, Orthodox from secular, wealthy from poor, partisan from neutral, manipulate those cues, identify the principles each coalition invokes, and swap the groups. Most of all it would predict the present realignment before the outcome is known. Mearsheimer and Walt wrote an adjacent version from outside and learned how combustible the ground was. Pinsof has the tools to approach it from inside political psychology with a more precise question: why particular coalitions become interested in particular descriptions of Jews at particular moments. That is the Jewish test of Alliance Theory. If the same principles survive the swaps, the theory loses one of its most provocative claims. If the principles move with the allies and rivals, antisemitism becomes one of the theory’s hardest tests, because the machinery has been turned on a conflict close enough to its author’s own world that there is no comfortable side from which to watch it run.
Alliance Theory Applied to Cards Against Humanity
What happens when Alliance Theory is applied to the coalition that pays David Pinsof?
Cards Against Humanity frees Pinsof from one institutional coalition while embedding him in another. The independence that lets him criticize academia rests on material interdependence with a company he built with seven friends from one high school. Alliance Theory predicts that interdependence strengthens allegiance, that shared allies and rivals strengthen it, and that long association and similarity strengthen it. Cards Against Humanity bundles all three.
The company also supplies laboratory-grade material. It has never presented itself as a manufacturer of dirty jokes on cardboard. It has attached the brand to causes, raising money for abortion access, funding a scholarship for women and nonbinary students in science, supporting LGBTQ health care, and running political promotions that encouraged voting and mocked Republican policy. None of that establishes hypocrisy. The donations are real, the causes may be good, and the people involved may believe every word. That is why the company is interesting under Alliance Theory, which does not require moral language to be fraudulent. Its claim is that moral principles are selectively activated by alliance structure. A person can care about fairness and notice unfairness more readily when an ally suffers it. A company can care about inclusion and use inclusion as a marker of coalition membership. A political gesture can accomplish something, advertise who belongs, and establish who does not, all at once. Sincerity and strategy are partners in Pinsof’s model.
So the company offers an intimate test. Which causes become Cards Against Humanity causes? Which injustices are funny enough, urgent enough, or obvious enough to deserve company money? Which groups are presumed to deserve protection and which are acceptable objects of ridicule? Which political violations are existential and which merely stupid? Which inequalities are obscene and which are the rewards of successful cultural production?
Pinsof has supplied one awkward clue. “Strange Bedfellows” cites Pulse of the Nation polling, from the survey he wrote the questions for, in which liberals were more offended by enormous pay for corporate executives than by enormous pay for Hollywood movie stars. Because movie stars belong to the liberal coalition, the authors argue, the principle about economic inequality changes with the identity of the rich person. That example demands an investigation of the entertainment industry. Instead the entertainment coalition stays scenery.
He gets close on occasion. In “Charisma Is Bullshit” he observes that people make supposedly subversive art for high-status audiences and rebel against conformity in conformist ways. In “A Big Misunderstanding” he pushes hard against his academic world: social science takes people’s official explanations too literally, institutions proclaim goals while human beings pursue status, coalition, reputation, and control underneath them, and academics may be climbing hierarchies and dominating rivals under moralistic language. The unbound version of Alliance Theory would do the same to comedy and entertainment.
Hollywood is built for the theory. It offers enormous rewards and few objective measures of desert. Jobs move through relationships, recommendations, agents, executives, writers, producers, festivals, fellowships, critics, and social circles. Status is intensely desired and naked status seeking is discrediting, so success has to be narrated in another vocabulary. Talent is one. Authenticity is another. Representation, harm, safety, courage, transgression, inclusion, and punching up or down are others. Alliance Theory does not say in advance that those concepts are bogus. It says to perform substitutions. Take the same joke and change the comedian. Take the same hiring preference and change the group it favors. Take the same offensive statement and move it from an ally to a rival. Take the same paycheck and give it to a performer, a technology executive, an oil executive, or a conservative broadcaster. Take the same refusal to hire someone whose politics colleagues find objectionable and move the target from right to left. Then measure whether the principle survives. That is the method Pinsof has used elsewhere. The comedy industry needs no new theory. It needs the existing theory without geographic restrictions.
The Kilroys make the omission sharper. Pinsof’s wife, Annah Feinberg, was a founding member of the Kilroys, a theater collective created to address the underrepresentation of women, trans, and nonbinary playwrights. The organization is candid about its method. It describes itself as a collective that will mobilize others in the field and leverage its own power to support its members, and its best-known intervention, The List, uses nominations from theater professionals to identify plays by writers from underrepresented genders and raise their visibility with institutions able to produce them.
A coalition identifies an institutional outcome it regards as unjust, organizes people who already hold influence, creates common knowledge about deserving members of an underrepresented category, coordinates attention, changes the information available to gatekeepers, and tries to alter the distribution of opportunities by altering which candidates become salient. That is a translation of the Kilroys’ own account into Pinsof’s vocabulary, and it is where Alliance Theory becomes more interesting than the usual argument about diversity programs. The usual argument asks whether the Kilroys are right. Alliance Theory asks what happens to merit, representation, discrimination, and fairness when the identities of the beneficiaries change. The question is testable. Remove authors’ names and demographic information from plays and obtain evaluations; restore them and repeat with another sample. Map who nominates whom, and measure whether professional relationships predict nominations after controlling for blind assessment of the work. Compare reactions to a collective built to advance women and nonbinary playwrights with reactions to an identically structured collective built around a culturally rival group. Track whether the defended principle holds when the alliance changes. The result might vindicate the Kilroys; neglected writers may have been producing stronger work, and the coalition may have corrected an information failure. Alliance Theory requires the question, not the answer.
The personal connection raises the value of the experiment. On Pinsof’s theory a spouse is close to the limiting case of interdependence. Careers overlap with friendships, friendships with reputation, reputation with access to communities from which both spouses draw professional and social benefit. An attack on one person’s coalition imposes costs on the other with nobody issuing a threat. That yields a subtler prediction than the crude one that Pinsof is afraid to criticize his wife. The theory predicts that no prohibition is needed. Questions that threaten one’s coalition feel worse. They seem less interesting, less urgent, less fair, less publishable, less funny, or needlessly cruel. The scholar does not experience himself choosing loyalty over truth. He experiences himself exercising judgment about what deserves attention. That innocence is what the theory predicts: allies imitate one another, share preferences, avoid betrayal, and acquire common readings of a conflict because those tendencies make alliances work.
“Money Is Bullshit” is where Pinsof comes closest to turning the machinery on his own class. The essay argues that money is entangled with status and with the meaning of impersonal exchange. He notices that progressive elites compete for status while developing norms that make talk about money distasteful. He repeats the executive-versus-movie-star inconsistency. He ties beliefs about who deserves money to the desire to raise allies and lower rivals. Then something happens. When he needs examples of relationships that escape cold exchange, he reaches for the wife, the friend, and the tribesman, whose interests become intertwined with his own, and marriage becomes evidence that human relationships cannot be understood through price alone. The argument may be right. Structurally it is telling. The status analysis reaches the perimeter of Pinsof’s world and the object of analysis changes. Hollywood wealth stays a clue; progressive status games stay a clue; his most intimate alliances become the solution. The essay becomes a defense of warm, durable, nonmarket relationships, with marriage among its examples. It does not become an ethnography of how Pinsof’s own durable relationships distribute money, prestige, opportunity, moral approval, and reputational protection. That is the unwritten paper.
It has a ready-made episode. In 2020 former employees made public allegations about the company’s workplace culture. The company acknowledged failures, announced changes, and said Max Temkin had stepped down. Pinsof has said nothing about it in public. There is no need to prejudge the dispute; Alliance Theory supplies better questions than another prosecution or defense. Who believed whom? Which kinds of evidence changed minds? How did friendships of twenty years compete with newer institutional loyalties? When did criticism count as betrayal and when as reform? How were responsibility and mitigation allocated? Which reputations were defended and which harms amplified? Which outsiders were recruited? How did the company’s public moral identity shape the vocabulary available for explaining an internal conflict? Those are operationalizations of perpetrator bias, victim bias, attributional bias, transitivity, and alliance maintenance, the concepts of his own paper.
Pinsof could study such a world better than almost anyone because he has what social psychologists usually lack, which is standing. He knows the jokes before publication, the arguments about taste, the playtesting, the commercial pressure, and the relationships among people who built something together. He works with the company’s data. He has spent years on status and alliance in theory. Through his household he has proximity to theater, television, and comedy networks. The disadvantage is the same fact. An insider has more information and less freedom, and that tension is the theory’s best possible demonstration.
The paper insists that alliance psychology is symmetric across humanity. It rejects the idea that one faction distorts reality while another sees clearly. It applies the same account to conservatives and liberals, and it turns briefly on social psychologists, citing evidence that liberal academics will discriminate against ideological rivals. The logical destination is recursion, since nobody is neutral in this account. Study the scientist studying alliances. Map his alliances. Identify the people whose welfare is interdependent with his own, the organizations that fund his independence, the moral vocabulary those organizations favor, and their allies and rivals. Run the substitutions. See which principles survive.
The ownership of Cards Against Humanity in 2026 is not visible enough in the public record to say that Pinsof has seven co-owners today. What is visible is more interesting. He created the company with seven other men, remains attached to it, and says its income makes his career possible. The patron and the peer group are the same institution.
That makes the silence analytically valuable and evidentially worthless. His failure to write this paper proves nothing about coalition protection; treating every omission as hidden allegiance would be the cheapest use of the theory and would make it unfalsifiable. Writing it would be the test. Alliance Theory has shown it can explain why Republicans discover principles that help Republicans and Democrats discover principles that help Democrats. The harder question is whether the machinery keeps running when the groups in the diagram are Cards Against Humanity founders, progressive comedians, playwrights, entertainment executives, spouses, and friends.
Pinsof’s theory says people are worst at recognizing political strategy when it arrives wearing the language of morality, and his best work applies that insight where other scholars would rather not look. The next step is to look where he would rather not. A theory of alliances becomes more convincing when it survives contact with the alliance that pays its author.
Applying Alliance Theory to Everything Is Bullshit Readers
The fifth extension of Alliance Theory examines the Everything Is Bullshit constituency: heterodox, heavily male, rationalist-adjacent, Darwinian, skeptical of progressive social science, amused by moral debunking, and comfortable with explanations that reduce lofty principles to incentives, status, and coalition management.
What are the badges of Pinsof’s own coalition?
He has supplied most of the machinery needed to answer it. In “Opinions Are Bullshit” he argues that opinions are instruments in a status competition, that people use arguments to convert preferences into norms, and that people with platforms hold an advantage because they can broadcast their judgments. In “Everything Is Signaling” he goes further. Humans continuously model how observers will judge them, and much signaling is defensive: we try to avoid looking stupid, immoral, low-status, or disloyal before the people whose judgments count. For a writer with a defined readership the implication is immediate. His audience is one of the social environments the theory is supposed to explain.
Pinsof knows this. In “Bullshit Is a Choice” he asks whether his own desire to write interesting things is bullshit and answers yes. He adds that he does not consciously select beliefs for their social rewards, and the distinction does little for him inside his own framework. Alliance Theory is a theory of minds that generate sincere arguments while remaining sensitive to allies, rivals, prestige, and incentive. A man can believe every word he writes and still be producing the beliefs that prosper in his niche.
The revealing episode is AI. In “AI Doomerism Is Bullshit” Pinsof attacked a movement he had spent years admiring. He calls rationalists and effective altruists brilliant, truth-obsessed people, then argues that AI doomerism has become a self-aggrandizing myth and a shibboleth for nerds. Whether he was right about artificial intelligence matters less here than what happened when he aimed his apparatus at people who thought of themselves as members of the anti-bullshit coalition. The response was fiercer than anything drawn by his attacks on conventional morality, happiness, politics, or progressive social science, and he said so. In “Another Year of Bullshit” he marveled that readers could calmly entertain the proposition that morality is a tool of social domination and erupt when told that robots were not going to kill everyone. His explanation was a miniature Alliance Theory of his own newsletter: everybody believes everything is bullshit except his own thing. That sentence deserves more weight than he gives it. It names the experiment. Every intellectual subculture has the proposition it treats as proof that its members are the kind of people who can handle the truth.
The rationalists were one wing of the readership. The harder target is the larger heterodox coalition around them, whose sacred vocabulary differs from progressive academia’s and is sacred nonetheless. Free inquiry. Viewpoint diversity. Intellectual courage. Open debate. Following the evidence wherever it leads. Saying what everyone else is afraid to say. Opposition to censorship. Resistance to conformity. All of those may name real goods, and Alliance Theory does not require them to be fake. It requires asking why particular people invoke particular principles in particular conflicts, why some violations become defining scandals while others vanish, and why universal principles line up so reliably with coalition boundaries.
That is the operation Pinsof performs on his political opponents. Progressive invocations of equality, compassion, inclusion, and harm reduction become suspicious when the same people drop them whenever they would help a rival coalition; Alliance Theory reads those reversals as evidence that principles are recruited to help friends and attack enemies. The fifth extension applies the same test to the heterodox world. When does free speech disappear because the speaker is regarded as contemptible? When does viewpoint diversity lose its appeal because the excluded viewpoint threatens the coalition? When does suspicion of institutional power become enthusiasm for it because the institution has changed hands? When does “follow the science” become admirable because the science favors one’s side?
The claim would stop short of calling free inquiry worthless; Pinsof’s better work rarely depends on that inference. Effective altruism can be bullshit and still do good. Signaling can be useful. Moral motives can conceal status motives without moral behavior losing its meaning. The proposition is subtler. Free inquiry can be a real norm and a coalition instrument at once. A person can favor intellectual freedom while becoming attentive to violations by his enemies, forgiving of violations by his friends, and eloquent about the principle when eloquence raises his standing with his audience.
The heterodox coalition is vulnerable to the analysis because its status system rests on the denial that it has one. Progressive academics acquire prestige by displaying sensitivity, solidarity, and correct moral concern. Heterodox intellectuals acquire prestige by displaying independence from those things. One side shows it knows what respectable people must never say. The other shows it is brave enough to say it. Both performances identify trustworthy allies, both build hierarchies, and both reward fluency in a local moral language. The rebel can be as conformist as the commissar so long as he conforms to a community that pays for looking rebellious.
This is where “A Big Misunderstanding” stops one step short of its best conclusion. Pinsof attacks intellectuals who believe the world is broken because ordinary people misunderstand it, and shows that the belief makes intellectuals indispensable: if misunderstanding causes social problems, the people who understand things become the repair technicians. A profession’s theory of society happens to make the profession necessary. The same move works against the dissident. Suppose the central pathology of modern society is ideological conformity; the brave heretic becomes important. Suppose universities are captured by dogma; the scholar who defies his colleagues becomes heroic. Suppose censorship is the central threat to civilization; the person who says forbidden things acquires moral stature. The progressive intellectual tells a flattering story in which society needs enlightened experts. The heterodox intellectual tells a flattering story in which society needs heterodox intellectuals.
Pinsof is well placed to write this because he cannot exempt himself. He runs a platform devoted to puncturing flattering stories and his readers reward him for the puncturing, so cynicism becomes a status-bearing product. Knowing that morality is status competition can confer moral status. Ridiculing virtue signaling can signal sophistication. Announcing that everyone is tribal identifies you with the tribe that has noticed tribalism. The phrase “everything is bullshit” is itself a badge, separating people who can handle Darwinian truths from sentimentalists who still believe the press release. The progressive earns status by showing that he cares. The cynic earns it by showing why the progressive cares. The next cynic earns it by showing that the first cynic’s exposure was itself status seeking. Pinsof understands that recursion better than most writers, and his unfinished task is to keep it running past the level occupied by David Pinsof and his readers. He has one entry on the ledger already. The AI essay cost him a wing of his audience.
The extension would also develop his argument with John Jost. The dispute goes back to Mark Hoffarth and Jost’s “When Ideology Contradicts Self-Interest” and Pinsof and Martie Haselton’s “When Self-Interest Contradicts Ideology: A Reply to Hoffarth and Jost”. Jost’s program emphasizes psychological differences between left and right, including distinctive motivational roots for conservatism, and his 2024 attack on both-sideology, which never names Pinsof, aims at the posture Alliance Theory adopts: liberals and conservatives recruit principles around allies and enemies because coalition psychology is human psychology. The trouble is that symmetry argued mostly with examples from the progressive coalition looks like asymmetric politics in symmetric dress. The strongest answer to Jost would be an empirical demonstration using the people most likely to applaud Pinsof’s existing conclusions. Take a population of self-described free-speech advocates, heterodox academics, rationalists, and anti-woke intellectuals. Measure support for speech protection while varying only the coalition identity of the speaker. Measure enthusiasm for academic freedom when a progressive professor is punished by a conservative institution and when a conservative professor is punished by a progressive one. Measure confidence in institutional neutrality when one’s coalition controls the institution and when its rival does. Measure tolerance of censorship when the prohibited speech attacks one’s allies. Then compare the size of the double standards.
If the effects appear, Alliance Theory wins something better than another victory over progressives: it shows the machinery running in people who built an identity around resisting it. If the effects do not appear, Pinsof has found a boundary condition for his theory. Either result is useful. One is comfortable.
That is why this extension is harder than attacking the academy. The academics who dislike “A Big Misunderstanding” do not sustain Everything Is Bullshit. Rationalists were closer, which is why “AI Doomerism Is Bullshit” drew the response it did. The rest of the heterodox readership is closer still. A paper showing that their cherished norms behave like coalition instruments would challenge the audience’s account of what kind of people they are, which is a heavier thing than a belief.
That is why Pinsof should write it. Alliance Theory becomes dangerous only when it predicts the bullshit produced by the people who love Alliance Theory. Otherwise it becomes one more flattering theory of error in which the theorist’s enemies have motivated minds and the theorist and his admirers have clear ones. Pinsof has spent years arguing that nobody gets that exemption. The fifth extension is where he would show that he means it.
Does Alliance Theory Predict Realignment?
The sixth paper examines realignment. When does an alliance of strange bedfellows stop? Which member leaves first? Does the loss of a common rival weaken a coalition at once or after a generation? How much ideological distance can be tolerated for electoral strength, institutional access, protection, money, prestige, or safety in numbers? What happens when the leader who served as focal point disappears? How fast do people discover new principles once the coalition that made the old ones useful is gone? Those questions are the moving half of Alliance Theory, and the theory has answered none of them.
The encouraging part is that Pinsof is moving toward them. UCLA describes his research as including “mathematical models of alliance formation,” the UCLA Library hosted his 2025 presentation through the Jacob Marschak Interdisciplinary Colloquium on Mathematics in the Behavioral Sciences, and the 2025 Human Behavior and Evolution Society program lists his talk “Whose Side Are You On? The Evolutionary Game Theory of Alliances.” His starting point was the simple majority game, an n-player game in which actors coordinate to form an alliance of insiders who receive benefits while outsiders bear costs, and his stated goal was to find its evolutionarily stable strategies and use them to guide psychological theory. He has begun formalizing alliance formation. The next problem is alliance death.
The simple majority game contains the seeds of both. Peter DeScioli and Robert Kurzban’s alliance hypothesis for human friendship uses the same logic. Alliance problems differ from ordinary reciprocity because helping one party to a conflict can require harming another, and in a three-player majority game any two players can exclude the third. Alliances are relational. Your value to me depends on what you can give me and on whom else you might join. The missing model takes that logic through time.
Imagine every politically relevant group as a node in a network. Evangelicals, union members, academics, rural voters, technology executives, environmentalists, police, ethnic groups, religious groups, professional associations, media institutions, activists, and parties are players with weighted relationships. A positive edge is alliance, a negative edge rivalry, and the weight is what it would cost for the relationship to disappear. That resembles the signed-network tradition of structural balance theory, in which friendship and enmity settle toward configurations where the friend of my friend is my friend and the enemy of my enemy is my friend. Pinsof already puts transitivity at the center of Alliance Theory. A formal version turns the verbal principle into a network that moves.
There is precedent inside psychology. Kurt Gray, David Rand, Eyal Ert, Kevin Lewis, Steve Hershman, and Michael Norton showed in “The Emergence of ‘Us and Them’ in 80 Lines of Code” that robust groups emerge in agent-based simulations from almost nothing. Their agents began with no races, religions, or ideologies; reciprocity and transitivity were enough to produce groups, cooperation inside them, and defection against outsiders. Pinsof cites that literature. The next step is to run the film backward and ask what parameter changes make us and them disappear.
The simplest model needs few ingredients. Each group draws benefits from an alliance, protection, access to a winning coalition, institutional influence, status, resources, information, the power to exclude opponents, and pays costs, policy concessions, reputational contamination, economic loss, ideological discomfort, conflict with other allies, and the opportunities forgone by not joining someone else. The coalition survives while the expected value of staying exceeds the expected value of leaving. That sounds trivial until every term has to be measured. How valuable is the common enemy? How expensive is the disagreement? How replaceable is a partner? How good is the defector’s outside option? How much does history raise switching costs, and how much stability comes from twenty years of the same allies? How much coordination does a leader supply, and how fast does his value decay when he leaves? Once those quantities exist, Alliance Theory acquires what it lacks, a hazard rate.
Political science has a large literature on that problem. Michael Laver’s review “Government Termination” separates the observed duration of governments from their theoretical durability and describes the event-history and survival models used to estimate the risk that a government collapses at a given moment. The quantity of interest is the conditional probability of dissolution given survival to this point, since everything dies. Alliance Theory needs the equivalent. The outcome is the probability that an alliance edge disappears in the next month, year, cycle, or generation. The predictors are rival strength, ideological distance, resource interdependence, leader centrality, coalition size, alternative partners, recent betrayals, electoral usefulness, status competition, demographic change, and the number of other relationships that depend on the edge holding. At that point “ever-shifting alliances” is a quantitative proposition.
The common enemy is the place to start. Economists have documented the common-enemy effect, an external opponent raising cooperation among actors with little other reason to cooperate, and the literature is messy: external threat does not always produce unity. The mess is the point. The formal question is under what parameter values enemies unite people. Alliance Theory says transitivity matters, so the strength of A’s alliance with B should depend partly on their shared hostility toward C, and reducing the threat from C should weaken the A-B edge. How fast? Some alliances should dissolve almost as soon as the enemy goes. Others should last decades because they have generated institutions, friendships, marriages, donor networks, media habits, legal structures, professional identities, and moral narratives with independent value. Instrumental interdependence becomes path dependence. That yields a concept the theory needs, coalition half-life: after the original rival disappears, how long until half the alliance strength attributable to that rival has decayed, and does the answer differ for a single-issue alliance and one spread across dozens of institutions?
Leadership is a second parameter. Coalitions often contain groups that disagree with each other more than with outsiders, and a leader holds them together because everyone knows everyone else coordinates through him. He is a Schelling point. Remove him and the coalition need not dissolve; the identities and institutions remain. But uncertainty about everyone else’s future allegiance rises, and transitivity becomes recursively unstable. A supports B partly because B supports the leader; C supports A partly because A does. If the leader goes, nobody needs to change a substantive view for the network to become fragile; each actor only needs to become unsure of the others. The model can therefore separate a leader’s ideological role from his coordinating role. If he matters because followers share his positions, a similar successor preserves the coalition. If he matters as the point through which incompatible groups coordinate, similarity in the successor will not be enough. That is testable, and so is the order in which principles change.
That may be the most valuable prediction. Alliance Theory currently shows reversals after they happen. If principles are partly alliance-serving tools, they should not be equally vulnerable at every moment; some should lead coalition change. The model could assign each principle an elasticity, measured experimentally by presenting identical scenarios with the group labels randomized. Free speech, equality, institutional deference, military intervention, immigration, fiscal restraint, punishment, bodily autonomy, local control, skepticism of authority, and concern about discrimination would each get a coefficient. Then watch coalitions change. The most coalition-dependent principles should move first. A party does not reverse its worldview overnight. It begins at the edges: a rule gets an exception, the exception gets an explanation, the explanation becomes a principle, and the original principle is remembered as something the other side believed.
David Karol’s Party Position Change in American Politics supplies the adjacent account: parties are coalitions of groups whose demands politicians manage, and positions change when members change their preferences, new constituencies enter, or politicians maneuver on issues no group yet owns. Kathleen Bawn, Martin Cohen, Karol, Seth Masket, Hans Noel, and John Zaller argue in “A Theory of Political Parties: Groups, Policy Demands and Nominations in American Politics” that parties are coalitions of intense policy demanders. Alliance Theory could add psychology to that political science and motion to its own psychology.
It would also settle, or at least specify, Pinsof’s disagreement with Yoav Ganzach and Yaacov Schul. Their “Partisan Ideological Attitudes: Liberals Are Tolerant; the Intelligent Are Intolerant” found conservatives more intolerant of ideological outgroups than liberals, especially among extremists, and Pinsof, Sears, and Haselton cite it as an exception to the literature they use for symmetric hostility. The verbal argument forces a binary: either political psychology is symmetric or the two sides have different dispositions. A formal model offers a third possibility. The cognitive machinery can be symmetric while the equilibrium outcomes are asymmetric. Give identical agents different networks and they behave differently. One coalition faces a more threatening rival, contains more dissimilar groups, controls institutions that lower the cost of punishing opponents, has a focal leader, depends on extremists rather than moderates, or faces coordinated rather than fragmented opposition. None of that requires different evolved psychology. So the formal way to handle the Ganzach problem is to specify identical alliance psychology for every agent and ask whether the observed asymmetries emerge from coalition structure. If they do, Pinsof has an explanation for Ganzach’s finding. If they do not, the theory needs an individual-difference parameter. Either result advances it, and that is what formalization buys. A verbal theory protects itself by explaining every result. A formal theory has parameter ranges in which it loses.
The 2025 game points this way. The simple majority game asks which strategies are stable. The dissolution paper would ask what shocks move a population from one stable arrangement to another, which connects Alliance Theory to William Riker’s The Theory of Political Coalitions, the founding attempt to force coalition politics into a game. Riker’s size principle asked why rational actors form coalitions larger than needed to win when victory then has to be shared. Pinsof’s problem is richer because the game is repeated, players are uncertain, groups differ in strength, today’s surplus partner is tomorrow’s indispensable one, membership changes reputations, and the same coalitions operate in elections, universities, businesses, media, neighborhoods, marriages, churches, and status hierarchies. Riker’s achievement was to take what politicians said in rhetoric and force it into a game. That is what Alliance Theory now needs.
The Trump coalition is the visible American test, and it requires no position on Trump, only the observation that the coalition has changed. Pew’s analysis of validated 2024 voters found that the 2024 Trump coalition was more racially and ethnically diverse than the 2016 or 2020 coalitions: White voters without college degrees fell from 63 percent of Trump voters in 2016 to 51 percent in 2024, while Hispanic, Black, Asian, and other non-White voters rose from 11 percent to about 20 percent. That is an alliance structure changing in real time, and the disagreement over its future already has two named parties. Ganzach has told me the Trump alliances rest on short-term interest organized around one man and are doomed to disintegrate. Alliance Theory says a coalition lasts as long as its rival does and rewrites its principles for the next leader. A serious project would freeze the data now, map the constituent groups, measure their affinity for one another, estimate how much of each relationship is explained by the leader, by shared hostility, by policy agreement, religion, economic interest, cultural identity, or institutional exclusion, estimate switching costs, identify alternative alliances, and preregister what should happen after a specified shock. The target is the coalition. Twelve months after the focal leader leaves active politics, which pairwise alliances weaken, which strengthen, which voter groups decorrelate on issues, which bundled beliefs separate first, and at what horizon does continued cohesion count against the model? That is what a forecast with a date means. Anyone can predict eventually. The paper should say that under stated observable conditions a named alliance edge will cross a stated threshold within a stated number of months, and then the clock starts.
Israel makes the same machinery close to a laboratory, because coalition bargaining is explicit and governments depend on party combinations renegotiated continuously among religious, nationalist, centrist, Arab, and secular constituencies whose alliances fit no single axis. The judicial and conscription fights are useful because coalition membership and preference can be observed separately. In a March 2026 Israel Democracy Institute survey, coalition voters were far more supportive of advancing the judicial overhaul than opposition voters, and support varied widely inside the coalition. On the Haredi draft, IDI polling in 2025 found large majorities of non-Haredi Jews, including majorities of Likud and Religious Zionist voters, supporting sanctions on Haredi men who avoid service. That is the cross-pressure the model needs. If coalition membership determined principles completely, voters would move to whatever position keeps the coalition. If material and identity costs overwhelmed coalition value, they would resist. If both forces operate, there is a threshold where loyalty stops buying accommodation, and Israel lets Pinsof observe it repeatedly. The model can compute the expected durability of an alliance between parties whose voters disagree sharply on conscription but cooperate profitably in parliament and compare it with an alliance between parties that agree substantively but compete for the same voters. The strange-bedfellows thesis becomes a comparison of costs, and Israeli turnover supplies out-of-sample tests: fit on one period, freeze the parameters, predict the next, with no retrospective reinterpretation.
The academy is a third test and perhaps the cleanest, since elections do not muddy the outcome. José Duarte, Jarret Crawford, Charlotta Stern, Jonathan Haidt, Lee Jussim, and Philip Tetlock argued in “Political Diversity Will Improve Social Psychological Science” that homogeneity shapes research questions, methods, and the treatment of uncomfortable topics; Alice Eagly answered in “Mischaracterizing Social Psychology to Support the Laudable Goal of Increasing Its Political Diversity” that the diagnosis mischaracterized areas including stereotype accuracy. That dispute runs on argument. Alliance Theory could run it as a time series. Take a sensitive domain, group differences, stereotype accuracy, sex differences, discrimination, immigration, policing, intelligence. Measure the institutional coalition around it: who funds the research, who edits the journals, which associations issue statements, which populations academics code as allies and rivals, which findings pay and which cost. Then observe changes in that network. Jussim, Crawford, and Rachel Rubinstein’s “Stereotype (In)Accuracy in Perceptions of Groups and Individuals” argues that accuracy is a large, replicable finding the field underemphasized; one need not accept every part of that to see the test. If academic positions partly follow alliances, a shift in the coalition should precede a shift in what is publishable, sayable, grantable, or safe. The prediction concerns the institutional boundary around acceptable hypotheses, not the truth. The boundary is measurable: count papers, code abstracts, track conference programs, citations, terminology, and grant solicitations, and survey scholars anonymously each year about which hypotheses they believe, which they would investigate, and which they would state in public. Then see whether the boundary moves where the model said. This would be a rare theory of academic bias that exposes itself to the danger it attributes to others. If the line does not move when the alliance structure changes, the theory loses. If it moves before the predicted coalition change, the arrow may run the other way. If it moves only after new evidence, ordinary updating explains it without coalition psychology. A model that can be wrong in three ways is a good model.
The deepest variable may be stored coalition capital. Alliances are not recreated each morning. Years of cooperation produce trust, networks, institutions, rituals, myths, donor lists, marriages, friendships, media habits, and reputational markers, and those assets make yesterday’s alliance cheaper to continue than a superior alternative. So an Alliance Theory of realignment cannot be static game theory; it needs hysteresis. The coalition a society selects moving from state A to state B need not be the one it selects moving back. That is how strange bedfellows become ordinary ones. Two groups ally against a rival; twenty years later the rival is gone but their children share schools, their intellectuals share magazines, their politicians share donors, their members marry, and their moral vocabularies have converged. Strategic transitivity has become similarity and interdependence. The paper already says its three criteria reinforce one another, so the formal model should produce lock-in, and lock-in yields another risky prediction: coalition age should matter after current interests are controlled, and old alliances should need larger shocks than young ones.
The model would also explain why realignments look slow and then sudden. An alliance’s underlying value can decline continuously while switching costs hold everyone in place, so nothing visible happens until one group crosses its exit threshold. Its departure changes everyone else’s outside options; a second group moves; a third move becomes profitable; the network cascades. For years nothing happens and then everything does. A static theory is worst at explaining that and a network model is best, and the order of ideological revision should follow the cascade: first the positions attached to the departing group, then those involving its allies and rivals, then the broader principles rebuilt to make the new arrangement look coherent. What journalists later call an ideological transformation is a cascade with rhetoric trailing behind.
This is where the comparison with Vilfredo Pareto earns its place. His distinction between residues and derivations treated explicit political reasoning as justification layered over deeper motive, and his theory of the circulation of elites tried to explain why ruling combinations form, decay, and get replaced. His weakness was that it stayed historical interpretation. He could look across centuries and tell you what the lions and foxes had been doing; he could not give you a stopwatch. Pinsof can. Survey panels, voter files, text embeddings, social networks, experiments, agent-based models, game theory, longitudinal polling, event-history analysis, and preregistration make it possible to turn a Paretian theory of rationalization into a model that forecasts its own failure date.
That is the standard. The mature Alliance Theory would specify the state variables that govern coalition stability, the shock needed to break an alliance, the delay between structural change and ideological change, the belief that changes first, and the differences among leader-dependent and rival-dependent coalitions, interest-based and identity-based ones, young and institutionalized ones. Then it would publish the predictions before the realignment. The Trump coalition is one test, Israel another, the academy a third. Historical party switches serve for estimation; validation has to come afterward. At that point the theory stops being a powerful way to reread political history and becomes a machine that can be wrong about the future.
That is why this paper is less dangerous than the others and worth more. Pinsof would accuse no one of hypocrisy, publish no forbidden stereotype, expose no industry, and tell no reader that his principles are coalition instruments. He would write down players, payoffs, ties, switching costs, threat levels, and transition rules. A model is not exempt from politics, and it is hard to denounce for taking the wrong side when its central claim is a conditional prediction anyone can test. He is halfway there. “Whose Side Are You On?” supplies the game, Alliance Theory the psychology, the coalition-survival literature the hazard function, network theory the cascades, political science the cases, Israel the repeated shocks, American partisan change the live longitudinal case, and the academy an institutional case whose outcomes are publications. What is missing is time. Add time to Alliance Theory and everything becomes harder, and science in a stronger sense.
The audacious version would end with a table of named relationships, measurable thresholds, specified shocks, confidence intervals, and expiration dates. If the relationships survive, the model was wrong. If the principles fail to move, the model was wrong. If the predicted defectors stay and the supposedly stable partners leave, the model was wrong. That paper would cost Pinsof less socially than almost anything else he could write, and it would cost the theory more, which is why it is the valuable one.
