For most of my career, political psychology kept what seems, in retrospect, like a running gag going. The gag was that the left-wing authoritarian, a creature everyone had heard of and nobody could photograph, did not exist.
A few people swore they’d seen it in the wild. The sightings never held up... or got published. Bob Altemeyer, who built the field’s standard measure of the authoritarian personality, gave the joke its shape in 1996, asking whether the authoritarian on the left was like the Loch Ness Monster: “an occasional shadow, but no monster.” The irony, of course, is that the man who wrote the instrument for measuring the creature was the one certifying that four decades of looking had turned up nothing.
Quelle surprise: he wasn’t lying. That’s what makes this story worth telling, I suppose. So, join me, dear reader, as I take you down this rabbit hole. Some of this I knew, some of this I learned while researching. It’s pretty fascinating (to me anyway).
In 1998 Altemeyer reported that the left-wing authoritarian had been “as scarce as hens’ teeth in my samples,” and by his lights, and his measures, it had been. His scale, the Right-Wing Authoritarianism (RWA) measure, drafted in 1981 and still the workhorse of the subfield, captured submission to authority, aggression toward sanctioned targets, deference to convention. It found those things on the right, over and over, across samples and decades.
On the left, well, it came up mostly empty. So the field drew a conclusion about human nature, that authoritarianism was a right-wing trait.
That ended up being a bigger leap than it looked then. The scale measured something real, something that predicts real behavior, and that something does concentrate on the right in the samples where it was tested and with the measure that was used.
But “our instrument finds it on the right” and “it exists only on the right” are different sentences, and political psychology slid from the first to the second and stayed there. Nobody built a second net, one shaped for whatever authoritarianism might look like on the left. So, for four decades the discipline cast one net, hauled up one kind of fish, and announced the sea contained only that fish.
This essay is about why the second net took forty years, and the answer comes in layers. The top layer is a story about measurement, and, well, it’s damning enough. The bottom layer is worse. It’s about who still feels authoritarianism as a threat at all, and it runs through a fight two German philosophers had in 1969, a strategy memo published in The Nation in 1966, and an argument that happened on X this week.
Start at the origin, because the origin explains more than it should.
The concept of the authoritarian personality came out of the Frankfurt School in American exile. Max Horkheimer, directing the Institute for Social Research’s American operation, ran the American Jewish Committee’s Department of Scientific Research; under his direction, the AJC’s Studies in Prejudice series set out to understand the catastrophe that had just consumed Europe. Its flagship, published in 1950, was The Authoritarian Personality, with the Frankfurt School’s Theodor Adorno as senior theorist alongside three Berkeley psychologists, Else Frenkel-Brunswik, Daniel Levinson, and Nevitt Sanford. Its instrument was the F-scale. (Yes, the F is for fascism.) It was built to detect the pre-fascist character, the personality type the European disaster grew from, and to detect nothing else.
Now, we’re not going down the definitional road here, fascism versus National Socialism versus communism versus the rest of the isms. Another day, perhaps. This part is a story about measurement, and well, deciding to measure.
That said, a measure engineered to find the incipient fascist was never going to find an authoritarian on the left—they’re different, they have different beliefs, different ideology. So, this wasn’t a flaw someone failed to catch. It was the design brief, the measurement decisions, and those governed everything downstream. The mission came before the instrument. And the mission was anti-fascism because of Hitler’s Third Reich’s horrifying assault on the world, not anti-authoritarianism in general. In 1950 those two commitments looked identical. They aren’t.
And the distinction was visible at the time: the year after the Berkeley study appeared, Hannah Arendt published The Origins of Totalitarianism, which treated Nazism and Stalinism as two species of a single genus. The framework that saw both was sitting on the same shelf as the instrument that could only see one, and the field picked the instrument.
People noticed immediately. In 1954, four years after publication, the sociologist Edward Shils wrote that the Berkeley group had proceeded “as if there were a unilinear scale of political and social attitudes at the extreme right of which stands the Fascist... and at the other end what the authors call the complete democrat who... actually holds the views of the non-Stalinist Leninist.” The objection wasn’t buried in a footnote of history. It was on the table in year four of this story, in a volume of methodological critiques everyone read and cited.
An alternative was even in hand. Milton Rokeach spent the late fifties building the dogmatism scale as a deliberately ideology-neutral measure of the closed mind, published in The Open and Closed Mind in 1960; his 1956 monograph carried the subtitle “an alternative to the authoritarian personality.” His own results were equivocal, small communist samples, differences that didn’t reach significance, and later work found dogmatism scores correlating with the right anyway. The road not taken existed. Everyone looked at it, shrugged, and went back to the F-scale’s descendants.
The shrug had good credentials, and the credentials are the trap. Why?
Psychology is hard if you haven’t tried it. You can’t see an attitude. You can’t weigh one. You ask (or you hook people up to brain measuring machines, which is expensive and not something I’ve ever had access or resources to do, so to the survey I go). Questions, statements to agree or disagree with, feeling thermometers, some format for coaxing a mind into a number, and even the number is a ranking, more agreement or less, not an amount of anything. Then you check your work, repeatedly, and the checks pass or they don’t.
Reliability: the answers hang together, this month and next. The RWA measurement scale has always been superb here.
Face validity: whatever you asked looks like it’s about authoritarianism. Altemeyer’s statements did, to the room that wrote them.
Neither check can tell you whether you sampled the whole concept or only the corner you already had in mind. A bathroom scale that reads the same number every morning is reliable. Set it on thick carpet and it’s reliably wrong. The methods textbooks have more types of validity where that came from, content, criterion, construct in its convergent and discriminant flavors, and researchers ran those checks too, for decades. Every one goes through the same bottleneck: a scale gets validated against what it should relate to, and the should comes from somebody’s theory of the construct.
The theory here drew the fascist. All the credentials are real. All of them attach to the questions someone wrote down, the questions came from a mental picture, and nothing in the machinery of measurement checks the question you didn’t think to ask.
What happened next is the part the discipline remembers least honestly, at least as I read the record. The blind spot didn’t stay in the instrument. It grew into a whole intellectual tradition.
Richard Hofstadter turned it into history. His 1954 essay “The Pseudo-Conservative Revolt” took its central concept straight from the source, and said so: “I borrow the term from the study of The Authoritarian Personality published five years ago by Theodore W. Adorno and his associates.” A decade later, “The Paranoid Style in American Politics” gave the postwar right its permanent diagnosis, “heated exaggeration, suspiciousness, and conspiratorial fantasy.”
Hofstadter was more careful than his inheritors; he wrote that the style was “not necessarily right-wing” and listed left examples, and he flagged reservations about the Adorno study’s methods. He also admitted the term was “pejorative, and it is meant to be.” The hedges got forgotten. The diagnosis stuck. And to be fair to that generation, the school wasn’t monolithic. For example, Lipset located authoritarian predispositions in the working class, not the right per se, an inconvenient placement the tradition also declined to pursue.
Altemeyer gave the hunch a scale of its own in 1981. Jost, Glaser, Kruglanski, and Sulloway built it into a general theory in 2003, conservatism as motivated social cognition: the threat-sensitive, closure-seeking, dogmatic mind sorting itself rightward. That paper also disposed of Rokeach’s alternative in a sentence, noting that “dogmatism has been no more useful than the construct of authoritarianism for identifying rigidity of the left.” Christopher Federico spent much of a distinguished career mapping the mechanism, how those needs become right-leaning belief, and how political sophistication sharpens the effect rather than washing it out. His later work, to its credit, narrowed the claim, finding the needs-and-ideology link strong for cultural conservatism and weak for economics.
But the tradition’s through-line held for seventy years: an apparatus for treating one side’s politics as a symptom, with the other side’s politics left standing as simply true.
William Stone said the quiet part in a title in 1980: “The Myth of Left-Wing Authoritarianism.” His specific target was the claim of equal prevalence, and he allowed that individual cases existed. But titles do the cultural work titles do, and his verdict, that it “seems a waste of time to continue to test” the hypothesis, arrived before any instrument existed that could have tested it properly. The conclusion came first. The measure that might have checked it came forty years later.
The field did, in the end, find the monster. Conway and colleagues built a measure in 2018 and titled the paper, cheerfully, “Finding the Loch Ness Monster.” Costello, with a team that included the late Scott Lilienfeld, did it comprehensively in 2022: six samples, more than seven thousand people, in the Journal of Personality and Social Psychology, the discipline’s flagship. Left-wing authoritarianism turned out to be real, measurable, and structured: anti-hierarchical aggression, an appetite for top-down censorship, anti-conventionalism. More uncomfortable still, its bearers shared the classic kit, dogmatism, need for closure, belief in a dangerous world, an appetite for aggression against political enemies. The core the tradition had spent decades attributing to the conservative mind showed up on the other wing, attached to different targets.
That looks like self-correction working. It is, right up until you check the clock.
None of this proves the left is more authoritarian than the right, and I’m not arguing so. Costello’s team is careful; the profiles overlap at the core but differ at the edges; prevalence is genuinely contested, and Costello himself has suggested that right-wing authoritarians outnumber left-wing ones in the United States by something like three to one. The LWA researchers never claimed parity. Neither do I. And some of the field’s early caution was earned, because the first attempts at measuring left authoritarianism really were bad.
But honest difficulty doesn’t explain a forty-year gap, because difficulty isn’t distributed by accident. Over those same decades the journals filled with ever-sharper instruments for prejudice, for right-wing authoritarianism, for every attitude the discipline expected to find and wanted to explain. The measures a discipline sharpens and the measures it leaves unbuilt are a map of its curiosity, and the unbuilt measure sat next to a construct that would have implicated the researchers’ own political coalitions. As late as February 2020, the Society for Personality and Social Psychology was staging a symposium asking, in its actual title, whether left-wing authoritarianism was real, barely eighteen months before seven thousand subjects’ worth of yes arrived at the flagship journal.
The proximate machinery of a gap like that is familiar, and none of it requires a conspiracy. It’s how groups kill (maybe good) new ideas in the room, identity protection run in slow motion across a generation: the shared prior nobody defends because nobody in the room holds the other one, the questions that never get funded because they don’t strike anyone as “interesting enough.”
Expertise helps less than you’d hope; Dan Kahan’s work suggests that on identity-charged questions, more sophistication often means sharper defense of the conclusion you already hold, though the strongest version of that claim has taken hits and earns its caveats. And the room’s composition is not in dispute: when Inbar and Lammers surveyed their own field, around six percent of social psychologists identified as conservative, and more than a third admitted that, between two equally qualified candidates, they’d favor the liberal. Alvin Gouldner saw the deep pattern half a century ago, and I’ve worked through it before: when a field’s authority stops resting on getting things right, it rests instead on protecting its self-image, its identities, and a measure you never build is a cheap way to keep the image intact.
So the verdict on the correction is not health. A forty-year lag, broken only after outside pressure, the replication crisis, a decade of heterodox needling, made not-looking cost more than looking. A system barely working, grinding against its own self-generated friction.
Not a clean bill. The diagnosis.
And yet. Monoculture, motivated reasoning, reputational risk: all of that explains how a blind spot survives. It doesn’t quite explain why this blind spot had this shape. For that you have to go one layer down, to the part nobody wants to say out loud.
Who actually feels authoritarianism, as such, as a threat?
Every instrument in this story was built by somebody with an answer.
The anti-authoritarian project was a liberal project. Small-l liberal: proceduralist, pluralist, open-society liberal. The postwar settlement that built the F-scale, the UN, the civil-liberties consensus, and the whole architecture of “never again” rested on one commitment: coercion is the enemy no matter whose hand holds it, and process, rights, procedure, contestability, is the only reliable defense, because anyone’s hand can hold the whip eventually. In that frame, an authoritarian of the left is not a category error. He’s an expected finding. The whole point is that the whip doesn’t care who swings it.
But there is another frame, and the academy has spent fifty years drifting toward it. In the power frame, the master category isn’t coercion; it’s oppression. The question isn’t “is force being used,” it’s “who is using force on whom, and on whose behalf.” Coercion downward is domination. Coercion upward is liberation, or at least self-defense. Within that frame, “left-wing authoritarianism” doesn’t parse as a discovery. It parses as a trick, an attempt to smuggle the master’s categories into the liberation movement’s luggage. You don’t build an instrument to catch a danger you don’t experience as one.
The receipts on this exist, and they are remarkable. The same school that built the F-scale produced, fifteen years later, the explicit theory of the second frame. Herbert Marcuse joined Horkheimer’s Institute in 1933 and remained one of the Frankfurt School’s most famous names all his life. In 1965 he published “Repressive Tolerance,” and there is no need to characterize it, because it characterizes itself: “Liberating tolerance, then, would mean intolerance against movements from the Right and toleration of movements from the Left.” He called for “withdrawal of tolerance from regressive movements before they can become active; intolerance even toward thought, opinion, and word.” He knew exactly what he was proposing; he called his own notions “anti-democratic” in the text and justified them on emergency grounds: “I maintain that our society is in such an emergency situation, and that it has become the normal state of affairs.” In the 1968 postscript he sharpened it to “restraining the liberty of the Right.” The tradition that engineered the instrument against fascism also wrote, in plain declarative sentences, the permission slip for asymmetric coercion. Same milieu. Same shelf.
And then the school fought about it, in letters that deserve to be famous. In 1969, with student radicals occupying university buildings, Adorno called the police on the SDS students occupying his own Institute. Marcuse was appalled, and their letters survive: “if the alternative is the police or left-wing students, then I am with the students,” Marcuse wrote. Adorno’s reply is the whole essay in one sentence: “I take much more seriously than you the danger of the student movement flipping over into fascism.” Marcuse dismissed the phrase “left fascism” as a contradiction in terms. Adorno pressed: “But you are a dialectician, aren’t you? ... might not a movement, by the force of its immanent antinomies, transform itself into its opposite?” The two most famous members of the school that invented authoritarianism research spent 1969 arguing, in private, about whether the anti-authoritarian commitment applied to their own side. Adorno, the man whose name is on the F-scale study, died that August still insisting that it did. The field that inherited his instrument spent the next forty years proceeding as if it didn’t.
The fork never really closed. This week it was running on X, in a way that struck me. A widely shared post from the left put the indictment plainly: “Liberalism is so depleted it’s insane. No ambition, no new answers, no adjustments. Just 1992 forever. A Clintonite infinity. We must imagine Sisyphus balancing the budget.” And Kelsey Piper’s reply is the proceduralist confession, guilty as charged and proud of it: “I think the fundamental disconnect here is that you want[s] politics to solve man’s search for meaning and I want politics to deliver a prosperous, safe, free society in which everyone can go find meaning in something other than politics.”
Read the replies under it and the frame war performs itself. “You don’t get to have a neat barrier between the moral/existential and politics,” says one. “It’s a distinction with no difference,” says another. And one reply hands over the historical hinge with both hands: the real issue, it says, is “rejecting the liberal consensus formed after the WW2... No fresh ideas, no fresh approaches. Your thought leaders are intellectually and morally bankrupt.” There it is. “1992 forever” is a complaint about 1945. The consensus being rejected as bankrupt is the postwar anti-authoritarian settlement itself, the one the F-scale was built to defend.
Adorno and Marcuse, fifty-seven years on, in twitter prose.
Piper asked the question that sorts everyone: “is it a knock against politics if they merely deliver a prosperous, free society? yes or no?” I’d put the same fork in social-thermodynamic terms, because I keep ending up there. Liberalism, institutionalism, proceduralism, whatever you call the entropy-resisting posture, accepts that social order degrades by default and treats politics as maintenance: unglamorous, repetitive, Sisyphean, and the only thing standing between a society and its own second law.
The Sisyphus jab only works if maintenance is contemptible. The liberal answer is to accept the boulder earnestly, Camus said one must imagine Sisyphus happy, because the boulder is the job. The rival posture treats degradation as opportunity: the system is the oppression, so hastening its collapse is progress, and what comes after will be better. That posture has a published pedigree too. In 1966, two Columbia academics, Richard Cloward and Frances Fox Piven, laid it out in The Nation as an explicit strategy: overload the welfare system past its breaking point to force “a profound financial and political crisis” and compel its replacement, in their design, with a federal guaranteed income. Crisis as a plan, in print, with footnotes. (Read the 1966 article, not the conspiracy literature that later grew around it; the primary text is doing all the work I need here.)
So anyway. The net. An entropy-resister builds alarms against coercion from every direction, because coercion is what entropy looks like when it wears a uniform. An entropy-accelerator has no use for such an alarm; in that frame the system’s coercion is the threat and the movement’s coercion is the remedy, so an instrument that treats the two symmetrically is not just pointless but hostile.
Political psychology built its alarm in 1950, when its center of gravity sat inside the postwar liberal settlement. It stopped extending the alarm as the second frame gained ground. The field didn’t lose the ability to see the left-wing authoritarian. It lost the motivation to look.
The right objection arrives on schedule here: most social psychologists are not Marcusians. Inbar and Lammers found a room full of ordinary liberals, Democrats, not revolutionaries. True, and it rescues nobody, because the transmission runs through deference, not doctrine. In the academy’s moral economy, the flank that can accuse you of standing with the oppressor sits to your left, and no equivalent tribunal sits to your right. So the liberal researcher polices rightward and defers leftward, and the deference does the rest. Nobody in the room has to hold the power frame for the room to enact its priorities. Watch what happened this week to Alexander Kustov, a center-left political scientist at Notre Dame, for a live demonstration: he raised the question and was not refuted but classified, recast in the replies as “the noted conservative Notre Dame professor,” which he isn’t. That’s the enforcement mechanism working exactly as designed. You don’t need commissars when classification does the job. Anti-authoritarianism is the liberal’s master category. Power is the left’s. Nobody can even agree what “liberal” means anymore, but this is the working definition I keep coming back to: a liberal is someone for whom Piper’s question has an easy answer, and for whom the whip is the enemy in whoever’s hand it rides.
As I said above, Kustov is not a conservative. That’s what makes him the right messenger for at least this part of the story. He studies public opinion on migration, sits somewhere on an idiosyncratic center-left, and, by his own account, used to assume the people complaining about viewpoint discrimination in the academy were exaggerating. Then he spent a week watching his own colleagues, in his own replies, defend the thing they had long insisted wasn’t happening. He wrote it up as “Reality May Have a Liberal Bias, but the Social Sciences Still Need Conservatives,” and the title concedes the maximum and still wins the argument. Grant that reality tilts left. You still need people in the room who begin from other assumptions, or the questions your assumptions hide never get asked.
Kustov’s mechanism is a good one to read and think about, and the RWA scale is the cleanest illustration of it. A couple of conservatives in the room in 1981 would have flagged the loaded items the first afternoon, not because conservatives reason better, but because they’d have felt the itch of a measure tuned against their side while everyone else felt nothing. The mechanism has nothing to do with hiring people who deny evidence; it runs on which questions get asked at all. A room that shares a prior doesn’t think to build the instrument that would test it. Homogeneity mostly leaves the answers alone. It works upstream, on which questions exist at all, where nobody’s watching. And the homogeneity is not subtle: when Musa al-Gharbi ran the numbers for Heterodox Academy, ideological underrepresentation in the professoriate dwarfed the gaps for race, gender, and sexuality. We spent a generation, rightly, treating demographic homogeneity as a limit on what a field can see. Point the identical logic at viewpoint and it gets dismissed as a talking point.
The scope matters, and it’s what moved me off my own skepticism about this argument. The blind spot isn’t universal. Physics is ideologically homogeneous and corrects itself viciously, because reality slaps a wrong answer down inside a week. The mechanism bites where reality’s feedback is slow, in the value-laden territory where a wrong answer can sit comfortably for forty years because nothing external forces the issue. Left-wing authoritarianism might be the cleanest case in the social sciences: slow feedback gave the monoculture’s prior time to harden into consensus, and the consensus sat like a dam across everything downstream.
Worse, the stakes don’t stay penned inside a methods dispute.
Wall off a set of questions and they don’t evaporate. They get answered elsewhere, by people with worse intentions and lower standards, and the people who walked away have no standing left to object. Watch it happen in real time: this past month a careful, peer-reviewed finding about how partisans caricature each other, one side reaching for images of savagery, the other for images of childishness, got boiled down, inside a day, into a bumper sticker reading “conservatives think, liberals feel.” Nobody in the discipline is positioned to contest that laundering with much credibility, because contesting it means owning the terrain of left illiberalism, and forty years of “myth” titles and Loch Ness jokes said the terrain was empty. The same abdication, one field over. You can’t leave ground unoccupied for forty years and then be persuasive about who moves onto it.
And here is the bill’s cruelest line item. Does left-wing authoritarianism feel like it’s on the rise right now? A lot of people, on several sides, seem to think so. Is it? Nobody can say. Measuring a rise requires a baseline, and the baseline was never built. There is no 1985 LWA number, no 2005 number, nothing to plot 2022 against. The field spent forty years making sure that the one question everyone now wants answered is a question nothing in the literature can answer. Trend lines don’t come retroactively. That’s what the incuriosity actually cost, and we’re paying it at the worst possible moment.
Objections, in the order they’ll arrive.
“Nobody ever denied LWA existed; that’s a strawman.” Partly fair, which is why this essay claims something more specific. Altemeyer himself acknowledged the concept, found “wild-card” authoritarians high on both scales, and allowed in 2006 that left-wing authoritarians exist “here and there.” He even built a left-wing scale for the 1996 book. Look at how, though: it defined the creature as submission to an established revolutionary movement, a species that barely exists in North America, and when it caught nothing, the nothing was read as a fact about the ocean rather than the net. The serious measurement work that did happen was deployed abroad, in post-communist Europe, where the left held the establishment. So the claim is not that anyone said “it doesn’t exist.” The claim is forty years of motivated incuriosity at home: the question ridiculed, “myth” in the titles, the one domestic net cut so the local creature couldn’t fit it, and SPSP still symposium-debating whether the construct was real in 2020. The measurement record is the exhibit.
“What Costello measured isn’t authoritarianism; real authoritarianism means submission to established authority.” This is the strongest objection, made most carefully by Artur Nilsson, and it deserves a real answer. Three, actually. The tradition’s own founding instrument wasn’t a submission-meter; the F-scale measured “implicit antidemocratic tendencies,” fascist potential, and by that founding standard an anti-democratic left qualifies on its face. Second, one of the three LWA factors is top-down censorship, which is not the mirror of authority but the use of it. Third, read the definitional defense closely and notice what it concedes: Nilsson’s own formulation distinguishes right antidemocracy (norm enforcement) from left antidemocracy, which runs through “prioritization of superordinate ideological values over established societal norms and democratic principles.” Values over procedure. That is precisely the power frame this essay has been describing, offered by the construct’s chief critic as a definition. So the fight is over the name. Nobody serious is still disputing that the thing exists.
“The asymmetry never rested on one scale. Social dominance orientation, system justification, need for closure, the whole nomological network converges on the right.” This is the version the methodologists will bring, and it proves less than it seems to. Convergence among instruments is only as informative as the instruments’ coverage. The network’s tools were built by the same people in the same era, and the politically pointed ones all pointed the same way; the neutral ones, like need for closure, got their political meaning from how psychologists applied them. Nets that converge on the same waters converge; that is not evidence about the waters they never entered. The test of the convergence defense arrived when someone finally cut a net for the other side, and the result is the whole story: Costello’s left-wing authoritarians carried the same core kit, the dogmatism, the need for closure, the dangerous-world belief, the appetite for aggression, that the network had spent decades certifying as the conservative mind. The network didn’t converge on the right. It converged on wherever it was pointed. The cleanest-looking counterexample makes the same point: the child-rearing measure, which asks nothing about politics at all, only which virtues matter more in a kid, obedience or independence, good manners or curiosity. No political content anywhere in the items, and it still finds authoritarianism on the right, which is why it gets offered as the neutral instrument. But the politics isn’t in the items. It’s in the scoring key. Valuing obedience counts as authoritarian; valuing independence counts as its opposite. That’s the 1950 construct decision again, submission to the established order, hiding in a question about children. A Costello-style left authoritarian, the one who wants the heretic deplatformed but raises her kids to question everything, sails through it without ringing a single bell.
The field spent forty years watching the rather large raptor in front of it. “Clever girl,” indeed.
“LWA is rarer and less dangerous; the field allocated attention rationally.” Maybe, and remember Costello’s own three-to-one estimate; nobody here is claiming parity in the general population. But rational allocation doesn’t produce zero validated instruments across four decades, and there’s a wrinkle: prevalence in the general population isn’t prevalence everywhere. Nilsson himself grants that left-wing authoritarianism is common in contexts “where the socially sanctioned authorities and norm systems are left leaning.” Now apply that sentence to the institution doing the measuring. It describes the modern academy.
John Jost skipped defense entirely, and I appreciate the clarity. His recent Lewin Award address is titled, in full seriousness, “Both-Sideology Endangers Democracy and Social Science.” His argument, to state it fairly, is that asserting left-right equivalence without evidence is false balance, and that the evidence shows asymmetry. But notice what the framing does: it treats the symmetric question as the hazard, names the very act of both-sides comparison as a threat to democracy and to the science, in the field’s own journal, from its most decorated chair. When “we never said it doesn’t exist” gives way to an award address arguing that both-sides framing itself endangers the discipline, the retreat from bailey to motte has ended and the bailey has been rebuilt in public. The forty-year missing instrument stops being a mystery at that point. It starts looking like policy.
One more thing, and it cuts at me, not them. Everything in this essay is subject to the mechanism it describes. I hold the liberal commitment; the finding stings me in a way it will not sting a reader in the power frame, and my sense that this all matters enormously is exactly what worldview-protective cognition feels like from the inside. That’s not false modesty; it’s the test the essay proposes. Whether left-wing authoritarianism reads to you as a real problem, a category error, or a right-wing talking point is less a fact about the evidence than a reading on which frame you’re standing in, and mine is showing.
(Check the citations. All of them. That’s what they’re for.)
Here the serious people disagree and I won’t fake my own certainty on this. It’s rather complicated.
Quotas are a trap; a conservative-headcount metric would be gamed into self-declared heterodoxy long before it produced a better study. Reverse discrimination swaps one political test for another. At the harder end, the philosopher Matt Lutz, defending the Vanderbilt report’s conclusion, argues that scholarship premised on rejecting objective truth forfeits the protection of academic freedom altogether. I’m with him on the metaphysics. The remedy makes me flinch, for the reason this essay has spent five thousand words supplying: a forfeits-clause needs a gatekeeper, and the monster is what happened the last time this particular guild kept a gate. Kustov settles on the modest version, and it may be the most anyone can defend: stop punishing the questions. Make it survivable to hold a few unfashionable priors out loud, not because balance is owed, but because the room needs those priors to see straight.
This solution, however, is not for our ideologically pure-of-heart colleagues, friends.
I doubt it will happen through reason or virtue, and the monster is the reason. The field didn’t correct from the inside; it corrected forty years late, too late to have built any internal defenses, after outside pressure raised the price of the blind spot, and the same bill is now arriving across the academy, from outside, for the same reason.
Let’s be precise about what “outside” means, because the distinction matters right now: the pressure that actually found the monster was replication-crisis methodology (and that’s just for the things, you know, that could actually be replicated), heterodox critics, and reputational cost, not a government with funding threats and viewpoint conditions. That version just swaps one political test for another, and nothing in this essay supports it. But a discipline whose leading figure argues that symmetry itself endangers democracy is not about to reform because someone asked nicely, either. It will get corrected the way it got corrected on the monster: from outside, against resistance, and later than it should have been.
The creature was right there in the water the whole time. Adorno saw its shadow from the beginning: he died in August 1969 with the argument still open, having spent his last months warning Marcuse that a movement can transform into its opposite, and the field that inherited his instrument spent the next half century declining to check. When somebody finally did, the monster was exactly where the joke said it couldn’t be.
What should trouble the discipline now isn’t the miss. Misses happen, this whole thing is a flawed, human endeavor, but correction is what good, well done, science is for.
The trouble is that the looking only started when someone outside the boat forced the question, that the argument, even now, is still about the net we use to catch the beast, and that nobody inside has yet asked the next question: what else is in the water that no one has wanted to measure?