Dan Sperber (b. 1942) was born in Cagnes-sur-Mer on the French Riviera. His father, the Austrian-French novelist Manès Sperber (1905-1984), broke from the Communist Party in the 1930s and wrote about ideological capture and political faith. His parents, both non-religious Ashkenazi Jews, raised him an atheist but passed on a respect for his rabbinic ancestors and for serious religious thinkers in general.
Sperber says he came to anthropology because he wanted to understand how rational people end up holding mistaken beliefs about the supernatural. The puzzle stayed with him for the next sixty years and shaped most of his work.
He studied anthropology at the Sorbonne and at the University of Oxford. At Oxford, Rodney Needham (1923-2006) introduced him to structural anthropology. Back in Paris he attended the seminar of Claude Lévi-Strauss (1908-2009), the founder of structuralism, who encouraged what Sperber later called his “untypical theoretical musings.” In 1965 he joined the CNRS as a researcher in the African studies laboratory and conducted fieldwork among the Dorze people of Ethiopia.
His first book, Le structuralisme en anthropologie (1973), grew out of his early enthusiasm for the structuralist program. By the time it appeared, he had begun to turn against it. In Rethinking Symbolism (1975) he attacked the dominant semiological treatment of culture associated with Victor Turner (1920-1983) and Clifford Geertz (1926-2006). The semiological view treats ritual and myth as systems of meaning to be decoded. Sperber argued that this got the cognitive picture wrong. People do not decode symbols. They process them. Symbolic thought is not a code but a way the mind handles information that ordinary inferential thought cannot fully digest. The book proposed a cognitive psychology of religion and ritual rather than a semiotics of culture.
In On Anthropological Knowledge (1985) he turned to ethnographic method. Anthropologists, he argued, often present interpretive readings of native belief as if those readings were objective records of what informants think. The interpreter’s voice gets confused with the informant’s. The result is a literature that resists falsification and lacks explanatory power.
The same impulse drove Explaining Culture: A Cognitive Approach (1996), the book that made his reputation in the cognitive sciences. Here he laid out his alternative to symbolic anthropology and to the meme theory of Richard Dawkins (b. 1941). He called it the epidemiology of representations, later renamed cultural attraction theory. The picture goes like this. Mental representations spread through populations the way pathogens spread, by transmission from mind to mind, but unlike pathogens they almost never copy with high fidelity. Each act of transmission is a reconstruction. Cultural items survive across generations not because minds are good copying machines, which they are not, but because some contents are catchy. They line up with how the mind works. Cognitive attractors pull noisy reconstructions back toward stable shapes. Religious beliefs, folktales, and rituals owe their persistence to that pull. The picture is Darwinian without being memetic. It explains population-level patterns through individual cognitive attractors operating over time.
Kim Sterelny (b. 1950) later named this body of work the Paris School of cultural evolution and contrasted it with the California School of Robert Boyd (b. 1948) and Peter Richerson (b. 1943), who lean more on dual-inheritance models of high-fidelity transmission. The two schools agree that culture evolves and disagree about the cognitive picture under the evolution.
Sperber’s other major line of work began in the late 1970s with Deirdre Wilson (b. 1941), the British linguist and philosopher he met at University College London. Together they developed relevance theory, set out in Relevance: Communication and Cognition (1986) and refined in Meaning and Relevance (2012). Relevance theory is a theory of pragmatics, the part of linguistics that asks how hearers move from what a sentence says to what a speaker means. The theory rests on a claim about cognition. The mind seeks the largest cognitive payoff for the smallest processing cost. When a hearer interprets an utterance, he picks the interpretation that gives him the best ratio of effects to effort. The communicative principle of relevance follows: a speaker who chooses to speak signals that his utterance is worth the effort of interpretation. The theory has become a standard frame in pragmatics and has shaped work in linguistics, artificial intelligence, and cognitive psychology.
His third major project, developed with Hugo Mercier (b. 1974), came out of his work on epistemic vigilance, the cognitive capacities that let people resist deceptive communication. The Enigma of Reason by Hugo Mercier and Dan Sperber argues that reason did not evolve to help us find truth or make better individual decisions. It evolved to help us justify our actions to others and evaluate the justifications others provide. Reason works well in adversarial group settings, where two or more people exchange arguments, and works poorly in solitary deliberation, where the lone reasoner falls prey to confirmation bias. The classic findings of the heuristics-and-biases tradition look less like flaws of an imperfect truth-tracker and more like features of a tool built for argument.
The argumentative theory connects back to relevance theory and to the epidemiology of representations. All three projects share a common picture. Communication is not a transparent transfer of meaning. It is a cognitive process between minds that have evolved to produce, evaluate, and resist messages. Reason is a tool inside that process.
Sperber spent most of his career at the CNRS, moving through the African studies laboratory, the Laboratoire d’ethnologie et de sociologie comparative, the Centre de Recherche en Epistémologie Appliquée, and from 2001 the Institut Jean Nicod, the cognitive science institute affiliated with the École Normale Supérieure. He later took a chair at Central European University, in the departments of cognitive science and philosophy. He directs the International Cognition and Culture Institute, an online research and discussion site.
He has held visiting posts at Princeton, Michigan, Hong Kong, Chicago, the London School of Economics, University College London, and the University of Bologna. He is a Corresponding Fellow of the British Academy. He won the Rivers Memorial Medal of the Royal Anthropological Institute in 1991, the Silver Medal of the CNRS in 2002, the Mind and Brain Prize in 2009, and the inaugural Claude Lévi-Strauss Prize for French research in the humanities and social sciences in 2009. He has delivered the Malinowski Memorial Lecture at the LSE, the Mircea Eliade Lectures at Western Michigan, the Radcliffe-Brown Lecture at the British Academy, the Robert Hertz Lecture at the EHESS, the Lurcy Lecture at Chicago, and the Carl Hempel Lectures at Princeton.
His career sits at an unusual crossroads. He trained as an anthropologist and ended up a cognitive scientist. He fought structuralism and symbolic anthropology from inside French intellectual culture and built an alternative drawing on Anglo-American cognitive psychology. He worked on linguistics with a partner from across the Channel and on reasoning with a younger collaborator from the next generation of French cognitive science. The throughline is the question he started with as a young man in the 1960s: how do rational minds come to hold the beliefs they hold, and how does culture move through populations of such minds? The answer he built over five decades treats culture as the visible pattern of countless small cognitive events, each one a reconstruction, each one shaped by the architecture of the mind doing the reconstructing.
Sperber’s convenient beliefs are easy to list. Naturalism in the sciences of mind pays well at the CNRS and Institut Jean Nicod. Methodological individualism applied to culture pays well in cognitive science, costs little outside humanistic anthropology. Atheism is the default European academic stance. He has avoided the third rails. He does not write about race, sex differences, group genetic variation, or Jewish ethnocentrism. His coalition stays intact.
Now the inconvenient ones.
First, his attack on anthropological relativism. In his 1982 essay “Apparently Irrational Beliefs” and in his book Explaining Culture (1996), he argued that the protected category of “different rationalities” was a sentimental defense of cultural anthropology’s professional turf. The Dorze of Ethiopia who claim that the leopard is a Christian animal fasting on Wednesdays and Fridays are not operating with a different logic. They hold the belief in a different register, with different commitments, but the laws of thought do not vary by culture. Cultural anthropologists treat that claim as imperialism. Sperber accepted the cost.
Second, the argumentative theory of reasoning, developed with Hugo Mercier (b. 1974). The Enigma of Reason by Hugo Mercier and Dan Sperber argues that reason did not evolve to help us find truth or make better individual decisions. It evolved to help us justify our actions to others and evaluate the justifications others provide. The claim is inconvenient for several coalitions at once. Liberal democratic theorists need reason to track truth through deliberation. Educators need reason to be teachable as a truth-finding tool. Philosophers need reason as their professional warrant. Sperber’s account treats reason as a coalition tool. He kept his job because cognitive science can absorb the claim as an empirical hypothesis, but the broader implication corrodes a great deal of received wisdom about expertise, deliberation, and democratic discourse.
Third, epistemic vigilance. Sperber and his collaborators argue that humans have evolved defenses against deception. Audiences calibrate trust by source, message, and context. The implication is awkward for the trust-the-experts script. If audiences resist expert claims, that resistance is not a bug to be corrected by better science communication. It is a feature of human cognition. Public health authorities, climate communicators, and university administrators do not want to hear that lay distrust of official sources reflects something cognitively reasonable.
Fourth, his treatment of religion and ritual. He argues that religious commitments live in a reflective register different from ordinary factual belief, and that ritual practices spread through cognitive attractors rather than conscious adoption. This pleases neither devout believers nor the New Atheists who want religion to be a simple cognitive error.
Fifth, his methodological individualism in cultural explanation. Cultural attraction theory says some contents catch on because of how human minds work, not because of social construction. Blank-slate constructivism dominates much of the humanities. Sperber’s account says the slate is not blank and that the catchiness of certain ideas reflects cognitive structure.
Sixth, his critique of the meme concept. Richard Dawkins’s memetics treats cultural items as replicators that copy with high fidelity. Sperber argued that cultural transmission is reconstructive rather than replicative, and that the high-fidelity assumption fails empirically. This put him at odds with a popular framework that had powerful advocates.
Why did these inconvenient beliefs survive in his career? Sperber writes from cognitive science, not anthropology, so his attack on anthropological relativism came from outside the threatened guild. He has institutional protection at the CNRS and Institut Jean Nicod, where philosophy of mind sets the terms. His heterodoxies are intellectual rather than political, which makes them tolerable in a way that, say, Amy Wax’s (b. 1953) heterodoxies are not. Turner’s frame predicts that an academic with several inconvenient beliefs will compensate by holding standard convenient beliefs in adjacent areas. Sperber fits the pattern. He is heterodox on reasoning and culture and orthodox on the third rails. The portfolio works.
The most inconvenient of his beliefs, by Turner’s measure, is the argumentative theory of reasoning. It corrodes the warrant for academic authority. Most academics depend on the premise that their reasoning tracks truth and that their disagreements with the public reflect superior epistemic discipline. Sperber says reasoning is a coalition tool that produces better outcomes mostly in adversarial group settings, and that the lone reasoner is prone to motivated confirmation. If his readers took the claim seriously, they might demand of professors what professors demand of others: produce the adversarial test, not the credentialed pronouncement. That demand has not been made, which is evidence of the theory.
Dan Sperber builds his career on the claim that reasoning did not evolve for truth. It evolved for argument. People reason to justify themselves and to evaluate the justifications others offer. The Enigma of Reason by Hugo Mercier and Dan Sperber lays this out.
Apply Pinsof to Sperber and a problem opens.
If arguing is not about persuasion, then Sperber’s career-long argument is also not about persuasion. He has spent fifty years arguing for naturalism in social science, against memetics, against structuralism, against Gricean pragmatics, against the standard picture of reason. What was he doing the whole time?
Pinsof’s answer: tribal work.
Sperber rallies the naturalistic cognitive science coalition. He gives them flags. Cultural epidemiology is a flag. Relevance theory is a flag. The argumentative theory of reasoning is a flag. Members of the coalition wave these flags at conferences, in graduate seminars, in citation chains. The flags say: we are the serious naturalists, they are the soft humanists.
Sperber rationalizes the coalition’s superiority. The continental humanities lack rigor. Memetics oversimplifies. Pure relativism cannot explain cultural stability. Each rationalization makes the coalition look smarter than its rivals.
Sperber spars. He does not engage the strongest version of structuralism. He does not steelman pure relativism. He picks easier targets. He answers some objections and skips others. Pinsof’s warning signs of pseudoargument apply: targets softened, alternatives dismissed, opponents’ questions half-answered.
Sperber defends his status. A senior chair at Jean Nicod, an appointment at Central European University, decades of co-authored papers, a long line of students. The argumentative theory of reasoning lands late in this career as a capstone. It says: your old picture of reason was wrong, ours is right. That sentence raises the speaker’s status.
Sperber covers the operation in the language of science. The cover story: I am describing how the human mind works. The cover story works because the descriptions are partly true. Pinsof’s point holds anyway. Partly true descriptions can still serve coalition purposes. The descriptions that get airtime, get funded, get cited, are the ones that flatter the coalition. The descriptions that might embarrass the coalition stay quiet.
Now the recursion bites.
Sperber cannot apply his own theory to his own career without losing the career. If The Enigma of Reason is a coalition product, then The Enigma of Reason is not a proof of anything. It is an argument made to a tribe. The tribe accepts it because the tribe wants it. Sperber knows this and does not say it. Saying it might cost the chair, the citations, the standing.
Pinsof says it for him. That is why Pinsof writes on Substack and not at CNRS.
Sperber argues that arguing is not about truth. Sperber’s argument for that claim is also not about truth. It is coalition work for the naturalistic cognitive science tribe. The argument succeeds inside the coalition and stalls outside it. Form fits function.
The Set
The Paris School includes Pascal Boyer (b. 1957), who wrote Religion Explained: The Evolutionary Origins of Religious Thought, and Scott Atran (b. 1952), who worked on folk biology and later on sacred values and the men who kill and die for them. The linguistics wing runs through Deirdre Wilson (b. 1941) at University College London, Sperber’s partner in relevance theory, and her students Robyn Carston and Diane Blakemore. Relevance theory extends and corrects Paul Grice (1913-1988). The younger generation includes Hugo Mercier, co-author of The Enigma of Reason, along with Olivier Morin, Nicolas Claidière, Christophe Heintz, Nicolas Baumard, Thom Scott-Phillips, Olivier Mascaro, Fabrice Clément, and the philosopher Gloria Origgi. Lawrence Hirschfeld worked the social-categorization side. Maurice Bloch (b. 1939), a Marxist anthropologist, argued with the set as a friendly antagonist. Their homes are the CNRS, the Institut Jean Nicod on the rue d’Ulm, and the Central European University. Sperber built them a hub, the International Cognition and Culture Institute.
What they value most is explanation over interpretation. The reigning anthropology of their youth, the symbolic anthropology of Victor Turner (1920-1983) and Clifford Geertz (1926-2006), read a culture as a text, a web of meanings the analyst interprets with care. Sperber’s first book, Rethinking Symbolism, rejected that. He wanted causes. He wanted to say why a symbol spreads and holds, not what it means to a sensitive reader. The set inherits that hunger. They want anthropology to behave like a science, to find regularities across cultures, to make predictions, to run experiments. The younger men do run them.
They value naturalism. Mind and culture belong to the same world the natural sciences describe, and no separate social level floats free of psychology and biology. Explaining Culture: A Cognitive Approach states it plainly: culture is the distribution of representations across a population, ideas in heads and the public signs that carry them, and you explain that distribution by studying how minds take in, store, and pass on ideas. They also value clarity and they detest the guru. Sperber attacked the prestige economy of French theory, the rewarding of profound-sounding obscurity, and the set prizes plain argument and claims a critic can test.
Their hero takes something the humanities had walled off as sacred or beyond reach and gives it a cognitive account. Religion is the prize. Sperber set the puzzle. Boyer answered a large part of it with the claim that religious concepts succeed because they are minimally counterintuitive, a familiar template with one strange feature, easy to remember and pleasant to repeat. Atran carried the work toward why men treat certain values as beyond trade and price. The heroic act is the naturalizing of the holy. A second hero appears in the circle, the iconoclast who breaks a reigning paradigm without contempt for his teachers. Sperber is the model again, and the lineage reads his career as the pattern.
The status games run on three tracks. The first prize comes when your framework stops being one option and becomes the standard others assume. Relevance theory reached that in pragmatics. It is taught, applied, and fought over, and the field can no longer ignore it. The second track is the Paris-California contest. The California School of Robert Boyd, Peter Richerson, Joseph Henrich, and Michael Tomasello says culture spreads by high-fidelity copying and selection on variants. The Paris School says representations get rebuilt each time they pass between minds, and they stabilize because cognitive attractors pull them toward the same shapes. Cinderella survives because minds reconstruct her, not because the tale travels word for word. Both sides claim Darwin, and the contest decides who counts as the better Darwinian of culture. The third track is distinction from memetics. Richard Dawkins (b. 1941) gave the public the meme, a unit of culture that replicates the way a gene does. Sperber rejected the replicator picture and took standing as the sophisticated alternative to a popular idea. Membership in the public-science world counts too, the Edge.org orbit of John Brockman, the adjacency to Steven Pinker (b. 1954) and Daniel Dennett (1942-2024). And inside the set, the founder who produces disciples outranks the lone theorist. Sperber’s standing rests in part on a living lineage that carries the program and cites him as it goes.
Their normative claims follow from the values. Social science ought to be causal and naturalistic, and interpretation by itself does not count as explanation. Anthropology ought to seek what recurs across cultures rather than retreat into the uniqueness of each. Reasoning ought to be understood as a social tool. Mercier and Sperber argue that reason evolved to produce arguments and to weigh the arguments of others, so a lone man reasons poorly and a group that argues well reasons better, and the practical lesson is to build settings where people argue. They also hold that listeners ought to screen what they are told, and that they do. The 2010 paper “Epistemic Vigilance,” written by Sperber with Clément, Heintz, Mascaro, Mercier, Origgi, and Wilson, argues that trust runs calibrated rather than automatic, and that communication holds together because people watch their sources.
Their essentialist claims sit underneath all of it. The mind has a fixed architecture, a set of domain-specific systems present in every human: folk biology, folk psychology, a sense for physical objects, a system for sorting people into kinds. These belong to human nature rather than to culture, and they set the channels culture runs in. Human nature is real and shared, against the blank slate. Culture has no separate existence apart from the minds that make and remake it. Religious belief has a profile, the small violation of an intuitive category. Reason has a function, justification and evaluation in company. Sacred values, on Atran’s account, form a real kind of commitment that resists exchange. Each claim says what something is at bottom, and the set treats such claims as the proper aim of a science of man.
The strength and the exposure share an address. They promise a causal account of culture, which wins them respect among scientists and scorn from interpretive anthropologists who think they crush meaning down into cognition. Whether attraction beats selection, whether the mind’s parts are as fixed as the program asserts, whether reason’s purpose is argument, all stay open. The set has built durable theory in pragmatics and a serious research line in cultural evolution. It has not yet written the full naturalized science of culture it set out to produce.
Vigilance: The Hero System of Dan Sperber
A Dorze man in southern Ethiopia knows two things about leopards. He knows the leopard keeps the Coptic fasts, Wednesday and Friday and the long stretch before Easter, because the leopard is a Christian animal. He also knows the leopard will take his goats on a fast day as fast as on any other, and so he watches the herd on Wednesday and Friday with the same care he gives the rest of the week. Both claims live in the one man at the one time. Neither troubles the other. Dan Sperber (b. 1942) spent a year among the Dorze and came home with the puzzle that held him for sixty years. The man guards his goats well. He reasons about rain and cattle and debt as well as anyone. So what is the leopard doing in his head, and why does it cost him nothing to keep it there beside its own contradiction?
Ernest Becker (1924-1974) gives a way to read what the puzzle did to the man who could not put it down. Becker says a culture is a hero system. It hands each man a script for how to be of lasting worth, how to count, how to earn a place that death cannot cancel. The soldier dies for the flag and lives in the regiment’s memory. The mother pours herself into children who carry her forward. The scholar adds a brick to a wall that will stand after him. Under every script runs the same dread: that a man is an animal who rots, and that his rot means nothing. The hero system answers the dread. It tells him his life is an act in a drama large enough to outlast his body. Take the script away and the dread comes back raw.
Sperber’s script is the rarest kind, because his hero is the man who is not fooled.
Look at where he stands. His father, the novelist Manès Sperber (1905-1984), gave his heart to the Communist Party and then tore himself out of it and wrote for the rest of his life about what it is to believe with your whole soul and be wrong. The son grew up next to a man who had been captured and had cut himself free, and who carried the scar. The puzzle the son chose, how do rational men come to hold mistaken beliefs, is the father’s wound turned into a research program. Manès paid for his clarity in grief. Dan set out to build the science of how the capture happens, so that he might see the next one coming. A boy raised beside a man who escaped a faith does not want to spend his one life inside a faith he cannot see. He wants the apparatus that tells him which of his own beliefs are leopards.
So he builds it. Against the structuralists he argues that people do not decode symbols, they process them. Against Clifford Geertz (1926-2006) and the interpreters he argues that the anthropologist who reads meaning into a rite often reads in his own voice and calls it the native’s. Against the meme he argues that minds copy nothing cleanly, that every act of passing a belief along is a fresh reconstruction, and that the beliefs which survive are the catchy ones, the ones that fit the grooves the mind already has. Religion lasts because it sits well in the head, not because the head is a good copier. In Relevance: Communication and Cognition he and Deirdre Wilson (b. 1941) argue that a hearer reaches for the reading that pays the most for the least effort. In The Enigma of Reason he and Hugo Mercier (b. 1974) argue that reason did not grow to find truth on its own. Reason grew to make arguments and to test the arguments of others. Alone, a man reasons his way deeper into what he already wants. In a group that pushes back, reason works.
Sperber spends his career showing that the beliefs men cherish are catchy reconstructions, that the reason men trust runs crooked in solitude, that the meanings men decode are often their own faces in the water. He is the great deflater of other men’s certainties. And the deflater needs a hero too. His is the man who stands at the population level and watches the beliefs spread like weather and does not catch them. The man who knows the leopard is in his own head and can name it. The unfooled man.
Sperber presents this work as a clearing. He says he has subtracted the errors, the interpreter’s projection, the structuralist’s hidden grammar, the meme’s false fidelity, and that what remains is the bare cognitive truth of how minds handle information. Take away the priests and the poets and the wishful copiers and you are left with the architecture. The architecture is not a creed. It is what is there when the creeds are gone.
That is the oldest move in the deflationary book. The man who subtracts is sure that his subtraction reaches bottom and stops at the real. He has faith that the mind’s machinery is the floor and that meaning is the froth on top. He has built a place to stand that he insists is not a place but the absence of one. Becker would say is a hero system that denies it is a hero system, which is the strongest kind, because you cannot mourn a faith you have told yourself you never had. Look at the prizes and the named lectures stacked through his life, the medal and the chair and the institute. A man does not amass that to be forgotten. The brick goes in the wall. The work outlasts the body. The denial of the immortality project is run by an immortality project.
Sperber followed the puzzle when it stopped being fashionable, turned against his own first book when the evidence turned, and built where most men only knock down.
Now take a single word and watch it change shape as it crosses from his hero system into others. Take reason.
For Sperber reason is a social tool, sharp in company and dull alone, vindicated in the room where men hand each other arguments and tear at them. Two minds checking each other catch what one mind misses. The lone thinker is the dangerous one. Reason for him is the friction between heads.
His ancestors held a different word under the same spelling. The men he descends from sat in pairs over the law, one putting the case and one breaking it, machloket for the sake of Heaven, the argument a form of worship and the pair, the chavruta, the place where the mind got honest. Mercier’s claim that reason wakes up in twos is the yeshiva’s claim with the prayer taken out. For the scholar of the law the friction is holy. For Sperber the friction is a tool that happens to work. The word sits in the same groove. The reverence drains out.
For his father reason was the blade that cut him out of the Party and left him bleeding. Reason for Manès was the thing that cost you your comrades, your certainty, your place in the march of history, and gave you in return only the truth and the loneliness of holding it. The son inherited the blade and not the wound, and built it into an instrument he could use without paying the price in the same coin.
For the Dorze man the leopard claim never enters the court where reason rules at all. He is not making a bet on how the world is. He is holding a thing his people hold, in the register where such things live, and his goats are safe because in the other register, the one that touches his livelihood, he knows leopards do not fast. The word reason, swung at his leopard, hits nothing, because he was never offering the leopard as a claim about leopards.
For the engineer building a machine that talks, reason is a function to implement and relevance is a cost to minimize, and Sperber’s pragmatics arrives as a spec. The mind is a thing to rebuild in silicon. The word loses its body and becomes an objective to optimize.
And for one more man the word means something Sperber’s whole life works against.
Set him beside a man whose hero system is the tribe. Call him the covenantal man, the one who holds the faith of his fathers, the land, the dead behind him and the unborn ahead, the people whose name he carries and means to pass on. For this man a belief is not a proposition he audits for accuracy. A belief is a loyalty. The myth of his people is not a catchy reconstruction with poor fidelity. It is the cord that ties him to the men who died holding it and binds him to the children who will hold it after he is gone. Reason, to him, is the servant of the people, not the judge of them. A reason that dissolves the covenant is not reason doing its work. It is reason gone feral, a son turning the family blade on the family.
These two cannot hear each other, and Becker says why. Each man is built against a terror, and each man’s safety is the other’s nightmare.
Sperber is built against two terrors at once, and his hero stands in the narrow band between them. On one side is the terror of the credulous crowd, the dread of being one of the fooled, of dying inside a belief that fit his brain but not the world, indistinguishable from the millions who hold their leopards and never know it. On the other side is the terror of the man alone, the solitary reasoner who talks himself into anything because no one is there to push back, who mistakes the warmth of his own conviction for the light of truth. The credulous tribe on one hand, the self-deceiving hermit on the other. His hero, the small adversarial company of minds that argue and catch each other’s errors, is the one safe ground between the crowd and the cell. Enough company to escape the hermit’s drift. Enough friction to escape the crowd’s fusion.
The covenantal man lives in the place Sperber dreads. To Sperber the tribe fused around its inherited faith is the credulous crowd, the leopard held by millions. To the covenantal man that same fusion is the only thing that carries a man across death, the people who keep his name when his body is gone. What looks to Sperber like capture looks to the covenantal man like home. And what looks to Sperber like freedom, the lone vigilant mind that will not be fooled, looks to the covenantal man like the worst death there is, the deracinated atom with no people behind him and no one to say the prayers when he dies. Each man has built his shelter against the other’s idea of safety. The covenantal man hears Sperber say that his God persists because it is catchy and he hears a son turning the blade on the fathers. Sperber hears the covenantal man defend the inherited faith and he hears his own father in 1932, about to be captured, about to spend twenty years getting free.
How aware is Sperber of the trade he has made? More than most. He knows reason fails the lone man, and he has said so in print, and that is a rare honesty. He turns the deflation on reason itself, which is more than the structuralists or the interpreters ever managed. But the apparatus has a corner it cannot light, and the corner holds his whole position up.
His theory says reason is a tool for argument, not a tracker of truth, except in the adversarial group, where it works. He needs that exception. Without it The Enigma of Reason is only another catchy reconstruction, an attractor that fits the grooves of a certain kind of academic head, with no claim on anyone outside the groove. He needs one room in the house where reason really reaches the world, and that room is the community of scientists arguing in good faith, his own room. The exception is not deflated. It cannot be, because everything else stands on it. The man who showed that men believe what fits their heads must believe, on something close to faith, that his own head and the heads around his table are the one place where fit and truth line up. That is the leopard he guards least.
His apparatus can price a belief by its fit, its catchiness, its fidelity in transmission, its relevance to a hearer. It cannot price what the covenantal man receives, which is to be carried. To be held by a people who do not reconstruct you into noise, who keep your name on purpose, who stand at your grave because you stood at theirs. Sperber can explain why men want that and why the wanting spreads. He cannot, inside the frame, give a man that thing, and his frame contains no slot for the man who chooses to be wrong with his people over right and alone. The unfooled man buys his clearing at the price of the cord. He sees the leopard for what it is. He stands, like his father after the Party, in the truth and the cold of it.
Three coordinates, then, for the man. His hero is the small company of minds that argue and catch each other, the one safe ground between the crowd that fuses and the hermit who fools himself. The rival he fights without naming is the covenantal man, whom he meets only as irrationality or bad epistemics, never as a brother heroism built against a terror as old as his own. And the one cost his ledger cannot price is the cord itself, the carrying, the thing his father lost when he chose the truth, the thing the science of belief can map from the outside and never hand to the man who built it.
The Voice
He is French and English is his second language, but the register stays clean and unfussy. He keeps the jargon for the titles on the page, the epidemiology of representations and the rest, and talks in kitchen words.
His basic move is to set up the received view first and then turn it. He gives you the classical picture of reason, the one where reason lifts man above the animals and lets the lone thinker reach truth on his own, and grants it its full weight before he takes it apart. In a 2013 interview in Budapest he laid out that classical view and then offered the social alternative he built with Hugo Mercier. He persuades by reframing. The signature turn is the one where a flaw becomes a feature: confirmation bias looks like a defect if reason hunts for truth alone, and looks like good design once you take reason as an advocate’s tool. He hands you the puzzle, then the answer.
He thinks in cases. Ask him anything abstract and he reaches for a homely scene. A plane full of strangers who stay polite where chimpanzees would maim each other, the real estate agent who lies about the flat and keeps his job, the academic whose career ends the day someone shows he faked his data, the Dorze man who puts butter on his head because his fathers did. He piles up examples until the idea holds.
And he stages them. He quotes speech, builds small scenes, watches status. The seminar where he handed his book to his colleagues and not one of them mentioned it, until a single man pulled him aside, away from the others, and whispered that he liked it. The dream among the Dorze that told him to listen to his informants and stop pushing the wrong question at them. Tom Wolfe would know the method.
He hedges. In a way, of course, more or less, I’m not saying, on the whole. He marks the edge of every claim. He concedes ground and credits collaborators by name, Deirdre Wilson and Mercier and Nicolas Baumard, and he tells you where he parts from his own former student. He says of Maurice Bloch that each man has failed to convince the other and the conversation runs on anyway.
The humor runs dry and self-aimed. He notes that his French sales peaked with his first structuralist book in the sixties and have fallen ever since. He calls Descartes’s project, the man rebuilding all knowledge alone from nothing, a little crazy.
He describes the sacred in the same even voice he gives goats and flats and office life. The leopard that keeps the Coptic fast and the rite with butter get the same tone. The naturalist takes belief as weather, a thing to explain, and the explaining voice never lifts.
His theory says reason is the advocate, biased when it produces arguments, sharp only when a second mind pushes back. His own manner is the other pole of that theory, the fair evaluator he describes, the man who weighs the case he is handed and gives ground. He talks the way he says reason works at its best, in company, against friction, with the other man’s argument granted its due. He practices in person the cooperation he explains on the page.
