Your taste for fat is doing exactly what selection built it to do. The fact that doing exactly what selection built it to do is now slowly killing you is not a malfunction. It is a mismatch.
For most of the millions of years that produced human metabolism, fat was rare and unreliable. People who craved it ate more of it when it appeared, stored more of it as energy reserve, survived more lean periods, and had more children. Your tongue, your dopamine reward circuits, the gut-brain axis that signals satiety — all of it was calibrated to chase a resource that was hard to find. The calibration worked. You are descended from people in whom it worked.
Then someone invented the cheeseburger. Or the doughnut. Or the milkshake, the pizza, the family-size bag of chips. None of these were in the EEA. They are concentrated, cheap, available everywhere, formulated to ring every bell the calibrated taste system was built to ring. The mechanism cannot tell that the environment changed. The mechanism is doing what it was designed to do, in a context where doing what it was designed to do produces clogged arteries, type 2 diabetes, and a heart attack at 55. That gap — between what the mechanism was tuned for and what the mechanism is now firing on — is evolutionary mismatch.1
Buss has a phrase for it that is now the canonical compact formulation: "we carry around a Stone Age brain in a modern environment."2 The brain in the phrase is metaphorical for the whole evolved psychological apparatus — fears, hungers, mating preferences, status sensitivities, kinship attachments, social-comparison machinery. The modern environment is what the brain wakes up in every morning. Mismatch is everywhere there's a gap between the two, and the gap is wide.
Mismatch follows from two ordinary facts about how selection works.
The first fact: selection is slow. Evolutionary change requires thousands of generations of recurrent selection pressure. Useful new variants arise rarely, spread through populations gradually, and cumulate over deep time. A mechanism shaped over a million years cannot retune itself in a hundred. Modern environments — the ones produced by agriculture, cities, industry, mass media, networked computing — have arrived too recently for selection to have done much retuning at all. Buss calls this the evolutionary time lag.3
The second fact: adaptations carry costs. Selection does not produce optimally designed mechanisms. Selection produces mechanisms whose benefits exceed their costs on average across recurrent activations.4 The fear-of-snakes mechanism could in principle be set so high that humans never went outdoors; that would prevent all snake bites but would also prevent foraging, hunting, mating, and survival. The actual setting is a compromise — high enough to produce caution, not so high that it disables the rest of life. The compromise is calibrated against a particular distribution of risks. Change the distribution and the calibration becomes either too lax or too strict.
Combine the two facts. Mechanisms calibrated against ancestral distributions of risks and rewards are sitting inside organisms whose current environments have changed the distributions. The fear distribution shifted: snakes became rare, cars became deadly. The hunger distribution shifted: calories became abundant, scarcity became rare. The status distribution shifted: visible peer comparisons jumped from fifty band-members to seven billion strangers. The mating distribution shifted: stable pair-bonds with one well-known partner gave way to optional comparison against thousands of curated images. Each shift is a mismatch waiting to fire.
Three patterns of mismatch follow.
Type one mismatch is when an adaptation fires on inputs that are now harmful, even though the inputs match what the adaptation was tuned to detect. Fat-craving is one. Sugar-craving is another. The mechanism reads the input correctly. The input is now a hazard. The result is overconsumption that the EEA didn't punish but the modern environment does.5
Type two mismatch is when an adaptation does not fire on inputs that are now harmful, because the inputs don't match what the adaptation was tuned to detect. Children fail to develop fear of cars or electrical outlets despite their lethality.6 The fear-acquisition system is looking for slither and predator-shape; it does not respond to chrome and copper. The result is danger that goes unrecognized until experience produces it the hard way.
Type three mismatch is when an adaptation fires on inputs engineered to mimic ancestral cues, even though the cues are now decoupled from what they originally tracked. This is the supernormal-stimulus pattern. Modern advertising, dating apps, fast food, casinos, and pornography all work by producing cue-rich stimuli that fire mechanisms harder than the ancestral inputs ever did. The mechanisms are doing what they were built to do, on inputs designed to maximize their activation.7
Each pattern produces different downstream pathology. Type-one mismatches produce slow-onset chronic problems (obesity, diabetes, heart disease). Type-two mismatches produce avoidable accidents (car crashes, electrical injuries). Type-three mismatches produce compulsive consumption and felt suffering (addiction, eating disorders, status anxiety, social-media-driven loneliness).
Take obesity as a worked case.5 Cross-cultural data show that traditional hunter-gatherer populations — Hadza, !Kung, Ache — have low rates of obesity despite consuming whatever caloric food they can get. The mechanism is the same; the environment provides different inputs. When the same populations are introduced to modern food environments through colonization, urbanization, or relocation, obesity rates climb rapidly. The Pima Indians of Arizona, who experienced a rapid environmental shift from traditional desert agriculture to industrial food access, now have one of the highest rates of obesity and type 2 diabetes in the world. Their genes did not change. Their food did. The mismatch fired and continues firing.
Take phobia distribution as a second case.6 The most common clinical phobias are snakes, spiders, heights, darkness, blood, and small animals. None of these kills many modern Americans. Cars, falls in the home, prescription drug interactions, refined sugar, and stationary lifestyles kill millions. Almost no one has clinical phobias of these. The fear-acquisition system is looking for ancestral hazard cues — slither, fang, fall, blood, ambush — and it doesn't read modern hazard cues. The mismatch produces a population that is simultaneously over-afraid of statistically minor risks and under-afraid of statistically major ones.
Take mate-value contrast effects as a third case.8 Kenrick and colleagues showed men photographs of either highly attractive women or women of average attractiveness, then asked them to evaluate their commitment to their current romantic partners. Men who had viewed the highly attractive images rated their actual partners as less attractive and rated themselves as less committed, less satisfied, less close than men who had viewed the average images. The relevant mechanism is mate-value comparison — a system tuned in the EEA to compare a current partner against the few dozen potential alternatives in a band. The system is now being fed thousands of curated highlight reels from people who exist only as images. The mechanism does what it was built to do — compares, ranks, updates expected mate value — and produces dissatisfaction with reality calibrated against an impossible distribution.
Take eating disorders as a fourth case.8 Buss documents how the same contrast-effect mechanism, applied to women's intrasexual competition for mate value, helps explain rising rates of anorexia and bulimia in modern environments. Women's competition for mate value runs partly on physical attractiveness, calibrated against local rivals. Modern media broadcasts thousands of professionally produced images of extreme thinness as the mate-value standard. The competitive mechanism reads these as the local distribution and recalibrates accordingly. Women's bodies cannot in fact reach those proportions through normal eating. Compulsive restriction follows. The mechanism is not pathological. It is the same mate-value-calibration mechanism that produced normal selectivity in the EEA, fed inputs the EEA didn't have.
A fifth case is loneliness in a connected world. Humans evolved in small groups of fifty to a hundred fifty people whose relationships persisted over years and decades. The coalition-tracking mechanisms are calibrated for recurring relationships with reliable kin and allies. Modern cities, modern workplaces, and modern social networks expose people to thousands of one-shot interactions and curated personas. The mechanism reads the volume of contact correctly — many people — but cannot find the recurrence and reliability the mechanism was tuned to detect. The output is a felt sense of being unknown despite being seen, of being alone in a crowd. The mechanism is firing correctly. The input does not contain the structure the mechanism was looking for.
A sixth case is contained in the basic finding that the modern environment can only produce some kinds of suffering through specific adaptations responding to specific changed inputs. Buss's textbook treats this systematically across mating, parenting, kinship, cooperation, aggression, and status chapters. Each chapter ends with examples of how the chapter's mechanisms fire on modern inputs in ways that produce mismatch pathology. The volume of cases is part of the case for the framework. Mismatch is not one phenomenon. Mismatch is the general pattern produced by recent environmental change running over slow-changing equipment.
The hardest tension in mismatch theory is empirical specification. To diagnose a behavior or a felt state as mismatch, you need to identify the underlying adaptation, identify the EEA conditions it was calibrated against, identify the changed inputs in the modern environment, and show that the gap produces the observed pathology. Each step is contestable. Each step requires reconstruction of conditions we cannot directly observe.
A second tension is selection bias in mismatch claims. Researchers tend to identify mismatches that fit familiar pathologies and ignore mismatches that don't. Modern obesity gets the mismatch story; modern infant survival rates do not. Modern anxiety gets the mismatch story; modern absence of routine warfare violence does not. The framework is real, but its application is asymmetric — it tends to diagnose modern problems and not modern improvements. A more honest mismatch analysis would catalog both, and would recognize that some modern environments are better matches to ancestral mechanisms than the EEA was. Modern infant nutrition is better than ancestral infant nutrition. Modern medicine matches the body's repair mechanisms better than ancestral environments did. A balanced mismatch frame includes the wins.
A third tension is recent evolution. The conventional mismatch picture treats human equipment as Pleistocene-calibrated and modern environments as too recent for selection to have updated the equipment. Hawks and colleagues have shown that human adaptive evolution accelerated over the past 40,000 years, especially in the Holocene.9 Lactase persistence in dairy-farming populations is a clean case of recent adaptation. The mismatch frame holds, but the time horizons need calibration per adaptation. Some adaptations are still in the Pleistocene. Others are catching up.
A fourth tension is normative. If a felt state is diagnosed as mismatch, what follows? Some readers take the diagnosis as license to ignore the felt state ("it's just my Stone Age brain misfiring"). Others take it as a call to remove the modern stimulus ("I should not look at Instagram"). Neither follows automatically. The diagnosis explains where the suffering comes from. It does not prescribe what to do about it. Some mismatches are best handled by changing the environment (closing the social media app). Others are best handled by training the mechanism (cognitive reframing of the comparisons). Others are best handled by accepting the felt state as part of being human (loneliness in cities will not fully go away). The framework supplies the diagnosis; the response is a separate question.
A fifth tension is whether "mismatch" is the right frame at all. Some critics argue that humans are an exceptionally flexible species precisely because our mechanisms are not narrowly tuned to the EEA — that we have always lived in changing environments and have built-in plasticity to handle them. The strongest version of this position treats "mismatch" as a category error: the EEA itself was not stable, so there is no single calibration to mismatch against. Buss's response is that even with environmental variation in the EEA, the mechanisms were calibrated against statistical patterns that bear little resemblance to modern industrial conditions. The patterns matter even if they were not constant. The dispute reduces to how variable the EEA actually was, and the empirical answer is: variable, but bounded, and much less variable than the post-1800 environment.
Tooby and Cosmides supplied the formal framing. The EEA-mismatch concept is implicit in their 1992 framework: every adaptation is calibrated to a particular EEA, every modern environment is some distance from that EEA, and the distance produces mismatch. Buss treats this framing as standard EP machinery.3 Where Buss adds emphasis is on the practical implications. He uses mismatch reasoning throughout the textbook to explain modern pathologies that proximate-cause psychology cannot fully account for. Obesity gets a mismatch story. Phobia distribution gets a mismatch story. Mate-value contrast effects get a mismatch story. Eating disorders get a mismatch story.
Where Buss is silent — and where the vault has work to do — is on the asymmetry of mismatch diagnosis. Modern environments produce both more pathology than the EEA in some dimensions and less in others. Modern lives are longer, healthier in many ways, less violent, with better infant survival. The textbook does not foreground these wins. The reader is left with a narrative in which mismatch produces mostly costs, when in fact the modern environment is also a better fit for some mechanisms than the EEA was. A complete mismatch analysis is harder to write than a partial one because it has to track the gains alongside the losses.
A more substantive disagreement runs between the conventional EP picture and the recent-evolution picture. Hawks, Cochran, and Harpending's work on accelerated Holocene evolution suggests that some adaptations have been updating in the past 10,000 years.9 The mismatch frame still holds, but the gap between EEA calibration and current calibration is smaller for some mechanisms than the textbook framing implies. Buss reports the Hawks finding without integrating it into the rest of his mismatch reasoning. The reader who takes both findings seriously ends up with a more nuanced picture: mismatch is real, but the time horizon over which it operates differs by mechanism, and some mechanisms have already partly retuned to recent environments.
A third tension runs to the broader literature on mismatch — Lieberman's Story of the Human Body, Konner's The Tangled Wing, Nesse's Darwinian medicine framework. All converge on the basic claim: most modern chronic illness is mismatch pathology. Where they differ is in scope. Lieberman emphasizes physical mismatch (obesity, diabetes, myopia, sedentism). Nesse emphasizes psychological mismatch (anxiety, depression, addiction). Buss spans both with more emphasis on psychological mechanisms. The vault implication: mismatch is not narrowly an EP concept. It's a general framework that any discipline studying the gap between evolved equipment and current environments uses, and the EP version is one application among several.
Behavioral-mechanics, looked at through the mismatch frame, is the systematic engineering of type-three mismatches — the deliberate construction of cue-rich stimuli that fire ancestral mechanisms harder than the ancestral inputs ever did. Almost every modern persuasion technique can be redescribed this way. Advertising imagery exploits visual-attention mechanisms that were tuned to detect ripe fruit, fertile mates, and threatening predators by serving up amplified versions of these inputs. Dating apps exploit mate-evaluation mechanisms that were tuned for tens of potential partners by serving up thousands. Casino floors exploit reward-prediction mechanisms tuned for variable foraging success by engineering near-miss frequencies that maximize dopamine release without maximizing actual rewards. Fast food exploits taste-preference mechanisms tuned for caloric scarcity by engineering hyperdense fat-sugar-salt combinations.
The connection to existing BM pages: the manipulation and influence hub catalogs hundreds of techniques that look heterogeneous on the surface but become coherent when re-read as mismatch engineering. Each technique presents a cue the target's evolved mechanism reads as ancestral, when in fact the cue has been engineered. The technique works on the EEA's mechanism in the modern environment. The mechanism cannot tell that the cue has been engineered.
The insight neither domain generates alone: most modern manipulation is not a new attack on a vulnerable system. It is a familiar attack — kin recognition, status anxiety, threat detection, reward prediction — that the EEA mechanism was never built to defend against in this volume. The EEA didn't have ad agencies, A/B testing, attention engineers, dating algorithms, or processed-food chemists. The mechanism doesn't have defenses against tools it never met. Mismatch is the substrate. Manipulation is the engineering on top of it. Defense against manipulation requires not just willpower but environmental control — keeping the engineered cues out of the mechanism's input range. The vault's BM pages get sharper when each technique is annotated with which mismatch pattern it exploits.
A second handshake runs to history and the long arc of environmental change. Three transitions stand out as mismatch events: the agricultural revolution (~10,000 years ago), the industrial revolution (~250 years ago), and the digital revolution (~30 years ago). Each one introduced a wave of new conditions that EEA-calibrated machinery had not been selected against. Each produced specific new pathologies and reduced other pathologies in turn.
The agricultural revolution introduced sedentism, surplus calories, hierarchical bureaucracies, dense settlement, and continuous exposure to domesticated-animal pathogens. Diamond's Guns, Germs, and Steel documents the resulting health collapse — agricultural populations were shorter, sicker, and lived shorter lives than the hunter-gatherers they replaced for the first millennia after the transition. Eventual gains came through immune-system co-evolution and improved farming. The mismatch was real and consequential.10
The industrial revolution introduced wage labor, urbanization, separation of work from home, mass-produced food, mass media, and continuous artificial light. Each of these put new pressure on mechanisms tuned for different conditions. Sleep mechanisms tuned for sunset-to-sunrise darkness now operate in environments with continuous light. Reward mechanisms tuned for episodic foraging success now operate in environments with continuous purchasable indulgence. The pathology profile shifted.
The digital revolution introduced infinite-scroll attention environments, instant feedback at planetary scale, and the decoupling of social comparison from local groups. Buss's mate-value contrast effect work was published before social media's full development; the mechanism it described has only become more relevant since 2014.8
The connection to the geographic-historical determinism hub and to history-domain pages on industrialization: large-scale historical change is partly a story of mismatch accumulation. Each transition adds environmental novelty without removing previous novelty. The cumulative gap between EEA conditions and current conditions has been growing for ten thousand years, and the rate is accelerating. The insight neither domain generates alone: the felt experience of "modernity is broken" maps onto specific mismatch patterns produced by specific historical transitions, and the diagnosis is testable rather than rhetorical. History's "great transitions" are not just changes in productive capacity. They are changes in the gap between human equipment and human environment, and the gap predicts which kinds of suffering and which kinds of social pathology each transition produced.
The Sharpest Implication.
A category of suffering you experience as a verdict on your life is in fact a mismatch artifact. The status anxiety you feel scrolling Instagram is not a signal that your status is bad. It is a status-tracking mechanism designed to monitor a few dozen rivals in a small band, firing in an environment where you can see thousands of curated highlights from people you'll never meet. The chronic hunger you feel after a balanced meal is not a signal that you're undernourished. It is a calorie-acquisition mechanism designed to motivate foraging in an environment where calories were uncertain, firing in a kitchen where they aren't. The loneliness you feel surrounded by people on a subway is not a signal that you have no friends. It is a coalition-monitoring mechanism designed to track stable kin and ally relationships, firing in an environment where the people you see don't recur and don't know your name.
This does not make the suffering less real. It does change what the suffering means. It is not your verdict on yourself. It is your equipment misfiring on inputs it cannot read.
The corollary cuts the other way. Many of the things that genuinely indicate flourishing in modern life don't trigger the felt-good signals you'd expect. Your friend group did not save you from a leopard last week. Your stable career does not register as the kind of resource accumulation your status mechanisms were tuned to track. Your meaningful work does not feel as rewarding as a single notification of approval from a stranger online. The mechanism that should be lighting up isn't, because the input doesn't match what the mechanism was watching for. The mechanism that lights up easily is responding to engineered cues. Reality and felt-reality have decoupled, and the decoupling is mismatch.
The work — and a lot of contemplative and therapeutic traditions are doing this work in different vocabularies — is to learn which of your felt signals are reading current inputs accurately and which are mismatch noise. The practice is not to suppress the signals. The practice is to read them with the right frame. Body-craving says "fat-rich food is rare" when fat-rich food is everywhere. Loneliness says "no stable allies present" when stable allies are present but happen to be invisible to a mechanism tuned for face-to-face recurrence. Status anxiety says "rivals are gaining" when the "rivals" are strangers whose lives you cannot actually compete in.
Generative Questions.
If a felt experience is most likely a mismatch artifact, what evidence would distinguish that case from one in which the felt experience is reading the modern environment correctly? Some signals are correctly read — physical pain from injury, hunger after no food. Others are mismatch — felt loneliness in a busy city, status anxiety scrolling social media. Is there a methodology for diagnosing which is which, or is it always retrospective?
The mismatch frame implies a kind of "natural" living that is incoherent. The conditions a mechanism was tuned against no longer exist and were never a stable thing to return to. What replaces "natural" as a normative anchor in arguments about how humans should live, given that the conditions our equipment was built for are gone?
Different individuals have different mismatch profiles. Two people in the same modern environment can have the same felt status anxiety, but one might be running a high-baseline status-comparison mechanism and the other might be in a particularly comparison-rich micro-environment. Is there a personalized mismatch diagnostic that would let people identify which mechanisms are misfiring most in their specific lives?