The awkward thing about studying great apes is that the questions are never really about them.
Nobody funds a chimpanzee cognition laboratory to find out what chimpanzees are like. They fund it to find out what we are like, by subtraction. The whole enterprise runs on a comparison in which one species is the reference standard and the other four are measured against it, which means the research program has a structural bias built into its foundations: whenever an ape turns out to do something previously considered uniquely human, the usual response is not to revise the picture of humans but to redefine the capacity so that it remains uniquely human.
Tool use was the line, until Jane Goodall watched a chimpanzee strip a twig and fish for termites in 1960 and Louis Leakey observed that we would now have to redefine tool, redefine man, or accept chimpanzees as human. Culture was the line, until between-population behavioral differences turned out to be extensive. Theory of mind was the line, until an eye-tracking result in 2016. Each retreat has been orderly and each has been accompanied by a reformulation that preserves the boundary somewhere slightly further back.
The interesting thing about great ape cognition is not the individual capacities. It is that the boundary has moved this many times and the people moving it are the same people who keep insisting it is real. What follows is where the line currently sits, what the recent evidence actually establishes, and which of the retreats were justified.
Who is actually in the great ape cognition comparison
The living great apes are chimpanzees, bonobos, gorillas, orangutans, and us, and the phylogeny matters for reading everything downstream.
Chimpanzees and bonobos are our closest living relatives, sharing a common ancestor with humans somewhere in the range of six to eight million years ago, and they split from each other far more recently, likely under two million years ago, apparently when the Congo River formed and separated populations. Gorillas branched earlier, around eight to ten million years ago, and incomplete lineage sorting means a meaningful fraction of the human genome is actually closer to gorilla than to chimpanzee at particular loci, which is a useful antidote to treating any single similarity figure as meaningful. Orangutans are the outgroup, diverging perhaps twelve to sixteen million years ago and the only great apes outside Africa.
That structure produces a specific inferential tool. If a capacity appears in chimpanzees and in humans, it might be shared inheritance or convergence. If it appears in chimpanzees, bonobos, gorillas, orangutans, and humans, the most economical explanation is that the common ancestor of all of them had it, which pushes the origin back well past ten million years and makes it a great ape trait rather than a human one that apes happen to approximate.
The species differences matter as much as the similarities. Chimpanzees live in fission-fusion communities with male philopatry, marked status competition, cooperative hunting, and lethal intergroup aggression that has been documented across multiple long-term field sites. Bonobos live in communities with far less severe aggression, no confirmed lethal intergroup killing, female social bonds that constrain male behavior, and a great deal of sexual behavior deployed in non-reproductive social contexts. Two species, nearly identical genetically, separated by a river, running incompatible social systems.
The leading explanation for that divergence is ecological rather than mysterious. Bonobos live south of the Congo River in habitat with no gorillas competing for terrestrial herbaceous vegetation and with more reliable year-round food, which reduces the payoff to contest competition and permits females to travel together. Females traveling together can form coalitions. Coalitions constrain male aggression. Once female bonds are the organizing structure, selective pressure on male size, aggression, and coalitionary violence drops away. A river changed a food distribution, a food distribution changed a social system, and a social system changed a mind. That causal chain is the best-documented case in great ape cognition of ecology producing psychology.
Gorillas live in stable groups with a single dominant silverback and are the most folivorous, which shapes their foraging cognition. Orangutans are semi-solitary, the least social of the group, and the only great apes where a substantial fraction of adult life is spent alone. That last point is the one worth flagging early, because it is a natural experiment on whether sociality is the engine of ape cognition, and orangutans keep performing well on tasks the social intelligence hypothesis says they should fail.
The ecological alternative is worth stating because it competes directly. Orangutans are extractive foragers in a habitat with severe fruit unpredictability, requiring them to remember where widely dispersed trees are, when each fruits, and how to open a large number of protected foods. That is a cognitive load with no social component, and it predicts exactly the profile orangutans show: strong physical problem-solving, excellent spatial memory, high innovation rates, and comparatively modest social cognition. Great ape cognition therefore has at least two candidate engines running in different species, and the field has not cleanly separated them.
Theory of mind, and the result that moved the line
Whether any non-human animal understands that others have mental states is the longest-running argument in comparative psychology, and it started with apes. A 1978 paper asked whether the chimpanzee has a theory of mind, generating four decades of experiments and a persistent negative result on the crucial test.
The crucial test is false belief. Understanding that another individual can hold a belief that is wrong, and that they will act on the wrong belief rather than on reality, requires representing someone else’s mind as a thing separate from the world. Human children pass explicit versions around age four. Apes failed every version for decades.
The 2016 result changed that by changing the measure. Rather than asking apes to perform an action, researchers borrowed the anticipatory looking paradigm developed for pre-verbal human infants: show a scenario on video and track where the subject looks in anticipation. In the scenarios, a person watched an object being hidden, then the object was moved while the person was away. When the person returned, chimpanzees, bonobos, and orangutans looked toward the location where the person falsely believed the object to be, rather than where the object actually was.
The design detail that made it work is worth noting. The scenarios were built to be meaningful to apes rather than to children: a human in a gorilla suit stealing an object, competitive interactions over status and food. Apes had been failing tests partly because the tests were boring to them.
The 2025 follow-up pushed it into communication. Working with three bonobos, researchers ran a preregistered study in which a human partner needed to find a hidden food item, and the bonobo could see where it was. When the partner had watched the hiding, the bonobos mostly did nothing. When the partner was ignorant, the bonobos pointed at the correct location, more often and more quickly. The finding that bonobos point more for ignorant than knowledgeable partners is the first evidence that a non-human primate tailors communication to a partner’s knowledge state in order to coordinate.
The caveats deserve equal billing. The sample was three bonobos at a single facility, including Kanzi, an animal with an extraordinarily unusual history of human interaction. The authors themselves flagged the open question: the study shows apes communicate to change a partner’s behavior, not that they are trying to change a partner’s beliefs. And a competing account, submentalizing, holds that anticipatory looking results can be produced by simpler mechanisms tracking behavioral cues rather than mental states, though control experiments have been run against it.
Where this leaves the field is genuinely unsettled, and the honest summary is that apes track what others can and cannot see, use that information flexibly, and pass a test designed for infants, while whether they represent belief as belief remains open.
There is a related capacity where the evidence is older and much less disputed, and it involves deception. Wild chimpanzees suppress copulation calls when a dominant male is nearby, conceal food discoveries from competitors, and take circuitous routes to hidden resources when others are watching. Subordinate individuals in food-competition experiments reliably choose items a dominant cannot see, which requires tracking another individual’s line of sight and acting on it. Tactical deception is not a laboratory artifact in this group, and it is worth noting that the capacity showed up in competitive contexts decades before it showed up in cooperative ones. Apes read minds most readily when there is something to be gained by it.
Gestures, and a vocabulary we can mostly read
The communication work has produced the most surprising recent finding in great ape cognition, and it concerns us rather than them.
Wild chimpanzees use a repertoire of roughly seventy to eighty distinct gestures, produced intentionally in the technical sense: directed at a specific recipient, adjusted for whether that recipient is looking, persisted with and elaborated when the first attempt fails. Cataloguing what they mean took years of watching what outcome apparently satisfied the signaller.
Bonobo repertoires overlap with chimpanzee repertoires by roughly ninety percent in physical form, and the meanings overlap substantially as well. Chimpanzee and gorilla repertoires overlap around sixty percent, chimpanzee and orangutan around eighty. Great apes use a small subset, on the order of seventy to ninety, of the thousand-plus gestures that are morphologically possible, and the same small subset keeps appearing across species that separated millions of years ago.
Then the part that reframes it. Researchers built an online task showing video of ape gestures to people with no relevant expertise and asked them to select the meaning. Untrained humans performed well above chance. The demonstration that inexperienced humans understand common nonhuman ape gestures suggests the repertoire is not a foreign language that has to be learned but a signalling system we retain access to, presumably because we inherited it.
Human infants use gestures from that same repertoire before they acquire speech, and drop most of them as language comes online. The reading offered is that the great ape gestural system is ancestral, that humans still have it, and that it is largely obscured in adults by a communication channel that arrived later.
The intentionality criteria used in this work are stricter than most people assume, which is what makes the results carry weight. To count as intentional communication, a gesture must be directed at a specific recipient, produced when that recipient is attending or preceded by an attention-getting behavior, followed by a pause during which the signaller waits for a response, and elaborated or repeated if no satisfactory response arrives. Apes meet all four criteria. Those standards were developed to separate genuine communication from emotional expression and reflex, and they are the same ones applied to pre-linguistic human infants.
The vocal side has moved recently too. A 2025 study reported extensive compositionality in bonobo vocalizations, meaning call combinations whose meaning is derived from the meanings of the components rather than being arbitrary, which is a property long treated as a signature of language. Whether it constitutes compositionality in the linguistically demanding sense is contested, and the appropriate posture is that the gap between ape communication and language is narrower than the textbook version and still real.
Culture, and what the between-site differences show
Behavioral variation between chimpanzee populations is extensive and cannot be explained by genetics or ecology alone, which is the standard criterion for culture. A landmark synthesis pooling data across long-term field sites identified dozens of behaviors present in some communities and absent in others despite ecological availability.
The variation covers tool techniques, grooming postures, courtship displays, and social conventions with no obvious function. The grooming handclasp, where two chimpanzees clasp hands overhead and groom with the free hand, is present at some sites and absent at others, with local variants in exactly how the hands are held, and it does nothing except be the way it is done there. That is arbitrary convention, which is a demanding criterion.
The long-term study populations in the Mahale mountains maintain their own repertoire distinct from neighboring communities, and comparable variation shows up in orangutans across Bornean and Sumatran sites, where distinct populations maintain their own nest-building refinements, tool repertoires, and vocal signals. The birds whose regional song dialects can be mapped between neighborhoods and the cetacean clans whose call types mark group membership are running the same phenomenon on different signal channels, which is the strongest available argument that culture is a general property of social animals with adequate transmission fidelity rather than a primate specialty.
What ape culture appears to lack is cumulative ratcheting. Human technology accumulates: each generation inherits, modifies, and passes on something more complex, and no individual could reinvent it. Chimpanzee traditions are largely stable rather than accumulating. A chimpanzee alive today fishes for termites the way chimpanzees did when Goodall arrived, and there is no evidence anywhere in the record of a technique becoming progressively more elaborate across generations. The cockatoos whose bin-opening technique spread geographically with local variants show the same ceiling: transmission without ratcheting. The proposed reason is a difference in social learning mechanism, with apes relying more on emulation, reproducing an outcome, and humans relying more on imitation, reproducing the method including steps whose purpose is unclear.
That difference shows up in a specific experimental result. Given a demonstration of a puzzle box containing both necessary and unnecessary actions, human children copy everything including the useless steps. Chimpanzees skip the useless steps and go straight to the outcome. This is usually reported as children being irrational, and it is closer to the opposite: high-fidelity copying of methods you do not understand is the mechanism that permits accumulation, and skipping what looks pointless is the thing that prevents it. The macaque troop whose innovation spread through a population without accumulating illustrates the same ceiling in a different primate.
There is a live counter-argument worth registering. Some researchers hold that the imitation-emulation distinction is overdrawn, that chimpanzees imitate when the task rewards it, and that the real constraint on ape cumulative culture is demographic rather than cognitive: small, fragmented populations with limited contact between communities cannot sustain the transmission chains accumulation requires, and any innovation is likely to be lost before it spreads. On that account ape culture is capped by ape population structure rather than by ape minds, which is a claim with uncomfortable implications given how much smaller those populations now are.
Cooperation, fairness, and the limits of ape prosociality
Chimpanzees cooperate, and the shape of that cooperation is informative about what changed in our lineage.
They hunt colobus monkeys in coordinated groups with apparent role differentiation, and meat is subsequently shared, though the sharing is heavily influenced by harassment and by social relationships rather than by anything resembling equity. They form coalitions to contest status, reconcile after conflicts through affiliative contact, console distressed third parties, and recruit specific partners for specific tasks, choosing effective collaborators over ineffective ones.
Where they diverge from us is in the structure of joint action. Human cooperation typically involves a shared goal that both parties represent as shared, with commitment to a joint task and expectations about the partner’s role. Chimpanzee cooperation looks more like parallel individual goals that happen to require another body. A chimpanzee that has obtained its share tends to leave.
The fairness literature is where overclaiming has been most severe. The famous capuchin experiment, in which a monkey rejects a cucumber slice after seeing a neighbor receive a grape, has been enormously influential and is genuinely contested. Alternative explanations include frustration at the visible presence of better food regardless of who receives it, and replication attempts have produced mixed results with some finding the effect requires a social partner and others not. Apes tested in ultimatum-style games generally accept any non-zero offer, which is what a rational self-interested agent does and not what a human does.
The conclusion that holds is narrow: apes are sensitive to what others get and adjust behavior accordingly, and the elaborate norm-enforcement and third-party punishment machinery humans run appears to be ours.
The helping literature runs the same way. Chimpanzees will hand a tool to a conspecific who requests it, and will open a door for an individual trying to reach food, which establishes instrumental helping. They largely will not spontaneously provision food to a partner at any cost to themselves, and they do not reliably choose an option that benefits both over one that benefits only themselves when the two cost the same. The reasonable reading is that apes help when the cost is near zero and the request is explicit, which is a real prosocial capacity with a narrow operating range. The cooperative breeders whose entire social system runs on costly help to non-offspring exceed apes on exactly this axis, which is a useful corrective to any ranking that puts primates at the top by default.
Faces, memory, and individuals across decades
Great apes recognize individual conspecifics and remember them for extraordinary periods.
Work using eye-tracking found that chimpanzees and bonobos looked significantly longer at photographs of former groupmates than at strangers, with the effect detectable for individuals not seen in over twenty-five years, and stronger for individuals with whom the subject had positive relationships. That is social memory persisting across most of a lifespan, in animals that live thirty to fifty years in the wild and considerably longer in captivity. The bowerbirds whose display structures encode accumulated individual effort hold their information outside the body; apes hold it inside, for decades, without rehearsal.
Chimpanzees also show configural face processing, the same holistic mechanism humans use, and they show an inversion effect for conspecific faces, meaning upside-down faces become disproportionately hard to recognize. That is a signature of specialized face machinery rather than general object recognition.
Working memory produced one of the field’s genuinely surprising results. In a task where numerals appear briefly on a touchscreen and are then masked, young chimpanzees at Kyoto have outperformed human adults at recalling the spatial arrangement, with performance holding at presentation durations too brief for humans to manage. Attempts to explain this away as pure training effect have not fully succeeded, though the human comparison samples have been criticized. Whatever the resolution, the default assumption that human cognition dominates on every axis does not survive contact with a chimpanzee doing a rapid spatial memory task.
The proposed explanation is a trade rather than a mystery. On the cognitive tradeoff account, human language acquisition consumed neural resources previously allocated to rapid visuospatial processing, and what looks like a chimpanzee advantage is a human loss. That hypothesis is difficult to test and it fits a pattern visible elsewhere in this subject: capacities are rarely added without something being reallocated, and the animals whose sensory systems were tuned hard toward one channel at the expense of others show the same accounting in a different domain.
The elephants whose social knowledge accumulates in the oldest individuals and the cetacean populations maintaining recognition and affiliation across decades are running the same long-horizon social memory, and the convergence across three unrelated mammal groups with large brains and long lives is itself the argument. The long-lived birds maintaining individual recognition on a fraction of the neural hardware complicate the tidy version, since whatever social memory costs, it evidently does not cost a primate brain.
Brains, genes, and what actually differs
The neuroanatomy is where the comparison gets quantitative, and the numbers are less dramatic than expected in some places and more in others.
A chimpanzee brain runs around three hundred and eighty grams against a human average near thirteen hundred. Neuron counts follow: humans carry roughly eighty-six billion, chimpanzees something under thirty billion. But the scaling relationship between brain size and neuron number is the same in humans as in other primates, which means the human brain is a primate brain of the expected composition for its size rather than an exceptional design. What differs is the size, and behind that, the developmental schedule that produces it.
Prefrontal cortex is proportionally larger in humans, though by less than older estimates claimed, and the more robust differences are in connectivity, in the extent of cortical asymmetry, and in the protracted timeline of human brain development. Human synaptic pruning in prefrontal regions continues into the late twenties, far longer than in chimpanzees, which extends the window during which experience shapes circuitry.
The genetic comparison has become more interesting as it has become more precise. The commonly cited figure of ninety-eight to ninety-nine percent similarity depends heavily on how insertions and deletions are counted, and comparisons including structural variation give lower numbers. More useful than any percentage is the identification of specific regions: human accelerated regions, sequences conserved across mammals that changed rapidly in the human lineage, are enriched for regulatory elements active in neural development. The pattern points at regulation and timing rather than at novel genes.
Organoid work has begun to test this directly, with cerebral organoids grown from human and chimpanzee cells showing differences in the timing of progenitor cell maturation, with human progenitors dividing longer before differentiating, which yields more neurons. That is a timing difference producing a size difference producing a capacity difference, which is a far more tractable story than a search for uniquely human genes.
Metabolism sets the constraint underneath all of it. A human brain consumes roughly twenty percent of resting energy budget against something closer to eight or nine percent in other primates, and the leading account of how that became affordable involves changes in diet quality and in gut size, with the digestive tract shrinking as the brain grew. That is a trade rather than an upgrade, and it is worth remembering whenever great ape cognition gets discussed as though our lineage simply added capability. Something was given up to pay for it.
Self-recognition, death, and what apes appear to understand about themselves
Two lines of evidence bear on whether great apes represent themselves as objects in the world, and both are messier than the summaries suggest.
Mirror self-recognition was first demonstrated in chimpanzees in 1970, using the mark test: apply a dye mark to a place the animal cannot see without a mirror, and observe whether it touches the mark on its own body rather than on the reflection. Chimpanzees, bonobos, and orangutans pass reliably. Gorillas mostly do not, which was long treated as a puzzle and is now generally attributed to gorillas finding direct eye contact aversive, since gorillas raised in unusual circumstances have passed and modified procedures improve performance. That explanation is plausible and also a reminder that any negative result on a test with a social component is difficult to interpret.
What passing the mark test actually establishes is narrower than the popular framing. It demonstrates that an animal recognizes the reflection as itself rather than another individual, which requires a body representation and the ability to update it. It does not establish introspective self-awareness, and the inferential distance between the two is large enough that the test has generated more philosophy than it can support.
The responses to death are harder to categorize and harder to dismiss. Chimpanzee mothers have been observed carrying dead infants for days or weeks, continuing to groom and transport bodies well past decomposition. Group members have been documented sitting quietly with a dying individual, and at one sanctuary an entire group gathered around a dying elderly female, with several individuals attempting to rouse her and the group remaining subdued for days afterward. Chimpanzees have been observed cleaning the teeth of a dead groupmate with tools.
Whether any of this constitutes a concept of death is unresolved and probably unresolvable with current methods. What is not in question is that the behavior is specific to death rather than being general distress, that it varies between individuals in ways that track prior relationships, and that it is inconvenient for anyone who wants the boundary drawn tidily.
The language projects, audited
No part of great ape cognition has generated more heat and less durable evidence than the attempts to teach apes language, and the record deserves an honest accounting.
The chimpanzee Washoe was reported to have acquired a substantial sign vocabulary. Koko the gorilla was reported to have over a thousand signs and became internationally famous. Nim Chimpsky was raised in a sign-language project explicitly designed to test the claims, and the subsequent analysis of the video record by the project’s own director concluded that Nim’s utterances were largely prompted, repetitive, and lacked grammatical structure, with the appearance of conversation produced substantially by the teachers’ cueing.
The methodological problems were serious and general: interpretation by invested researchers, absence of blinded scoring, ambiguity between a sign and an ordinary gesture, and enormous unpublished data. The Koko work in particular was never subjected to the peer-reviewed reporting its fame implied.
It is also worth being clear about why the projects were attempted at all, since the motivation was reasonable even where the execution failed. If apes could acquire something language-like, the question of what makes human cognition distinctive would have a much sharper answer, and if they could not, the boundary would be located precisely. Neither outcome arrived, because the methodology could not support either conclusion, and a genuinely important question was left unanswered for a generation by work that was too eager to answer it.
Kanzi is the strongest case and stands apart for a specific reason. He acquired lexigram use spontaneously while his adoptive mother was being trained, rather than through explicit reward-based instruction, and his comprehension of spoken English was tested with novel sentences under conditions designed to prevent cueing, including requests to do unusual things with familiar objects. He performed at a level compared to a two-and-a-half-year-old child on those comprehension trials. That comprehension result is more robust than the production claims, and comprehension and production are different capacities.
The reasonable summary is that apes can acquire symbol-referent relationships and use them communicatively, that comprehension outstrips production substantially, that nothing in the record demonstrates syntax, and that a large fraction of the popular impression rests on work that would not pass current standards. The field mostly moved on to studying natural communication for exactly this reason, which was the right call.
The episode also left a useful methodological legacy. Nim was specifically designed as a check on claims that had been accepted too readily, and the check worked, at considerable cost to the animals involved and to several careers. Comparative cognition has been more careful about blinded coding, preregistration, and cueing controls ever since, and the recent bonobo pointing work being preregistered is a direct descendant of that correction. The research programs that produced enormous public enthusiasm on thin evidence are a recurring hazard in this field, and the discipline it forced was worth having.
The claims that do not hold up
An audit, since this domain is unusually contaminated by both overclaiming and reflexive dismissal.
Chimpanzees are ninety-nine percent human genetically oversimplifies a figure that depends on methodology, and more importantly implies that percentage similarity predicts phenotypic similarity, which it does not.
Bonobos are peaceful hippie apes flattens a real difference into a fiction. Bonobo aggression exists, including serious wounding, and female coalitions enforce social outcomes with force. The genuine finding is the absence of confirmed lethal intergroup killing and lower overall severity, which is interesting without being pacifism.
Chimpanzee warfare is a human projection was argued for decades on the grounds that observed intergroup killing was an artifact of provisioning by researchers. A large multi-site analysis found the pattern tracks ecological and demographic variables rather than human interference, which settled it against the projection hypothesis.
Koko understood language and discussed her emotions rests on an evidentiary base that does not support it.
Apes cannot understand pointing is the deflationary error running the other direction, and the recent bonobo work indicates they both understand and produce it in knowledge-sensitive ways.
Gorillas are gentle giants and orangutans are the smart ones are folk rankings without support. Cross-species cognitive comparison is task-dependent, and the database work compiling nearly two decades of great ape testing has found domain-specific rather than general differences.
Apes are just like us in a fur suit is the error running opposite to the boundary-defending one, and it does its own damage. Chimpanzee social life includes infanticide, coalitionary killing, and severe wounding at rates that would be unrecognizable in most human communities, and reading great ape cognition through a lens of similarity produces expectations that get people and animals hurt. The pet-chimpanzee cases that end in catastrophic injury are the practical version of this error.
Chimpanzees are five times stronger than humans is inflated. Measured differences are real but modest, on the order of one and a half times in muscle-specific force, attributable largely to muscle fiber composition.
The aquatic ape hypothesis and similar single-cause accounts of human divergence remain unsupported by the fossil, genetic, or comparative record.
Where the great ape cognition line sits now
Assemble the current evidence and the pattern is not a list of things apes cannot do. It is a set of capacities present in reduced or differently-organized form, with a small number of genuine discontinuities.
Apes track what others perceive and know, and use it. They communicate intentionally with a shared gestural vocabulary we can partly read without training. They maintain arbitrary local conventions that qualify as culture. They cooperate, recognize individuals for decades, form long-term relationships, and console. They plan, they deceive, and they solve physical problems flexibly. They pass a self-recognition test, respond to death in ways specific to death, and remember individuals across a quarter century.
What appears genuinely different in humans is a smaller list than the traditional one and more specific. Cumulative culture, enabled by high-fidelity imitation of methods rather than emulation of outcomes. Shared intentionality, the representation of a goal as jointly held. Syntax. Norm enforcement including third-party punishment. And an extended developmental window that keeps circuitry plastic for decades.
Notice that most of those are about transmission rather than about individual cognition. The animals whose knowledge visibly moves between individuals are running the same machinery at lower fidelity, and the difference in outcome is enormous because fidelity compounds. That is the same lesson the corvids and parrots that built comparable cognition from unrelated forebrain tissue deliver from a different branch, and the same one the parrots whose innovations spread through urban populations within a decade deliver from a third.
The methodological lesson is the one worth keeping. Apes failed false belief tests for four decades and passed when someone stopped requiring them to perform, which is exactly the pattern that turned up when working dogs were assessed on capacities nobody had designed a test for and when cooperative hunters were finally studied in the field rather than in enclosures. A negative result in comparative cognition is a statement about a task at least as much as about an animal.
The 24-lecture Neurozoology course works the tree of life on that basis throughout, alongside the first edition’s survey of nervous systems and the working animals whose capacities were discovered by the people relying on them. The long-lived social mammals whose knowledge dies with specific individuals and the populations whose behavioral traditions vanished with the animals carrying them are the same argument about transmission from further out on the tree, as are the migratory routes that had to be re-taught by aircraft once the birds who knew them were gone.
All four non-human great ape species are endangered, several critically, from habitat loss, hunting, and disease. Which means the comparison that has organized this entire research program, the one that keeps forcing revisions to what we think we are, is being conducted on populations that may not persist long enough to answer the questions. The populations whose accumulated knowledge disappeared along with the individuals holding it are the version of this that has already happened elsewhere, and with apes the loss would be of something additionally specific: not only the animals, but the comparison itself.
Every retreat of the line has been forced by watching an animal do something we had just finished explaining it could not do. There is a limit to how long that will remain possible.
Leave a Reply