There is a control experiment sitting on the couch, and almost nobody runs it.
Comparative cognition spends enormous effort trying to isolate variables that cannot be isolated. You cannot rerun the evolution of corvids with a different social structure. You cannot give an octopus a longer lifespan and see what accumulates. But two carnivores currently share human households in the hundreds of millions, both arrived there through domestication, and they arrived through domestication processes so different that they function as a natural comparison on what domestication actually does to a mind.
Dogs entered the arrangement first, somewhere in the range of fifteen to forty thousand years ago, from wolves, in a relationship that selected relentlessly on responsiveness to human social signals. Cats entered roughly ten thousand years ago, from a solitary desert ancestor, through a process nobody designed, driven by rodents in grain stores. One was bred for cooperation with a species that hunts in groups. The other was tolerated for pest control by a species it had no ancestral reason to cooperate with at all.
The result is a matched pair. Same house, same food, same human, two completely different cognitive profiles, and the differences track the domestication histories closely enough that dog and cat cognition is the most useful natural experiment available on the question of what living alongside people does to an animal’s mind. The recent literature has also produced a specific and slightly embarrassing finding: the gap between them is considerably smaller than anyone expected, and most of what looked like a gap was an artifact of the fact that cats will not cooperate with experiments.
Two domestications, and what they did to cognition
Wolf to dog is the more studied transition and the more contested one. Genomic evidence places the divergence somewhere between fifteen and forty thousand years ago, well before agriculture, which means the founding relationship was between hunter-gatherers and a large social predator rather than between farmers and livestock. Whether humans deliberately raised wolf pups or whether tolerant wolves self-selected by scavenging around camps remains argued, and the answer is probably both at different times and places.
What is not argued is the selective pressure. Whatever produced dogs selected hard on tolerance of humans, on reduced fear and aggression, and on attention to human behavior, and it did so in an animal whose ancestor already lived in cooperative groups with coordinated hunting, social hierarchy, and the ability to read conspecific intent. The raw material was a social species. Domestication redirected an existing capacity toward a new target.
The physical consequences arrived as a package, and the package is itself informative. Dogs show the domestication syndrome: floppy ears, curled tails, patchy coats, shortened muzzles, reduced sexual dimorphism, retention of juvenile behaviors into adulthood, and a brain roughly a quarter smaller than a wolf’s of equivalent body mass. The leading explanation ties most of these to neural crest cells, the embryonic population that contributes to adrenal tissue, pigment cells, cartilage, and parts of the peripheral nervous system, on the theory that selecting for reduced fear response selects for mildly reduced neural crest activity and everything else comes along uninvited. The Russian farm-fox experiment produced a version of the same suite within decades by selecting on tameness alone, though the interpretation of that experiment has been complicated by evidence that the founding population was already partly domesticated.
Cats are the opposite case in nearly every respect. The ancestor is Felis silvestris lybica, the African wildcat, which is solitary, territorial, and does not form groups even where food is abundant. Domestication began in the Fertile Crescent roughly ten thousand years ago when grain storage created dense rodent populations, and the cats that exploited that niche were those least disturbed by human proximity. Nobody was breeding for anything. The selection was for tolerance and nothing else, and it ran on an animal with no ancestral social cognition to redirect.
That produced something genuinely unusual. Domestic cats form social groups, which their wild ancestor does not, when resource density permits. Sociality here is not a redirected ancestral trait. It appears to be new, arising during domestication in a lineage where every other subspecies remains solitary regardless of food availability, which makes the domestic cat one of the cleaner examples available of a social capacity emerging rather than being inherited.
Cats also went through the process far less thoroughly, and their genome shows it. Comparisons between domestic cats and wildcats find far fewer differences than the equivalent dog-wolf comparison, concentrated in genes associated with fear response, reward-seeking, and neural crest development. Domestic cats remain capable of surviving independently and interbreeding freely with wild populations, which is not true of most domesticates. The reasonable description is that cats are semi-domesticated: changed in temperament and in tolerance, largely unchanged in body plan and ecology, and never subjected to the intensive functional breeding that produced sheepdogs and retrievers and sighthounds. Most cat breeds are a nineteenth-century invention and select on coat rather than behavior.
The consequence for dog and cat cognition is that the two species should differ specifically in social attention and specifically in the direction that reflects their starting material, and they do, but less than the folk model predicts.
The dog side: reading humans as the specialty
The finding that established dog social cognition as a serious field is embarrassingly simple. Hide food under one of two containers, then point at the correct one. Dogs follow the point. Chimpanzees, our closest relatives, largely do not, at least not spontaneously and not with the same fluency.
That result has held up across many variations and it is more specific than it looks. Dogs follow pointing gestures, gaze direction, and head orientation, they use them flexibly, and puppies do it before extensive experience with humans, which argues for a heritable predisposition rather than pure learning. Wolves raised identically by humans do it worse, and hand-raised wolves famously do not look back at a human face when a task becomes unsolvable, while dogs do so readily. That looking-back behavior is the signature: when a dog cannot solve a problem, its next move is to consult a person.
That specialization has a cost that rarely gets mentioned alongside the celebration of it. On physical problem-solving tasks with no social component, dogs frequently perform worse than wolves. Given a puzzle requiring persistence and manipulation, wolves work at it and dogs give up and look at the nearest human. Dogs are also more susceptible to being led astray by a human demonstrating an inefficient solution, copying the unnecessary steps where a wolf will skip them. Domestication did not make dogs generally smarter than wolves. It made them dependent on a social channel that is usually right, and dependence on a channel is a strategy rather than a capability.
The oxytocin work supplies a candidate mechanism. Mutual gazing between dog and owner produces increases in oxytocin in both, and administering oxytocin to dogs increases gazing, which increases owner oxytocin, which is a positive feedback loop of the kind that normally operates between parents and infants. Wolves do not show it. The interpretation offered is that domestication co-opted an existing mammalian attachment system and pointed it across a species boundary, which is a specific and testable claim rather than a sentiment.
Attachment testing supports it. Dogs tested in adaptations of the Strange Situation procedure, designed for human infants, show the behavioral signature of secure attachment: using the owner as a safe base for exploration, distress on separation, and specific greeting on reunion. That is the same measure, producing the same pattern, in an animal that is not a primate.
Two further findings sit alongside it. Dogs discriminate human facial expressions and process them with a hemispheric asymmetry comparable to the one humans show, and functional imaging in awake unrestrained dogs has identified regions responsive to human faces and to the emotional valence of human vocalizations, with some evidence of separate processing for praise intonation and word content. Dogs also perform above chance on tasks requiring them to choose between a person who has previously been helpful and one who has not, and to avoid taking food from someone who has behaved unfairly toward another dog. They show something resembling contagious yawning in response to human yawns and elevated stress markers when hearing recordings of human infants crying, which is thin evidence for empathy in any strong sense and reasonable evidence for emotional contagion.
The word-learning problem, and what the EEG actually showed
The most contested question in dog cognition is whether dogs understand words as referring to things, or whether they have learned that certain sounds predict certain outcomes. Those are different claims and the behavioral evidence could not cleanly separate them.
The famous cases are real and rare. A border collie named Rico learned around two hundred object labels and appeared to use inference by exclusion, retrieving a novel object when given a novel name. Chaser, another border collie, reached over a thousand. These animals exist, they are documented, and they are unusual enough that the field named the category Gifted Word Learners and started studying what makes them different, with recent work indicating the ability is not simply a product of training intensity but tracks individual cognitive differences.
The problem is that typical dogs perform poorly on the same tasks. Surveys of owners suggest an average comprehension around eighty-nine words, and controlled tests of ordinary dogs with a handful of claimed object names have found performance that does not clearly exceed chance once the owner’s ability to inadvertently cue is controlled. That gap between owner report and laboratory performance is where the argument lived.
Then a group in Budapest took a different approach and stopped asking dogs to perform. Using scalp electroencephalography on awake, cooperating dogs, they ran a semantic expectancy violation paradigm: the owner says a word the dog knows, then presents either the matching object or a mismatched one. The event-related potential evidence for referential understanding of object labels in dogs, published in Current Biology, showed that responses to the visual object differed depending on whether the preceding word was semantically congruent or incongruent, which is the same signature used to demonstrate semantic processing in humans.
The methodological point matters more than the result. A capacity that behavioral testing had failed to reveal in typical dogs was detectable when the measure did not require the animal to do anything. Performance-based tests conflate knowing with being willing and able to demonstrate knowing, and for an animal that may not care about the task, that conflation hides real capacities. Hold that thought, because it is the entire explanation for the cat literature.
More recent work has pushed further, reporting that dogs extend verbal labels according to object function rather than only to specific trained items, and that dogs with large label vocabularies can learn new labels by overhearing rather than by direct training, with the authors describing sociocognitive skills functionally parallel to those of eighteen-month-old children.
The soundboard buttons deserve their own sentence, because they are everywhere online and the research is more careful than the videos. A controlled investigation of soundboard-trained dogs found that the animals responded appropriately to button presses made by humans, which addresses the most obvious deflationary explanation, that the dogs are simply pressing buttons in patterns their owners reward. It does not establish that dogs are composing meaning, and nobody involved has claimed it does.
The cat side, and the experiments cats refused to take
For decades the comparative literature said cats were less socially cognitive than dogs, and it said so on the basis of studies cats declined to complete.
The pattern in the methods sections is consistent and slightly comic. Studies begin with a target sample, cats withdraw from the experiment, and the analysis proceeds on whoever stayed. Cats leave the testing area, refuse to approach the apparatus, fall asleep, or simply sit down. A dog that does not want to do a task will usually do it anyway because a human asked. A cat will not, and the resulting data made cats look incapable when they were mostly uninterested.
Once researchers redesigned around that, the picture changed substantially. The productive shift was toward measures that require nothing from the animal except looking, since where a cat looks and for how long is measurable whether or not the cat is cooperating.
The results have come quickly. Cats discriminate their own names from similar-sounding words. They learn the names of other cats in their household through daily exposure alone, with no training, and show surprise when a name is paired with the wrong cat. They match human voices to faces, indicating cross-modal representation of individuals. They mentally track their owner’s location from voice alone, showing surprise when the owner’s voice suddenly comes from a place the owner could not have reached. They discriminate human emotional expressions and adjust behavior accordingly.
The 2024 result is the one that inverted the ranking. Using a switched-stimuli looking-time task, researchers presented cats with arbitrary picture-word pairings and then swapped them. The rapid formation of picture-word association in cats found that cats looked longer at the switched combinations, indicating they had formed the association, after two nine-second exposures. Human infants in comparable paradigms have required roughly four trials at twenty seconds.
The social qualifier in that study is the part worth keeping. When the sounds were electronic rather than human speech, the effect did not reach significance. Cats formed the association with a human voice and not clearly with a machine tone, which suggests the capacity is entangled with social attention rather than being a general auditory-visual pairing ability.
Attachment testing has produced the same reversal. Applying secure base tests to cats found that a majority display secure attachment to their owners, in proportions closely comparable to those found in human infants and in dogs.
What the dog and cat cognition gap turned out to be
Put the two literatures side by side and the honest summary is that the difference in dog and cat cognition is smaller than a century of assumption, and that most of the apparent gap was measurement.
There is a real remaining difference and it is worth stating precisely. Dogs are better at using human communicative signals in cooperative problem-solving contexts, they look to humans when stuck, and they were selected for exactly that. Cats do not reliably follow pointing in the same way, do not look back at humans when a task becomes impossible, and generally do not treat a human as a collaborator on a problem.
But the perceptual and representational capacities look comparable. Both species recognize individual humans across modalities. Both form associations between arbitrary sounds and objects. Both attach securely to a primary human. Both read human emotional expressions. Both track social information about their household.
Which reframes the difference as motivational rather than cognitive. A cat can represent what a human is doing. It just has no ancestral reason to organize its behavior around helping, because its ancestor never cooperated with anything. Cats also have a communication asymmetry that supports this reading: adult wildcats meow at each other essentially not at all, and domestic cats meow at humans constantly, with individual cats developing vocalizations specific to particular people. The cat did not inherit a social communication channel. It built one, aimed exclusively at us. The acoustic detail supports the reading: analysis of domestic cat meows finds them shorter and higher-pitched than wildcat vocalizations, and humans rate them as more pleasant, which is what a signal shaped by human response rather than by feline biology should look like. There is also the solicitation purr, in which a cat embeds a high-frequency component in the purr that overlaps the frequency range of an infant cry, and which humans reliably rate as more urgent. Whatever else that is, it is a signal tuned to a receiver of a different species.
That is a more interesting finding than cats being aloof, and it is a warning about comparative methodology generally. Any test that requires an animal to want to participate is measuring motivation and capacity together and reporting the product as capacity. That failure mode is not confined to cats. Every negative result in comparative cognition carries the same ambiguity, and the species that decline to cooperate with laboratory paradigms generally have literatures shaped by the same bias, which is why field observation and passive measures keep overturning conclusions that behavioral testing had settled.
Where the aging brains converge
There is a practical reason the comparison matters beyond comparative psychology, and it involves what happens at the end.
Dogs develop canine cognitive dysfunction, a syndrome of disorientation, altered social interaction, disrupted sleep-wake cycles, house-soiling, and reduced activity, and the neuropathology includes beta-amyloid accumulation in patterns resembling human Alzheimer’s disease. Cats develop a comparable syndrome with a different pathological signature involving tau. Both species live alongside humans, share environmental exposures, receive medical care, and reach old age in numbers that laboratory rodents on controlled diets never do.
That makes household carnivores an unusual epidemiological resource. A dog shares your air, your household chemicals, your noise environment, and roughly your activity pattern, on a compressed lifespan that lets a fifteen-year study run in fifteen years rather than eighty. Large-scale longitudinal work following thousands of companion dogs through life is now producing data on how environment, activity, and body size interact with cognitive aging, and the results feed back into human questions rather than only veterinary ones. The contrast with long-lived wild animals whose cognitive aging is nearly impossible to study is stark, since following an individual whale or elephant across a full lifespan requires a research program longer than most careers.
The size relationship remains the strangest part. Across mammals generally, larger species live longer. Within dogs the relationship inverts sharply, with giant breeds reaching old age at seven and small breeds at fifteen or more, and cognitive decline tracking that compressed schedule. Nothing about that is fully explained, and it is one of the more accessible open problems in aging biology sitting in plain sight in millions of homes. The birds whose lifespans run decades on a fraction of the body mass sit at the opposite corner of the same puzzle.
The dog nose, and a sense nobody can benchmark
Olfaction is where dogs are not merely better than us but operating in a different regime, and the numbers are worth getting right because the commonly cited ones are inflated.
Dogs have on the order of two to three hundred million olfactory receptor neurons against roughly six million in humans, with the olfactory epithelium spread across a turbinate structure that maximizes surface area, and a proportionally much larger olfactory bulb. The frequently repeated claim that dog smell is ten thousand to a hundred thousand times better than human smell traces to a casual estimate rather than a measurement, and detection threshold varies enormously by compound. For some odorants dogs are spectacularly more sensitive. For others the difference is modest.
What makes canine olfaction genuinely different is not only sensitivity but sampling. Dogs sniff in rapid bouts, several per second, and the nasal anatomy separates airflow into a respiratory path and a dedicated olfactory path, so sniffing is not breathing. Exhaled air exits through side slits rather than back over the sensory epithelium, which prevents the outgoing breath from disturbing the odor being sampled. The animal is running an active sampling strategy with a probe rate under motor control, in the same architectural sense that echolocating animals control the timing and shape of their own signal, and the dolphins whose biosonar was put to military use because nothing built could match it were selected for the same reason working dogs were: a sensory instrument nobody has managed to replicate.
Dogs also track odor gradients over time, which lets them determine direction of travel from a trail by comparing the age of successive footprints, and they discriminate individual humans by scent reliably enough for the capacity to be used operationally. The working dogs whose jobs depend entirely on that capacity are deployed on the basis of performance nobody has fully characterized mechanistically.
Medical detection is the frontier and it deserves calibration. Dogs have been trained to indicate on samples from people with various cancers, on impending seizures, and on hypoglycemia, with published sensitivity figures that are sometimes impressive. The problems are reproducibility across laboratories, the difficulty of controlling for handler cueing, and the fact that a dog trained on a particular sample set may be detecting something specific to that set. The capacity is real. The reliability required for clinical deployment has been harder to demonstrate than early results suggested. The Clever Hans problem is unusually severe here because a detection dog is by design attending closely to a handler who frequently knows which sample is which, and double-blinding a scent trial is harder than it sounds when the dog can smell the people running it.
Cat sensing, and the whiskers as an instrument
Cats get less attention on the sensory side and the hardware is worth describing because it is a coherent design for a different job.
The eyes are optimized for low light rather than for detail or color. A reflective tapetum lucidum behind the retina bounces unabsorbed photons back through the photoreceptors, roughly doubling the chance of capture and producing eyeshine. Rod density is high and cone density is low, which yields excellent dim-light performance and poor visual acuity, on the order of a tenth of human acuity at distance. Color vision is dichromatic. The pupil is a vertical slit, which permits a much larger dynamic range of aperture than a round pupil and which is characteristic of ambush predators active across a wide range of light levels.
Hearing extends well into the ultrasonic, past sixty kilohertz, which covers rodent vocalizations, and the pinnae rotate independently through a wide arc under the control of a large number of muscles, allowing directional scanning without head movement.
The whiskers are the piece most people underrate. Vibrissae are embedded in follicles with dense mechanoreceptor innervation and a dedicated cortical representation, and cats actively position them, sweeping them forward during close approach and pinning them back during conflict. Whisker position is also readable as a state indicator by anyone who knows what to look for, which is one of several channels cats use that humans systematically miss. Ear position, pupil dilation, tail posture and tail-tip movement, and the slow blink all carry information, and controlled work has found that humans reciprocating a slow blink increases the likelihood of a cat approaching, which is a rare case of a deliberate cross-species signal being experimentally validated rather than asserted. Cats also display over two hundred distinct facial expressions in interactions with other cats, a repertoire nobody had catalogued until recently because nobody had looked closely at animals long assumed to have nothing to say. The species whose signal repertoires turned out to be far larger than assumed once somebody recorded properly are a recurring pattern rather than an exception.
The whiskers function as a near-field spatial sensor at exactly the range where the eyes cannot focus, which matters because a cat’s minimum focal distance leaves it effectively unable to see prey held in its own jaws. The whiskers are how the animal knows what it is holding, and the mystacial pad arrangement provides enough spatial resolution to determine orientation, which is the information required to deliver a killing bite between two vertebrae without looking.
That is active sensing in the technical sense: a self-positioned probe, under motor control, sampling a region and adjusting based on the return.
Hunting, play, and behavior with no outlet
A domesticated predator carries a behavioral program its situation no longer requires, and what happens to that program explains a large fraction of what owners actually experience.
The cat’s predatory sequence runs stalk, chase, pounce, grab, kill-bite, and each element is separately motivated rather than being a single chain that runs to completion. A well-fed cat still hunts, because the motivation attaches to the earlier elements and satiation only suppresses consumption. That decoupling is why play with a toy is genuinely satisfying to a cat and why a cat will kill things it has no intention of eating, and it is also why the ecological damage from free-roaming domestic cats is substantial and unrelated to whether the animals are fed. The scale of that damage is one of the genuinely uncomfortable findings in the field, with island extinctions attributable to introduced cats and continental bird and small-mammal mortality estimates running into the billions annually, which sits awkwardly against the same animal’s status as a household companion. The seabird and ground-nesting populations most affected evolved with no equivalent predator, and no amount of feeding changes the motivation.
Object play in cats is predatory behavior with the terminal element removed, which is why toys that move erratically and are roughly prey-sized work and why a toy that does not eventually get caught produces frustration rather than satisfaction. The same logic runs through dog play, where the sequence is chase and grab-bite rather than kill-bite, and where breed differences map onto which elements of the ancestral hunting sequence were amplified or suppressed. Herding breeds are running a modified stalk and chase with the grab suppressed. Retrievers have an amplified grab-carry with an inhibited bite. Livestock guardian breeds have most of the sequence suppressed entirely, which is why they can live among animals that every other part of their ancestry says to eat.
Play between dogs also carries formal signals that make it a useful case in animal communication. The play bow functions as a metacommunicative marker, indicating that what follows should be interpreted as play rather than aggression, and dogs self-handicap during play with smaller or weaker partners, reducing the force of their behavior in ways that suggest an assessment of the partner rather than a fixed motor program. That combination, a signal about how to interpret subsequent signals plus adjustment to a specific partner, is more sophisticated than the behavior looks.
The claims that do not hold up
An audit, because this is the domain where folk belief and research diverge most.
Dogs are colorblind is wrong in the way people mean it. Dogs are dichromatic, seeing blues and yellows and confusing reds and greens, which is comparable to human red-green colorblindness rather than to monochrome vision.
One dog year equals seven human years is a rule of thumb with no biological basis. The relationship is nonlinear, dogs mature much faster early and then slow, and lifespan varies enormously with size in a pattern that runs opposite to the usual mammalian relationship: large dogs die younger, which is unusual and not fully explained.
Dogs feel guilt when they look guilty is the best-studied case of human misreading. Experimental work found the guilty look appears in response to owner scolding rather than to the dog having actually transgressed, with dogs who had done nothing wrong displaying it just as readily when scolded. The expression is an appeasement signal, and it is a response to the human’s behavior rather than to the dog’s own.
Cats are asocial is contradicted by group formation, allogrooming, allorubbing, and secure attachment. What is true is that the ancestor was solitary and that cat sociality is facultative rather than obligate.
Cats are not affectionate, they only want food is unsupported. Controlled preference testing has found that a substantial share of cats prefer social interaction with humans over food, and the attachment findings point the same direction.
Cats do not respond to their names because they cannot recognize them is refuted by the discrimination work. Cats recognize their names. Response is a separate question, and the failure to respond is the behavior people are actually describing.
Cats are low-maintenance compared to dogs understates their requirements in a way that has welfare consequences. An animal running an intact predatory sequence with no outlet, in a territory smaller than any wild felid would tolerate, frequently sharing that territory with unrelated conspecifics it did not choose, is in a situation with real potential for chronic stress, and much of what gets labeled behavioral problems in cats is a housing problem rather than a temperament one.
Cats always land on their feet is true enough to be interesting and false enough to matter. The righting reflex is real, developing by about seven weeks and working through a sequence of body rotations that conserves angular momentum, and it requires a minimum falling distance to complete. Veterinary case series on falls from height describe a pattern in which injury severity does not increase monotonically with the number of stories, which has generated a great deal of confident explanation and which is substantially confounded by the fact that cats that die on impact are less likely to be brought to a clinic at all.
Purring means a cat is happy is incomplete. Cats purr when content and also when injured, frightened, or dying, and the leading hypothesis is that purring is a self-soothing and possibly tissue-beneficial behavior rather than a happiness signal.
Dogs see the owner as a pack leader and the associated dominance training framework rest on a wolf model that has been repudiated by the researcher whose captive-wolf work generated it. Wild wolf packs are family groups led by breeding parents rather than by individuals who fought their way up, and applying a captive-artifact hierarchy model to dog training has no support. The social carnivores studied in the wild where the packs are family units run the same structure, with cooperative breeding and reproductive suppression rather than dominance contests, which is the arrangement the captive studies obscured.
Dogs understand hundreds of words in the way humans do is not established for typical dogs, and the gifted individuals are exceptional rather than representative.
What the dog and cat cognition pair demonstrates
The reason dog and cat cognition earns a place in a comparative course is not that these animals are clever. Plenty of animals are cleverer, and the cockatoos that manufacture drumsticks and beat out individual rhythms or the reef fish coordinating hunts with a different species will outperform either on the specific things they were built for. It is that they are the only natural experiment available on a specific question: what does living with humans do to a mind, and how much of the result depends on what the mind was before it started.
The answer that emerges is layered. Dogs show that domestication can take an existing capacity, in this case group-living social cognition inherited from wolves, and retarget it at a different species with astonishing precision, producing an animal that reads human gestures better than our closest primate relatives do. Cats show that domestication can produce social behavior in a lineage that had none, which is a harder thing and which produced a narrower and more selective result: a communication channel aimed only at humans, sitting on top of an animal that still hunts alone.
Both results also demonstrate how fast this can happen. Ten to fifteen thousand years is nothing on an evolutionary timescale, and in that window one lineage acquired a cross-species attachment system and the other invented a vocalization aimed at a species it had never previously addressed. Whatever capacity for change a mammalian nervous system has, it is larger than the timescales usually invoked would suggest.
Both outcomes are versions of the same principle the comparative literature keeps producing. Capacities are rarely built from nothing and rarely absent entirely. The corvids and parrots that constructed executive function from non-homologous forebrain tissue, the parrots whose problem-solving keeps outrunning what their brain size predicts, and the primate societies whose tool traditions differ between neighboring valleys are all instances of an existing substrate being recruited for a new job under a new pressure.
What the household pair adds is a control on the pressure. The wild populations whose behavior has been tracked for decades and the ones studied under different ecological conditions show what selection does over evolutionary time in the absence of us. Dogs and cats show what happens when we are the selection pressure, applied to two starting points, over roughly comparable spans. The macaque troop whose innovation spread socially and the birds whose regional dialects mark where they were raised are running the transmission side of the same question, and the working animals whose capacities were discovered by people who needed something from them are the applied version.
The methodological lesson is the one worth carrying furthest, and it generalizes past pets. For a century, dog and cat cognition research reported that cats were less capable, and the reason was that cats would not take the tests. When the measurement changed to something requiring no cooperation, the cats turned out to be doing most of the same things. Every comparative claim rests on a task, every task requires participation, and participation is motivation rather than capacity. The 24-lecture Neurozoology course applies that skepticism throughout, alongside the first edition’s survey of nervous systems and the study of how knowledge moves between animals.
Somewhere in your house there is an animal descended from a pack hunter that has spent fifteen thousand years learning to read your face, and possibly another descended from a solitary desert cat that invented a vocalization it uses on nobody but you. Neither of them is a lesser version of anything. They are two different answers to the same question, and one of them has been refusing to take the exam this entire time.
Leave a Reply