Cat Psychology & Domestication: Are We Good Owners?

Extended book review of Bradshaw 2013 (Cat Sense) on the connections between cat psychology, evolution/genetics, history of domestication or lack thereof, & possible dysgenics, highlighting modern maladaptivity of cat psychology, with fulltext bibliography of key references.
reviews⁠, biology⁠, genetics⁠, cats⁠, bibliography⁠, insight-porn
2018-11-032021-10-17 finished certainty: possible importance: 5 backlinks / bibliography

I review John Bradshaw’s book on psychology, Cat Sense, after difficulties dealing with my own cat. Bradshaw reviews the history of domestic cats from their apparent Middle Eastern origins as a small solitary desert predator to their domestication in Ancient Egypt where breeding millions of cats for sacrifice may have played a critical role (as opposed to any unique role as a vermin exterminator) through to the modern day and psychological studies of the learning abilities and personalities of cats, with particular emphasis on cat social skills in “cat colonies” & plasticity in kittenhood. As Bradshaw diagnoses it, these are responsible for what ability they have to modern pet life, even though they are not bred for this like dogs; every tame cat still has the feral cat in them, and are in many ways unsuited for contemporary living, with disturbing hints that human lack of selective breeding plus recent large-scale spay/​​​neuter population control efforts may be producing a subtle dysgenic effect on domestication, and this double neglect & backfire may be responsible for disturbingly high rates of cat maladaptation & chronic stress diseases.

Cat researcher John Bradshaw’s 2013 book Cat Sense: How the New Feline Science Can Make You a Better Friend to Your Pet⁠, Bradshaw 2013: Popularization of scientific research on domestic cat psychology, starting with its history & evolution. (Bradshaw has researched cats for decades, written a similar book on dogs called Dog Sense, and co-authored in 2016 with Ellis The Trainable Cat⁠.)

While published in 2013, I regret to say that, unlike a human behavioral/​ book like 2014 A Troublesome Inheritance (review) or 2018 Who We Are And How We Got Here⁠, which were obsolete either before they were published or shortly thereafter1⁠, Bradshaw’s book is, as far as I am aware, a useful overview of cat genetics & psychology research, such is the dormancy of the field. (I could only point to ⁠, Montague et al 2014 or ⁠, Ottoni et al 2017, as being relevant updates—also interesting in their own right.) Despite their overwhelming popularity as the #1 pet & in popular culture, cats aren’t researched much—cat appear to be largely useless for heritability purposes (unlike dog pedigrees which are often analyzed), there is no equivalent of 23andMe or for cats the way there is for the dogs (Embark⁠, which can do impressive research like or ), and large-scale dog breeding programs or experiments in dog cloning for military use or behavioral genetics experiments like & ’s ⁠. Cat research being such a backwater, the sample sizes are absolutely tiny with scarcely any replication, and I’m sure that a decent fraction of what Bradshaw claims or speculates is wrong2⁠, unfortunately, I don’t know which ones. We must do what we can what with we have.

Far From the Madding Crowd

I picked up Bradshaw for 2 reasons.

First, I’ve been interested in the catnip response for a while, curious how common it is, why it varies from country to country⁠, and what alternatives there are, and thought Bradshaw might be useful.

The second was my cat.

My cat.

Back in 2017, my cat began having problems, urinating bloody urine and then not urinating at all; coming right before I was about to take a month-long trip and having some experience as a kid with the swiftly fatal consequences of a cat not urinating from my previous cat (which killed her on Christmas day, just to twist the knife), I rushed him back and forth to the vet’s, a task complicated by the fact that during the first trip he was so upset he urinated in the cat carrier and a urine sample couldn’t be obtained, but where the eventual diagnosis was 3⁠, temporarily patched by painkiller injections to allow urination, mostly caused by the standard dry cat food I fed him and fixed by never again feeding him dry food but wet cat food, and in helpfully specific advice founded no doubt on extensive empirical research, noted that the cystitis was possibly stress-exacerbated in some fashion or other (maybe by relatives visiting & the neighboring cat). Thankfully, switching in the prescription wet food “Crystal Diet” did the trick; in time, he rejected it, refusing to eat4⁠, so I replaced it with regular wet food which is only somewhat more expensive (although eventually he began turning his nose up to some degree at that, preferring only the salmon pate, so eventually I gave up and began ordering only that5). I thought I was a good owner. Given how expensive and stressful the experience was for the both of us, I decided some insight into cat psychology was worth obtaining and Bradshaw came reasonably recommended.

Bradshaw, it turns out, does not once mention catnip or cat psychoactives. That was a little disappointing.

But he does have a lot of advice about cat psychology & & how to keep a cat sane. He emphasizes the degree to which a cat reacts to its environment and is stressed by it.

As an example of their sensitivity, he mentions (pg134) that cats will notice and investigate any changes in their environment. To test this claim, I began making small harmless changes while he was outside, such as moving a box or bowl to the side, and sure enough, upon jumping back through the cat flap, much of the time he would look around the room and immediately jump down to investigate the changed object. I was impressed how observant he turned out to be. I would never have noticed those changes.

Based on Bradshaw’s advice, I tried a number of things: I replaced a noisy box fan & put rubber vibration-absorbing pads on the bottoms of all the fans/​air-filters/​dehumidifiers/​computers (unclear efficacy); I bought a Feliway diffuser spray (no apparent benefit and the Feliway-sponsored studies left me skeptical); I bought a large water bowl to encourage drinking and eventually began adding water to the wet food; moved his feeding station to a more hidden corner; I bought two ‘puzzle treat’ balls, simple and complex⁠, to put dry food or treat bits in for him to play with (a big hit, although use must be strictly rationed to avoid triggering cystitis again, and the simple treat ball, which was an empty sphere, turned out to be far too easy to get treats out of6); bought 2 ‘cat condos’ which are cubes similar to cat perches (which he frequently sleeps on although again I don’t know how much difference that makes); red/​blue/​purple/​green laser pointers off eBay to supplement the wand for playing chase (initially highly effective but he quickly lost interest & I think limitations of may make some colors much less effective7); a Mini (cool toy but he remained afraid of it so after a year I gave it to my sister for her ferret); a replacement cat flap door (previous one was breaking down & jamming & sometimes he struggled to get in); “oat grass”⁠/​“wheat grass” seeds for growing oat/​wheat shoots to nibble on8 (which he sometimes did but it wasn’t worth the trouble & I had problems with over-watering causing mold); bought a large shelving unit with extra shelves in the hope that he would find it a useful hiding-place and would jump up to various levels (he sometimes sleeps at the bottom but my efforts to get him to experiment with higher levels failed); the vet, during the next visit which went poorly (as usual: ⁠/​)9⁠, gave me a 3-pill sample of 10 to try during relatives’ visits or future appointments (when I tried one dose, it made a little difference but not a lot, so I reserved the remaining 2 for the next vet visit, and 2×100mg turned out to be much more effective although the visit was still difficult for both fo us); and replaced the opaque cardboard/​foam insulation around the cat-flap with an acrylic sheet from Lowe’s I cut to fit so he could more easily watch outside while laying on the window ledge. I also began periodically putting him in the cat carrier and carrying him around either by hand or in my car to try to gradually reduce his aversion to it. One particular success was using : I had a hard time getting him to pay attention to the computer monitor long enough to realize that it was displaying videos of birds, but when I turned on the sound, he noticed and instantly became addicted. Of these, the most worthwhile changes seem to be the wet cat food, puzzle treats, cat condos, videos-for-cats and exposure therapy.

One case study matched him almost exactly, but most of the relevant solutions were inapplicable: he already has his own litter box, water bowl, food dish, apartment, and places to hide, and the outside was visible only from the cat perch & ledge. To do any better, I would have to eliminate visitors and nearby cats entirely—I can’t do much about those major stressors, unfortunately, and in fact another neighborhood cat has been coming around occasionally at night, which doesn’t help. But while he has taken a disliking to the litter box in favor of peeing in my bath tub, the cystitis has not repeated itself. I’ll take what I can get.


One of Bradshaw’s most interesting points about domestication is about Egypt. If wildcats are so untameable even adopted as kittens (as documented by case-studies like Mary Frances Pitt & Smithers & Tomkins’s attempts with Scottish wildcats), and we have not been domesticating cats ourselves, where does the original domestication come from? Whose seed-corn have we been eating?

He points to the dizzyingly long history of cats in Ancient Egypt—in few places other than ancient Egypt can authors so casually skip 1000 or 2000 years between examples—well displayed in Malek’s The Cat in Ancient Egypt illustrations, which I have since read & scanned, and does include the humorous cartoons of cats Bradshaw mentions. Cats back then were, apparently 15% larger than contemporary domestic cats (one of at least two anomalous changes in domestic cat sizes, the other being a “staggering and unexplained 70% reduction in bone-lengths…between the 11th & 14th centuries”), but more intriguing is the role of cats in Egyptian religion.

The common assumption I shared, that cats were naturals for domestication because they are such good vermin exterminators, is apparently not well-supported as there were many alternatives, some superior to cats in ways. Instead, the key to their domestication may be—and this is speculative, I should caution—their essentially arbitrary role as popular sacrifices, requiring countless ‘catteries’ attached to temples, and at least millions of sacrifices on a scale staggering to contemplate: “more than 4 million mummified ibis, a medium-sized wading bird that the Egyptians bred in captivity, were recovered from catacombs at Tuna el-Gebel, and an additional 1.5 million from Saqqara.” This is only the tip of a grisly iceberg:

We will never know how many cats were sacrificed this way. The archaeologists who discovered these sites wrote of vast heaps of white cat bones, and dust from disintegrating plaster and linen blowing across the desert. Several other cemeteries were excavated wholesale, and their contents ground up and used as fertilizer—some was used locally, some was exported. One shipment of cat mummies alone, sent to London, weighed nineteen tons, out of which just one cat was removed and presented to the British Museum before the remainder were ground into powder. Out of the millions that were mummified, only a few hundred now survive in museums, and these come from a mere handful of the many cemeteries constructed over a period of several hundred years.

The iceberg becomes even more gruesomely impressive when you remember that cats are obligate carnivores (unable to be fed on non-meat products which lack the amino acids they have lost the ability to synthesize such as taurine which is ), and the Egyptian cat mummies were kept well-nourished & healthy right up to being carefully strangled, implying expensive meat consumption, and perhaps most astonishing of all to a cynic like myself, Bradshaw specifically notes “almost all the mummies that were produced to look like cats actually do contain a complete cat skeleton”. So:

The cat had rivals for the role of vermin exterminator: other carnivores of similar size were tamed, including various members of the weasel family, and the genet and its cousin the Egyptian mongoose…Very probably, cats did not have the edge as vermin controllers. The answer, therefore, must lie elsewhere, possibly in the cat’s biology. The connection between cats and religion is unlikely to have been crucial, since Egyptians sometimes venerated both mongoose and genet as well.

More likely, the cat managed to become more profoundly domesticated than any of its rivals. However, which was the cause and which the effect? Are cats more “trustworthy” and predictable than ferrets because they have evolved ways of communicating with humans, or is it the other way around? Since we do not know precisely how the domestic cat’s direct ancestors behaved, such questions are impossible to answer. Nevertheless, the cat’s capacity to evolve not only into a pest controller but also into a pet animal—its present-day roles—must have been central to its success in the first 2,000 years of its domestication. So what set the cat apart on its millennia-long journey into our homes?

Here, the cat’s involvement in Egyptian religion may well have been crucial. It is possible that the Egyptians’ veneration of cats that gave the cat the time required to evolve fully from wild hunter to domestic pet; otherwise, it might have remained a satellite of human society and not an intrinsic part. It is even possible that the factories that produced the cat mummies forced the evolution of cats that could tolerate being kept in confined spaces and in close proximity with other cats, both qualities that are signally absent from today’s strongly territorial wildcats but that are essential to life as an urban pet. Although of course most of the cats that carried the relevant genes died young—that was how they were being bred, after all—some must have escaped into the general population, where their descendants would have inherited an improved ability to deal with the close confines of urban society. Such changes take only a few decades in captive carnivores, as exemplified by the that turned wild animals docile in ⁠.

Is it possible that today’s apartment-dwelling cat owes its very adaptability to the inhabitants of those gruesome Egyptian catteries?

Cats gradually spread post-Egypt, particularly by sea (leading to some interesting cases like islands of all-black cats due to founder effects), but without any signal events aside from an unfortunate period of persecution by the Catholic Church (itself anomalous inasmuch as cats were long popular with ecclesiastics) and occasional mistakes like 200,000 cats being killed because of London’s Great Plague of 1665–66.

At this point, Bradshaw turns from the history of cats, such as it is understood, to what sort of psychology we expect from a creature evolved for the cat niche, with its uniquely difficult nutritional needs, and then the practical implications of this psychology for cats as pets. Other reviews of Cat Sense complain that it is repetitive, which I didn’t really notice (there is a difference between being repetitive, and revisiting a central point from different angles) or that it focused too much on summarizing past research (a positive aspect for me) or Bradshaw’s research (write what you know, and such criticisms would be more convincing if they mentioned what key papers Bradshaw was neglecting in favor of his own), or that they already knew everything Bradshaw said and were deeply bored by it (in which case they must be far more expert on cats than I am and perhaps should be writing the papers themselves), or seem to be reading the wrong book entirely and should be reading the later The Trainable Cat.


Miscellaneous points I learned about cat psychology & biology:

  • the Greek for cat, aielouros, is literally “waving tail”; Egyptian girls were being named miw not long after miws started being pets

  • long-haired cats are disadvantaged less by heat than by fur matting, causing infections/​​infestations

  • cat fur colors are something of a mystery; despite & pleiotropic associations with various phenotypic traits, they have little practical effect:

    • cats appear to have no fur color preferences in mating

    • there is no known explanation for why black cats are so common, especially given that humans dislike black coats and black cats are seen as less friendly11

      However, this prevalence of black cats does imply that net human control over cat genetics is near-zero—perhaps because the cats we favor are immediately neutered and therefore we genetically destroy the cats we prefer (Clark 1975)

  • cats are obligate carnivores, unable, among other things, to make vitamin A or vitamin D and needing extra niacin & thiamine, and actually require meat/​​fat to make hormones to reproduce at all, while they can smell tomcat urine for the sulfuric ‘felinine’ to gauge hunting success; because of this, it is quite difficult to make a healthy vegetarian diet for a cat, and nutritionally-complete cat food apparently only became “widely available for only 35 years or so” (I wonder how well urban cat’s-meat men compare?)

  • due to their desert descent, they have small water-efficient kidneys (though this has unfortunate consequences like quickly succumbing to kidney failure and being easily killed by home pollutants like essential oils)

  • kittens have a much shorter window of plasticity than puppies, who can tolerate lack of human contact for up to seven weeks with any harm, but by that point, kittens have already been damaged. By seven weeks, a kitten should already be learning play signals and mock fights and signals to stop playing (apparently, the ‘stop play’ signal is to “arch their back, their tail upward, and then leap off the ground”; I don’t think I’ve ever seen that one before). Similarly, kittens raised as litter-mates will get along closely, as indicated by things like grooming each other or laying touching another or happily eating side by side; while cats who met as adults will never, rarely, or never (respectively) do those behaviors no matter how cordial. And kittens raised with friendly dogs will be fine with dogs in the future, which is not something which could be said of all (or even most?) cats who encountered dogs as adults…

    • Cat research has publication problems. Of the plasticity window, Bradshaw then notes: “Remarkably, this revolutionary work has never been published in peer-reviewed journals; however, no one since has fundamentally disagreed with its conclusions.” On another occasion, “The late Penny Bernstein conducted a detailed study of stroking, details of which sadly remained unpublished when she died in 2012.” And Bradshaw himself omitted the key behavioral details from his 1999 cat dysgenics paper⁠.

      Cat research has issues with publishing & propagating knowledge—among catnip researchers, consider the ignorance of Villani 2011 or the never-published catnip GWAS⁠.

  • the is a quick (and amusing) way to immobilize a cat; I tested it myself on several cats. (My vet prefers to use a method of wrapping the whole cat tightly in a towel so they can’t see anything.)

    Incidentally: Do male cats ever carry around kittens/​​other cats? Do females do this instinctually, or does it only happen because of motherhood leading to learning the behavior via trial-and-error? It occurs to me that I’ve never seen a spayed/​​neuter cat do it that I know of, either in person or via media; all the cats carrying around kittens are either mothers or I don’t know their history/​​status.

  • ” is one of the major killers of kittens, rather than starvation or accident, especially in cat colonies, and a perennial problem in animal shelters (getting in the way of good socialization)

  • Mother cats can’t/​​don’t recognize their young, and are perfectly happy to suckle a random kitten quietly slipped into their litter. Apparently cuckoos are not a problem; anything which acts like a kitten gets milk. (This suggests that cat colonies are rare enough or genetically homogenous enough to not evolve strong offspring recognition skills as the occasional mistake doesn’t reduce inclusive fitness enough to matter.)

  • in addition to the famous reflecting layer, cat eyes achieve night vision by having 10× fewer nerves, each nerve accounting for a larger bundle of rods spread further out, allowing greater sensitivity to dim light, at the cost of precision in bright light; they see only blue & yellow but even there appear bad at discriminating. “Any other difference between objects—brightness, pattern, shape, or size—seems to matter more to cats than does color.” Peculiarly, lacking muscles to focus their eyes, they ‘scan’ objects instead, and Bradshaw suggests that “perhaps because it is just too much effort, they often don’t bother to focus at all”. The inability to see things nearby is presumably compensated by whiskers. This farsightedness has a peculiar psychological consequence: cats appear to have a special memory system to memorize what is in front of them, in order to step over obstacles; if they haven’t stepped over it and are distracted, they then forget (and need to look around before they can start walking again), but if they have stepped over it, they automatically remember for the next 10 minutes and their hindlegs will automatically step over it. (A decerebrated cat can still walk⁠, so there seems to be quite a bit going on in the rest of the feline nervous system.) I found that quite bemusing. (One thing Bradshaw doesn’t cover is how cats step in precisely the same place with their back paw as the front paw, which is called “direct registering” and is useful in quietly navigating cluttered environments; this explains why cats can jump on and off desks for years without knocking off a single thing—accidentally knocking things off would be shamefully dog-like. Cats, of course, only knock things off intentionally.)

    On the other hand, they can hear ultrasound and unusually low-frequency noises because of an unusual double-chambered chamber in their head, and localize even ultrasonic sounds—elegantly, if the ears’ repointing doesn’t work because of high frequencies, the chamber then starts to work

  • when showing a cat something interesting, like leaves (a much rarer cat psychoactive), I’ve noticed them do an odd rictus grin to breath deeply; this turns out to be a way of invoking the for hormone smelling, called the response. Primarily used in social situations, such as the cheek rubbing, which marks things with the glands in the cheeks. (If you’ve ever wondered why your cat likes rubbing its cheeks on you…)

  • cat toys simulate hunting, and boredom appears to reflect a combination of habituation and frustration. Changing color of a cat toy will restore interest, and being given a toy they can change or destroy also maintains interest.

    Commercial toys do neither. Bradshaw notes that cats appear to be cautious around large toys, and never learn that they are in fact harmless, which matches my experience with trying to get him to play with the Sphero Mini, his great interest in the puzzle treats (getting a treat after fighting with it is isomorphic to hunting, after all, and eating the pellet definitely changes it), and with being able to restore his flagging interest in a laser pointer by switching to a different color. It seems like commercial cat toys may be suboptimal. (One exception I haven’t tried yet is the Ripple Rug⁠, a 2-layer carpet, which is reminiscent of how cats play with things under sheets or behind curtains and which is inherently scrunchable & rearrangeable for novelty; the positive reviews of it describe it in ways which line up with this mutability theory of cat play.)

  • there are a : domestic cats can hybridize with a remarkable number of other species in Felidae, including South American cats with different chromosome counts (and the domestic cat × = ‘Safari’ cats are even fertile!)

  • cats can be trained, but it is difficult. (Tellingly, one often refers to ‘untrained’ or ‘poorly-trained dogs’, but never to an ‘untrained cat’. It is just assumed that cats will go along improvising regardless of how maladaptive a behavior is, and cannot be taught better.)

    It is just harder: training & reinforcement are inherently ambiguous/​​under-determined12⁠; unlike dogs, who can be trained with fairly sloppy mapping between action and reward, are eager to please, and are rewarded just by human praise, cats have no particular drive to please or human theory of mind (people think that meows and other sounds are communication but as cat owners are unable to interpret other cats’ sounds, the meows generally seem to represent an arbitrary language learned by each cat by trial-and-error) and require food treats, must be trained backwards starting from the final step, and require very tight action/​​reward feedback—a reward must be provided within seconds. As it’s not easy to dispense a cat treat within seconds of an action, the easiest way to do this is : use a small gizmo which makes click sounds, and give them a treat every time you click it; after a while, the click itself becomes a reward, and it’s easy to make a click quickly, perhaps at a distance, delivering instant feedback.13

    I found this useful to know for how to ‘train’ him but also for how to not train him to do things. One thing people often do with pets or small children is accidentally train them to do bad things. The pet/​​child wants something, makes trouble to get it, the person gives in and gives the thing, and one iteration of training has occurred. Dogs barking to get something is one I hear a lot, but cats can do it too—my grandparents have ‘trained’ their cat to wake them up at 6AM so he can go out. The problem, of course, is that they gave in too quickly, and he learned the lesson well. Now, when he tries out being a nuisance (to get some food, usually), and it’s a legitimate need, I avoid reinforcing him by deliberately waiting a few minutes until he’s given up and enough time has passed that then giving him whatever would not lead to any learning.

    • the lack of trainability apparently has an exception, Bradshaw states: food can trigger learning of powerful associations even hours after consumption. This would make sense as an anti-bad-food defense, but unfortunately, this is yet another maladaptation in the modern context: “…this mechanism occasionally has unexpected consequences: a cat that succumbs to a virus may then go off its regular food even after it has recovered, because it has incorrectly associated the illness with the meal that happened to precede it.”1415

    • people ascribe vague & indefinite mental powers to cats and think they are more intelligent than they are, simply because cats’ trial-and-error can go further than one thinks (example).

      for example, mothers can teach kittens simply by interacting with an object, and then the kittens will spend time interacting with it too—not imitating her actions, just acting at random on their own—and may reinvent a reward (in the cited experiment, a food lever), while ignoring unrelated female cats. (I’ve long noticed that if I want a cat to learn something, putting them into the situation works better than trying to demonstrate it. To teach him to use the cat flap, I just shoved him through it several times in both directions. He got the idea.) I was amused to recall some instances of : for example, when I ran out of my original blue bottle of Purina tuna treats and happened to buy various other fish treats, he refused them all, including similar-seeming tuna treats; eventually I realized that it wasn’t the flavor that was the problem, the problem was that the treats were coming out of the wrong container! When I dumped them all into the blue bottle, he was happy to eat them all. I exploited this to train him to tolerate nail-clipping by starting with as little clipping as possible and reinforcing with treats within a second of letting go, and patiently expanding over a year to clipping all of his front claws before treats.

      This use of trial-and-error fuzz testing leads to a signature blend of stupidity & genius familiar to anyone who has spent much time in statistics or AI or computer security—I was amused to think while reading Cat Sense that, as described, cats have all the strengths (& weaknesses) of contemporary AI research’s deep (especially ): generalized, sample-inefficient, demonstrating the perverse genius of trial-and-error, sleep()s all the time while (the door), their poor exploration benefits from expert demonstrations, they are reward hackers (who optimize for getting the treat not what you wanted), hairy to train, black boxes, who are unable to generalize, solve short-term credit assignment problems only, are unable to model or plan, and overly-averse to states which were harmful without realizing which disentangled factor was responsible (cat/​​​hot stove etc), as put it (ch11 of , 1897):16

      We should be careful to get out of an experience only the wisdom that is in it—and stop there; lest we be like the cat that sits down on a hot stove-lid. She will never sit down on a hot stove-lid again—and that is well; but also she will never sit down on a cold one any more.

  • while cats have few abstractions or mental models of the world (and are merely very good at associations/​​trial-and-error)17⁠, they do have mental maps as they can take physical shortcuts which go out of sight of a target or the ‘wrong’ direction for a while

  • cats are deeply inflexible about their territory and not adaptable like dogs, deeply anxious about establishing a territory and stopping other cats from entering their territory, and this appears to be a major cause of cat behavioral problems and disease like, incidentally, cystitis, and according to the UK Cat Protection, inability to get along is one of the most common reasons for a cat to be returned to their pounds. (This point made me think back to when I visited the pound and the cat room: festooned in cat trees and shelves and nooks and crannies, the ~20 cats there had all carefully themselves equidistantly—which made it easy for me to play with each one separately, at least.)

    Famous cat photographer needed to acclimate cats to his photography studio over days⁠, and gingerly them to strobe lights, while his wife, a ‘cat whisperer’ coaxed the into relaxing & posing; Chandoha quit photographing cats after she died. (To highlight how severe the consequences of poor adaptation can be to both the cat and owner, I would have mentioned how frequently cat bites from even extremely friendly cats will send people, including my grandmother, to the emergency room, due to the narrowness of cat fangs & ease of blood infection.) Maren Huck describes her informal observations of her ‘catcam’ research ():

    Cats are seen as relatively lazy, especially compared to dogs. But we saw that when they were outside, they became superalert.18 They scanned their surroundings, sometimes for a half-hour or more on end. And even though cats are highly territorial, they didn’t always fight with other cats they encountered. Often, they just sat a couple of meters away from each other for up to a half an hour. They may have been sizing each other up. Sometimes they would engage in a greeting, briefly touching noses. When they were in their homes, the cats spent a lot of time following their humans around. They liked to be in the same room. A lot of my students were surprised at how attached cats were to people.


    …cats’ emotional needs are still the cause of widespread misapprehensions. Cats are widely perceived as being far more socially adaptable than they actually are. Owners polled for a recent survey said that half of pet cats avoid (human) visitors to the house; almost all pet cats either get into fights with cats from neighboring houses, or avoid any contact with them; and half of the cats that share households with other cats either fight or avoid one another.1 Research confirms that cats find such conflicts highly stressful: they experience fear during the event itself, and anxiety in anticipation of the next encounter. They are constantly hypervigilant through cues we are unaware of, such as the odor of a rival cat. Chronic anxiety can lead to deteriorating health and may reduce life expectancy. Unfortunately, we do not know enough about how to mitigate this situation, made worse by the ever-increasing number of cats kept as pets.

    …Simply replacing random-bred cats with cats from today’s pedigree breeds will not only perpetuate those genetic problems that already exist, it also cannot solve the problems that the cat is facing as a species. A reduced motivation to hunt and kill prey is just one of several factors that will enable cats to adapt better to twenty-first-century living. Allowing a little anthropomorphism: if cats could write themselves a wish list for self-improvement, a set of goals to allow them to adapt to the demands we place on them, it might look something like this:

    • To get along better with other cats, so that social encounters are no longer a source of anxiety.
    • To understand human behavior better, so that encounters with unfamiliar people no longer feel like a threat.
    • To overcome the compulsion to hunt even on a full stomach.

    The corresponding requests from owners:

    • I’d like to have more than one cat at a time, and for my cats to be company—not just for me, but also for one another.
    • I wish my cat didn’t disappear into the bedroom to urinate on the carpet every time I have visitors.
    • I wish my cat didn’t bring gory “presents” through the cat-flap.

    Suffering is all the more cruel when those suffering do not & cannot understand why.

    A lot of otherwise-amusing cat behaviors look like hyper-active fear and danger responses. Consider the famous “cats and cucumbers” prank.

  • Cat territories do not need to be large: I’ve been surprised how rarely I see him venture far from home considering how much undeveloped area there is around my house which should be perfect for exploring/​​hunting, but small ranges are common for domestic cats (eg ), and being less than 91 meters is the usual even in natural areas, something like 10,000 times smaller than a wild or feral cat’s range, apparently without major psychological issue. (I’ve thought of getting a tracking collar because I am curious where he does go, but at >$120/​​year, I’m not that curious.) Bradshaw muses of the tracking data from his own cat that

    Splodge rarely ventured beyond the trees—to my great relief, since there was a busy road not far beyond. He would sometimes remain in the same location for hours at a time, usually one of a few favored vantage points such as a branch of a fallen tree, before moving on to another site or returning home. He rarely seemed to be hunting: occasionally, he caught a mouse or a young rat, but would let birds fly past him without batting an eyelid. I often wondered, and still do, what was going through his mind as he maintained his surveillance of the same small area, day after day, year after year.

    This seems strikingly connected to the lack of plasticity outside kittenhood and inability to befriend other cats as adults. The tragedy of domestic cats is their minds are as rigid as their bodies are flexible. But perhaps if cats are like humans, for whom anticipation is so often better than realization, then being safe while watching the world go by is a kind of heaven.

  • cats prefer their food and water bowls strictly separate; this apparently was well-known among cat researchers (although I can’t find any good experiments) and is explained as a anti-fouling behavior.

    I had somehow never heard of this! My family had always put out pairs of bowls for our cats just like our dogs, I’d always put out pairs for my cats & dogs, it seemed to work fine, the cats didn’t seem to mind…19 I tried separating them for my cat, and sure enough, he stopped dragging chunks of wet food away to eat and seemed to be eating more as well. I tested this further: I’d been adding water to his saucer of wet food because I was concerned he wasn’t drinking enough water & adding water is something some cat owners do, and the saucer sometimes wound up dry, so I thought it was working. But the separate-food/​​water claim suggests it was actually bad. So, I split his food into two saucers with half each and water/​​no-water, and put them down simultaneously (alternating left/​​right); out of 4 trials, he showed a strong preference for the no-water one—he always ate the no-water-added wet food saucer first and left the water-added saucer for eating only hours later (or not at all), dragging food away from the water one or licking the dry saucer instead. Thus, I stopped doing that too.

    • I also read that cats often prefer their wet food microwaved briefly, which makes the smell much stronger. The smell is indeed overwhelming, but in 8⁄10 pairwise trials, my cat preferred the cold wet food he’d had for the past 4 years. (So it may well be true but his preferences are locked-in now.)
  • the coolest term in cat biology is “primordial pouch” (the baggy part of their stomach all cats have); the uncoolest is definitely “Flehmen response” (the facial expression you make when offered sauerkrat while watching men in leiderhosen play Glockenspiels)

  • One undiscussed question: “What is it like to be a cat (tail)?”

    Anyone who has seen a cat (or dog) whack itself in the face with its ⁠, attack its tail after being surprised by it, or lay with their tail lashing only to abruptly lick it (apparently “trying to calm it down”), must have wondered what is going on & if tails have minds of their own. This seems plausible.

    If we look at patients, (whose arms contain 3⁄5ths its neurons, and which when cut off, can still grab food & attempt to pass it back to the mouth), problems like & treatment like (or using VR to induce fake-limb mental ownership), the (who, Peter Watts notes⁠, avoid using the pronoun “we”), the inability of the sane to without ⁠, PRISM⁠/​​​pain etc, one gets the inescapable impression that personal identity is a continuum closely linked to a triangle of latency/​​​bandwidth/​​​predictability: brains perceive subsystems as part of themselves, rather than external objects, only if they can be sufficiently accurately modeled & controlled. At a certain point (akin to VR ), identity “snaps” into place.

    A well-designed tool simply feels unconsciously part of ourselves, while a poor tool constantly causes errors & problems & is hard to control quickly, repeatedly feeling alien to us, forcing its way into consciousness to be laboriously processed (a good tool from to ). Our hands feel so much closer to us than our feet (cf the sensory homunculus), and our eyes feel so much more us than our big toe because it provides such rich information to the brain, with such low latency (being but inches away), and the of our eyes are the fastest and most precise muscular movements in the body—we are entirely unconscious of eye focusing & movement, or the blind spot, because our eyes work so well. The 2 hemispheres of our brain are likewise so tightly linked by such high-speed & high-bandwidth neural connections that we have no conscious experience of laterialization, and cannot even conceive ‘what it would be like’ after they are surgically severed—to be just the left hemisphere or just the right hemisphere. Or more fundamentally, we are unable to notice or perceive the extensive processing brains must do to laboriously extract colors, edges, objects, temporal persistence, even when an optical illusion reveals to us the tricks and we stare at it trying to will it away: you still can’t feel the vision processing as separate from you. It just works too well, too fast, too often.

    How do cat tails fit in with these other examples? A cat tail is as far away from the cat brain possible (a tail might be as long as the body itself!), and is an extension of the neuron-filled spine. A cat tail exists mostly for the sake of balancing (as opposed to being an escape decoy, distraction, stinger, spare hand for tree climbing etc). Balancing in feline acrobatics like pounces is an extremely rapid process which must happen in milliseconds, with potentially fatal errors; there is no time to chat back and forth with the main brain, the tail must be relatively autonomous. (Similar to how even cats can still do things like walk or trot. The peripheral nervous system must be quite capable.) The tail is also quite separate from the rest of the body: tail adjustments have little to do with the neighboring rear legs, eg, much less the rest. A cat tail doesn’t contain remotely as many neurons as an octopus arm does (as far as I can find out—but since octopuses have as cats & put an absurd amount into their limbs, it seems highly likely), but this is a continuum.

    Given the tail’s role and distance, and how personal identity works, it seems entirely possible that a cat doesn’t quite ‘own’ its tail the way it does more tightly-integrated limbs like its forepaws, and genuinely can be surprised by what its tail ‘decides’ to do.

    So, what is it like to be a cat tail? It would be an extremely limited sort of mind (far above a thermostat, but far below the main brain): little or no memory beyond the seconds-scale, with senses confined to the tactile; always trying to find a more comfortable position based on signals about being in equilibrium coming from without (the main brain), sometimes succesfully adjusting within milliseconds, but still periodically bothered by intense feelings of discomfort or predicted pain, which it attempts to fix with new positions (usually futile), until equilibrium signals are restored for unknown reasons.

    Eh. It’s a living.
  • incidentally, here’s one I rediscovered myself (and may have solved): why do cats like human earwax so much?

Are Cats Domesticated?

“Anyone who considers protocol unimportant has never dealt with a cat.”

Robert Heinlein,

This is all well and good, but what I took away from Cat Sense, which I was not expecting, was a deeper appreciation of cat problems and a pessimism. Cats are not acceptable cats. They can be much better.

Schopenhauer on cat humor. “We relate to this, for it is how we appear before the eyes of God.”

Bradshaw’s overall thesis is this: think of a cat as a small solitary desert ambush predator which happens to have some limited cognitive plasticity in kittenhood and some basic social skills enabling it to, on rare occasions, live in unstable ‘colonies’ of usually-related individuals focused around a rich food source; this desert predator happens to have become insinuated throughout human society but it has no real understanding of humans or “theory of mind” and to survive, relies heavily on what learning it manages about humans & other cats in that short window of plasticity, the simple social skills of a cat colony, trial and error, and treating humans as either dangerously unpredictable predators or large mother cats.

As pet cats are universally ‘fixed’, cat reproduction is now primarily done by those feral cats who are able to escape the traps and being neutered, with the most secretive surviving toms endlessly searching suburbia for the rare fertile female in heat. To be reproductively fit, a cat must always be able to survive on its own and evade humans, and remain hyper-alert to foreign cats or predators which might damage it (leaving it unable to hunt effectively on its own causing a downward spiral of starvation), sensitive to dangerous foods & places & people, passing up nutritious food if there’s even a small chance of being disablingly poisonous, and is unable to cope with too much novelty and may be entirely unable to understand humans or other cats if it was not properly acculturated as a kitten during its window of plasticity.

Bradshaw remarks that a cat raised only by humans, without experience with kittens, reacts to meeting another cat for the first time with “a bizarre combination of fascination and fear…Others…barely seem to realize they are cats at all…Hand-reared kittens may develop extreme personalities” (pg236; the word ‘fear’ comes up a lot in Cat Sense), which made me recall that my neighbor’s cat, who was found as a tiny kitten in a cat nest of 5 kittens in a field by a farmer while mowing, and then adopted by them when the farmer took them to his vet. Thus, he was raised solely by humans & dogs; he is Maine Coon-ish, and friendly to humans & dogs, and we often remark that “he seems to think he’s a human”, just as predicted. But he also reacted as just described when I brought my own cat home, who seems to have been the first cat he ever encountered close up, and was an obsessive bully initially—years later, he still seems to both loathe & be fascinated by my cat, while he is much less interested in dogs. (He also declared me dead to him after I brought a new cat home, despite our previously excellent relationship and, almost 7 years later, still refuses to be petted by me.)

In other words, cats are barely domesticated.

Domestic cat admixture appears to contaminate ‘wild cat’ populations21 and vice versa, which aside from slowing any kind of domestication, also indicates the successful flow of ‘domestic’ cats back into the wild. We could also note that before the in the 1940s, keeping a cat indoors was challenging due to the awful stench of their urine & difficulty feeding them (where does one get all the meat if the cat is not hunting for itself and there are no cat’s-meat men?) and most people had outdoor cats, and farm cats were more tolerated than anything. Cats don’t exhibit a clear ‘’ of small skulls, floppy ears, no longer going into heat, piebald coats, etc. (The do; more obscurely, may too, with the “dumbo” mutation⁠.) Mice may be more domesticated than cats—“today’s house mice rarely breed successfully away from human habitation, especially where there are wild competitors, such as wood mice”! They have no theory of mind, and won’t look to a human for help and training must be done with food rewards. A cat without any human contact before 10 weeks will be near-feral (except in the case of severe trauma, apparently, where it may bond with its rescuer, analogous perhaps with “hitting bottom”). This is still better than true wild cats, where even fairly young kittens are aggressive and difficult to handle adults, but it is still indicative of deep fragility, dependent on just the right rearing environment else lapsing back to wild-type. (Puppies, in contrast, are more robust to lack of human experience.)

On a behavioral level, cats aren’t very domesticated compared to even, say, the Russian silver foxes. The behavioral assay for the foxes was whether they would eagerly approach a quiet human; after half a century, they all do and are very friendly. While anyone who has visited many people with cats knows that the modal reaction to a stranger in a house is to run away & hide until they’re gone (and Bradshaw cites a survey to that effect). While looking for a cat after my dog died, I visited the local animal shelter where most of the cats, perhaps a good 30, were all in a play room, and I spent an hour inside it, sitting & waiting for cats to decide to come to me; despite being the only person in the room the entire time (and the only person to adopt a cat that day)—suggesting that they were not exactly blessed with a surfeit of human-recreation opportunities, unlike the cats in the San Francisco cat cafe I visited in early 2018—of that 30+, only a handful did, and I had to approach them to get any impression. I eventually coaxed my future cat out of his hidey-hole to try him, upon which he was very friendly & starved for affection (but in retrospect should’ve been a warning sign to me that I was not picking a bold cat which was well-equipped to deal with the stresses of the modern world).

I also note that most people are bad at dealing with cats, making what should be clear errors (again, despite cats being the #1 or #2 most common pet in the world & so ignorance should not be a problem): I don’t know how many times I have seen someone try to touch a cat’s belly, start petting it with a full-body stroke rather than a chin or head scratch, insist on scratching them at the butt-tail point, interpret tail-lashing (the opposite of dogs, ‘tail wagging’ is a bad thing!), or ears pinned back as good things, try to pick them up, or startle them by abrupt untelegraphed movements. In contrast, while I have seen many people with poor dog manners (such as not presenting their hand to be sniffed or even people oblivious enough to attempt to pet a growling dog with teeth bared), people generally seem to make fewer mistakes, the dogs more clearly communicate with the humans, and the dogs tolerate the inevitable mistakes better (rather than running away or biting).

Apropos of vet visits & dogs vs cats, Kirk 2019 investigated the long-standing disparity in spending on cats & dogs’ care & medicine, and in her survey experiments, found that the ‘willingness to pay’ for dogs but not cats is not because owners just like dogs qua dogs but the difference is mediated largely by a sense of control22—which may sound bad (“human owners are power-mad narcissists!”) but makes perfect sense if we think about the point of this control. It would be irrational to try to take cats to the vet as often as a dog given that we can’t control them & the visit may be completely wasted. We need that control for their own good in a modern context.

So, ‘domesticated’ might not be the right word. Perhaps it would be more accurate to describe cats as ‘half-domesticated’, or merely, ‘tame’. (I increasingly think of cats as acting like small children—with & autism, especially given how much vets restraining cats by ‘squishing’ them under a blanket looks like ⁠. Or would work better?)


The cat status quo.

In Like Engend’ring Like⁠, one of the definitions given of domestication is a creature whose reproduction is controlled by humans; Bradshaw points out that humans have never exercised control over cat reproduction anywhere like we do over dog or cow or sheep reproduction, and what control we did exercise (thereby selecting for domestication), we have wasted on cat-fancier fripperies & follies like coat colors, or forfeited by our otherwise-successful population control measures—abdicating cat reproduction to the worst possible cats, the cats so fearful and averse to humans that they are feral strays who can’t be caught (but which we feed lavishly anyway because when it comes to pets, we “love not wisely but too well”), creating what we might call ‘feline dysgenics’:

We must also ask whether the cat is being inadvertently and subtly altered by those who hold cat welfare closest to their hearts. Paradoxically, the drive to neuter as many cats as possible, with its laudable aim of reducing the suffering of unwanted kittens, may be gradually eliminating the characteristics of the very cats best suited to living in harmony with humankind: many of the cats that avoid neutering are those that are most suspicious of people and the best at hunting. The friendliest, most docile cats are nowadays neutered before leaving any descendants, while the wildest, meanest ferals are likely to escape the attention of cat rescuers and breed at will, thus pushing the cat’s evolution away from, rather than toward, better integration with human society.

…The tomcats must therefore roam as widely as possible, endlessly straining their senses for the yowl and odor of the rare female that is coming into season. Such toms are shadowy animals; some are theoretically “owned”—though their owners rarely see them—and many feral. Because they make themselves inconspicuous except when they have located a prize female, there are probably far more of them than most people realize. When it first became possible to obtain a cat’s DNA fingerprint from just a few hairs, my research team attempted to locate every litter born in homes in a couple of districts of Southampton, UK. From what we’d read, we expected to find that just a few “dominant” tomcats had sired most of the litters in each district; instead, we found that out of more than 70 kittens, virtually all litters had different fathers, only one of which we were able to locate.23

…the widespread adoption of early neutering by the most responsible cat owners risks pushing the domestic cat’s genetics back gradually toward the wild, away from their current domesticated state. A study that I conducted in 1999 suggests that such extrapolation cannot be dismissed as science fiction. [See also Clark 1975⁠.] In one area of Southampton (UK), we found that more than 98% of pet cat population had been neutered. So few kittens were being born that potential cat owners had to travel outside the city to obtain their cats. This situation had clearly existed for some time: from talking to the owners of the older cats, we calculated that the cat population in that area had last been self-sustaining some 10 years previously, in the late 1980s.

We located 10 female pets in the area that were still being allowed to breed and tested the temperament of their kittens after homing, when the kittens were 6 months old. Our hypothesis was that feral males must have fathered many of these kittens, since so few intact males were being kept as pets in the area, and all of these were young and unlikely to compete effectively with the more wily ferals. We found that on average, the kittens in those 10 litters were much less willing to settle on their owners’ laps than kittens born in another area of the city that still had a significant number of undoctored pet tomcats. There was no systematic difference in the way these two groups of kittens had been socialized, and the mother cats in the two areas were indistinguishable in temperament. We therefore deduced that even if only one of the two parents comes from a long line of ferals, the kittens will be less easy to socialize than if both parents are pets. The study was too small to draw any firm conclusions, but in the years since it was carried out, blanket neutering has become more widespread, and so the cumulative effects of this on the temperament of kittens should be becoming more obvious. Neutering is an extremely powerful selection pressure, the effects of which have been given little consideration. At present, it is the only humane way of ensuring that there are as few unwanted cats as possible, and it is unlikely ever to become so widely adopted that the house cat population begins to shrink. However, over time it will likely have unintended consequences.

A similar set of observations is made about dogs as well, by ⁠.24

Where do cats come from? Given that we sterilize almost all our pet cats and hardly buy from cat breeders, pedigree or otherwise, they must come from somewhere.

I don’t know where my family’s two cats or my cat came from, beyond “the animal shelter”; my neighbor’s cat was definitely a feral cat’s offspring; my aunt’s cat was from a pet cat’s litter but almost certainly had a feral father; on the other side, my uncle’s farm cat was definitely a semi-feral cat tolerated for its assumed pest hunting; more pointedly, as far as I know, no one within two degrees of separation of me has ever bought a purebred or ⁠, while I can offhandedly recall 10 purebred dogs which were bought specifically from dog breeders and 3 or 4 of which were even ⁠. (I eventually asked my grandmother, “has anyone in our family ever bought a pedigree cat, or from a cat breeder at all?” She could think of no examples either and agreed that there were at least a dozen dogs bought from breeders.) Those dogs definitely were not accidents or fathered by stray feral dogs. I do not know how much reproduction feral cats account for or how much de-domestication it is responsible for, but it does seem like it could be a lot, and could be enough to drastically slow any domesticating process or even reverse domestication. Bradshaw’s estimate of only 15% being “planned mating” (apparently inferred from ‘total cats—cat breeder sales’ estimates, which is, if anything, a loose upper bound as cat breeders do not necessarily breed for much or well) seems reasonable to me, which allows great scope for bad effects. (Indeed, given the neutering & lack of breeding, how could there not be dysgenics?)

How would we know if cat domestication was reversing? Veterinary medical science has advanced enormously over the past century and spending on pet medical treatment increases even faster than human medical treatment (doubling since 200025) so we would not expect any clear long-term trend (perhaps the health gains are eaten by cat dysgenics), the environment has changed enormously (consider urbanization or the several-fold increase in suburban house sizes) which makes comparisons harder, and no one maintains systematic records of cat behavioral problems or health to do such comparisons in the first place. Pedigree cats are presumably immune if breeders are being honest and not enrolling half-stray matings as purebloods, and there is stabilizing selection for all pedigree cats in the sense that cat breeders will not breed or buy cats which cannot endure being transported to cat shows & sitting in a cage in the middle of thousands of cats for hours & being judged. (Are the ancient city cats of Istanbul, recently made famous by the 2016 documentary ⁠, so friendly to strangers because they are rarely neutered and to a considerable extent hand-fed by locals/​tourists & do not subsist purely on hunting?)

So there should be a slowly growing gap over time—but no one is formally measuring pedigree cats against regular pet cats or feral ancestry admixture, and some pedigree cats are crazy themselves (eg Siamese cats). And there are no large cat genetic datasets which could be examined to see if the domestication has been decreasing recently or is lower than in ancient cat DNA (of which there is a smidgen).

We are now willing to put up with far more than our ancestors just a century ago will: only a monster would euthanize their fur-baby rather than spend $5k on medical treatment or therapy. (‘fur-baby’ here is not a joke and $5k is not a hypothetical but based on a real case: that is the total spending of my neighbor on her Labrador’s last 2 years, who had terminal cancer, whose treatment involved, among other things, amputating her leg to buy a few months; after she finally died, my neighbor was so distraught she was prescribed sedatives. And as mentioned, my aunt for years spent $1k a month on drugs for her dog, so the lifetime total there scarcely bears thinking on.)

So in other words, if cats were steadily de-domesticating and becoming sicker, between all the environmental interventions and additional spending & tolerance for crazy cats, the world would look… much as it does now.

What is to be done?

The simple answer for feline dysgenics is to simply select the other way. There’s no canine equivalent because we control their reproduction and get all our dogs from either breeders or from domestic dogs being permitted to mate by lax owners, both of which inherently select against neurotic or unfriendly dogs; in most places in the West, there are no strays to speak of and they certainly don’t make up the overwhelming majority of reproduction. (Dogs have other genetic issues, largely stemming from recent population bottlenecks, which would benefit from better & more systematic breeding, but there is no threat of them de-domesticating.)

We are not trying to select on subtle or hidden traits here—it is clear to owners how fearful of strangers or affectionate a cat is, and those traits are definitely heritable (Braastad et al 1999⁠, among others).

Bradshaw notes that one way to get started would be to more deliberately aim for : choose an already highly domesticated breed, such as the or or ⁠/​⁠, the last of which, perhaps because of pleiotropy or perhaps because Ragamuffin breeders say “The only extreme allowed in this breed is its friendly, sociable and intelligent nature”, are so friendly Bradshaw mentions they aren’t allowed outside (because they will be too naive & be attacked by other more aggressive/​territorial/​paranoid cats); another point in Ragamuffins’ favor is their many coat/​eye colors, and their large size, which might be useful for fixing their kidney vulnerabilities. Another interesting possibility is to tap into the many feline hybrids, to greatly increase the amount of genetic diversity available to select on—probably more useful for health issues than domestication itself.

Pedigree cats sometimes have health problems26⁠, but they appear unrelated to the domestication per se, and due to inbreeding & population bottlenecks (eg Makino et al 2018⁠; many pedigree breeds trace back to a handful of cats or just one cat), and limited use of more advanced breeding techniques like direct genetic testing; all of this could be avoided by starting with a large diverse founding population of a few hundred cats, and focusing their selection on the important things like health.

Another objection people raise against this is that they don’t want cats who are too “dog-like”. I sympathize with this and acknowledge that Ragamuffins might be too dog-like for many people, but I do not think that things like stressing oneself to death via cystitis or being terrified of strangers are intrinsic to cats’ appeal, nor do I think any owner actively desires those things, and it should be possible to improve social skills & plasticity & anxiety while preserving the things we value about cats, like their perennial curiosity, watchfulness, clever trial-and-error, enjoyment of playing chase, purring etc without having to keep their problems like exploding kidneys or adult cats’ inability to befriend.

Jorge Luis Borges comments on cats and men.

Bradshaw is pessimistic that the cultural norm of getting cats for free and not paying for them like we do dogs will be impossible to break. How can high-quality cats outcompete ‘free’ kittens (even ones which are ticking time bombs)? How can we get people to see that there is any problem, or even that there might be a problem & we need to research cats more rather than leaving cat research the crumbs from dog research? While it may be difficult to change cultural patterns, it is not impossible: the social stigma among the middle/​upper-class of buying from “puppy mills” did a number on them and successfully shifted the Western norm to buying either directly from the dog breeder when a specific breed is desired (eg for hypoallergenic) or getting pets from the pound; admitting to peers that your new puppy was bought from a ‘puppy mill’ will earn one the stinging rebuke of glances askance, shuftis, and raised eyebrows. Many other practices were changed for reasons like their purported environmental friendliness, and I see no reason why one could not breed a better more domesticated cat which hunts less and, entirely truthfully, flip the environmentalist anti-cat narrative on its head. One could without much trouble work out how long a breeding program would take given the preliminary observed heritabilities and behavioral distributions (eg for catnip response, I estimate it ought to take under 10 generations in the worst case of random initial selection of cats in order to make catnip response essentially universal, which since cats mature sexually at ~1y, means hardly a decade).

Rather, the problem is no one considers it a problem. Reading Cat Sense, I was repeatedly struck by a double-standard: “if cats were the size of large dogs and acted like normal cats do now27⁠, they would be considered more dangerous than Rottweilers or Dobermanns and feared & outlawed” (indeed, many hybrids are outlawed in many places); “if a dog breed were as unhealthy, neurotic, unable to adapt, and stressed out by interaction to the point of routine life-threatening kidney failure, as normal cats are now, buying such a dog would be considered more immoral than buying a pug or English bulldog is now”; and so on. But because they are cats, it’s taken for granted and just the ‘catus quo’. (“Oh cats—isn’t it so funny how cats spend all that time staring out the window? Or won’t stay in the same room as the family dog? Or hide whenever someone visits? Or pee in your bed? Or bite you for no reason? Adorable!” No. No, not really.)

Overall, I learned much more than I expected from Cat Sense, and while much of it was not good news, some of it was useful, and I now find it easier to understand & forgive cats. I truly believe, that, in the strange artificial (often all too small) worlds in which they have been brought, so far from their ancestral land, filled with confusing clamorous clumsy giants, whether biting the hand that feeds them or lolling in the light or patiently observing, they, to the extent a breast filled with the heart of a small solitary desert predator little meant for such things can, love us. They are not at fault; we are.


Relevant cites in rough order of appearance in Cat Sense:

Further reading:


Fuzz Testing

“If builders built houses the way programmers built programs, the first woodpecker to come along would destroy civilization.”

(attributed in Conrad Schneiker 1975)

Software is poorly tested, and not robust even after decades of development using techniques like “fuzz testing” (bruteforcing random inputs to trigger problems); as an example, I log all instances of a different kind of fuzz testing—when my cat walks over my computer keyboard and causes problem! This has done everything from frozen my monitor to deleted files to segfaulted statistical software to both DoSed & crashed X. Other cats have shut down datacenters, crashed Mac/​​​Window/​​​Ubuntu systems by pressing a button too long, broken printers (somehow), and watched cat videos on YouTube.

Things my cat has done walking across my keyboard:

  • ordered extra socks from

  • ‘favorited’ tweets or deleted draft tweets on Twitter

  • clipped whole pages in Evernote

  • deleted a dozen photos

  • turned off sound

  • chatted on IRC

  • killed Emacs

  • killed Firefox

  • crashed R inside Urxvt

    • crashed Stan inside R inside Urxvt
  • made Urxvt unusable (terminal escapes?)

  • deleted directories of docs (but fortunately, versioned)

  • hard-locked (by opening arbitrarily many windows)

  • permanently disabled weather forecast widget in MATE status bar, which remained broken even after being manually re-added

  • crashed X (the Synaptic driver for my Acer laptop would segfault whenever he stepped on the trackpad; I never figured out why but chalk it up to perhaps he was touching it in too many spots and triggered bugs)

  • turned off monitor via keyboard (replicating this, I disabled my monitor a second time but it didn’t come back up when I pushed the hotkey apparently responsible; I had to push the power button on the PC itself to bring it back up, and network was disabled afterwards; a check of dmesg suggests that the system had fully suspended and apparently that can’t be undone from the keyboard?)

  • made external monitor a mirror display

  • turned off external monitor, forcing reboot of MATE to restore access to both external & laptop monitors

  • shut down laptop

Things other peoples’ cats have done:

  • crashed a datacenter:

    In March 2012, Google responded to an unusual power outage at one of its Belgian datacenters that ultimately led to local data corruption. Investigation revealed that a cat had damaged a nearby external power supply, triggering a series of cascading failures in the building’s power systems.

    Eaton’s Blackout Tracker Annual Report for 2013 notes additional power-related examples:

    On Jan. 20, a cat shorted out a Philadelphia transformer. The four-legged feline caused power to go out in all three buildings of the West Park Housing Complex…On Nov. 11, a cat got into a place where it shouldn’t have been and caused an equipment failure, knocking out power to 1,800 in Harper’s Ferry.

  • broke Macbook laptop screen by laying on it

  • disabled Mac, Window & Ubuntu OSes by apparently pressing & holding the “Print Screen” button long enough

  • damaged NASA’s ultra-smooth by playing on it at night

  • Twitter thread: 1 examples include: printed 58 pages of a Twitter feed; flipped laptop display & changed language to Arabic; delete everything on desktop, open YouTube to cat videos, & texted mother via WhatsApp; broke a brandnew printer

Stan segfaulting. Somehow.

This , aside from demonstrating that the first AI with the intelligence of a cat will take over civilization (crashes like Stan/​R/​X simply should not happen in the first place much less be triggered by legal input like random keypresses), also shows a lot of bad user-experience (UX) design in FLOSS programs. Here, the problematic software includes ⁠, Urxvt, MATE shutdown interface, MATE external monitor handling, and the Twitter UI. Destructive irreversible operations like shutting down a computer should require prompts which cannot be faked by a cat sitting on a keyboard; in the case of MATE’s external monitor handling, it is reversible by the same key pressed—but only partially, as MATE forgets the external display settings like the orientation & relative positioning of the two monitors, requiring me to again manually set all the options.

Perhaps more tech companies & software developers should do fuzz testing.


Are we doing cat toys wrong? Some research suggests that cat play is inherently about hunting simulation, and using a static toy fails to imitate the consummation of a successful hunt, and is unsatisfying. ⁠, Hall et al 2002:

We have investigated the role of habituation and disinhibition in the control of object (predatory) play by adult domestic cats Felis silvestris catus both with and without prior experience of hunting. We hypothesised that object play is terminated by rapid habituation to the sensory characteristics of the object played with, and therefore should be disinhibited if the sensory characteristics of the object are changed. Three sequential sessions of play with an unchanging object (a toy) caused almost complete habituation of the play response; replacing the toy with one of contrasting colours in a fourth session elicited intense disinhibited play, suggesting that motivation for play itself had not diminished substantially during the first three sessions. The time interval between sessions affected the extent of disinhibition. After a long delay (25±45 min) between each session play was less intense in the fourth session than in the first; if the interval was 5 min, it was more intense, indicative of post- inhibitory rebound, possibly caused by initial positive feedback of play on its own performance. We suggest that object play by adult cats is controlled by two mechanisms derived from predatory behaviour: one responds to prey-like stimulus characteristics, such as texture and small size, which elicit play, while the second detects change in the toy. The behavioural default towards any object is initial interest if it possesses relevant stimulus characteristics, followed by rapid habituation unless these stimulus characteristics change.

Extended discussion by Bradshaw from Cat Sense 2013:

My graduate student Sarah Hall and I found that habituation is the main underlying reason for this apparent boredom. We presented cats with toys—mouse-sized, fake-fur-covered “pillows” tied to a piece of cord—and at first they usually played intensely, appearing to treat the toy as if it was indeed a mouse.

However, many cats stopped playing within a matter of a couple minutes. When we took the toys away for a while and then presented them again, most of the cats started playing again, but neither as intensely nor for as long as the first time. By the third presentation, many of the cats would scarcely even begin to play. They clearly became “bored” with the toy.

If we switched the toy for a slightly different one—a different color (say, black to white, since cats’ perception of colors is different from ours), texture, or odor—almost all of the cats would start playing again. Thus, they were “bored” not by the game, but by the toy itself. In fact, the frustration of being offered the same toy repeatedly actually increased their desire to play. If the interval between the last game with the original toy and the first game with the new toy was about five minutes, they attacked the second toy with even more vigor than they did the first one.28

To understand why playing with a toy would make a cat frustrated, we considered what might motivate cats to play in the first place. Kittens sometimes play with toys as if they were fellow kittens, but adult cats invariably treat toys as if they were prey: they chase, bite, claw, and pounce on toys just as if the toys were mice or rats. To test the idea that cats think of toys in the same way they think of prey, we tried different kinds of toys to see which ones cats prefer. Our findings showed that, unsurprisingly, they like mouse-sized toys that are furry, feathered, or multi-legged-toy spiders, for example. Even indoor cats that had never hunted showed these preferences, so they must be hardwired in the cat’s brain. The cats played with rat-sized toys covered in fake fur in a subtly different way from the mouse-sized toys. Instead of holding them in their front paws and biting them, most cats would hold the rat-sized toys at arm’s length and rake them with their hind claws—just as hunting cats do with real rats. The cats were apparently thinking of their toys as if they were real animals, and as if their size, texture, and any simulated movement (such as our pulling on the toy’s string) had triggered hunting instincts.

We then examined whether a cat’s appetite has similar effects on the way it hunts and the way it plays with toys. If cats play with toys just for their own amusement, as many people assume they do, then they should be less inclined to play when they are hungry, since their minds should be focused instead on how to get something to eat. Conversely, as a hunting cat gets hungrier, it will hunt more intensely and become more inclined to take on larger prey than usual. We found exactly the latter when we offered toys to our cats. If their first meal of the day had been delayed, they played more intensely than usual with a mouse-sized toy—for example, biting it more frequently.

Moreover, many of the cats that normally refused to play with a rat-sized toy at all were now prepared to attack it. 6 This convinced us that adult cats do think that they are hunting when they’re playing with toys.

Cats don’t easily get “bored” with hunting, so we were still puzzled as to why our cats stopped playing with most toys so quickly. Indeed, they appeared to get “bored” with most commercially available toys and with the kinds of toys we made for our first experiments. The few toys that sustained our cats’ interest all shared one quality: they fell apart as the cat was playing with them.29 Although we had to abandon experiments that involved these toys, which came apart at the seams as our cats batted them about, we noticed that several of the cats were extremely reluctant to give them up. We then realized that our original swapping experiments mimicked one aspect of what happens when a cat rips a toy apart: when we exchanged the toy for a slightly different one, the cat’s senses told it that the toy had changed. It didn’t seem to matter to the cat that it had not caused the change itself; what was important was that a change seemed to have occurred.

We deduced that not only do cats think they are hunting when they’re playing with toys, but their behavior is being controlled by the same four mechanisms whether they’re hunting or playing. One of these mechanisms is affected by hunger, and the same one that makes a cat more likely to play with a toy makes it likely to make a kill when it’s hungry.8 The second is triggered by the appearance—and presumably the smell and sound—of prey, and certain specific features, such as fur, feathers, and legs, that the cat recognizes instinctively are likely to belong to prey animals. The third mechanism is affected by the size of the toy or prey. Attacking a mouse puts the cat in much less danger than attacking a rat, so the cat attacks the rat much more carefully; likewise, cats treat large toys much more circumspectly than small toys, as if they were capable of fighting back. Even though cats should quickly learn that the toys are unlikely to retaliate, most cats don’t seem to do so. The fourth mechanism is the source of the cat’s apparent frustration: if all that biting and clawing doesn’t seem to have any effect on its target, then either the target wasn’t a meal, or if it is prey, then it’s proving difficult to subdue. A toy that starts to disintegrate, or is taken away but looks different when it comes back (as in our original experiment), mimics the early stages of a kill, thus encouraging the cat to persist.

Most cat toys don’t change. Is this a serious problem? What about a color-changing moving ball? It could change colors and turn red when it ‘dies’. It could then follow a different movement pattern when it comes ‘back to life’, cycling through a bunch of procedurally-generated patterns… Maybe something like a Sphero with different-colored LEDs in it? When we use a laser pointer, should we always end the game by putting a treat on the floor and leading the cat to it?

  1. Particularly striking in Reich’s case as his book sent population geneticists into a frenzy by revealing unpublished secrets of the Reich lab, but such is the pace of ancient human genetics research 2010–2020 that by 2019, it was already… ancient history.↩︎

  2. An example: -response was a Mendelian dominant genetic trait for 50 years, based on a cursory analysis of in a single paper back in the 1960s, until (bonus nominative determinism: the graduate student in question worked at the lab of Leslie Lyons) to check with more plausible contemporary genetic models, and found catnip-response is just an ordinary polygenic trait.↩︎

  3. Where ‘urinary’ means ‘pee’, ‘cystitis’ mean ‘swollen bladder’, and ‘idiopathic’ means ‘dunno why’.↩︎

  4. Given Bradshaw’s comment about cats going off their feed due to misattributing a health issue to the food they eat, I wonder if there was a hidden recurrence of cystitis?↩︎

  5. To rant a little about this topic: Purina’s salmon pate turns out to be oddly hard to get at a reasonable price, despite being included in their standard “Seafood Variety” pack.

    I can buy the seafood variety pack locally no problem, but not an all-salmon pate. Online, all-salmon packs exist but are easily 2–3x the cost—unacceptable.

    Finally, after a good deal of hunting online for alternatives, I found offered them (batches of solely salmon pate) at hardly a 15% markup and as a monthly subscription as well—perfect! I was pleased with myself. I signed up, set up a subscription, and got a shipment once a month with no additional effort, my cat was much happier, less food was wasted, and it cost about the same. And it did indeed work perfectly. For about 3 months.

    Yu know what Petco decided to do? They banned all Purina products from their store⁠, on the grounds of banning “all dog and cat food and treats that contain artificial ingredients”. So instead they shipped me a package of ‘organic’ salmon pate which cost like 50% more and my cat instantly hated and would eat only after starving overnight. (Yeah, screw you and your virtue-signaling pseudo-science bullshit too, Petco. Is there even a single real study demonstrating all-cause mortality changes? I doubt it. And you know what’s worse for my cat than ‘artificial ingredients’? Eating food he hates.)

    So after all that, it wound up being an almost total waste of time: I then had to go and cancel, and find some alternative. I haven’t found any alternatives, so as of June 2019, I simply go to my local Walmart and buying every single salmon pate they have in stock.↩︎

  6. Concerningly, finds that using puzzle feeders to provide all a cat’s food (as opposed to treats), was the only intervention (out of 5 interventions) to actually increase wildlife predation. This suggests that there is a limit to how much puzzling/​​hunting a cat wants to do, beyond which is stressful and hunger-inducing. (Perhaps because ?)↩︎

  7. Judging from descriptions, probably a blue or purple laser pointer is best. I did get a purple one, but it was not as bright as the others and less reliable. Bradshaw notes that in exchange for the loss of color, cats appear much more attuned to shapes—so perhaps what laser pointers need is to be able to change the shape of the dot, with a rotating filter/​​cutout?↩︎

  8. Why do cats & dogs eat grass or other plants, and occasionally barf? (My cat does this so regularly that I keep an eye out when I note him eating rass outside, lest he barf on my keyboard/​​monitor from his perch.) Are they sick, or their diet nutritionally-deficient? The best theory so far is that it’s simply : leaves & grass help physically push any intestinal worms out.

    See ⁠, Huffman et al 1996/​​⁠, Huffman & Caton 2001; ⁠, Sueda et al 2008/​​⁠, Hart 2008; ⁠, Hart et al 2019; ⁠, Hart & Hart 2018. ( tests out a rival explanation for plant-eating, that it helps expel hairballs, and finds little relevant correlation in snow leopards.)↩︎

  9. A (apocryphal?) veterinary medicine class handout summarizes cat handling for vets:

    The cat is faster and has sharper teeth and nails than you do. It has no ‘code of ethics’ or considerations for its own future. In a fair fight it will win.

    3. USE DRUGS
  10. Suggestions to use gabapentin on cats appear to be based on van Haaften et al 2018 where the gabapentin reduced a mean “fearful” response to merely “very tense”, although I was particularly struck by how they note—which apparently is just as expected to them & merely a detail—that a fifth of the pet cats could not be examined by the vet without being drugged, so traumatizing is the experience:

    The veterinarian was able to complete the physical examination on at least 1 visit for 19 of the 20 cats. One cat could not be removed from the carrier after either treatment because of aggression. For 4 cats after placebo administration, the examination could not be completed; however, after gabapentin administration, the examination could be completed.

  11. Perhaps related to emotion reading: finds subjects who rate black cat photos as being hard to emotionally read also rate them as being less friendly & more aggressive, and less ‘adoptable’.↩︎

  12. Suppose I come across him peeing in my bath tub and spritz him with a spray I have in the bathroom. What would he learn? poses a similar question ():

    A certain mother habitually rewards her small son with ice cream after he eats his spinach. What additional information would you need to be able to predict whether the child will:

    1. Come to love or hate spinach,
    2. Love or hate ice cream, or
    3. Love or hate Mother?

    If I spritzed him would he learn to: fear & avoid the spray bottle, fear & avoid the bath tub, fear & avoid the bathroom, fear & avoid me, fear & avoid urinating—or all of the above?↩︎

  13. I have been experimenting with clicker training for getting him into his cat carrier with mixed success. (He is not a smart cat.)↩︎

  14. This phenomenon apparently can cause some odd market dynamics:

    But sometimes the Amazon app, acting as a Geiger counter of consumer demand, will light up on something strange, and it’s time to chase a product. [Freelance product buyer] Anderson recently hit half a dozen Walmarts buying Game of Thrones Oreos. Baron discovered the Oreos, too: “We had to hustle really hard, just driving from city to city, filling up the vehicle with every one of these Oreos we could get.”…Discontinued nail polish, Pop-Tarts, hair curling products: Anderson has chased them all when the scanner has shown them fetching multiples of their normal price. He once hunted a particular brand of discontinued dental floss across the Big Lots of America, buying six-packs for 99 cents and selling them on Amazon for over $100 apiece.

    He has no idea why someone would pay so much for such things, but the scanner tells him people do. His best guesses are melancholy ones. Discontinued cat food is a big seller, which he didn’t understand until his mom’s cat grew old and senile and refused to eat any of the new flavors. He once saw a post from a parent whose son was autistic and drank from the same plastic cup every day for 20 years. The cup eventually disintegrated, and he didn’t want to drink from any other vessel. “I’ve always wondered if it’s something like that”, Anderson says.

  15. While plausible & matching my experience & many other anecdotes, Bradshaw provides no primary reference for the existence of this specialized food-phobia learning mechanism, and tracing citations only turned up an echo chamber of secondary/​​​​tertiary references. So this claim seems to just be conventional wisdom and not based on any specific experiment/​​​​research.↩︎

  16. If you’re wondering about the exact citation for this famous and oft-misquoted Twain quote and you diligently checked Following the Equator and noted that it actually ascribes the chapter epigraphs (of which the cat/​​​​stove quote is one) to a Puddn’head Wilson’s New Calendar, fear not—Puddn’head Wilson was just a pseudonym for Mark Twain for publishing a few popular collections of folksy almanac-style wisdom⁠, and this ‘new calendar’ was apparently only in Following the Equator.↩︎

  17. I noticed it seemed to take him a long time to realize I controlled the laser pointer’s dots, and I’m not sure how much he grasps that the laser pointer causes the dot across the room.↩︎

  18. I can confirm that while I have seen many cats doing many things outside, whether rolling around or laying down or rubbing their backs on the ground or hiding in tall grass or stalking me, I have yet to see one actually asleep.↩︎

  19. Which illustrates the problem. To paraphrase a anecdote:

    Wittgenstein: “But what would it have looked like if it had looked as if cats were unhappy with water bowls near their food or water being added to their wet foods?”

    Me: [sad puppy dog eyes while whimpering at him]

    Wittgenstein: “Exactly!”

  20. “Cats: A Scientific Inventory”
  21. Bradshaw’s list: 8 of 24 in captured South African wildcats genetics; 10 of 12 with suspiciously domestic behavior in zoos; 5 of 7 in Mongolia; & a third of French wildcats.↩︎

  22. See previously Jones & Hart 2019 on emotion readability & adoptability.↩︎

  23. This appears to be further unpublished research, as Bradshaw gives no footnote or reference to it, and Google Scholar turns up nothing which looks relevant.↩︎

  24. Dawson et al 2019a note that current practices are potentially dangerous on the genetic level: commercial & hobbyist dog breeders tend to select dogs for future breeding when quite young (when behavioral measures are even more unstable & unreliable than usual, see also dog behavioral heritabilities) and neuter the rest, while many of the rest of dogs are bred by amateurs in the general population who select even more haphazardly. Even on an environmental level, puppies (like kittens) are heavily influenced by their rearing environment and need the proper stimuli to develop their genetic potentials (as demonstrated by various behavioral genetics research, particularly Scott & Fuller 1965), which may not be adequately provided by either commercial or hobbyist breeders as compared to being reared in ordinary households similar to the ones they are destined to live in as pets.↩︎

  25. If this seems improbable, think about the greatly intensified medical care and the increasing willingness of people to pay luxury-prices for pets. My aunt once mentioned that the drugs for her little dog, who despite coming from a breeder had always been messed up with a number of problems, cost $1k—per month; I was flabbergasted that even if she could easily afford it, she would.↩︎

  26. For example, I notice in one of the few large datasets of cat survivals by breeds, a Swedish pet life insurance database for presumably mostly pedigree cats, Ragdolls/​​Persians/​​Siamese appear to have the highest mortality rates: “Mortality of Life-Insured Swedish Cats during 1999–2006: Age, Breed, Sex, and Diagnosis”⁠, Egenvall et al 2009. Did their selective breeding from a handful of individuals come at a steep genetic cost for overall health?

    On the other hand, the authors note that it’s hard to be sure what the numbers mean—the Maine Coons/​​Norwegian Forest Cats in their dataset live longer but then often die from falls/​​accidents, consistent with being more likely to be outdoors cats, and as already noted, Ragdolls/​​Ragamuffins aren’t supposed to be allowed outside (and Persians/​​Siamese presumably are much more likely to be kept indoors as well), so perhaps the real meaning is that living indoors is bad for cats (due to overfeeding/​​lack of exercise)?↩︎

  27. Although cats of more leonine proportions might act even more aggressively than they do now—how much of the modal cat reaction of ‘running away’ is simply because they realize how much smaller they are? They’re still highly territorial…↩︎

  28. Sarah L. Hall, John W. S. Bradshaw, and Ian Robinson, “Object Play in Adult Domestic Cats: The Roles of Habituation and Disinhibition”⁠, Applied Animal Behaviour Science 79 (2002): 263–71. Compared to “classic” habituation as studied in laboratory rats, the timescale over which cats remain habituated to toys is very long-minutes, rather than seconds. We subsequently found that the same applies to dogs.↩︎

  29. Commercially available toys don’t come apart for a good reason: occasionally a cat can choke on a piece of toy, or get fragments lodged in its gut.↩︎