Your cat climbs onto your chest, treads your sweater with alternating paws, rumbles like a small engine, and finishes by driving her forehead into your chin. Three behaviours, three different explanations, and only one of them means what most people think it means.
Some of this is well-established science. Some of it is reasonable inference. And some of the confident claims circulating online are not supported by much at all, which is worth separating out.
Purring is not a happiness meter
The single most useful correction about purring is that it is not exclusively contentment. Cats purr while being petted, yes, but they also purr while injured, while being examined by a vet, while giving birth, and sometimes while dying. Any veterinary nurse can tell you about a cat purring steadily on an exam table in obvious distress.
The current best reading is that purring is a signal of low-threat intent and a request for proximity or care, rather than a readout of mood. A kitten purrs while nursing and a queen purrs back; the sound functions as reassurance in both directions. That interpretation covers the awkward cases that pure happiness cannot.
Practically, this means you cannot use purring to rule out pain. A cat who is purring but also hiding, eating less, or sitting in a hunched posture needs attention regardless of the soundtrack. The purr is not a clean bill of health, and body language carries far more reliable information.
How a purr is actually made
This is a genuinely open question, and it got considerably more interesting in 2023.
The long-standing explanation was active muscle contraction: rapid, cyclical twitching of the laryngeal muscles, driven by continuous neural input, opening and closing the glottis around 25 to 30 times a second. The puzzle was always the frequency. A cat’s larynx is small, and small vibrating structures normally produce high-pitched sound, not a rumble at 25 Hz.
Then Herbst and colleagues published work in Current Biology using excised larynges from eight domestic cats. All eight produced sustained oscillation in the 25 to 30 Hz purring range with no neural input at all. Histology found masses of connective tissue up to 4 millimetres across embedded in the vocal folds, which the authors nicknamed purring pads and suggested could allow those unusually low frequencies passively.
Be careful how far you take this. The researchers were explicit that their data do not fully reject the active-muscle hypothesis, and these were excised larynges rather than living cats. Press coverage declared the old theory debunked; the paper was more modest, showing that a passive tissue mechanism is plausible and could contribute. The honest summary is that cats may purr with less active neural control than we assumed, and the question is not closed.
One related claim deserves pushing back on. You will often read that purring at around 25 Hz heals bones and accelerates tissue repair. This is an extrapolation from laboratory work on low-frequency mechanical vibration applied to bone, not a finding about cats. Nobody has demonstrated that a cat’s own purr heals the cat, let alone the person it is sitting on. It is a nice idea with no direct evidence behind it.
The purr that was engineered for you
Some purrs are aimed squarely at human beings, and there is good evidence for it.
Karen McComb’s team at Sussex, publishing in Current Biology in 2009, recorded purrs from ten cats in food-soliciting and non-soliciting contexts and played them to 50 human listeners. Listeners consistently rated the food-seeking purrs as more urgent and less pleasant, and the difference tracked an unusual high-frequency component embedded in the low rumble, peaking somewhere between about 220 and 520 Hz. Even people who had never owned a cat heard the difference.
That embedded element sits in a range humans associate with an infant’s cry, which is a hard sound for our species to ignore. McComb’s suggestion was that a cat has good reason to bury the cry inside a purr rather than simply yowling, since outright demands at 5am tend to get you shut out of the bedroom. The behaviour also appeared most often in cats living one-to-one with their owners rather than in large households, which fits a learned, relationship-specific tactic.
So if you have ever felt that one particular purr is somehow unignorable, you are responding to a measurable acoustic feature rather than imagining things.
Kneading: the leading theory, and the one that is shakier
Kneading, or making biscuits, is the rhythmic alternate treading of the front paws against a soft surface, often with a glazed expression and sometimes with drooling or suckling at the fabric.
The best-supported explanation is that it is retained nursing behaviour. Kittens tread the queen’s body while suckling, which helps stimulate milk flow, and the adult version looks nearly identical. The drooling and fabric-suckling that often accompany it strengthen that reading considerably, since both belong to the nursing complex rather than to anything else a cat does.
The other explanation you will see everywhere is scent marking via glands in the paw pads. Cats do have scent glands between the toes and they do deposit scent when they scratch. But applying that to kneading is less settled, and at least one behaviour specialist draws a clear distinction: kneading happens when a cat is relaxed and settled on a lap, whereas scratching a tree trunk or a door frame is the behaviour more clearly tied to depositing scent. Treat the marking explanation as possible rather than established.
What you can take as reliable is the context. Kneading happens when a cat feels safe. It is one of the better indicators you have that a cat is genuinely comfortable with you, which is useful information with a new or nervous rescue.
Head-butting and the shared smell of home
Bunting, the deliberate press of the forehead or cheek against you, belongs to a family of behaviours researchers call allorubbing: cats rubbing face and flank along other cats, people, and objects.
Crowell-Davis and colleagues, reviewing feline social behaviour, argue that allorubbing probably serves several functions at once, with the intense face-to-body contact facilitating the exchange of scent between individuals. Their proposal is that cats in a colony build a shared colony odour maintained by these exchanges and by frequent mutual sniffing. Allorubbing is typically preceded by a tail-up approach and is directed at preferred associates rather than distributed at random, and it occurs more often in feral cats than in housecats.
Work on captive African lions supports the social reading: head rubbing there was reciprocal and fitted a social-bond explanation better than alternatives based on status display or tension reduction.
So when your cat head-butts you, the most defensible interpretation is that she is treating you as a colony member and folding your scent into the household’s. It is an affiliative act directed at someone she has chosen, which is a reasonable thing to feel good about.
What to do with this
Three small things follow. Do not treat purring as proof a cat is well, because it is not. Treat kneading as a trust signal and let it happen, trimming claws rather than pushing the cat off. And lean into bunting by offering a hand or a knuckle at cheek height, which lets a cat complete the greeting on her own terms.
All three behaviours are social overtures from an animal still widely described as aloof. A cat who purrs at you, kneads your lap, and marks you with her cheek glands is doing the feline equivalent of a long conversation. Give her plenty to do between those conversations, because a well-occupied cat is a more sociable one, and the overtures get more frequent.
