Chapter 1 — The Instrument of Taste

Chapter 1 — The Instrument of Taste

"The tongue certainly plays a prominent part in the mechanism of degustation — for, being endued with great muscular power, it enfolds, turns, presses and swallows food." — Jean Anthelme Brillat-Savarin, The Physiology of Taste, 1825 (trans. Fayette Robinson, 1854)

Put a finger to your own tongue for a moment — the reader is permitted a little indecorum in the privacy of a library chair — and consider what you are touching. It feels, to the finger, like nothing more than damp velvet: a muscular slab, pink and unremarkable, doing nothing in particular. And yet this modest organ, weighing a few ounces and rarely admired even by its owner, is one of the most heavily instrumented surfaces on the human body: a landscape stippled with thousands of microscopic sensing organs, wired by three separate cranial nerves to a brain that stands ready, within a fraction of a second, to issue a verdict of pleasure or alarm. It is time this instrument received the close, unhurried examination a naturalist would give any specimen of comparable ingenuity — and time, too, to correct a diagram that has misled schoolchildren for the better part of a century.

The tongue, examined

Run the same finger now toward the tip and sides of the tongue, and the velvet resolves, under a lens, into something considerably stranger: a dense stubble of small projections called papillae — from the Latin for a nipple or pimple, and no more dignified for the etymology.[^1] Papillae come in four distinct patterns, and it is the first order of business to keep them straight, for a great deal of confusion in the popular understanding of taste flows from muddling them.

The most numerous by far are the filiform papillae — slender, cone-shaped, arranged in rows that sweep backward across the whole surface of the tongue like wheat laid flat by wind. They give the tongue its characteristic roughness, the better to grip and manipulate food, and in cats and other obligate carnivores they are stiffened further still, into the rasping file that allows a cat to groom flesh from a bone with unsettling efficiency. But the filiform papillae, for all their number, are sensory impostors: they contain no taste buds whatsoever, and register only pressure and texture, never flavour of any kind. A tongue coated entirely in filiform papillae would be a perfectly serviceable trowel, and nothing more.

Taste proper is the business of the other three kinds. The fungiform papillae — mushroom-shaped, as the name declares — are scattered across the front two-thirds of the tongue, thickest at the tip and edges, and are visible to the naked eye as small pink or red dots, especially after a sip of something hot. A typical tongue carries on the order of two hundred of them, though — and this will matter a great deal later in this chapter — the true number varies enormously from one person to the next. Each fungiform papilla carries, on average, a modest handful of taste buds sunk into its upper surface. Further back, where the tongue rises toward the throat, the foliate papillae sit in vertical folds along the two rear edges, like the pages of a slightly gathered fan, their crevices lined with taste buds sheltered from the direct scrape of chewing. And at the very root of the tongue, arranged in a shallow, backward-pointing V that anatomists call the sulcus terminalis, sit the circumvallate papillae — only a dozen or so, but the grandest of the four, each one a walled island a few millimetres across, moated by a narrow trench into which saliva and dissolved food are free to pool and linger over the taste buds lining the trench's inner wall. It is a wonderfully engineered piece of plumbing: by delaying food at the very back of the mouth, the circumvallate papillae give the tongue one last, lingering opportunity to assess what is about to be swallowed — a final customs inspection before the border.

Taste buds, it should be said, are not the tongue's exclusive property. A modest scattering of them persists on the soft palate at the roof of the mouth, on the epiglottis that guards the entrance to the windpipe, and along the upper reaches of the throat — a fact anatomists have known since taste buds were first properly described, in the tongues of rabbits and other mammals, by the German anatomists Gustav Schwalbe and Christian Lovén, working independently and arriving, within a year of one another around 1867, at the same small, flask-shaped discovery. The tongue remains the instrument's principal keyboard, but it is not, strictly speaking, the only place the music is heard.

The architecture of a taste bud

Zoom closer still — closer than any Victorian microscope could comfortably manage, though the imagination will serve — and a single taste bud resolves into an object of real elegance: a cluster of fifty to a hundred slender cells, bundled together like the segments of a peeled orange, or the staves of a tiny barrel standing on end within the full thickness of the tongue's outer skin. The whole structure is barely the width of a coarse human hair and perhaps twice that in height, and it opens to the world through a single narrow taste pore, through which the topmost tips of the receptor cells — dressed in fine, hair-like microvilli — reach out to meet whatever has been dissolved in saliva.

The cells within are not all of one kind, and the division of labour among them is one of the taste bud's more elegant economies. Some — called, with pleasing bluntness, Type I cells — perform the housekeeping: they wrap around their neighbours as a kind of glial support staff, clear away spent neurotransmitter, and are thought to handle a portion of the sensation of plain salt. Others, Type II cells, carry the receptors for sweetness, umami and bitterness — three quite different chemical alarms sharing a common style of machinery, of which more shortly. Still others, Type III cells, are the taste bud's true nerve endings in miniature, forming direct synapses with the sensory nerve fibres below and handling, chiefly, the sensation of sourness. And beneath them all sit basal cells, a modest reserve of stem cells whose sole function is replacement: for taste receptor cells, unlike most of the neurons in the body, do not last. Battered by heat, abrasion, and the passing insult of a too-hot cup of tea, they wear out and die within one to three weeks and are just as regularly replaced — a fact that offers real comfort to anyone who has ever scalded their tongue on a mouthful of soup taken too eagerly, in a hurry to reach the second course. The burn is real, but it is not, on the scale of the body's own reckoning, a lasting injury; the instrument rebuilds itself, cell by cell, several times a year, so quietly that its owner never notices the renovation taking place.

<figure> <svg xmlns="http://www.w3.org/2000/svg" width="100%" viewBox="0 0 460 640" role="img"> <title>The Taste-Bud in Section — a single taste bud shown in cross-section within the lingual epithelium, with receptor, support and basal cells, the taste pore and nerve fibres</title> <defs> <pattern id="tb-stip" width="6" height="6" patternUnits="userSpaceOnUse"><circle cx="1" cy="1" r="0.5" fill="#a9906c"/></pattern> </defs> <rect x="12" y="12" width="436" height="616" fill="none" stroke="#43331f" stroke-width="1"/> <rect x="15.5" y="15.5" width="429" height="609" fill="none" stroke="#43331f" stroke-width="0.5"/> <g fill="#43331f"> <path d="M12,7 L17,12 L12,17 L7,12 Z"/><path d="M448,7 L453,12 L448,17 L443,12 Z"/> <path d="M12,623 L17,628 L12,633 L7,628 Z"/><path d="M448,623 L453,628 L448,633 L443,628 Z"/> </g> <text x="230" y="41" text-anchor="middle" font-family="Georgia, 'Times New Roman', serif" font-size="14" letter-spacing="2" fill="#43331f">THE TASTE-BUD IN SECTION</text> <line x1="90" y1="56" x2="370" y2="56" stroke="#6f5a40" stroke-width="0.5"/> <path d="M55,150 Q100,136 145,150 Q190,162 230,152 Q270,142 315,150 Q360,160 405,148 L405,500 L55,500 Z M230,172 Q195,225 190,300 Q195,378 218,452 Q230,458 242,452 Q265,378 270,300 Q265,225 230,172 Z" fill="url(#tb-stip)" fill-rule="evenodd" stroke="#6f5a40" stroke-width="0.7"/> <path d="M230,172 Q195,225 190,300 Q195,378 218,452 Q230,458 242,452 Q265,378 270,300 Q265,225 230,172 Z" fill="none" stroke="#43331f" stroke-width="1"/> <g fill="none" stroke="#6f5a40" stroke-width="0.6"> <path d="M230,174 Q212,225 209,300 Q212,375 220,447"/> <path d="M230,174 Q220,225 218,300 Q220,372 224,446"/> <path d="M230,174 Q230,225 228,300 Q229,372 229,447"/> <path d="M230,174 Q240,225 241,300 Q239,372 235,446"/> <path d="M230,174 Q250,225 250,300 Q248,375 240,447"/> <path d="M230,174 Q259,225 259,300 Q255,378 244,449"/> </g> <ellipse cx="230" cy="449" rx="16" ry="11" fill="none" stroke="#43331f" stroke-width="0.8"/> <circle cx="230" cy="172" r="4.5" fill="none" stroke="#9c5a2c" stroke-width="1"/> <g fill="none" stroke="#6f5a40" stroke-width="0.6"> <path d="M220,168 Q218,157 216,148"/><path d="M225,166 Q225,154 224,146"/> <path d="M230,165 Q230,152 231,144"/><path d="M235,166 Q237,154 239,147"/> <path d="M240,168 Q243,158 246,150"/> </g> <g fill="none" stroke="#6f5a40" stroke-width="0.6"> <path d="M215,458 C195,472 185,485 168,505"/> <path d="M222,461 C208,478 200,492 190,518"/> <path d="M230,463 C228,482 230,500 226,528"/> <path d="M239,461 C248,480 250,496 258,520"/> <path d="M246,458 C262,474 272,486 288,504"/> </g> <g stroke="#6f5a40" stroke-width="0.5" fill="none"> <path d="M215,240 L82,222"/><path d="M252,270 L392,250"/> <path d="M222,448 L150,494"/><path d="M230,166 L230,102"/><path d="M258,502 L328,536"/> </g> <g font-family="Georgia, 'Times New Roman', serif" font-style="italic" font-size="12" fill="#43331f"> <text x="76" y="220">a</text><text x="396" y="248">b</text> <text x="144" y="500">c</text><text x="227" y="96">d</text><text x="332" y="540">e</text> </g> <g font-family="Georgia, 'Times New Roman', serif" font-style="italic" font-size="11" fill="#43331f"> <text x="60" y="562">a — receptor cell</text> <text x="60" y="578">b — support (sustentacular) cell</text> <text x="60" y="594">c — basal (stem) cell, at the foot</text> <text x="240" y="562">d — the taste pore</text> <text x="240" y="578">e — nerve fibres, braiding at base</text> </g> <line x1="330" y1="606" x2="378" y2="606" stroke="#43331f" stroke-width="0.7"/> <line x1="330" y1="602" x2="330" y2="610" stroke="#43331f" stroke-width="0.7"/> <line x1="378" y1="602" x2="378" y2="610" stroke="#43331f" stroke-width="0.7"/> <text x="354" y="618" text-anchor="middle" font-family="Georgia, 'Times New Roman', serif" font-style="italic" font-size="10" fill="#43331f">~30–70 µm</text> </svg> <figcaption><em>Plate 1. Anatomy of a taste bud in cross-section: the onion-like bundle of receptor, support and basal cells, the taste pore and its microvilli, and the nerve fibres departing at its base</em></figcaption> </figure>

Field Note — The Taste Bud

Under a sufficiently patient microscope — the kind of instrument this compendium's compiler keeps, quite unreasonably, on the breakfast table — a single taste bud presents itself less as a piece of anatomy than as a small jewel, set flush into the epithelium like a sapphire in a ring's bezel. Sectioned crosswise, it shows the barrel-stave arrangement of its fifty-odd cells with startling regularity, each one tapering to a fine microvillus at the taste pore, so that the whole assembly resembles nothing so much as a tulip bulb sliced from crown to root, or a tightly furled anemone waiting for the tide.

Specimen notes, for the collector's card: diameter at the widest point, some thirty to seventy micrometres — roughly the breadth of a coarse human hair, and quite invisible without magnification, though several hundred of them together, massed within a fungiform papilla, will just betray themselves to the naked eye as a faint reddening of the tongue's surface. Life expectancy of the individual receptor cell: one to three weeks, among the shortest-lived of all the body's specialised cells, and constantly renewed from a modest population of basal stem cells at the bud's foot — an anatomical structure, in other words, that is still under construction on the very day one dies. Distribution: concentrated on the tongue, but not confined to it, with modest colonies on the soft palate, the epiglottis and the upper throat, silent witnesses to every swallow. And perhaps the most delightful specimen note of all: no two collectors' cabinets — no two tongues — are alike. The density of these jewels varies genuinely and measurably from one person to the next, by a factor of two or three, a fact this chapter will return to before its close, and which explains, rather better than manners do, why one guest at a dinner table reaches eagerly for a second helping of something the guest beside him quietly pushes to the rim of the plate.

Wires to the brain

A taste bud, for all its jewel-like intricacy, is mute without a listener, and the business of listening falls to three cranial nerves working in strict geographic partnership, rather like three district officers each responsible for their own parish of the tongue. The front two-thirds — the fungiform territory — reports to a branch of the facial nerve, the seventh cranial nerve, called the chorda tympani, a name that betrays an anatomical eccentricity worth pausing over: this nerve, on its way from tongue to brainstem, takes a most improbable detour directly through the middle ear, threading between two of the tiny bones of hearing, the malleus and the incus, entirely unprotected by bone. It is a piece of wiring any engineer would consider reckless, and surgeons operating on the middle ear for quite unrelated reasons — to repair a perforated eardrum, say, or replace a diseased stapes — must take care not to sever it, for its casualties report a lingering metallic taste or numbness on one side of the tongue for months afterward. Nature, having found a convenient conduit through an already crowded piece of anatomy, saw no reason not to use it twice.

The rear third of the tongue, circumvallate papillae included, reports instead to the glossopharyngeal nerve, the ninth cranial nerve, whose name — "tongue-and-throat" in its Greek roots — describes its territory with admirable economy. And a small, easily forgotten contribution comes from the tenth cranial nerve, the vagus, which picks up taste signals from the epiglottis and the upper throat before continuing on its famously wandering course through the chest and abdomen. Between the three of them, facial, glossopharyngeal and vagus nerves account for every taste bud in the mouth without exception or overlap dispute — a piece of anatomical tidiness the reader may find satisfying, given the muddle to come in a later section of this chapter.

All three converge, first, on a single relay station in the brainstem called the nucleus of the solitary tract — a fitting name, since it is here that the solitary business of tasting first becomes a co-ordinated committee report. From the solitary nucleus the signal travels up to the thalamus, the brain's great switchboard, and specifically to a way-station called the ventral posteromedial nucleus, before arriving at last at the gustatory cortex proper: a modest patch of tissue folded into the anterior insula and the neighbouring frontal operculum, buried just behind the temple, roughly where a Victorian phrenologist might have gone looking for the organ of appetite and, for once, not been very far wrong. It is only here, deep in the brain and a considerable journey from the tongue that first raised the alarm, that a pattern of nerve impulses is finally read as sweet, or bitter, or anything at all. And it is here, too — though this is a story for a later chapter — that taste first begins to keep company with smell and texture, in the orbitofrontal cortex just beyond, where the brain performs the remarkable trick of binding several quite separate senses into the single, seamless experience we call flavour.

The five doors of taste

Whatever a taste bud's location, its repertoire is the same throughout the mouth, and it is a shorter repertoire than most people suppose. Despite the vast vocabulary of flavour talked about at any decent dinner table — the grassiness of good olive oil, the mineral tang of an oyster, the burnt-sugar depth of a dark ale — the tongue itself is capable of registering only five fundamentally different qualities. Everything else, as the Preface to this volume has already confessed, is smell, texture, and the busy embroidery of the mind. But those five are not arbitrary. Each was assembled by a very long and unsentimental process of natural selection, and each answers a specific question that every animal with a mouth has needed answered since long before there were mouths capable of asking it in words: is this food, and how much of it should I eat?

Sweetness answers the first and friendliest of these questions. It is the tongue's detector for sugars — for the simple, energy-dense carbohydrates that ripe fruit, and little besides ripe fruit, offers in quantity in the wild — and it is registered by a single receptor, a paired protein called T1R2 and T1R3 that clasps together into a single lock broad enough to be triggered by an enormous range of different sugar molecules, and, as any dieter reaching for an artificial sweetener discovers, by a good many entirely unrelated molecules that merely happen to fit the same lock. The preference for sweetness needs no education: a newborn infant, who has never eaten a solid meal in its life, will purse and relax its lips with visible pleasure at a drop of sugar water on the tongue, and grimace at a drop of something bitter, within hours of being born. Three million years of ancestors who found ripe fruit delightful, and left more descendants for the finding, are not easily argued with.

Umami — a word that entered the English dictionary only in this century, though the sensation itself has presumably been on every human tongue since the first — answers a hungrier and more urgent question: does this food contain protein? It is the taste of glutamate, an amino acid that is one of the fundamental building blocks of every protein in every living cell, and its detection was, for a hunting-and-gathering animal, a matter of considerable practical importance: meat, and the amino acids meat is built from, are scarcer and more valuable than sugar in almost any wild landscape. The taste was identified, named and very nearly patented by a single scholar: Kikunae Ikeda, a chemistry professor at Tokyo Imperial University, who in 1908 set out to discover why a broth of kombu, the kelp used to make Japanese dashi, tasted so satisfying in a way that fitted none of the four tastes then recognised by Western science. Ikeda isolated the responsible compound as glutamic acid, coined the word umami — a compound of umai, "delicious," and mi, "taste" — and within the same year filed a patent for a practical method of manufacturing the crystalline sodium salt of glutamate: monosodium glutamate, or MSG, soon licensed to a small Tokyo trading firm that would grow into the company Ajinomoto, "the essence of taste." The receptor Ikeda's discovery implied would not be confirmed for nearly a century; in 2002 the neurobiologist Charles Zuker and his colleagues showed that a pairing of the proteins T1R1 and T1R3 — cousins of the sweet receptor, sharing one subunit with it — responds specifically to glutamate, closing a case that cooks in every glutamate-rich cuisine on earth had been proving by practice for very much longer than science had been paying attention.

Umami possesses one further trick worth a naturalist's particular admiration: synergy. Glutamate alone produces a pleasant savoury taste; glutamate in the presence of certain ribonucleotides — inosinate, abundant in meat and dried bonito flakes, or guanylate, which forms in quantity as shiitake mushrooms are dried — produces a savoury taste dramatically more intense than the two compounds' separate strengths would predict, a chemical duet in which one voice makes the other very much louder. It is no accident that classical Japanese dashi is built from precisely this pairing, kombu's glutamate meeting the inosinate of dried bonito or the guanylate of dried shiitake, nor that so many of the world's other deeply savoury preparations — a good stock built from both bones and vegetables, a ragù that unites meat with tomato, a scattering of Parmesan over a mushroom risotto — turn out, whether or not their cooks ever heard the word synergy, to be independent rediscoveries of the same happy chemistry.

It would be an oversight, at this point, not to address the question the mere mention of MSG reliably raises in a certain kind of reader: is it safe? The honest answer, arrived at after several decades of exactly the controlled trials the question deserves, is yes, without meaningful qualification. The glutamate in a shaker of MSG is chemically indistinguishable from the glutamate your own palate has been enjoying, quite unlabelled, in every ripe tomato, aged cheese and mushroom you have ever eaten; the body's cells cannot tell the two apart, because there is no difference to tell. The notion of a "Chinese restaurant syndrome" traces to a single speculative letter published in an American medical journal in April 1968, in which a physician described symptoms he had noticed after eating at Chinese restaurants and speculated, without evidence, that the cause might be salt, cooking wine, or possibly MSG. The letter was not a scientific study, and no controlled trial in the decades since has managed to reproduce the effect it described under conditions where subjects did not know whether they had received MSG or a placebo. Food-safety authorities on both sides of the Atlantic and the Pacific — the American Food and Drug Administration among them — classify MSG as safe at ordinary levels of consumption. That a seasoning associated with Chinese cooking, and almost no other cuisine's, should have borne the brunt of half a century's unfounded suspicion is a fact this compendium notes and leaves the reader to judge.

Saltiness answers a narrower but no less vital question: does this food contain the electrolytes — chiefly sodium — that nerves and muscles require to function at all? Of the five tastes, salt is detected by the simplest mechanism of all: sodium ions pass directly through a channel in the receptor cell's membrane, called, with impressive literalness, the epithelial sodium channel, and their arrival alone is enough to trigger the cell, without any of the elaborate molecular lock-and-key machinery sweetness or bitterness requires. At the modest concentrations found in a well-seasoned dish, salt is purely appetitive — which is a scientific way of saying moreish — but the appetite has a governor built into it. Raise the concentration high enough, as any diner who has taken an incautious mouthful of seawater or an oversalted broth can testify, and the tongue's salt-sensing cells are joined by cells belonging to the sour and bitter pathways, recruited specifically to make dangerously concentrated salt taste unpleasant rather than merely intense. This double mechanism, confirmed by Zuker's laboratory only in 2013, is a fine example of evolution's frugality: rather than engineer an entirely new alarm for a problem — too much salt — that arises only occasionally, the body simply presses its existing danger signals into a second job.

Sourness is the tongue's acidity meter, and its message is deliberately ambivalent, because acidity itself is ambivalent evidence. A sharp, sour taste can mean a fruit is unripe and best left on the branch a little longer; it can equally mean that food has spoiled and begun to ferment in an unwelcome way, its sugars broken down by unwanted microbes into acid. Both are useful things for a foraging animal to know, and both come flagged by the same mechanism: hydrogen ions, the chemical signature of any acid, pass into specialised receptor cells through a channel called OTOP1 — identified, remarkably recently, only in 2018, by Emily Liman's team at the University of Southern California, who had spent years narrowing down a list of candidate genes before finding the one that, transplanted into an ordinary cell with no prior interest in acid, made that cell flinch at a drop of vinegar. Yet sourness is not always a rejection: humankind has spent millennia deliberately courting the very spoilage sourness warns against, coaxing milk into yoghurt, cabbage into sauerkraut, grape juice into wine, and grain into sourdough, learning through long culinary experience to distinguish the sourness of ruin from the sourness of transformation. The taste bud cannot make this distinction; the cook, mercifully, can.

Bitterness, last of the five, is the tongue's burglar alarm, and like most alarms it is designed to err heavily on the side of false positives. An extraordinary number of the toxic compounds that plants manufacture to discourage being eaten — alkaloids, glycosides, and a long chemical rogues' gallery besides — share the single quality of tasting intensely bitter to a mammalian tongue, and evolution has responded by building not one bitter receptor but some twenty-five distinct ones, the largest family of taste receptors by far, each tuned to a different shape of molecular threat. A single bitter-sensing cell may carry several of these receptor types at once, sacrificing precision for coverage: it cannot tell the diner which poison has arrived, only that one has, much as a smoke alarm cannot say whether the fire began in the toast or the curtains. This is also why bitterness tends to announce itself at far lower concentrations than sweetness needs to persuade: a few parts per million of a bitter alkaloid will register clearly, where sugar must be present in proportions closer to a tablespoon per cup before the tongue troubles to notice. An animal that discovers a poison only after swallowing a generous portion of it has learned an expensive lesson; better, evolution appears to have decided, to be needlessly suspicious of a thousand harmless bitter leaves than trusting of the one deadly one.

The reformed map

It is time, as promised, to correct the diagram. Almost every reader of this volume will have encountered, at some point in a school textbook or a well-meaning teacher's chalk drawing, a map of the tongue divided into four cordoned territories: sweetness at the tip, saltiness and sourness along the two sides, bitterness banished to the back, each taste confined to its own tidy province like a colour on a political map. It is among the most widely reproduced diagrams in elementary science education. It is also, this compendium must report with the particular satisfaction naturalists reserve for a persistent error finally run to ground, entirely wrong.

The story of how it came to be believed is nearly as interesting as the correction itself, and worth telling in full, for it is a small parable about how carelessly a respectable fact can curdle into a popular myth. In 1901 a German researcher named David P. Hänig published a study, "Zur Psychophysik des Geschmackssinnes" ("On the Psychophysics of the Sense of Taste"), in which he measured how much of a given tastant — a drop of sugar solution, of salt, of acid, of a bitter compound — was required before a subject could detect it at various points around the rim of the tongue. Hänig found real, if small, differences: the tip of the tongue, for instance, needed a slightly weaker concentration of sugar to register sweetness than the sides did, and the back needed a slightly weaker concentration of the bitter compound than the front. These were genuine, measurable, and entirely unremarkable variations in sensitivity — the sort of minor regional difference one might expect to find on almost any biological surface — and Hänig never claimed that any region of the tongue was blind to any taste. He merely charted where each was detected most easily.

The trouble began with the chart itself. Hänig's graph plotted relative sensitivity along the rim of the tongue for each taste separately, on its own vertical scale, and when the four separate curves were later redrawn and compared side by side — most influentially by the Harvard psychologist Edwin G. Boring, in his 1942 textbook Sensation and Perception in the History of Experimental Psychology — the different scales vanished from the retelling, and with them, the modesty of the original differences. What had been small variations in threshold, a matter of a few percentage points of sensitivity, was flattened by decades of simplified retelling into the crisp cartoon zones every schoolchild has since been shown, with each taste's peak of sensitivity redrawn as its exclusive territory. Hänig's careful, minor observation had, by the time it reached the textbooks, been quietly and thoroughly rewritten by other hands.

The correction has, in fact, been available for just as long as the myth has been current. In 1965 a surgeon named T. R. Bull reported on patients who had undergone surgical section of the chorda tympani — the very nerve, the reader will recall, that reports taste from the front two-thirds of the tongue — and found that they lost no meaningful ability to taste sweetness at all, though the tongue map would have predicted total sweet-blindness at the tip. In 1974 a researcher named Virginia Collings tested all five tastes methodically across every taste-bearing region of the mouth, including the soft palate and throat, and confirmed that every taste could be detected everywhere taste buds existed, some regions merely a shade more sensitive than others, in rough agreement with Hänig's original, modest numbers. And in perhaps the most delightfully counter-intuitive finding of all, the taste researcher Linda Bartoshuk demonstrated in 1993 that anaesthetising the chorda tympani nerve did not abolish sweetness at the front of the tongue, as the old map would insist, but in some subjects made sweetness taste more intense elsewhere — because the nerves serving different regions of the tongue turn out to inhibit one another slightly, and silencing one voice in the choir makes its neighbours sing out more clearly. It is a finding the tongue map, with its tidy fences, could never have predicted, because the tongue map's fences do not exist.

The verdict of modern molecular biology is more decisive still. Every receptor type this chapter has described — the sweet-sensing T1R2 and T1R3, the sour-sensing OTOP1, the twenty-five bitter T2Rs, all the rest — has been found, wherever researchers have thought to look for it, scattered across fungiform, foliate and circumvallate papillae alike, tip to root, with no exclusive zone anywhere. The reader need not take any of this on faith, nor even on the authority of a German paper filed away since 1901. A cup of black coffee touched first to the tip of the tongue and then to the sides will taste bitter in both places; a salted pretzel will taste salty wherever it is pressed; and a wedge of lemon, drawn slowly from the very tip of the tongue to its furthest reachable corner, will draw the same pucker of sourness the entire way. The map was never real. The tongue, happily, was simply doing its job.

<figure> <svg xmlns="http://www.w3.org/2000/svg" width="100%" viewBox="0 0 640 460" role="img"> <title>The Tongue Reformed — the retired four-zone taste map beside the truth of taste receptors distributed evenly across the tongue</title> <defs> <path id="tongueShape" d="M175,110 C130,112 100,130 90,170 C82,205 85,240 100,275 C115,310 140,345 175,378 C210,345 235,310 250,275 C265,240 268,205 260,170 C250,130 220,112 175,110 Z"/> <clipPath id="tongueClip"><use href="#tongueShape"/></clipPath> <pattern id="fine-stip" width="7" height="7" patternUnits="userSpaceOnUse"><circle cx="1.2" cy="1.2" r="0.7" fill="#6f5a40"/><circle cx="4.6" cy="4.8" r="0.6" fill="#6f5a40"/></pattern> </defs> <rect x="12" y="12" width="616" height="436" fill="none" stroke="#43331f" stroke-width="1"/> <rect x="15.5" y="15.5" width="609" height="429" fill="none" stroke="#43331f" stroke-width="0.5"/> <g fill="#43331f"> <path d="M12,7 L17,12 L12,17 L7,12 Z"/><path d="M628,7 L633,12 L628,17 L623,12 Z"/> <path d="M12,443 L17,448 L12,453 L7,448 Z"/><path d="M628,443 L633,448 L628,453 L623,448 Z"/> </g> <text x="320" y="41" text-anchor="middle" font-family="Georgia, 'Times New Roman', serif" font-size="14" letter-spacing="2" fill="#43331f">THE TONGUE REFORMED</text> <line x1="130" y1="56" x2="510" y2="56" stroke="#6f5a40" stroke-width="0.5"/> <line x1="320" y1="90" x2="320" y2="392" stroke="#c8b79a" stroke-width="0.6"/> <use href="#tongueShape" fill="none" stroke="#43331f" stroke-width="1"/> <g fill="none" stroke="#6f5a40" stroke-width="0.7" stroke-dasharray="4,3"> <path d="M95,163 Q175,185 255,163"/> <path d="M112,312 Q175,335 238,312"/> <path d="M175,178 Q168,245 175,322"/> </g> <g stroke="#6f5a40" stroke-width="0.6"> <line x1="120" y1="163" x2="120" y2="171"/><line x1="230" y1="163" x2="230" y2="171"/> <line x1="128" y1="313" x2="128" y2="321"/><line x1="222" y1="313" x2="222" y2="321"/> </g> <g font-family="Georgia, 'Times New Roman', serif" font-style="italic" font-size="11" fill="#43331f" text-anchor="middle"> <text x="175" y="145">bitter</text><text x="175" y="358">sweet</text> <text x="130" y="248">salty</text><text x="220" y="248">sour</text> </g> <g fill="none" stroke="#9c5a2c" stroke-width="1"> <circle cx="240" cy="128" r="15"/> <line x1="232" y1="120" x2="248" y2="136"/><line x1="248" y1="120" x2="232" y2="136"/> </g> <text x="240" y="150" text-anchor="middle" font-family="Georgia, 'Times New Roman', serif" font-style="italic" font-size="7" fill="#9c5a2c">retired</text> <text x="175" y="405" text-anchor="middle" font-family="Georgia, 'Times New Roman', serif" font-style="italic" font-size="12" fill="#43331f">the old map</text> <g transform="translate(290,0)"> <use href="#tongueShape" fill="url(#fine-stip)" clip-path="url(#tongueClip)"/> <use href="#tongueShape" fill="none" stroke="#43331f" stroke-width="1"/> </g> <text x="465" y="405" text-anchor="middle" font-family="Georgia, 'Times New Roman', serif" font-style="italic" font-size="12" fill="#43331f">the truth</text> <text x="465" y="421" text-anchor="middle" font-family="Georgia, 'Times New Roman', serif" font-style="italic" font-size="9" fill="#43331f">every taste, everywhere — only the density varies</text> </svg> <figcaption><em>Plate 2. The tongue reformed: the old cordoned map of exclusive zones retired at last, set beside the truth — every receptor type present, in varying density, across the whole tasting surface</em></figcaption> </figure>

The measure of a tongue

If no two regions of a single tongue differ very much in what they can taste, the same cannot be said of two different tongues compared against one another — and here the chapter arrives at one of the more entertaining corners of the whole subject. Individual human beings vary, quite dramatically and quite measurably, in how intensely they experience the same mouthful, and the difference has a name, a history, and — appropriately for a compendium of this kind — an origin story that began with an accident in a chemical laboratory.

In 1931 a DuPont chemist named Arthur Fox was transferring a quantity of a newly synthesised compound called phenylthiocarbamide, or PTC, into a bottle when a little of the powder drifted into the air. A colleague working nearby complained at once of a strongly bitter taste; Fox, standing closer to the spill and by rights the more exposed of the two, tasted nothing whatsoever. Curious rather than alarmed, Fox began testing family, friends and eventually large populations for their sensitivity to PTC, and within the year the geneticist L. H. Snyder had shown that the ability to taste it followed a recognisable inherited pattern, PTC-blindness behaving as a recessive trait passed down in an orderly Mendelian fashion.[^2] A chemically milder cousin of PTC called propylthiouracil, or PROP, has since become the standard laboratory tool for the same test, and it is now understood that the difference traces very largely to variation in a single gene, TAS2R38, which comes in a small number of common forms: individuals carrying two copies of the sensitive form taste PTC or PROP as intensely, unpleasantly bitter; those with two copies of the insensitive form taste next to nothing; and those with one of each fall somewhere in between.

Roughly a quarter of any given population turns out to be a "non-taster" of this kind, roughly a quarter a highly sensitive taster, and the remaining half somewhere comfortably in the middle. But bitterness genetics is only half the story, for the psychologist Linda Bartoshuk, building on decades of this research in the early 1990s, established that the most sensitive tasters — the group she christened supertasters, a term that has since escaped happily into popular use — owe their sensitivity not only to their TAS2R38 genes but to simple anatomy: a measurably higher density of fungiform papillae on the tongue, and so a genuinely higher density of taste buds packed into the same square centimetre of tissue. Where a typical non-taster's tongue carries on the order of thirty-odd fungiform papillae per square centimetre, a supertaster's may carry sixty or more — much the same difference, roughly, as between a lawn and a well-fertilised one. Supertasters, who make up perhaps a quarter of the population and are found rather more often among women than men, do not merely taste bitterness more sharply; they experience sweetness, the burn of chilli and the sting of alcohol with greater intensity across the board, and they tend, on the whole, to be considerably fussier eaters, for the plain reason that a mouthful an ordinary taster finds pleasantly bitter — a well-grown kale, a bitter melon, a strong black coffee — can arrive on a supertaster's tongue as something closer to an assault.

This, in turn, offers the tidiest possible explanation for a phenomenon every parent of a small child will recognise without needing a laboratory to confirm it: the near-universal, deeply stubborn childhood loathing of bitter vegetables. It is not, whatever the old wives' tale insists, that children possess more taste buds than adults; the number of taste buds a person carries is set early in life and does not meaningfully increase thereafter. What does change is sensitivity: children are, on the whole, considerably more responsive to bitterness than adults of the very same genetic makeup, a heightened alarm that serves an obvious protective purpose in a small animal still learning which leaves are safe to put in its mouth, and one that reliably fades as the child grows into adolescence. A sprout that tastes merely assertive to a grown palate can taste frankly punishing to a seven-year-old one, and the seven-year-old, on the available evidence, is not being unreasonable.

It is worth pausing, before leaving this subject, to record that none of it would have surprised Brillat-Savarin, who arrived at a rough version of the same discovery in 1825 by nothing more sophisticated than close observation at his own dinner table. "Anatomy tells us," he wrote, "that all tongues are not exactly alike, there being three times as many feelers in some tongues as in others. This circumstance will explain why one of two guests, sitting at the same table, is delighted, while the other seems to eat from constraint." It would be a hundred and seventy years before Linda Bartoshuk's laboratory supplied him with a number rather more precise than "three times," and a genetic mechanism he could not possibly have imagined; but the essential observation — that some tongues are simply more thickly furnished than others, and that no amount of good manners will make a difference-blind guest enjoy a dish a keener tongue relishes — was, remarkably, already his.

Beyond the five

No chapter on the instrument of taste would be honest if it pretended the last word had been spoken on the number five. Science, as this compendium will insist throughout its volumes, is not a museum of settled facts but a very slow argument, and two candidates currently stand at the argument's edge, neither yet fully admitted to the club of five, both worth the reader's attention.

The first is fat. For a very long time, the pleasure of fat on the tongue was filed away entirely under texture: the silkiness of good olive oil, the slip of butter, the coating richness of cream, all attributed to mouthfeel alone, with no taste component of its own. That tidy division came under serious challenge in 2015, when a research team at Purdue University led by Richard Mattes demonstrated that human subjects, presented with solutions of non-esterified fatty acids from which every trace of smell and texture had been carefully stripped away, could nonetheless identify a taste distinct from all five recognised categories — one most subjects did not much enjoy, describing it as faintly bitter or sour but ultimately its own unpleasant thing, closer to the flavour of oil gone rancid than to any dish deliberately cooked. Mattes and his colleagues proposed the name oleogustus, built plainly enough from the Latin for oil and for taste, and the case for a receptor to match has grown steadily since. Whether oleogustus will ultimately join sweet, sour, salty, bitter and umami as an official sixth member remains, at the time of this volume's writing, a genuinely open question among taste researchers rather than a settled fact — but the direction of the evidence points, on balance, toward yes.

The second candidate arrives from Japan, and its name, kokumi, resists easy translation: it is built from koku, suggesting richness, depth and a long, lingering fullness, and mi, "taste." Japanese food chemists working for the flavour company Ajinomoto — the same firm, by a pleasing symmetry, that commercialised Ikeda's umami discovery generations earlier — noticed in the late 1980s that certain compounds, most notably a small sulphur-containing molecule called glutathione, abundant in garlic, onions, aged cheese and slow-cooked meat, produced no distinct taste of their own and yet made every other taste in a dish seem rounder, fuller and more persistent, rather as a skilled accompanist can make a soloist sound better without ever playing a note the audience consciously registers. The mechanism now proposed involves a receptor called the calcium-sensing receptor, found on certain taste cells, which kokumi compounds appear to switch on as an amplifier rather than a signal in its own right. Most researchers therefore hesitate to call kokumi a taste at all, preferring to describe it as a taste modulator — a backstage technician rather than a performer — but cooks the world over have been achieving its effect for centuries without needing the vocabulary, in the long-simmered stock, the well-aged cheese, and the unhurried stew that somehow tastes deeper than the sum of its labelled ingredients.

Fat and kokumi are offered here not as settled additions to the list of five, but as evidence that the list itself is a living, working document rather than a stone tablet — a reminder, appropriate to a naturalist's compendium, that the map of any living system is always a little behind the territory it describes.

And so the instrument itself has been unpacked, examined, and — the reader may take some quiet satisfaction in this — its most persistent myth retired for good. Five doors, or perhaps a door or two more yet to be confirmed, stand upon the tongue, each answering an ancient question about what is safe and what is nourishing to eat, each wired with unhurried certainty to a brain now waiting for its verdict. And yet if the reader has followed this chapter attentively, a nagging discrepancy ought by now to have presented itself: five tastes, however finely calibrated, cannot possibly account for the near-infinite vocabulary of flavour a good cook commands — the difference between a Bramley and a Cox, between a Puy lentil and a common brown one, between this Tuesday's stew and last Tuesday's, made from the same recipe by the same tired hand. The tongue, for all its jewelled machinery, has told us remarkably little of what makes dinner taste like dinner. For the truth of that, as the next chapter will show, we must follow the plume of scent up and behind the palate, to an organ this compendium has scarcely mentioned yet, and which turns out, on close naturalist's inspection, to have been doing most of the real work all along.

[^1]: The word survives, rather charmingly, in papillae mammae — the anatomist's term for the nipples proper — a small etymological joke the body has been keeping to itself for centuries. [^2]: So reliable did this pattern of inheritance prove that PTC-tasting tests were later used, in a rough and ready way, as supporting evidence in paternity disputes before the advent of DNA fingerprinting.