Chapter 6 — Texture, Mouthfeel & the Trigeminal Thrill

"Guinea Pepper is extreme hot and dry even in the fourth degree: that is to say, far hotter and drier then Avicenna showeth Dog's Ginger to be." — John Gerard, The Herball, or Generall Historie of Plantes, 1597

Consider two mouthfuls, taken a moment apart. The first is the outer inch of a good baguette, or a wafer of shortbread, or the first bite from a cold russet apple: the teeth close, there is a small clean fracture, a sound arrives at the ear a fraction of a second before the sweetness arrives at the tongue, and the brain — on the strength of that sound alone, before any chemistry has been consulted at all — pronounces the word fresh. The second mouthful is a sliver of red chilli, offered to a diner new to the sport. The teeth close exactly as before, but what follows is not a report from the tongue in any sense we have yet discussed. It is a rising, spreading, wholly convincing heat, as though a small coal had been posted to the roof of the mouth, and it has nothing whatever to do with sweet, sour, salt, bitter or umami. Both mouthfuls belong to this chapter, and both have been poorly served by the old, tidy notion that eating is chiefly a business of the tongue. The tongue, we now understand, is a mere reporter of five dispatches. The mouth entire — its teeth, its jaw joints, its lining, its widely branching fifth cranial nerve — is a reporter of a great many more, and this chapter is a field guide to two of them: first, texture, the mechanical drama of biting, chewing and swallowing, which the tongue and palate read as a kind of Braille; and second, a stranger and more recently mapped territory that scientists call chemesthesis, in which certain molecules in food masquerade as heat, cold, or an electric buzzing, by trespassing directly on the nerves that were built to feel the world, not to taste it. Reader, brace yourself: some of what follows will sting.

The mouth as a laboratory of touch

Long before flavour chemists had names for volatile aroma compounds, cooks knew that a dish could fail on texture alone. A perfectly seasoned purée that has been left to congeal, a fish cooked one minute past tender, a loaf gone leathery in its bag — none of these has lost a single molecule of flavour in the chemist's sense, and yet each has become, in the mouth, a disappointment. That is because the mouth is first and foremost an organ of touch, generously supplied with mechanoreceptors in the lips, tongue, gums, cheeks and the roof of the palate, and further served by the muscles of the jaw and the stretch receptors buried in them, which report on resistance, on give, on the rate at which a solid surrenders to the bite. This whole apparatus is capable of registering forces and deformations far too small to see, and it does so continuously, so that what we call mouthfeel is in fact a moving picture, assembled frame by frame between the first bite and the final swallow, rather than a single snapshot taken at the moment of contact.

The vocabulary that has grown up around this sense is unusually rich, and worth pausing over, for each word names a distinct physical event. Crisp and crunchy describe the sudden, clean fracture of a rigid, largely air-filled structure — the cell walls of a raw carrot, taut with water under pressure (a state botanists call turgor), or the glassy, dehydrated starch matrix of a good biscuit — a fracture violent enough to produce an audible report, of which more in a moment. Snap is crispness compressed into an instant, the signature of a fresh bean pod or a well-tempered piece of chocolate, whose cocoa butter has crystallised into a dense, brittle lattice that shatters rather than bends. Chewy and tender sit at opposite ends of a single spectrum governed largely by collagen, the connective tissue that laces through muscle: raw or briefly cooked, collagen is tough and elastic, giving meat its chewy resistance; held long and gently in a moist heat, that same collagen dissolves into gelatin, and the tissue that was once resilient becomes tender enough to part at the touch of a fork. Creamy names a lubricating smoothness we shall return to directly, and juicy names the sensation of a food's own liquor being released under the pressure of the bite and washing across the tongue — an event as much about timing as volume, since a food that leaks its juice too early, before the teeth have engaged, reads as merely wet rather than juicy. None of these words, the reader will notice, refers to a taste in the technical sense. They are the mouth's report on architecture: on how a food is built, and how that building comes apart.

Nor is any of this a single, instantaneous verdict. Texture is read as a narrative, unfolding in stages between the first bite and the moment of swallowing, in a process food scientists call oral processing. A raw almond begins hard and dry; thirty seconds of chewing, moistened by saliva and warmed by the mouth, reduce it to a soft, cohesive paste, or bolus, judged ready for swallowing only once its particles have been broken down and lubricated to a fairly consistent standard — a standard the brain enforces with some rigour, since a poorly formed bolus, too coarse or too dry, is uncomfortable and occasionally dangerous to swallow. This is why the rate of textural change matters nearly as much as any single quality: a food that starts crisp and dissolves quickly, such as a meringue or a well-fried tempura batter, is prized precisely for the drama of that transformation, while a food that stays stubbornly, wearisomely the same from first bite to last is generally judged the poorer for it. It is also worth the naturalist's noting that no texture is universally loved or loathed: the pleasing chew of a good mochi, the deliberate slipperiness of a raw oyster or a strand of natto, the crackle of a fried locust, the yielding wobble of a properly made egg custard, are all textures that some culinary traditions prize precisely for qualities — sliminess, resistance, a total absence of crunch — that others have been trained since infancy to avoid. Texture, perhaps more than any other department of flavour, reveals how much of eating is learned rather than given.

The sound of a good bite

It was long assumed that this architecture was assessed by touch alone, until an Italian psychologist named Massimiliano Zampini, working with the Oxford experimental psychologist Charles Spence, performed one of the most delightful demonstrations in the whole literature of the senses. They had volunteers bite into Pringle's potato crisps while wearing headphones, through which the sound of their own chewing was fed back to them in real time — but doctored. When the researchers boosted the volume of the bite, or specifically amplified the higher frequencies of the crunch (the crackling range between roughly two and twenty kilohertz), the very same crisp was rated as significantly crisper and fresher than when the sound was left alone or muffled.[^1] Nothing about the crisp itself had changed; only the echo of its own destruction, piped back into the eater's ears, had been turned up. The finding tells us something important and slightly unsettling: crispness is not simply detected in the mouth and then reported to the brain. It is assembled by the brain from at least two channels of evidence, tactile and acoustic, arriving together and corroborating one another, so convincingly that the eater never suspects a duet is taking place.

This is not a frivolous finding, whatever its charmingly improbable methodology, and it did not go unrecognised: Zampini and Spence were awarded the Ig Nobel Prize in Nutrition in 2008, that most useful of honours, awarded for work that first makes people laugh, and then makes them think. The deeper reason a crunch persuades us of freshness is biological rather than aesthetic. Crispness is, almost without exception, a sign of low moisture and structural integrity — a state that spoilage, staling and bruising all conspire to destroy. An apple that has begun to soften, a biscuit that has drawn moisture from the air, a lettuce leaf a day past its best: all lose their crispness before they lose anything else a chemist could measure, because the collapse of a rigid, air-filled structure is one of the very first symptoms of decay. Evolution has therefore wired the ear, by way of the brain's habit of pooling evidence, into the front line of freshness detection. We hear whether our food is trustworthy before we have properly tasted it at all.

Rheology in plain words

Sound explains the crisp half of texture; a rather more sedate branch of physics explains nearly everything else, and it goes by the forbidding name of rheology — the study of how materials flow and deform under stress. It sounds like a subject for engineers, and indeed it was largely developed by them, but its vocabulary maps with startling precision onto the sensations of eating, and it is worth borrowing two of its central terms, because once understood they cannot be unlearned.

The first is viscosity, which is simply a material's resistance to flowing. Water has low viscosity and pours from a jug in an instant; honey has high viscosity and takes its leisurely time. On the tongue, viscosity is read directly as body or richness: a thin soup and a velouté may share an identical list of ingredients in different proportion, and it is viscosity, more than any single flavour compound, that persuades the palate one is a broth and the other a sauce. The second term is elasticity, a material's tendency to spring back to its original shape once a deforming force is removed. Bread dough is elastic because of gluten, a network of wheat proteins that stretches under the rolling pin and slowly resists it; a properly made jelly is elastic because of the three-dimensional mesh formed by gelatin or pectin, which yields to the spoon and then, released, trembles back toward its original form. Most foods are neither purely viscous liquids nor purely elastic solids but some combination of the two — viscoelastic, in the jargon — and the precise ratio of flow to spring is a great part of what separates a good caramel from a bad one, or a properly reduced sauce from a merely thickened one.

Viscosity and its cousin, lubrication, also explain a texture that the tongue prizes above nearly all others: creaminess. For years the folk explanation was simply "more fat, more cream," but the truth is more interesting and rather more useful to the cook. What the mouth actually detects is the size of the fat particles suspended in a food, not merely their quantity. Food scientists studying emulsions — mixtures of fat dispersed as tiny droplets through a watery base, of which cream, mayonnaise, and a well-made ice cream are all examples — have found that once those droplets shrink below roughly ten microns across, a barely perceptible graininess gives way to a silken smoothness; below about one micron, the sensation reads as unmistakably creamy, regardless of the food's total fat content.[^2] This is a matter of tribology as much as chemistry: fine fat droplets act as tiny ball-bearings, reducing the friction between tongue and palate as they roll and coalesce, and saliva itself plays an active role in that process, helping to break the droplets loose from the food's structure and spread them evenly across the oral surfaces. A skilled ice-cream maker, churning air and fat and ice crystals into submission, is in the most literal sense an engineer of particle size — which is why a homemade ice cream, however good the cream, can taste grainy where a commercial one feels like silk: industrial churns simply shear the fat globules finer.

Rheology also, at last, explains why a good biscuit does not stay good. The crispness we discussed above depends on the starch and sugar matrix of a baked good existing in what physicists call a glassy state: rigid, low in moisture, and brittle, rather like window glass, which is itself a liquid frozen too quickly to crystallise. Biscuits, crackers and crisps are all hygroscopic, meaning they are thirsty for the water vapour that surrounds them in an ordinary kitchen, and as they draw that moisture in from the air, their sugars and starches soften and the whole structure crosses what is called its glass transition — from a hard, glassy solid to a pliable, rubbery one. The biscuit has not spoiled in any sense a bacteriologist would recognise; its chemistry is largely unchanged. It has simply hydrated its way out of the crisp state and into a chewy one nobody asked for, which is why the humble tin, keeping the air at bay, is as important a piece of kitchen apparatus as any knife.

A second nervous system at the table

Everything discussed so far — crispness, creaminess, chewiness, the whole rheological orchestra — is perceived through what physiologists call the somatosensory system: ordinary touch, pressure and stretch receptors of the kind that also report the texture of a coat sleeve or a stair rail. But the mouth, nose and eyes share a second, quite different nerve supply, the trigeminal nerve, the fifth and largest of the twelve cranial nerves. It divides, early in its course, into three great branches — ophthalmic, maxillary and mandibular, named for the regions of the face they serve — which between them carry sensations of ordinary touch, temperature and pain from the eyes, the nasal cavity, the lips, the gums, the tongue and the lining of the entire mouth: the ache of a bumped tooth, the sting of shampoo in the eye, the sharp cold of an ice cube pressed too long against a cheek. It is this same nerve, it turns out, that certain plants have learned to hijack. A handful of molecules found in chillies, mustard, horseradish, wasabi, mint, and even the carbon dioxide fizzing in a glass of soda water, do not register on the tongue's five taste-receptor families at all. Instead they walk directly up to specific ion channels on the endings of the trigeminal nerve and switch them on, precisely as a genuine burn, a genuine chill, or a genuine vibration would. The sensation that results is not a hallucination and not, strictly, a taste: it is real touch information, called into being by a false alarm. This whole category of sensation has been given the rather beautiful technical name chemesthesis — chemical touch — and no discovery has done more to explain it than the identification, within the last thirty years, of the actual molecular doorbells these compounds ring.

The fire that is not there: capsaicin and TRPV1

Foremost among these doorbells is a protein called TRPV1, and the story of its discovery is one of the finest in modern physiology. In 1997 a young researcher at the University of California, San Francisco, named David Julius, set out to answer an old question with new tools: precisely how does capsaicin, the pungent compound concentrated in chilli peppers, produce its unmistakable sensation of burning? Working with colleagues including Michael Caterina, Julius built a library of DNA fragments from the sensory neurons known to respond to capsaicin, inserted them individually into cells that were normally quite indifferent to the compound, and searched, patiently, for the single fragment that would make an ordinary cell wince. He found it: a gene encoding an entirely new ion channel, later named the transient receptor potential vanilloid 1 receptor, or TRPV1, sitting in the membranes of pain-sensing neurons.[^3] What made the discovery so satisfying, rather than merely clever, is what came next: when Julius's team tested the same channel against heat alone, with no capsaicin anywhere in sight, it sprang open at a threshold of roughly forty-three degrees Celsius — precisely the temperature at which ordinary tissue begins to register pain. Capsaicin, it transpired, had not built a private door of its own. It had found the door already fitted for detecting dangerous heat, and simply let itself in.

This is the elegant, slightly vertiginous truth at the centre of every chilli-eater's experience: the burn of a jalapeño is not a metaphorical heat, nor an illusion, nor a taste dressed up in borrowed language. It is the very same signal that a lit match or a mouthful of scalding soup would send, arriving by the very same nerve, interpreted by the very same pain centres of the brain — with the singular difference that no tissue is actually being damaged, and no thermometer in the room would register any change at all. The mouth is, quite literally, lying to the brain, and the brain, quite reasonably, believes it. For this discovery, and its extension to a related family of temperature- and touch-sensing channels, David Julius shared the 2021 Nobel Prize in Physiology or Medicine with Ardem Patapoutian, the Nobel committee noting drily that the pair had answered one of the oldest questions in physiology: how does a nerve ending know what is happening to it at all.[^4]

The Scoville scale, and the story it tells

Long before Julius named the receptor, a Detroit pharmacist named Wilbur Scoville had already set about measuring the sensation the receptor produces, and the method he devised in 1912, while working for the pharmaceutical firm Parke, Davis & Company, remains one of the most charming pieces of unimprovable science of the whole gastronomic literature. Scoville's problem was practical: capsicum extracts were used in medicinal preparations, and no reliable way existed to standardise their strength. His solution, now known as the Scoville Organoleptic Test, was to dissolve a precise weight of dried pepper in alcohol, then dilute the tincture in sugar water, drop by drop, and hand it to a panel of tasters, asking them at each dilution whether they could still detect any heat at all. The dilution at which the last taster's tongue surrendered became the pepper's rating in Scoville Heat Units — a bell pepper needing no dilution at all to taste of nothing, a jalapeño needing to be diluted several thousandfold before its heat vanished, the ferocious Carolina Reaper needing more than a million.

It is worth dwelling on how peculiar this method is, and how honestly it wears its peculiarity. It is not a chemical assay at all, but a democracy of human tongues, its precision limited by the training, tolerance, and, one imagines, the mood of Scoville's panel on any given afternoon — and yet the scale that resulted proved sturdy enough to survive more than a century of use. Today the actual capsaicinoid content of a pepper is more often measured by high-performance liquid chromatography, a method blind to fatigue and personal taste, and the results are still routinely converted back into Scoville Heat Units, in tribute to the pharmacist who first thought to ask a room full of people how much sugar water it would take to make them stop noticing they were on fire.

Field Note — The Chili Pepper

Slice a chilli lengthwise and the eye goes, by habit, to the seeds, which folklore has long blamed for the fruit's fire. The blame is misplaced. The seeds themselves manufacture no capsaicin at all; what heat they carry they have merely borrowed from touching the true source, a pale, spongy ridge of tissue called the placenta, which runs down the pod's interior wall and to which the seeds are attached. It is here, in specialised glands on the placental surface, that the plant manufactures and stores the great majority of its capsaicinoids — by some estimates, the better part of ninety per cent. A cook wishing to tame a chilli need not fuss over every seed; scraping away the pale pith is nine-tenths of the labour.

Why should a fruit — whose entire evolutionary purpose is to be eaten and its seeds thereby dispersed — arm itself against being eaten at all? The American ecologist Joshua Tewksbury supplied the answer in 2001, in a short paper that reads like a small detective story. Studying wild chiltepin peppers in the Arizona desert, he found that birds ate the fruit eagerly and passed the seeds unharmed, while rodents that sampled the same fruit avoided it thereafter, and, crucially, tended to crush any seeds they did swallow with their molars, destroying them in the process. Capsaicin, Tewksbury proposed, is not a general defence at all, but a directed one: a chemical instruction, addressed specifically to mammals, that reads pass this fruit by, while remaining wholly illegible to birds — which, lacking a crop full of grinding teeth, are far better seed-dispersers in any case. The molecular reason for this selective deafness was supplied the following year by David Julius's own laboratory, which found that the bird version of the TRPV1 receptor simply will not bind capsaicin, differing from the mammalian version by a handful of amino acids. The chilli, in other words, has not merely evolved a weapon. It has evolved a weapon with a lock that only mammalian jaws can spring — and then, by a final irony neither bird nor rodent could have anticipated, one particular mammal learned to treasure the very sting it was designed to repel, and carried the fruit to every inhabited continent on earth.

<figure> <svg xmlns="http://www.w3.org/2000/svg" width="100%" viewBox="0 0 460 640" role="img"> <title>The Chilli, Dissected</title> <defs> <pattern id="hs1" width="4" height="4" patternUnits="userSpaceOnUse" patternTransform="rotate(45)"><line x1="0" y1="0" x2="0" y2="4" stroke="#a9906c" stroke-width="0.5"/></pattern> <symbol id="fl" overflow="visible"><path d="M0,-5 L5,0 L0,5 L-5,0 Z" fill="#43331f"/></symbol> </defs> <rect x="12" y="12" width="436" height="616" fill="none" stroke="#43331f" stroke-width="1"/> <rect x="17" y="17" width="426" height="606" fill="none" stroke="#6f5a40" stroke-width="0.5"/> <use href="#fl" x="12" y="12"/><use href="#fl" x="448" y="12"/><use href="#fl" x="12" y="628"/><use href="#fl" x="448" y="628"/> <text x="230" y="38" text-anchor="middle" font-family="Georgia, 'Times New Roman', serif" font-size="15" letter-spacing="2" fill="#43331f">THE CHILLI, DISSECTED</text> <line x1="150" y1="49" x2="310" y2="49" stroke="#6f5a40" stroke-width="0.5"/> <text x="230" y="64" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#6f5a40">Fig. II — Capsicum, in longitudinal section</text>  <path d="M225,86 L233,86 L235,132 L223,132 Z" fill="url(#hs1)" stroke="#43331f" stroke-width="1"/> <g stroke="#43331f" stroke-width="1.1"> <line x1="229" y1="134" x2="197" y2="123"/><line x1="229" y1="134" x2="210" y2="107"/><line x1="229" y1="134" x2="229" y2="101"/><line x1="229" y1="134" x2="248" y2="107"/><line x1="229" y1="134" x2="261" y2="123"/> </g> <circle cx="229" cy="135" r="4" fill="url(#hs1)" stroke="#43331f" stroke-width="0.8"/>  <path d="M208,140 Q174,230 174,320 Q174,410 290,505 Q206,410 206,320 Q206,230 252,140 Z M214,149 Q185,233 185,319 Q185,399 286,497 Q201,399 201,319 Q201,233 246,149 Z" fill="url(#hs1)" fill-rule="evenodd" stroke="#43331f" stroke-width="1.1"/>  <path d="M228,156 Q196,240 196,318 Q196,392 267,463" fill="none" stroke="#9c5a2c" stroke-width="1.2"/> <path d="M228,156 Q196,240 196,318 Q196,392 267,463" fill="none" stroke="#43331f" stroke-width="0.4" opacity="0.5"/>  <g fill="#a9906c" stroke="#43331f" stroke-width="0.6"> <circle cx="213" cy="170" r="3.4"/><circle cx="224" cy="163" r="3.4"/><circle cx="232" cy="174" r="3.4"/><circle cx="215" cy="185" r="3.4"/><circle cx="228" cy="190" r="3.4"/><circle cx="207" cy="193" r="3.4"/><circle cx="220" cy="203" r="3.4"/><circle cx="234" cy="204" r="3.4"/><circle cx="211" cy="215" r="3.4"/> </g>  <line x1="216" y1="248" x2="330" y2="222" stroke="#6f5a40" stroke-width="0.5"/> <text x="333" y="225" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">outer wall</text> <line x1="196" y1="330" x2="90" y2="345" stroke="#6f5a40" stroke-width="0.5"/> <text x="87" y="348" text-anchor="end" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">hollow interior</text> <line x1="228" y1="180" x2="335" y2="168" stroke="#6f5a40" stroke-width="0.5"/> <text x="338" y="171" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">seeds</text> <line x1="240" y1="380" x2="335" y2="392" stroke="#9c5a2c" stroke-width="0.5"/> <text x="338" y="396" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#9c5a2c">placenta</text> <line x1="267" y1="463" x2="267" y2="558" stroke="#9c5a2c" stroke-width="0.5"/> <text x="230" y="578" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11.5" fill="#43331f">the seat of capsaicin —</text> <text x="230" y="596" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11.5" fill="#43331f">the <tspan fill="#9c5a2c">placenta</tspan>, not the seeds</text> </svg> <figcaption><em>Plate 11. A chilli pepper in section, showing the placenta and pith — not the seeds — as the true seat of capsaicin</em></figcaption> </figure>

Menthol's borrowed frost

If capsaicin persuades the mouth that it is burning, a smaller family of molecules performs the opposite trick with equal conviction. Menthol, the crystalline compound distilled from peppermint and its relatives, produces an unmistakable sensation of coolness on the tongue, lips, and skin, even when applied at room temperature — a fact confirmed by any reader who has bitten into a strong mint and felt the draught of an open window that is not, in fact, open. The receptor responsible, TRPM8, was identified in 2002, in the same extraordinarily fertile season of research that had already yielded TRPV1, by two research groups working in friendly parallel — one including David Julius's collaborator David McKemy, the other led by Ardem Patapoutian's team, which would go on to share the Nobel eighteen years later.[^5] TRPM8 sits in the membranes of the same class of sensory neurons as its heat-sensing cousin, but tuned to open not at scalding temperatures but at cool ones, below roughly twenty-five degrees Celsius — and menthol, along with eucalyptol and a synthetic compound called icilin, springs the same lock at any temperature at all. A menthol sweet eaten on a warm afternoon persuades the mouth of a chill that no thermometer would find, for precisely the mirror-image reason that a chilli persuades it of a fire.

The buzz of Sichuan pepper

Neither burning nor chilling, but stranger than both, is the sensation produced by the dried husks of Zanthoxylum, known in the West as Sichuan or Szechuan pepper, and prized across western China for a quality the Chinese language names precisely and English does not: málà, numb-and-spicy, the numbness being every bit as deliberate an effect as the heat. The active compound, hydroxy-alpha-sanshool, produces neither burn nor chill but a curious, buzzing, faintly electrical tingle, most often compared by first-time eaters to touching the terminals of a small battery to the tongue, or to the after-tingle of a limb that has briefly gone to sleep.

The mechanism, worked out by Diana Bautista's laboratory at the University of California in 2008, is unlike either of the receptors discussed so far. Rather than opening a channel that shouts fire or frost, hydroxy-alpha-sanshool closes a different set of channels — so-called two-pore potassium channels, which ordinarily act as a kind of resting brake on nerve firing, holding sensory neurons quiet between genuine stimuli.[^6] With the brake released, the nerves nearest the sanshool molecule begin firing spontaneously, entirely unprompted by any real touch or vibration at all, and it is this spontaneous, rhythmic firing that the brain interprets as a physical buzz. In 2013, researchers at University College London led by Nobuhiro Hagura put a number to that rhythm with a wonderfully simple experiment: they applied Sichuan pepper to volunteers' lips, then asked them to compare the sensation against a mechanical vibrator applied to a fingertip, tuned until the two felt identical. The match, with remarkable consistency, came in at almost exactly fifty cycles a second — the pepper, in effect, dialling the lips' own touch nerves to a fixed and surprisingly precise frequency, indistinguishable from a real, external vibration the mouth is not actually receiving.[^7] Chew a Sichuan peppercorn, and your lips are being telephoned by a plant, in a frequency they cannot tell from the truth.

Mustard, horseradish, wasabi — and why the fire goes up, not along

A third pungent family — the mustards, in the broadest sense, encompassing horseradish and true wasabi as well as the more familiar table condiment — works through yet a third receptor, TRPA1, activated chiefly by a volatile compound called allyl isothiocyanate.[^8] This compound does not exist ready-made in the intact plant; it is manufactured on the spot, defensively, the instant a cell wall is broken by chewing, cutting or grating, when an enzyme called myrosinase is released from one cellular compartment and meets its substrate, a sulphur-containing precursor, in another. This is worth knowing for entirely practical reasons: a horseradish root left whole is nearly odourless, and only becomes fearsome in the seconds after grating, its heat fading again within an hour or two as the volatile compound escapes — which is why a wise cook grates it at the very last possible moment.

The most distinctive feature of this family's pungency, familiar to anyone who has taken too large a dab of wasabi with a piece of sushi, is that its heat does not sit placidly on the tongue as capsaicin's does, but rockets upward into the sinuses within seconds, then vanishes almost as quickly as it arrived. The explanation lies in simple volatility. Capsaicin is a large, oily, essentially non-volatile molecule, which stays put wherever it lands on moist tissue and lingers there, releasing its slow, sustained burn for as long as ten or fifteen minutes. Allyl isothiocyanate is a small, volatile molecule that evaporates readily at body temperature, rising as a vapour up the back of the throat and into the nasal cavity by the retronasal route we met in an earlier chapter — and the nasal lining, like the mouth, is generously supplied with its own trigeminal nerve endings. The result is a pungency that arrives in the nose as much as the mouth, strikes hard, and departs within a minute or two as the volatile compound disperses. The chilli-eater settles in for a long siege; the wasabi-eater endures a short, sharp raid on the sinuses, and is generally wise to breathe out through the nose rather than in, the moment the dose proves too large.

The prickle in the glass

A gentler member of this same family of sensations arrives every time a bottle is opened. The bite of a carbonated drink was long assumed to be a purely mechanical matter — the physical bursting of countless tiny bubbles against the soft tissues of the mouth — and bubbles do indeed play some small part. But research led by Jayaram Chandrashekar in Charles Zuker's laboratory in 2009 established that the sensation is chiefly chemical rather than mechanical, and arises by a genuinely elegant route.[^9] An enzyme called carbonic anhydrase, present on the surface of the tongue's sour-sensing taste cells, converts dissolved carbon dioxide into carbonic acid at considerable speed; the acid in turn releases free protons, and it is those protons — the same currency of sourness discussed in an earlier chapter — that excite the sour cells and, at higher concentrations, spill over to excite the trigeminal nerve endings nearby, producing the characteristic tingling bite. This is why flat cola, however sweet, tastes faintly wrong even before the tongue can say precisely how: it has lost not merely its fizz but its acid prickle, the small chemical sting that had been quietly seasoning every sip.

The drying grip of tannin

One further sensation belongs in this catalogue of the mouth's touch-senses, though it is neither hot, cold, nor buzzing, but dry: the astringency of a young red wine, an unripe persimmon, or an over-stewed pot of tea. Astringency is caused by tannins, a broad class of plant polyphenols, and its mechanism is a matter not of nerve endings at all, but of chemistry acting on the mouth's own lubrication system. Saliva is kept slippery in large part by a family of proteins unusually rich in the amino acid proline; tannins bind to these proline-rich proteins with great enthusiasm, clumping them together into complexes that precipitate out of solution, much as milk curdles when lemon juice is added to it.[^10] Stripped of its lubricating proteins, the saliva coating the tongue and cheeks can no longer glide smoothly across the tissue beneath, and the resulting friction is registered by the mouth's mechanoreceptors as a distinctive puckering, drying grip — astringency being, in the strictest sense, less a flavour than the sensory experience of one's own saliva failing at its job. It is why a young, tannic wine can be softened by food rich in fat or protein, which offer the tannins an alternative target to bind before they ever reach one's own saliva, and why the same wine, given ten years in a dark cellar, mellows on its own: its tannin molecules link slowly together into ever-larger chains, until they grow too large to dissolve at all, and fall to the bottom of the bottle as the sediment a careful decanter is built to leave behind.

<figure> <svg xmlns="http://www.w3.org/2000/svg" width="100%" viewBox="0 0 640 440" role="img"> <title>The Map of Chemesthesis</title> <defs><symbol id="fl" overflow="visible"><path d="M0,-5 L5,0 L0,5 L-5,0 Z" fill="#43331f"/></symbol></defs> <rect x="12" y="12" width="616" height="416" fill="none" stroke="#43331f" stroke-width="1"/> <rect x="17" y="17" width="606" height="406" fill="none" stroke="#6f5a40" stroke-width="0.5"/> <use href="#fl" x="12" y="12"/><use href="#fl" x="628" y="12"/><use href="#fl" x="12" y="428"/><use href="#fl" x="628" y="428"/> <text x="320" y="38" text-anchor="middle" font-family="Georgia, 'Times New Roman', serif" font-size="15" letter-spacing="2" fill="#43331f">THE MAP OF CHEMESTHESIS</text> <line x1="230" y1="49" x2="410" y2="49" stroke="#6f5a40" stroke-width="0.5"/> <text x="320" y="63" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#6f5a40">a key to five sensations of the trigeminal nerve</text> <line x1="40" y1="95" x2="600" y2="95" stroke="#c8b79a" stroke-width="0.5"/> <line x1="40" y1="158.5" x2="600" y2="158.5" stroke="#c8b79a" stroke-width="0.5"/> <line x1="40" y1="221.5" x2="600" y2="221.5" stroke="#c8b79a" stroke-width="0.5"/> <line x1="40" y1="284.5" x2="600" y2="284.5" stroke="#c8b79a" stroke-width="0.5"/> <line x1="40" y1="347.5" x2="600" y2="347.5" stroke="#c8b79a" stroke-width="0.5"/> <line x1="40" y1="410" x2="600" y2="410" stroke="#c8b79a" stroke-width="0.5"/> <line x1="112" y1="95" x2="112" y2="410" stroke="#c8b79a" stroke-width="0.4"/> <line x1="228" y1="95" x2="228" y2="410" stroke="#c8b79a" stroke-width="0.4"/>  <path d="M80,141 C70,141 65,128 72,119 C67,115 74,103 80,101 C86,103 93,115 88,119 C95,128 90,141 80,141 Z" fill="#9c5a2c" stroke="#43331f" stroke-width="0.8"/> <text x="130" y="123" font-family="Georgia, serif" font-size="12.5" letter-spacing="1.5" fill="#43331f">HEAT</text> <text x="238" y="123" font-family="Georgia, serif" font-size="12" font-style="italic" fill="#6f5a40">capsaicin <tspan fill="#6f5a40" font-style="normal">—</tspan> <tspan fill="#43331f">TRPV1</tspan></text>  <g stroke="#43331f" stroke-width="0.9"><line x1="67" y1="190" x2="93" y2="190"/><line x1="80" y1="177" x2="80" y2="203"/><line x1="71" y1="181" x2="89" y2="199"/><line x1="71" y1="199" x2="89" y2="181"/></g> <circle cx="80" cy="190" r="2" fill="#43331f"/> <text x="130" y="186" font-family="Georgia, serif" font-size="12.5" letter-spacing="1.5" fill="#43331f">COOL</text> <text x="238" y="186" font-family="Georgia, serif" font-size="12" font-style="italic" fill="#6f5a40">menthol <tspan fill="#6f5a40" font-style="normal">—</tspan> <tspan fill="#43331f">TRPM8</tspan></text>  <circle cx="76" cy="253" r="2.6" fill="#43331f"/> <path d="M86,243 A14,14 0 0 1 86,263" fill="none" stroke="#6f5a40" stroke-width="0.7"/> <path d="M92,238 A20,20 0 0 1 92,268" fill="none" stroke="#a9906c" stroke-width="0.6"/> <text x="130" y="249" font-family="Georgia, serif" font-size="12.5" letter-spacing="1.5" fill="#43331f">TINGLE</text> <text x="238" y="249" font-family="Georgia, serif" font-size="12" font-style="italic" fill="#6f5a40">hydroxy-α-sanshool <tspan fill="#6f5a40" font-style="normal">—</tspan> <tspan fill="#43331f">(K⁺ channels)</tspan></text>  <path d="M80,328 C80,314 80,308 78,300" fill="none" stroke="#43331f" stroke-width="1"/> <ellipse cx="73" cy="311" rx="6" ry="3" fill="none" stroke="#43331f" stroke-width="0.7" transform="rotate(-30 73 311)"/> <ellipse cx="86" cy="316" rx="6" ry="3" fill="none" stroke="#43331f" stroke-width="0.7" transform="rotate(25 86 316)"/> <ellipse cx="78" cy="299" rx="2.4" ry="5" fill="#a9906c" stroke="#43331f" stroke-width="0.7"/> <text x="130" y="312" font-family="Georgia, serif" font-size="12.5" letter-spacing="1.5" fill="#43331f">PUNGENCY</text> <text x="238" y="312" font-family="Georgia, serif" font-size="12" font-style="italic" fill="#6f5a40">allyl isothiocyanate <tspan fill="#6f5a40" font-style="normal">—</tspan> <tspan fill="#43331f">TRPA1</tspan></text>  <g fill="none" stroke="#43331f" stroke-width="0.8"><circle cx="74" cy="388" r="3"/><circle cx="87" cy="384" r="4"/><circle cx="78" cy="371" r="2.6"/><circle cx="90" cy="374" r="2"/></g> <text x="130" y="375" font-family="Georgia, serif" font-size="12.5" letter-spacing="1.5" fill="#43331f">PRICKLE</text> <text x="238" y="375" font-family="Georgia, serif" font-size="12" font-style="italic" fill="#6f5a40">carbon dioxide <tspan fill="#6f5a40" font-style="normal">—</tspan> <tspan fill="#43331f">(trigeminal)</tspan></text> </svg> <figcaption><em>Plate 12. The chemesthesis map: heat, cold, tingle, pungency and prickle, traced to their molecules and receptors</em></figcaption> </figure>

Benign masochism, or why we alone ask for more

We are left with a genuine puzzle, and it is worth stating plainly before this chapter closes. Every sensation catalogued above — the burn of capsaicin, the electric buzz of sanshool, the sinus-clearing jolt of mustard oil — is, from the nervous system's point of view, indistinguishable from an injury. These are not appetites in the ordinary sense, built by evolution to draw us toward nourishment; they are alarms, built to drive every other creature away. And yet no rational observer of a Sichuan hotpot, a jar of English mustard, or a plate of Nashville hot chicken could conclude that the diners gathered around it are suffering. They are, by every visible measure, delighted, and they have gone out of their way, at some expense and no small discomfort, to be delighted in precisely this fashion. No other species does this voluntarily. Rats, offered capsaicin-laced food in the laboratory, avoid it as reliably as any sensible creature should.

The psychologist Paul Rozin spent decades investigating this puzzle, beginning with fieldwork in a Mexican village in the 1970s, where he and his colleague Deborah Schiller documented, systematically, how children who initially rejected chilli-laced food, exactly as their biology instructed them to, gradually came to prefer it — not despite the burn, but in some sense because of it, once repeated safe exposure had taught them that the sensation, however alarming, carried no genuine threat.[^11] Rozin later broadened this observation into a general theory he named, with a certain relish, benign masochism: the enjoyment of sensations that the body still correctly registers as unpleasant, or even as mildly threatening, once the mind has become thoroughly convinced that no real harm is on offer. In a wide survey of such preferences published in 2013 with Lily Guillot and other colleagues, Rozin found chilli heat keeping close company with a whole family of other voluntarily sought discomforts — the plunge into icy water, the crying jag induced by a sad film, the nauseating plummet of a roller-coaster, the ache of a punishing massage — all of them activities the body flags as aversive in the moment, and the mind savours precisely because it knows better.[^12] The key, in every case, is a constant, background awareness of safety: remove that awareness, tell the diner the chilli might genuinely poison them, and the pleasure collapses instantly into ordinary fear.

There is something almost philosophical in this, and it deserves to close the chapter, for chemesthesis returns us, by the most unexpected of routes, to a fact about eating we had perhaps allowed ourselves to forget. The mouth does not exist merely to identify nourishment and reject poison, sorting the world with the flat efficiency of a customs officer. It is also, uniquely among the human sensory organs, a place where danger can be safely rehearsed, tasted, and enjoyed on our own terms — a small, recurring theatre of controlled risk, staged three times a day, in which we are at once the audience and, gloriously, the fire.

We have now followed flavour a very long way from the tongue: into the architecture of a biscuit, the pitch of a crunch, the borrowed fires and frosts of a nerve that was never built to taste anything at all. And yet at every turn in this chapter, the sensation itself has proved to be only half the story — a signal that means nothing until some listening mind decides what to make of it. A crunch is not fresh; a brain, hearing it, decides fresh. A burn is not dangerous; a brain, feeling it, decides — eventually, happily — that it is not. It is time, then, to follow the trail to its true terminus, and ask what becomes of all these dispatches once they arrive at their final address. The tongue tastes, the nose smells, the trigeminal nerve tingles and burns; but it is the mind that dines, and it is to the mind, hungry and endlessly persuadable, that we turn next.

[^1]: The volunteers were, appropriately enough, eating in what the researchers dryly termed an "anechoic chamber" — a room engineered to have no echo whatsoever, the better to control exactly what reached the ear. Few crisps have ever been eaten in more scientifically rigorous silence. [^2]: The commercial fat-replacer Simplesse, developed in the 1980s, works on precisely this principle, forming microscopic protein spheres in the same size range as fat droplets, fine enough to read as creamy on the tongue despite containing almost no fat at all. [^3]: Caterina, M. J., Schumacher, M. A., Tominaga, M., Rosen, T. A., Levine, J. D., & Julius, D. (1997). "The capsaicin receptor: a heat-activated ion channel in the pain pathway." Nature, 389, 816–824. [^4]: The Nobel Prize in Physiology or Medicine 2021, awarded jointly to David Julius and Ardem Patapoutian "for their discoveries of receptors for temperature and touch." [^5]: McKemy, D. D., Neuhausser, W. M., & Julius, D. (2002). "Identification of a cold receptor reveals a general role for TRP channels in thermosensation." Nature, 416, 52–58; Peier, A. M. et al. (2002). "A TRP channel that senses cold stimuli and menthol." Cell, 108, 705–715. [^6]: Bautista, D. M. et al. (2008). "Pungent agents from Szechuan peppers excite sensory neurons by inhibiting two-pore potassium channels." Nature Neuroscience, 11, 772–779. [^7]: Hagura, N., Barber, H., & Haggard, P. (2013). "Food vibrations: Asian spice sets lips trembling." Proceedings of the Royal Society B, 280, 20131680. [^8]: Bandell, M. et al. (2004) and Jordt, S.-E. et al. (2004) independently established allyl isothiocyanate as a potent activator of TRPA1. [^9]: Chandrashekar, J. et al. (2009). "The taste of carbonation." Science, 326, 443–445. [^10]: The proteins in question are the salivary proline-rich proteins; their precipitation by tannins is the best-supported mechanism of astringency, though researchers continue to debate the contribution of direct tannin binding to the cells and membranes of the oral lining itself. [^11]: Rozin, P., & Schiller, D. (1980). "The nature and acquisition of a preference for chili pepper by humans." Motivation and Emotion, 4, 77–101. [^12]: Rozin, P., Guillot, L., Fincher, K., Rozin, A., & Tsukayama, E. (2013). "Glad to be sad, and other examples of benign masochism." Judgment and Decision Making, 8(4), 439–447.

บทถัดไป · 10 / 16Chapter 7 — The Hungry Mind