Chapter 5 — The Fifth Taste

"Osmazome is the most meritorious ingredient of all good soups." — Jean Anthelme Brillat-Savarin, The Physiology of Taste, 1825

In the autumn of 1907, a professor of chemistry at the Imperial University of Tokyo named Kikunae Ikeda sat down to a bowl of tofu simmering in a broth his wife had made in the ordinary way, from a length of dried kelp, and noticed — as he had noticed once before, years earlier, in the tomatoes, asparagus, cheese and meat broths of a Europe where he had studied — a taste that fit into none of the four boxes any chemist of his day was trained to recognise. It was not sweet, nor sour, nor salty, nor bitter, and yet it was unmistakably there: a fullness, a savoury depth, faint and easily overshadowed by louder tastes but present all the same, and present, he began to suspect, far more often across the world's cuisines than anyone had troubled to name. Ikeda did what any good naturalist does with an unclassified specimen: he went to the bench and began taking it apart.

He was not, in truth, the first person in the modern era to sense that something was there. Some eighty years earlier, on the other side of the world, the French gastronome whose motto opens every volume of this compendium had already given a name — the wrong name, drawn from an outdated chemistry, but a name nonetheless — to a savoury essence he was certain lay hidden in a good bouillon. Brillat-Savarin called it osmazome, "the purely sapid portion of flesh," and pronounced it, rightly if imprecisely, the most meritorious ingredient of any good soup. He could isolate it only crudely, by dissolving meat in cold rather than boiling water, and had no means whatsoever of identifying the molecule responsible; the word osmazome did not survive its century, quietly retired once chemistry moved on to finer instruments. But the intuition behind it — that somewhere in a good broth lurks a savoury substance separate from salt, separate from sugar, worth extracting and worth naming — was entirely sound, and it makes Brillat-Savarin's guess a kind of ghost that haunted European kitchens for the better part of a hundred years before a Japanese chemist, working from the opposite end of the world with a bowl of seaweed broth rather than a pot of bouillon, finally ran the substance to ground and gave it a name that stuck.

Ikeda's crystals

The work took the better part of a year. Ikeda began with some forty kilograms of dried kelp — konbu, the same broad kelp still used in Japanese kitchens today — and reduced it, stage by stage, evaporation by evaporation, filtration by filtration, down through a dark, tarry residue until, in the early months of 1908, he had isolated a mere thirty grams — a single ounce — of pale brown crystals. Tasted alone, dissolved in water, they reproduced exactly the savoury fullness he had first noticed in his wife's broth. Chemically, they proved to be nothing more exotic than a salt of glutamic acid, an amino acid already well known to chemistry as one of the ordinary building blocks of protein, but never before recognised as a taste in its own right. Ikeda coined a new word for what he had found, welding together two existing Japanese words: umai, delicious, and mi, taste. Umami. He filed a patent in April 1908 for a method of producing the substance as a commercial seasoning and received it that July; within a year he had gone into partnership with the businessman Saburōsuke Suzuki, and by the spring of 1909 the resulting company — in time to become the Ajinomoto Corporation — had brought the world's first purified umami seasoning to market under the name Aji-no-moto, "the essence of taste." Ikeda's own stated motive, worth recording plainly rather than skipping past, was substantially a nutritional and practical one: a nation whose traditional diet leaned heavily on rice could be made to eat more cheaply and more satisfyingly, he reasoned, if a small quantity of concentrated savouriness could season a plain meal the way a much larger quantity of meat or fish otherwise would.

The kelp itself deserves a word, since Ikeda could hardly have chosen a better specimen to interrogate. Kombu harvested from the cold waters off Hokkaido comes in several named grades — Rishiri, Ma-kombu, Rausu and Hidaka among them — prized chiefly for evenness of thickness and cleanliness of edge rather than for any real difference in savour, since a side-by-side tasting generally finds little to separate a cosmetically blemished lower grade from a pristine top one. What does change the flavour dramatically is age: the finest kombu is traditionally cellar-aged for one to three years in a practice called kuragakoi, kept cool and dim rather as a cellar keeps a cask of wine, during which its harshest, most bitter notes mellow away and its savour deepens considerably. The fine white powder that dusts a good dried sheet of kombu, and which the careful cook is instructed never to wash off before use, is not mould or road-dust at all but a bloom of mannitol and glutamic acid rising to the surface as the kelp dries — a visible dusting of the very substance Ikeda spent a year chasing in solution.

For most of the twentieth century, umami's status outside Japan remained curiously unsettled — treated by Western food science, when it was treated at all, as a mere flavour enhancer rather than a genuine fifth taste standing alongside sweet, sour, salt and bitter. Part of the trouble was linguistic: no European language had ever needed a single word for the sensation, so Western tasters, trained since childhood to sort every flavour into one of four familiar boxes, tended simply not to notice a fifth box was missing. It took the tools of molecular biology, not the tongue alone, to finally close the case Ikeda had opened almost a century earlier. In the early 2000s, researchers identified the actual receptor responsible: a pairing of two proteins called T1R1 and T1R3, sitting together on the surface of certain taste cells and forming what biologists call a heterodimer — a matched pair that functions as a receptor only when working together, rather like a two-handed grip. T1R1/T1R3 belongs to the same broad family of taste receptors as the sweet receptor described in Chapter 1, which pairs a different partner, T1R2, with that same T1R3; each is equipped with a large, hinged binding pocket, often likened to a Venus flytrap, that closes around a glutamate molecule and, in closing, fires the signal the brain reads as umami. Ikeda had been right about the sensation in 1908; it merely took the rest of the world's science the better part of a century to catch up with his tongue.

The word itself did not become the settled international term of art until later still. In October 1985, taste researchers from Japan, the United States, England, France, Switzerland, Israel and Mexico convened at a symposium on the Hawaiian island of Kauai and formally adopted umami as the standard scientific name for the taste Ikeda had identified seventy-seven years before — a small act of diplomacy that mattered more than it sounds, since a sensation with no agreed name is difficult for any science to compare notes on. The molecular receptor evidence of the early 2000s supplied the proof; the Kauai symposium supplied, rather more simply, the paperwork.

The multiplying trick

On its own, dissolved glutamate is a comparatively quiet taste — noticeable, pleasant, but rarely the thunderclap umami is capable of producing. The thunderclap requires a second ingredient, and the discovery of what that second ingredient was, and how it worked, is as much a piece of detective work as Ikeda's own.

The trail picked up again within a few years of Ikeda's patent, when his own student, Shintarō Kodama, went looking for the source of a related savoury punch in katsuobushi — the dried, smoked bonito that gives Japanese dashi its other half — and in 1913 isolated a second savoury compound entirely, a ribonucleotide called inosinate, or IMP. It sat quietly in the literature for decades until 1957, when a chemist at the Yamasa Corporation named Akira Kuninaka, investigating why dried shiitake mushrooms tasted so disproportionately savoury for a vegetable, isolated a third compound, guanylate, or GMP, and made the discovery that mattered most: mixed together in roughly equal measure, glutamate and a ribonucleotide such as inosinate or guanylate do not simply add their two savoury signals together. They multiply them.

The multiplication is not a small effect, and it has been measured carefully and repeatedly since Kuninaka's original work, in a research tradition reaching back to a landmark 1967 study by the food scientist Shizuo Yamaguchi. Combined at something close to a one-to-one ratio, glutamate and a ribonucleotide have been shown, again and again, to produce a perceived umami intensity some seven or eight times greater than either compound could achieve alone at the same concentration — a relationship researchers in the field still like to summarise, with evident relish, as "one plus one equals eight." The proportion between the two compounds matters as much as their presence: Yamaguchi's own data traced a bell-shaped curve rather than a straight line, the synergy peaking when glutamate and ribonucleotide are present in roughly similar amounts and tapering off again if either one comes to dominate the mixture completely — a finding with an unglamorous but very practical consequence in the professional kitchen, where a stock built from nothing but bones (all ribonucleotide, comparatively little free glutamate) or nothing but tomato (very nearly the reverse) will generally taste flatter than one that deliberately combines the two. It is worth flagging honestly, besides, that some later studies of particular compound pairs and concentrations have reported multipliers wider still — inosinate has been shown, under some conditions, to lower the threshold at which the tongue can first detect glutamate at all by a factor of fifteen or more — so seven- or eightfold is best understood as the sturdy, oft-replicated headline figure for a roughly equal mixture, not a hard ceiling the effect never exceeds. The molecular reason, so far as structural biology has been able to establish it, is elegant: glutamate and the ribonucleotide do not compete for the same docking site on the T1R1/T1R3 receptor at all, but bind at two separate pockets, and the ribonucleotide's role, once glutamate is already seated, is essentially to reach across and clamp the receptor's hinge shut — prolonging and intensifying a signal that would otherwise release almost as soon as it began, rather than adding some fresh signal of its own.

This single piece of chemistry is the reason dashi — kelp and bonito flakes steeped together in hot water for a matter of minutes — tastes so peculiarly, disproportionately alive for something so simple, and Japanese cooks arrived at the combination by instinct and generations of repetition centuries before Kodama ever isolated the second half of the equation. Kelp alone makes a mild, faintly savoury stock; bonito flakes alone make another; the two together make something that tastes, on the first sip, like considerably more than the sum of two mild things. The same trick, discovered independently and by instinct rather than by chemistry, turns up wherever a cuisine has stumbled onto a glutamate-rich ingredient and a ribonucleotide-rich one and pushed them together: grated Parmesan, thick with free glutamate from months of ageing, stirred into a ragù built on slow-cooked, inosinate-rich meat; a tomato sauce, its glutamate concentrating as it reduces, simmered for hours with that same meat; dried mushroom, rich in guanylate, dropped into a stock already carrying inosinate of its own. Food scientists have even proposed the same synergistic logic at work in a pairing that looks, on paper, nothing like a bowl of dashi: a 2020 study argued that the celebrated match of oysters and Champagne rests on precisely this chemistry, the oyster's natural glutamate and the wine's own trace amino acids and nucleotides meeting on the tongue and, however unwittingly, doing exactly what kombu and katsuobushi do in a Kyoto kitchen.

<figure> <svg xmlns="http://www.w3.org/2000/svg" width="100%" viewBox="0 0 640 440" role="img"> <title>The Synergy of Savour</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> <pattern id="hs2" width="3" height="3" patternUnits="userSpaceOnUse" patternTransform="rotate(45)"><line x1="0" y1="0" x2="0" y2="3" stroke="#c8b79a" stroke-width="0.4"/></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="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 SYNERGY OF SAVOUR</text> <line x1="245" y1="49" x2="395" 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">Fig. I — the umami receptor T1R1–T1R3, and its amplification</text>  <g> <text x="165" y="95" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">glutamate alone</text> <line x1="165" y1="100" x2="165" y2="112" stroke="#6f5a40" stroke-width="0.5"/> <path d="M154,126 A13,13 0 1 1 176,126 L172,144 L165,168 L158,144 Z" fill="url(#hs1)" stroke="#43331f" stroke-width="1"/> <line x1="120" y1="112" x2="150" y2="118" stroke="#6f5a40" stroke-width="0.5"/> <text x="118" y="110" text-anchor="end" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">T1R1–T1R3</text>  <circle cx="95" cy="126" r="9" fill="none" stroke="#43331f" stroke-width="1"/> <line x1="104" y1="126" x2="153" y2="126" stroke="#43331f" stroke-width="1"/> <line x1="143" y1="126" x2="143" y2="133" stroke="#43331f" stroke-width="1"/> <line x1="149" y1="126" x2="149" y2="130" stroke="#43331f" stroke-width="1"/> <text x="95" y="150" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="10" fill="#6f5a40">glutamate</text>  <line x1="95" y1="230" x2="235" y2="230" stroke="#c8b79a" stroke-width="0.5"/> <path d="M100,230 q6,-7 12,0 t12,0 12,0 12,0 12,0 12,0 12,0 12,0" fill="none" stroke="#43331f" stroke-width="1"/> <text x="165" y="258" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="10" fill="#6f5a40">response, modest</text> </g>  <line x1="320" y1="80" x2="320" y2="270" stroke="#c8b79a" stroke-width="0.6"/> <text x="320" y="175" text-anchor="middle" font-family="Georgia, serif" font-size="18" fill="#9c5a2c">+</text>  <g> <text x="475" y="95" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">glutamate & inosinate together</text> <line x1="475" y1="100" x2="475" y2="112" stroke="#6f5a40" stroke-width="0.5"/> <path d="M464,126 A13,13 0 1 1 486,126 L479,146 L500,158 L479,150 L475,168 L471,150 L450,158 L471,146 Z" fill="url(#hs1)" stroke="#43331f" stroke-width="1"/> <line x1="518" y1="112" x2="490" y2="120" stroke="#6f5a40" stroke-width="0.5"/> <text x="520" y="110" text-anchor="start" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">T1R1–T1R3</text> <text x="520" y="163" text-anchor="start" font-family="Georgia, serif" font-style="italic" font-size="10" fill="#9c5a2c">inosinate (IMP)</text> <text x="520" y="175" text-anchor="start" font-family="Georgia, serif" font-style="italic" font-size="9" fill="#6f5a40">clamps the lock wider</text> <line x1="518" y1="158" x2="500" y2="158" stroke="#9c5a2c" stroke-width="0.5"/>  <circle cx="405" cy="126" r="9" fill="none" stroke="#43331f" stroke-width="1"/> <line x1="414" y1="126" x2="463" y2="126" stroke="#43331f" stroke-width="1"/> <line x1="453" y1="126" x2="453" y2="133" stroke="#43331f" stroke-width="1"/> <line x1="459" y1="126" x2="459" y2="130" stroke="#43331f" stroke-width="1"/> <text x="405" y="150" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="10" fill="#6f5a40">glutamate</text>  <line x1="405" y1="230" x2="545" y2="230" stroke="#c8b79a" stroke-width="0.5"/> <path d="M410,230 q9,-30 18,0 t18,0 18,0 18,0 18,0" fill="none" stroke="#9c5a2c" stroke-width="1.1"/> <text x="475" y="258" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="10" fill="#6f5a40">response, greatly amplified</text> </g>  <g> <path d="M250,300 h140 l14,14 l-14,14 h-140 l-14,-14 Z" fill="none" stroke="#43331f" stroke-width="0.8"/> <text x="320" y="322" text-anchor="middle" font-family="Georgia, serif" font-size="16" fill="#43331f">1 + 1 = 8</text> </g> <text x="320" y="345" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="10.5" fill="#6f5a40">synergy, not sum</text> <line x1="180" y1="380" x2="460" y2="380" stroke="#c8b79a" stroke-width="0.5"/> <text x="320" y="403" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="10" fill="#6f5a40">the ribonucleotide does not merely add its own savour — it multiplies the receptor's answer</text> </svg> <figcaption><em>Plate 9. The synergy of taste: glutamate and ribonucleotide binding together at the umami receptor, multiplying the signal rather than merely adding to it</em></figcaption> </figure>

Savouriness without the word

No cuisine on earth needed Ikeda's word to go looking for what the word describes, and a naturalist surveying the world's larders finds umami's fingerprints on virtually every table, arrived at by entirely independent roads — fermentation, drying, curing, and the sheer patience of long cooking, each one a different method of doing the same underlying chemistry: breaking bound protein down into the free amino acids, glutamate chief among them, that the tongue can actually taste.

Aged, hard cheeses are perhaps the most immediate example. A well-aged Parmigiano-Reggiano carries an exceptional load of free glutamate, released slowly over many months as the cheese's proteins are broken down by its own enzymes — worth stating plainly, because a persistent piece of kitchen folklore gets the chemistry backwards. The white, gritty crystals scattered through the interior of a well-aged wedge, the ones that crunch faintly between the teeth, are not, as is sometimes claimed, crystallised glutamate; they are tyrosine, a different amino acid entirely, one that happens to be poorly soluble and so precipitates into visible crystals once concentrated enough, while the actual umami-bearing glutamate remains fully dissolved throughout the cheese and entirely invisible to the eye. The crunch and the savouriness are near neighbours, products of the same slow proteolysis, but they are not one and the same phenomenon — a distinction worth drawing with the same care this compendium gave the tongue map in Chapter 1.

Fish, too, has been fermented into pure umami since antiquity, by a route with an unusually long and well-documented history. The ancient Greeks and Romans built an entire condiment industry around garum and its cheaper cousin liquamen — small fish layered with salt in open vats and left to break down slowly in the sun for weeks or months, their proteins surrendering free amino acids as they liquefied — and Roman cooks, to judge from the surviving recipes of Apicius, splashed it onto nearly everything. Garum itself largely vanished from European kitchens by around the ninth century, but the identical principle re-emerged, whether by independent invention or by a chain of contact historians still debate, across South-East Asia, where it survives today as Thai nam pla and Vietnamese nước mắm; a single Mediterranean holdout, colatura di alici from Italy's Amalfi coast, still ferments anchovies in salt by a method its makers freely describe as all but identical to the Roman original. The English kitchen, without ever borrowing the Roman word, keeps a quieter version of the same inheritance: Worcestershire sauce, built on a base of anchovies fermented in salt for as long as three to five years before being barrelled with vinegar, molasses and tamarind, owes its savoury backbone to precisely the same slow breakdown of fish protein into free glutamate that built garum two thousand years before it — a small, unadvertised piece of the Roman larder sitting, to this day, in most British kitchen cupboards.

Soy sauce and miso perform the same transformation on soybeans rather than fish, using a cultivated mould, koji, to do the enzymatic labour that salt and sun perform on garum. A well-brewed soy sauce may spend six months to a year fermenting and maturing in cedar vats before bottling, its koji enzymes patiently unlocking the soybean's bound protein into free amino acids by the same slow arithmetic as a good cheese or a good ham; the cheapest soy sauces on any supermarket shelf achieve a broadly similar amino acid profile by breaking the same protein down with hydrochloric acid in a day or two rather than months, a shortcut that yields much of the savour but few of the several hundred further aroma compounds a genuine fermentation generates along the way, which is why the practised palate can usually tell the two apart at a glance down the ingredients label, let alone at the table.

The tomato, so central to a cuisine that never named umami either, turns out to be a small glutamate factory in its own right, and an instructive one, because its savouriness visibly builds rather than simply being present or absent. A hard, underripe tomato carries comparatively little free glutamate; as it ripens on the vine the amount rises sharply, which is one reason a properly vine-ripened tomato tastes so much rounder and more savoury, not merely sweeter, than one picked green for shipping. Drying concentrates the effect dramatically further — a sun-dried tomato has been measured carrying roughly four times the free glutamate of the fresh fruit it was made from — which is the real, chemical reason a spoonful of tomato paste or a scatter of sun-dried tomato can anchor an entire sauce that would otherwise taste thin.

Mushrooms perform a similar trick with the other half of the synergy. Fresh shiitake carry very little guanylate; drying is what creates it, the mushroom's own enzymes converting other compounds into the ribonucleotide as moisture leaves the flesh, which is why a handful of dried shiitake, reconstituted in a meat or vegetable stock already rich in glutamate, produces a savouriness well beyond what either ingredient could manage alone — Kuninaka's 1957 discovery, still working quietly in kitchens that never heard his name.

Cured meats complete the picture by the slowest route of all. A jamón or a Parma ham, hung and aged for a year or more, undergoes months of gradual proteolysis as the meat's own enzymes patiently disassemble its proteins into peptides and free amino acids, glutamate prominent among them; the deep, savoury, faintly sweet complexity prized in a well-aged ham is, in significant part, literally the taste of protein that has already been broken halfway towards digestion before it ever reaches the plate — a phrase worth remembering, since it will matter again before this chapter closes.

And then there is Marmite, and its yeast-extract cousins around the world, which arrived at concentrated umami by an entirely different and more recently invented door. The German chemist Justus von Liebig discovered in the late nineteenth century that brewer's yeast, broken down by its own enzymes in a process called autolysis, yielded a dark, intensely savoury paste rich in both free glutamate and ribonucleotides together — the synergy pairing again, this time arrived at industrially rather than through a cuisine's slow trial and error. Marmite itself has been sold in Britain since 1902, and the loyalty and disgust it provokes in roughly equal measure are, chemically speaking, an argument about the intensity of a single taste more than about any of its individual notes.

<figure> <svg xmlns="http://www.w3.org/2000/svg" width="100%" viewBox="0 0 640 440" role="img"> <title>A Cabinet of Savoury Specimens</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> <pattern id="hs2" width="3" height="3" patternUnits="userSpaceOnUse" patternTransform="rotate(45)"><line x1="0" y1="0" x2="0" y2="3" stroke="#c8b79a" stroke-width="0.4"/></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="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="36" text-anchor="middle" font-family="Georgia, 'Times New Roman', serif" font-size="14" letter-spacing="1.5" fill="#43331f">A CABINET OF SAVOURY SPECIMENS</text> <line x1="215" y1="46" x2="425" y2="46" stroke="#6f5a40" stroke-width="0.5"/>  <rect x="30" y="66" width="580" height="340" fill="none" stroke="#43331f" stroke-width="0.8"/> <line x1="223" y1="66" x2="223" y2="406" stroke="#6f5a40" stroke-width="0.5"/> <line x1="416" y1="66" x2="416" y2="406" stroke="#6f5a40" stroke-width="0.5"/> <line x1="30" y1="236" x2="610" y2="236" stroke="#6f5a40" stroke-width="0.5"/>  <g> <text x="40" y="80" font-family="Georgia, serif" font-size="9" fill="#6f5a40">I</text> <path d="M124,90 C136,108 114,126 128,146 C140,164 116,182 124,204 L132,204 C126,184 148,166 136,146 C124,128 146,110 132,90 Z" fill="url(#hs2)" stroke="#43331f" stroke-width="0.8"/> <path d="M126,94 C132,112 116,128 128,146 C138,164 120,182 126,200" fill="none" stroke="#6f5a40" stroke-width="0.5"/> <line x1="119" y1="110" x2="112" y2="106" stroke="#6f5a40" stroke-width="0.5"/> <line x1="132" y1="140" x2="140" y2="137" stroke="#6f5a40" stroke-width="0.5"/> <line x1="119" y1="172" x2="111" y2="169" stroke="#6f5a40" stroke-width="0.5"/> <text x="126" y="222" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">kombu kelp</text> </g>  <g> <text x="233" y="80" font-family="Georgia, serif" font-size="9" fill="#6f5a40">II</text> <rect x="292" y="132" width="34" height="26" fill="url(#hs1)" stroke="#43331f" stroke-width="0.8"/> <path d="M292,132 L304,120 L338,120 L326,132 Z" fill="url(#hs2)" stroke="#43331f" stroke-width="0.7"/> <path d="M326,132 L338,120 L338,146 L326,158 Z" fill="url(#hs1)" stroke="#43331f" stroke-width="0.7"/> <path d="M348,138 q10,-6 3,-15" fill="none" stroke="#6f5a40" stroke-width="0.6"/> <path d="M352,150 q11,-4 6,-14" fill="none" stroke="#6f5a40" stroke-width="0.6"/> <path d="M347,160 q10,-3 6,-12" fill="none" stroke="#6f5a40" stroke-width="0.6"/> <text x="319" y="222" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">katsuobushi, shaved</text> </g>  <g> <text x="426" y="80" font-family="Georgia, serif" font-size="9" fill="#6f5a40">III</text> <circle cx="513" cy="148" r="22" fill="url(#hs2)" stroke="#43331f" stroke-width="1"/> <path d="M513,148 L500,134 M513,148 L506,128 M513,148 L522,127 M513,148 L529,136 M513,148 L531,148 M513,148 L497,146" fill="none" stroke="#6f5a40" stroke-width="0.5"/> <path d="M506,170 L520,170 L516,186 L510,186 Z" fill="url(#hs1)" stroke="#43331f" stroke-width="0.8"/> <text x="513" y="222" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">dried shiitake</text> </g>  <g> <text x="40" y="250" font-family="Georgia, serif" font-size="9" fill="#6f5a40">IV</text> <path d="M96,326 L148,298 A50,50 0 0 1 148,354 Z" fill="url(#hs2)" stroke="#43331f" stroke-width="0.8"/> <path d="M148,298 A50,50 0 0 1 148,354" fill="none" stroke="#43331f" stroke-width="1.2"/> <circle cx="118" cy="320" r="0.9" fill="#43331f"/><circle cx="126" cy="332" r="0.9" fill="#43331f"/><circle cx="112" cy="334" r="0.9" fill="#43331f"/> <circle cx="132" cy="316" r="0.9" fill="#43331f"/><circle cx="122" cy="342" r="0.9" fill="#43331f"/><circle cx="106" cy="322" r="0.9" fill="#43331f"/> <circle cx="136" cy="336" r="0.9" fill="#43331f"/> <text x="126" y="392" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">parmesan, aged</text> </g>  <g> <text x="233" y="250" font-family="Georgia, serif" font-size="9" fill="#6f5a40">V</text> <path d="M295,305 C310,295 335,297 345,310 C352,320 348,335 336,345 C322,353 302,350 293,338 C286,328 288,313 295,305 Z" fill="url(#hs1)" stroke="#43331f" stroke-width="0.8"/> <path d="M298,312 C308,314 320,310 332,316" fill="none" stroke="#c8b79a" stroke-width="0.6"/> <path d="M296,325 C310,329 326,323 340,328" fill="none" stroke="#c8b79a" stroke-width="0.6"/> <path d="M300,337 C312,340 324,336 334,340" fill="none" stroke="#c8b79a" stroke-width="0.6"/> <text x="319" y="392" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">cured ham, sliced</text> </g>  <g> <text x="426" y="250" font-family="Georgia, serif" font-size="9" fill="#6f5a40">VI</text> <circle cx="513" cy="326" r="24" fill="url(#hs2)" stroke="#43331f" stroke-width="1"/> <circle cx="513" cy="326" r="5" fill="none" stroke="#6f5a40" stroke-width="0.5"/> <line x1="513" y1="326" x2="513" y2="303" stroke="#6f5a40" stroke-width="0.5"/> <line x1="513" y1="326" x2="533" y2="316" stroke="#6f5a40" stroke-width="0.5"/> <line x1="513" y1="326" x2="533" y2="337" stroke="#6f5a40" stroke-width="0.5"/> <line x1="513" y1="326" x2="513" y2="349" stroke="#6f5a40" stroke-width="0.5"/> <line x1="513" y1="326" x2="493" y2="337" stroke="#6f5a40" stroke-width="0.5"/> <line x1="513" y1="326" x2="493" y2="316" stroke="#6f5a40" stroke-width="0.5"/> <circle cx="513" cy="312" r="0.9" fill="#9c5a2c"/><circle cx="524" cy="320" r="0.9" fill="#9c5a2c"/><circle cx="520" cy="334" r="0.9" fill="#9c5a2c"/> <circle cx="503" cy="331" r="0.9" fill="#9c5a2c"/><circle cx="506" cy="317" r="0.9" fill="#9c5a2c"/> <text x="513" y="392" text-anchor="middle" font-family="Georgia, serif" font-style="italic" font-size="11" fill="#43331f">tomato, halved</text> </g> </svg> <figcaption><em>Plate 10. A cabinet of savoury specimens: kombu, katsuobushi, dried shiitake, Parmesan, cured ham and the ripening tomato — umami's many disguises around the world</em></figcaption> </figure>

The Ajinomoto story, told fairly

No account of umami is complete, or honest, without following Ikeda's own invention forward into the one controversy that has dogged it for more than half a century, and this compendium's own rule on the matter is worth restating before we begin: state the science calmly, and do not present a settled question as though it remained open merely because it is emotionally loaded.

Ikeda's seasoning, monosodium glutamate — the sodium salt of the very glutamic acid he had isolated from kelp — was produced by direct extraction from wheat or soy protein for its first half-century on the market, a laborious process descended from Ikeda's own. That changed in 1957, when the Japanese microbiologists Shukuo Kinoshita and Sadao Udaka identified a soil bacterium, Corynebacterium glutamicum, that could be coaxed to excrete glutamic acid directly as it fermented ordinary sugars, much as yeast excretes alcohol or a souring dough's lactobacilli excrete lactic acid; commercial fermentation production began the following year, and fermentation has been the standard industrial method ever since. The great majority of the world's MSG today is brewed, in other words, rather than extracted — kin, in method if not in glamour, to the soy sauce and miso already discussed in this chapter. However it was made, the seasoning spread through Japanese and then global cooking across the twentieth century as a cheap, reliable way of adding the savoury dimension a stock or broth otherwise took hours to build. It attracted little controversy at all until the fourth of April 1968, when a physician named Robert Ho Man Kwok wrote a short letter to the New England Journal of Medicine, wondering in print about a cluster of symptoms — numbness, weakness, palpitations — he experienced roughly twenty minutes after eating at Chinese-American restaurants, and speculating that the cause might be the cooking wine, the salt, or possibly the MSG used liberally in that style of cooking. The journal soon attached a memorable label to the letter, "Chinese restaurant syndrome," and within a month a flurry of further anecdotal letters from other physicians, each reporting similar unverified experiences, had turned a single speculative note into a durable piece of popular medical folklore.

It is worth pausing on what that folklore actually did, because the pattern is instructive as well as unfair. Suspicion attached itself overwhelmingly to Chinese cooking specifically, rather than to the compound wherever it appeared — a can of condensed soup, a bag of flavoured crisps, an Italian dish thick with Parmesan, all containing glutamate in comparable or greater quantity, drew none of the same alarm — and food historians, along with Ajinomoto's own later public-education campaigns, have both pointed out, reasonably, that a purely chemical fear rarely distributes itself so selectively by cuisine unless something other than chemistry is doing part of the work. None of which is to say Kwok's original letter was written in bad faith; he described a real, if never satisfactorily explained, personal experience, and did no more than speculate about its cause in a short note plainly framed as a question rather than a finding.

The half-century of research that followed answers that question about as clearly as nutritional science ever manages to answer anything. Repeated double-blind, placebo-controlled trials — the only design capable of separating a genuine chemical effect from expectation — have consistently failed to reproduce "Chinese restaurant syndrome" reliably under controlled conditions; a small number of subjects, given very large doses of MSG on an empty stomach with no food at all, have occasionally reported mild, transient symptoms, but no consistent, reproducible pattern linking ordinary dietary use to real harm has ever been established. Regulatory bodies, including America's Food and Drug Administration, classify MSG as generally recognised as safe, and state plainly that no scientific evidence establishes it as causing severe or dangerous reactions at the levels found in food. And the underlying chemical fact is the one this compendium's own accuracy rules insist on stating without hedging: the glutamate ion released into solution by a spoonful of MSG is structurally, chemically, and in every measurable respect identical to the glutamate ion released by a splash of soy sauce, a simmering stock, or a well-aged wedge of Parmesan — a tongue, or a gut, has no means whatsoever of telling the three apart, because there is nothing to tell apart. Weight for weight, a single tablespoon of grated Parmesan commonly carries more free glutamate than a typical restaurant portion seasoned with added MSG. Ikeda's crystal, in other words, was never a foreign substance smuggled into food; it was, and remains, the same molecule the tongue has been quietly enjoying since long before anyone thought to isolate it, purify it, or worry about it.

Umami, satiety, and the moreish

Umami's reach turns out not to stop at the tongue. The same T1R1/T1R3 receptor pairing that fires when glutamate meets a taste bud has also been found further down, along the lining of the stomach and intestine, where it appears to serve a related but distinct purpose: helping the gut register that protein, specifically, has arrived in a meal, rather than merely calories in general. When these gut receptors are activated, the evidence so far suggests they contribute to a chain of hormonal signalling — prompting the release of cholecystokinin, a hormone long associated with the sensation of fullness, and feeding information back to the brain along the vagus nerve, the same long, wandering nerve that carries a great deal of the gut's running commentary to the brain in Chapter 7's account of how flavour is finally assembled. This is an active and still-developing area of research rather than a closed one, and the honest picture it currently offers is genuinely two-sided: some studies find that a savoury, umami-rich course can increase a diner's appetite and enjoyment of a meal in the moment, while others find that the same umami content promotes a stronger sense of fullness somewhat later on — both effects appearing to be real, rather than one cancelling the other out, which is a more interesting and more honest finding than either claim alone would be.

There is, underneath the biochemistry, a simpler and older logic worth returning to, one already sketched in Chapter 1's account of why the basic tastes exist at all: sweetness evolved as an energy detector, bitterness as a poison alarm, and umami, on the same reading, as a protein detector — a tongue's way of confirming that a food contains the amino acids an animal's body cannot manufacture for itself and must, therefore, eat. A steak, a good stock, a well-aged cheese, a bowl of proper dashi: each is, in this light, less a flavour experience than a nutritional announcement, the tongue's oldest and most reliable way of saying yes, this is what you were built to seek out. It is not a bad working theory for why deeply savoury food so often produces the particular, slightly helpless pleasure English speakers reach for the word "moreish" to describe — a pleasure less like enjoying a treat than like being, in some small and ancient sense, fed.

There is a more immediate, and more practical, consequence of all this for the working kitchen, and it returns the argument directly to the seasoning compass of the previous chapter. Because umami-rich ingredients measurably increase a dish's perceived saltiness and overall palatability independent of its actual sodium content, food scientists have begun studying them deliberately as a lever for reducing salt without reducing satisfaction. In one Japanese trial, a clear soup's salt concentration was cut by two-thirds — from a fairly standard 0.9 or 0.6 per cent down to 0.3 per cent — without any loss in tasters' enjoyment, simply by supplementing the broth with free glutamate; the result was a sodium reduction of over sixty per cent for no perceptible sacrifice in flavour. Comparable trials have reported reductions of thirty per cent or more in rice and in ground-meat dishes seasoned the same way, with no fall in acceptability. None of this licenses a cook to empty the salt cellar into a drawer — glutamate sharpens savouriness, but it is no substitute for salt's own particular tricks of suppression and revelation explored in the previous chapter — only that the two levers, salt and umami, appear to support one another closely enough that a well-judged dose of the second can quietly stand in for a meaningful fraction of the first.

Ichiban Dashi — First Stock

The simplest possible demonstration of everything argued in this chapter: two mild, dried, unglamorous ingredients, one carrying glutamate and the other inosinate, combined in nothing more than hot water for a matter of minutes. Dashi underpins Japanese cooking the way a good stock underpins French cooking, and it takes a tenth of the time.

Serves 4, as a soup base (makes about 1 litre/4 cups) · Time 20 minutes, mostly waiting

Ingredients

10g dried kombu (one piece roughly 10cm/4in square)
1 litre (4 cups) cold water
15–20g katsuobushi (dried bonito flakes; a generous handful, loosely packed)

Method

Wipe the kombu gently with a dry or barely damp cloth to remove any grit, but do not scrub away the fine pale bloom on its surface — much of its savour lives there. Submerge it in the cold water and leave to stand at least 30 minutes, or overnight in the refrigerator, giving the glutamate time to migrate slowly into the water.
Set the pan over medium heat and warm gently until small bubbles just begin to gather at the edges — the point just short of a simmer. Lift out the kombu at this moment; boiling it hard draws out bitter, gluey compounds along with the flavour wanted. (The spent kombu can be simmered again for a second, weaker dashi, or sliced thinly into a side dish.)
Bring the kombu-infused water to a full boil, then remove the pan from the heat entirely and scatter in the katsuobushi.
Let the flakes steep undisturbed for about 2 minutes and no longer — do not stir, or the dashi will turn cloudy and faintly bitter as fine particles break apart.
Strain through a fine sieve, or a sieve lined with muslin, into a clean bowl, pressing the flakes very lightly rather than wringing them.

Cook's note: What has just happened in the pan is this chapter's entire argument, performed in twenty minutes rather than explained in several thousand words: glutamate from the kombu and inosinate from the katsuobushi have gone into solution together and multiplied each other several-fold, so that a broth built from two humble, inexpensive and individually rather quiet ingredients tastes, on the first spoonful, disproportionately alive. Dashi keeps three or four days under refrigeration, and underlies miso soup, simmered vegetable and fish dishes, and the dipping sauces served alongside cold noodles.

Field Note — Katsuobushi

Few ingredients anywhere earn a nickname as improbable as "the hardest food in the world," and katsuobushi — dried, smoked and, in its finest grades, fermented skipjack tuna — wears the claim more or less literally.[^1] The making of it is a small marvel of patience even by the standards of a cuisine that prizes patience highly: fresh bonito is filleted, simmered for the better part of two hours, deboned, and then smoked over a wood fire in repeated sessions — sometimes a dozen or more, spread across several weeks — each smoking followed by a careful scraping and repair of the surface, until the flesh has lost the greater part of its moisture and taken on the dense, almost mineral hardness of old timber. The finest grade, karebushi, goes a step further still: the smoked block is inoculated with a cultivated mould related to the ones used in sake and soy production, then left to age for months in a cycle of mould growth, sun-drying and brushing, the mould quietly consuming the last of the surface fat and moisture and, in doing so, concentrating the savour within. What began as a fish the size of a forearm ends, after half a year or more, as a blackened block that rings almost like wood when struck, and must be shaved, not sliced, with a special wooden plane called a kezuriki to yield anything usable at all. The shavings that fall from that plane bear only a passing resemblance to the pre-packaged flakes most cooks outside Japan will ever meet: freshly cut katsuobushi carries a smoky, faintly sweet cloud of aroma that begins fading within minutes of exposure to the air, oxidising quietly away, so that a bag of flakes bought weeks earlier — however carefully sealed — is tasting the ghost of a much livelier thing. It was this same block, or an ancestor of it, that Ikeda's student Shintarō Kodama placed under scrutiny in 1913 and found, to nobody's great surprise in hindsight, harbouring the second half of umami's great synergy.

Five tastes, now, stand confirmed on the naturalist's list — sweet, sour, salt, bitter, and this deep, savoury fifth that waited a hundred years for its passport. But taste, as the reader will recall from the very first chapter, was always the lesser part of the story: five doors on the tongue, set against some four hundred kinds of receptor in the nose alone. What remains unaccounted for is everything the mouth feels rather than tastes — the crackle of a good crust, the cool of menthol, the outright assault of a chilli pepper upon nerves that were never built to taste anything at all. It is to those nerves, and their particular variety of benign masochism, that we turn next.

[^1]: The comparison most often repeated is a Mohs hardness of roughly 7–8, harder than iron — a striking figure that turns up across a great many popular accounts of katsuobushi, though this compendium was unable to trace it to a formal, citable measurement, and it is offered here in the spirit in which it usually circulates: as a vivid piece of kitchen folklore rather than a verified laboratory fact.

บทถัดไป · 9 / 16Chapter 6 — Texture, Mouthfeel & the Trigeminal Thrill