With so much going for it, you would think water alone would be more than adequate as a braising liquid. And it is, sometimes. In fact, Samin was of the opinion that tap water was underrated as a braising liquid, while chicken stock, the default in most kitchens, was used way too much.*

"I don't understand why you would want everything you braise to taste like chicken, unless you're braising chicken," Samin mentioned one afternoon, when we were getting ready to put a Moroccan lamb stew into the oven. The dish already promised plenty of flavor. To its base of mirepoix and garlic, we had added a bunch of toasted Moroccan spices, and then laid out some orange peels, dried apricots, cilantro stems, and, on top of that fragrant bed, the well-browned cuts of lamb. So we dispensed with stock and used some water, and a splash of white wine, instead. "Eventually that liquid is going to turn into something rich and delicious-it doesn't need to taste like chicken!"

As the continuous phase in our lamb stew, water's role is to blend and balance some pretty wild flavors, forging them into a familiar sense experience: the flavor of Moroccan food. Most of us instantly recognize such basic flavor profiles, and indeed depend on them to figure out what we're eating and to feel comfortable doing so. If the omnivore's dilemma is to determine what is good and safe to eat amid the myriad and occasionally risky choices nature puts before us, then familiar flavor profiles can serve as a useful guide, a sensory signal of the tried and true. To an extent, these familiar blends of flavor take the place of the hardwired taste preferences that guide most other species in their food choices. They have instincts to steer them; we have cuisines.

This at least is the theory of culinary flavor advanced by Elisabeth Rozin, the cookbook writer, and her former husband, Paul Rozin, the social psychologist. "Flavoring a dish with soy sauce, for example, almost automatically identifies it as Oriental," she points out in her book Ethnic Cuisine: The Flavor-Principle Cookbook. But the sprawling Eastern empire of soy has many nations within it: "If you add garlic, mola.s.ses, ground peanuts and chilies to the basic soy sauce, you will create a taste characteristically Indonesian," she points out. And if fish sauce and coconut milk are added instead, the dish becomes Laotian, and so forth. Every cuisine has its characteristic "flavor principle," Rozin contends, whether it is tomato-lemon-oregano in Greece; lime-chili in Mexico; onion-lard-paprika in Hungary, or, in Samin's Moroccan dish, c.u.min-coriander-cinnamon-ginger-onion-fruit. (And in America? Well, we do have Heinz ketchup, a flavor principle in a bottle that kids, or their parents, use to domesticate every imaginable kind of food. We also now have the familiar salty-umami taste of fast food, which I would guess is based on salt, soy oil, and MSG.) But as soon as we encounter a familiar flavor principle, we know what we're eating and can relax in the knowledge that our dinner has been prepared according to a set of time-tested rules, and so probably won't kill us or make us sick.

These flavor principles always involve the marriage of at least two aromatic plants and often many more. That may be because no single seasoning can ever mark a food as having completed the necessary journey from the risky realm of raw nature to the safety of cooked culture. What we seem drawn to is the combination of flavors that only h.o.m.o sapiens, experimenting over time, could concoct from whatever nature has to offer locally. And much like any other artifact of culture-a vase, a melody-these combinations most appeal to us when they exhibit a kind of balance or symmetry-in this case, between sweet and sour, say, or bitter and salty.

Particularly in the case of the more elaborate combinations of flavor, such as in our Moroccan stew, the greatest conductor of flavor principles is the element of water: it is what weaves together the differently colored threads of taste into a familiar pattern, gives them their unity. A cooking oil can achieve somewhat similar results (and is often itself an important element in a flavor principle), but water is the principle medium of taste; indeed, for the tongue to taste any molecule, it must be soluble in water. (Strictly speaking, "taste" is limited to one or more of the five senses perceptible to the tongue: sweet, salt, sour, bitter, and umami. Flavor is a broader category, encompa.s.sing smells as well as taste, with the result that our response to it depends less on our genes than on our experience.)

But if plain old water can do all this for the flavor of a stew or soup or sauce, then why do so many cuisines resort so often to animal-enhanced waters, in the form of a stock or broth? Cooks will tell you that stocks add a quality of richness or intensity or "depth" to a braise or stew or sauce, making a savory dish that much more savory. It also adds "body," or substance. "Stock is everything in cooking," the great French chef Auguste Escoffier famously declared, "Without it, nothing can be done." This is why many great restaurants employ a "saucier," a cook whose entire job consists of making stocks. To buy such a foundational ingredient would be out of the question.

It's curious that this one ingredient of a dish consists, in effect, of a whole other dish-one with its own recipe, its own cook pot, its own liquid, and its own foundation of aromatic vegetables, notably including our old friends onion, carrot, and celery. To make a stock to add to a braise or a sauce, which Samin and I did on several occasions, feels like embarking on an infinite culinary regression, taking us all the way back, again, to the chopping of onions, browning of meat, and adding of liquid. But this process of repeating reductions-cooking things down in water, extracting their essences, and then reducing them yet again-seems to be how the deepest, purest layers of flavor are formed.

So what exactly is it about stock that makes it so indispensable? What does it really mean to say it gives "body" or "depth" to a pot dish or sauce, or makes something taste more "savory," as stocks undeniably do? What, in other words, is so special about this particular liquid we call stock?

I suspect it's something more than the flavor of the animal (or vegetables) on which the stock is based. As Samin's feelings about chicken stock suggest, the flavor of chicken is not necessarily a plus, and often goes unnoticed in the finished dish anyway. Indeed, one of the reasons that chicken or veal is usually the default stock owes to their relative lack of flavor, at least when compared with beef or pork, as well as to the fact that their young bones contribute comparatively more gelatin to a dish or sauce, thereby adding to their body. But there had to be something more to it than that, and after I spent some time researching the chemistry of meat stocks and the physiology of our sense of taste, it became (you'll forgive me) clear as consomme: The most important quality that a long-simmered stock contributes to any dish to which it has been added is the seductive and still somewhat mysterious fifth taste called umami.

Umami has been recognized as a full-fledged taste in j.a.pan since 1908. That was the year that a chemist named Kikunae Ikeda discovered that the white crystals that form on dried kombu, a kind of seaweed that has been used as a base for soups and stocks in j.a.pan for more than a thousand years, contained large amounts of glutamate, and that the savory taste of this molecule was sui generis-was not sweet or sour or bitter or salt. Ikeda decided to call the taste umami-the j.a.panese word for "deliciousness." Today, most of us encounter glutamate on ingredient labels in the form of a salt called monosodium glutamate, or MSG.*

The idea of a fifth taste was controversial in the West until 2001, when American scientists identified a dedicated taste receptor for glutamate on the human tongue. Now umami is generally recognized as a distinct taste, as is the fact that, in addition to glutamate, at least two other molecules, including the nucleotides inosine (found in fish) and guanosine (found in mushrooms) also contribute to a perception of umami. When combined, these chemicals seem to have a synergistic effect, dramatically intensifying the umami taste.

Like the other four tastes that have been identified in mammals, umami is actually a discrete sense. In each case, we are born with dedicated receptors that are wired to regions of the brain primed to respond in a specific way. Thus no one needs to "learn" the taste of sweetness or recognize it as positive. It is innate. Olfaction operates quite differently: Humans can sniff out some ten thousand smells and how we respond to each of them is largely the result of learning, individual and cultural. A smell that is appealing in one culture-that of rotted tofu, say, to which I was treated in China-may be absolutely disgusting in another. The difference between innate taste and learned smell is encoded in our language, which makes clear that smell is more a.s.sociative, or metaphoric, than taste: We say something smells "like" something else, while we say that something simply is sweet or bitter or whatever-no simile required.

Each of the five tastes has been selected by evolution for its survival value. Either it guides us toward nutrients we need to survive, or it steers us away from ingesting things that might endanger us. For example, the taste of sweetness steers us toward particularly dense sources of energy in our environment, which is what sugars are. Salt is an essential nutrient we have been wired to like. Bitter happens to be how many plant toxins taste, which probably explains why babies instinctively frown when it is introduced on their tongues. (And why pregnant women in particular are often repelled by bitter foods.) Sour elicits an instinctive negative reaction, too, perhaps because when food rots it generally becomes more acidic, and rotten food is generally a risky thing to eat (stinky tofu notwithstanding). But even though they are innate, these last two responses can be "overridden": many of us learn to like sour or bitter foods.

So what about the taste of umami, or savoriness? Like salt and sugar, it evokes a universally positive response and, also like them, it signals the presence of an essential nutrient, in this case protein. Curiously, umami receptors have been found in the stomach as well as on the tongue. Their purpose, presumably, is to prepare the body to digest meat, alerting it to produce the necessary enzymes, hormones, and digestive acids. The most important chemical known to stimulate the umami receptors is the amino acid glutamate and the nucleotides inosine and guanosine, all of which are by-products of the breakdown of protein.

Which is of course precisely what is going on in a long-simmered stock: The long protein chains in the meat are breaking down into their various amino acid building blocks, glutamate chief among them. In fact, chicken stock is loaded with glutamate, which has been contributed not only by the protein-rich meat but by the slow cooking of the aromatic vegetables as well. Also present in meat stocks is inosinate, which when combined with glutumate creates a perception of umami much greater than the sum of its chemical parts.

But though umami can make a food taste "meaty," meat is only one of the many sources of glutamate. (That's why "savoriness" is probably a better translation for umami than "meatiness" or "brothiness.") Ripe tomatoes, dried mushrooms, Parmesan cheese, cured anchovies, and a great many fermented foods (including soy sauce and miso paste) contain high levels of glutamate, and can be added to a dish to boost its quotient of umami. This property of ripe tomatoes surely explains why so many of the braises I made with Samin called for a "tomato product"-canned tomatoes, or tomato paste-in addition to stock or wine. Occasionally we threw in a Parmesan rind, too, or some dried porcini or a squirt of anchovy paste. (And the reason we sometimes, a la Julia Child, browned our meat in bacon fat? Because bacon is a veritable umami bomb, containing all of the umami compounds that have thus far been identified.) I didn't know it at the time, and nor did Samin, but all these additions were ways to augment the umami in our dish, and the reason there was always more than one of them-tomato plus Parmesan, or stock plus dried mushrooms-was no doubt to exploit the synergistic properties of this particular taste. Umami, I realized, is the quasi-secret heart and soul of almost every braise, stew, and soup.

I say "secret" because umami works in somewhat mysterious ways, at least compared with sweet, salt, bitter, and sour. Encountered in the purified form of monosodium glutamate (MSG), glutamate doesn't taste particularly good, or for that matter much like anything at all. To work its magic, umami needs to be in the company of other ingredients. A bit like salt, glutamate seems to italicize the taste of foods, but, unlike salt, it doesn't have an instantly recognizable taste of its own.

The other mystery about umami is how it alters the texture as well as the taste of many foods-or, more accurately, our perception of its texture. Add umami to a soup and eaters will report it is not just "heartier" but actually thicker, too; the umami taste appears to have synesthetic properties. It makes a liquid seem less like water and more like food. It's possible that the umami chemicals activate not only the sense of taste in our mouths, but also trip the sense of touch as well, creating an illusion of "body."

What I learned about the properties of umami made me want to run some experiments with dashi, the cla.s.sic j.a.panese stock. If pot dishes owe so much of their power to umami, then making them with dashi-a cooking water designed, albeit unwittingly, to contain as much umami and as little of anything else as possible-seemed worth trying. It sounded to me like the Urcooking liquid. So, naturally, I wanted to make some.

At least until you understand the science of umami, dashi seems like a thoroughly improbable concept: a stock made from dried seaweed, shavings from a cured fish, and, optionally, a dried mushroom or two. But it so happens that each of those items contains a different one of the three princ.i.p.al umami chemicals. Put all three together in water and you get synergies that vastly amplify the umami effect. Dashi, which has been made in j.a.pan for more than a thousand years, is a cla.s.sic example of the wisdom of cuisines: how, strictly by trial and error, a traditional culture can perfect a chemistry in food that is not fully appreciated until long after the fact.

With my dashi experiments, I was venturing well outside Samin's culinary orbit. She doesn't have much experience of Eastern foodways. But she was able to direct me to someone who did: a young j.a.panese American cook by the (unsurpa.s.sed) name of Sylvan Mishima Brackett. When I e-mailed him to say I was interested in learning how to make dashi, Sylvan invited me over to the tiny, converted garage behind his house where he cooks, using little more than a hot plate.

What Sylvan did have, and what is difficult to find in this country, is a block of katsuobushi, or cured bonito, that he had brought back with him from a recent trip to j.a.pan. Katsuobushi looks like a toy submarine carved out of a block of hardwood, perhaps walnut. It is as hard and fine-grained as walnut, too, making it impossible to cut with any tool less sharp than a woodworker's plane. Which is in fact what is traditionally used to coax shavings from Katsuobushi.

Sylvan had been to a katsuobushi factory in j.a.pan, and he described the absurdly laborious process by which it is made. After the bonito is filleted into quarters, the fillets are simmered in water for two hours and then put on racks in a room in which an oak fire is burned for part of each day for a minimum of ten days. After that, the dried blocks of fish are sc.r.a.ped, taken out in the sun, and then inoculated with a fungus called koji (Aspergillus oryzae), before spending ten more days in a "molding room." That process-sc.r.a.ping, sun-drying, inoculating-is repeated three times, before the block-now completely desiccated and as hard as rock-is ready to be used. Here was an extreme instance of a pot-dish ingredient that was itself a complicated dish with a long recipe that called for an ingredient that itself had an unbelievably complicated recipe.

Sylvan used a whetstone to sharpen the blade on his plane and put me to work shaving katsuobushi. The block was actually considerably harder than wood, and it took a strenuous effort to acc.u.mulate even a small pile of shavings. The grain that the plane raised was a beautiful shade of salmony pink; how is it, I wondered as I worked, that the flesh of a fish and a tree could have so similar a structure? Meanwhile, Sylvan cranked up his hot plate and put a pot of water on to boil, to which he added a foot-long section of kombu. Kombu is air-dried kelp, one of the richest sources of glutamate in nature. Out of the package, it wears a cloak of white salt that is basically monosodium glutamate. Sylvan said the very best-quality kombu comes from (wouldn't you know it?) a specific beach on the northern coast of Hokkaido. He also mentioned that soft water was best for extracting the maximum flavor from the ingredients, and that in fact the word "dashi" means "extraction."

But if the backstory of the ingredients in dashi is complicated, the recipe for making it is fairly straightforward and, for a stock, quick and easy-less than ten minutes from start to finish. Sylvan dropped a sheet of kombu into a pot of cold water, heated it to a point just shy of the boil, and then removed the now green and floppy length of kombu with a pair of tongs. If the kombu reaches a full boil, he explained, the dashi will turn out bitter. At this point, the liquid gave off only the faintest scent of brininess. Unlike the kombu, the bonito flakes need to be boiled to release their flavor, so, as soon as the pot began to roil, Sylvan dropped in a big handful. The pinkish shavings danced crazily on the surface and then, as they rehydrated, began to sink to the bottom. They had only been in the water five or six minutes when Sylvan poured the stock through cheesecloth and discarded the residual flakes. The resulting liquid resembled very weak tea, an almost perfectly transparent pale gold. As the liquid cools, you have the option of adding a dried shiitake mushroom. But that's all there is to it.

I bent over to smell the finished dashi. It reminded me of a tide pool: briny, with the faintest suggestion of decay-the beach at low tide. I dipped a finger into the cooling liquid. It had very little taste to speak of; some saltiness, but not too much-sort of like a freshwater version of the ocean. Brackish. Compared with a real stock, it was pallid stuff; you would never think of sipping it as a soup. But the pale liquid contained large amounts of the three main umami chemicals-glutamate from the kombu, inosinate from the bonito, and guanylate from the mushroom-each of them extracted by the water.

Sylvan gave me some bonito flakes and kombu to take home, and over the course of the next several days I made my own dashi and experimented with it. The first thing I tried was a dipping sauce. To a small bowl of dashi I added a tablespoon each of soy sauce, mirin, and rice wine vinegar, as well as a small handful of chopped scallions and ginger. It was remarkable stuff: Anything dipped in it-a chicken breast, some soba noodles, a piece of pork-received an uncanny boost in flavor, somehow tasted more platonically itself. (And more platonically j.a.panese.) Next, I tried the dashi as a braising liquid, for beef short ribs and then for a pork loin, combining it, again, with some soy sauce, mirin, vinegar, and sake, as well as some miso paste. The result in both cases was a rich and satisfying dish, somewhat lighter than the braises that Samin and I had made, though no less intensely flavored. I have not yet tried dashi in a non-Asian dish; that might be crazy, I don't know, and Samin would probably flip if I proposed it. But dashi itself is not a flavor principle, exactly-it's more like an italicizer of flavors-so it might well work with another cuisine. Nothing about dashi, when tasted by itself, prepares you for what it does in concert with other flavors. I'm starting to think of it as magic water: hydrogen and oxygen and amino acids and something no one knows.

One curious fact I stumbled on in my umami research was that human breast milk is rich in this particular taste, and contains relatively large amounts of glutamate-as it happens, nearly the same amount of glutamate as an equivalent amount of dashi. It stands to reason that everything in milk is there for an evolutionary reason; since every chemical compound in it comes at a metabolic cost to the mother, natural selection would quickly dispense with any const.i.tuent of milk that didn't do the infant some good. So what good does all that glutamate do?

There are a couple of possible explanations. Bruce German, a food chemist at the University of California, Davis, who a.n.a.lyzes human milk in order to better understand our nutritional needs, believes that the glutamate supplies an important nutrient to the growing infant. Besides being a flavor, this particular amino acid is a cellular fuel and molecular building block of special value to the stomach and intestines of the growing infant. In the same way that glucose is an ideal food for the brain, glutamate is a perfect food for the gut, which might explain why we're born with taste buds in the stomach that can detect it.

All that glutamate in breast milk might be doing something else as well: conditioning the baby to like the taste of umami, that being (along with sweetness) one of the first and most abundant tastes it encounters in mother's milk. This preference is highly adaptive for h.o.m.o sapiens, since we require a diet rich in the proteins that umami helps us to recognize and seek out.

But could it be that, for us, the taste of foods rich in umami also sounds deep Proustian echoes, bearing us back to memories, however faint, of our very first food? Is it merely a coincidence that so many of the things we think of as "comfort foods"-everything from ice cream to chicken soup-traffic in tastes of either sweetness or umami, the two big tastes first encountered on the breast?

This bit of speculation was very much on my mind during a recent Sunday afternoon with Samin, when we set out to make an ancient Roman dish called maiale al latte-pork braised in milk. I was skeptical about this one, and not only because it was so radically unkosher. The fact that I eat pork should by now be well established, but there does seem something slightly perverse about cooking it in milk, and I wondered if there might not be some good practical reason behind the Old Testament's taboo on mixing milk and meat. But apparently not: The rabbinical commentators say that particular taboo falls under the heading of "hukkim," which are laws for which there is no obvious explanation.

My guess? The kosher laws are all about drawing and defending crisp lines of demarcation between various realms, and what line is sharper than the one between life and death? You don't mix a symbol of death, such as animal flesh, with a symbol of life as powerful as mother's milk. Also, cooking meat in milk mingles the male realm of the hunt with the female realm of nurturing-a taboo in many cultures. As the anthropologist Mary Douglas has written, a rule against mixing meat and milk "honors the procreative function."

Well, not today. "This is one of my all-time favorite dishes," Samin said, when I expressed my doubts about it. "I know, it sounds really weird, and I have to prepare you: it looks sort of gross when you're cooking it. But I promise, it will be the most delicious, succulent comfort food you've ever tasted in your life!"

As a cooking liquid, milk presents special challenges. Of all the pot dishes we cooked, this one had to be watched the most closely, lest the sugars in the milk begin to burn on the bottom of the pot. Yet, at the same time, maiale al latte was also the very simplest recipe we'd made. In fact, it can be written out in a sentence: Brown chunks of pork in b.u.t.ter; add some milk, a few cloves of garlic, a handful of sage leaves, and the juice (and peel) of a lemon; simmer for several hours. That's it. No soffritto? I asked Samin. No chopped onions?

"Nope. Weird, I know. But I think this dish must be even older than soffritto is. It might even go back to Etruscan times."

The biggest challenge is keeping the milk at a gentle simmer just below the boil-the braising liquid should merely "smile," as the French say, rather than bubble. So we peeked in at regular intervals, taking advantage of the established fact that a watched pot will never boil. (Probably because in order to watch you lift the lid, which drops the temperature.) After a while the milk began to yellow slightly and form curds-and to look very much like baby vomit. Which is not at all unlike what it was: warm milk that has curdled after having been exposed to an acid. The age-old conceit of the cauldron as an external organ of digestion had never seemed so apt, but that of course was precisely what was going on here, the proteins in the milk being broken down and rearranged by the acids.

"I know, it's sort of disgusting," Samin allowed. "But this is exactly what we want. You'll see, those curds are going to be super-delicious."

And so they were, eventually. After several hours the cooking liquid turned a gorgeous shade of ochre, and the golden curds no longer looked like mistakes. The lemony milk had gone to work on the proteins in the meat, breaking it down until it was so tender it fell apart at the prodding of a fork. The meat was as succulent and tasty as Samin had promised, but it was the sauce that was most incredible, with its creamy layers of savory and sweet. Actually, all five tastes were represented in that silky liquid: besides the savory-saltiness from the meat and the sweetness from the milk, the sauce bore traces of sourness and bitterness contributed by the lemon peel and sage leaves, all of it harmoniously dispersed in the milk. To concoct so much flavor from such a small number of ordinary ingredients-pork, garlic, lemon, sage, and milk-seemed like a miracle of transubstantiation. "The transformation which occurs in the cauldron is quintessential and wondrous, subtle and delicate," wrote a Chinese chef named I Yin in 239 B.C., no doubt moved by a similar eating experience. "The mouth cannot express it in words."

Gaston Bachelard, the somewhat obtuse French philosopher of the elements, wrote a book called Water and Dreams, in which he attempts to "psychoa.n.a.lyze" water and other liquids in much the same way he attempted a psychoa.n.a.lysis of fire. "For the imagination, everything that flows is water," Bachelard writes in a chapter called "Maternal Water and Feminine Water." Water is always feminine in the imagination, he contends, just as its opposite, fire, is always masculine. But he then goes a step further, suggesting that, to the imagination, "all water is a kind of milk," though a moment later he confines this claim to the kinds of water we like: "More precisely, every joyful drink is mother's milk," and, a bit farther on, "water is a milk as soon as it is extolled fervently."

As an example Bachelard offers an image of the "nourishing waters" of the sea, in which the resident fish effortlessly feed themselves from the particles of fat and other nutrients dispersed in the liquid medium, floating along without care as if in amniotic fluids. "For the material imagination, water, like milk, is a complete food."

Bachelard has little else to say about food in Water and Dreams, and nothing at all about stews and soups, but my guess is that they would all qualify in his imagination as "milks"-as a medium much like the nourishing sea, in which the fish, like babies on the breast, never want for anything they need or desire. The nourishing liquid that forms in a pot dish starts out as thin and transparent as water, then clouds and colors as it absorbs and disperses substance and flavor, ending up eventually as a more or less complete and milklike food. In the imagination, at least, this kind of cooking qualifies as a transubstantiation of matter, in this instance not of water into wine but of something no less miraculous: water into milk.

"Stone Soup" is the ancient parable of this everyday miracle, of turning water into food. In the story, which has been told for centuries in many different cultures (sometimes as "Nail Soup" or "b.u.t.ton Soup" or "Ax Soup"), some poor, hungry strangers come to town with nothing but an empty pot. The villagers refuse them food, so the strangers fill their pot with water, drop a stone in it, and put it on to boil in the town square. This arouses the curiosity of the villagers, who ask the strangers what it is they're making.

"Stone soup," the strangers explain. "It's delicious, as you'll soon see, but it would taste even better if you could spare a little garnish to help flavor it." So one villager gives them a sprig of parsley. Then another remembers she has some potato peelings at home, which she fetches and drops into the pot. Someone else throws in an onion and a carrot, and then another villager offers a bone. As the kettle boils, one villager after another comes by to throw in a sc.r.a.p of this, a bit of that, until the soup had thickened into something nourishing and wonderful that everyone-villagers and strangers-sits down to enjoy together at a great feast.

"You have given us the greatest gift," one of the village elders declares, "the secret of how to make soup from stones."

VI.

Step Six: Simmer, Below the Boil, for a Long Time