Proust Was a Neuroscientist - Jonah Lehrer (2007)

Chapter 3. Auguste Escoffier

The Essence of Taste

So it happens that when I write of hunger, I am really writing about love and the hunger for it, and warmth and the love of it and the hunger for it ... and then the warmth and richness and fine reality of hunger satisfied ... and it is all one.

—M.F.K. Fisher, The Gastronomical Me

AUGUSTE ESCOFFIER INVENTED veal stock. Others had boiled bones before, but no one had codified the recipe. Before Escoffier, the best way to make veal stock was cloaked in mystery; cooking was like alchemy, a semimystical endeavor. But Escoffier came of age during the late stages of positivism, a time when knowledge—some of it true, some of it false—was disseminated at a dizzying rate. Encyclopedias were the books of the day. Escoffier took this scientific ethos to heart; he wanted to do for fancy food what Lavoisier had done for chemistry, and replace the old superstitions of the kitchen with a new science of cooking.

At the heart of Escoffier's insight (and the source of more than a few heart failures) was his use of stock. He put it in everything. He reduced it to gelatinous jelly, made it the base of pureed soups, and enriched it with butter and booze for sauces. While French women had created homemade stock for centuries—pot-au-feu (beef boiled in water) was practically the national dish—Escoffier gave their protein broth a professional flair. In the first chapter of his Guide Culinaire (1903), Escoffier lectured cooks on the importance of extracting flavor from bones: "Indeed, stock is everything in cooking. Without it, nothing can be done. If one's stock is good, what remains of the work is easy; if, on the other hand, it is bad or merely mediocre, it is quite hopeless to expect anything approaching a satisfactory meal." What every other chef was throwing away—the scraps of tendon and oxtail, the tops of celery, the ends of onion, and the irregular corners of carrot—Escoffier was simmering into sublimity.

Although Escoffier introduced his Guide Culinaire with the lofty claim that his recipes were "based upon the modern science of gastronomy," in reality he ignored modern science. At the time, scientists were trying to create a prissy nouvelle cuisine based on their odd, and totally incorrect, notions of what was healthy. Pig blood was good for you. So was tripe. Broccoli, on the other hand, caused indigestion. The same with peaches and garlic. Escoffier ignored this bad science (he invented peach Melba), and sauteed away to his heart's malcontent, trusting the pleasures of his tongue over the abstractions of theory. He believed that the greatest threat to public health was the modern transformation of dining from a "pleasurable occasion into an unnecessary chore."

The form of Escoffier's encyclopedic cookbook reflects his romantic bent. Although he was fond of calling sauciers (the cooks responsible for making the sauces) "enlightened chemists," his actual recipes rarely specify quantities of butter or flour or truffles or salt. Instead, they are descriptions of his cooking process: melt the fat, add the meat, listen "for the sound of crackling," pour in the stock, and reduce. It sounds so easy: all you have to do is obey the whims of your senses. This isn't a science experiment, Escoffier seems to be saying, this is hedonism. Let pleasure be your guide.

Escoffier's emphasis on the tongue was the source of his culinary revolution. In his kitchen, a proper cook was a man of exquisite sensitivity, "carefully studying the trifling details of each separate flavor before he sends his masterpiece of culinary art before his patrons." Escoffier's cookbook warns again and again that the experience of the dish—what it actually tastes like—is the only thing that matters: "Experience alone can guide the cook." A chef must be that artist on whom no taste is lost.

But Escoffier knew that he couldn't just serve up some grilled meat and call it a day. His hedonism had to taste haute. After all, he was a chef in the hotels of César Ritz, and his customers expected their food to be worthy of the gilded surroundings and astonishing expense. This is where Escoffier turned to his precious collection of stocks. He used his stocks to ennoble the ordinary sauté, to give the dish a depth and density of flavor. After the meat was cooked in the hot pan (Escoffier preferred a heavy, flat-bottomed poele), the meat was taken out to rest, and the dirty pan, full of delicious grease and meat scraps, was deglazed.

Deglazing was the secret of Escoffier's success. The process itself is extremely simple: a piece of meat is cooked at a very high temperature—to produce a nice seared Maillard crust, a cross-linking and caramelizing of amino acids—and then a liquid, such as a rich veal stock, is added.* As the liquid evaporates, it loosens the fronde, the burned bits of protein stuck to the bottom of the pan (deglazing also makes life easier for the dishwasher). The dissolved fronde is what gives Escoffier's sauces their divine depth; it's what makes boeuf bourguignon, bourguignon. A little butter is added for varnish, and, voilà! the sauce is complet.

The Secret of Deliciousness

Escoffier's basic technique is still indispensable. Few other cultural forms have survived the twentieth century so intact. Just about every fancy restaurant still serves up variations of his dishes, recycling their bones and vegetable tops into meat stocks. From espagnole sauce to sole Véronique, we eat the way he told us to eat. And since what Brillat-Savarin said is probably true—"The discovery of a new dish does more for the happiness of the human race than the discovery of a new star"—it is hard to overestimate Escoffier's importance. Clearly, there is something about his culinary method—about stocks and deglazing and those last-minute swirls of butter—that makes some primal part of us very, very happy.

The place to begin looking for Escoffier's ingenuity is in his cookbooks. The first recipe he gives us is for brown stock (estouffade), which he says is "the humble foundation for all that follows." Escoffier begins with the browning of beef and veal bones in the oven. Then, says Escoffier, fry a carrot and an onion in a stockpot. Add cold water, your baked bones, a little pork rind, and a bouquet garni of parsley, thyme, bay leaf, and a clove of garlic. Simmer gently for twelve hours, making sure to keep the water at a constant level. Once the bones have given up their secrets, sauté some meat scraps in hot fat in a saucepan. Deglaze with your bone water and reduce. Repeat. Do it yet again. Then slowly add the remainder of your stock. Carefully skim off the fat (a stock should be virtually fat-free) and simmer for a few more hours. Strain through a fine chinois. After a full day of stock-making, you are now ready to start cooking.

In Escoffier's labor-intensive recipe, there seems to be little to interest the tongue. After all, everybody knows that the tongue can taste only four flavors: sweet, salty, bitter, and sour. Escoffier's recipe for stock seems to deliberately avoid adding any of these tastes. It contains very little sugar, salt, or acid, and unless one burns the bones (not recommended), there is no bitterness. Why, then, is stock so essential? Why is it the "mother" of Escoffier's cuisine? What do we sense when we eat a profound beef daube, its deglazed bits simmered in stock until the sinewy meat is fit for a spoon? Or, for that matter, when we slurp a bowl of chicken soup, which is just another name for chicken stock? What is it about denatured protein (denaturing is what happens to meat and bones when you cook them Escoffier's way) that we find so inexplicably appealing?

The answer is umami, the Japanese word for "delicious." Umami is what you taste when you eat everything from steak to soy sauce. It's what makes stock more than dirty water and deglazing the essential process of French cooking. To be precise, umami is actually the taste of L-glutamate (C5H9NO4), the dominant amino acid in the composition of life. L-glutamate is released from life-forms by proteolysis (a shy scientific word for death, rot, and the cooking process). While scientists were still theorizing about the health benefits of tripe, Escoffier was busy learning how we taste food. His genius was getting as much L-glutamate on the plate as possible. The emulsified butter didn't hurt either.

The story of umami begins at about the same time Escoffier invented tournedos Rossini, a filet mignon served with foie gras and sauced with a reduced veal stock and a scattering of black truffles. The year was 1907, and Japanese chemist Kikunae Ikeda asked himself a simple question: What does dashi taste like? Dashi is a classic Japanese broth made from kombu, a dried form of kelp. Since at least A.D. 797, dashi has been used in Japanese cooking the same way Escoffier used stock, as a universal solvent, a base for every dish. But to Ikeda, the dashi his wife cooked for him every night didn't taste like any of the four classic tastes or even like some unique combination of them. It was simply delicious. Or, as the Japanese would say, it was umami.

And so Ikeda began his quixotic quest for this unknown taste. He distilled fields of seaweed, searching for the essence that might trigger the same mysterious sensation as a steaming bowl of seaweed broth. He also explored other cuisines. "There is a taste," Ikeda declared, "which is common to asparagus, tomatoes, cheese and meat but which is not one of the four well-known tastes." Finally, after patient years of lonely chemistry, during which he tried to distill the secret ingredient that dashi and veal stock had in common, Ikeda found his secret molecule. It was glutamic acid, the precursor of L-glutamate. He announced his discovery in the Journal of the Chemical Society of Tokyo.

Glutamic acid is itself tasteless. Only when the protein is broken down by cooking, fermentation, or a little ripening in the sun does the molecule degenerate into L-glutamate, an amino acid that the tongue can taste. "This study has discovered two facts," Ikeda wrote in his conclusion, "one is that the broth of seaweed contains glutamate and the other that glutamate causes the taste sensation 'umami.'"

But Ikeda still had a problem. Glutamate is an unstable molecule, eager to meld itself to a range of other chemicals, most of which are decidedly not delicious. Ikeda knew that he had to bind glutamate to a stable molecule that the tongue did enjoy. His ingenious solution? Salt. After a few years of patient experimentation, Ikeda was able to distill a metallic salt from brown kelp. The chemical acronym of this odorless white powder was MSG, or monosodium glutamate. It was salty, but not like salt. It also wasn't sweet, sour, or bitter. But it sure was delicious.

Ikeda's research, although a seminal finding in the physiology of taste, was completely ignored. Science thought it had the tongue solved. Ever since Democritus hypothesized in the fourth century B.C. that the sensation of taste was an effect of the shape of food particles, the tongue has been seen as a simple muscle. Sweet things, according to Democritus, were "round and large in their atoms," while "the astringently sour is that which is large in its atoms but rough, angular and not spherical." Saltiness was caused by isosceles atoms, while bitterness was "spherical, smooth, scalene and small." Plato believed Democritus, and wrote in Timaeus that differences in taste were caused by atoms on the tongue entering the small veins that traveled to the heart. Aristotle, in turn, believed Plato. In De Anima, the four primary tastes Aristotle described were the already classic sweet, sour, salty, and bitter.

Over the ensuing millennia, this ancient theory remained largely unquestioned. The tongue was seen as a mechanical organ in which the qualities of foods were impressed upon its papillaed surface. The discovery of taste buds in the nineteenth century gave new credence to this theory. Under a microscope, these cells looked like little keyholes into which our chewed food might fit, thus triggering a taste sensation. By the start of the twentieth century, scientists were beginning to map the tongue, assigning each of the four flavors to a specific area. The tip of the tongue loved sweet things, while the sides preferred sour. The back of the tongue was sensitive to bitter flavors, and saltiness was sensed everywhere. The sensation of taste was that simple.

Unfortunately for Ikeda, there seemed to be no space left on the tongue for his delicious flavor. Umami, these Western scientists said, was an idle theory unique to Japanese food, a silly idea concerned with something called deliciousness, whatever that was. And so while cooks the world over continued to base entire cuisines on dashi, Parmesan cheese, tomato sauce, meat stock, and soy sauce (all potent sources of L-glutamate), science persisted in its naÏve and unscientific belief in four, and only four, tastes.

Despite the willful ignorance of science, Ikeda's idea gained a certain cult following. His salty white substance, MSG, a powder that science said couldn't work because we had no means to taste it, nevertheless became an overused staple in everything from cheap Chinese food to bouillon cubes, which used glutamate to simulate the taste of real stock. MSG was even sold in America under the labels Super Seasoning and Accent.* As food products became ever more processed and industrial, adding a dash of MSG became an easy way to create the illusion of flavor. A dish cooked in the microwave tasted as if it had been simmered for hours on the stovetop. Besides, who had time to make meat stock from scratch?

With time, other pioneers began investigating their local cuisines and found their own densities of L-glutamate. Everything from aged cheese to ketchup was rich in this magic little amino acid. Umami even seemed to explain some of the more perplexing idiosyncrasies of the cooking world: why do so many cultures, beginning with ancient Rome, have a fish sauce? (Salted, slightly rotting anchovies are like glutamate speedballs. They are pure umami.) Why do we dip sushi in soy sauce? (The raw fish, being raw, is low in umami, since its glutamate is not yet unraveled. A touch of soy sauce gives the tongue the burst of umami that we crave.) Umami even explains (although it doesn't excuse) Marmite, the British spread made of yeast extract,* which is just another name for L-glutamate. (Marmite has more than 1750 mg of glutamate per 100 g, giving it a higher concentration of glutamate than any other manufactured product.)

Of course, umami is also the reason that meat—which is nothing but amino acid—tastes so darn good. If cooked properly, the glutamate in meat is converted into its free form and can then be tasted. This also applies to cured meats and cheeses. As a leg of prosciutto ages, the amino acid that increases the most is glutamate. Parmesan, meanwhile, is one of the most concentrated forms of glutamate, weighing in at more than 1200 mg per 100 g. (Only Roquefort cheese has more.) When we add an aged cheese to a pasta, the umami in the cheese somehow exaggerates the umami present elsewhere in the dish. (This is why tomato sauce and Parmesan are such a perfect pair. The cheese makes the tomatoes more tomatolike.) A little umami goes a long way.

And of course, umami also explains Escoffier's genius. The burned bits of meat in the bottom of a pan are unraveled protein, rich in L-glutamate. Dissolved in the stock, which is little more than umami water, these browned scraps fill your mouth with a deep sense of deliciousness, the profound taste of life in a state of decay.

The culture of the kitchen articulated a biological truth of the tongue long before science did because it was forced to feed us. For the ambitious Escoffier, the tongue was a practical problem, and understanding how it worked was a necessary part of creating delicious dishes. Each dinner menu was a new experiment, a way of empirically verifying his culinary instincts. In his cookbook, he wrote down what every home cook already knew. Protein tastes good, especially when it's been broken apart. Aged cheese isn't just rotten milk. Bones contain flavor. But despite the abundance of experiential evidence, experimental science continued to deny umami's reality. The deliciousness of a stock, said these haughty lab coats, was all in our imagination. The tongue couldn't taste it.

What Ikeda needed before science would believe him was anatomical evidence that we could actually taste glutamate. Anecdotal data from cookbooks, as well as all those people who added fish sauce to their pho, Parmesan to their pasta, and soy sauce to their sushi, wasn't enough.

Finally, more than ninety years after Ikeda first distilled MSG from seaweed, his theory was unequivocally confirmed. Molecular biologists discovered two distinct receptors on the tongue that sense only glutamate and L-amino acids. In honor of Ikeda, they were named the umami receptors. The first receptor was discovered in 2000, when a team of scientists noticed that the tongue contains a modified form of a glutamate receptor already identified in neurons in the brain (glutamate is also a neurotransmitter). The second sighting occurred in 2002, when another umami receptor was identified, this one a derivative of our sweet taste receptors.*

These two separate discoveries of umami receptors on the tongue demonstrated once and for all that umami is not a figment of a hedonist's imagination. We actually have a sense that responds only to veal stock, steak, and dashi. Furthermore, as Ikeda insisted, the tongue uses the taste of umami as its definition of deliciousness. Unlike the tastes of sweet, sour, bitter, and salty, which are sensed relative to one another (this is why a touch of salt is always added to chocolate, and why melon is gussied up with ham), umami is sensed all by itself. It is that important.

This, of course, is perfectly logical. Why wouldn't we have a specific taste for protein? We love the flavor of denatured protein because, being protein and water ourselves, we need it. Our human body produces more than forty grams of glutamate a day, so we constantly crave an amino acid refill. (Species that are naturally vegetarian find the taste of umami repellent. Unfortunately for vegans, humans are omnivores.) In fact, we are trained from birth to savor umami: breast milk has ten times more glutamate than cow milk. The tongue loves what the body needs.

The Smell of an Idea

Veal stock was not always the glutamate-rich secret of French food. In fact, haute cuisine was not always delicious, or even edible. Before Escoffier began cooking in the new restaurants of the bourgeoisie (unlike his predecessors, he was never a private chef for an aristocrat), fancy cooking was synonymous with ostentation. As long as dinner looked decadent, its actual taste was pretty irrelevant. Appearance was everything. Marie-Antoine Carême (1783-1833), the world's first celebrity chef—he cooked for Talleyrand and Czar Alexander I, and he baked Napoleon's wedding cake—epitomized this culinary style. Although Carême is often credited with inventing French cuisine, his food was normally served cold and arranged in epic buffets comprising dozens, sometimes hundreds, of rococo dishes. In Carême's Paris, fancy food was a form of sculpture, and Carême was justifiably famous for his pièces montées, which were detailed carvings made of marzipan, pork fat, or spun sugar. Although these sculptures were pretty, they were also inedible. Carême didn't care. "A well-displayed meal," Carême once said, "is improved one hundred percent in my eyes." Such insipid lavishness typified nineteenth-century service à la française.

Escoffier considered all this pomp and circumstance ridiculous. Food was meant to be eaten. He favored service à la russe—the Russian style—a system in which the meal was broken down into numbered courses. Unlike Carême's ornate buffets, serviceà la russe featured a single dish per course, which was delivered fresh out of the kitchen. The meal was staggered, and it unfolded in a leisurely culinary narrative: soup was followed by fish, which was followed by meat. And although the chef wrote the menu, the client dictated the tempo and content of the meal. Dessert was the guaranteed happy ending.

This revolution in restaurant service required a parallel revolution in the kitchen. No longer could cooks afford to spend days sculpting marzipan, or molding aspic, or concocting one of Carême's toxically rich stews.*Everything on the menu now had to be à la minute, and cooked to order. Flavor had to be manufactured fast. This new level of speed led Escoffier to make his cooking mantra "Faites simple." Every dish, he said, must consist of its necessary ingredients only, and those ingredients must be perfect. A veal stock must contain the very quintessence of veal. Asparagus soup must taste like asparagus, only more so.

There was one added bonus to this new culinary method founded upon simplicity and velocity: food was served hot. While Carême feared heat (his lard sculptures tended to melt), Escoffier conditioned his diners to expect a steaming bowl of soup. They wanted their fillets sizzling and the sauce fresh from the deglazed frying pan. In fact, Escoffier's recipes required this efficiency: when meals were served lukewarm, the flavors became disconcertingly one-dimensional. "The customer," Escoffier warned in his cookbook, "finds that the dish is flat and insipid unless it is served absolutely boiling hot."

What Escoffier inadvertently discovered when he started serving food fresh off the stovetop was the importance of our sense of smell. When food is hot, its molecules are volatile and evaporate into the air. A slowly simmering stock or a clove of garlic sautéed in olive oil can fill an entire kitchen with its alluring odor. Cold food, however, is earth-bound. It relies almost entirely upon the taste buds; the nose is not involved.

But as anyone with a stuffy nose knows, the pleasure of food largely depends on its aroma. In fact, neuroscientists estimate that up to 90 percent of what we perceive as taste is actually smell. The scent of something not only prepares us for eating it (our salivary glands become active), but gives the food a complexity that our five different taste sensations alone can only hint at. If the tongue is the frame for the food—providing us with crucial information about texture, mouth-feel, and the rudiments of taste—the sensations of the nose are what make the food worth framing in the first place.

Escoffier was the first chef to take full advantage of the acutely sensitive nose. Although his food presciently catered to the needs of the tongue (especially its lust for umami), Escoffier aspired to a level of artistry that the tongue couldn't comprehend. As a result, Escoffier's capacious recipes depend entirely upon the flourishes of flavor that we inhale. In fact, all the culinary nuances that so obsessed him—the hint of tarragon in a lobster velouté, the whisper of vanilla in a crème anglaise, the leaf of chervil floating in a carrot soup—are precisely what the unsubtle tongue can't detect. The taste of most flavors is smell.

When we eat, air circulates through the mouth and rises up into the nasal passages, where the gaseous particles of hot food bind to 10 million odor receptors arrayed in an area the size of a thumbprint. When a smell particle binds to the receptor (how they bind no one knows), a surge of ionic energy is created; it travels down the wiry axon, courses through the skull, and connects directly to the brain.

Of course, the receptors in the nose are just the beginning of our sense of smell. The world is an aromatic place. It seems preposterous to expect a nose to have a different receptor for each of the 10,000 to 100,000 discrete odors we can detect. Indeed, our sense of smell—like all of our senses—indulges in a bit of a cellular shortcut, exchanging accuracy for efficiency. And while this might seem stingy, evolution has actually been extremely generous to our sense of smell: odor receptors take up more than 3 percent of the human genome.

Why does the sense of smell require so much DNA? Because our nasal passages are equipped with more than 350 different receptor types, each of which expresses a single receptor gene. These receptor neurons can be activated by many different odorants, including ones that belong to completely distinct chemical families. As a result, before the brain can generate the sensation of any specific scent, the receptors have to work together. They have to transform their scattered bits of information into a coherent representation.

To figure out how this happens, Nobel laureate Richard Axel's lab engineered a fruit fly with a glowing brain, each of its neurons like a little neon light. This was done through the careful insertion of a fluorescent protein in all of the insect's olfactory nerves. But the glow wasn't constant. Axel engineered the fly so that the fluorescent protein turned itself on only when calcium was present in high concentrations inside the cell (active neurons have more calcium). Using some fancy microscopy, Axel's lab group was able to watch—in real time—the patterns of activity within the fly brain whenever it experienced an odor. They could trace the ascent of the smell, how it began as a flicker in a receptor and within milliseconds inflated into a loom of excited cells within the tiny fly nervous system. Furthermore, when the fluorescent fly was exposed to different odors, different areas of its brain lit up. The scent of almonds activated a different electrical grid than the scent of a ripe banana. Axel had found the functional map of smell.

But this imaging of insects, for all of its technical splendor, leaves the real mystery of scent unanswered. Using his neon neurons, Axel can look at the fly's brain and, with shocking accuracy, discern what smell the fly is smelling. He performs this act of mind reading by looking at the fly brain from the outside. But how does the fly know what it's experiencing? Unless you believe in a little drosophila ghost inside the fly machine, reconstructing its deconstructed smell, this mystery seems impossible to explain. As Axel notes, "No matter how high we get in the fly brain when we map this sensory circuit, the question remains: who in the fly brain is looking down? Who reads the olfactory map? This is our profound and basic problem."

To illustrate the seriousness of this paradox, take an example from our own experience. Imagine you have just inhaled the smell of a deep, dark demi-glace, a veal stock that is slowly simmered until it becomes viscous with the gelatin released by its bones. Although Escoffier wanted his sauce to be the distilled essence of veal, he knew that such purity required a lengthy shopping list. You couldn't simply simmer some cow bones in water. The irony behind Escof-fier's cooking is that his pursuit of the unadorned dish—a sauce that tasted simply of itself—led him to add a cavalcade of other ingredients, none of which had anything at all to do with the essence of veal. As a result, the demi-glace our nose knows is actually composed of many different aromas. Neurons all over the brain light up, reflecting the hodgepodge of smells simultaneously activating our odor receptors. There is the carnal odor of roasted meat; the woodsy scent of the bouquet garni; and the soothing smell of the roux, the flour turning brown with the butter. Underneath those prominent notes, we get the sweetly vegetal waft of the mirepoix, the caramelized tang of tomato paste, and the nutty vapor of evaporated sherry. But how do all of these different ingredients become the specific smell of a demi-glace, which tastes like the essence of veal? From the perspective of the nervous system, the challenge is daunting. Within a few milliseconds of being served the demi-glace, the mind must bind together the activity of hundreds of distinct smell receptors into a coherent sensation. This is known as the binding problem.

But wait: it gets worse. The binding problem occurs when we experience a sensation that is actually represented as a network of separate neurons distributed across the brain. In the real world, however, reality doesn't trickle in one smell at a time. The brain is constantly confronted with a pandemonium of different odors. As a result, it not only has to bind together its various sensations, it has to decipher which neurons belong to which sensation. For example, that demi-glace was probably served as a sauce for a tender fillet of beef, with a side of buttery mashed potatoes. This Escoffier-inspired dish instantly fills the nose with a barrage of distinct scents, from the umber notes of reduced veal stock to the starchy smell of whipped russets. Faced with such a delicious meal, we can either inhale the odor of the dish as a whole—experiencing the overlapping smells as a sort of culinary symphony—or choose to smell each of the items separately. In other words, we can parse our own inputs and, if we so desire, choose to focus on just the smell of potatoes, or demi-glace, or a piece of beef served medium rare. Although this act of selective attention seems effortless, neuroscience has no clue how it happens. This is known as the parsing problem.

Parsing and binding are problems because they can't be explained from the bottom up. No matter how detailed our maps of the mind become, the maps still won't explain how a cacophony of cells is bound into the unified perception of a sauce. Or how, at any given moment, we are able to shuttle between our different sensations and separate the smell of the sauce from the smell of the steak. Neuroscience excels at dissecting the bottom of sensation. What our dinner demonstrates is that the mind needs a top.

A Sense of Subjectivity

To make matters even more complicated, what we experience is never limited to our actual sensations. Impressions are always incomplete and require a dash of subjectivity to render them whole. When we bind or parse our sensations, what we are really doing is making judgments about what we think we are sensing. This unconscious act of interpretation is largely driven by contextual cues. If you encounter a sensation in an unusual situation—such as the smell of demi-glace in a McDonald's—your brain secretly begins altering its sensory verdict. Ambiguous inputs are bound together into a different sensation. The fancy scent of veal stock becomes a Quarter Pounder.

Our sense of smell is particularly vulnerable to this sort of outside influence. Since many odors differ only in their molecular details—and we long ago traded away nasal acuity for better color vision—the brain is often forced to decipher smells based upon non-olfactory information. Parmesan cheese and vomit, for example, are both full of butyric acid, which has a pungent top note and a sweetish linger. As a result, blindfolded subjects in experiments will often confuse the two stimuli. In real life, however, such sensory mistakes are extremely rare. Common sense overrules our actual senses.

On an anatomical level, this is because the olfactory bulb is inundated with feedback from higher brain regions. This feedback continually modulates and refines the information garnered by smell receptors. A team of scientists at Oxford has shown that a simple word label can profoundly alter what we think our noses are telling us. When an experimental subject is given odorless air to smell but told he is smelling cheddar cheese, his olfactory areas light up in hungry anticipation. But when that same air arrives with a "body-odor" label, the subject unwittingly shuts down the smell areas of his brain. Although the sensation hasn't changed—it's still just purified air—the mind has completely revised its olfactory response. We unknowingly deceive ourselves.

Escoffier understood this psychological fact. His restaurants were all about the power of suggestion. He insisted that his dishes have fancy names and be served on gilded silver platters. His porcelain was from Limoges, his wineglasses from Austria, and his formidable collection of polished utensils from the estate sales of aristocrats. Escoffier didn't serve steak with gravy; he served filet de boeuf Richelieu (which was steak with a well-strained gravy). He made his waiters wear tuxedos, and he helped oversee the rococo decoration of his dining room. A perfect dish, after all, required a perfect mood. Although Escoffier spent eighteen hours a day behind a hot stove, crafting his collection of sauces, he realized that what we taste is ultimately an idea, and that our sensations are strongly influenced by their context. "Even horsemeat," Escoffier quipped, "can be delicious when one is in the right circumstances to appreciate it."

This is a suspicious-sounding concept. It smacks of solipsism, relativism, and all those other postmodern -isms. But it's our neurological reality. When we sense something, that sensation is immediately analyzed in terms of previous experiences. A demi-glace gets filed under sauce, meat, and ways to serve a filet mignon. As the brain figures out what to tell us about this particular demi-glace, those previous experiences help us decipher the information being received from the tongue and nose. Is this a good sauce? How does it compare to our memories of other sauces? Do we feel guilty for ordering veal? Was this dish worth the price? Was the waiter rude?

The answers to this cavalcade of unconscious questions determine what we actually experience. Before we've even reached for a second forkful, the demi-glace has been ranked and judged, our subjectivity emulsified into our sensation. Thus, what we think we are tasting is only partially about the morsel of matter in the mouth. Equally important is the sum of past experiences enclosed within the brain, for these memories are what frame the sensation.

The most persuasive proof of this concept comes from the world of wine. In 2001, Frederic Brochet, of the University of Bordeaux, conducted two separate and very mischievous experiments. In the first test, Brochet invited fifty-seven wine experts and asked them to give their impressions of what looked like two glasses of red and white wine. The wines were actually the same white wine, one of which had been tinted red with food coloring. But that didn't stop the experts from describing the "red" wine in language typically used to describe red wines. One expert praised its "jamminess," while another enjoyed its "crushed red fruit." Not a single one noticed it was actually a white wine.

The second test Brochet conducted was even more damning. He took a middling Bordeaux and served it in two different bottles. One bottle was labeled as a fancy Grand Cru. The other bottle was labeled as an ordinary vin du table. Despite the fact that they were served the exact same wine, the experts gave the differently labeled bottles nearly opposite ratings. The Grand Cru was "agreeable, woody, complex, balanced, and rounded," while the vin du table was "weak, short, light, flat, and faulty." Forty experts said the wine with the fancy label was worth drinking, while only twelve said the cheap wine was.

What these wine experiments illuminate is the omnipresence of subjectivity. When we take a sip of wine, we don't taste the wine first and the cheapness or redness second. We taste everything all at once, in a single gulp of this-wine-is-red, or this-wine-is-expensive. As a result, the wine experts sincerely believed that the white wine was red, and that the cheap wine was expensive. And while they were pitifully mistaken, the mistakes weren't entirely their fault. Our human brain has been designed to believe itself, wired so that prejudices feel like facts, opinions are indistinguishable from the actual sensation. If we think a wine is cheap, it will taste cheap. And if we think we are tasting a Grand Cru, then we will taste a Grand Cru. Our senses are vague in their instructions, and we parse their suggestions based on whatever other knowledge we can summon to the surface. As Brochet himself notes, our expectations of what the wine will taste like "can be much more powerful in determining how you taste a wine than the actual physical qualities of the wine itself."

The fallibility of our senses—their susceptibility to our mental biases and beliefs—poses a special problem for neural reductionism. The taste of a wine, like the taste of everything, is not merely the sum of our inputs and cannot be solved in a bottom-up fashion. It cannot be deduced by beginning with our simplest sensations and extrapolating upward. This is because what we experience is not what we sense. Rather, experience is what happens when sensations are interpreted by the subjective brain, which brings to the moment its entire library of personal memories and idiosyncratic desires.* As the philosopher Donald Davidson argued, it is ultimately impossible to distinguish between a subjective contribution to knowledge that comes from our selves (what he calls our "scheme") and an objective contribution that comes from the outside world ("the content"). In Davidson's influential epistemology, the "organizing system and something waiting to be organized" are hopelessly interdependent. Without our subjectivity we could never decipher our sensations, and without our sensations we would have nothing about which to be subjective. Before you can taste the wine you have to judge it.

But even if we could—by some miracle of Robert Parkeresque objectivity—taste the wine as it is (without the distortions of scheming subjectivity), we would still all experience a different wine. Science has long known that our sensitivity to certain smells and tastes varies by as much as 1,000 percent between individuals. On a cellular level, this is because the human olfactory cortex, the part of the brain that interprets information from the tongue and nose, is extremely plastic, free to arrange itself around the content of our individual experiences. Long after our other senses have settled down, our senses of taste and smell remain in total neural flux. Nature designed us this way: the olfactory bulb is full of new neurons. Fresh cells are constantly being born, and the survival of these cells depends upon their activity. Only cells that respond to the smells and tastes we are actually exposed to survive. Everything else withers away. The end result is that our brains begin to reflect what we eat.

The best documented example of this peripheral plasticity concerns the chemical androstenone, a steroid that occurs in urine and sweat and has been proposed as a human pheromone. When it comes to smelling androstenone, humans fall into three separate categories. The first group simply can't smell it. The second group is made up of very sensitive smellers who can detect fewer than ten parts per trillion of androstenone and who find the odor extremely unpleasant (it smells like urine to them). And the third group consists of smellers who are less sensitive to the odor but perceive it in oddly pleasant ways, such as sweet, musky, or perfume-like. What makes these differences in sensory experience even more interesting is that experience modulates sensitivity. Subjects repeatedly exposed to androstenone become more sensitive to it, thanks to feedback from the brain. This feedback causes the stem cells in nasal passages to create more androstenone-sensitive odor receptors. The new abundance of cells alters the sensory experience. What was once a perfume becomes piss.

Of course, in the real world (as opposed to the laboratory) it is we who control our experiences. We choose what to eat for dinner. Escoffier understood this first: he wanted his customers to order their own meals precisely because he never knew what they might want. Would they order a beef stew or some salmon quenelles? A bowl of bouillon or some sweetbreads larded with truffles? While Escoffier made his customers obey a few basic rules (no white wine with beef, no smoking between courses, no blanquette after a creamy soup, and so forth), he realized that every diner had a separate set of idiosyncratic desires. This is why he invented the menu: so his guests could choose dinner for themselves.

And every time a customer devoured one of Escoffier's dishes, choosing the fillet over the rouget, the sensations of that person's tongue were altered. When Escoffier was working at the Savoy in London (a joint venture with César Ritz), he had faith that he could educate even the palates of the British. At first, Escoffier was horrified by how his new patrons defied his carefully arranged menu. (He refused to learn English out of fear, he later said, that he would come to cook like the English.) Some customers would order two creamy dishes (a real faux pas), while others would want meat without a sauce, or would only have a little soup for supper. To teach his British guests how to eat a meal properly, Escoffier decided that any party larger than four people dining in his London restaurant could only have whatever dishes he put in front of them. He invented the chef's tasting menu as an educational tool, for he was confident that people could learn how to eat. Over time, the English could become more French. He was right: because the sense of taste is extremely plastic, it can be remodeled by new experiences. It's never too late to become a gourmet.

Since the publication of his encyclopedic cookbook in 1903, Escof-fier's culinary inventions have gone on to modify untold numbers of olfactory cortices, noses, and tongues. His recipes have literally changed our sensations, teaching us what to want and how our most coveted dishes should be served. This is the power of good cooking: it invents a new kind of desire. From Escoffier, we learned how to love flavorful stocks, viscous sauces, and all the silver accoutrements of fancy French food. He led us to expect food to taste like its essence, no matter how many extra ingredients that required. And while his love of butter may have shortened our own life spans, the wisdom of his recipes has made our short lives a little bit happier.

How did Escoffier invent such a potent collection of dishes? By taking his own experiences seriously. He knew that deliciousness was deeply personal, and that any analysis of taste must begin with the first-person perspective. Like Ikeda, he listened not to the science of his time, which treated the tongue like a stranger, but to the diversity of our cravings and the whims of our wants. Our pleasure was his experimental guide. As Escoffier warned at the start of his cookbook, "No theory, no formula, and no recipe can take the place of experience."

Of course, the individuality of our experiences is what science will never be able to solve. The fact is, each of us literally inhabits a different brain, tuned to the tenor of our private desires. These desires have been molded—at the level of our neurons—by a lifetime of eating. Escoffier's Guide Culinaire is more than six hundred pages long because he knew that there was no single recipe, no matter how much umami and cream it contained, that would satisfy everyone. The individuality of taste, which is, in a way, the only aspect of taste that really matters, cannot be explained by science. The subjective experience is irreducible. Cooking is a science and an art. As the chef Mario Batali once said about one of his recipes, "If it works, it is true."