A New York Times story about a dessert prepared by the French actor Gérard Depardieu brings together several themes of neurogastronomy:
“The other day we made a fondant of apples with three different kinds of apples—Canada, Granny Smith and Calville,” Depardieu said. “We did a white caramel sauce, put that on the plate. Then in the compote of apples, slightly reduced, I put a bit of butter and an egg yolk, and put that in the oven. It’s not at all like a compote because it’s lightly gratineed on top. We served it with a creme Anglaise and a caramel ice cream, and I said: ‘It’s far too sweet, too rich. Nothing explodes, there is no subtlety.’ So what we did is put an apple sorbet and a drop of rum—old rum that Fidel Castro took from the cave of Battista 50 years ago. He gave me a few bottles. And that”—he paused with a theatrical flourish—“that is God in velvet underwear.”
The quote shows the central role that flavor plays in our lives. It shows how fascinated people are by the procedures of cooking, one of our defining human characteristics. No other animal would be so crazy as to attempt to combine all these elements just to have a meal. There are at least 10 ingredients in this little dessert dish, ranging from raw apples to 50-year-old rum. They are subjected to several methods of preparation, from freezing ice cream and sorbet to heating a compote in the oven. There is testing at each stage all along the way for just the right flavor combinations.
And there is language. The dessert could not have emerged without language: words to describe and discuss each flavor combination at each stage; words to express the disappointment with the excessive sweetness and lack of subtlety and to find the solution; and finally, words to express the joy at the right result.
Human Language Is Essential for Human Flavor
In the action section of the human brain flavor system (see figure 18.2), language is therefore an essential part. It is one of the main reasons our flavor system is unique among all animals. If cooking is the defining characteristic of humans, as James Boswell averred, flavor is its joy and language is its handmaiden.
This essential role of language in relation to smell and flavor may come as a surprise. The received wisdom (I have believed it myself) is that we are poor at describing our world of smells. There are no words that characterize smells in specific ways, the way we use words to describe colors or simple shapes, for example. Language seems to fail, to resort to analogy, metaphor, or reference to the object that produces the smell. Thus, things smell “sweet,” like the taste of sugar, or, for the scent of bananas, “like bananas.” For an organic chemist, there may be meaning in describing the scent of bananas as the smell of amyl acetate, but if you ask how amyl acetate smells the answer will probably be “like bananas.”
The argument against a close link between smell and language is also made on scientific grounds. As we have seen, the cortical receiving area for smell is in the most anterior part of the frontal lobe. By contrast, the cortical area for the motor control of language is Broca’s area in the posterior part of this lobe, and the area for the sensory integration of language, Wernicke’s area, is in the temporal lobe even further away (see figure 13.1). This shows, it is argued, the tenuous connection smell perception has to language. A similar argument would apply to the relation between flavor and language.
Wernicke’s area is close to the auditory receiving area, which seems logical because of its direct involvement in interpreting the sounds of the spoken language, and Broca’s area is close to the motor area for control of the muscles of speech. Apart from that, the two language areas are related by long-distance connections to many parts of the cerebral cortex in order to form the spoken language in the service of all the sensory and motor systems of the brain. Just as the frontal lobes are connected to the visual areas in the occipital cortex, so are the prefrontal areas connected to distant parts of the cortex, including areas related to the speech areas. What matters is whether the connections can subserve not only the sensory functions related to flavor but also the learning required in relating the perceptions to the neural machinery of language.
There are strong indications that they can. We have already seen that the ingestion of food, and the retronasal smells involved in flavor, occur through the mouth, the same orifice used in the production of language. Food and language are not only close neighbors in this sense; they occupy the same house. Further reasons for a close link between smell and language are, first, the strong evolutionary link between the rise of cooking and the rise of language, and second, the extraordinary vocabulary we actually have developed for flavors (witness the Depardieu dessert).
From the earliest documents of recorded history, it is obvious that foods and their preparation have been high priorities of human societies. It is from these times that the traditional cuisines from around the world have emerged.
In perusing some of the history of those times from the perspective of the sense of smell and its role in food flavors, I have been particularly struck by the richness of the cuisines and the associated richness of the cuisine vocabularies. For example, by Roman times there was a huge demand for herbs and spices imported from Arab traders, who brought them through the silk and spice routes that extended from the Far East (this was a thousand years before Marco Polo). In The Spice Trade of the Roman Empire, J. Innes Miller lists the spices that were mentioned in the cookbooks of the time; they totaled 87. By comparison, a recent edition of The Joy of Cooking lists only 18. This vast array of herbs and spices, each with a name and a characteristic aroma and flavor, indicates the richness of both the cuisines and the languages to describe them from earliest times.
The Vocabulary of Flavor
Today, the vocabulary of food aromas and flavors seems overwhelming. It would help if we could classify all smells and flavors as arising from the mixing of a few “primary” smells, in analogy with how colors arise from just three primary colors, but smell and flavor are too complex. There are in fact several vocabularies. Organic chemists have produced thousands of terms for the smells and flavors of the molecules they synthesize. Psychophysicists have their vocabulary for the flavor components isolated by mass spectrometry, as described in chapter 4. And food scientists have their terms for characterizing the flavors of foods, as do wine tasters for their wines.
An example may be found on the Flavornet Web site, maintained by the leading psychophysicist Terry Acree and his colleagues at Cornell University: a compilation of more than 700 odorous food components, with data on their chemistry and sensory properties, organized into 25 classes, such as fruity, cooked meat, dairy, fishy, spicy, and so on.
This list provides dramatic evidence that the large universe of smell molecules produces a corresponding large universe of smell perceptions and the vocabulary to give them identities. This is amplified by the associations each of the food smells has with the flavor of which it is a part.
The intriguing idea is that all this vocabulary plus the syntax and grammar to communicate it reflect the attempt on the part of humans to describe their world of smell and flavor. Some claim that humans can discriminate among 10,000 odors. However many it is, there must be a corresponding number of words to describe them. Moreover, we have seen that the sense of smell involves not only the perception of a scent, but also the associated memories that are evoked and the emotions that are attached to it. These amplify the vocabulary, so that as we saw with the incident of Marcel Proust and the madeleine, a perception can bring back the whole scene of a bygone time and the emotions connected with it, all requiring the use of language in order to identify the memory, describe the emotion, and communicate it to others.
A major hypothesis of this book is that the reason it is so difficult to describe these smell and flavor perceptions in words may be that they are represented in the brain as arbitrary irregular patterns of activity, what we have called “smell images.” As we argued in chapter 8, it is difficult to describe in words a nongeometrical visual image such as a face, even though we can identify it unerringly. In like manner, it may be postulated that it is difficult to describe in words a smell image, even though we can identify it unerringly as well.
Thus, connecting smells and flavor with language may be difficult, but it is a uniquely human endeavor. That we require effort to do it, using all the linguistic tricks at our disposal (analogies, metaphors, similes, metonyms, and figures of speech) qualified by the entire vocabulary of emotion (joy, despair, hate, revulsion, craving, and love) should not come, therefore, as a surprise. Gérard Depardieu was only doing what came naturally.
Among the best challenges to the use of language to evaluate flavors is wine-tasting by experts. In chapter 15, we saw the importance of language in assessing the flavors of red and white wines when complicated by the factor of color. Here are two further examples of using language to characterize wine flavors.
Ann Noble and the Wine Aroma Wheel
Ann Noble is such an expert, having devoted her life to the scientific analysis of wine tasting in her laboratory at the University of California, Davis. I visited Noble in her home in Davis several years ago, and also her laboratory for wine tasting. For this work she has developed “The Wine Aroma Wheel,” with terms to describe wines organized in three concentric circles, starting at the center with the most general terms (fruity, earthy, and so on), to more specific (berry, citrus, and the like), to the most specific in the outer circle (blueberry, other specific fruits, and flavors). She proposes that this provides a logical system for using language tags to work one’s way through a hierarchy of classification of percepts. It enables even the beginner to recognize varietals (based on the type of grape) and the “notes” of different wines.
Robert Parker and the Language of Wine
Robert Parker, renowned for his million-dollar nose, stands in contrast. In his Parker’s Wine Buyer’s Guide (from my 1995 edition), he catalogues more than 7,500 wines, based on his ability to discriminate among different labels and different years. How does he do it? Let us look at some examples of his descriptions.
His summary of the qualities of one of the great wines, a 1994 Petrus Bordeaux, starts with an overall comment on it being a “powerful, tannic, backward looking wine.” Powerful presumably means strong-tasting, tannic refers to the astringency due to the binding of proteins (chapter 14), and backward looking presumably means a long aftertaste. The color he describes as “deep, dark ruby/purple,” the darkness increasing with age and complexity. He then goes on to comment that “with coaxing” (slow careful swishing of the wine in the mouth?), the “closed nose [presumably meaning retronasal smell with the wine held in the mouth] offers up scents of coffee, herb-tinged, jammy black cherries, and toasty new oak.” On his scale of 100 points being best, he rates it a 91–93-plus.
For comparison, the next wine in the list is a 1994 Phélan-Ségur (Saint-Estèphe), which has an “impressive saturated purple color, sweet, jammy cassis aromas, medium body, moderate tannin, and fine ripeness and chewiness in the mouth and finish.” Here we have almost all the senses combined in the evaluation: purple color, sweet taste, cassis aromas, and tannin feel, plus even the motor quality of “chewiness” in the mouth and finish. From this perspective one can begin to see the logic in Parker’s approach, using language to differentiate the qualities of all the sensations and motor qualities of the wine. In addition he notes the “finish,” which includes the dimension of time for the evolution of the whole effect.
For comparison, let’s go to California, where Parker has built up enthusiasm for wines that stand equal to the best of France. He especially likes the Robert Mondavi wines from the Napa Valley. Among the choicest is the Cabernet Sauvignon Reserve. He describes the 1990 vintage as exhibiting an “opaque dark ruby/purple color.” There is a “huge” nose, full of “abundant aromas of sweet, toasty, new oak, jammy cassis, and roasted nuts,” with “a concentration of cassis fruit welded to sweet, soft tannins.” The finish is “velvety until some tannins emerge as the wine sits in the glass.” Here again we see the evaluation emerge from a combination of color, taste, nose and mouth-feel, and the evolution of the flavor with time, not only from the lingering flavor from what is left in the mouth and throat but also from the wine simply sitting in the glass.
In 2000, Frèdèric Brochet and Denis Dubourdieu in Bordeaux analyzed the language used by Parker and three other wine experts. They found that the specific sensory qualities were evaluated within the context of whether the taster liked those properties, singly or together. This is the same axis along which Noam Sobel and his colleagues arrange smells (chapter 12). Brochet and Dubourdieu conclude:
Although many efforts have been made to characterize the quality and flavor of the compounds in wine by gas chromatography and other chemical techniques, tasting remains the single universal test used to assess properly wine sensory properties…. The main cognitive concern regarding flavors is whether they are good or not. This concern is so strong that even experts cannot ignore it and it is what drives the organization of their descriptive language. In this way experts are not so different from novices…. P[arker] evaluates as would a novice, and this could be one reason for the extraordinary success of this writer.
Language Is Difficult for Faces, Art, and Music, Too
Because the concepts of smell images and flavor images, as well as images of desire, are so new, it is understandable that there have been no studies on how to describe a wine in terms of these images. We can therefore only speculate. If we recall that the core problem is describing a highly irregular smell image, it may help to compare this with the equivalent problem in other sensory systems. We think we can use language in very precise ways in vision and hearing, but is this true? Surprisingly, I think not, if we use our visual analogy again.
In vision, familiar objects—tables, houses, flowers—are so easily described that it seems that this is the system best connected to language. However, is this true for irregular objects? For example, a face is a complex irregular spatial object. As noted in chapter 8, humans are extremely good at recognizing faces. This ability is of critical adaptive value in recognizing not only a particular individual but also the nuances of facial expressions, conveying all the emotions that enter into human relations.
Recognizing a face, as we have discussed, is a case of pattern recognition. It is often claimed that this is one of the things humans are best at. We’ve already indicated how, if you are shown a number of pictures of grandmothers’ faces, you will have no difficulty picking your own grandmother out very quickly. But can you describe to someone else in words how you identify her? Try to describe for yourself your grandmother’s face in a way that would tell others exactly what she looks like. You might agree that we are at loss for words to describe a face, just as most of us (although apparently not Robert Parker) are at loss for words to describe a wine.
Another example from vision is painting. Representational art is easy; describing in words a painting of a landscape by Constable or Cézanne or a sunflower by van Gogh can be done by anyone. However, Picasso and the cubists present more difficulty, and current modern painters even more so.
Take, for example, Cy Twombly. An American artist born in 1928, he developed as an Abstract Expressionist whose influences came from Europe, Africa, classical art, and mythology. Many of his works contain differently colored scratchy blobs seemingly scattered over the canvas, with, however, a disciplined structure that is impenetrable to the casual viewer. Here is a description of his style from a recent exhibition:
Twombly’s “literariness” is something that has consistently told against him, along with his fancy foreign ways and his “insinuating elegance.” But his art, as [Nicholas] Serota acknowledges in the catalogue, has always been elusive and, for many people, even enthusiasts of contemporary art, unfathomable. Twombly himself has maintained an unusual reticence. In the mid-50s, he wrote a short statement for the Italian art journal L’Esperienza moderna: “To paint involves a certain crisis, or at least a crucial moment of sensation or release; and by crisis it should by no means be limited to a morbid state, but could just as well be one ecstatic impulse.”
Literariness, insinuating elegance, and unfathomable are words that would be in some people’s wine vocabularies. It appears that observing the complex visual image of a vintage Twombly brings forth the same challenges to language as sensing the flavor image of a vintage wine. This is exactly what was implied when introducing the concept of smell images in chapter 9.
Another example comes from the world of music. We can all recognize and reproduce a simple melody. However, classical and jazz composers seek much more. Bruno Walter (1876–1962) conducted many of the great symphony orchestras around the world in the first half of the twentieth century. In his memoir Of Music and Music Making, he describes the music he conducted: “Music springs from and is replenished by a hidden source which lies outside the world or reality. Music ever spoke to me of a mysterious world beyond, which moved my heart deeply and eloquently intimated its transcendental nature.”
Here, too, a wine connoisseur could have used some of those same words. It appears that the complex temporal images of music are as difficult to describe in words as the complex spatial images of art and the complex brain images of wine. A picture created within the brain is worth a thousand words, whether it is representing an abstract painting, chords of music, or the molecules of smell.