Category Archives: umami

Feeling “umami”: On taste, subjectivity, and metaphor

The Modern Four Taste Orthodoxy

The idea that there are four basic tastes—sour, salty, bitter, and sweet—was widely taken to be gospel truth until 2002, when the taste receptors for glutamate were identified. Glutamate, and the "umami" flavor it imparts to foods like seaweed, bacon, parmesan cheese, and Doritos, was first identified and isolated by a Japanese chemist named Kikunae Ikeda in 1908. But the four-taste model was so dominant that umami’s status as a distinct taste was considered "controversial" until its molecular basis was confirmed almost a century later. Here’s the description of umami in a book titled Sensory Processes published in 2002 (before the taste receptors were identified):

Umami originated from a glutamate derived from seaweed. The chemical substance is commonly known as MSG, monosodium glutamate, and, by itself, has no odor and an unusual taste that is approximated, so they say, by appropriate combinations of the four primary taste qualities. Whether umami is a result of the unique combination of the four tastes or an independent classification of is own is open to debate (176).

Oh, they and the things they say.

Umami’s been gaining traction—Kikkkoman’s current advertising campaign is "discover umami"(.com)—but it hasn’t quite arrived. This past Sunday on Iron Chef America, one of the judges said he detected some "umami" in a coconut-based soup, and then he had to define it for the other judges (and perhaps the audience?). His first stab was to call it an "illusory" taste, although he did follow that up by ranking it with "sweet, salty and sour," so perhaps he actually meant something more like "ineffable." Either way it shows how dominant the classical four tastes still are.

Except "classical" isn’t quite right word, because it turns out the idea that we only experience four distinct tastes is actually a pretty recent invention—more recent than Ikeda’s "discovery" of umami, actually. Traditional Chinese medicine named five tastes: sour, bitter, sweet, pungent, and salty, each one corresponding with one of the five elements or movements that are omnipresent in early Chinese thought. Aristotle claimed there were only two, which doesn’t come as that much of a surprise given his characteristic love of dualities. Just as he divided visual perception into the "fundamental colors" black and white, which contain all the elements of all the other colors we perceive, Aristotle thought the whole range of gustatory sensations derived from the "fundamental tastes" of sweetness and bitterness. He also proposed a second-order classification of seven "primary flavors" that corresponded with his rainbow of seven "primary colors": sweet (which included fatty or oily), bitter, salty, harsh, pungent, astringent, and acidic or sour. Hard to say now what exactly the difference was between harsh, pungent, and astringent—I suppose the latter might be something like the tannins in tea and red wine while "pungent" instantly evokes blue cheese although it often just acts as a modifier rather than a descriptor—a pungent smell is strong, not necessarily strongly any particular thing. It’s hard to even think of those things as taste categories on the same level as "sweet" or "sour."

But one question that raises is whether or not it’s hard to think of them that way because there’s some objective difference between sweetness and astringency or because it’s just unfamiliar to think of "astringent" as a primary taste category. Certainly tannins cause a particular reaction on people’s tongues—is that less of a distinct taste experience than the reaction caused by sugars?

Colors are a useful parallel, again. A linguistics professor I had at NYU told us about this experiment that my casual googling is not coming up with, but here’s the gist: if you give children a set of colored tiles and tell them sort them into as many piles as they want, by color, there are predictable, reliable differences between the number of piles they make that correspond to the number of primary colors in their primary language. So, for example, English-speaking kids generally put all hues of blue in one pile while Russian-speaking kids usually separate lighter blues from darker blues because they have two "primary color" words for those shades. It’s one of the classic examples of how language can shape how we perceive the world rather than just reflecting it. Also a reason why translation is always imperfect. 

Henning taste tetrahedronGetting back to the four taste orthodoxy, that was something a German psychologist named Hans Henning  came up with in 1916. He devised a ""taste tetrahedron" with each of the four tastes he thought were primary at the four vertexes. The idea was that flavors could be conceptually mapped onto geometric plane based on which of the primary flavors they were comprised of—a flavor relying on two of the primary tastes would be located on the edge between those two vertexes. Flavors that used three would be on the surface between the relevant three points. And the tetrahedron was hollow, according to Henning, because no substance could produce all four taste sensations. So while taste itself was three-dimensional, tastes were two-dimensional at best.

Minor digression: I suspect Jean Anthelme Brillat-Savarin, the eighteenth century French gourmand who’s responsible for the cliche "you are what you eat" (well, ish, the problem of translation rears its head again; what he wrote was Dis-moi ce que tu manges, je te dirai ce que tu es, which is closer to, "Tell me what you eat, I will tell you what you are") would disagree, and name as exhibit A something involving veal stock (a sweet-sour-salty-bitter cabbage soup maybe?). Speaking of veal stock, Brillat-Savarin also sought to identify the special savory quality of "the purely sapid portion of flesh soluble in cold water," or functionally, umami. His word was osmazome. And surely Brillat-Savarin wasn’t the first to muse on the particular deliciousness of dashi, truffles, and tomatoes—especially after they’ve been cooked down into a rich sauce. So the 1908 "discovery" of umami turns out to be, like so many "discoveries" (*cough* America *cough*), basically a trumped-up (re)-christening.

Henning pruned away tastes like "astringent" and all the shades of flavor produced by the almost infinite variety of aromas we can detect because those aren’t, strictly speaking, "taste" sensations. The prickling and burning sensations caused by capcaicins (which make peppers "hot"), the puckering induced by tannins, and the palpable richness of unctuous or viscous foods are all tactile sensations. And aromas, obviously, are processed by the separate-but-related chemical sense of smell. As most people know, particularly if they’ve ever been seriously congested, smells are what turn functionally one or two-dimensional taste sensations into much more complicated (and enjoyable) perceptual experiences. Laboratory experiments in the early twentieth century that involved delicately swabbing the tongues of blindfolded, noseplugged subjects confirmed Henning’s taste quartet. In those conditions, the only things most tasters could reliably identify were sour, salty, sweet, and bitter.

More science to the rescue?

Setting aside for a moment the question of whether or not "taste" should actually be seen as a multi-sensory experience, those experiments were also limited by the descriptive vocabularies of the participants and the kinds of compounds applied to their tongues. Did the researchers try compounds that would have tasted alkaline or metallic or umami to most people? Even if they did, maybe people wouldn’t identify something like "umami" (or "osmazone") because they didn’t have a word for it, not because they didn’t taste it. Or maybe if they did, they called it "savory," and that was conflated with "salty."

Umami is, if not illusory, still notoriously difficult to isolate when tasting food and even harder to characterize. Perhaps part of that is because the compound itself has no smell, although neither do sugar or salt, and perhaps another part is that its description always ends up sounding more evaluative than descriptive—I mean, the Japanese term literally means "delicious." On the Iron Chef America episode I mentioned earlier, another judge playing with his new vocabulary joked, "Ooh, mommy!" which was dumb and irritating, but not actually completely off-base in terms of trying to describe the particular sensation imparted by glutamate. Try eating msg plain sometime (it’s available as Accent(tm) in the spice section of most grocery stores and in bags ranging from a few ounces to multiple pounds in Asian markets. And no, it won’t kill you, which is a post I’ll get to some other time)—it tastes like the exact intersection between Doritos, instant ramen, and Heinz ketchup. How do you describe that? It’s like the taste of tastiness.

More recent research, most of it enabled by the sequencing of the human genome, suggests that four tastes is far two few. While scientists haven’t identified them all yet, they now estimate that there are probably about 40 distinct taste receptors (and 300 distinct olfactory receptors), at least half of which are devoted to detecting bitter tastes. They’ve also discovered a number of complications in terms of how chemicals react with those receptors, and how the response triggered by different chemicals is perceived and processed in the brain. Even with other senses muffled, it now seems that people can indeed identify metallic, alkaline, and umami in addition to the big four. If we have trouble differentiating between all the different kinds of "bitter" we can taste, that’s probably because we tend to be averse to those tastes and have little practice trying to distinguish or name them, not a lack of complexity at the level of the tongue.

Additionally, tastes interact, even without the "interference" of smell or feeling. Sour flavors dampen bitter flavors (the role of the lime in a gin and tonic) and if you have a lot of something sour, it may make other foods or even relatively "neutral" substances like water taste sweet. There are also substances in specific foods that can mess with your taste receptors, which is why artichokes give anything else you eat for a while after an additional subtle sweetness (part of why it’s a great appetizer ingredient). The most dramatic example is the "miracle fruit" people got all excited about last year that contains a protein that temporarily binds to "sweet" taste receptors and reacts with acidic compounds, meaning you can eat plain raw lemons and they taste like the sweetest candy. (Note: it is kind of cool to be able to scarf down lime wedges like potato chips, but eating that much acid turned out to be sort of a regrettable decision for a lot of the people at our miracle fruit party. The remainder of the tablets we bought have languished in a drawer for over a year.)

To complicate matters even further, not all people taste the same things the same way. "Supertasters" are highly sensitive to bitter and spicy compounds, and some people really do have a "sweet tooth" that makes them inclined like sweeter foods more than most people, which could reflect a different perception of sugars anywhere between their tongue and their brain. Related: some fifteen percent of the population thinks cilantro tastes like soap.

And, as most people know from personal experience, the way people respond to the same foods may change over time—I drank about a hundred mochas when I was fifteen and had an irrepressible crush on a Starbucks employee, which stopped me from ordering the hot chocolate I actually wanted because I didn’t want to seem like a little girl. Gradually, aided no doubt by the fact that mochas are as close to hot chocolate as a coffee drink can possibly be, I came to like the coffee flavor. Within another year or so, I honestly preferred my coffee black (and personally delivered by a particular waiter at my local Denny’s).

But all that said, I can’t quite get behind the idea that we’re all special snowflakes and taste is an entirely, or even primarily subjective, individual experience. I just don’t buy that in five years or so, I will be able to "tell [a restaurant] my flavor type on the Internet at the time I make my reservation and [have them] design a meal just for my DNA," the way this Gourmet magazine article suggests.

Fill in the blank: My, you’re ____________! A) sour B) salty C) bitter D) sweet E) umami

Surely it’s not a purely arbitrary coincidence that the four tastes Henning settled on were part of basically every attempt to classify the primary tastes from ancient China to ancient Greece to Restoration England. Would anyone, ever, propose a four-taste system that included only harsh, pungent, astringent, and sour? Or metallic, umami, spicy, and oily?

It seems odd, and potentially significant that the four tastes Henning canonized are also the tastes with the most widespread metaphorical use. They’re not just taste sensations, they’re part of our basic descriptive currency for emotional states, facial expressions, personalities, reactions, gestures, and the things that prompt those things. Sourpuss. Salty humor. A bitter pill. A sweet smile. And they’re used similarly in other languages—the French refer to someone being overly polite or affable as "Etre tout sucre tout miel," or "being all sugar all honey," and the phrase "sweet as sugar/honey" in Arabic (ahla-mina s’sukkar/l’asal) means the same thing it does in English. I’m sure there are a million more examples, but foreign languages aren’t my strong suit.

Other flavors can be, and often are, used metaphorically, especially "spicy" (which, remember, was a common candidate for the fifth spot in many pre-20th C. classification schemes), but few of them have permeated to the point of idiom and cliche the way the big four have. You might describe a person as astringent or a prospect as savory, but both involve a taking a little poetic license. Others are even further afield—I guess I can imagine a metallic facial expression or alkaline feeling, but I think you’d need other context to help you out there and I’m really not sure what it would mean to describe something other than food as "umami."

At least one thing the metaphorical use of taste seems to suggest is a minimum amount of shared taste perception. When Shakespeare used the phrase "honey tongue," no one in his audience needed to have heard the phrase before to understand what he was getting at. With all due respect to individual genetic and cultural differences, it seems to make more sense that  people would largely share the same taste experiences than that they would differ, at least in extreme ways. It seems only natural that we should all be repulsed by the bitter poisons that would kill us, and that our understanding of what sorts of feelings are involved in and communicated by a "bitter glance" is part of that common taste experience.

So, by way of explaining the title of the blog, I didn’t leave out tastes like umami and metallic and alkaline (and however many more have yet to be named) because I wanted to reify the outdated idea that we only have four basic tastes. Instead, I wanted to invoke the dimensions of taste that seem most central to our experiences with food and also impart clear metaphorical connotations. The twinned subjects of this blog—the food I make and eat and my experiences and concerns as a cook and an eater—will both be sour sometimes, salty often, bitter occasionally, and hopefully sweet at least from time to time. I would try to be and feel and say umami things too, but I’m not sure I know how.

References not linked above:

Aristotle and William Alexander Hammond. 1902. Aristotle’s psychology: a treatise on the principle of life (Cambridge, MA: Harvard University Press). Digitized 2006 by Google.

Brillat-Savarin, Jean Anthelme. 1825. The Physiology of Taste or Transcendental Gastronomy, trans. Fayette Robinson. Digitized 2004 by ebooks@Adelaide and Project Gutenberg

Korsmeyer, Carolyn. 1999. Making Sense of Taste: Food & Philosophy. Ithaca, NY: Cornell University Press.