Smells (and the people who write about them)

by Rishidev Chaudhuri

ScreenHunter_08 Feb. 20 12.38Luca Turin and Tania Sanchez's “Perfumes” is an intriguing (and rather wonderful) collection of reviews of various scents. It is hard to write about smell while avoiding cliches: smell is almost always seen as a primitive, noble-savage sort of sense, pre-verbal and inextricably linked to sex and memory. They, on the other hand, start with the assumption that smell should be taken seriously as an artistic medium, and that viewing perfume simply as bottled memory or barely sublimated sexual enticement is misleading. Perfume is not simply mimetic, not simply trying to smell like the natural world, and we should take olfactory abstraction as seriously as we do visual abstraction. This makes for an often odd collection of perfume reviews. After all, what does one make of reviews like these:

“The result was the powderiest, rootiest, most sinister iris imaginable, a huge gray ostrich-feather boa to wear with purple devore velvet at a poet's funeral”

“The surprise effect of Le Feu d'Issey is total. Smelling it is like pressing the play button on a frantic video clip of unconnected objects that fly past one's nose at warp speed: fresh baguette, lime peel, clean wet linen, shower soap, hot stone, salty skin, even a fleeting touch of vitamin B, and no doubt a few other UFOs that this reviewer failed to catch the first few times.”

“Maurice Roucel has a knack for putting together perfumes that feel haunted by the ghostly presence of a woman: Lyra was a compact, husky-voiced Parisienne, Tocade a tanned, free-as-air Amazon…. However it did constrain the woman inside Envy to be at once seraphic and sub-urban, complete with the sort of suppressed anger that such a creature would feel at being reincarnated as a florist in Eastern New Jersey.”

“When it comes to arranging a folk theme, the early twentieth century had Mahler and L'Heure Bleue, and we have Enya and Fleurs d'Oranger. The brilliance of L'Heure Bleue is to cast fresh, shallow, sunny orange flower in a huge role, flanked by two giants: eugenol (cloves, carnation) and ionones (woody violets). Their chorus inexplicably achieves a mouthwatering praline effect close to that lethal Torino delicacy, the Gianduja. This is Guerlain the virtual pastry chef at his best, with a fragrance that teeters on the edge of the edible for hours without missing a step. If you're Red Hot Riding Hood and a hungry wolf just rang the bell, this is the one for you.”

These reviews are quite different in kind both from the review of something functional, like a deodorant (imagined review: “product did not work as advertised, did not cover up my smell, drove away my lover”) and of the sort of review that one might write out of a more naturalistic view of perfume (“faithful reproduction of the smell of rose”). How are we to make sense of what they're saying? Part of the reviews are of the “this is how this made me feel” sort; but a large part is attempts to convey what something smelt like. Are they actually doing this? Is something being said here, in this odd oscillation between scientific-chemical descriptions and impressionistic ones?1

Smell frustrates in the paucity of the language we have attached to it. Perhaps this is a consequence of the way it intuitively seems primitive and direct and strongly evocative of memory, or perhaps this is the cause, but it is very hard for most people to describe a smell with any degree of precision or even to remember complex smells vividly, the way one can imagine a piece of music or a color. Very few people grow up learning precise names for the smells of things, and no one is sent for smell lessons the same way they are for music or art training. Something similar applies to the combination of smell and taste we call flavor. For the most part, we don't really grow up learning a flavor vocabulary, and while we know how to classify tastes in the narrow sense (salty, sweet, etc.), this is only a small aspect of what we “taste”, which is much more about odor (and somewhat about texture). There is something more of a shared flavor vocabulary, because we grow up cooking and eating a range of things, and talking and thinking about them, and so start to see similarities and form somewhat more abstract concepts. But this is still imprecise. At some point you might realize that anise and tarragon have something in common, for example, and if you're scientifically minded you might dig up the chemical name, but this transfers only partially to everyday experience and doesn't really tell you which flavors will be perceived similarly and which flavors to pair with which.

The questions about language that writing about smell raises are not new or unique, but they are especially acute examples of a general problem. For most people, this is probably commonly encountered in wine reviews and accounts for a common suspicion about wine descriptions and wine snobbery. When reading a wine review, how are we to make sense of what the words signify, and how does training lead us to understand their meaning? There are shared criteria that are verifiable in principle, but only in a somewhat weak and perhaps artificial sense, and it often seems that what every wine connoisseur is smelling is something different or sufficiently different as to be unconstrained. Or, at the very least, that it is so shaped by expectation and environment (e.g. the studies of the effect of the believed price of wine on the words used to describe it and its perceived pleasurableness) that the descriptors are much less stable than, say, describing something as “blue”.

So how does one person's “cat piss” become the other person's? On one hand, it seems hard to argue that these descriptors are transparently expressing a shared cognitive architecture of smell and flavor. While there certainly is a shared sensory apparatus, any real smell is a complex mixture that is more than the sum of its parts and is a combination that we haven't grown up agreeing on how to decompose. On the other hand, the argument that our words are standing in for particular configurations of behavior and particular responses to things, and that it is these that are shared and that underlie meaning seems much more tenuous here than when referring to, say, pain or even to color which can be isolated and where particular sub-parts can be pointed at.

We often fall back on scientific language when describing smell, as Turin and Sanchez's book often does. In some ways, this is a stand-in for our lack of an actual shared cultural language. But the problems we have speaking about smell are paralleled scientifically. For our other senses we understand the units of composition and how they fit together. Color vision works by combining red, green and blue, and so these are the basic words with which we make color sentences. Two colors are similar to each other if they have similar fractions of red, green and blue, and so on. Similarly, we understand audition in terms of frequencies and sounds that have similar frequencies sound similar. But we don't have clean ways of understanding how smell is organized or categorized. Similarly (or equivalently, depending on what sort of Kantian you are), we don't really understand how the brain decomposes smell or why two things smell alike. So we're left with a set of fragmentary relationships and incomplete analogies and the scientific language we use is that of particular chemicals and families of chemicals rather than the particular properties of these chemicals that make them smell similar or different.

One of my favorite bars briefly took the extreme expressionist tack and completely eschewed general wine terminology or any attempt to describe particular flavors in their wine reviews; instead each wine was described by an extended rambling meditation on what the wine made the person writing the menu think of. One, for example, was described as being like the feeling you get when you stumble into your grandfather's cupboard and smell his coats. Again, what is a description like this doing? The puzzling thing is that it is certainly doing something. I know, at least vaguely, what is meant here and, even more vaguely, the sort of wine this probably corresponds to but I can't quite put it into a coherent description and it isn't immediately clear to me whether this understanding is shared or particular.

This strange range of possibilities makes smell and flavor a lot of fun to write about. They often feel only loosely tethered to language or like we're inventing a language as we write about them. Of course this tells us nothing about how an invented language is possible: how an invented language can make sense to anyone besides its creator or, from a more Wittgensteinian angle, how it is coherent to speak of an invented, private language.

It's also fascinating to look at more phenomenological attempts to create or refine the language we use to speak about smell and flavor – to formalize the process that chefs, perfumers and wine critics find themselves going through. How would one do this? The most interesting suggestions I can think of focus on elaborating basic categories of similarity and difference. So, for example, if you were trying to explore the structural space wine occupies, you might start with two wines, putting one of them into two glasses and the second into one, and then do a blind test to tell the odd one out. This seems to get at the most primitive structural operation – when do two things qualify as the same or as different? You might then move on to tasting similar wines that differ on some particular character (e.g. oaked vs not) as a way to highlight that difference and so on, constructing a vocabulary by drawing out basic structural oppositions.

Both of these approaches – attempting to invent new languages or attempting to train the senses – replace an attempt at understanding exactly what we're doing when we smell and describe smell with the cultivation of smell and of descriptions of smell. As in so many other spheres, the two endeavors are most interesting when hand in hand (compare the oft theorized switch back and forth between understanding what we are doing when we have and talk about sex and cultivating the ways we have and talk about sex). In the next decade or two it'll be interesting to see how these two approaches inform each other. We learn more every year about how our brains make sense of smell and it is at least plausible that we'll soon have a much more scientifically grounded decomposition of the space in which we smell. What will this do to the way we think about and talk about and experience smell in the everyday world? I find myself hoping that the puzzles of smell won't quite go away and that our various senses, apart from the trivial fact of allowing us to experience the world in different ways, will continue to allow us to think about the world in different ways and to allow for different metaphorical possibilities rather than simply being variations on a similar sensory theme.

1The book is certainly worth reading. Even if you don't care about smell or think that their approach is misguided and self-indulgent, there are enough whimsical and opinionated prose moments to delight most people.

Like what you're reading? Don't keep it to yourself!
Share on Facebook
Tweet about this on Twitter
Share on Reddit
Share on LinkedIn
Email this to someone