"No one can describe a taste" - Why?

Via Andrew Sullivan, Jonah Lehrer channels Proust to explore why smells (and tastes) are so emotionally laden:

Why is smell so sentimental? One possibility, which is supported by this recent experiment, is that the olfactory cortex has a direct neural link to the hippocampus. In contrast, all of our other senses (sight, touch and hearing) are first processed somewhere else - they go to the thalamus - and only then make their way to our memory center. This helps explain why we're so dependent on metaphors to describe taste and smell. We always describe foods by comparing them to something else, which we've tasted before. ("These madeleines taste just like my grandmother's madeleines!" Or: "These madeleines taste like the inside of a lemon poppy seed cake!") In contrast, we have a rich language of adjectives to describe what we see and hear, which allows us to define the sensory stimulus in lucid detail. As a result, we don't have to lean so heavily on simile and comparison.

C.S. Lewis, in the final Narnia chronicle The Last Battle, casually delivers a different (though not really contradictory) explanation for why we depend on metaphors to describe taste and smell. When his newly-dead protaganists taste the first fruits of paradise, Lewis writes:

What was the fruit like? Unfortunately, no one can describe a taste. All I can say is that, compared with those fruits, the freshest grapefruit you've ever eaten was dull, and the juiciest orange was dry, and the most melting pear was hard and woody, and the sweetest wild strawberry was sour (Ch. 13, my emphasis).

So according to CSL, it's not that we're less inclined to create descriptive language about taste and smell: for some reason we just can't.

Why is that? We generally describe taste and smell, as Lehrer suggests via Proust, by reference to another taste or smell. But isn't the same true of sight? Start with colors: each color name is a giant analogy -- or scientifically speaking, a classification, grouping objects the surfaces of which really do reflect light in the same range of the spectrum. In fact, re that "rich language of adjectives" we use for sight and hearing -- every adjective is a compressed or aggregated simile or analogy, a classification.

I think that we describe sights more precisely than smells and tastes not because smell and taste are more emotionally laden but because they're less precise senses than sight. (Maybe because of some processing in that trip to the thalamus that smells don't make?) You can say of a tree's appearance that it's thirty feet tall, has a spear-shaped leaf crown, reddish bark in fishlike scales, and needle foliage; all you can say about the experience of eating a grapefruit is that it tastes like a more sour orange and smells fragrant and pungent. All language is ultimately relative, comparative -- but our range of comparison is much richer with visual data.

Dogs are supposed to rely far less on sight and more on smell than humans. I have a blind dog, half beagle to boot, and I can report that although he gets along pretty well, even for a dog smell is no substitute for sight.


While he'll always find food that you toss him, he does it by elaborate, slow elimination -- and he can walk right over it, more than once, before putting his nose (and mouth) to it. Smell is time-limited: when someone runs by he gets very excited, but he has no idea where they are. His hearing actually seems like a nearer sight-substitute than his sense of smell: when he's chasing a bouncing ball you'd think he could see until the bouncing noise stops, at which point he's relatively helpless.

Smell and taste may go straight to our emotional core. But their superior impact seems part and parcel with inferior discernment.