Language of smells

I'd been musing on the question of whether we could more easily distinguish smells if we developed a language for them: that having words for a scent might make it easier to identify. I see a bird and think "bird" but Eldest Son notices details of plumage because he likes to identify the kinds--so he registers things I don't.

I vaguely remembered an experiment in which people were asked to track the scent of some target object dragged across a field, and were remarkably successful (though they had to go on hands and knees to do it).

So I googled for it, and as usual other things turned up. It turns out we do seem to have a good sense of smell after all, and Shepherd beat me to the speculation that we need a vocabulary: "In the enlarged processing capacity for perceiving and discriminating odors, language plays a critical role. This seems paradoxical, for we have great difficulty describing a smell in words."

I wonder how you'd go about such a thing. I guess the first step is to try to collect the work perfumers and wine tasters have done along the way and try to test some "standard scents" (concocting brand new names with minimal associations), and see how far one can get with combinations and how far one has to expand the set of "standard scents". If the set numbers more than a few hundred, only children are likely to be able to learn the new language. Which also might be interesting to look into--can children easily learn scent language?

UPDATE: 8-July-2013: See the New Yorker article about wine tasters. There's a lot of expectation in the evaluation, but we can do fairly well...