An article on a new study which seems to support the Sapir-Whorf hypothesis, which says in so many words that one's language determines what it is possible to think.
Whorf's idea does not seem to have had a lot of respect lately. The standard old saw about Inuit having a large number of words for "snow" (once mentioned by Whorf himself as evidence for his hypothesis) has turned out to be inaccurate, if not completely irrelevant, and certain older studies seem to have disproved the hypothesis.
This one set out to test numeric ability among the Pirahã, whose language has only two or three numbers, and seems to suggest that humans may innately recognize numbers up to three, but that beyond this numbers are not distinguished unless one's language defines them.
This seems to be making a lot of people happy, but I don't believe that this is the vindication of Whorf that some think it is. In one study of color recognition, researchers tested the ability to distinguish colors among people who spoke a language with only two color-words which roughly corresponded to "black/dark" and "white/light". It turned out the subjects could distinguish between many more colors than they had names for. Not only that, the subjects had a much easier time learning to distinguish primary colors than other hues, suggesting that in this case the categories are not linguistically determined at all, and that some categories are in some sense "natural". (Abstracts for some interesting color studies by the same anthropologist are here.)
It seems reasonable to suppose that the enthusiasm greeting this study has something to do with the fact that the hypothesis is seen by some as a basis for a strong form of relativism.
Posted by Alan Hogue at August 26, 2004 12:54 PMThe subjects had a much easier time learning to distinguish primary colors than other hues, suggesting that in this case the categories are not linguistically determined at all, and that some categories are in some sense "natural".
Surely this was never in doubt, was it? It's a biological issue - we're trichomats.
Posted by: Alan Allport at August 26, 2004 02:04 PMIt shouldn't have been, but judging by the fame of the study in question I suppose it was.
But this brings up a good distinction between the two studies. The color studies are testing something based mainly on perception (meaning based on not just the structure of sensory organs but on how the brain processes the data), whereas the new study focuses more on activities which as such require some amount of skill and therefore practice. If you have no words for numbers, you are never going to learn to count. Does that really prove much?
Posted by: Alan Hogue at August 26, 2004 02:17 PMWell, it's a commonplace of political discussion (and of Orwell's *1984* comments on language of course) that people can't think about a concept coherently if they don't have a word for it. That's why it's so useful to keep inventing new political terms. My favorite recent one, for example, is "astroturf."
But then there's the cultural-barrier phenomenon, which I wonder about. Remember the part in the Raoul Ruiz film "On Top of the Whale," where the anthropologist questions two South American Indians? It's fiction, but interesting fiction. The anthropologist holds up a variety of traditional objects, and they reply in each case, "Yama Scuma," in varying pronunciations and tones of voice. You think at first that their language consists of these two words varied only by inflection -- but it turns out the words mean, "Give me that," which happens to be their response each time the questioner holds up an object.
Posted by: Martha Bridegam at August 26, 2004 03:00 PM""The subjects had a much easier time learning to distinguish primary colors than other hues, suggesting that in this case the categories are not linguistically determined at all, and that some categories are in some sense "natural".""
"Surely this was never in doubt, was it? It's a biological issue - we're trichomats."
Yes, but...
At the time I studied this stuff, many moons ago, the _locus classicus_ for colour recognition investigation was a monograph by Berlin & Kay. They showed their subjects charts with very subtle gradations of shading and asked (among other things) 'which of these samples represents to you the purest example of _green_?' They then went to other places and asked 'which of these samples etc. is _verde_?' 'Vert'? 'Grun"? And so on.
The interesting thing was that the area of the chart where English speakers said 'this best typifies GREEN' was NOT the same as that chosen by Spanish speakers for _verde_, French speakers for _vert_, etc.
Thus, while the structure of the human eye might be trich[r]omat, there does appear to be some linguistic relativism at play too.
Posted by: Henry at August 27, 2004 03:40 AMAnd perhaps this relativism has a biological origin, with your average English, French, or Spanish person defining their typical green in relation to their local flora. What would be the typical green for desert dwelling nomad?
Paul
(Who suffers from red-green colour blindness and finds these questions more confusing than most people.)