Scientific American revisits Whorf
The April/04 issue of Scientific American has an article titled Draining the Language out of Color which revisits the Sapir-Whorf hypothesis (though poor Sapir does not even get mentioned in the article). This hypothesis advances a position with respect to the relation between thought and language, and the question of which shapes which, that, in Whorf's own words, can be summarized as follows:
We cut nature up, organize it into concepts, and ascribe significances as we do largely because we are parties to an agreement to organize it in this way -- an agreement that holds throughout our speech community and is codified in the patterns of our language. The agreement is, of course, an implicit and unstated one, but its terms are absolutely obligatory; we cannot talk at all except by subscribing to the organisation and classification of data which the agreement decrees.
I am perpetually fascinated by the question. I don't particularly have a strong inclination to a position on it yet (or at least I didn't until I lived in Germany and had to deal with German and Germans 1), but I always welcome a new article on it, especially as I think we are witnessing a resurgence of interest in the topic, as evidenced by the work of Boroditsky at MIT and other colleagues.
But the article doesn't dwell too much on this newer research (with one exceptional reference to work coming out of the Max Planck Institute). It mostly concentrates on the work of Paul Kay, now emeritus professor of Linguistics at UC Berkely, who has investigated color terms in vastly diffent languages and concluded that there are some universal tendencies in the way humans discretize the (continuous) color spectrum and allocate lexical terms to describe portions of it. Kay himself denies that his work should be viewed as a refutation of the (Sapir-)Whorf hypothesis: his work applies to the very restricted domain of color perception and does not rule out that to some extent language may condition the way we think in other domains.
Kay's raw findings are interesting: Though languages differ greatly in their inventory of color terms (English has about 11 basic terms; Dani, a language from New Guinea, has two!), the way they go about carving out the spectrum shows some tendencies Kay claims are universal. Languages that have two color terms, as per Kay, tend to group what we would call black/green/blue (and other "cool") colors and oppose this group to a white/red/yellow (and other "warm") colors. When a language has a three-way distinction, the grouping tends to go as black/cool, white/light, and red/yellow/warm. These distinctions are then further refined in other languages, leading to a more dense sampling of the spectrum, but... you get the idea: no language seems to arbitrarily group all colors of the spectrum under the same rubric except for shades of mauve, to which it devotes 20 different words.
I don't know how uncontroversial Kay's findings are. However, while reading the article, I find bits like
[i]n any case, Kay notes, the degree to which the perceived world is man-made seems to explain the variation in the number of color words. Hunter-gatherers need fewer color words because color data rarely provide much crucially distinguishing information about a natural object or scene. Industrial societies get a bigger informational payoff from color words.
This makes me somewhat skeptical and uncomfortable because these kinds of cutsey, Rousseauesque divisions always do. I also find perverse the idea that in a hunter-gathering environment color distinctions provided by, say, the basic English color terms would seem redundant. And if we accept the assertion that degree of industrialization correlates with the number of color terms in a language, what are we to make of the fact (a fact mentioned in the article) that Russian, compared to English, has an extra color term to make a two-way distinction between darker and lighter blue, a distinction which probably predates Russia's feudal days? (Of course, whether Kay or the journalist made this assertion remains open to interpretation or sleuthing.) I grant the author of that remark that he meant industrialization in a very broad sense since I'm pretty sure most of our 11 English color terms didn't enter the lexicon in the last 300 years. But even then, I'm skeptical that we have acquired a richer color palette as we have moved out of the forests (and of course, the artificial synthesis of color terms to describe items in a J. Crew catalog does not count). I don't know whether we possess the records to investigate this further, but it might be easily refutable.
For more on the Sapir-Whorf hypothesis, including treatment of the strong and weak versions of linguistic determinism and relativity, you can check out this site.
1This is a joke. Maybe. Kinda. I told you I'm still agnostic on this issue.
We cut nature up, organize it into concepts, and ascribe significances as we do largely because we are parties to an agreement to organize it in this way -- an agreement that holds throughout our speech community and is codified in the patterns of our language. The agreement is, of course, an implicit and unstated one, but its terms are absolutely obligatory; we cannot talk at all except by subscribing to the organisation and classification of data which the agreement decrees.
I am perpetually fascinated by the question. I don't particularly have a strong inclination to a position on it yet (or at least I didn't until I lived in Germany and had to deal with German and Germans 1), but I always welcome a new article on it, especially as I think we are witnessing a resurgence of interest in the topic, as evidenced by the work of Boroditsky at MIT and other colleagues.
But the article doesn't dwell too much on this newer research (with one exceptional reference to work coming out of the Max Planck Institute). It mostly concentrates on the work of Paul Kay, now emeritus professor of Linguistics at UC Berkely, who has investigated color terms in vastly diffent languages and concluded that there are some universal tendencies in the way humans discretize the (continuous) color spectrum and allocate lexical terms to describe portions of it. Kay himself denies that his work should be viewed as a refutation of the (Sapir-)Whorf hypothesis: his work applies to the very restricted domain of color perception and does not rule out that to some extent language may condition the way we think in other domains.
Kay's raw findings are interesting: Though languages differ greatly in their inventory of color terms (English has about 11 basic terms; Dani, a language from New Guinea, has two!), the way they go about carving out the spectrum shows some tendencies Kay claims are universal. Languages that have two color terms, as per Kay, tend to group what we would call black/green/blue (and other "cool") colors and oppose this group to a white/red/yellow (and other "warm") colors. When a language has a three-way distinction, the grouping tends to go as black/cool, white/light, and red/yellow/warm. These distinctions are then further refined in other languages, leading to a more dense sampling of the spectrum, but... you get the idea: no language seems to arbitrarily group all colors of the spectrum under the same rubric except for shades of mauve, to which it devotes 20 different words.
I don't know how uncontroversial Kay's findings are. However, while reading the article, I find bits like
[i]n any case, Kay notes, the degree to which the perceived world is man-made seems to explain the variation in the number of color words. Hunter-gatherers need fewer color words because color data rarely provide much crucially distinguishing information about a natural object or scene. Industrial societies get a bigger informational payoff from color words.
This makes me somewhat skeptical and uncomfortable because these kinds of cutsey, Rousseauesque divisions always do. I also find perverse the idea that in a hunter-gathering environment color distinctions provided by, say, the basic English color terms would seem redundant. And if we accept the assertion that degree of industrialization correlates with the number of color terms in a language, what are we to make of the fact (a fact mentioned in the article) that Russian, compared to English, has an extra color term to make a two-way distinction between darker and lighter blue, a distinction which probably predates Russia's feudal days? (Of course, whether Kay or the journalist made this assertion remains open to interpretation or sleuthing.) I grant the author of that remark that he meant industrialization in a very broad sense since I'm pretty sure most of our 11 English color terms didn't enter the lexicon in the last 300 years. But even then, I'm skeptical that we have acquired a richer color palette as we have moved out of the forests (and of course, the artificial synthesis of color terms to describe items in a J. Crew catalog does not count). I don't know whether we possess the records to investigate this further, but it might be easily refutable.
For more on the Sapir-Whorf hypothesis, including treatment of the strong and weak versions of linguistic determinism and relativity, you can check out this site.
1This is a joke. Maybe. Kinda. I told you I'm still agnostic on this issue.