Wednesday, May 28, 2008

Color naming

While trying to explain the color of fuchsia to my five-year-old daughter recently, I was reminded of something I learned in a linguistics class at Harvard many years ago about how color terms in a given language are predictable based on the number of color terms that exist. Though I don't remember the source, it's likely that the work presented was that of Berlin and Kay, in Basic Color Terms: Their Universality and Evolution. According to the authors, languages that have only two color terms specify only black and white. If a third color exists, it is red, followed by green, yellow, blue, and brown.

The idea is plausible, though I haven't read the book itself, and the study does have its critics. I've known since I was a child that Taiwanese and Mandarin make no distinction between the colors of green and blue. And according to the Wikipedia article on color naming, dark and light blue are considered separate "basic colors" in Russian and Italian, much the same way that red and pink are in English.

The engineer in me finds the precision of the RGB color model quite appealing, but I can also spend hours paging through Avon catalogs looking at all the different names given to subtly varying shades of pink, purple, and brown. I can't say that "Sinful Passion" is terribly descriptive, though.

2 comments: said...

That is somehow absolutely fascinating to me.

When I read this post, I immediately accepted the theory. It confirms some long-held hunches that I've had about the development of languages.

One would think that this phenomenon must extend to categories other than colors. I'm still trying to digest all the ramifications.

Think of what this idea could do for the reverse-engineering of extinct languages. Surely someone has already thought of this.

Language Lover said...

Yes, it is fascinating. I've tried to think of categories other than colors, but can't seem to come up with anything other than flora and fauna, where one can make finer and finer distinctions. Of course, that's somewhat complicated by the fact that there are biological ways to classify things that don't necessarily parallel what people guess from appearances.

Interesting idea about applying this to reverse engineering of extinct languages. I don't know what, if anything, has been done in this area.