Colors, wavelengths, brains

Mar 04, 2004 22:15

Why do red and blue mix to make purple? When you mix two materials (of the same type), one of which reflects red and one of which reflects yellow, the resulting material reflects orange, a middling wavelength. Why is it that when you mix red and blue materials, the resulting material reflects something we view as purple/violet/thereabouts, rather ( Read more... )

brain, colors, cogsci

Leave a comment

amoken March 5 2004, 08:11:49 UTC
Actually, after that study came out there were like a billion incensed anthropologists and a famous counter-study by Berlin and Kay, among several others that cast serious doubt on that conclusion. It does not show that people can't tell them apart; it shows that people choose to categorize among the category names they've learned.

It has been shown that people can tell them apart exactly as easily when told not to categorize, but to say which two of three color samples are closer in color. Everyone gives the same answer; even when they might call one blue and the other two green, they would still group a closer blue and green together.

In a survey of color terms, it has been shown that if a culture has a certain number of color terms, they will be for certain colors or groups of colors. Consulting a CogAnthro book, here are the color terms a culture has, given the number of terms they have for colors ("black" and "white" basically refer to dark and light colors until you have enough color names to limit them):

2: black, white
3: black, white, red
4: black, white, red, {yellow or green}
5: black, white, red, yellow, green
6: black, white, red, yellow, green, blue
7: black, white, red, yellow, green, blue, brown
>7: black, white, red, yellow, green, blue, brown, purple, pink, orange, grey
And after this point you get into color names that not everyone in the culture knows or agrees upon the precise boundaries.

Reply


Leave a comment

Up