2 ms·
> There is an excellent Radiolab podcast on "Colors" [1] which has (for me) the fairly stunning revelation that you can't see a color you can't name. As I unde
by gwern 13y ago
> There is an excellent Radiolab podcast on "Colors" [1] which has (for me) the fairly stunning revelation that you can't see a color you can't name.
As I understand, this has been tested and found wanting: you can ask people to take a group of things colored aoi / blue-green, when they have words for neither blue nor green, and ask them to come up with some additional categorization. Instead of categorizing at random or something, they form two distinct groups of... blue and green. And you can predict what colors a language will have based on the number of words it allocates to color (eg. if there's only two color words, they're always going to be black and white https://en.wikipedia.org/wiki/Linguistic_relativity_and_the_color_naming_debate#Revisions_of_the_Berlin_.26_Kay_Hypothesis https://en.wikipedia.org/wiki/Linguistic_relativity_and_the_... )