Abstract
Background: A striking feature of color naming in the World Color Survey (WCS) (www.icsi.berkeley.edu/wcs) is the great diversity among the color vocabularies of informants speaking each of these unwritten, third-world languages. In this project, we examined English speakers to determine whether this prominent diversity in color naming applies to a written, Indo-European language.
Methods: We collected free color-naming data on educated native English speakers (N=23) using the same 330-color stimulus set and the same instructions as were used for the WCS. Then we followed up with color names restricted to the 11 English “Basic” Color Terms (BCTs).
Results: English-speaking subjects showed remarkable diversity in their usage of color terms. The average pair-wise similarity under restricted instructions was 0.80 ±0.04, reflecting variations in the locations of boundaries between BCTs. Average similarity under WCS instructions dropped to 0.68 ±0.06 (85th percentile of the WCS, which averaged 0.54 ±0.13), reflecting non-BCT (nBCT) usage. Subjects averaged 8 nBCTs (males averaging fewer color terms than females, pDiscussion: Our results indicate that the low concordances observed in the WCS are a general feature of world languages, not restricted to unwritten languages spoken far from the influence of Western culture. We do not know why pink is a BCT, used by all speakers of English, whereas peach and lavender are not, or why no language (in the WCS or elsewhere) has a one-word term for *pale-green.
This research was supported by the Ohio Lions Eye Research Foundation.