A study of colour grouping in three languages: A test of the linguistic relativity hypothesis

Abstract

We report a cross‐cultural study of the relationship between language and colour cognition that tests the linguistic relativity hypothesis. We compared speakers of English, Russian and Setswana—languages with different numbers of ‘basic’ colour terms—on a colour grouping task in order to see whether choice of colour groups reflected the differences among the colour lexicons of the three languages. Participants sorted a representative set of 65 colours into ‘N’ groups (where N ranged from 2 through to 12) based on their perceptual similarity. We assessed the similarities and differences among the choices of the three language samples, and focused in particular on two more specific questions. First, we tested the conjecture that levels of consensus over which tiles to group together should peak when N equalled the number of basic colour terms in the language. Second, we focused on possible linguistic influences on colour grouping in the GREEN–BLUE region of colour space. Setswana uses a single term for this region (botala), whereas English uses two terms (green and blue) and Russian uses three terms (Zelenyj ‘green’, sinij ‘dark blue’ and goluboj ‘light blue’). The most striking feature of the results was the marked similarity of the groups chosen across the three language samples. In addition there were small but reliable differences in grouping associated with linguistic differences. Specifically, maximum levels of consensus occurred at lower values of N for Setswana speakers than for the other two languages, and they were more likely to group GREEN with BLUE than speakers of the other languages. However, Russian speakers were no more likely than speakers of the other languages to form separate LIGHT BLUE and DARK BLUE groups. The results are consistent with universal perceptual processes modulated at the margin by linguistic or cultural influences.