Abstract
The current thesis presents a series of experiments examining the relationship between language and colour cognition. The main focus is on the categorical perception (CP) of colour, the finding that discrimination of colours that straddle a category boundary is more accurate than within category colour discrimination. It has recently been argued that CP is not a perceptual effect, but rather a direct effect of language due to the comparison of stimulus labels (Roberson & Davidoff, 2000). The current thesis tests this account in a series of colour discrimination tasks with verbal and visual interference, a visual search task and a cross-cultural comparison of speakers of languages that make different categorical colour distinctions. The results of the colour discrimination tasks suggest that the effect of verbal interference on CP is dependent on task design. CP was found to survive verbal interference when the type of interference was not predictable, suggesting that the task was open to encoding strategies. CP did not survive when the colour target was presented with incongruent Stroop interference. An account of CP is proposed in which target name generation is necessary for CP. However, incongruent Stroop interference at test stimuli presentation did not selectively affect CP, suggesting that the process leading to CP is more complex than a simple matching-to-labels account would suggest. Target name generation may activate or reinforce a category code, which in turn facilitates cross-category discrimination. However, as evidence for CP was also found in the visual search task - suggesting that CP can be a perceptual effect - the account of CP proposed here may be specific to certain types of tasks only. The possibility that CP is a memory effect due to a shift towards prototype (Huttenlocher, Hedges & Vevea, 2000) was also considered. The findings presented here suggest that CP is dissociated from a shift towards prototype. The cross-cultural comparison of English and Owambo speakers on a triads and a visual search task allowed a further test of the role of language in colour CP. Language effects were found on both tasks, however, the differences were not all predicted by differences in colour naming. The results suggest that the representation of colour may also differ cross-culturally. Overall, the results presented here suggest that colour perception is not independent of colour language.