Cite as:
Ozgen E, Davies I, 1998, "Categorical effects in colour perception: cross-language differences and category learning" Perception 27 ECVP Abstract Supplement

Categorical effects in colour perception: cross-language differences and category learning

E Ozgen, I Davies

The hallmarks of categorical perception are enhanced cross-category discriminability and reduced within-category discriminability. We have previously shown, using a successive same - different task, that speakers of English show categorical effects across the blue - green boundary, whereas speakers of languages with a generic blue - green (grue) category show no such effect (Davies et al, 1996 Perception 25 Supplement, 102). We explored the nature of the categorical effect further by using both a successive and a simultaneous same - different task. We assume that the perceptual component of the task is highest for simultaneous presentation, whereas the memory load is highest for successive presentation. English speakers showed a categorical effect under both conditions, whereas speakers of languages with a grue term showed no such effect in either condition. In experiment 2 we tried to induce a category effect within a pre-existing category (either blue or green). Subjects were trained on four successive days to categorise the stimuli into two adjacent non-overlapping sets in colour space based on either hue differences or lightness differences. On the fifth day we sought for categorical effects using the same - different task and for transfer of learning to a new stimulus set. Subjects showed more accurate judgments to the dimension they were trained on (hue or lightness) than to the other dimension, and they showed enhanced cross-category discrimination. Learning did not transfer to new stimuli. Colour perception in adults can be changed through training, and category effects can be induced within well-established categories.

These web-based abstracts are provided for ease of seaching and access, but certain aspects (such as as mathematics) may not appear in their optimum form. For the final published version of this abstract, please see
ECVP 1998 Abstract Supplement (complete) size: 2107 Kb