Color Naming

Universality of Color Names

Color names in different languages are remarkably consistent. It has been proposed that 11 basic color terms are strikingly similar across different languages and cultures. The English words for these basic color terms are black, white, red, yellow, green, blue, gray, orange, brown, pink, and purple. These color names may be grouped and ordered to approximately reflect their appearance during child development and cultural development [Page 265]as follows: (black, white) → (red) → (yellow, green) → (blue) → (gray, orange, brown, pink, and purple). An analysis of data from 110 languages of nonindustrialized societies gathered in the World Color Survey confirmed the idea of universal color names. Although there are some small variations of the focal colors (those colors that provide the best example of a basic color term) across languages, the focal color fell on average close to the colors found for English speakers. Despite these similarities, different languages may differ in the number of basic color terms. For example, some languages such as Russian, Polish, and Italian have a basic color term to distinguish between “light blue” and “dark blue”; some languages merge green and blue into a single “grue” category, and members of the Dani of New Guinea have only two basic color terms to distinguish between light-warm and dark-cool colors. Interestingly, when asked to remember colors, the Dani were better across the same color boundaries (e.g., blue-green) than English observers were, even though their language does not have any corresponding color terms. Another New Guinea language, Berinmo, deviates from the universal pattern: Berinmo has a “nol”/“wor” category that goes right across the green category of English speakers. Overall, the issue of the universality of color names and categories is still controversial.

Psychophysical Experiments

Color naming used as a method in psychophysical experiments gives a direct report of the color sensation evoked by the stimulus. Color naming seems to be an inherently coarse method because there are only a few color names but millions of distinguishable colors. One way to overcome this limitation is to use a technique called hue scaling, where observers rate the amount of red, green, blue, and yellow in the stimulus. The shapes of the hue scaling curves have been found to be highly similar across observers, and only the amplitude differed. Further, observers almost never rated a color as being both red and green, or both yellow and blue, in agreement with the notion that red-green and blue-yellow are opponent colors. Another way to overcome the limitations of color naming is to let observers name a large number of color samples and to infer the partitioning of color space based on the naming results of all color samples arranged on the color plane according to the chromaticities of the color samples. This method has been used to investigate color constancy, showing that observers can achieve almost perfect color constancy (the ability to see objects in the same color, independent of the illumination).

...