Collaborative Research:Universals and Variation in Cross-language Color Naming

University Of Chicago, Chicago IL

Investigators

Abstract

Does our language determine the way we see color or does our perception of color determine the way we talk about it? This question constitutes a major research focus of the broader 'linguistic relativity' question of the influence of language on perception and of perception on language. It has been established by prior research that the color naming systems of different languages (a) are not all alike but (b) differ from one another in a constrained fashion. In particular, there appear to be a few universally privileged or salient percepts -- ideal black, white, red, yellow, green, blue, and perhaps purple -- around which almost all the major color terms of almost all the world's languages are formed. But the boundaries of the color terms of different languages can differ dramatically. For example, the color words of most languages of the world do not draw a boundary between green and blue; that is, green and blue are included in the reference of a single word. With National Science Foundation support, Dr. Paul Kay and Dr. Terry Regier will investigate the degree to which the boundaries of named color categories can be predicted from the locations of the privileged or 'focal' colors, using existing models of how categories are generalized from given examples. The research involves (1) discovering the universal focal colors by analyzing the results of the World Color Survey, in which the reference and best examples of the major color words were collected in the field for 110 unwritten languages, (2) assessing the boundaries of the named color categories in the same sample of languages from the same data base and (3) applying three different general models of category formation from the cognitive psychology literature to see if the data in (2) can be predicted from the data in (1). To the degree this effort is successful, it will both validate the hypothesis of universal focal colors and demonstrate the interaction of domain specific constants (the focal colors) with general cognitive principles (of category formation). It will also show that variable cognitive behavior need not be lawless, but may result from the subtle interaction of a variety of constraints.