Abstract
Adults are shown to be better in distinguishing a pair of colors from two adjacent
categories (e.g., a blue and a green) rather than a pair of colors from the same color category (e.g., two shades of blue) although sychophysically
these are all equi-distant. This phenomenon has been referred to as “Categorical
Perception” (CP) in literature (Harnad, 1987). The issue of whether CP is inborn or is
induced by learning has been a much debated topic (Lane, 1965 on speech perception; Goldstone, et. al., 2001 on category learning; Calder, et. al., 1996 on facial expressions, and Roberson, et. al., 2000, 2007; Franklin, et. al., 2005, 2008 on color categorization). Previous research did not reach consensus about the role of language learning in CP of color. While some researchers suggested that color term knowledge affects CP (Roberson,
et. al., 2000, 2007), others argued that it does not (Franklin, et. al., 2005, 2008).
If CP is an artifact of verbal labeling, then pre-linguistic infants should not show CP to
color. Our study examined categorical responding to color in 9-month old infants by recording eye movements on a target detection task. Our findings provide support for the existence of an innate basis for CP or at least for a pre-linguistic bias in color categorization. Further studies are needed to discover what exactly infant color categories are, how they interact with language, and if language may
introduce more categories or weaken some of the existing pre-linguistic categories if they do not exist in language.
