Abstract
<p>Color science has spent most of its history explaining one thing: the conscious sensation of hue, saturation, and brightness. Yet beside that success sits a growing list of inconvenient findings. Examples include ambiguous colors (one photograph seen as white-and-gold by some and blue-and-black by others); memory colors (a gray banana that looks faintly yellow); color before form (newly sighted patients who see color before objects); imagined or dreamed colors (chromatic experiences generated without light or eyes); aphantasia (people who cannot summon a color in imagination); synesthesia (a printed letter flooded with color); and large, stable individual differences in the colors people report. These are not marginal curiosities or isolated anomalies. They are the ordinary output of a system that the percept-centered framework leaves out. A chromatic stimulus does not simply produce a color percept; it initiates a broad physiological and psychological event, of which the percept is only one part. That percept is constructed at the moment of seeing by memory, language, attention, affect, and culture, and it shapes them in turn, a reciprocal determinism of percept and event. I develop this framework at three levels, as two nested theories: a new theory of color contained within a constructivist model of the entire chromatic event. The account retains trichromatic and some opponent-process theory rather than replacing them, but extends them by predicting phenomena the traditional framework treats as noise. Chief among these are the individual differences that, on this view, become the primary evidence for how color is constructed.</p>