The distinction between perceptual color and the color of objects: Neuropsychological and anatomical data

Damaged knowledge of perceptual color in the presence of spared knowledge of color as a property of an object has been repeatedly documented, often in the presence of spared perception of form (Zeki, Brain 1990). By contrast, the reverse dissociation of spared perceptual color and impaired object color (Luz/atti and Davidoff, Neuropsychologia 1994), was reported on in only two patients, who also showed damage to object form. They named color probes, but could not provide the canonical color of an object, nor distinguish real and unreal objects. We describe two further subjects with damage to object color and sparing of perceptual color. In these subjects, we also evaluate the relationships between knowledge of color and of other (visual and functional) properties of objects. PCO is a 34-year-old, ambidextrous female, with bilateral damage to the frontal, temporal, and parietal lobes (from probable HVE). She suffers from severe semantic damage. IOC is a 76-year-old right-hander, with ischemic damage to the calcarine cortex, part of the fusiform and the lingual gyri, and of the infero-lateral temporal lobe. She is severely anomic and suffers from mild semantic damage. Both subjects scored flawlessly on the Farnsworth-Munsell hue task and had no difficulty naming, sorting, and recognizing color probes. However, when asked to say the name of the typical color for an object, or to pick a pencil in order to color a black-and-white picture, both demonstrated a marked and quantitatively comparable impairment (between 35% and 55% incorrect responses across tasks). For example, PCO said that strawberries are blue, and IOC selected a green crayon to color an elephant. Their behavior diverged on object decision tasks, and on questions tapping knowledge of object size and of other visual (noncolor) and functional properties of objects, as PCO was again severely impaired, whereas IOC performed flawlessly. These results: (a) confirm that object color can be damaged independently of perceptual color; (b) demonstrate that color is represented independently of other visual and nonvisual properties of objects (e.g., size, shape, function, etc.); and (c) suggest that abstract object properties (i.e., object color) are grounded in elementary, perceptually based information. Lesion localization in PCO and IOC suggest that perceptual color and object color are represented in distinct neural substrates, and that the same is true for object color as opposed to other properties of objects.