ABSTRACT
We have briefly reviewed some of the main findings from studies of braindamaged patients with apparent category-specific impairments. While some cases may be explained in terms of confounding variables, this possibility cannot account for all of the behavioural data. The existence of a double dissociation, and of cases of category-specific impairment across controlled sets of stimulus materials, suggests that these deficits can reflect genuine effects of semantic category. Furthermore, although there is evidence that some patients show a disproportionate impairment with living relative to nonliving things (and vice versa), category-specific impairments do not always break down neatly in terms of a strict living/nonliving distinction. One question that can be asked is how conceptual knowledge is represented in the brain, such that damage leads to the specific patterns of impaired and preserved performance that is found in these patients.
