The investigation of language processing following brain damage may be used to constrain models of normal language processing, We review the literature on semantic and lexical processing deficits, focusing on issues of representation of semantic knowledge and the mechanisms of lexical access, The results broadly support a componential organization of lexical knowledge-the semantic component is independent of phonological and orthographic form knowledge, and the latter are independent of each other. Furthermore, the results do not support the hypothesis that word meaning is organized into modality-specific subcomponents. Wealso discuss converging evidence from functional imaging studies in relation to neuropsychological results.The investigation of cognitive deficits in brain-damaged individuals can provide powerful constraints for theories of normal cognition. Brain damage does not lead to an undifferentiated loss of cognitive abilities but to richly structured patterns of impaired and spared performance. Thus, for example, a patient might show severe difficulties in orally producing the names ofpictures but show no comparable difficulties in writing them. Or a patient might show great difficulty in producing nouns but not verbs and adjectives. Furthermore, in most cases impaired performance deviates from normal performance in instructive ways. For example, in a picture naming task a patient might produce phonological distortions of the target response (e.g., chair~"share, chail"), suggesting the possibility that the correct lexical node has been accessed but that subsequent processes are damaged. Or the patient might produce well-formed, semantically related responses (e.g., chair~"table, not that but close, something you sit on"), suggesting that motor programming and articulatory processes are intact but that lexical selection mechanisms are malfunctioning. The analysis of the dissociation and association of symptoms and their related error patterns severely constrains possible interpretations of the locus of functional impairment in the normal language production system. The logic of inferences from impaired performance to normal cognition is straightforward: We prefer a theory of normal cognition over alternative theories if that theory can explain both normal performance and the various ways in which a system breaks down in conditions ofbrain damage. One area 5 that has benefited greatly from neuropsychological investigations is lexical processing.A basic assumption shared by almost all current theories of the lexicon is that it is composed of several distinct components. This assumption supports the expectation that brain damage could result in the selective disruption of each of the components that constitutes the lexical system. The ways in which a cognitive system can be damaged depends on how it is organized in the brain. Functionally independent lexical components can be damaged selectively only if they are neuroanatomically separable.' But, within the constraints imposed by the neuroanatomical organ...