ABSTRACT
Does language possess its own distinct set of processing mechanisms, or does it share mechanisms with other cognitive processes? Traditional symbolic models of language comprehension have assumed a set of distinct processing components within a modular system (e.g., the lexicon or grammar), each subserving a language-specific function and operating on language-specific information and representations (e.g., Fodor, 1983; Grodzinsky, 1995a, 1995b, 2000; Pinker, 1994; Pinker & Ullman, 2002). According to such models, the mechanisms responsible for language comprehension are essentially separate and distinct from the mechanisms responsible for other cognitive processes, and do not share general resources with them, Further, the operation of domain-specific processing modules is thought to be impenetrable to attention and cognitive control, as well as to sources of information outside of the symbol system of that particular module (Fodor, 1983; Pylyshyn, 1980, 1984). Such models assume a static base of linguistic knowledge-i.e., linguistic competence-which is associated with distinct neural structures and may
therefore be selectively disrupted by localized brain injury (Fodor, 1983). The apparent dissociations among various language impairments observed in patients with brain damage has been offered as evidence in favor of domainspecific models, in particular the frequently cited separation of lexicalsemantic and grammatical processing in certain aphasic patients (e.g., Pinker & Ullman, 1994; Ullman, 2001).
