ABSTRACT
Over a century of neuroanatomy and neuropsychology has taught us that the adult cerebral cortex is composed of a number of distinct cytoarchitectonic areas (e.g., Brodman, 1905). It is widely assumed that these structural differences result in areas with different computational properties, and correspondingly there is now a vast cognitive neuroscience literature attempting to localize cognitive functions to particular cortical areas (Posner, Petersen, Fox, & Raichle, 1988). Yet, the more fundamental question of
how these structural and functional specializations arise within an individual remains unresolved. Over the past decade, developmental neuroscientists have debated whether cortical specificity arises through molecular and genetic interactions, or through activity-dependent proc-esses (O’Leary, 1989; Rakic, 1988). A parallel debate has raged among cognitive scientists between those who believe that the newborn infant has “innate modules” for cognitive functions such as language acquisition (Pinker, 1994, 1997) and object knowledge (Spelke, 1994), and those who regard such specialization as an emergent product of prenatal and postnatal activity-dependent processes (e.g., Elman, Bates, Johnson, Karmiloff-Smith, Parisi, & Plunkett, 1996; Munakata, McClelland, Johnson, & Siegler, 1997). In fact, much recent evidence from developmental neuroscience appears inconsistent with the assumption of extensive prespecification, and instead supports the view that there is likely to be some degree of postnatal activity-dependent specialization in humans, especially for more cognitive functions (see Johnson, 1997a, 1997b for review).
