Is cognitive neuropsychology plausible? The perils of sitting on a one-legged stool.

We distinguish between strong and weak cognitive neuropsychology, with the former attempting to provide direct insights into the nature of information processing and the latter having the more modest goal of providing constraints on such theories. We argue that strong cognitive neuropsychology, although possible, is unlikely to succeed and that researchers will fare better by combining behavioral, computational, and neural investigations. Arguments offered by Caramazza (1992) in defense of strong neuropsychology are analyzed, and examples are offered to illustrate the power of alternative points of view.