Bruce Umbaugh

This work argues that even modest naturalism requires giving a pluralistic account of rational action, belief, choice, desire, et cetera. Of the features I attribute to the "Standard View" advanced by such authors as Savage, Hempel, Rawls, Ellis, and Forrest, virtually all must be rejected. Following Simon, the key feature I retain is the goal-directed character of rationality. ;Experimental psychology strongly intimates a substantial gap between actual cognitive performance and the idealized cognitive competence attributed on the Standard View. Further, because humans have limited cognitive resources, there are some belief-state transitions the Standard View requires that normal humans cannot execute. In addition, evidence from anthropology and developmental psychology suggests there is in fact cognitive diversity both "between subjects," e.g., that a U.S. college student reasons differently from a Zande, and "within subjects," e.g., that the same individual reasons differently at different developmental stages. ;The central contention of the present work is that there are some normatively appropriate forms of cognitive diversity. I depart from others who have defended pluralism, e.g., Stich and Roth, in distinguishing two issues: the diversity of cognitive methods or means and the diversity of cognitive outputs or consequences. I defend a "robust pluralism" that maintains that an adequate account of rationality must comprise both method pluralism and consequence pluralism. ;The program of the thesis necessitates the Adoptability Requirement : cognitive practices endorsed by a theory of rationality must be such that normal humans at the time in question can acquire and employ them. Contrary to the arguments of Quine, Strawson, Davidson, Dennett, Hollis, and Kekes, I show that it is neither obligatory nor desirable to attribute substantial cognitive uniformity to all humans. Further, because of the scarcity of cognitive resources, the Standard View's demand for cognitive optimality violates AReq. ;Together with descriptive evidence of diversity, these two conclusions prepare the way for a two-part defense of robust pluralism: First, I argue for the normative appropriateness of diverse cognitive methods and diverse cognitive consequences. Second, I defend the position against various challenges, including nihilism, scalar rationality, regress, and psychologism.