Representationalist theories of consciousness, such as that advocated by Michael Tye, identify conscious experiences with a subset of the mental representations involved in our cognitive processing, and they provide a satisfying account of many otherwise puzzling aspects of conscious experience. However, current versions of representationalism remain unsatisfactory in that they fail to provide a convincing solution to the "hard problem." (Tye, in fact, holds that the "hard problem" is a pseudo-problem, and thus needs no solution, but his arguments on this point are unconvincing.)
I argue that this failure of current versions of representationalism arises from the assumption that mental representations, and in particular those representations that are conscious experiences, are token identical with brain states. It seems possible (even probable) for all brain processes to occur "in the dark". We lack a convincing scientific account of what might enable brain states to "light up" with the glow of consciousness (or, indeed, of intentionality, which, for Tye, is a necessary condition for consciousness). To be sure, the identity assumption is widely shared by cognitive scientists and neuroscientists, but in fact the most that the neuroscientific evidence shows is that certain sorts of brain states or activities are necessary for certain sorts of conscious experience to occur, not that any brain states or activities are sufficient for consciousness, still less that any brain states are conscious states.
I will sketch an alternative view of mental representation, rooted in the sensorimotor approach to cognition of O'Regan and others. It equates mental representations (consciously experienced ones, anyway) not with brain states or processes, but with actions of the organism. To be sure, the brain plays a crucial role in orchestrating such activities, but the relevant brain activity should no more be equated with the representing or experiencing itself than should the brain activity that controls walking be equated with the act of walking itself. Walking necessarily involves the legs and their musculature as well as the neural structures and activities that control them, and the surface that walked upon. By the same token, visual experience (for example) necessarily involves the eyes and their musculature, as well as the neural structures and activities that control them, and whatever it is that is being seen. I will argue (as I have argued in detail elsewhere) that quasi-visual experience (mental imagery) is also best accounted for within this sort of theoretical framework.
It may be objected that sensorimotor theories also fail to overcome the "hard problem" because they do not give an adequate account of the qualia of experience. However, this failure is different from and, I believe, complementary to, the failure of representationalism. When the two theories are appropriately combined, I believe they provide the resources for a scientifically satisfying account of consciousness that is genuinely physicalistic and non-Cartesian.
