Recent scientific studies of consciousness reveal the challenges involved in striking the proper balance between concrete fact and abstract theory. Christof Koch believes he is on the road to a scientific understanding of consciousness because he has a research paradigm which appears to create solvable puzzles. Unfortunately, work in other areas of neuroscience reveal that these puzzles rely too heavily on Aristotelian common sense to account for the unique nature of connectionist processing. William Ramsey claims that biological versions of traditional theories of representation, of the sort embraced by Koch, Fodor, and others, are so inadequate that we should give up all attempts to create high level abstract theories about human cognition. I argue, however, that what is needed is more abstract theorizing, not less. The intelligible entities described by Dynamic Systems Theory should be seen as embodiments of mental representations. These embodiments are physical in the sense that they are comprehensible in terms of modern physics, even if they are not material items that can be directly perceived.