The computational metaphor for mind is still the central guiding idea in cognitive science despite many insightfuland well-founded rejections of it. There is good reason for its staying power: when we are at our cognitive best, we reasonabout our world with our concepts. But the challengers are right, I argue, in insisting that no reductive account of that capacityis forthcoming. Here I describe an externalist account that grounds representations in organism-level engagement with itsenvironment, not in its neural activity.

The current orthodoxy in cognitive science, what I describe as a commitment to deep representationalism, faces intractable problems. If we take these objections seriously, and I will argue that we should, there are two possible responses: 1. We are mistaken that representation is the locus of our cognitive capacities — we manage to be the successful cognitive agents in some other, non-representational, way; or, 2. Our representational capacities do give us critical cognitive advantages, but they are not fundamental to us qua human beings. As Andy Clark has convincingly argued, anti- representationalism, option one, is explanatorily weak. Consequently, I will argue, we need to take the second option seriously. In the first half of the paper I rehearse the problems with the current representational view and in the second half of the paper I defend and give a positive sketch of a two-systems view of cognition – a non-representational perceptual system coupled with a representational language-dependent one – and look at some consequences of the view.