Computation as the boundary of the cognitive

Khalidi identifies cognition with Marrian computation. He further argues that Marrian levels of inquiry should be interpreted ontologically as corresponding to distinct semi-closed causal domains. But this counterintuitively places the causal domain of representations outside of cognition proper. A closer look at Khalidi's account of concepts shows that these allegedly separate Marrian domains are more tightly integrated than he allows. Theories of concepts converge on algorithmic-representational models rather than computational ones. This suggests that we should reject the wholesale identification of cognition with computation.