The University of Sussex

The ontological status of computational states

Ronald L. Chrisley

Both Putnam and Searle have presented arguments for the claim that computational states are universally realisable, in the sense that we could interpret any physical system as instantiating any computational characterization. They both argue that this has dire consequences for the computational view of the brain and mind that is a working hypothesis in cognitive science. But whereas Searle admits that the threat of universal realisability could be avoided if our notion of computation involved causal and counterfactual notions (implying that these are lacking at present), Putnam thinks that the universality, and hence vacuousness, of the notion of computation remains, even if one requires computational state transitions to be causal. In the following, I analyse Putnam's argument and find it inadequate, because it employs a notion of causation that is too weak. I suggest that if one were to augment formal notions of computation with constraints involving embeddedness and full causality, one would arrive at a notion that directly addresses the worry of universality. As a result, the claim that the brain is a digital computer will be one with empirical content.

This paper is not available online