Alberto Bardi
Walter Pitts is usually remembered as the mathematical co-author of the 1943 McCulloch-Pitts paper and as a precursor of artificial neural networks. That placement is accurate, but it is too narrow to capture the theoretical significance of his work for biological information systems. This article reconstructs Pitts' contribution as part of an unfinished program of biological computationalism: an approach in which computation becomes explanatory when formally characterizable relations are physically realized by organized biological processes. In the original McCulloch-Pitts case, the relevant realization is neural: threshold, inhibition, connectivity, recurrence, and temporal delay provide a causal architecture capable of instantiating logical relations. The article further argues that the importance of this program lies not only in the threshold neuron but in the conjunction of neural realization, temporal organization, recurrent structure, and the problem of invariance. It also distinguishes Pitts' own neural project from later extensions of McCulloch-Pitts-style computation to proteins, cellular signaling, bacteria, morphogenesis, and other non-neural systems. This broader uptake shows that Pitts' framework was not simply forgotten; rather, the conceptual unity of his biological computationalism has been comparatively underemphasized in dominant histories of cognitive science. By connecting the 1943 logical calculus, the 1947 treatment of universals, cybernetic recurrence, biological time, and subsequent non-neural developments, the article identifies both the power and the incompleteness of the Pittsian program. Its continuing relevance lies in a still-open question: under what conditions can formal computation be realized, constrained, and transformed by living organization?