This site contains materials related to debates in Cognitive Science. It is intended to support the preparation of a book directed at the general public, and also as a resource for other scholars. Contributions are welcome and encouraged, especially from those who took part in or witnessed these debates. These contributions might be archival materials, or observations – I’d be particularly happy to have them in a public-facing form that I can use on this website or in the book. I can be reached at pater@umass.edu, and my own research website is here.
Seidenberg and Plaut (2014: 1217):
For nativists, the Pinker and Prince critique of Rumelhart and McClelland was a replay of Chomsky’s (1959) critique of Skinner. From this perspective, both episodes involved turning back attempts to reduce language to “associations.”
On the PDP side, the past tense debate was a replay of a different event, Minsky and Papert’s (1969) critique of perceptrons. From this perspective, both episodes involved authoritative but ultimately mistaken overstatements about the limitations of neural networks based on analyses of specific cases.
Minsky and Papert (p. xv 1988 ed. of Perceptrons):
The movement of research interest between the poles of connectionist learning and symbolic reasoning may provide a fascinating subject for the sociology of science, but workers in those fields must understand that these poles are artificial simplifications. It can be most revealing to study neural nets in their purest forms, or to do the same with elegant theories about formal reasoning. Such isolated studies often help in the disentangling of different types of mechanisms, insights and principles. But it never makes any sense to choose either of these two views as one’s only model of the mind. Both are partial and manifestly useful views of a reality of which science is still far from a comprehensive understanding.
Update June 2 2021: This project has been on indefinite hiatus for a while now. I published a paper with some of this material and theme as:
Pater, Joe. 2019. Generative linguistics and neural networks at 60: foundation, friction, and fusion. Language 95/1, pp. e41-e74. Project MUSE, doi:10.1353/lan.2019.0009.