Hybrid encoding: Constraints on addressing structure
Cognitive models often make recourse to locality constraints as an important determinant of mental functioning. Such constraints are generally assumed to have their source at the hardware, or neuronal, level. However, the paper shows that certain cognitive limitations arise from the hybrid nature of cognitive architecture. In particular, the way symbols address constituent structure represented at the connectionist level limits their access to the encoded information. These limitations are expressed as the constructs of local and address domain and can be shown to provide an explanatory basis for a wide range of cognitive constraints.
Unable to display preview. Download preview PDF.
- Fodor, J.A., and McLaughlin, B. P. 1990. Connectionism and the problem of systematicity: Why Smolensky's solution doesn't work. Cognition 35: 183–204.Google Scholar
- Fodor, J.A., and Pylyshyn, Z.W. 1988. Connectionism and cognitive architecture: A critical analysis. Cognition 28: 3–71.Google Scholar
- Hendler, J. 1991. Special issue of Connectionist Science on hybrid symbolic-connectionist architectures. Forthcoming.Google Scholar
- Hinton, G.E. 1990. Mapping part-whole hierarchies into connectionist networks. Artificial Intelligence 46: 47–77.Google Scholar
- Marr, D. 1982. Vision. San Fransisco:Freeman.Google Scholar
- Pollack, J.B. 1990. Recursive distributed representations. Artificial Intelligence 46: 77–107.Google Scholar
- Slack, J.M. 1990. Unbounded Dependency: Tying Strings to Rings. In Proceedings of COLING-90, Helsinki, Finland.Google Scholar
- Slack, J.M. 1991 Hybrid Encoding as a Source of Cognitive Constraints. Forthcoming.Google Scholar
- Smolensky, P. 1990. Tensor product variable binding and the representation of symbolic structures in connectionist systems. Artificial Intelligence 46: 159–217.Google Scholar
- Touretzky, D.S. 1990. BoltzCONS: Dynamic symbol structures in a connectionist network. Artificial Intelligence 46: 5–47.Google Scholar