Friday, October 15, 2004

12 noon

Redwood Neuroscience Institute

 

Title "Vector Symbolic Architectures answer Jackendoff's challenges for cognitive neuroscience."

 

Ross Gayler

Melbourne, Australia

 

Abstract:

Vector Symbolic Architectures (Gayler, 1998; Kanerva, 1997; Plate, 1994) are a little-known class of connectionist models that can directly implement functions usually taken to form the kernel of symbolic processing.  They are an enhancement of tensor product variable binding networks (Smolensky, 1990).

 

Like tensor product networks, VSA's can create and manipulate recursively-structured representations in a natural and direct connectionist fashion without requiring lengthy training.  However, unlike tensor product networks, VSA's afford a practical basis for implementations because they require only fixed dimension vector representations.  The fact that VSA's relate directly, without training, to both simple, practical vector implementations and core symbolic processing functionality suggests that they would provide a fruitful connectionist basis for the implementation of cognitive functionality.

 

Ray Jackendoff (2002) posed four challenges that linguistic combinatoriality and rules of language present to theories of brain function.  These challenges are: the massiveness of the binding problem, the problem of dealing with multiple instances, the problem of variables, and the compatibility of representations in working memory and long-term memory. The essence of these problems is the question of how to neurally instantiate the rapid construction and transformation of the compositional structures that are typically taken to be the domain of symbolic processing.

 

Jackendoff contended that these challenges had not been widely recognised in the cognitive neuroscience community and that the dialogue between linguistic theory and neural network modelling would be relatively unproductive until the challenges were answered by some technical innovation in connectionist models.  Jerome Feldman (2002) broadcast these challenges to the neural network modelling community via the Connectionists Mailing List.  The few responses he received were unable to convince Feldman that any standard connectionist techniques would meet Jackendoff's criteria.

 

I argue that Vector Symbolic Architectures are able to meet Jackendoff's challenges.

 

References

 

Feldman, J. (2002).  Neural binding.  Posted to Connectionists Mailing List, 5th August, 2002. (http://www-2.cs.cmu.edu/afs/cs.cmu.edu/project/connect/connect-archives/arc

h.2002-08.gz 0005.txt see also 8, 9, 18, and 21)

 

Gayler, R. W. (1998).  Multiplicative binding, representation operators, and analogy.  In K. Holyoak, D. Gentner & B. Kokinov (Eds.), Advances in analogy

research: Integration of theory and data from the cognitive, computational, and neural sciences (p. 405).  Sofia, Bulgaria: New Bulgarian University. (http://cogprints.ecs.soton.ac.uk/archive/00000502/ (see also 500 and 501)

 

Jackendoff, R. (2002).  Foundations of language: Brain, meaning, grammar, evolution. Oxford: Oxford University Press.

 

Kanerva, P. (1997).  Fully distributed representation.  In Proceedings Real World Computing Symposium (RWC'97, Tokyo).  Report TR-96001 (pp. 358-365). Tsukuba-city, Japan: Real World Computing Partnership. (http://www.rni.org/kanerva/rwc97.ps.gz see also

http://www.rni.org/kanerva/pubs.html)

 

Plate, T. A. (1994).  Distributed representations and nested compositional structure.  Ph.D. thesis, Department of Computer Science, University of Toronto. (http://pws.prserv.net/tap/papers/plate.thesis.ps.gz see also

http://pws.prserv.net/tap/)

 

Smolensky, P. (1990). Tensor product variable binding and the representation of symbolic structures in connectionist systems. Artificial Intelligence, 46, 159-216.