The long-term, overall goal is to understand the neural bases of mental processes. Though it is probably impossible to underestimate the difficulties that might impede progress toward this goal, if we do not explicitly try to reach it, then progress will be at best random. I have been encouraged to work on this problem because of the complimentary convergence of three disciplines. First, is the increasing incidence over the last 15 years of cognitive theories that use neuron-like elements as building blocks. These theories attempt to model psychophysical-like experiments in human pattern recognition and concept infomration. Often the neuron-like building blocks involve hypothetical properties that are as yet unknown to neuroscientists. Second is work like my own that studies the role of well-defined neural actiovity in associatively based synaptic modification. These studies are able to test microscopically the reasonableness of the hypothesized neural properties. Third is the existence of what are necessarily precisely defined theories of statistical pattern recognition produced by engineers. The mathematical groundwork their theories provide seems eminently suited to provide a rigorous bridge for evaluating cognitive theories and the discoveries of synaptic modification. Because the hypothesized rules of synaptic modification seem to distinguish among the various neural-like cognitive theories and because so little is really known about synaptic modification issues, our studies concentrate on constructing well controlled, easily interpreted experimental situations which allow the comparison of various theories of synaptic modification in a context amenable to both electrophysiological and electron microscopic analysis. The research is a continuation of such studies that identify, as quantitatively as possible, the characteristics of synaptic modification. In addition, I would like to produce theories which better harmonize the cognitive and neural experimental data. I eagerly anticipate increased interactions with experimental cognitive scientists interested in neural-like theories.
Showing the most recent 10 out of 37 publications
