Agent-based Models of Language Acquisition and Evolution
John Holland
Professor of Psychology; Computer Science and Engineering
The University of Michigan, Ann Arbor MI
jholland@umich.edu
Thursday, October 30th, 2008
White 411: 11:30 – 12:30 p.m.
Abstract
Can we exhibit a model of language acquisition and evolution that does not require its agents to have a 'wired in' grammar? This presentation proposes an exploratory agent-based, signal-processing model that suggests an affirmative answer. It is a social model depending on interacting agents situated in an environment that exhibits perpetual novelty. The agents have only primitive cognitive capacities, but they can learn. Their survival, and hence the survival of any proto-linguistic communication, depends upon their ability to collect resources from the environment.
Biographical Sketch
Dr. Holland's main research interests are genetic algorithms, complex adaptive systems (natural and artificial), computer-based models of cognitive processes, and the construction of models for computer-based thought experiments. 49 students have completed doctorates under his direction or co-direction. He is a member of the Board of Trustees of the Santa Fe Institute and a board member of the International Society for Genetic and Evolutionary Computation. He has been named a MacArthur Fellow and is a Fellow of the World Economic Forum. His two most recent books are Hidden Order: How Adaptation Builds Complexity and Emergence: From Chaos to Order.
Hosted by the Department of Electrical Engineering and Computer Science, School of Engineering.