Universiteit van Amsterdam

Events

Institute for Logic, Language and Computation

Please note that this newsitem has been archived, and may contain outdated information or links.

18 February 2002, First Seminar Cognitive Science Center Amsterdam (CSCA)

Date: Monday 18 February 2002
Time: 10:00
Location: Cognitive Science Center Amsterdam, Nieuwe Prinsengracht 130,
Room G.018

On February 18 the newly established Cognitive Science Center Amsterdam (CSCA) will hold its first scientific meeting, which will be hosted by ILLC.

Program

10.00-11.00 Herbert Clark (Psychology & CSLI,Stanford)
Methods of meaning, modes of thinking
11.15-12.15 Stan Rosenschein (Quindi Corp., Palo Alto)
title to be announced
14.00-15.00 John Perry (philosophy & CSLI, Stanford)
title to be announced
15.15-16.15 Reinhard Blutner (ILLC)
Nonmonotonic inferences and neural networks
16.30-17.30 Keith Stenning (Cognitive Science, Edinburgh)
Evolution, communication and exchange: integrating formal and empirical evidence in the cognitive science of human reasoning

Drinks will be served afterwards.

Abstracts

Herbert Clark (Psychology & CSLI,Stanford)
Methods of meaning, modes of thinking

Communication relies on three basic methods of signaling. In describing-as, speakers use symbols (such as words and sentences) to categorize objects for their addressees. In indicating, speakers create indexes (such as pointing) to locate objects for their addressees. And in demonstrating, speakers create iconic signs (such as quotations and iconic gestures) to get their addressees to imagine how objects appear. Most work in cognitive science has focused on describing-as, even though most forms of communication rely on a combination of the three methods. Moreover, the three methods are based on fundamentally different modes of thinking: categorical, locational, andimaginal. It is important to see how people combine the three methods of signaling as based on the three modes of thinking.

Reinhard Blutner (ILLC)
Nonmonotonic inferences and neural networks

It is well known that there exists a hard-to-bridge gap between symbolic and subsymbolic/neural processing. We intend to close the gap by viewing symbolism as a high-level language for the desciption of properties of neural networks. The main thesis is that certain activities of neural networks can be interpreted as nomonotonic inferences. In particular there is a strict correspondence between certain types and particular nonmonotonic inference systems.

Keith Stenning (Cognitive Science, Edinburgh)
Evolution, communication and exchange: integrating formal and empirical evidence in the cognitive science of human reasoning

`Evolutionary' psychologists, notably Cosmides (1989), have interpreted evidence from human conditional reasoning as showing that human economic exchange is evolutionarily more fundamental than generalised communication capacities. The first part of the talk analyses the basis of this argument as a paradigm case of failure of disciplinary integration.

The second part of the talk shows how our prexisting logical understanding of conditional reasoning is a good basis for designing a program of empirical research into the psychological processes in question, and presents results which throw a rather different light on the mental processes involved.

The third part of the talk argues that inasmuch as this laboratory evidence bears on human evolution, it suggests a diametrically opposed ordering of the evolution of human communication and exchange.

 

For more information, contact Michiel van Lambalgen ().

Please note that this newsitem has been archived, and may contain outdated information or links.