A cognitive agent model for mindreading

From Master Projects
Jump to: navigation, search


has title::A cognitive agent model for mindreading
status: ongoing
Master: project within::Cognitive Science
Student name: student name::Monique Hendriks
Dates
Start start date:=2010/01/01
End end date:=2010/06/02
Supervision
Supervisor: Jan Treur
Second reader: has second reader::Zulfiqar Ali Memon
Poster: has poster::Media:A_cognitive_agent_model_for_mindreading_-_collage.pdf‎

Signature supervisor



..................................

Abstract

Humans are social beings. They can appreciate each other not only as sensory objects, but also as subjects. Social interaction is made possible through the understanding that others have mental states and that these mental states may differ from those of ourselves. This capability of assigning to others mental states that di�er from the mental states of ourselves is termed mindreading. The question of how we understand other minds is as yet unresolved. Neurolog- ical research and the discovery of mirror neurons have provided more concrete grounds on which the two dominating theories of mindreading, Simulation The- ory and Theory-Theory can be based. However, most attempts at full- edged theories of mindreading contain gaps that result in the permission of multiple interpretations of the theory. Allowing multiple interpretations makes it more diffcult to falsify a theory, and therefore, the theory becomes less robust.

This thesis provides a formalised hypothesis regarding mindreading that is based on existing theories. Formalisation of the hypothesis indicates gaps in the existing theories. These gaps are filled in by axioms that indicate my interpretation of the theories on which the hypothesis is based. These axioms provide a basis for deriving predictions from the formalised hypothesis that need to be tested in future research in order to gather support for the hypothesis or to falsify (and adjust) it. Formalisation is done in the form of a cognitive agent model. The cognitive agent model is implemented in the language LEADSTO. The LEADSTO tool is used to run simulations of the model. A case study is used to demonstrate the functionality of the model. Predictions from this case study were verified, providing evidence in support of the hypothesized model of mindreading.

The final version of my Master's thesis:

A cognitive agent model for mindreading(pdf)