A feedforward Connectionist Model of Cognitive Dissonance

This project aims to explore a feedforward connectionist model of
cognitive disso-nance reduction. Festinger (1957) proposed that cognitive
dissonance occurs when a person holds conflicting cognitions or engages in
behavior that conflicts with his or her cognitions. He argued that this
dissonant state leads to some intriguing, counter-intuitive phenomena of
attitude change. On the basis of recent advancements in research on
causal attribution (Coo-per and Fazio, 1984), we propose that cognitive
dissonance reduction is driven by a feedfor-ward connectionist mechanism,
as implemented by the component-cue model (Gluck & Bower, 1988) or pattern
association model (McClelland & Rumelhart, 1988). Van Overwalle (1997)
demonstrated that this feedforward connectionist model can account for
many empiri-cal findings documented in various research paradigms.
Another connectionist model was re-cently put forward by Shultz and Lepper
(1996). However, their approach has some severe shortcomings that the
present feedforward model does not have. The aim of this project is to
test some unique predictions of the feedforward model with empirical data.