Marilynn B. Brewer Ohio State University
Six years ago, the editors of Advances in Social Cognition gave me the opportunity to write the lead article for the first volume in this series. It proved to be a valuable experience, permitting me to try out some new and (what seemed then) radical ideas about two separate representational systems for encoding and storing information about individual persons. I am pleased now to have a second opportunity to contribute to this series, this time in the role of commentator on an article that is yet more radical and far more elegant than the rough model I proposed in 1988. After reading Carlston's integrative chapter (this volume), I think it is safe to say we have come a long way in the last 6 years.
As Carlston acknowledges, in some respects Associated Systems Theory (AST) is an extension and elaboration of various dual-representation models proposed by myself ( Brewer, 1988) and others (e.g., Andersen … Klatzky, 1987; Fiske … Pavelchak, 1986; Ostrom, Carpenter, Sedikides, … Li, 1993; Wyer … Martin, 1986). Although these prior models share the general idea that person information can be represented in different modes, each suggests a different basis for dichotomizing the primary representational systems. In effect, AST integrates different versions of dualrepresentation systems into a single multirepresentation framework. I hope to illustrate the value of this integrative effort by focusing on points of contact between AST and my own dual-process model -- pointing out where the two models are mutually supportive, where AST can clarify and enrich concepts introduced in the earlier model, and where differences between the two approaches might suggest extensions and modifications of the AST framework.