ADAPTER: A Conceptual Model of Category-Driven Analogical Retrieval.

Abstract

Research on analogy-making agrees that mapping allows one to find structural similarities when comparing two situations. However, whether retrieval of past events from memory is guided by surface or structural similarities remains subject to empirical debate. The current contribution is aimed at dissolving this controversy by reviewing experimental evidence showing that the determinants of analogical retrieval primarily depend on the encoding of the situations, which is itself modulated by prior categories available to the participants. Based on this review, a conceptual model is introduced (ADAPTER, As Deep As Possible Target Encoding and Retrieval), in which available categories determine the level of abstraction characterizing encoding as well as the type of retrieval that can be implemented. The model also incorporates the impact of encoding contexts and characteristics of the target descriptions on the likelihood of a relational encoding, which in turn influence the determinants of retrieval. This framework elucidates prior findings within a unified account and provides avenues for advancing the debate on the determinants of analogical retrieval by generating empirical predictions. The model also provides novel insights into the developmental trajectory of structurally based retrievals and suggests promising educational interventions aimed at promoting spontaneous transfer.