Dual recollection in episodic memory.

Abstract

In the mainstream memory literature, recollection is conceptualized as a univariate process that involves conscious reinstatement of contextual details that accompanied earlier events. That conception predominates in several domains other than basic memory research-such as cognitive development, cognitive neuroscience, dementia, and forensic interviewing. According to the dual-recollection hypothesis, however, there are 2 distinct forms of recollection: conscious reinstatement of contextual details (context recollection) and conscious reinstatement of target events per se (target recollection). We review existing lines of evidence that favor the dual-recollection hypothesis, and we describe a source-monitoring paradigm with an accompanying model that separates the 2 recollections from each other and from familiarity. Some experiments are reported whose aims were to determine how measures of target and context recollection react to a series of theoretically motivated manipulations and to assess the validity of the modeling tool that supplies those measurements. The manipulations produced a series of single and double dissociations between target recollection, context recollection, and familiarity, and subsequent state-trace analyses revealed that the 3 retrieval processes were jointly independent. Fit analyses showed that the model gave acceptable accounts of the data of all experiments, but that fit was unacceptable when either the target recollection process or the context recollection process was removed from the model.