How Retrieval Shapes Episodic Memories

During his PhD years, Alf Zimmer worked on fundamental research in episodic memory, cumulating in his dissertation on the creation of order in human memory (Zimmer, 1973). Since then he kept up this interest in the field, although his focus shifted a bit from more fundamental to more applied issues in memory research. In particular, Alf Zimmer became interested in research on eyewitness testimony and, for instance, the question of how witnesses should be interrogated to extract as much information as possible from their memory representation of an observed incident. This is an important issue in applied memory research and so, like Alf Zimmer, a number of researchers got interested in this question in the 70-ies and 80ies of the past century. The research led to the publication of a new technique to help police officers, and others, to interrogate witnesses efficiently. The technique was called the cognitive interview (Geiselman, Fisher, MacKinnon, & Holland, 1985). The cognitive interview is based on four general retrieval mnemonics: (i) mentally reinstating a witness’s internal and external context that occurred at the time the incident was observed, (ii) reporting every detail that comes to mind, regardless of how trivial or irrelevant it may appear, (iii) attempting to recount the incident in several different temporal orders, e.g., both forwards and backwards, (iv) attempting to report the incident from a range of different perspectives, e.g., the own perspective and that of other witnesses of the incident. Indeed, results from several studies showed that witnesses often show better recall of an observed incident when the cognitive interview was employed for interrogation than when so-called standard