When items ‘pop into mind’: variability in temporal-context reinstatement in free-recall

It is well established that performance in free-recall is mediated by an individual’s ability to reinstate the study-context during retrieval. This notion is supported by an abundance of evidence and is reflected in prominent models of memory. Introspectively, however, we often feel that a memory just ‘pops into mind’ and its recall is not accompanied by contextual detail. Here we ask whether this introspection is honored by the cognitive system. Namely, do items one recalls vary in the extent to which their contexts are reinstated? Previous research has provided evidence that indeed recall of some items relies on only little, if any, contextual reinstatement. This evidence pertains to one aspect of context: the concurrent, static encoding context of items, as tapped by the source-memory paradigm. However, because real-life events are strongly embedded in time, it is crucial to also investigate the dynamic, temporal aspects of context. To do so, we capitalized on one of the seminal findings linking recall with temporal-context: the temporal-contiguity effect, whereby the closer two items at study, the higher the probability that they will be retrieved one after the other during test. Using the Remember/Know paradigm, we show that in free-recall, ‘Remember’ retrievals, which are supposedly accompanied by contextual reinstatement, produce a larger temporal-contiguity effect as compared to ‘Know’ retrievals. Furthermore, ‘Know’ retrievals are more likely to be followed by retrieval errors (e.g., intrusions) than ‘Remember’ retrievals. These findings provide evidence that recalled items vary in the degree to which their temporal-context is reinstated.