Abstract
Mounting evidence implicates sleep in the consolidation of learnt memories. It has to be shown if sleep helps consolidate memory for face identity, a form of memory critical for appropriate social interactions in humans. We investigated how sleep facilitates the consolidation of face recognition memory and whether it can selectively consolidate memories of behaviorally significant stimuli. In the presentation phase, observers passively viewed faces on green or blue backgrounds, which were respectively categorized as highly significant (S) or not (nS). Remembering correctly the S (nS) faces on the test gained the observer 20 (1) points. In the test phase, observers saw old and new faces randomly mixed on a gray background and judged if they remembered seeing each before. Sixty-four participants were classified, according to the times of presentation and test, into four equal-sized groups—i) PM-AM, ii) AM-PM, iii) PM-PM (24 hours), and iv) AM-AM (24 hours). Memory strength (d') was evaluated for each group. We tested the following explanatory factors for our results: sleep occurrence, occurrence of sleep after presentation, occurrence of sleep before test, duration between presentation and test, occurrence of visual stimulation, occurrence of visual stimulation after presentation, and occurrence of visual stimulation before test. We found that the occurrence of sleep before test was the key factor in explaining the facilitatory effect of sleep on memory for face recognition: PM-AM and AM-AM groups had similar d's that were significantly higher than the d's of AM-PM and PM-PM groups, which were similar to each other's. We also found that assigning behavioral significance to certain faces did not improve their recognition. Our findings thus suggest that sleep does not consolidate one's memory for faces but rather sleep prior to test shelters the memory for faces from interference (exposure), and exposure to faces while awake might disrupt it.