Memory consolidation for duration

Abstract

Humans and animals encode and store in memory elapsed intervals as evidenced through their temporal expectancies. However, there are very few experimental studies on long-term memory of duration. The aim of this original study was to examine the consolidation process for duration and its effect on time judgement. In our study, memory of duration was tested in humans with a temporal generalization task. Consolidation was assessed by means of a 15-min nontemporal interference task introduced at different delays after the initial learning of a 4-s standard duration. The results showed that (a) when tested 24 hours after the learning phase, memory of the 4-s duration was disrupted (less precision and lengthening effect) if the interference task was introduced 30-45 min after learning; (b) no disruption was observed when memory was tested immediately after the interference task; and (c) there was a temporal gradient of the disruptive interference effect within the first hour after learning. Overall, these results fulfil the key criteria for the inference of a synaptic/cellular consolidation process and thus demonstrate that, as is the case for other memories, memory of duration undergoes a consolidation process that lasts at least one hour.

Keywords:

• Time
• Timing
• Memory
• Consolidation

This manuscript was supported by a French National Research Agency (ANR) grant [grant number ANR-11-EMCO-01201] and by the “Emotion & Time” Partner University Fund (PUF) project between France and United-States given to Valérie Doyère and Regina Sullivan, in which S. Droit-Volet is a co-PI (principal investigator). This manuscript was also supported by Capes-Cofecub (Brazil–France) cooperation program to J. L. O. Bueno and S. Droit-Volet, a grant from National Council of Scientific and Technological Development–Brazil (CNPq-), and a Doctoral Scholarship from Capes-Br for R. Cocenas-Silva (PhD).

Notes

¹ One participant in the “immediate interference/test at 24 hours” group was excluded from the results because, unlike the other participants, he produced a generalization gradient that was flat.