Electrophysiological dissociation of recency and recognition memory

Abstract

Event-related brain potentials (ERPs) were recorded during the test phases of two experiments. In experiment 1, subjects first studied two consecutively presented word lists. At test, they were presented with pairs of words, and were required to judge which word had been presented most recently. The test pairs were composed of two previously studied words, one drawn from each list, (Old+Old pairs), one previously studied and one new word (Old+New pairs), or two unstudied words (New+New pairs). At temporo-parietal electrodes, ERPs to Old+Old and Old+New pairs were both reliably more positive-going than those to New+New pairs. At electrode sites overlying prefrontal cortex, ERPs to Old+Old pairs attracting correct recency judgements were more positive, from around 300 ms onwards, than those elicited by the other classes of item, which did not differ from one another. In experiment 2, the test task was changed to one that required discrimination between Old+New items on the one hand, and Old+Old and New+New pairs on the other. ERPs to Old+Old and Old+New pairs once again differed from those to New+New pairs at temporo-parietal sites, but no differences were evident between the ERPs from frontal electrode sites. In line with the evidence from lesion studies, these findings suggest that judgements of relative recency depend upon processes, supported by the prefrontal cortex, additional to those that are necessary for recognition memory. They further suggest that these processes are activated rapidly and selectively in response to pairs of studied items when these must be discriminated on the basis of their relative recency of occurrence.