P3, Positive slow wave and working memory load: a study on the functional correlates of slow wave activity

Parietal positivities of the `slow wave' type are known to emerge after the P300 whenever target detection leads to a complex subsidiary task. Although the functional correlates of these `positive slow waves' (PSW) are not known, it has been suggested that they may index (a) the selection or decision processes, (b) the preparation of the response or (c) the evaluation of its correctness. We investigated whether PSW could be dissociated from each of these putative steps of information processing by means of a paradigm devoid of motor components and needing very long reaction times. In our protocol, target stimuli acted as the triggering signal to perform silently one of 4 different tasks, namely (a) simple updating of a target count; (b) counting backward in threes; (c) simultaneous updating of two items (day of the week and ordinal of the month) and (d) updating of 3 items (the two above plus the month of the year). Reaction times to the same stimuli were obtained in 5 subjects during separate sessions. The different tasks did not modify the latencies of N2 or P3b components, but attenuated the amplitude of P3 as a mirror image of the subjective difficulty scores. A conspicuous parietal PSW appeared in conditions where two or 3 items had to be updated. This PSW developed 1–2 s earlier than the reaction times to the same experiments and could be therefore dissociated from the selection and decision processes. PSW latency was correlated with the number of items to be updated, but not with subjective difficulty. In the present paradigm PSW appeared to index the retrieval of information from working memory; however, in more general terms our results suggest that PSW is a non-specific activity that signals the completion of any synchronized operation immediately following target detection. Our data suggest a functional link between P3 and PSW, also supported by the similarity of their respective scalp topographies. The present paradigm proved to be easy to implement and suitable to study the `executive' functions governing attentional and working-memory control during the performance of multiple tasks.