Abstract
Two largely separate lines of research suggest that detection performance can be improved by cueing the location of a target either shortly before (attentional cueing) or shortly after target presentation (iconic memory). Here we attempted to directly compare these two mechanisms. In a behavioral experiment, we investigated performance on an auditorially cued discrimination task, where observers had to determine for bilateral displays (presented for 80 ms) how many of three circles on the cued target side (right or left) contained a gap. Cues were administered in four different temporal conditions (cue-target SOA's of −400, −200, 200, and 400 ms), allowing us to contrast performance on trials with pre-cues versus post-cues. Performance improved following both pre-cues and the shorter SOA post-cue, while there was a significant decrease in performance with the longer SOA post-cue, consistent with previous results on iconic storage. In order to identify common and distinct neuronal processes underlying the pre- and post-cueing conditions, we subsequently performed an fMRI experiment using the same task, but only 200 ms pre-cues and -200 ms post-cues. Initial analyses show that in comparison with a “passive” baseline, both pre-cue and post-cue trials resulted in increased activations in bilateral superior parietal cortex and frontal eye-fields. These common activations may reflect attentional control processes involved in retrieval from iconic storage as well as attentional cueing. However, irrespective of cued side, pre-cue trials yielded larger target-related activations in bilateral occipital cortex than post-cue trials. These different activations possibly reflect that attentional cueing, in contrast to iconic retrieval, enhances the receptivity of occipital cortices for sensory input.
AK is supported by a Long Term Fellowship from the Human Frontiers Science Program. JD is supported by the MRC and the Wellcome Trust