Abstract

A near continuous stream of information reaches our eyes and the task of the visual system is to make sense of it. Visual stimuli arrive in quick succession: sometimes it is necessary to integrate stimuli over time, but other times it is necessary to segregate them. These opposing processes of integration and segregation are two forms of temporal processing, a topic which is underrepresented in the literature as compared to spatial processing. In particular, there is limited understanding about how temporal processing is influenced by selective attention. Selective attention refers to a family of mechanisms by which the limited resources of our mental architecture are diverted to preferentially process stimuli more likely to be relevant. Here, a number of empirical investigations into the effect of endogenous, covert spatial attention on temporal processing are presented. This includes use of a task in which visual stimulation is held constant and only the temporal processing goal (integration versus segregation) and the spatial location of attention are manipulated. There were strong spatial cueing effects: a benefit to performance when validly cued and a cost when invalidly cued. These attentional effects are reliable for both opposing processes of temporal integration and segregation. Furthermore, these spatial cueing effects are robust even when the cues provide no implicit temporal expectations. We used magnetoencephalography and the same task to measure changes in the brain signal associated with these effects, namely shifts in peak alpha frequency for integration as compared to segregation, as well as spatially specific modulations in this metric relative to the locus of attention. These findings of robust spatial cueing effects on temporal processing and of strategic shifts in oscillatory frequency associated with temporal processing goals and allocation of attention are discussed within a temporal windows framework and in the context of other candidate mechanisms. The empirical evidence reported here can be accounted for by the idea of a flexible adaptation of the size of temporal windows, essentially changing the sampling rate of perception in line with task demands.