another. A key distinction between these studies of within-modality switches and our between-modality study is that the two tasks are typically afforded by the same stimuli in the former,
whereas in the current design the participants switch between both the task and the stimuli affording those tasks. When one switches between auditory and visual inputs, the suppression of the potentially distracting sensory inputs can putatively be achieved by a relatively learn more indiscriminate suppression of a large swath of cortex, probably involving early sensory regions. On the other hand, when both tasks are afforded by the same object (e.g. the printed words in a Stroop task), then the suppression mechanisms would need to target much more specific, feature-level representations. In a recent study, we assessed this issue by asking individuals to switch between a color and a motion task, where the two features were afforded by the same random dot field arrays (Snyder & Foxe, 2010). Consistent MG-132 molecular weight with a feature-based suppression account, we found that alpha power increased within dorsal visual regions when
motion was to be suppressed (i.e. when color was the relevant feature), whereas alpha power increased in ventral visual regions when color was irrelevant. One could certainly argue that, in the current experiment, the auditory and visual inputs to be acted upon had no natural relationship to each other. Thus, although they are presented simultaneously and compete for resources, they may be perceived as separable objects, and the Acetophenone level of competition between them would probably then be less than if the tasks were afforded by features of the same object. It may be of considerable interest, in future work, to employ audiovisual stimuli where there is a clear semantic relationship between the constituent inputs (e.g. animals and their related vocalisations; Molholm et al., 2004, 2007; Fiebelkorn et al., 2010). We observed clear behavioral mixing costs in a cued audiovisual task, but no apparent switching costs, suggesting that preparatory processes during the cue-target period allowed for the entirely successful
resolution of competition among the two task-sets. We argue that, within our design, the competing tasks are held in close states of readiness, and then ‘tipped’ in favor of one or the other of the tasks by neural biasing mechanisms. Our findings support the contention that one of these mechanisms very probably involves the distribution of alpha oscillations among relevant cortical regions. Further work is required to fully tease apart the contribution of alpha synchronisations and desynchronisations to task-set reconfigurations. This work was primarily supported by a grant from the U.S. National Science Foundation (NSF) to J.J.F. (BCS1228595). The authors thank Mr Jason Adler and Ms Sarah Walkley for help with initial data collection and analysis. Additional support for the work of J.J.F.