A variety of new cognitive neuroscience shows how our ability to ignore distractions – to “perceptually filter”, in a sense – is based on a ventral attentional network, is related to working memory, and may be involved in putative inhibitory tasks.
First, a little background. In 2004, Vogel & Machizawa showed that some people may appear to have a lower working memory capacity merely because they are unable to filter distractions from their environment. The authors found a particular wave of electrical activity on the scalp – over the parietal cortex – which corresponded to subjects’ working memory capacity, and revealed exactly to what extent they were mistakenly updating their working memory with distractors.
Using a somewhat similar task, McNab & Klingberg have used fMRI to show that this probably reflects activity in posterior parietal cortex, which itself predicts both behavioral working memory span as well as activity in the globus pallidus of the basal ganglia (although not in the frontal cortex).
More recent work, described a few weeks ago, has similarly shown that activity over the parietal cortex predicts the speed with which subjects can identify a briefly presented target. The idea here is the same: subjects who can more effectively filter distractions from their environment show a particular pattern of scalp electrical activity which reflects the result of that filtering, and predicts success on cognitive tasks.
These findings are compatible with those of Zhang & Luck, who recently showed that working memory provides only a certain number of “slots” for the temporary online maintenance of information, and that those slots have equal precision. Thus, filtering distractors and redundantly updating only those items which are most relevant to the task will yield an apparent increase in the precision of working memory.
These effects can be understood with respect to a dualistic architecture of attention in the brain – as proposed by Corbetta, Shulman and colleagues, a dorsal attentional network (DAN; involving the dorsolateral prefrontal cortex and the superior parietal lobe) may be involved in top-down, controlled attention towards current goals and goal-relevant stimuli, whereas a ventral attentional network (VAN; including the right inferior frontal gyrus and the temporal parietal junction) interrupts or “circuit-breaks” the DAN.
A recent paper from the same group (Shulman et al., 2007) has gone farther to show that decreased activity in the right temporo-parietal junction (rTPJ) prior to the onset of a trial (in either of two simple target detection tasks) can predict whether a target is successfully detected on that trial.
Basically, the idea is that reduced activity in this region of the VAN reflects that there are relatively few influences in the VAN which might drive the reorienting of attention to stimuli that are outside the current focus. A similar effect was found in the right middle frontal gyrus, which is also technically a part of the VAN (although it’s job is theoretically more of a middleman role, in which it communicates goals from dlPFC to rIFG, and reorienting signals in the other direction). It is possible that they may have also detected effects in the rIFG had they not limited their search to regions which showed deactivation during the search task. This limitation seems unnecessarily restrictive, since one might expect the neural area actually accomplishing the filtering (as opposed to manifesting its effects) to show increased search related activity.
Shulman et al. note that Weissman et al. and Hester et al showed similar effects insofar as deactivation of a “default network” (brain areas that are typically active when subjects have no task) appeared to be important in target detection and response inhibition tasks, respectively. They imply that “broad monitoring of the environment” may be an important aspect of default functioning, and that such a function needs to be itself tuned down in order to function effectively in most cognitive laboratory tasks.