Two sorts of perceived items must compete for attention: items that are required for on-going tasks and items from the environment that are surprising or very conspicuous. We do not want to be hit by a bus because we are solving a little problem, nor do we want to be distracted from our concentration by every little change in our surroundings. How is the compromise accomplished?
In a recent paper (see citation) Mazaheri and others have studied this question. They used tracking of eye movements and EEG recordings to follow the choice between a target that was part of the ‘task’ and a distraction (one of: none, a distraction that was no more conspicuous than the target, and a highly conspicuous distraction). Would the eyes cascade towards the target or the distraction first? And was there a difference between the EEG events before a move towards the target compared with one towards the distraction?
Not too surprisingly, we found that people differed in their ability to concentrate and ignore the distraction. Despite this there was a clear pattern of activity. When there was distraction there appeared to be a prior disengagement from the task.
We found that an increase in pre-stimulus alpha activity over frontal-central regions was predictive of subsequent attentional capture by a salient distractor. Previous studies have found an alpha increase in a particular region to be indicative of the functional inhibition/disengagement of that region. … could reflect the disengagement of the frontal-eye fields (FEF). FEF is involved in top-down voluntary control of saccades and attention.
Also there was a decrease in the N1 wave response when the target won over the distractor for the first saccade. This is assumed to be due to a gating that favours items at the current spatial focus. An increase in N1 would imply greater task-relevant processing locked to the appearance of the stimulus in the trials where the target was looked at first.
As well as this locking effect tied to the stimulus (a fixed time after the stimulus appeared), there is also locking tied to the saccade (a fixed time before the eyes moved).
there was a central-parietal alpha burst just preceding the onset of the first saccade that was greater in amplitude for saccades to the target. … could index the transient inhibition of the prepotent response to saccade to the more salient distractor. … intraparietal sulcus contains an attentional priority map and is involved in saccadic control and is a good candidate for being the source of the inhibitory control signal seen here.
They also found evidence of a locking effect of a negative wave occurring before the eye movement.
A qualitative inspection of the saccade locked ERPs (event related potentials) suggests that this negative deflection is due to a latency shift in the slow negative drift building up to a potential preceding the saccade to a salient distractor.
It would be interesting to see how these indications varied if the task targets and the conspicuous distractors had varying importance/strength. What does it take for the top-down control to overcome bottom-up and vice versa?
Mazaheri, A., DiQuattro, N., Bengson, J., & Geng, J. (2011). Pre-Stimulus Activity Predicts the Winner of Top-Down vs. Bottom-Up Attentional Selection PLoS ONE, 6 (2) DOI: 10.1371/journal.pone.0016243