The Role of Working Memory in Attentional Capture

Article excerpt

Much previous research has demonstrated that visual search is typically disrupted by the presence of a unique "singleton" distractor in the search display. Here we show that attentional capture by an irrelevant color singleton during shape search critically depends on availability of working memory to the search task: When working memory is loaded in a concurrent yet unrelated verbal short-term memory task, capture increases. These findings converge with previous demonstrations that increasing working memory load results in greater distractor interference in Stroop-like tasks (de Fockert, Rees, Frith, & Lavie, 2001; Lavie, Hirst, de Fockert, & Viding, 2004), which support the hypothesis that working memory provides goal-directed control of visual selective attention allowing to minimize interference by goal-irrelevant distractors.

Coherent goal-directed behavior requires top-down control of attention, so that attention is allocated to goalrelevant stimuli rather than to goal-irrelevant distractors. Executive cognitive control functions that are typically associated with the frontal lobe, such as working memory, have long been thought to play a major role in such goaldirected control of attention (Baddeley, 1996; Desimone & Duncan, 1995). However, although neuropsychological studies have generally implicated the frontal lobe in selective attention, behavioral evidence for a causal role for working memory in goal-directed control of visual selective attention has been rather scarce.1

In a series of studies, Logan (1978) failed to find any effect of verbal working memory load on efficiency of search as expressed in the slopes of search set size functions. Woodman, Vogel, and Luck (2001) failed to find any effect of visual working memory load on the efficiency of search performance. More recently, Woodman and Luck (2004) found an effect of spatial memory load on efficiency of search. This, however, may be attributed to the shared content between the tasks used (the search task required to specify the location of a target attribute) rather than to higher level, content-independent cognitive control of visual search by working memory.

More positive suggestions of a cognitive control role come from studies showing that individual differences in working memory span correlate with performance in Stroop-like paradigms. For example, Kane and Engle (2003) showed that low-span subjects make more erroneous responses to a distracting incongruent word in the Stroop task than high-span subjects, suggesting involvement of working memory capacity in control of distractor responses. However, because these findings are correlative, they cannot inform about any causal role of working memory in attention. It is possible, for example, that better selective attention ability allows for better performance in working memory tasks.

Working memory was found to play a causal role in determining negative priming (NP) by distractors that are subsequently presented as targets (Engle, Conway, Tuholski, & Shisler, 1995). Engle et al. presented a word after every pair of prime-probe trials and found that NP from prime distractors was reduced with more than one word to remember. It is not clear, however, whether high working memory load reduced inhibition of distractors (an effect on selective attention) or reduced their encoding into memory and the retrieval of previous episodes (both effects on long-term memory).

The first clear evidence for a causal role of working memory in control of interference by irrelevant visual distractors has been provided in a series of studies by Lavie and colleagues (de Fockert, Rees, Frith, & Lavie, 2001; Lavie, 2000; Lavie, Hirst, de Fockert, & Viding, 2004). Lavie and colleagues have suggested that working memory serves to actively maintain processing priorities, specifying which stimuli are currently task relevant and which are irrelevant. From this claim, it follows that reducing availability of working memory for a selective attention task (by loading working memory in a concurrent, yet unrelated task) should result in reduced efficiency of focusing attention on the relevant stimuli, with greater interference by irrelevant distractors. …