Immediate serial recall is seriously disrupted by to-be-ignored sound. According to the embedded-processes model, auditory distractors elicit attentional orienting that draws processing resources away from the recall task. The model predicts that interference should be attenuated after repeated exposure to the auditory distractors. Previous failures to observe evidence for habituation can be explained by assuming that habituation to complex distractor features depends on the availability of working memory resources. Here we demonstrate that the irrelevant sound effect is attenuated after passive listening to the auditory distractors during a preexposure phase prior to the serial recall task. Experiment 1 shows that the irrelevant sound effect is abolished after 20 min of passive listening to the distractor speech. Experiments 2-4 show that irrelevant sound interference is significantly reduced after listening to distractors for 45 s. As predicted by the habituation hypothesis, an attenuation of interference occurs only when the distractor material matches the material played in the preexposure phase (Experiment 5). The results support an attentional conceptualization of the irrelevant sound effect.

No abstract

Working memory theories make opposing predictions as to whether the disruptive effect of task-irrelevant sound on serial recall should be attenuated after repeated exposure to the auditory distractors. Although evidence of habituation has emerged after a passive listening phase, previous attempts to observe habituation to to-be ignored distractors on a trial-by-trial basis have proven to be fruitless. With the present study, we suggest that habituation to auditory distractors occurs, but has often been overlooked because past attempts to measure habituation in the irrelevant-sound paradigm were not sensitive enough. In a series of four experiments, the disruptive effects of to-be-ignored speech and music relative to a quiet control condition were markedly reduced after eight repetitions, regardless of whether trials were presented in blocks (Exp. 1) or in a random order (Exp. 2). The auditory distractor's playback direction (forward, backward) had no effect (Exp. 3). The same results were obtained when the auditory distractors were only presented in a retention interval after the presentation of the to-be-remembered items (Exp. 4). This pattern is only consistent with theoretical accounts that allow for attentional processes to interfere with the maintenance of information in working memory.

In a series of experiments, it was tested whether distraction by changing-state irrelevant speech is inevitable or can be modulated by foreknowledge of an imminent to-be-ignored distractor sequence. Participants were required to remember visually presented digits while ignoring background speech. In the foreknowledge condition of Experiment 1, the upcoming to-be-ignored sentence was presented auditorily and visually before each trial. With specific foreknowledge, the changing-state irrelevant sound effect (here, increased disruption by sentences compared with repeated words) was significantly attenuated relative to a condition without foreknowledge. This finding was replicated in Experiment 2, in which the information about the upcoming auditory distractor speech was presented only in the visual modality. Experiment 3 showed that only specific foreknowledge of the auditory distractor material has beneficial effects on the ability to ignore distraction. The mere notification that an unspecified distractor sentence would be presented next had no effect on distraction. In Experiment 4, there was only a small and not statistically significant reduction of the irrelevant speech effect when lists of randomly selected words were used as distractor material, suggesting that foreknowledge effects are more pronounced for highly variable, meaningful distractor material. We conclude that the disruption of short-term memory by irrelevant speech is not purely a stimulus-driven process that is immune to top-down control. A significant proportion of the effect can be modulated by specific knowledge about an imminent distractor sequence.

It is well established that task-irrelevant, to-be-ignored speech adversely affects serial short-term memory (STM) for visually presented items compared with a quiet control condition. However, there is an ongoing debate about whether the semantic content of the speech has the capacity to capture attention and to disrupt memory performance. In the present article, we tested whether taboo words are more difficult to ignore than neutral words. Taboo words or neutral words were presented as (a) steady state sequences in which the same distractor word was repeated, (b) changing state sequences in which different distractor words were presented, and (c) auditory deviant sequences in which a single distractor word deviated from a sequence of repeated words. Experiments 1 and 2 showed that taboo words disrupted performance more than neutral words. This taboo effect did not habituate and it did not differ between individuals with high and low working memory capacity. In Experiments 3 and 4, in which only a single deviant taboo word was presented, no taboo effect was obtained. These results do not support the idea that the processing of the auditory distractors' semantic content is the result of occasional attention switches to the auditory modality. Instead, the overall pattern of results is more in line with a functional view of auditory distraction, according to which the to-be-ignored modality is routinely monitored for potentially important stimuli (e.g., self-relevant or threatening information), the detection of which draws processing resources away from the primary task. (PsycINFO Database Record