Past research on the effects of articulatory suppression on working memory for nonverbal sounds has been characterized by discrepant findings, which suggests that multiple mechanisms may be involved in the rehearsal of nonverbal sounds. In two experiments we examined the potential roles of two theoretical mechanisms of verbal working memory-articulatory rehearsal and attentional refreshing-in the maintenance of memory for short melodies. In both experiments, participants performed a same-different melody comparison task. During an 8-s retention interval, interference tasks were introduced to suppress articulatory rehearsal, attentional refreshing, or both. In Experiment 1, only the conditions that featured articulatory suppression resulted in worse memory performance than in a control condition, and the suppression of both attentional refreshing and articulatory rehearsal concurrently did not impair memory more than articulatory suppression alone. Experiment 2 reproduced these findings and also confirmed that the locus of interference was articulatory and not auditory (i.e., the interference was not attributable to the sound of participants' own voices during articulatory suppression). Both experiments suggested that articulatory rehearsal played a role in the maintenance of melodies in memory, whereas attentional refreshing did not. We discuss potential theoretical implications regarding the mechanisms used for the rehearsal of nonverbal sounds in working memory.

Memory for nonverbal sounds such as those used in sonifications has been recognized as a priority for cognitive-perceptual research in the field of auditory display. Yet memory processes for nonverbal sounds are not well understood, and existing theory and research have not provided a consensus on a mechanism of memory for nonverbal sounds. We report a new analysis of a qualitative question that asked participants to report the strategy they used to retain nonverbal sounds—both melodies and sounds discriminable primarily by timbre. The question was originally posed as part of the debriefing procedure for three separate memory experiments whose primary findings are reported elsewhere. Results of this new analysis suggested that auditory memory strategies — remembering acoustic properties of sounds — were common across both types of sounds but were more commonly reported for remembering melodies. Motor strategies were also more frequently reported for remembering melodies. Both verbal labeling of sounds and associative strategies — linking the sounds to existing information in memory — were more commonly reported as strategies for remembering sounds discriminable primarily by timbre. Implications for theory and future research are discussed.

Both subvocal articulatory rehearsal and attentional refreshing have been advanced as rehearsal mechanisms for verbal material (including speech sounds) in working memory, but the mechanism of rehearsal for nonverbal sounds-especially timbral attributes of sounds-remains unclear. Since timbral attributes of sounds often have been assumed to be unvocalizable, researchers have suggested that working memory for timbre must involve an attentional or sensory (rather than subvocal) maintenance mechanism. Two experiments examined the role of subvocal rehearsal and attentional refreshing in working memory for timbre using sounds that were designed to be abstract, unvocalizable, and devoid of apparent mnemonic semantic connections. Participants heard either three (Experiment 1) or two (Experiment 2) brief (250 ms), abstract sounds. Following an 8000 ms retention interval, participants heard a single probe sound and indicated whether the probe had been a member of the original set. During the retention interval, manipulations were introduced to suppress subvocal articulatory rehearsal, attentional refreshing, or both. Across both experiments, all conditions with articulatory suppression during the retention interval showed worse memory performance than a control condition. Effects of attentional refreshing suppression were inconsistent. These experiments suggested that subvocal imitation may be an important mechanism of rehearsal in working memory for timbre-a finding that presents complications for existing theoretical perspectives on both verbal and nonverbal working memory.