We have no known conflict of interest to disclose. We thank Yichen Wang, Yuhang Li and Yajing Wang for experimental assistance.

In contrast to background white noise, the detrimental effects of background speech on verbal working memory (WM) were often explained by speech interference in the same verbal modality. Yet, those results were confounded with potential differences between arousal levels induced by speech and white noise. To address the role of arousal, in the present study, we minimized the verbal interference by using a visual WM task to test the influence of background speech or white noise. Electrodermal activity (EDA) was recorded simultaneously to indicate participants' arousal levels. Results showed that both background speech and white noise significantly improved visual WM performance, which further correlated with individuals' changes of EDA signals. Taken together, our results suggest that background sounds of both speech and white noise facilitate visual WM by raising the arousal level. K E Y W O R D S arousal, background speech, background white noise, retro-cues, visual working memory | INTRODUCTIONPlaying background sound is a preferred working mode for many individuals. However, empirical studies suggest that such background sound influences human performance, either in a facilitating or a deficient way

Phonological loop and visuospatial sketchpad have been proposed as two independent systems in working memory (WM). Previous studies suggest that background speech but not white noise declines performance of verbal WM.However, whether the background sounds influence visual WM remains unclear. In the present study, participants performed an orientation reproduction task while we played background speech or white noise.Electrodermal activity (EDA) and Electromyography (EMG) were recorded simultaneously. Results indicated that both background speech and white noise significantly improved visual WM performance, and such behavioral enhancement was significantly correlated with changes in physiological signals linked with arousal or emotion expression. Taken together, our results suggest that the phonological loop and visuospatial sketchpad is not fully independent and background speech and white noise facilitate visual WM through physiological changes in arousal and emotion expression.

No abstract