Joint focus of attention between two individuals can influence the way that observers attend, encode, and value items. Using a nonpredictive gaze cuing task we previously found that working memory (WM) was better for jointly attended (validly cued) versus invalidly cued colored squares. Here we examine whether this influence of gaze on WM is driven by observers sharing the perspective of the face cue (mental state account), or simply by increased attention to the cued location (social attention account). To manipulate perspective taking, a closed barrier obstructed the cue face's view of the memoranda, while an open barrier allowed the cue face to "see" the colors. A central cue face flanked by two identical barriers looked left or right, followed 500 ms later by colored squares for encoding which appeared equally often in the validly and invalidly cued locations. After a blank 1,000 ms maintenance interval, participants stated whether a probe color was present or not in the preceding display. When the barrier was open, WM was significantly impaired for invalidly versus validly cued items. When the barrier was closed, the effect of gaze cues on WM was abolished. In contrast, further experiments showed a significant cuing effect on the speed of simple target localization and color discrimination regardless of barrier type. These findings support the mental state account of joint attention in WM, whereby the attentional focus of another alters WM via higher level engagement with the second person perspective. A goal-specific model of perspective taking is proposed.
Gregory, S. E. A., & Jackson, M. C. (2019). Barriers block the effect of joint attention on working memory: Perspective taking matters. Journal of Experimental Psychology: Learning Memory and Cognition, 45(5), 795–806. https://doi.org/10.1037/xlm0000622