Saliency and trial context in embodied cues of spatial attention: Pointing to context

Recent work suggests that the pointing hand on an outstretched arm is possibly a more powerful cue than the gaze-cue, suggesting these embodied social cues are not equal. The aim of this study is to investigate differences between gaze- and pointing-cue, looking specifically at saliency, spatial proximity, and trial context. A cartoon figure was used to present four types of cues: 1) a gaze-cue, 2) a peripheral pointing cue on an outstretched arm, 3) a central pointing cue presented over the torso of the body, and 4) a flower cue matched for low-level features to the peripheral pointing cue. Validity was non-predictive. To test the impact of trial context on the impact of the cues, different cue types were presented randomly within blocks (Experiment 1) or tested in separate blocks (Experiment 2). In Experiment 3, trials were mixed within blocks but the total number of gaze- and gesture cues were balanced. The results showed that the pointing cue was much more effective in directing attention than the gaze-cue, especially when cues could not be predicted (Experiment 1 and 3). The impact of the pointing cue could not be explained by low-level salience or spatial proximity to the target. Trial context did affect the effectiveness of the gaze-and pointing cue, suggesting that spatial cues are influenced by overall context and shaped by expectations. Taken together, these results suggest that not all embodied cues operate equivalently; the pointing cue exerts a stronger influence on attentional allocation than the gaze cue.