Abstract
Humans are able to make inferences about objects’ potential animacy (i.e., are alive) and intentionality (i.e., have goals that they are attempting to meet) based upon reasonably simple visual movement patterns. One open question is whether humans are aware of the visual cues that they use to make such inferences, and if so, can humans employ these cues to mask their own agency and intentionality from other humans. We hypothesized that when participating in a competitive chasing task on a computer (one human chasing another), the “chaser” would avoid producing movement patterns that tend to trigger detection of intentionality by the “chasee.” In separate blocks of trials, participants (N=40 pairs) played the role of chaser (the “wolf”) and chasee (the “sheep”). The setup involved one red circle (the sheep; always controlled by a human participant) and many white circles (one of which was the wolf, the others were distractors), all moving in a large rectangular field. In some blocks, the wolf was controlled by a computer algorithm, and in other blocks, it was controlled by another human being. The wolf’s task was to catch the sheep. The sheep’s task was to avoid being caught. Consistent with previous work, when playing against a computer-controlled wolf, participants playing as the sheep often detected and avoided the wolf (in particular when the wolf was controlled by a direct-chasing or trajectory-interception algorithm). Conversely, when playing against a human-controlled wolf, participants very rarely escaped, with, as expected, the human-controlled wolves showing more complex movement patterns (and thus better disguised among the distractors) than the simple computer-controlled wolves. These results suggest that humans have some degree of awareness of the visual cues they use to make inferences about animacy and intentionality and are capable of masking those cues in their own behaviors.