Abstract

A series of experiments show that attribution of intentionality to figures depends on the interaction between the type of movement-Theory of Mind (ToM), Goal-Directed (GD), Random (R)-with the presence of human attributes, the way these figures are labeled, and their apparent velocity. In addition, the effect of these conditions or their interaction varies when the use of human nouns-present in the participant's responses-is statistically controlled. In Experiment 1, one group of participants observed triangular figures (n=46) and another observed humanized figures, called Stickman figures (n=38). In ToM movements, participants attributed more intentionality to triangular figures than to Stickman figures. However, in R movements, the opposite trend was observed. In Experiment 2 (n=42), triangular figures were presented as if they were people and compared to triangular figures presented in Experiment 1. Here, when the figures were labeled as people the attribution of intentionality only increased in R and GD movements, but not in ToM movements. Finally, in Experiment 3, Stickman figures (n=45) move at a higher (unnatural) speed with higher frames per second (fps) than the Stickman figures of Experiment 1. This manipulation decreased the attribution of intentionality in R and GD movements but not in ToM movements. In general terms, it was found that the human attributes and labels promote the use of human nouns in participants' responses, while a high apparent speed reduces their use. The use of human nouns was associated to intentionality scores significantly in R movements, but at a lesser extent in GD and ToM movements. We conclude that, although the type of movement is the most important cue in this sort of task, the tendency to attribute intentionality to figures is affected by the interaction between perceptual and semantic cues (figure shape, label, and apparent speed).