Implied object direction from eye location enhances animacy ratings but not detection of chasing behavior

Abstract

Understanding how humans perceive animacy in dynamic visual stimuli is fundamental to elucidating the mechanisms underlying visual social cognition. While both object geometry and eye-like features are known to independently influence animacy impressions, their interactive effects remain insufficiently explored. This study investigates how the combination of object geometry and the location of eye-like patterns modulates the perception of animacy in non-living moving objects. In Experiment 1, we manipulated the pointing direction of triangular objects and the location of eye-like features (near the vertex, near the edge, or absent), and found that animacy impressions were enhanced when the direction of motion aligned with the gaze-implied object direction, irrespective of the object's geometrical shape. Experiment 2 examined whether this effect generalizes to an objective task, in which participants identify a target triangle chasing a green disk among distractors. Although the gaze-implied object direction did not significantly influence detection sensitivity, the spatial location of the eye pattern within the triangular shape continued to weakly modulate participants' performance. These findings suggest that cues related to eye location contribute to subjective animacy at multiple stages of cognitive processing, with their impact varying depending on task demands and perceptual context.

Keywords: Animacy; Eye location; Intention; Motion; Object shape; Wolfpack effect.

Fig. 1

(a) Stimulus conditions used in Experiment 1. The experiment manipulated two factors: three types of eye locations and the relative direction between the motion direction and the pointing direction of the triangular object. (b, c) Box plots of rating scores for (b) the impression of a live animal and (c) the impression of intentionality. The horizontal thick lines within the boxes represent the median scores for each condition, while the top and bottom edges of the boxes indicate the 75th and 25th percentiles of the rating scores, respectively. Lines above and below the graph panels indicate significant comparisons among the experimental conditions. Asterisks have been placed above these lines to denote significance (p < 0.05).

Fig. 2

(a) Example snapshots from stimulus clips for each experimental condition. (b) A′ as a function of relative orientation for each eye location. (c) B″D as a function of relative orientation for each eye location.