Abstract
When and why does the visual analysis of human movement differ from the visual analysis of other types of motion? Previous research suggests that the visual analysis of human motion does not always differ from the visual analysis of non-human animal motion (Chouchourelou et al., 2002). Instead, unusual human movements and common animal movements appear to be similarly analyzed. In this experiment, we examined some of the factors that might perceptually differentiate categories of human movement. Social relevance may ultimately differentiate human action from other motions (e.g., Brothers, 1997). Using assessments of apparent motion quality as a window into the temporal aspects of visual motion analyses, we compared the perception of apparent human motions across systematic variations in social context. Two-frame apparent motion sequences were created from digital videos of two human actors interacting with each other. In the “interaction” condition, both actors were visible but only one moved. In the “isolated” condition, the stationary actor was removed. Thus, identical displacements were shown in both conditions. Naïve observers saw either interactive or isolated picture pairs in a between-subjects design and were asked to make qualitative assessments of the smoothness of the apparent motion they experienced at each of seven interstimulus intervals (ISI). Across trials, stimulus duration was fixed at 100ms and ISI randomly varied between 0 and 600ms. Qualitative assessments were rendered with a 7-point scale in which 1 represented no apparent motion and 7 indicated perfectly smooth motion. Identical displays of human displacement produced significantly different percepts of apparent motion quality as a function of the social context of the displacement. Control studies with non-interactive and inconsistently interactive contexts support the idea that social interactions might actually shape the visual analysis of human action.