Perception of a Point-Light Walker Produced by Eight Lights Attached to the Back of the Walker

Abstract

The Johansson demonstration, achieved with point-lights placed on the main joints of a walker, gives a vivid impression of an invisible human being. An experiment was devised to test the degree to which gender and age of an individual could be correctly identified by observing a point-light-walker (PLW).

Eight point-lights were attached to the backs of each of eight actors, four men and four women, aged 18 to 62 years, at the shoulders, hips, knees, and heels. In a dimly lit room, the actors stood on a treadmill and walked for 1 min on an endless belt moving at speeds from 1.0 to 3.0 km h⁻¹. The movements made by the point-lights were recorded with a video camera and projected on a screen (96 cm high and 120 cm wide). The PLW demonstrations were viewed by 144 undergraduates, divided into five groups, at observation distances of 4.8 to 13.3 m. They were asked to identify and write the gender and age of the perceived gait. On average, there was a 74% correct identification of the genders, especially of the PLWs of the older actors, which were 80% or more correctly identified. Recognition of age, however, was less robust and the PLWs were mostly judged as 20 to 40 years old. Of particular interest was that the PLW demonstrations by two actors of about the same age were quite differently judged; one as old and the other as young. Comparison of these two PLWs revealed that the point-lights corresponding to the shoulders and heels moved very differently. The extents of both the horizontal distance of the right shoulder and the vertical distance of the left heel were measured at intervals of 100 ms. When the obtained values were plotted on the abscissa (shoulder) and ordinate (heel), the relation between the movements could be represented as a pattern corresponding to the italic letters 8, O, and D; pattern 8 was typical of the PLWs judged as ‘young’, pattern O as ‘old’, and pattern D as in between. These findings lead to the suggestion that the correlative changes between the movements of point-lights could become a useful aid in the studies of gender and age recognition in the PLWs.