Abstract
Most previous studies on figure/ground have employed methods that measure figure/ground (f/g) in a global, explicit manner: most commonly, subjects are simply asked which region appears to be in front. Recently Kim and Feldman (2009) proposed a method that assesses figure/ground indirectly at isolated points along a boundary. In the motion-probe method, a small spatially circumscribed motion signal is introduced at a specific point on the boundary between two differently colored regions. The subject is then asked which color appeared to move. Because the figural region “owns” the boundary - and hence the motion - the response reflects local f/g. Our current study aims to (1) expand our understanding of local f/g by investigating circumstances under which it differs from global f/g; and (2) to investigate the scope of the motion probe method, studying where it agrees or disagrees with conventional methods. We collected motion-probe responses along the boundaries of shapes in a wide variety of configurations, with a variety of conventional f/g cues present. Some displays were globally consistent, meaning the contour was assigned an apparently uniform f/g polarity along its entire length, while others were not. For globally consistent stimuli, subject responses were qualitatively equivalent to conventional measures, indicating that motion-probe responses were consistent with subjective f/g organization. But for globally inconsistent stimuli, we confirmed, as in previous studies, that the motion probe method can pick up variations in f/g assignment along a single contour that are, in principle, impossible to assess using conventional global judgments (because they involve a single judgment per contour and thus force apparent consistency). Locally inconsistent f/g assignments are a potentially revealing phenomenon, because they reflect the operation of covert competitive processes underlying the determination of f/g; and the motion probe method is a useful tool for revealing them.