Academic journal article Perception and Psychophysics

The Edge Complex: Implicit Memory for Figure Assignment in Shape Perception

Academic journal article Perception and Psychophysics

The Edge Complex: Implicit Memory for Figure Assignment in Shape Perception

Article excerpt

Viewing a stepped edge is likely to prompt the perceptual assignment of one side of the edge as figure. This study demonstrates that even a single brief glance at a novel edge gives rise to an implicit memory regarding which side was seen as figure; this edge complex enters into the figure assignment process the next time the edge is encountered, both speeding same-different judgments when the figurai side is repeated and slowing these judgments when the new figurai side is identical to the former ground side (Experiments 1A and 1B). These results were obtained even when the facing direction of the repeated edge was mirror reversed (Experiment 2). This study shows that implicit measures can reveal the effects of past experience on figure assignment, following a single prior exposure to a novel shape, and supports a competitive model of figure assignment in which past experience serves as one of many figurai cues.

The assignment of figure and ground status to portions of an image is among the most seemingly effortless aspects of visual perception. Yet the consequences of this assignment are profound. A region (or a portion of a region) perceived as figure has a definite shape, meaning that its bounding edges are assigned as belonging to it. When figure assignment is determined, the polarity of the segments of the edge (e.g., which segments are convex or concave) is defined with respect to the figure. In contrast, a region (or a portion of a region) that is perceived as ground does not have a definite edge where it adjoins the figure, since the ground is treated as only the visible portion of a larger shape that extends behind the figure that partly occludes it.1

Early studies of figure assignment were based on the direct reports of observers viewing images consisting of black and white regions (e.g., Kanizsa & Gerbino, 1976; Rubin, 1915/1958). These studies identified factors-henceforth called classic configurai cues-that influence figure-ground assignment. For instance, image regions that are relatively smaller, more enclosed, and more convex are more likely to be seen as figure and, therefore, in ownership of the edge between regions (for reviews, see Hochberg, 1971; Palmer, 1999; Peterson, 2000; Pomerantz & Kubovy, 1986).

A traditional assumption has been that the classic configurai cues were used to determine which region was figure before memory traces of previously seen shapes or objects were accessed. On this view, figure assignment was immune to the effects of past experience, which had an influence only after figure assignment was complete. Before the early 1990s, very few experiments were even designed to investigate whether or not past experience could affect figure assignment, and those that did produced contradictory results (for a review, see Peterson, 1999). More recently, Peterson and her colleagues (e.g., Peterson & Gibson, 1994) have shown that past experience with particular shapes and objects does exert an influence on figure assignment. Those conclusions have not gone unchallenged, however. In what follows, we will briefly discuss the evidence pro and con the conclusion that past experience affects figure assignment. We then will present three experiments designed to resolve the debate.

Evidence That Past Experience Affects Figure Assignment

Peterson and her colleagues (see Peterson & Skow-Grant, 2003, for a review) used stimuli in which an edge bordering two regions depicted a portion of a well-known shape (e.g., a portion of a sea horse or a woman) along one side, but not along the other. They presented such stimuli to observers in two orientations: upright, in which the known object was portrayed in the orientation in which it is typically seen, and inverted, in which the known object was portrayed upside down. In a number of experiments, they consistently found that the observers were more likely to perceive the figure on the side of the edge where the known object was depicted when the stimuli were upright, rather than inverted. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.