• Chen & Scholl
• Meyerhoff & Scholl
• van Buren & Scholl

Perception, as represented by most vision science research, is the process of recovering the physical structure of the world from shifting patterns of retinal images. But actual visual experience nearly always transcends this characterization. An especially salient example involves the aesthetic qualities of perception: it is often impossible to see something without also liking or disliking it. It may be possible to explain some (perhaps small) percentage of such experiences in terms of underlying regularities of visual processing. Even when studied in this way, however, aesthetic experience is often treated as being a later, independent aspect of perception. Here, in contrast, we explore how aesthetic preferences may interact with other types of visual processing. We were inspired by the inward bias in aesthetic preferences: when an object with a salient "front" is placed near the border of a frame (say, in a photograph), observers find the image more aesthetically pleasing if the object faces inward (toward the center) vs. outward (away from the center). We employed framed stimuli that were ambiguous in terms of the direction they appeared to be facing. For example, an equilateral triangle can be seen as pointing in the direction of any of its three vertices. Our observers' percepts were influenced by the frames in a way that corresponded to the inward bias: when a triangle was placed near a frame's border, observers tended to see whichever interpretation was facing inward. The same observers also judged an unambiguous version of the figure -- an otherwise matched "drop" figure -- as more aesthetically pleasing when it pointed inward. This match between the inward bias in aesthetic perception and ambiguous-figure perception suggests new ways in which aesthetic factors may relate not only to what we like, but also to what we see in the first place.

When two discs move toward each other, superimpose, and continue moving afterwards, observers typically perceive them as streaming past each other. If a brief tone occurs at the moment of overlap, however, then the discs are perceived as bouncing off each other. Recent research has attributed this effect to decisional (rather than perceptual) processes by showing that auditory tones alter response biases but not the underlying sensitivity for detecting objective bounces. Here we explore the nature of this phenomenon using 'illusory causal crescents': if observers view disc A move until fully overlapped with disc B, after which A stops and B moves, they may perceive either streaming or launching - but when perceiving launching, they also see B move before being fully overlapped with A (i.e. leaving an uncovered crescent). In several experiments, we measure illusory crescents in bouncing/streaming displays with auditory tones. Participants adjusted two probe discs until they matched the perceived overlap of an ongoing streaming/bouncing event. We first show that an illusory crescent emerges when the onset of a brief tone is synchronized with the moment of overlap between the two discs. We then show that the timing of this tone matters: illusory crescents still arise for tones occurring slightly earlier than the moment of maximal visual overlap, but when the tone follows the moment of maximal overlap, the crescents disappear. Moreover, the perceived "coincidence" of the tone timing is critical: illusory crescents also disappear when a perfectly-synchronized tone is heard as part of a larger repeating perceptual group of sounds. The presence of illusory crescents at all in such displays explains why observers have difficulty distinguishing objective streaming vs. bouncing. And collectively, these experiments suggest that sound-induced bouncing is a perceptual (rather than a cognitive) phenomenon, resulting from changes in visual (rather than decisional) processing.

For decades, work on perceived animacy has emphasized that the currency of perception consists not only of simple features such as color and shape, but also seemingly higher-level properties such as intentionality and goal-directedness. This work has typically been treated as an end-point of perception, but here we explore how perceiving intentions may interact with other aspects of visual processing. In particular, we demonstrate that when an object is seen to move intentionally, observers perceive its orientation with greater precision. We made use of the 'wolfpack effect': if a moving object remains oriented toward a particular target over time, observers will perceive its motion as intentional (e.g. as chasing) -- even if its actual movements are random. In a detection task, observers viewed a moving disc sporting two dots that were perceived as eyes. The disc moved randomly, facing toward or away from another 'target' shape. Observers had to detect when (on half of trials) this 'facing' regularity became broken, such that the disc began rotating independently. Perceived intentionality boosted orientation processing: observers were more sensitive to the disc's rotation when it initially faced the target. This improvement also occurred in a reproduction task, with a task-irrelevant target. On each trial, observers viewed a randomly moving shape facing toward or away from (or independent of) a moving target. Unpredictably, both the target and the shape's orientation cue disappeared, the shape stopped moving, and observers had to reproduce the shape's final orientation. Responses were more accurate on trials where the shape faced the target, compared to when it faced away or rotated independently. Thus intentional motion improves orientation sensitivity even when the visual cue to intentionality is incidental. Collectively, these experiments demonstrate that the perception of intention from motion interacts with other visual processes in rich and hitherto unknown ways.