摘要：In the past decades, the notion of
amodal completion has developed into a well-known concept in perception theory. The enigmatic completion phenomena raise fundamental theoretical questions about the nature of perception and cognition, and although significant progress has been made, they remain a conceptual challenge for perception theory in general. We have only begun to explore the rich conceptual playground it provides.
The term
amodal completion is widely used in the literature, but is also rather puzzling. What does it mean for a perceptual interpretation to be amodal—not constrained to the given modality of perception? Phenomenologically, the adjective
amodal refers to the fact we actually do not
see the partly occluding parts of an object or scenery. The appearance of any distal stimulus obviously belongs to the outcome of the perceptual process, revealing sensory qualities for some parts, but lacking for other parts—while still being inextricable blended with the overall percept. However, regarding the underlying processes themselves, the term
amodal appears less clear-cut. Note that the “complements amodeaux” in the original French title of
Michotte’s et al. (1964) famous booklet indeed referred to the result of the completion process, not so much the process itself (e.g.,
Van Lier & Gerbino, 2015).
Ever since the pioneering work of Burke, Michotte, and Kanizsa (
Burke, 1952;
Kanizsa, 1979;
Michotte & Burke, 1951;
Michotte et al., 1964), the phenomenon of amodal completion has posed intriguing opportunities and challenges for cognitive science. Because the perception of occluded scene regions involves quite an extreme level of stimulus incompleteness and ambiguity, it appears particularly well-suited for studying the rules and principles of perception. As
Gerbino (2020) highlights in his historical review, paraphrasing
Koffka’s (1935) original ideas, “perceptual completions (both modal and amodal) are key phenomena because they reveal inner forces of organization, when outer forces are weak or absent.” Perhaps most importantly, amodal completion challenges our naïve ideas about the distinction between seeing and thinking. Since amodal completion refers to mental experiences of occluded, and hence invisible regions in a visible scene, it appears odd to conceptualize it as seeing, yet these mental experiences tend to behave like visual impressions in many ways. They often force themselves upon the viewer in an automatic and strangely compelling way, and like most visual illusions, they tend to persist even when they are in direct conflict with the viewer’s conscious knowledge.
Interestingly, among the most ardent proponents of a strict dichotomy between perception and cognition, we find the major pioneers of research on amodal completion, as well as other major theorists who marshal observations from research on amodal completion as arguments for their position. For example, referring to
Kanizsa’s (1985) classic work,
Pylyshyn (1999) notes that particularly “revealing examples of the difference between the organizing principles of vision and the principles of inference are to be found in the phenomenon of ‘amodal completion’” (p. 344). Similarly,
Firestone and Scholl (2015) note that perhaps “nobody has elucidated the empirical foundations and theoretical consequences of this observation [that what you see can be different from what you think
