A brain theory of visual perception

Gestalt psychology, best known for its richness of perceptual phenomena and rigor of observation, was more than just a school of phenomenology. It was an early and systematic approach to infer brain processes from percepts. In his seminal article on apparent motion between two successive stimuli, Max Wertheimer (1912) writes: “Es handelt sich um bestimmte, zentrale Vorgänge, physiologische ‘Querfunktionen’ besonderer Art, die als das physiologische Korrelat der Phi-Phänomene dienen” (We are dealing with certain, central processes, physiological lateral interactions of a special kind that serve as the physiological correlate of the phi phenomena). Accordingly, apparent motion was thought to be based on interactions between two brain states, resulting in a perceptual filling-in of the unstimulated interspace (Felderfüllung), not unlike an electrical short circuit. Thus, far from being content with his discovery of phi motion, Wertheimer was interested in the way in which it originated in the brain.

In 1923, using simple dot and line figures, Wertheimer further investigated the rules of perceptual organization, that is, the Gestalt factors or principles inherent in the visual system. Both of his papers had a lasting impact. They showed that our perception cannot be understood on the basis of stimuli alone, but also requires active mechanisms by which they are processed. More than 80 years later and notwithstanding passionate debates on linking propositions (Teller, 1990), we are now witnessing a physiology that aims to understand these principles in terms of interactions within neural networks.

Inspired by Wertheimer's approach, Wolfgang Köhler (1920, 1923) systematically addressed the neural implications of the Gestalt concept. He postulated that the neuronal processes at the psychophysical level (psychophysisches Niveau), at which neuronal activity becomes conscious experience, have isomorphic structural properties to perceptual Gestalten. Consequently, perceptual organization was assumed to rely on neuronal rather than on stimulus properties.

Although Gestalt psychologists studied perception in general, they often selected phenomena demonstrating that we are not free to perceive as we choose. No matter how hard we try or how often we look at a given stimulus pattern, we are always the victims of the constraints that the visual system imposes on us. The principles that determine why we see the way we do (Koffka, 1935) have become known as laws of vision (Metzger, 1936/1953). They are the reason why, in a multitude of possible “solutions,” only one percept typically materializes. When there are several solutions, the task for the visual system becomes a challenge (Fig. 106.1).

Figure 106.1..  

With prolonged inspection, continuous changes in perceptual grouping may be experienced. New “flowers” arise spontaneously from the “meadow” and disappear again, testifying to the instability of figure-ground segregation. (From Robinson, 1972; see also Spillmann and Ehrenstein, 1996.)

Instead of providing a full historical account of Gestalt factors (Spillmann, 1999a), we concentrate in this chapter on select examples to examine where we stand and where the future might lead us. Excellent treatises of visual perception within a Gestalt psychological context may be found in the handbook chapter of Teuber (1960), in the handbook volume edited by Metzger (1966), and in subsequent books and chapters by Hochberg (1971), Kanizsa (1979), Kubovy and Pomerantz (1981), Beck (1982), Rock (1984), Spillmann and Ehrenstein (1996), and Palmer (1999). Readers will also enjoy the recent papers by Sekuler (1996) and Westheimer (1999), which focus on Max Wertheimer's early contributions and interpret them within the context of modern neuroscience.