Chapter 8
Organizing Objects and Scenes
Stephen E. Palmer
The Problem of Perceptual Organization The concept of perceptual organization
originated with Gestalt psychologists early in this century. It was one of the
central concepts in their attack on the atomistic assumption of Structuralism.
The Structuralists conceived of visual perception as a simple concatenation of
sensory ‘‘atoms’’ consisting of pointlike color sensations. This view of visual
perception is extremely local in the sense that each atom was defined by a par-
ticular retinal position and thought to be independent of all other atoms, at
least until they were bound together into larger spatial complexes by the pro-
cess of associative learning. The Gestaltists, in contrast, believed that visual
perception arose from global interactions within the visual nervous system and
resulted from the overall structure of visual stimulation itself. ‘‘Perceptual or-
ganization’’ was the name they used to refer both to this theoretical idea and to
the set of phenomena they discovered in support of it.
Max Wertheimer, one of the founding fathers of Gestalt psychology, first
posed the problem of perceptual organization. He asked how people are able to
perceive a coherent visual world that is organized into meaningful objects
rather than the chaotic juxtaposition of different colors that stimulate the indi-
vidual retinal receptors. His point can perhaps be most easily understood by
considering what the output of the retinal mosaic would be for a simple but
highly structured image. Figure 8.1A illustrates such an output as a numerical
array, in which each number represents the neural response of a single retinal
receptor. In this numerical form, it is nearly impossible to grasp the structure
and organization of the image without extensive scrutiny. This situation is a lot
like the one the visual system faces in trying to organize visual input, because
the structure we perceive so effortlessly is not explicitly given in the stimulus
image but must be discovered by the visual nervous system. In fact, there is a
potentially limitless number of possible organizations in an image, only one of
which we typically perceive. Which one we experience and why we perceive it
rather than others are thus questions that require explanations.
The structure of the numerical image becomes completely obvious when you
see these same values as luminance levels, as illustrated in figure 8.1B. It is a
picture of several black and white squares that are organized into four hori-
zontal rows on a gray background. But why is this simple structure so obvious
when we view the image and so obscure when we look at the array of num-
From chapter 6 inVision Science: Photons to Phenomenology(Cambridge, MA: MIT Press, 1999), 255–
- Reprinted with permission.