110 Foundations of Visual Perception
Figure 4.17 Threshold search times as a function of the contrast of the
letters against the background (Näsänen et al., 2001). Each point is the mean
of three threshold estimates. Source:From “Effect of stimulus contrast
on performance and eye movements in visual search,” by R. Näsänen,
H. Ojanpää, and I. Kojo, 2001, Vision Research, 41,Figure 2 (partial).
Copyright 2001 by Elsevier Science Ltd. Reprinted with permission.
THE “STRUCTURE” OF THE VISUAL
ENVIRONMENT AND PERCEPTION
Regularities of the Environment
As we saw earlier, the contemporary view of perception
maintains that perceptual theory requires that we understand
both our environment and the perceiver. In the preceding
section we reviewed some methods used to measure the per-
ceptual capacity of perceivers. In this section we turn our
attention to the environment and ask how one can determine
(a) the regularities of the environment and (b) the extent to
which perceivers use them.
The structure of the environment and the capacities of the
perceiver are not independent. When researchers look for sta-
tistical regularities in the environment, they are guided by be-
liefs about the aspects of the environment that are relevant to
perception. These beliefs are based on the phenomenology of
perception as well as on psychophysical and neural evidence.
We see that insights from the phenomenology and neuro-
science of vision interact to establish a correspondence be-
tween the structure of the environment and the mechanisms
of perception.
Thephenomenologyof perception, championed by Gestalt
psychologists and their successors in the twentieth century
(Ellis, 1936; Kanizsa, 1979; Koffka, 1935; Köhler, 1929;
Kubovy, 1999; Kubovy & Gepshtein, in press; Wertheimer,
1923), is a prominent source of ideas about the kinds of in-
formation the visual system seeks in the environment. The
Gestaltist program of research revealed many examples of
correlation between the relational properties of visual stimu-
lation and visual experience. The Gestalt psychologists
believed that the regularities of experience arise in the brain
by virtue of the intrinsic properties of the brain, indepen-
dent of the regularities of the environment. On this view,
the experience-environmental correlation occurs because the
brain is a physical system, just as the environment is, and
hence they operate along the same dynamic principles.
This Gestalt approach—known as psychophysical isomor-
phism—has been criticized by many, including Brunswik
(1969), who nevertheless considered the factors of perceptual
organization discovered by the Gestalt psychologists as
“guides to the life-relevant properties of the remote environ-
mental objects.” Brunswik and Kamiya (1953, pp. 20–21)
argued thatthe possibility of such an interpretation [of the factors of percep-
tual organization] hinges upon the “ecological validity” of these
factors, that is, their objective trustworthiness as potential indi-
cators of mechanical or other relatively essential or enduring
characteristics of our manipulable surroundings.Brunswik anticipated the modern interest in the statistical
regularities of the environment by several decades; he was
the first (Barlow, in press; Geisler, Perry, Super, & Gallogly,
2001) to propose ways of measuring these regularities
(Brunswik & Kamiya, 1953).
Another prominent champion of environmental factors in
perception was James J. Gibson, whose ecological realism
we reviewed earlier. We will only add here that Gibson de-
rived his ecological optics from an analysis of environment
that is hard to classify as other than phenomenological.
Epstein and Hatfield (1994, p. 174) put it clearly:We cannot shake the impression that “the world of ecological re-
ality” is largely coextensive with the world of phenomenal real-
ity, and that the description of ecological reality, although
couched in the language of “ecological physics,” nonetheless is
an exercise in phenomenology.... Gibson’s distinction between
ecological reality and physical reality parallels the Gestalt dis-
tinction between the behavioral environment and geographical
environment.Besides visual phenomenology, an important source of
ideas about the information relevant for visual perception is
visual neuroscience. The evidence of visual mechanisms
selective to particular “features” of stimulation (such as the
orientation, spatial frequency, or direction of motion of lumi-
nance edges) suggests the aspects of stimulation in which the
brain is most interested. As we mentioned earlier, this line of
thought can be challenged by the level of analysisargument:
Particular features could be optimal stimuli for single cells[Image not available in this electronic edition.]