senses. I am not claiming that the auditory system ‘‘tries’’ to achieve this result,
only that the processes have been selected by evolution because they did achieve
them.
The argument that I have made does not imply that Gestalt theory is wrong.
For the Gestaltists, the phenomena of perceptual grouping arose from the fact
that there were forces of attraction and segregation that operated in a percep-
tual field. This may indeed be the mechanism by which the grouping occurs. I
am simply arguing that even if this is the form of the computation, the partic-
ular grouping force given to each property of the sensory input and the way in
which the grouping forces are allowed to interact have been determined
(through evolution) to be ones that will tend to contribute to the successful so-
lution of the scene analysis problem.
Closure and Belongingness
Our senses of vision and audition, living in the same world, often face similar
problems. So we should not be surprised if we often find them using similar
approaches to overcome those problems. We have seen how the two systems
sometimes deal with fragmented views of a sequence of events by connecting
them in plausible ways. Another strong similarity between the sense modalities
can be seen in the phenomenon of ‘‘perceived continuity.’’ This is a phenome-
non that is sometimes said to be an example of ‘‘perceptual closure.’’
The tendency to close certain ‘‘strong’’ perceptual forms such as circles was
observed by the Gestalt psychologists. An example might be the drawing
showninfigure9.5inwhichwearelikelytoseeacirclepartlyobscuredbyan
irregular form. The circle, though its outer edge is incomplete in the picture, is
not seen as incomplete but as continuing on behind the other form. In other
words,thecirclehasclosedperceptually.
It is commonly said that the Gestalt principle of closure is concerned with
completing forms with gaps in them. But if it did that, we would not be able to
see any forms with gaps in them, which would be ridiculous. The principle is
really one for completingevidencewith gaps in it.
The Gestalt psychologists argued that closure would occur in an interrupted
form if the contour was ‘‘strong’’ or ‘‘good’’ at the point of interruption. This
would be true when the contours of the form continued smoothly on both sides
of the interruption so that a smooth continuation could be perceived. Presum-
ably laws of similarity would also hold so that if the regions on two sides of an
interruption were the same brightness, for instance, they would be more likely
to be seen as a single one continuing behind the interruption.
Like the perceptual grouping of discrete events, closure can also be seen as a
scene-analysis principle. This can be illustrated with figure 9.13 which shows a
number of fragments that are really parts of a familiar object or objects. The
fragments were obtained by taking the familiar display and laying an irregu-
larly shaped mask over it. Then the parts that were underneath the mask were
eliminated, leaving visible only those parts that had not been covered by it.
Why do the fragments not close up perceptually in this figure? A plausible
Gestalt answer might be that the forces of closure are not strong enough. The
contours of the fragments might not be similar enough or in good continuation
with one another. However, it is easy to show that these are not the basic rea-
232 Albert S. Bregman