strongly together into pairs. This factor of relative closeness, which Wertheimer
calledproximity, was the first of his famouslaws of grouping. (From now on, we
will refer to them as ‘‘principles’’ or ‘‘factors’’ of grouping because, as we will
see, they are considerably weaker than one would expect of scientific laws.) The
evidence that he offered for the potency of proximity as a factor in grouping
was purely phenomenological. He simply presented the array in figure 8.2B to
his readers and appealed directly to their experiences of which dots they saw as
‘‘going together.’’ Since nobody has ever seriously disputed Wertheimer’s claim
that the closer dots group perceptually, the principle of proximity was thereby
firmly established simply by demonstration, without any formal experiment.
It is perhaps worth making a brief digression here concerning the phenom-
enological methods employed by Gestalt psychologists. Their demonstrations
have often been criticized because they lack the rigorous experimental pro-
cedures adhered to by behaviorally oriented researchers (e.g., Pomerantz &
Kubovy, 1986). In actuality, however, the Gestaltists were often able to bypass
formal experiments simply because the phenomena that they discovered were
so powerful that no experiment was needed. If hundreds or even thousands
of people viewing their displays agree with their claims about the resulting
phenomenological impression, why bother with a formal experiment? As Irvin
Rock often remarked, the demonstrations of Gestalt psychologists, such as
those in figure 8.2, can actually be viewed as ongoing experiments with an
indefinitely large number of subjects—of which you are now one—virtually all
of whom ‘‘show the effect.’’ In cases for which the facts were less clear, Gestalt
psychologists often performed perfectly reasonable experiments and recorded
objective data, such as the number of observers who reported one percept ver-
sus another (e.g., Goldmeier, 1936/1972). Thus, their phenomenological meth-
odsarenotasfarremovedfrommodernbehavioralonesasisoftensuggested.
After demonstrating the effect of proximity, Wertheimer went on to illustrate
many of the other principles of grouping portrayed in figure 8.2. Figures 8.2C,
8.2D, and 8.2E, for example, demonstrate the principle ofsimilarity:Allelse
being equal, the most similar elements (in color, size, and orientation in these
examples) tend to be grouped together. Similarity can thus be considered a
very general principle of grouping because it covers many different properties.
Another powerful factor is what Wertheimer calledcommon fate:Allelse
being equal, elements that move in the same way tend to be grouped together.
Although this cannot be demonstrated in a static display, grouping by common
fate is indicated symbolically by the arrows in figure 8.2F. Notice that common
fate can actually be considered a special case of similarity grouping in which
the similar property is velocity of movement. It has even been claimed that
proximity can be considered a special case of similarity grouping in which the
underlying dimension of similarity is the position of the elements.
Not all possible similarities are equally effective, however, and some do not
produce much grouping at all. Consider the row of V’s in figure 8.3A, for ex-
ample. Adjacent pairs differ by 180in orientation, yet there is very little spon-
taneous grouping by similarity in this display. Figure 8.3B shows the same
figures in pairs that differ by only 45in orientation, and now the pairwise
grouping is immediately apparent. The visual system thus seems to be much
more sensitive to certain kinds of differences than to others. Even subtle differ-
Organizing Objects and Scenes 193