Gestalt psychology: The Study of Perception
Wertheimer’s paper of 1912 has usually been portrayed as the proximal event by which the Gestalt movement was brought into existence. There is truth to this, but it is a truth that must be qualified. Those who read Wertheimer’s paper at the time it was published were given little reason to suspect that in its wake, would soon come a movement calling itself Gestalt psychology. The truth of the matter is that the Gestalt movement, though surely by this time conceived, had not yet been visibly born.
The reason for this was that the basic working hypothesis of the movement -its major empirical generalization — had not yet been fully drawn. The orthodox theory held that stimulation of a sense organ gives rise, within the brain, to a mosaic pattern of discrete excitations. Wertheimer’s “physiological short circuit,” on the contrary, was intended to suggest that this is not the case at all. If we are to do justice to the phenomena of perceived motion, he argued, and then we must assume that the various excitations within the brain become integrated into a kind of “physiological whole-process [Gesamtprozess].” (Wertheimer, 1925) The implications of this view were perhaps not very obvious at the time, but in retrospect they are quite clear. If the various excitations within the brain are not discrete and punctiform, then the sensations arising from them cannot be discrete or punctiform either. Indeed, they cannot be “sensations” at all — at least, not in the sense in which the term had been used ever since Locke. It is still true, of course, that any given perception includes a number of distinguishable aspects, but these aspects can scarcely be spoken of as “sensations,” since they are in every instance determined by the configuration, or “whole-form” (Gesamtform),( Wertheimer, 1925) of the underlying neural process. The point to be made is this: If the state of excitation at a given place within the brain is influenced by the states of excitation of its neighbors, then it simply makes no sense to speak of “sensations” and “sensory elements”; for the doctrine of sensations requires that the sensory projection areas of the brain be influenced by nothing more than the respective sensory receptor units with which they are neurally connected.
The implicit within Wertheimer’s view of the workings of the nervous system was that basic working hypothesis of the Gestalt movement to which we alluded above. Now, following Wertheimer’s empirical researches and theoretical reflections, the way was paved for a kindred hypothesis which proved to be very novel indeed: namely, that perception as a whole is a function of Gestaltfaktoren. It was with the emergence of this hypothesis into explicitness that Gestalt psychology was born. We cannot say with certainty just when this occurred, nor even to which of the three founders of the movement it may be best attributed. Thus, as early as 1913, Wolfgang Köhler published an important theoretical paper in which he took to task the orthodox doctrines of “sensory elements” and “unconscious inference.” (Köhler, 1913) By the time Köhler published his physischen Gestalten around 1920, (Köhler, 1924) it may be said that the Gestalt movement was not only full born, but very nearly full grown as well.
One of the causes the Gestalt movement accumulated feverish tempo so speedily was that its chief hypothesis proved to be instantly relevant to a great number and multiplicity of perceptual experiences with which psychologists were previously aware. As regards to this, disciples of perception had known for many years, for example, of the variety of geometrical illusions. Yet, none had ever been able to conclude a satisfactory explanation of them. On the other hand, the Gestalt psychologists were capable to elucidate them effortlessly, persuasively, and almost without exemption in terms of configuration.
There were certain other, more recently discovered perceptual phenomena which also yielded readily to a configurational interpretation. A good example of this is found in a very thorough study of The Modes of Appearance of Colors, published by David Katz in 1911. (Katz, 1911) Katz’ study is usually described as a splendid example of the “phenomenological method.” There was, however, a great deal more to it than this. Indeed, a perceived color was taken by many of the advocates of Locke’s mental mechanism and Wundt’s “New Psychology” to be an exemplary instance of an elementary sensation.
What Katz did was to bring to light a number of color-perception phenomena which were clearly incompatible with the orthodox doctrine of sensations. He pointed out, for example, that any given hue can appear in any of three quite distinct “phenomenal modes.” Thus, a given shade and intensity of blue can have quite a different appearance depending upon whether it is the color of an opaque object, a translucent object, or of the sky. In the first instance it is a “surface color”, in the second a “volumic color”, and in the third a “film color”. Here, then, was a case in which the same individual color receptors were being excited by the same individual color stimuli, but yet in which the resulting overall perceptions were quite different. Moreover, the difference was clearly the result of configurational factors. Thus, as Katz observed, if one views a “surface color” out of context, it immediately becomes a “film color”; in so doing, it loses both its definite spatial localization and its tendency to appear as the same hue under varying conditions of illumination. In any case, so it was that a great many perceptual phenomena seemed from the outset to bear witness to the truth of the configurational hypothesis. That hypothesis, put into other words, held that the properties of any perceptual part are determined, in some measure, by the properties of the whole of which it is a member.
Gestalten: Perceptual, Physiological, and Physical
Perception does not always correspond exactly to its external stimulus. On the contrary, there is often discrepancy between the two; and this, it was held, is in every instance a product of configurational factors. But how or in what fashion is it a product of configurational factors? Gestalt psychology’s attempt to answer this question was contained in what has since come to be known as the “law of Prägnanz.” As an empirical generalization, what the law of Prägnanz held was this: Whenever a stimulus and a perception are discrepant, it will be found that the perception is more prägnant (terse, concise), einfach (simple), and regelmässig (regular, well-proportioned) than the stimulus; conversely, whenever a perception and a stimulus correspond exactly, it will be found that the stimulus is already as prägnant, einfach, and regelmässig as it could possibly be. This law of Prägnanz was also spoken of, at times, as the law of “good form.”
A good illustration of the law of Prägnanz is to be found in Kurt Koffka’s ingenious application of it to the perceptual illusion associated with the Necker cube. (Koffka, 1935)The Necker cube is a two-dimensional line drawing of a cube with transparent sides and opaque edges. Almost anyone who looks at this two dimensional drawing will readily agrees that it has a three-dimensional appearance. Now, up to this point, the law of Prägnanz was an empirical generalization and nothing more, for though it described the conditions under which stimulus perception discrepancy comes about, it did not go on to tell just how they serve to bring it about. Even granting, then, that the law of Prägnanz is descriptively accurate, we must still ask, how is it enforced? It should come as no surprise that Gestalt psychology sought to answer to this question in the workings of the central nervous system.
As we have noted, Gestalt perceptual theory was at its best in just those cases where the orthodox theory was at its worst — namely, in cases where there was a discrepancy between sensory “input” and perceptual “output.” The reason why orthodox perceptual theory found cases of this sort difficult to deal with should by this time be fairly clear. In cases where there is a discrepancy between the initiatory sensory-receptor process and the resulting perception, we must assume one or the other of two things about the brain process that stands intermediate between them. We must assume either than the mediating brain process is on a par with the sensory input, but not with the perceptual output, or that it is on a par with the perceptual output, but not with the sensory input. The orthodox theory had, of course, embraced the first of these assumptions; as we have seen, it held the mediating brain process to correspond point-for-point with the initiatory sensory-receptor process. In consequence, it was obliged to regard the discrepant perceptual output as being not perceptual at all, but rather aperceptual — the result of unconscious associative inference.
Consider, now, what happens if we reject the first assumption and embrace instead the second. That is, in cases where there is a discrepancy between sensory input and perceptual output, consider the consequences of assuming that the mediating brain process is on a par with the perception, but not with the initiatory sensory-receptor process. The first and most obvious consequence is that the task of explaining the discrepancy of the perception becomes a task of a very different order. Indeed, there is now no need of explaining perception at all; all that need be said of it is that it is the immediate correlate and correspondent of the underlying brain process. This, of course, was a conclusion well suited to the Gestalt point of view, for it carried with it the suggestion that perception is, after all, directly given. Thus, it was with this second assumption and all its attendant implications that the Gestalt theory of the nervous system began. It was spoken of under the title “psychoneural isomorphism,” and its status as an assumption was — at the outset, at least — freely admitted. (Henle, 1961)
Conversely, the Gestalt assumption of psychoneural isomorphism required no further explanation of the perception, but it did require a further explanation of the brain process. Specifically, Gestalt theory had to go on to explain the supposed discrepancy between the brain process and its initiatory sensory-receptor process. This, then, was the task, and the proponents of the Gestalt point of view set about it straightaway. As we have seen, orthodox theory took the brain to be a mere passive recipient of peripheral sensory excitation. Gestalt theory, of course, agreed that the brain is a recipient of peripheral excitation, but it went on to argue that it is not a mere passive recipient. On the contrary, it held, the role that the brain plays is an active one. It receives the peripheral excitation, to be sure, but it then goes on to transform this excitation in accordance with its own inherent dynamic properties. And it is precisely these, the inherent dynamic properties of the brain that are responsible for any disparity that might exist among the sensory-receptor process, the intermediate brain process, and the resulting perception. Considering that the term “dynamic” is tossed about with such abandon these days, we shall do well to point out that within Gestalt theory it had a fairly precise and unequivocal meaning. Orthodox theory had held that the workings of the brain are determined principally by the brain’s anatomical structure — that is, by the arrangement of its pathways and their various interconnections. Gestalt theory contended that these workings are determined also, and in large measure, by the distribution of forces within the brain — and this, moreover, in a fashion which is relatively independent of pathways and interconnections. (Watson, 1978)
Gestalt theory offered was simple and direct: The transformation must be of precisely the sort required to produce the observed discrepancy between sensory input and perceptual output. In truth, the argument was circular and would have remained so, but for one thing: the Gestalt theorists insisted that there are other considerations, independent of input output discrepancy, which leads to quite the same inference concerning the properties of the central nervous system. These were considerations supplied by none other than that old, familiar guiding light of psychological theory -physics. (Koffka, 1935)
It was a line of reasoning that took somewhat the following form: If “forces” can be propagated through space without the assistance of a material medium, it must be that they themselves have spatial properties; specifically, they must have spatial extension and spatial configuration. Indeed, we ought no longer to speak of “forces” at all, for what we are really dealing with are fields of force — motive agencies which are spatially extended and configured, and whose strength varies continuously from one part of the configuration to another. (Ellis, 1938) Take, for example, the familiar case of an iron magnet. If we place a sheet of paper over the magnet and then shake iron filings onto the top surface of the paper, we shall find that the filings become distributed on the paper in a characteristic pattern. The filings are not in contact with the magnet, but they are nonetheless affected by its extended and configured field of force. Magnetism, electricity, and light were the forces which seemed most clearly at the time to have such field properties. By the 1880s, there was good reason to believe that all three of these were but special cases of a single electromagnetic force. Indeed, there were even some who suspected that other varieties of force, or energy, might also be electromagnetic. All in all, then, physical field theory was just about ready to carry the day. (Wertheimer, 1925)
This characterization of a field of force as extended and configured should betray at once the relevance of physical field theory to Gestalt psychology. The brain is, after all, a physical system. As such, it is the locus of physical forces, of which at least some are apt to have field properties. Köhler began his physischen Gestalten with the observation that psychology, being a young science, would do well to find parallels whenever it can between its own phenomena and those of the older, better established sciences. He then pointed out that such parallels already exist, waiting only to be noticed, in the realm of physics. Certain psychological processes have configurational properties; so also do certain physical processes which we may assume take place within the brain. The task, then, is simply to determine the extent to which these two types of configurational properties correspond. We are already familiar with the configurational properties of psychological processes, so let us proceed to consider those of physical processes.
Though Köhler’s argument beyond this point was long and complex, it may be summarized fairly briefly. The processes associated with physical force fields, he observed, may exist in one or the other of two states: They may be either stationary (unchanging) or dynamic (changing). A stationary state will remain stationary until it is altered by some external force, whereupon it will become a dynamic process. On the other hand, a dynamic process will, if left to itself, eventually become a stationary one. Thus we may conclude that a dynamic process tends to become stationary, and that a stationary process tends to remain stationary.
What, now, is the difference between a stationary and a dynamic process, apart from the fact that one is changing and the other is not? It is simply this: A stationary process is one in which the several constituent forces have reached a state of equilibrium; a dynamic process is one in which the constituent forces are still in a state of non equilibrium. Thus, speaking more broadly, we may say that an unbalanced configuration of forces tends to change in the direction of equilibrium; further, once it has achieved equilibrium it will change no more.
These, then, are the general properties of physical force fields. For all who have eyes to see it, the parallel with psychological phenomena should be perfectly plain. Let us assume, Köhler argued, that sensory excitation gives rise to a force-field process within the brain. If the several constituent forces are balanced at the outset, then the process will be and remain a stationary one and the resulting perception will correspond exactly to the sensory input. Contrariwise, if the constituent forces are imbalanced, the process will change so as to achieve a state of equilibrium, and the resulting perception will be discrepant with sensory input. In the latter case two kinds of changes take place, and each is entirely parallel with the other: First, the constituent forces of the brain process have achieved a state of greater equilibrium; second, the resulting perception is more prägnant, einfach, and regelmässig than the initiatory sensory-receptor process. Surely we are not stretching things too far if we conclude that the perceptual tendency toward Prägnanz is but the phenomenal counterpart of the tendency of force fields within the brain to achieve equilibrium.
Köhler’s treatment of physical and psychological Gestalten naturally left a great many details to be worked out. As he noted, even though we may understand the generalities of the relationship between the physical and the psychological, we are still “not in the position to derive the respective physiological and phenomenal characteristics in individual cases.” Nevertheless, he went on to observe, “The directions taken by such processes show clearly enough that they involve Gestalten of the same basic character as are found in physics.” (Koffka, 1924) And so it is that the ultimate explanation of perceptual phenomena is to be found among the principles of physical field theory.
Gestalt psychology began as a theory of perception, but it did not long remain within that confine. Indeed, its pioneers began almost immediately to enlarge it into a theory of psychology in general. A good illustration of this attempt to expand the theory is found in Kurt Koffka’s Principles of Gestalt Psychology, published in 1935. Understandably, Koffka began with several chapters devoted to perception, principally visual perception. He then went on, though, to devote about an equal number of chapters to such diverse matters as “reflexes,” the “ego,” “adjusted behavior,” “attitudes,” “emotions,” the “will,” “memory,” “learning,” and even “society and personality.” Similar attempts to expand the theory were made by Wetheimer, Köhler, and others. We cannot examine these in detail here, so let us instead try to convey something of their rationale and general form.
It has often been said that Gestalt psychology differs from other major psychological theories, such as those of psychoanalysis and behaviorism, in that it is not mechanistic. Whether this is true depends upon how we define our terms. As applied to psychological theory, the term “mechanistic” can have one or the other of three quite separate meanings. First, it can mean that the psychological theory in question has been in part derived from mechanistic physical assumptions. (Murphy, 1949) Psychoanalysis would fall under this heading in large measure, behaviorism somewhat less so. Gestalt theory, of course, would not fall under it at all. Second, it can mean that the psychological theory in question inclines toward atomism. Behaviorism, with its tendency to analyze behavior into its elementary constituents, would of course be an example of this kind of mechanism, as would psychoanalysis and every other psychological theory that we have mentioned, with the sole exception of Gestalt theory. (Hartmann , 1935) Finally, the term “mechanism” can refer to what William James described as “automation theory,” the central assertion of which is simply that consciousness is causally irrelevant to behavior. Now it is true that Gestalt psychology was not mechanistic in either of the first two senses of the term. In this third sense, however, it most surely was. This is not to suggest that the Gestalt psychologists were so naïve as to deny the existence of consciousness, but it does mean that they, along with their psychoanalytic and behavioristic colleagues, denied its causal efficacy. They held that consciousness is nothing other than the isomorphic “phenomenal counterpart” of the underlying brain activity — which, though not “mechanical,” is nonetheless entirely automatic!
To appreciate this point, we have only to recall that Gestalt psychology was, after all, a physicalistic psychology. To be sure, its principal concern was with the phenomena of psychology, but the conceptions by which it sought to understand these phenomena were imported, virtually unaltered, from the realm of physics. In this, of course, Gestalt psychology has much company, for we have seen time and again how psychological theories, ever since the time of Descartes and Hobbes, have tended to incorporate physicalistic assumptions. There is, however, one very important difference. All such theories prior to Gestalt were guided by assumptions which came principally from Galilean or Newtonian physical theory. Gestalt theory was the first to be guided by assumptions deriving from physical field theory.
Ellis, W. D. , 1938. A Source Book of Gesltat Psychology ( New York: Harcourt, Brace).
Hartmann , G. W. 1935. Gestalt Psychology: A Survey of Facts and Principles (New York).
Henle, M. ed., 1961Documents of Gestalt Psychology ( Berkeley: University of California Press).
Katz, D. ( 1911). “Die Erscheinungsweisen der Farben und ihre Beeinflussung dutch die individuelle Erfahrung,” Zeitschrift für Psychologie, 7.
Koffka, K. 1935″Some Problems of Space Perception,” in Psychologies of 1930, ed. C. Murchison (Worcester, Mass.: Clark University Press, 1930), and K. Koffka, Principles of Gestalt psychology (New York: Harcourt, Brace), pp. 159.
Köhler, W. (1913) “Ueber unbemerkte Empfindungen and Urteiltäuschungen,” Zeitschrift für Psychologie. 66).
Köhler, W. 1924 Die physischen Gestalten in Ruhe und im stationären Zustand (Erlangen: Verlag der philosophischen Akademie,).
Köhler, W. 1938.The Place of Value in a World of Facts ( New York: Liveright).
Köhler, W. 1940 Dynamics in Psychology ( New York: Liveright).
Murphy, Ronald 1949.Historical Introduction to Modern Psychology ( New York: Harcourt, Brace, and World).
Watson, R. I. 1978 The Great Psychologists from Aristotle to Freud ( Philadelphia: Lippincott,).
Wertheimer, 1945 Productive Thinking ( New York: Harper).
Wertheimer, M., 1925″Experimentelle Studien über das Sehen von Bewegung,” in Drei Abhandlungen zur Gestalttheorie ( Erlangen: Verlag der philosophischen Akademie).