GESTALT THEORY

The Gestalt movement in psychology began early in the twentieth century; its founders were the German psychologists Max Wertheimer, Wolfgang Köhler, and Kurt Koffka. A Gestalt is essentially an organized whole whose parts belong together, as opposed to being simply juxtaposed or randomly distributed. As Wertheimer put it, "What happens to a part of the whole is determined by intrinsic laws inherent in this whole." The Gestalt theorists believed this principle to be of wide application and to be relevant to the psychology of perception in particular.

History

As early as 1890 Christian von Ehrenfels had pointed out that to appreciate a melody we need to be aware not of single tones in isolation but of a succession of tones

which combine in a particular way. If notes of the same pitch as those of the original melody are presented in a different temporal order, there will be a completely different effect, whereas the same melody played in a different key is immediately recognizable, even though the notes are different in pitch from the original ones. The melody as a whole was said by von Ehrenfels to have a Gestaltqualität independent of the qualities of the separate notes. Wertheimer, Köhler, and Koffka were concerned to apply the concept of Gestalt over a wide area and thus give a new direction to psychological research.

A central feature in their view was the doctrine of isomorphism, which asserts that our experiences have the same structure as the brain processes which underlie them. Thus, if the stimulus is a nearly complete circle which the subject sees as a complete circle the doctrine of isomorphism would assert that there must be some pattern in the brain that is isomorphic with the complete circle, as opposed to the incomplete one. The detailed neurological hypotheses which they put forward are of questionable value, but the general principle is still of interest, as is Köhler's demonstration that there are Gestalten in physical nature, for example, the soap bubble, whose spherical shape is the necessary result of the total forces in operation at any one time (see Koffka, Principles of Gestalt Psychology, p. 14).

The following are some typical examples of the Gestalt principle as applied to vision: Figure 1 appears as a cross; if, however, we consider the effect on the retina of

the eye of each of the dots in isolation, there is nothing to account for the way in which they are organized. Implicit in traditional thinking, according to Köhler and Koffka, is the so-called constancy hypothesis—the hypothesis that stimulation of a particular point on the retina has a constant effect regardless of the total pattern of stimulation. Yet if the constancy hypothesis were true, it would be hard to explain the obvious recognizability of the "4" in Figure 2a and its camouflage in Figure 2b, since the same retinal points are being stimulated in both cases. Similarly one cannot explain how a person who moves from twenty yards away to ten yards away continues to look approximately the same size, since the retinal stimulation must by the laws of optics be quite different. Indeed there are many characteristics of the perceived world (what Koffka terms the "behavioral," as opposed to the "geographical," environment) which do not bear a one-one relationship to anything in the pattern of stimulation. Thus, in Figure 3 we see the lines as four pairs, but the "togetherness" of each pair has no direct counterpart in the system of stimuli; and in Edgar Rubin's famous example (Figure 4), whether we see the white as "figure" and the black as "ground" or vice versa, there is no direct counterpart to the "thinglike" character of the figure and the absence of this character in the ground.

This thinglike character, the character of "productivity," which occurs in some causal transactions, the character of "being mine," which belongs to, say, my hand in contrast to an object on which my hand is resting, the character of anger present in someone's face—these are some of the many features which are present in the behavioral environment, even though they are necessarily absent from the world of physics because there is nothing in the stimulus situation directly corresponding to them.

On the Gestalt view what is all-important is the way in which the immediate, or "proximal," stimuli (for example, light waves or sound waves) combine in space and time; when these combinations are of a certain kind, certain perceptual organizations will arise (for example, two parts of a diagram will be seen as belonging together), and laws can be framed in terms of which such organizations can be predicted. Details of these laws have been set out by David Katz; they include the law of proximity, which states that, other things being equal, in a total stimulus situation those elements which are closest to each other tend to form groups, and the law of closed forms, which states that, other things being equal, lines which enclose a surface tend to be seen as a unit.

A law of a more general kind is that of Prägnanz. As formulated by Koffka (Principles of Gestalt Psychology, p. 110), this law states: "Psychological organisation will always be as 'good' as the prevailing conditions allow. In this definition the term 'good' is undefined. It embraces such properties as regularity and symmetry, simplicity and others." In other words, when the stimuli are of a certain kind, there are forces within the organism that operate in the direction of maximum simplicity; hence, we tend to see "good" figures—squares and circles, for example—rather than less regular ones. The word Prägnanz is of course ultimately connected with the Latin impregnare. The suggestion here, however, is not that of something being fertilized or made pregnant but rather of something being stamped or pressed into a particular shape (compare the word prägen, which is used primarily to refer to the minting of coins). Certain types of configurations, one might say, are particularly impressive; they carry a certain stamp or they strike us in particular ways.

Contrary to what has sometimes been said, the Gestalt psychologists did not dispute that past experience can influence perception; this is made plain by Katz in Gestalt Psychology (pp. 28–29). Their criticism was

directed against the view of perception that invoked past experience as a deus ex machina when observed results did not fit the constancy hypothesis. Thus a penny, unless its flat surface is directly in front of us, might be expected from the laws of optics to look elliptical. Since it does not, one can preserve the view that the "basic datum" is an ellipse by postulating a rapid process of inference based on past experience. On this view we infer that the penny is round because of our alleged previous experience of round pennies. According to the Gestalt psychologists, however, not only is there nothing in introspection to suggest such an inference; they would also have questioned whether in fact it is particularly common in ordinary life for pennies to have their flat surfaces directly in front of the observer—a condition that their opponents' theory seems to require. In contrast, their view was that when the proximal stimuli combine in a certain way a particular perceptual organization is forced upon us; thus a circle is a "good" figure, and hence the "internal forces" will operate in the direction of a circle rather than an ellipse.

Interesting experimental studies include those of Wertheimer ("Experimentelle Studien über das Sehen von Bewegung," in Zeitschrift für Psychologie ) on the perception of movement, those of Köhler (The Mentality of Apes ) on problem solving in apes, those of Wertheimer (Productive Thinking ) and K. Duncker ("On Problem Solving") on problem solving in humans, that of Katz (The World of Colour ) on the perception of color, that of Rubin (Synsoplevede Figurer ) on the figure-ground distinction, those of Kurt Lewin (Principles of Topological Psychology ), who has attempted to apply Gestalt principles to the study of social situations, and those of Albert Michotte (La perception de la causalité )—although he was not a member of the original group—on the conditions in which we receive an impression of causality. A recent interesting development is the attempt to relate figural goodness to the amount of information (in the mathematical

sense) needed to specify a particular pattern or figure (see especially Fred Attneave, Applications of Information Theory to Psychology, p. 82). The problem of perceptual Gestalten has arisen in an acute form in the programming of computers to carry out pattern recognition (see, for instance, Kirsch, "Computer Interpretation of English Text and Picture Patterns").

Science and Common Sense

The advance of science continually brings in its train a challenge to our commonsense beliefs about the world. At one time or another in the history of scientific thought it has been held, for example, that sense perception is unreliable, that the things around us are not really colored, that the floor on which we walk is not really solid, and that no two events are ever exactly simultaneous. In contrast with many other scientific systems, Gestalt theory involves the attempt to call us away from such paradoxes back to common sense; it invites us to consider the world as we in fact experience it, not as we might expect to experience it in the light of the latest scientific developments.

It does not, of course, follow, that philosophical paradoxes can be disposed of simply by pointing out that experience is in fact of such-and-such a kind. Thus, it is no argument against John Locke's account of substance or David Hume's account of causality to point out that the behavioral environment is found by experience to consist of things in causal interaction, any more than it is an argument against the distinction between primary and secondary qualities to point out that we are aware of greenness as being in the grass and not in our heads. Similarly, a philosopher who wishes to defend the sense-datum terminology cannot be refuted simply by an appeal to Rubin's claim that what we perceive is organized into figure and ground and is not just "a mosaic of sense data." As a corrective against those who mistake the point of philosophical arguments, it may be helpful on occasions to make explicit exactly what we are aware of at the commonsense level, but this does not prove the philosophical arguments to be wrong.

Despite the emphasis on naive judgment, however, the Gestalt program does not involve an uncritical return to naive realism. Rather, its claim is that the gulf between what common sense tells us and what science tells us is not, after all, as great as might be supposed; the world of nature is gestaltet no less than are our experiences. Moreover, although in some of their discussions Köhler and Koffka speak in traditional terms about "the relation between mind and matter," their views do not fit easily into the traditional categories of interactionism, epiphenomenalism, and parallelism; indeed, like many modern philosophers, they are critical of a starting point which forces us to decide between theories couched in these terms.

Positivism and Behaviorism

The prevailing scientific attitude of the time, which Koffka called "positivism," was mistaken, on the Gestalt view, because it allowed no place for the categories of meaning and value. The important fact for psychology is that the behavioral environment is organized—it is intelligible. Thus we are making sense out of a person's facial expression when we say that he is angry. Similarly, if a person listens to music, he is sometimes aware that a chord with the leading note (the seventh of the scale) at the top requires to be followed by the tonic chord of the original key; the cadence has its special meaning only if the second chord follows the first. This remains true even though such "requiredness," as Köhler terms it in The Place of Value in a World of Facts, can play no part in the world of physics. As physical science advances we are enabled to make continually more refined statements about the geographical environment; but in so doing, on the Gestalt view, we are in danger of losing sight of facts—those of the behavioral environment—which for the psychologist are of special interest. Koffka agreed that vitalism is "no solution but a mere re-naming of the problem"; but by taking seriously the concept of Gestalt one can, he held, be antimechanistic without being obscurantist.

There is also an attack on the allegedly "scientific" creed of behaviorism, whose development was almost contemporary with that of Gestalt theory. The term behaviorism, as Köhler understands it, implies a denial that there can be "a science of direct experience," either because there is no such thing as direct experience or because if it exists, it is not accessible to public scrutiny. In reply Köhler points out that no scientist can even begin to experiment unless he starts from his own experienced world. He also points out that one has as little or as much justification to be skeptical about the world of experience as one has to be skeptical about the world of physics; there is no good reason why the behaviorist should choose to ignore the world of experience while taking the world of physics on trust.

Isomorphism

It is far from clear whether the doctrine of isomorphism constitutes a radically new discovery, as the Gestalt theorists supposed, or whether it is a somewhat high-sounding way of asserting the obvious. Most modern psychologists, if asked, would doubtless express the hope that complete explanations of perception and learning will eventually be found in terms of brain processes. If, therefore, the contribution of Gestalt theory is to be distinctive, clearly some more far-reaching claim must be involved.

Koffka expressly pointed out that Gestalt theory does not stand or fall with the correctness of a particular theory about perceived movement. According to Koffka its more general objective is to contribute to "the integration of value, life, and mind. … The Gestalt concept … cuts across the division of realms of existence, being applicable in each of them." That is, there are Gestalten in nature (for example, the soap bubble); there are Gestalten in the living brain; and there are Gestalten in our conscious experience. In traditional discussions about the relation between mind and body, according to Köhler, it was "tacitly assumed that only microscopic events in the cortex can be the correlates of mental life." In contrast, the doctrine of isomorphism invites us, in Koffka's words, to "think of the physiological processes not as molecular but as molar phenomena. … If they are molar, their molar properties will be the same as those of the conscious processes which they are supposed to underlie." As has already been pointed out, this is not just an answer within the context of traditional mind-body dualism; it is an attempt to look at this whole family of problems afresh.

The distinction between the geographical and behavioral environments gives rise to difficulties of its own. One very reasonably asks, in the first place, what kind of duality is involved. Clearly, there are not two environments in the same sense that there are—or might be—two rooms in a country cottage. If a child specialist recommended a change of environment for a child, it would make no sense to reply "very well; we will change him from the geographical to the behavioral environment." One is reminded in this connection of Arthur Eddington's claim—which in fact involves the same kind of difficulty—that he is writing simultaneously at two tables. Second, if we take the idea of two environments at its face value, we are tempted to ascribe some kind of superior status or "reality" to one or the other. It is the geographical environment, according to Koffka, whose contents are "real": "The pen with which I am writing is a unit in my behavioural environment and so is the real pen in the geographical" (italics added). What is "real," however, does not apparently coincide with what is "given"; "every datum is a behavioural datum; physical reality is not a datum but a constructum." The suggestion that each of us somehow "constructs" a physical world out of his immediate experiences implies a phenomenalist view which in the last resort leads inevitably to solipsism. Clearly this was not Koffka's intention, but he gave no indication how such a conclusion can be avoided.

The important point, according to Köhler and Koffka, is that the concept of Gestalt cuts across these two different kinds of reality. In the words of Köhler, "Any actual consciousness is in every case not only blindly coupled to its corresponding psychophysical processes, but is akin to it in essential structural properties." The difficulty here is that anything can be regarded as "structurally akin" to anything else, provided enough rules are given. Many maps are structurally akin to landscapes in that they involve the same geometrical shapes; but if one is allowed sufficient rules for specifying what represents what, a map of England (as we now call it) could function without any misrepresentation as a map of France. What Köhler needs to argue for is some relatively clear-cut and uncomplicated structural relationship. Thus, it may well be that the shape of the areas stimulated in the cortex has something in common with the shape which we observe in an object, although it is hard to see how there can be any close parallel in the case of color, since, when X is looking at a green object, Y does not find anything green in X 's cortex. One must suppose that the use of the term psychophysical (instead of physical ) to describe processes in the brain is intended to emphasize these relatively close structural similarities.

Knowledge of Other Minds

In reply to the charge of solipsism, Köhler and Koffka could point out that they both discuss the problem of knowledge of other minds at some length. The main feature of philosophical interest in these discussions is that structural similarities are pointed out between behavior which is noticed by others and so-called inner states, which are discriminated only by the person himself. A person's wincing may have precisely the same temporal properties as his twinges of pain, and the sound of his rising voice may have the same movement properties as his inner feelings of rising anger. Koffka pointed out that a character in a Mickey Mouse cartoon can quite well look exuberant or dejected; and if we can directly observe such exuberance or dejection in these cases, there seems to be no reason why we should not directly observe it in our friends. Similarly, Michotte has argued that visual experiences of live movement are structurally similar to kinesthetic and other experiences that we have when we make an effort. These considerations are not, of course, sufficient to remove all possible skeptical doubts; but they at least make clear the conditions in which we can justifiably say of a person that, for example, he is angry, dejected, or making an effort.

"Seeing" and Inference

Koffka suggested that the word see should be used in a "purely phenomenological" sense, that is, in such a way that the words which follow are a simple description of our experience. Thus, if we look at a table that is partly covered by a book, we are not aware of any gap or hole in the table in the area where the book is; and Koffka therefore wanted to say that what we see is a complete table. On his view it is necessary to describe our experience without being influenced by considerations of what we might expect to see on the basis of scientific knowledge—in this case, by our knowledge about the characteristics of the light waves striking the retina.

In ordinary usage, when we say that we or someone else saw something, this is normally not just the report of a visual experience; there is also an implicit claim to correctness. If the person is in fact deceived, then one is wrong to use the word saw, just as one would be wrong to say that a person has "proved" something if his argument contains a fallacy. One might therefore express Koffka's difficulty by saying that, as our knowledge of physics increases, claims to correctness will force us further and further away from naive description. "Why," he asked rhetorically, "are we so hopelessly stupid as to call the colour of our table-cloth on the candle-lit dinner table white, when Helmholtz told us that it was yellow?" According to Koffka's proposed usage we actually see a white tablecloth, the evidence of the physicists notwithstanding.

This "purely phenomenological" sense of "see," however, is unnecessary. We already possess the expression "seeing-as" for situations where we do not wish to make claims to correctness. On this usage, we see the tablecloth as white even though we might have expected to see it as yellow. Similarly, a person who is exposed to an ambiguous diagram may see it as two-dimensional, even though on another occasion he sees it as three-dimensional. Unless one is an extreme skeptic and asserts that claims to correctness are never appropriate when visual perception is involved, it is surely useful to have a terminology which enables us to make such claims on some occasions and to withhold them on others.

Moreover, Koffka's proposed usage, if adopted consistently, carries the paradoxical consequence that we can never be sure or unsure about what we see, nor can we be right or wrong. If we look more carefully at something, or if others give us a verbal description or point to a contour line, we may, of course, see something new; but we were still neither right nor wrong about what we saw before.

There is the further paradoxical consequence that no two people can ever see the same thing. Köhler is apparently prepared to accept this, since he expressly tells us that no two scientific investigators ever see the same galvanometer. Moreover, if everything that we see (in Koffka's sense) is, by definition, part of the behavioral environment, this has the effect of turning the behavioral environment into a home for every erroneous perception which has been made. It is as bad as having to postulate false facts to ensure that false propositions refer to something. In general, the distinction between the geographical and behavioral environments involves many points of interest, but in philosophizing in this area, Koffka in particular was not successful in avoiding paradoxical consequences.

As far as psychology is concerned, the work of the Gestalt theorists has led to the discovery of a large number of new facts, particularly in the sphere of perception, and to a reinterpretation of facts which were already known. On the basis of the laws of structural organization, predictions can be made about what will be perceived when the proximal stimuli are of a particular kind, and these predictions normally work. The neurological explanations are inadequate by present standards, but even in this area the Gestalt theorists have at least called attention to problems which require to be solved, and, in particular, they have taken seriously the challenge presented by our ability to perceive spatial and temporal relationships.

On the broader theoretical issues, traditionally the province of philosophy, their main contribution has been to indicate the need for a change of emphasis. In the light of advancing scientific knowledge, it appeared to be the case that what we thought we were perceiving was not what we were really perceiving at all ("the grass is not really green," "the white tablecloth seen in candlelight is really yellow," and so on). For the Gestalt theorists, however, as for some modern philosophers, such claims were paradoxical and confused. Whatever physics tells us, the starting point, on the Gestalt theorists' view, must necessarily be the world as perceived by common sense; this is the world to which organisms respond, and it is therefore of special importance for the psychologist; moreover, if we ourselves did not perceive it in certain ways, physics could not even begin. This is to say, in effect, that language descriptive of ordinary experience can never be reduced to the language of physics—a thesis that has been held in many forms but which has seldom been defended in such a sustained and systematic way.

See also Behaviorism; Eddington, Arthur Stanley; Ehrenfels, Christian Freiherr Von; Hume, David; Koffka, Kurt; Köhler, Wolfgang; Locke, John; Perception; Phenomenological Psychology; Positivism; Psychology.

Bibliography

Attneave, Fred. Applications of Information Theory to Psychology. New York: Holt, 1959. For the relevance of information theory to Gestalt theory, see especially p. 82.

Duncker, K. "On Problem Solving." Translated by L. S. Lees. Psychological Monographs 58 (5) (1945). A short monograph with some interesting experiments on problem solving.

Ehrenfels, Christian von. "Über Gestaltqualitäten." Vierteljarsschrift für wissenschaftliche Philosophie 14 (1890): 249–292. Of historical interest.

Ellis, Willis Davis. A Source Book of Gestalt Psychology. London: K. Paul, Trench, Trubner, 1938. A helpful collection of the writings of Gestalt psychologists in one volume; recommended as a standard textbook.

Hamlyn, D. W. "Psychological Explanation and the Gestalt Hypothesis." Mind 60 (240) (1951): 506–520. A philosophical article in which, among other things, the concept of Gestalt is compared with that of substance.

Hamlyn, D. W. The Psychology of Perception. London: Routledge and Paul, 1957. Contains a critical examination of some of the concepts of Gestalt theory in the light of recent British philosophy.

Hartmann, George W. Gestalt Psychology; A Survey of Facts and Principles. New York: Ronald Press, 1935. A readable account of the history of the movement and its discoveries up to 1935. Contains brief biographical details of Wertheimer, Köhler, and Koffka and a list of important dates. Recommended as a standard textbook.

Henle, Mary, ed. Documents of Gestalt Psychology. Berkeley: University of California Press, 1961. A collection of papers on Gestalt psychology by various authors, mostly dating from 1950 onward.

Katz, David. Der Aufbau der Farbwelt. Translated by R. B. Macleod and C. W. Fox as The World of Colour. London: K. Paul, Trench, Trubner, 1935. Interesting discussions of the perception of color.

Katz, David. Gestalt Psychology; Its Nature and Significance. Translated by Robert Tyson. New York: Ronald Press, 1950. An excellent introductory textbook.

Kirsch, R. A. "Computer Interpretation of English Text and Picture Patterns." Institute of Electrical and Electronic Engineering Transactions on Electronic Computers, Electronic Devices 13 (4) (1964). Illustrates some of the difficulties in programming computers to be aware of perceptual Gestalten.

Koffka, Kurt. The Growth of the Mind. Translated by R. M. Ogden. London: K. Paul, Trench, Trubner, 1924. An introduction to child psychology from the Gestalt viewpoint.

Koffka, Kurt. Principles of Gestalt Psychology. New York: Harcourt Brace, 1935. A comprehensive survey of all the relevant experimental work up to 1935, accompanied by theoretical discussion of wide scope. Not easy reading, but a standard text. See pp. 655–661 for discussion of knowledge of other minds.

Köhler, Wolfgang. Dynamics in Psychology. London, 1942. An amplified version of lectures given at the University of Virginia in 1958.

Köhler, Wolfgang. Gestalt Psychology: An Introduction to New Concepts in Modern Psychology. New York: Liveright, 1947. An interesting although somewhat discursive discussion of general principles. See Ch. 17 for the problem of knowledge of other minds.

Köhler, Wolfgang. The Mentality of Apes. Translated from the 2nd rev. German ed. by Ella Winter. London: K. Paul, Trench, Trubner, 1925; paperback ed., Harmondsworth, U.K., 1957. A classic work on problem solving. This work and his Gestalt Psychology (see above) are recommended as standard texts.

Köhler, Wolfgang. The Place of Value in a World of Facts. New York: Liveright, 1938. William James lectures. A scholarly discussion of philosophical issues.

Lewin, Kurt. Principles of Topological Psychology. Translated by Fritz Heider and Grace M. Heider. New York and London: McGraw-Hill, 1936. An attempt to apply Gestalt principles to the study of social situations.

Madden, Edward H. Philosophical Problems of Psychology. New York: Odyssey Press, 1962. Contains a critical examination of some of the key concepts of Gestalt theory.

Michotte, Albert. La perception de la causalité. Louvain, Belgium: Institut Supérieur de Philosophie, 1946 and 1954. Translated by T. R. Miles and E. Miles as The Perception of Causality. London: Basic, 1963. A systematic study of moving Gestalten, with special reference to the conditions in which we receive an impression of causality. See Chs. 12–13 for the problem of knowledge of other minds.

Rubin, Edgar. Synsoplevede Figurer; Studier i psykologisk Analyse. Copenhagen and Oslo: Gyldendal, Nordisk, 1915. Translated by Peter Collett as Visuell wahrgenommene Figuren. Copenhagen, Berlin, and London, 1921. An abridged version, translated by Michael Wertheimer, appears in Readings in Perception, edited by D. C. Beardslee and Michael Wertheimer. Princeton, NJ: Van Nostrand, 1958. Pp. 194–203. Interesting pioneer work.

Wertheimer, Max. "Experimentelle Studien über das Sehen von Bewegung." Zeitschrift für Psychologie, Part I, 71 (1912): 161–265. Of historical interest; commonly regarded as marking the beginning of the Gestalt movement.

Wertheimer, Max. Productive Thinking. Edited by S. E. Asch et al.. New York and London: Harper, 1945. An important contribution to the study of problem solving.

T. R. Miles (1967)