by Morton Hunt
Learning
For many centuries the study of how knowledge is acquired had been one of the chief interests of psychologist-philosophers and psychologists. But with the advent of the physiologist-psychologists and Wundt, most of it was stored in the attic of culture with other obsolete mentalist topics.
What little the physiologists and followers of Wundt said about learning was mostly secondhand associationism; they saw it as merely the linking or joining of bits of experience. The behaviorists made learning the central topic of their research—but only the mindless learning of SR conditioning; the higher-level mental processes involved in much human learning were ignored in favor of such calculations as the relationship between the number of reinforced trials and the strength of the established habit.
Among the contributions of the Gestaltists, and perhaps their greatest, was the restoration of meaning and thought to the study of learning. Although the Gestalt movement flourished only briefly in Germany and did not replace behaviorism in the United States, it revived and renovated the cognitive tradition and prepared the way for the cognitive revolution of the 1960s.
It was not the human mind, however, but the mind of the hen that provided the first solid evidence that associationist and S-R theories of learning were seriously inadequate. Köhler, during his stay on Tenerife, conducted a tedious but enlightening experiment with four chickens. He allowed two of them to peck at grain scattered on a light gray square of paper but shooed them away whenever they tried to peck at grain on a darker gray square of paper. He gave the other two chickens the opposite treatment. Chickens are notoriously stupid, but after four hundred to six hundred trials the first two would peck only at grain on the lighter paper and the second two only at grain on the darker paper.
Köhler then altered both situations. He kept the background color the chickens had been trained to eat from but replaced the other one, substituting a still lighter paper in the first case, a still darker one in the second. Associationist and S-R theory would predict that since the chickens had learned to associate eating with a particular shade of gray, they would continue to do so, but in 70 percent of the trials they pecked at grain on the new backgrounds rather than the old ones. The pair that had been trained to eat from the lighter of two backgrounds now mostly chose the new, still lighter background; the two who had been trained to peck at the darker of two backgrounds now mostly chose the new, still darker background. Gestalt theory offered an answer: The chickens had learned to associate food not with a specific color but with a relationship— in one case the lighter, in the other case the darker, of two backgrounds.32
Köhler repeated the experiment with chimpanzees and with a three-year-old child. He presented each with two boxes, one of a dull color, the other of a bright color. When a chimpanzee was the subject, the bright-colored box had a bit of food in it; when the child was the subject, a bit of candy. After the chimpanzee and the child learned that the bright box contained the reward, Köhler eliminated the dull box and substituted a new one, even brighter than the reward box. This time he put a reward in both boxes so that there was no incentive for the subjects to choose either except its color relationship to the other—and in fact the chimpanzees and the child usually chose the new, brighter box.
Behaviorists and Wundt’s followers had known that an animal can be trained to choose one of two different-colored objects, but had refused to believe that what the animal learned was the relationship between the colors. To these “elementalist” psychologists, a relationship could not be a primary psychological fact. As Solomon Asch, a student of Wertheimer’s, observed, “This premise was sufficiently potent to blot out the ceaseless evidence of experience.”33
But Köhler’s experiment showed conclusively that the relationship between the colors was indeed the primary fact the animals had learned, since they transposed it to a different situation.34 It was an example of the general rule, said Asch, that animals and humans perceive and learn nearly everything in terms of relationships. This object stands on top of that one, is between two others, is bigger than, smaller than, earlier or later than another, and so on. Relations are the key to perception, learning, and memory. That truth had been excluded from psychology but was reinstated by the Gestaltists.
Wertheimer, Köhler, Koffka, and many of their students did research on learning, but much of the credit for promulgating their cognitive view of it goes to Koffka. That shy, self-doubting, homely little man with his odd, high-pitched voice was at his best when assembling facts and theory on paper; in print he could be masterful and scathing.
Koffka himself conducted no noteworthy research on learning; nearly all his experimental work was on the perception of depth, color, and motion. But because his English was excellent, the editor of the Psychological Bulletin, Robert M. Ogden (who had studied with Koffka at Würzburg), invited him to prepare the first account in English of Gestalt psychology. It appeared in 1922; from then on Koffka was the unofficial spokesman of the movement. Largely through his journal articles and two books, the research findings and ideas of the Gestaltists about learning became known to the profession.
In one of those books, The Growth of the Mind, published in German in 1921 and English in 1924, Koffka reviewed existing knowledge about mental development from a Gestaltist viewpoint. Of the many new ideas and interpretations he offered, two stand out.
The first: Instinctive behavior is not a chain of reflexive responses mechanically triggered by a stimulus; rather, it is a group or pattern of reflexes—a Gestalt imposed by the creature on its own actions—aimed at achieving a particular goal. A young chick pecks at certain things that it “knows” are edible, but the instinct is goal-oriented, driven by hunger, not a mechanical and automatic response to the sight of food.35 The chick does not peck when sated, despite the sight of food and the existence of the reflex.
The second: Against the behaviorist doctrine that all learning consists of chains of associations created by rewards, Koffka argued that much learning takes place through the processes of organization and reorganization in the mind in advance of reward; he offered as proof Köhler’s studies of problem solving by apes and comparable data on problem solving by children. But the exact cause of those organizing processes, he admitted, was not yet known.
Fourteen years later, in Principles of Gestalt Psychology (1935), a heroic attempt to review all existing knowledge of psychology from a Gestaltist viewpoint, Koffka was ready to offer a theory as to the cause of the organization and reorganization in the mind. The theory, elaborated from one originally proposed by Köhler, was that “psychophysical” forces inherent in the brain—neuronal energy fields—act like the force fields elsewhere in nature that always seek the simplest or best-fitting configuration (as in, again, the bubble, or the lines of force in a magnetic field). Hence the mind’s tendency to construct and reconstruct information in the form of “good Gestalten.”36
But are those good Gestalten faithful representations of the outside world? Koffka gave a resounding affirmative to this ancient question. He offered the theory, suggested by Wertheimer and developed by Köhler, that our thoughts about the world are isomorphic with the world itself— they are brain events that are, in some way, similar in structure to the external things they represent. If we see two separate lights, there are two separate areas of brain excitation; if we see movement, there is a corresponding movement in the field of arousal in the brain.37 The contents of the mind are not something wholly unlike the outside world but a neural simulacrum of it.
This solved the classic problem of how thought, a different kind of phenomenon from the material world, could represent that world. Or so it seemed to Koffka and his colleagues. But in the 1950s Karl Lashley and other neurophysiologists conducted experiments designed to interrupt the supposed electrical fields of isomorphic theory. They implanted mica plates in the visual cortex of some animals and in others placed silver foil on the surface of the brain, short-circuiting the different electrical pote
ntials that were supposed to simulate the perceived world. In neither case did the animals respond differently to visual experiences; isomorphism and force field theory was effectively scuttled.38
Yet if force field theory is viewed not as a physiological reality but as an illuminating metaphor, it has genuine value. It says that in a manner analogous to the operation of force fields, we group, categorize, and reorganize our experiences, always seeking the simplest and most meaningful constructs of the contents of our mind. As a guiding image, this comes closer than associationism, conditioning, or any earlier epistemological theory to describing how we perceive, learn, store, and utilize information. Field theory was not the ultimate truth, but it was a better approximation of the truth than earlier theories, and the basis of better approximations yet to come.
Memory is an aspect of epistemology about which Gestalt psychology offered some particularly useful and illuminating ideas.
One was the hypothesis, presented in some detail by Koffka, that the physiological basis of memory is the formation of “traces” in the central nervous system—permanent neural changes induced by experience. It was an acute guess; decades later, neurophysiologists would begin to discover the actual cellular and molecular changes that constitute traces.
Another keen guess dealt with the psychological basis of memory. Previously laid-down memory traces, Koffka said, influence how new experiences are perceived and remembered. Unlike associationism, which said that new experiences are merely added to old ones, Koffka said that new experiences interact with traces, traces with new experience, in ways not available to the mind early in life, and that this interaction is the cause of mental development.39 His idea would be borne out by a wealth of observational data that the Swiss child psychologist Jean Piaget was even then gathering.
Koffka marshaled a mass of experimental evidence to show that memory is not a mere sticking-together or aggregation of experiences, as in association theory, but a weaving together by means of the meaningful connections. Among the evidence he adduced was that of Ebbinghaus and his followers, that it is much harder to learn a string of nonsense syllables than a series of words connected by meaning. Koffka gave a simple and persuasive example: If every connection between items were merely one of association, these two lines would be equally easy to learn:
pud sol dap rus mik nom
A thing of beauty is a joy for ever.
Koffka’s comment: “It is not easy for association theory to explain why the second line is learned and retained so much more easily than the first, a difficulty which, as far as I know, was never explicitly mentioned by the associationists.”40
Like much else about Gestalt psychology, the truth illustrated by those two lines seems so obvious that one wonders why it needed to be rediscovered. But psychology has not moved in a steady course from ignorance toward knowledge; its progress has been more like that of an explorer of an unknown land who tentatively advances toward a distant goal by this valley or that, this river or that, and often must pursue a roundabout route or double back on his tracks when the chosen route proves a poor one. The followers of Wundt and the behaviorists made important headway toward the remote goal but went off into dead ends; the Gestaltists put psychology back on a truer course.
Boring made this point with a different metaphor in his magisterial history of psychology: “It appears that orthodoxy had been led astray along the straight and narrow path of sensory analysis. It is the wide gate and the broad way of phenomenology that lead to life.”41 Although the Gestalt psychologists were not the first or the only ones to make this discovery, it was they who made it in a form so convincing that it was amalgamated into the structure of scientific psychology.
Failure and Success
In Germany, as we saw, Gestalt psychology became a leading school in the 1920s but virtually disappeared in the mid-thirties after its three founders and many of their former students left Germany.
In the United States, after the publication of Koffka’s introductory article in 1922, Gestalt psychology met at first with great interest and even enthusiasm.42 Koffka and Köhler were asked to give seminars and colloquia at nearly all the important American research centers; Köhler was a visiting professor at Clark University in 1925; and Harvard later offered him a visiting professorship, which he had to decline.
But behaviorism was even then rapidly becoming the ruling brand of psychology in America, and there was no room in it for Gestaltist ideas. Most behaviorists saw Gestalt psychology as a regression to a discredited, unscientific nativism. To the extent that nativism means a belief in innate ideas, this was simply untrue. To the extent that nativism means a belief that the mind, by its very nature, imposes certain kinds of order on experience, it was correct. Gestalt theory was, in a way, a modern version of Kantian epistemology.43
Decades later this central tenet of Gestalt psychology would be strikingly confirmed by several forms of research. Studies of language acquisition, for instance, showed that children sense the grammatical structure of sentences and begin speaking in grammatical sentences long before they are taught anything about grammar. Even more remarkably, a study of deaf children who had not been taught any sign language found that when they were three or four years of age, they communicated by making up strings of gestures—quasi-sentences—that distinguished between agent, action, and object, just as verbal language does.44
The antipathy of behaviorists toward Gestalt psychology was reciprocated: Koffka, Köhler, and Wertheimer all were dismissive of behaviorism (and other psychologies) and presented their own approach as the only valid one, thereby offending many American psychologists. Reviewing the reception of Gestalt psychology in America, the psychologist Michael Sokal writes:
American psychologists were especially bothered by the attitude of the Gestaltists… Recently the term “Mandarin” has been used to characterize the attitudes and behavior of many of the German university professors of the period. In some ways the entire Gestalt movement represented a revolt against traditional German university culture, but in other, deeper ways the Gestaltists shared many traits typical of the faculties of German universities.45
The result was that by the early 1930s, Gestalt psychology, though it had become a definite part of the American psychological scene, remained a subordinate part; like the structuralists, functionalists, Freudians, and others, Gestaltists were a minority in a behaviorist-dominated establishment.46 Nevertheless, they had an influence on the development of psychology out of all proportion to their numbers and position.
Wertheimer, a warm and impassioned teacher, had a loyal but small following at the New School for Social Research, but no physical research facilities to speak of. Yet according to his distinguished student Abraham S. Luchins, during Wertheimer’s decade in America (he died in 1943) he was a “conspicuous and disquieting figure” in the behaviorist milieu.47
Koffka, though dry and overly theoretical as a teacher, was adored by the girls he taught at Smith. However, because the college’s emphasis was on undergraduate education, he supervised only one Ph.D. in his years there. But he did have an extensive effect on the psychological community through his writings, particularly the encyclopedic Principles of Gestalt Psychology, and he would undoubtedly have produced other influential works had his life not been cut short in 1941, at the age of fifty-five, by heart disease.
Köhler, despite his Germanic stiffness, was best able of the three to fit into the traditional academic framework. He created a center of psychological research and scholarship at Swarthmore that attracted a number of top-notch doctoral candidates, among them David Krech, Richard Crutchfield, Jacob Nachmias, and Ulric Neisser. Köhler retired in 1958 but remained active in research until his death at eighty, nine years later. After his retirement, he received the highest accolade of American psychology, election to the presidency of the American Psychological Association, an acknowledgment both of his personal achievements and of the contributions of the Gestalt movement to psych
ology.
For paradoxically, even though by midcentury the movement had lost its identity and was fading from view, its most important ideas had become part of the mainstream of psychology. Indeed, they remain a significant part of it today, although a number of Gestaltist ideas are now so taken for granted that they are rarely even identified as such when cited in textbooks of psychology.
The central Gestaltist doctrine, that the whole—the Gestalt— is greater than the sum of its parts and that it dominates our perceptions has stood the test of time and testing. In one recent experiment, psychologist David Navon measured the time it took observers to identify large and small letters in a display like this:
FIGURE 14
The “Forest Before Trees” Effect: It takes longer to identify the tiny letters than the bigger ones they make up.
The Gestaltists 347
Observers were able to name the large letters more swiftly than the small ones, whether or not the small ones were the same as the large ones they made up; in contrast, it took them longer to name the small letters when they were different from large ones made up of them.48 Evidently, the whole was recognized more easily than the parts it was made of.
Prägnanz, the tendency to see the simplest shape in complex patterns (see Figure 10, page 328), has held up as a valid perceptual principle. So has grouping (the Laws of Proximity and of Similarity, illustrated above on pages 326 and 327), although later research has extended and somewhat modified it.49
As for problem solving, although the reward-based, trial-and-error model espoused by behaviorism remains valid for many simpler animals, research with more intelligent animals and human beings has followed the direction taken by Köhler, Duncker, and Wertheimer. Newer models, based on information-processing theory, do not contradict Gestalt problem-solving theory so much as provide detailed programs of the step-by-step reasoning and searching for which Gestalt psychology had only such vague terms as “restructuring.”50