the organism in its environment and study all its reactions. The task of Gestalt psychologists is to bring the facts of the unanalyzed and analyzed phenomena together without using the hypothesis of substantial mental elements. This is done by adhering strictly to a functional point of view. The analyzed datum, as well as the original one, must be considered as a reaction of the organism and the question to be asked is, "What is the reason for the occurrence of either of them?" The answer is found in the assumption that the conditions of the two cases must have been different, and inasmuch as the objective conditions are supposed to have been identical, the subjective ones must be supposed to have changed. You were looking at this room and were interested in it at the beginning. You were introspecting and interested in psychology afterwards. This change of attitude is clearly a change of subjective conditions. . . . Current psychology explains the influence of memory on perception by "assimilation." The Gestalt explanation is that the first reaction to a certain stimulation leaves the organism in an altered condition. Consequently, in a similar stimulussituation the reaction of the organism will not be the same as the first, and will not be as it would have been without the previous occurrence of the first response; the perception will have features resulting from the change of subjective conditions in the organism. This is a description of fact without the many hypotheses implied in the older explanation, it is logically clearer, and has much higher explanatory value.

. . As a consequence, we cannot assume a point to point coordination between the objective and the phenomenal world. A line by itself and a line of the same color, direction and length, but forming one side of an oblong or a triangle, are two different things.

. . . If we want to prove the reality of the descriptions of these lines . . . we can do so by a procedure which is the keystone to all experimental psychology, and the test of all psychological descriptions: we must try to develop functional consequences from our descriptive terms and find out whether such consequences are confirmed by fact. For example, take Köhler's training of the chimpanzee to react to one member of a pair of different grays. What the animal acquires is not an association between each shade of gray separately and the striving or avoiding reaction respectively, but it learns to react positively to "a member of the pair," i.e., to some thing characterized by its place or function in the larger whole embracing the two

units. If the pair be changed in such manner as one of the old members is retained in the new pair but occupies the relative place which the other one held in the first-being now, for instance, the lighter, when it was the darker-it then determines the reaction according to its place in the pair and not as an isolated stimulus whose effect is carried over from one combination to another. This is the test. This gives us the functional method by which we may decide whether a system is uniform or not, for we can investigate the range over which we can transpose without destroying the result of training. . . .

Instead of describing the element into which it can be analyzed we (Gestalt psychologists) maintain that the wholes are what they are precisely in their specific character as wholes and that their parts-for they nearly always contain parts-are not pieces thrown together, but real organic members. . . . Thus, the wholes, instead of being composed of their parts, really determine what the parts are to be. They have their own laws of being, by which are determined the kind of wholes that are stable, the kind that are prone to become transformed and the direction of such transformation. Four different directions of transformation may be distinguished: (1) simplification; (2) structuration; (3) unification; (4) amplification and their opposites.

How can introspection reveal true facts if it changes the facts? Some changes are in conformity with the inner laws of the wholes observed and some are opposed to them. If our attitude is such that it causes a change of the first kind, then the product of our introspection will be, not the original datum itself, but a development of it. If our attitudes are opposed to the inner laws of wholes, we get a product which in its essential features may be different from the original datum. This is why it is wrong to attempt to maintain an analytical attitude at all costs as the method of psychology. The attitude which is legitimate has to be determined by the nature of each separate whole dealt with, and in this appears the art of introspection. We must not treat sensations as a mental element.

The first essential point in Gestalt psychology is the meaningful connection between the parts of a whole, a connection not based on mere coexistence, but on the very essence of the wholes involved. The second essential is a psycho-physical one. Those wholes are not particularities of mind. If we begin to look for them, we find them everywhere in nature, therefore, we are

forced to assume such wholes in the nervous system, to consider psycho-physical process as such wholes whenever we have reasons that suggest such a view. . . .

The reactions of our mind are really but a part of our total behavior, and even descriptively a phenomenon very often points beyond itself, either backwards or forwards-as in comparing two noise intensities. . . . Consequently, we cannot cut off conscious processes from processes not accompanied by consciousness. We must assume that conscious processes are part processes of larger wholes and that, by pointing to other parts of the same whole they give us evidence that the physiological process is just such a whole as the mental one.

To derive functional facts from descriptive concepts, therefore, means that we follow the conscious part of a larger processwhole beyond its conscious limits, and predict from the part we know the way in which it will be continued. Functional and descriptive facts, then, belong closely together, and we can use the former to test the latter. . . . This explains why observational results of the behaviorists are very meager and tell us very little of the real activity of the animal observed. The activity-bits composing, in enormous quantity, the bulk of the behaviorist's studies, are presented as so many independent isolated, elementary actions and not, as they ought to be, as part processes of an embracing whole. Thus, the description, "This animal running away from some danger" is more valuable than the behaviorist's explanation. ..

Bodily and mental behavior are parts of larger wholes-even for mental facts. These wholes, being real, are not open to analytic observation. The observability of real behavior is then no longer so mysterious-if he adopts a different attitude. Bodily and mental states belonging to the same whole-process will in many properties be very similar to each other. This is chiefly true for the dynamic aspects: rhythm, accentuation, rate uniformity or change of direction, etc..

Thus psychology as the science of Behavior is again allowed to describe real behavior without exposing itself to the reproof of anthropomorphism. And such "behaviorism" would have no need to banish real introspection as a method alien to all other sciences, for observation of behavior and of phenomenal (introspection) will be essentially alike and will yield essentially similar results.

9. Some Contributions of Gestalt Psychology to Education [OGDEN, Robert Morris, "The Need of Some New Conceptions in Educational Theory and Practice," School and Society, September 10, 1923, Vol. 18, pp. 343-348.] (Slightly adapted.)

This splendid selection was written by a well-known psychologist of Cornell University who is probably the leading exponent in America of the new Gestalt psychology. The author attempts to get away from the common practice of basing our educational theory and practice upon more or less mechanistic concepts that pervade much of contemporary psychology in which life is interpreted in terms of stimulus and response. He would interpret it more in terms of configuration, in which the members of a structure are what they are by virtue of their place in the whole.

The author considers (1) the conception of original nature; (2) the conception of learning; (3) the conception of individual differences; and (4) the conception of the school program. The author gives an alternative to the views commonly accepted on these four subjects.

The science of education is markedly affected by principles of mechanistic interpretation of life. There is another hypothesis equally consonant with scientific method and procedure. A mechanism is an assemblage of parts which work together by virtue of their. juxaposition. The criticism of mechanism as applied to life is similar to McDougall's, namely, that the (1) notion of mechanism is derived from our knowledges of the working of machines and that machines are an artificial product of man's ingenuity and not in any sense a product of unrestrained nature; (2) in that it is an outcome of a mathematical way of thinking about life, a way of adding and subtracting integers, which as a matter of fact we never find through observation. The idea that the universe is an arbitrary association of parts carries with it in the corollary that everything can be analyzed into ultimate or irreducible elements, and that these elements, by more or less chance concatenations, bundle themselves together and thus constitute both the things we perceive and our reactions to them.

The alternative: that nature works not with elements in arbitrary juxtaposition but with figures and with forms; these are the characteristic entities of all science and of all being,

and they cannot be fully explained as arbitrary combinations of parts. On the contrary, the conception of a figure or form is based upon an integration, not only of parts but of members whose reality involves the membership-character of belonging to some whole. Thus there are no entities which exist in their own right, since units are always relative to higher degrees of unity.

Original Nature.-The educator's point of departure is the organism as he finds it. Educating is a psycho-biological problem. In order to solve it, we must have some idea of the manifold influences of heredity and environment upon the organism we are educating. The potency and synaptic resistance theory makes the reflex arc the basic mechanism of behavior, and habit formation the goal. By virtue of nervous connections existing at birth the intricacy and number of possible connections between nerves are such that a multiple causation takes place with reference to many responses, and likewise multiple responses are aroused by a single type causation. The mechanists hold that this leaves room for "stamping in" of certain connections, while disuse permits others to atrophy and disappear.

The conditioned reflex theory attempts to show that a substitution of stimuli may take place whenever an alien stimulus is associated with another to which the organism by virtue of its inherited nature is predisposed to respond. The mechanist holds that all reactions of an organism are forced, and substitution through juxtaposition is the sole means of acquiring a modification of response.

As an alternative we suggest that the stimulus is never a discrete entity; that it operates only as a formal pattern whose configuration, though ill defined in the early experience of the individual, is effective, nevertheless, rather than as a discrete stimulus or as an aggregate of stimuli. Accordingly, the nervous structure of the organism is thrown into similar patterns of stresses and strains, and these corresponding patterns of nervous energy eventuate in the overt response or manifest behavior of the organism.

The characteristic of behavior is therefore one of uniformity rather than one of chaotic multiple responses. Similarly, the characteristic of perception, or the stimulation of the organism, is the pattern rather than a summation of stimuli of the multiple-causation type.

It follows that instinct is more typical of original nature than the reflex. The reflex approaches, though it never quite