The Gestalt theory and “Insight”
Professor Wolfgang Köhler was one of the founders of the Gestalt psychology. The term “Gestalt psychology” was first used by Christian von Ehrenfels, the Austrian philosopher, as early as 1890 in his paper “On Gestalt Qualities”. The Gestaltism was a reaction to some contemporary psychologists. The Gestalt psychology emphasized that behavior cannot be studied by analyzing its components but, rather, it must be studied as a whole, in all its complexity.
Köhler carried out a series of famous experiments during the First World War, studying the behavior of chimpanzees in solving problems. The results of the experiments led him to introduce the notion of “insight” (Einsicht), which is one of the central terms of the Gestalt psychology. Insight, according to Köhler, manifests itself in the behavior as a break in the process of learning, i.e. as a real emergence of a solution: the correct adaptive response appears suddenly without being visibly dependent direct on the tests and errors that preceded it. Once it appears, this response is stable, i.e. the subject (hominid or human) is able to reproduce it in an appropriate manner.
Theoretically, a Gestalt-adept’s “intuition” is described as a restructuring of the perceptive field: the subject perceives suddenly in his environment new possibilities to action that lead him almost immediately to the solution. The phenomena of conditioning and learning cannot explain this capacity of intuitive restructuring. On the one hand their sequential character actually is directly opposed to the suddenness of insight; on the other hand their associationist and analytical character does not allow the inclusion of global and structural properties of the perception-cognitive field. This conception thus confers on the brain a global capacity to process the environments perceived. To the explicative construction in terms of combination of elements, it substitutes the notion of aptitude to immediate changes of structure.
The concept of insight led Duncker (1945) to proposing the general theory of problem solving. Duncker’s theory (see a separate section above) is based on the concepts of insight and field. For Duncker, a problem situation comes about when an organism has a goal but does not know how to reach it. The solution, which eliminates this situation by enabling the realization of the goal, generally consists of several successive stages, each marking a progress relative to the foregoing stage. The basic question then is how to advance from the one stage to the other. To answer it, Duncker reuses certain concepts developed by Selz (Mayer, 1992).
Selz proposed some of the major ideas in cognitive psychology and formulated the firs non-associationist theory of thinking. Among his seven theses on thinking processes the most remarkable ones are the following: Directed association generates the unit of thought; Forming a structure leads to understanding a problem; Testing for conditions results in solving a problem. Selz gave the following general description to the solution of any problem: Let ?Rb be the structure of a situation in which the task is to find what ‘?’ covers. Perhaps an individual keeps in his memory the complex aRb, which partially correspond to ‘?Rb’. From this, the element a is disclosed. Reaching the solution thus takes place according to a phenomenon of “resonance” between some generalized objects.
In general, “resonance” can be described as an original combination of known elements. The origin of the combination may be internal (it is the result of imagination), or external (induced verbally). In fact, not every correspondence of a signal to an object implies necessarily a resonance between the object and the signal. It is necessary that the signal be adequate to the object, which means that it should satisfy several conditions: it must be, among others, complete, concise, conformist, etc.
The process of problem solution depends closely on the structure of the situation and above all on the “availability” of the elements of which the solution consists. The elements, in fact, are not distributed randomly, but according to a characteristic organization, whereby some of them stand out. Duncker illustrates this notion by the problem of rays: how can a tumor be destroyed by means of radiation without damaging neighboring tissue? Two test subjects were used to answer the question. To the first group of test subjects the problem was described in the active voice: “The radiation might destroy also the healthy tissue. How can this be avoided?” The second group of test subjects got the same task, but this time expressed in the passive voice: “The healthy tissue might also be destroyed. How could it be protected from being damaged by radiation?” The results show that in the first group, 43 % of test subjects link their solution to the intensity of the radiation used. In the second group only 14 % focus on the intensity of radiation. The simple verbal structure of the problem description is sufficient for making the component “intensity of radiation” less prominent in the latter case than in the former.
Last but not least, what is essential in the process of problem solving is the change manifesting itself in that the subject operates in a psychological structure of the situation perceived. This sudden restructuring is nothing else than the insight, the role of which is decisive.
How then can the thought handle a given data field to extract from it pertinent information? One could say that the relation between two sets of data A and B is “totally distinguishable” if it is possible to grasp directly that the affirmation of A implies that of B. A connection will be “partially distinguishable if only certain traits of B can be grasped or understood parting from A. To explain the existence of such relationships there are, according to Duncker, two traditional responses: either the set B already is contained in A and can be analytically separated from it, or the functioning of the thought is such that that it always must reconnect A and B (a synthetic explication). If different events that compose the universe were not mutually related according to some law, the thought would evidently have no practical efficacy. In reality, these relations must be described not as fixed couplings of defined elements, but rather as a constant structure of variables. The principal cause of an effect “b” may be grasped by abstraction of factors that all possible situations “b” have in common, as well as those factors that all possible situations “non-b” do not have in common. It is a process of “precipitation” of a common aspect of coverage, parting from the inspection of a set of situations belonging to the same class. The result of the precipitation is a change in the configuration the situations envisaged. These situations thus acquire intelligibility that resides in the fact that they are reducible to a general law. This phenomenon can be called, writes Duncker, “insight of the second degree”.
The various elements of a situation, as has been seen, can present very variable degrees of “availability”. The availability of an object is actually an inverse function of what the author calls its “fixedness”. For instance, a chimpanzee looking for a stick from which to make a tool, like in Köhler’s experiment, does not recognize it in a branch if the latter still is attached to the tree. In this context the branch constitutes a part of the perceptive unit called “tree” and, because of this fixedness that anchors it in a definite situation, it is less available for entry into another situation than when it is detached and lies on the ground.
Beside this perceptive dependence, the fixedness can also be conditioned functionally. Several of Duncker’s well-known experiments reveal that. For instance, two groups of test subjects must discover in a heterogeneous set of objects one that can serve as a counterweight: in this case a book of logarithms. But, before that, one of the groups had first to resolve a mathematical problem requiring the use of the logarithmic tables. It was found that the members of this group less often identify the book in point as a counterweight than do the others. Because of its foregoing usage, which has “fixed” it as part of a certain function, it is no longer available for a new re-structuring.
Another question emerges as a consequence of these results: how can one explain that the perceptive restructurings necessary for a discovery are, according to the individuals, more or less easy? Taking mathematical reasoning as an example, Duncker distinguishes two possible applications:
Subjects weak in mathematics are not capable of restructuring easily, because their thinking matter is too rigid, strongly fixed.
With individuals gifted in mathematics, there is an abstract level on which the perceptive functions do not interfere, but where only specific mathematical properties are considered.
The former hypothesis has a degree of plausibility but it cannot be upheld: in fact, it can be stated that with subjects who are strictly mathematically oriented the perceptive restructuring plays a considerably greater part than with “ordinary” subjects. Duncker supports, however, the second explanation: it is the different role of perceptive structuring, where the difference refers to the degree with which the material of the thought is organized in depth. This causes the variations in individual performance. It is known that there are subjects of the visual type, with whom the intellectual activities always repose on visual images. By the same token, according to Duncker, many people are unable to explore and treat the problem material if it is not totally permeated with perceptive structures.
It has been shown that the Gestalt theoretical notion of insight makes it undeniably possible to describe a plurality of phenomena that are allowed in the problem solving behavior. The notion has fulfilled a certain scientific function in the psychology of the last 100+ years. This advantage is, nevertheless, negatively compensated by a major explicative inability: the concept of intuition rests, almost by definition, out of reach for analysis, and therefore it cannot be included in the causal chain. It thus generates many questions that cannot be answered. Pavlov had noted it immediately and reproached Köhler for having introduced the “dualism” and the “mentalism” in the study of higher processes.
If the notion of intuition largely appears as a theoretical instance, is it possible at least to give it precise objective content? Westcott tried to find the answer to this question (see further down).
In his 1968 work, Westcott studied intuition by resorting exclusively to the experimental method. He conceived it initially as a form of inferential behavior and defined it operationally as the fact that an individual solver has attained a solution using less information than is usually necessary. If then for solving a particular problem P, the majority of the population p needs the quantity of information I, any individual who attains the solution using a quantity of information that is less than I displays intuitive behavior vis-à-vis P. This definition, which is perfectly satisfactory from the scientific point of view, introduces the adoption of differential perspective: the human subjects are more or less intuitive in the same manner as they are more or less intelligent or creative.
In his first experimental research Westcott used four kinds of tasks defined by the increasing variety of the type of material (verbal or numerical) and of the type of response (completing series or establishing analogies). These four categories of problems did not require any specialized knowledge, which assured that the test subjects (in Westcott’s case students) could be viewed as equals. In solving tasks involving completion of series, the test subjects had at their disposal a set of examples the use of which improved the chances of attaining a solution. The same goes for the problems involving analogies, where the subject was allowed to use as many examples of relations as he pleased. Westcott called every example or additional stage that led to the goal the clue.
Each test subject was required to use both the least possible number of clues, and rank, on a four-point scale, the certainty he attaches to each of his responses.
Under these experimental conditions, Westcott defined four measures of performance:
The quantity of the information used. Instead of the coarse number of clues used by an individual before proposing a solution, it is preferable to ponder the value of the clues used as a function of the mean and the typical spread in the population.
The number of correctly solved problems of those twenty that constitute a complete series.
The ratio of success (foregoing score) to information demand (first score). This ratio is called efficacy.
The mean value of ratings assigned by the subjects to their responses in the twenty problems.
The eleven samples studied did not differ with respect to these four measures. In particular, no significant differences are visible between girls and boys. On the other hand, each sample displays a large variability with respect to the scores of the individuals comprising the group, which gives credibility to the idea that individual capabilities differ vastly from person to person.
The principal results can be summarized as follows:
No correlation exists between successful solving of the tasks and the quantity of information demanded, which a priori could look like a little paradox. But, rather, this result shows that intuition constitutes a differentiating feature: the use of information collected varies greatly from person to person and is independent of the quantity alone. What is important is not the volume of what is received, but rather the specific treatment that the input data receive.
Westcott was interested in subjects whose scores were extreme with respect to the two dimensions specified previously. He thus put together four groups corresponding to as many groups of individuals:
- those who require little information, yet often are successful. Such persons are by definition intuitive subjects, representing no more than 6 % of the total population.
- those who require little information, but are seldom successful. They are the “guessers”, free from a real intuitive process.
- those who need a great deal of information and are often successful. Their performance is prudent, as they never attempt anything if they are under informed.
- finally, there are those who require a great deal of information and fail most often. They are unable to use information correctly even if the information available is abundant.
In all samples studied, a positive correlation exists between success and the degree of confidence in the solutions proposed, which indicates that the test subjects are sufficiently aware of their performance. By contrast, the correlations between confidence and the quantity or information demanded are all negative. It means that the more information a subject demands, the less confidence he has in the responses he gives. Finally, the correlations between confidence and the efficacy score are all positive.
Westcott then tried to establish relations that might exist between the four scores mentioned and various aptitudes or personality traits measured by means of various tests. He found a negative correlation between the information demand and the verbal and numerical factors of intelligence, while success, too, seems to be positively related. This confirms the results showing that using intuitive thought seems to go hand in hand with high intelligence. As for personality traits, success is negatively correlated with anxiety. The same goes for the efficacy score, which is also positively related to flexibility. In other words, the subjects who need little information seem to be less stable and more engaged in what they are doing than are subjects who demand much information. The former are also less conformist and less sociable. Finally, the subjects who often succeed tend to be more confident in themselves, while those who often fail are more anxious. The capacity to resolve problems intuitively thus belongs to a specific set of personality traits.
In the second part of his research, Westcott was interested in the behavior of perceptive inference, conceived a different field of manifestation of the intuitive thought that does not differ fundamentally from problem solving. How do individuals categorize their universe of perceptions? Here, too, the quantity of the information demanded will be considered on the one hand, and the ability to formulate valid conclusions in a perception related task on the other hand.
The experimental material used consisted of several images borrowed from children’s books. Each image was cut into a series of seven or eight blocks arranged in the increasing order of completeness. The elements of information occurring in a specific block reappeared evidently in the neighboring block, whereupon the original design gradually re-emerged in its totality.
The pictures were presented individually to each subject, beginning by the least complete block, such as a simple sketch of some features. The subjects got the instruction to identify the design as soon as they could, i.e. using the least possible number of blocks. There were thus two different passing conditions: first, the images were presented in the increasing degree of completeness and for as long a time as the subject desired. He could give as many responses as he pleased until he succeeded. Evidently, two measures can be applied here: the mean of the number of responses at which the first attempt to identification is made, and the mean of the number of responses at which the correct answer is given.
In the second procedure the subjects could only give one response. The sequence was, however, broken as soon as response, good or wrong, was given. Three scores can be defined here: the mean of the responses corresponding to the demand of information in attempts to resolve the problems; the total number of correct identifications corresponding to the measure of success; and the ratio of the second to the first measure of efficacy.
The experiments were carried out on 17 samples, where the age of the test subjects ranged from tender childhood to adulthood. It was thus possible to study individual differences at a given age on the one hand, and the differences between groups as a function of age and the level of education.
Westcott has, for instance, established a correlation between the success at examinations and the IQ. This relation seems positive, which confirms the results of the foregoing experiment. Yet, this coarse piece of data for aptitude in mathematics is differentiated according to age. With the youngest ones this aptitude accompanies the higher scores of success and efficiency, while the students’ results show that this aptitude depends to a lesser degree on efficiency and on a higher demand of information, which does not necessarily accompany success.
In other words, Westcott studied the relations between intuitive performance and creativity. The method used consisted of requiring that the teachers teaching the test subjects explain explicitly their own definition of creativity and then rank their students in accordance with this definition. In one of the samples, for instance, the teacher defined creativity by contrasting it to simple imagination and supposed that it requires aggressiveness, initiative, and a minimum of intelligence. Of the 26 test subjects, four most creative and four least creative subjects were selected using this concept. As for the mean number of correct responses obtained, it was found that the most creative students appeared among the top seven individuals, and the least creative ones occupied the bottom five posts.
In another sample, the teacher defined creativity by distinguishing it from intelligence, and associating it with flexibility and original expression of oneself. In view of this definition five very creative students along with four very little creative students could be identified. The results obtained confirm those of the foregoing sample: the less creative ones occupy the four lowest tiers as for the number of correct responses provided in the test of perceptive inference. Thus, in the ensemble there is a strong correlation between intuitive behavior and the estimated individual creativity.
According to Westcott, the opacity that characterizes the phenomenon of intuition may be due to three causes:
it can follow from the fact that certain elements included in the process are inconscient;
it can follow from the fact that the relationships between these elements are inconscient;
or it can follow from the fact that these elements are comprised in the multiple and complex contexts obtained subliminally, peripherally or accidentally.
It is also necessary to note that the intuitive behavior is useful and necessary only in certain situations. In general, our environment is very redundant, it is well informed. It could be said that the same thing is true about creative behavior the emergence of which mostly remains an exceptional phenomenon.
Westcott’s results make it possible to give a real content to the notion of intuition: the definition of explicit experimental operations permits, at least partially, to eliminate myths about intuition and treat it as a behavior. Of course, these initial results must be largely completed, but they already sketch the path of additional approach to the creative process.