Of late much work has been done in the field of the schizophrenic thought
disturbance. In comparing many of these investigations with the concrete
clinical material one often has the feeling of a strange contrast between the
academic thinness of the former and the full richness of the latter. The
investigations are thorough, but the strange, sometimes beautiful vitality of
the original spontaneous material often seems to have escaped.
Gestalt Psychology assumes that this result is due to a fundamental
assumption common to most of the current approaches to thinking. This is the
hypothesis that thinking is essentially the piecemeal addition, by
"association", of a sum of basic elements which have no objectively
understandable, intrinsic logical relation to each other but are linked together
merely by blind habit - simply because of the past experience of their equally
unintelligible frequent coincidence. The classic example of this approach is the
nonsense syllable experiment.
In contrast to this, Gestalt theory assumes that original thinking is a
process of achieving a clear structural understanding of the organization of its
object or problem as a genuine whole, the parts of which are defined
functionally by their place and role in the whole, and not in piecemeal identity
as basically unrelated bits. Thinking is assumed to have its own whole-structure
and whole-dynamics which are lost if one focusses only on the 'elements'.
In the following pages an attempt has been made to understand a few features
of the schizophrenic thought disturbance on the basis of this theory.
"Can an introvert ever be an extrovert?" he answered.
This was said with a somewhat defiant, sarcastic grin. The doctor felt
uncomfortable. There was something strange in this answer.
How can such a case be completely understood? What factors decide that one
answer is right, another wrong? Certainly before dealing with so queer an
example one has to know something about the normal, the good case of question
and answer. What happens when a question is asked?
An ordinary question intends its answer. It calls for it, requires it.
In itself it is incomplete and establishes a vector towards completion.
Once a proper answer is given, question and answer form a complete closed
whole.  As long as the answer is missing the whole is incomplete, has
a gap which is not simply a hole but is a dynamic gap that needs and
wants to be filled. The question is not an isolated piece but the opening part
of an intended whole.
The questioner may not know the answer. A number of answers may be possible,
but not just any answer at all will fit into the gap.  If the question is "How is your health?" the answer, "Thank
you, two times two is four," does not fit.
Q = question
? = gap
Q: How is your health?
A: Two times two is four.
Obviously the question contains factors which determine what answer is
consistent and what is not. Firstly, the answer must have something to do with
the question, it must deal with the question's topic. But that does not
suffice. The answer, "My health depends on the number of calories I get," is
concerned with the same topic as the question, but still it does not fit. It
deviates from the direction of the question and is not a "good
continuation"  of this direction.  The vector set up by the question really tends in a different
logical direction, and the direction of the answer must be in good continuation
of the question in order to achieve its closure. The answers, "My health is
fine," or, "I have terrible pains," fit into the gap both with regard to the
identity of the topic and direction of the question. They meet the requirements
of the whole and complete it. In the other two cases the gap is not fittingly
filled, continues to be sensed, and the whole remains incomplete.
Q: How is your health?
A: It depends on the calories.
Q: How is your health?
A: It is fine.
This is true only for simple cases. In more complicated cases, as, for
instance, in that of a scientific question, the answer to which requires a
lengthy paper, detours involving temporary changes of the topic as well as of
the direction may become necessary. But what these changes are is not arbitrary
but determined by the inner nature and structure of the problem and by the
whole-structure of the problem of which each detour is a part. And, too, these
detours must fit into the complex question-answer system as a whole; they are
determined by, and must be consistent with, the structural requirements of the
Sometimes a change of topic and direction may be sensible, if, for instance,
the question itself does not go to the heart of the problem. The question may be
just too peripheral, too unessential, it may not fit right, it may not face the
problem squarely enough. If the answer improves upon the question in the
direction of the structural requirements of the problem situation it is a good
answer even if, or just because, it does not stick to the topic and direction of
the original question.
Glancing back at the patient's answer it can now be said that it is
irrelevant, wrong, queer, because it does not meet the structural requirements
of the intended question-answer system either with regard to the identity of
topic or with regard to the factor of good continuation. This answer is a case
of Fig. 2.
The question of what determined this strange answer remains. Why did he give
it? Do such answers also occur in normal cases and under what circumstances?
"How do you do, Dr. X.?"
"Thank you, I am having a drink right now."
Here one feels that somehow the answer makes sense, is possible, although
somewhat peculiar. Yet, in itself, and as long as no other data are given, the
answer neither fits the question with regard to the topic nor with regard to
good continuation of direction. It should sound irrelevant, but it does not. In
this case one feels at once that the question-answer system points to a
surrounding social situation as a part of which it must have occurred. It is
unnatural to look at it piecemeal, in artificial isolation. Actually this
conversation took place when two gentlemen met at a party which was already well
under way. The man who answered was just having a drink. Now the answer jumps
into place and fits. In its setting as a part of this situation it simply means
"I'm having a drink and I'm very well as you see."
It seems that in many cases one must not, and frequently simply is not able
to, look at such a question-answer system in a piecemeal fashion, in isolation
from the concrete social situation in which it arises. In these cases the question-answer
system is not an independently closed whole but essentially a functional part
of the field factors and field events which play a role in determining what
questions are being asked at a certain moment and what answers will fit. In
extreme cases a question-answer system may appear completely meaningless and
nonsensical as long as it is taken in isolation, while one grasps its meaning
at once if it is seen in its place and role within its social field. The inner
logic of the system remains hidden unless it is experienced as part of the dynamic
structure of the field. 
Go to 2nd part of this paper
 Only simple cases are covered by this simple
formulation. In more complex situations the formulation may have to be changed
without effecting the principle. [-> back to text]
 The features of structural fitting and requirement,
and of the gap, will be dealt with extensively in a forthcoming book by Max
WERTHEIMER on productive thinking. [-> back to text]
 WERTHEIMER, Max. Untersuchungen zur Lehre
von der Gestalt, II, Psychol. Forsch. (1923) 4:301-350 - in particular,
p. 324. There is an English abstract in Ellis, Willis D., A Source Book of
Gestalt Psychology; New York, Harcourt, Brace, 1938 (xiv and 403 pp); pp.
71-88 - in particular, pp. 81-83.
[-> back to text]
 MAIER, Norman R. F., Reasoning in Humans.
I. On Direction. J. Comparative Psychol. (1930) 10:115-143. [-> back to text]
 ELLIS, Willis D., A Source Book, reference
footnote 3; pp. 1-11 - in particular, p.6.
KOFFKA, Kurt, Principles of Gestalt Psychology; New York, Harcourt, Brace, 1935 (xi and 720 pp.) - in particular, p. 42.
LEWIN, Kurt, Principles of Topological Psychology; New York and London, McGraw-Hill, 1936 (xv and 231 pp.)
SCHULTE, Heinrich, Versuch einer Theorie der paranoischen Eigenbeziehung und Wahnbildung. Psychol. Forsch. (1924) 5:1-23. There is an English abstract in Ellis, Willis D., reference footnote 3, pp. 362-369.
LEVY, Erwin, A Case of Mania with Its Social Implications, Social Research (1936) 3:488-493.
[-> back to text]