GESTALT PSYCHOLOGY TODAY 1
WOLFGANG KOHLER
Dartmouth College
I
N 1949, the late Herbert Langfeld gave a lec- their individual stimuli and must therefore change
ture in Europe in which he described what when these are greatly changed. How, then, could
appeared to him to be the major trends in any characteristics of the perceptual situation re-
American psychology. He also mentioned Gestalt main constant under these conditions? Where did
psychology; but he added that the main observa- the Gestalt qualities come from? Ehrenfels' quali-
tions, questions, and principles characteristic of this ties are not fancy ingredients of this or that par-
school had become part of every American psy- ticular situation which we might safely ignore.
chologist's mental equipment. I was not so opti- Both positive and negative esthetic characteristics
mistic. And, in fact, the very next year attempts of the world around us, not only of ornaments,
were made to explain the molar units in perception paintings, sculptures, tunes, and so forth, but also
by processes which gradually connect neural ele- of trees, landscapes, houses, cars—and other per-
ments. Soon afterwards, a theory of conditioning sons—belong to this class. That relations between
was developed, according to which more and more the sexes largely depend on specimens of the same
components of a stimulus object are gradually con- class need hardly be emphasized. It is, therefore,
ditioned, and the course of the whole process can not safe to deal with problems of psychology as
be explained in this fashion. Such theories may though there were no such qualities. And yet, be-
prove to be very useful, but one can hardly say ginning with Ehrenfels himself, psychologists have
that, at the time, their authors were greatly influ- not been able to explain their nature.
enced by Gestalt psychology. It is for this and This holds also for the men who were later called
similar reasons that a new discussion of old ques- Gestalt psychologists, including the present speaker.
tions seems to me indicated. Wertheimer's ideas and investigations developed in
a different direction. His thinking was also more
radical than that of Ehrenfels. He did not ask:
I should like to begin with a few remarks about How are Gestalt qualities possible when, basically,
the history of Gestalt psychology—because not all the perceptual scene consists of separate elements?
chapters of this history are generally known. In Rather, he objected to this premise, the thesis that
the eighties of the past century, psychologists in the psychologist's thinking must begin with a con-
Europe were greatly disturbed by von Ehrenfels' sideration of such elements. From a subjective
claim that thousands of percepts have character- point of view, he felt, it may be tempting to as-
istics which cannot be derived from the character- sume that all perceptual situations consist of inde-
istics of their ultimate components, the so-called pendent, very small components. For, on this as-
sensations. Chords and melodies in hearing, the sumption, we obtain a maximally clear picture of
shape characteristics of visual objects, the rough- what lies behind the observed facts. But, how do
ness or the smoothness of tactual impressions, and we know that a subjective clarity of this kind
so forth were used as examples. All these "Gestalt agrees with the nature of what we have before us?
qualities" have one thing in common. When the Perhaps we pay for the subjective clearness of the
physical stimuli in question are considerably customary picture by ignoring all processes, all
changed, while their relations are kept constant, functional interrelations, which may have operated
the Gestalt qualities remain about the same. But, before there is a perceptual scene and which thus
at the time, it was generally assumed that the influence the characteristics of this scene. Are we
sensations involved are individually determined by allowed to impose on perception an extreme sim-
1 plicity which, objectively, it may not possess?
Address of the President at the sixty-seventh Annual
Convention of the American Psychological Association, Wertheimer, we remember, began to reason in
Cincinnati, Ohio, September 6, 1959. this fashion when experimenting not with percep-
727
"J
•"*, , » ••' '
728 THE AMERICAN * PSYCHOLOGIST
tual situations which were stationary, and there- Gestalt psychologists were not satisfied with a quiet
fore comparatively silent, but with visual objects consideration of available facts. It seems that no
in motion when corresponding stimuli did not move. major new trend in a science ever is. We were
Such "apparent movements," we would now say, excited by what we found, and even more by the
occur when several visual objects appear or disap- prospect of finding further revealing facts. More-
pear in certain temporal relations. Again in our over, it was not only the stimulating newness of
present language, under these circumstances an in- our enterprise which inspired us. There was also
teraction takes place which, for instance, makes a a great wave of relief—as though we were escaping
second object appear too near, or coincident with, from a prison. The prison was psychology as
a first object which is just disappearing, so that taught at the universities when we still were stu-
only when the first object, and therefore the inter- dents. At the time, we had been shocked by the
action, really fades, the second object can move to- thesis that all psychological facts (not only those
ward its normal position. If this is interaction, it in perception) consist of unrelated inert atoms and
does not, as such, occur on the perceptual scene. that almost the only factors which combine these
On this scene, we merely observe a movement. atoms and thus introduce action are associations
That movements of this kind do not correspond to formed under the influence of mere contiguity.
real movements of the stimulus objects and must What had disturbed us was the utter senselessness
therefore be brought about by the sequence of the of this picture, and the implication that human life,
two objects, we can discover only by examining the apparently so colorful and so intensely dynamic, is
physical situation. It follows that, if the seen actually a frightful bore. This was not true of our
movement is the perceptual result of an interaction, new picture, and we felt that further discoveries
this interaction itself takes place outside the per- were bound to destroy what was left of the old
ceptual field. Thus, the apparent movement con- picture.
firmed Wertheimer's more general suspicion: we Soon further investigations, not all of them done
cannot assume that the perceptual scene is an ag- by Gestalt psychologists, reinforced the new trend.
gregate of unrelated elements because underlying Rubin called attention to the difference between
processes are already functionally interrelated when figure and ground. David Katz found ample evi-
that scene emerges, and now exhibits corresponding dence for the role of Gestalt factors in the field of
effects. touch as well as in color vision, and so forth. Why
Wertheimer did not offer a more specific physio- so much interest just in perception? Simply be-
logical explanation. At the time, this would have cause in no other part of psychology are facts so
been impossible. He next turned to the problem readily accessible to observation. It was the hope
of whether the characteristics of stationary percep- of everybody that, once some major functional
tual fields are also influenced by interactions. I principles had been revealed in this part of psy-
need not repeat how he investigated the formation chology, similar principles would prove to be rele-
of molar perceptual units, and more particularly of vant to other parts, such as memory, learning,
groups of such objects. Patterns which he used thinking, and motivation. In fact, Wertheimer
for this purpose are now reproduced in many text- and I undertook our early studies of intellectual
books. They clearly demonstrate that it is rela- processes precisely from this point of view; some-
tions among visual objects which decide what ob- what later, Kurt Lewin began his investigations of
jects become group members, and what others do motivation which, in part, followed the same line;
not, and where, therefore, one group separates itself and we also applied the concept of Gestaltung or
from another. This fact strongly suggests that per- organization to memory, to learning, and to recall.
ceptual groups are established by interactions; and, With developments in America, Wertheimer's fur-
since a naive observer is merely aware of the re- ther analysis of thinking, Asch's and Heider's in-
sult, the perceived groups, but not of their depend- vestigations in social psychology, our work on
ence upon particular relations, such interactions figural aftereffects, and eventually on currents of
would again occur among the underlying processes the brain, we are probably all familiar.
rather than within the perceptual field. In the meantime, unexpected support had come
Let me add a further remark about this early from natural science. To mention only one point:
stage of the development. Surely, in those years, Parts of molar perceptual units often have charac-
GESTALT PSYCHOLOGY TODAY 729
teristics which they do not exhibit when separated the same statement in one of his papers. And, only
from those units. Within a larger visual entity, a a few weeks ago, I read a paper in which Bridgman
part may, for instance, be a corner of this entity, of Harvard interprets Heisenberg's famous principle
another part its contour or boundary, and so on. It in such terms that I am tempted to call him, Bridg-
now seems obvious; but nobody in psychology had man, a Gestalt physicist.
seen it before: the same happens in any physical We will now return to psychology. More par-
system that is pervaded by interactions. These in- ticularly, we will inspect the situation in which
teractions affect the parts of the system until, even- American psychology finds itself today. The spirit
tually, in a steady state, the characteristics of all which we find here differs considerably from the
parts are such that remaining interactions balance one which characterized young Gestalt psychology.
one another. Hence, if processes in the central Let me try to formulate what members of this audi-
nervous system follow the same rule, the depend- ence may have been thinking while I described that
ence of local perceptual facts on conditions in European enterprise. "Enthusiasm?" they probably
larger entities could no longer be regarded as puz- thought. "Feelings of relief when certain assump-
zling. Comparisons of this kind greatly encour- tions were found less dreary than those of earlier
aged the Gestalt psychologists. psychologists in Europe? But this is an admis-
In America, it may seem surprising that enthusi- sion that emotional factors and extrascientific values
astic people such as the Gestalt psychologists were played a part in Gestalt psychology. We know
intensely interested in physics. Physics is gener- about the often pernicious effects of the emotions
ally assumed to be a particularly sober discipline. in ordinary life. How, then, could emotions be per-
And yet, this happened to us most naturally. mitted to influence scientific judgments and thus to
To be sure, our reasoning in physics involved no disturb the objectivity of research? As we see it,
changes in the laws of physics, and no new assump- the true spirit of science is a critical spirit. Our
tions in this field. Nevertheless, when we compared main obligation as scientists is that of avoiding
our psychological findings with the behavior of cer- mistakes. Hence our emphasis on strict method in
tain physical systems, some parts of natural science experimentation and on equally strict procedures in
began to look different. When reading the formulae the evaluation of results. The Gestalt psychologists
of the physicist, one may emphasize this or that as- seem to have been guilty of wishful thinking. Un-
pect of their content. The particular aspect of the der the circumstances, were not some of their find-
formulae in which the Gestalt psychologists became ings unreliable and some of their concepts vague?"
interested had, for decades, been given little atten- I will at once admit two facts. Almost from its
tion. No mistake had ever been made in applica- beginning, American psychology has given more at-
tions of the formulae, because what now fascinated tention to questions of method and strict proof than
us had all the time been present in their mathe- Gestalt psychology did in those years. In this re-
matical form. Hence, all calculations in physics spect, American psychology was clearly superior.
had come out right. But it does make a difference Secondly, sometimes the Gestalt psychologists did
whether you make explicit what a formula implies make mistakes. Not in all cases was the reliability
or merely use it as a reliable tool. We had, there- of their findings up to American standards, and
fore, good reasons for being surprised by what we some concepts which they used were not immedi-
found; and we naturally felt elated when the new ately quite clear. I myself once used a certain con-
reading of the formulae told us that organization is cept in a somewhat misleading fashion. I had bet-
as obvious in some parts of physics as it is in psy- ter explain this.
chology. What is insight? In its strict sense, the term
Incidentally, others were no less interested in this refers to the fact that, when we are aware of a re-
"new reading" than we were. These other people lation, of any relation, this relation is not experi-
were eminent physicists. Max Planck once told me enced as a fact by itself, but rather as something
that he expected our approach to clarify a difficult that follows from the characteristics of the objects
issue which had just arisen in quantum physics— under consideration. Now, when primates try to
if not the concept of the quantum itself. Several solve a problem, their behavior often shows that
years later, Max Born, the great physicist who gave they are aware of a certain important relation. But
quantum mechanics its present form, made almost when they now make use of this "insight," and thus
730 THE AMERICAN PSYCHOLOGIST
solve their problem, should this achievement be entist, as though little else counted. They are
called a solution by insight? No—it is by no means necessary in research, just as the brakes in our
clear that it was also insight which made that par- cars must be kept in order and their windshields
ticular relation emerge. In a given situation, we clean. But it is not because of the brakes or of the
or a monkey may become aware of a great many windshields that we drive. Similarly, caution and
relations. If, at a certain moment, we or a monkey a critical spirit are like tools. They ought to be
attend to the right one, this may happen for several kept ready during a scientific enterprise; however,
reasons, some entirely unrelated to insight. Conse- the main business of a science is gaining more and
quently, it is misleading to call the whole process a more new knowledge. I wonder why great men in
"solution by insight." physics do not call caution and a critical spirit the
This will be particularly obvious when the solu- most important characteristics of their behavior.
tion of the problem is arbitrarily chosen by the They seem to regard the testing of brakes and the
experimenter. Take Harlow's excellent experiments cleaning of windshields as mere precautions, but to
in which primates are expected to choose the odd look forward to the next trip as the business for
item in a group of objects. "Oddity" is a particular which they have cars. Why is it only in psychol-
relational fact. Once a monkey attends to it, he ogy that we hear the slightly discouraging story of
will perceive it with insight. But why should he mere caution over and over again? Why are just
do so during his first trials? His first choices will psychologists so inclined to greet the announcement
be determined by one factor or another, until he of a new fact (or a new working hypothesis) al-
happens to attend, once or repeatedly, to the oddity most with scorn? This is caution that has gone
relation just when he chooses (or does not choose) sour and has almost become negativism—which, of
the right object. Gradually, he will now attend to course, is no less an emotional attitude than is en-
this particular relation in all trials; and he may thusiasm. The enthusiasm of the early Gestalt psy-
do so even when entirely new objects are shown. chologists was a virtue, because it led to new ob-
Surely, such a process should not simply be called servations. But virtues, it has been said, tend to
"learning by insight." If Harlow were to say that, breed little accompanying vices. In their enthusi-
under the circumstances, it is learning of one kind asm, the Gestalt psychologists were not always suffi-
or another which gives the right relation and corre- ciently careful.
sponding insight their chance to operate, I should In American psychology, it is rightly regarded as
at once agree. What, I believe, the monkeys do a virtue if a man feels great respect for method and
not learn is insight into which object in a given for caution. But, if this virtue becomes too strong,
group is the odd one; but they must learn to pay it may bring forth a spirit of skepticism and thus
attention to the oddity factor in the first place. I prevent new work. Too many young psychologists,
hope that this will clarify matters. They have not it seems to me, either work only against something
always been so clear to me. done by others or merely vary slightly what others
When the solution of a problem is not arbitrarily have done before; in other words, preoccupation
chosen by the experimenter, but more directly re- with method may tend to limit the range of our re-
lated to the nature of the given situation, insight search. We are, of course, after clear evidence.
may play a more important role. But, even under But not in all parts of psychology can evidence im-
these circumstances, it is not insight alone which mediately be clear. In some, we cannot yet use
brings about the solution. The mere fact that solu- our most exact methods. Where this happens, we
tions often emerge to the subjects' own surprise is hesitate to proceed. Experimentalists in particular
clear proof that it cannot be insight alone which is tend to avoid work on new materials resistant to
responsible for their origin. approved methods and to the immediate applica-
tion of perfectly clear concepts. But concepts in a
new field can only be clarified by work in this field.
But I intended to discuss some trends in Ameri- Should we limit our studies to areas already fa-
can psychology. May I confess that I do not fully miliar from previous research? Obviously, this
approve of all these trends? would mean a kind of conservatism in psychology.
First, I doubt whether it is advisable to regard When I was his student, Max Planck repeated this
caution and a critical spirit as the virtues of a sci- warning over and over again in his lectures.
GESTALT PSYCHOLOGY TODAY 731
Our wish to use only perfect methods and clear term "force." And what did the physicists do?
concepts has led to Methodological Behaviorism. They worked and worked on it, until at last it did
Human experience in the phenomenological sense become perfectly clear. There is no other way of
cannot yet be treated with our most reliable meth- dealing with new, and therefore not yet perfect,
ods; and, when dealing with it, we may be forced concepts. Hence, if we refuse to study the phe-
to form new concepts which, at first, will often be nomenal scene, because, here, few concepts are so
a bit vague. Most experimentalists, therefore, re- far entirely clear, we thereby decide that this scene
frain from observing, or even from referring to, the will never be investigated—at least not by us, the
phenomenal scene. And yet, this is the-scene on psychologists.
which, so far as the actors are concerned, the drama Now, I had better return to Gestalt psychology.
of ordinary human living is being played all the Let me try to show you how Gestalt psychology
time. If we never study this scene, but insist on tends to work today by discussing a more specific
methods and concepts developed in research "from issue, an issue on which scores of American psy-
the outside," our results are likely to look strange chologists have worked for years. We shall thus
to those who intensely live "inside." be enabled to compare the way in which they ap-
To be sure, in many respects, the graphs and proach this issue with the Gestalt psychologists' ap-
tables obtained "from the outside" constitute a proach.
most satisfactory material; and, in animal psy- The issue in question refers to the concepts of
chology, we have no other material. But this ma- conditioning and motivation. One school seems to
terial as such contains no direct evidence as to the regard conditioning as almost the process with
processes by which it is brought about. In this re- which the psychologist has to deal. In a famous
spect it is a slightly defective, I am tempted to say, book with the general title Principles of Behavior,
a meager, material. For it owes its particular the late Clark Hull, then the most influential mem-
clearness to the fact that the data from which the ber of the school, actually dealt with little else—
graphs and tables are derived are severely selected although he often used other terms. He felt that
data. When subjects are told to say no more than even such facts as thinking, insight, intentions,
"louder," "softer," and perhaps "equal" in certain striving, and value would eventually be explained
experiments, or when we merely count how many by a consistent investigation of the various forms
items they recall in others, then we can surely apply of conditioning. We are all familiar with the basic
precise statistical techniques to what they do. But, concepts of his theory. Hence I will say only a
as a less attractive consequence, we never hear un- few words about it. When conditions in an ani-
der these circumstances how they do the comparing mal's tissue deviate from an optimal level, a state
in the first case and what happens when they try of need is said to exist in this tissue. Such needs
to recall in the second case. produce, or simply are, drives—which means that
Are such questions now to be ignored? After all, they tend to cause actions in the nervous system,
not all phenomenal experiences are entirely vague; some more or less prescribed by inherited neural
this Scheerer has rightly emphasized. And, if connections, others of a more random nature.
many are not yet accessible to quantitative pro- Drives are also called motivations. None of these
cedures, what of it? One of the most fascinating terms is to be understood in a phenomenological
disciplines, developmental physiology, the science sense. They always refer to assumed states of the
investigating the growth of an organism from one tissue. The main point is that, for biological rea-
cell, seldom uses quantitative techniques. And yet, sons, states of need must, if possible, be reduced
nobody can deny that its merely qualitative de- and that this may be achieved by certain responses
scription of morphogenesis has extraordinary scien- of the organism to the given situation. In case
tific value. In new fields, not only quantitative first responses are of a random character, learning
data are relevant. As to the initial vagueness of or conditioning will often select such responses as
concepts in a new field, I should like to add an do reduce the needs in question. In a simple for-
historical remark. When the concept of energy mulation, the well-known rule which governs such
was first introduced in physics, it was far from developments is as follows: when a response has re-
being a clear concept. For decades, its meaning peatedly occurred in temporal contiguity with the
could not be sharply distinguished from that of the neural effects of a certain stimulus, then this stimu-
732 THE AMERICAN PSYCHOLOGIST
lus will tend to evoke the same response in the fu- achieved cause in the present speaker a mild, in-
ture—provided the response has caused a reduc- credulous horror. It is not the business of science
tion of the need. I will not define such further con- to destroy evidence. Behaviorists would perhaps
cepts as habit strength, reaction potential, afferent answer that arguments which refer to human think-
stimulus interaction, reactive inhibition, and so ing as an experience are irrelevant, because science
forth, because they will play no role in my discus- is only concerned with facts observable from the
sion. outside, and therefore objective. This answer
But one term seems to me particularly impor- would hardly be acceptable. The Behaviorist's own
tant. Many recent, and important, investigations objective observations are invariably observations
are concerned with so-called "learned drives," an of facts in his perceptual field. No other form of
expression which has, of course, this meaning: if objective observation has ever been discovered.
a neutral stimulus is repeatedly followed by condi- Consequently, the Behaviorist cannot, without giv-
tions which cause a primary state of drive such as ing more particular reasons, reject reference to
pain, and the corresponding fear, then the fear with other individual experiences merely because they
its usual effects on behavior will gradually become are such experiences.
connected with that neutral stimulus, so that the Thus we are justified in considering a further
stimulus alone now evokes the fear and its overt example of human experience. A need or drive,
consequences. Certain drives are therefore said to we are sometimes told, is a motivation. I do not
be "learnable" in the sense that they can be at- entirely agree with this statement for the following
tached to facts which, as such, are not related to reasons. A need or drive, we remember, is sup-
the drive and hence would originally not evoke cor- posed to be a particular state in the tissue. There
responding responses. is no indication in Hull's writings that such a state
Some experiments in the field of conditioning in "points beyond itself" toward any objects—al-
general are most interesting. I will only discuss though it may, of course, cause movements, or ac-
the concepts used in the interpretation of this work tions of glands. Now it is true that the same holds
and the conclusions which it is said to justify. for certain needs as human experiences; because,
To begin with these conclusions: They refer to when a need is felt, it does not always point toward
certain human experiences which, if the conclusions an object, attainment of which would satisfy the
were justified, would have to be regarded as strange need. At the time, no such object may be in sight;
delusions. I mean our cognitive experiences. Sup- in fact, no such object may yet be known. But
pose somebody discovers by accident that, every when the proper object appears, or becomes known,
time he subtracts the square of a given integer then the situation changes. For, now the subject
from the square of the next integer in the series, feels attracted or (in certain instances) repelled by
the result is an odd number. A more learned friend this object. In other words, an object may have
now explains to him why this is a necessary rule, characteristics which establish a dynamic relation
undoubtedly valid beyond any tests ever done by between the subject and that object. According to
a person. The explanation refers to simple rela- common experience, it is this dynamic relation
tions and to relations among relations—all readily which makes the subject move toward, or away
understandable—and the final outcome is convinc- from, the object. We ought to use different terms
ing. Now, is the understanding of the relations in- for a mere need per se and the situation in which
volved to be explained in terms of conditioning? a subject is attracted or repelled by an object.
Nothing in conditioning seems to give us access to Otherwise, the dynamic aspect of the latter situa-
the psychological fact which I just called under- tion might easily be ignored. I suggest that we re-
standing; and, since an understanding of relations
serve the term "motivation" for this dynamic situa-
is essential to all cognitive achievements, the same
tion. Here we are, of course, on familiar ground.
applies to the whole field.
Explanation of our intellectual life in terms of Motivation as just described was Kurt Lewin's
conditioning would simply mean: its reduction to main concern in psychology. He clearly recognized
the operations of an often most practical, but in- the part which certain characteristics of an object
trinsically blind, connection of mere facts. Promises play in establishing the dynamic relation between
that such an explanation will nevertheless be this object and the subject. He called such charac-
GESTALT PSYCHOLOGY TODAY 733
teristics of objects Aufforderungscharaktere, a term We have just seen that the behavior of vectors in
which then became "valences" in English. motivational situations is the same as the behavior
So far as I know, there are no valences in objects, of forces in nature. Gestalt psychologists are, there-
no attractions and no repulsions between objects fore, inclined to interpret motivation in terms of
and subjects in the Behaviorist's vocabulary. I am such forces or, rather, of forces which operate be-
afraid that, in this fashion, he misses a point not tween certain perceptual processes and processes in
only important in human experience but also rele- another part of the brain, where a need may be
vant to what he regards as true science. physiologically represented. We have no time to
How would a Gestalt psychologist handle motiva- discuss the question how cortical fields or forces
tion in the present sense? He would begin with would cause overt movements of the organism in
the following psychological facts. I do not know the direction of these forces.
up to what point Lewin would have accepted what Now, not everybody likes the term "force." Its
I am now going to say. My facts are these: (a) meaning, it has been said, has anthropomorphic
In human experience, motivation is a dynamic connotations. But, in human psychology, we sim-
vector, that is, a fact which has a direction and ply must use terms which—if I may use this ex-
tends to cause a displacement in this direction. pression—"sound human." If we refused to do so,
(b) Unless there are obstacles in the way, this di- we would not do justice to our subject matter which
rection coincides with an imaginary straight line (to a high degree) is human experience. To be
drawn from the object to the subject, (c) The sure, in physics, Heinrich Hertz once tried to do
direction of the experienced vector is either that to- without the concept "force." He actually wrote
ward the object or away from it. In the first case, a treatise on mechanics in which he avoided this
the vector tends to reduce the distance in question; term. And what happened? He had to populate
in the second, to increase it. (d) The strength of the physical world with unobservable masses, in-
both the need present in the subject and of the troduced only in order to make their hidden pres-
valence exhibited by the object can vary. Both in ence substitute for the much simpler action of
man and in animals it has been observed that, forces. Ever since that time, physicists have hap-
when the strength of the valence is low, this reduc- pily returned to the old concept "force," and no-
tion can be compensated for by an increase of the body has ever been harmed by the fact.
need in the subject; and, conversely, that, when The present reasoning leads to a conclusion
the need is lowered, an increase of the strength of which distinguishes this reasoning from the treat-
the valence may compensate for this change. ment of motivation in the Behaviorist's system.
When considering these simple statements, any- Clark Hull was a great admirer of science; but, to
body familiar with the elements of physics will be my knowledge, he hardly ever used the concepts
reminded of the behavior of forces, (a) In phys- characteristic of field physics. The fundamental
ics, forces are dynamic vectors which tend to change distinction between physical facts which are scalars
the distance between one thing (or event) and an- (that is, facts which have a magnitude but no di-
other, (b) Unless there are obstacles in the way, rection) and vectors (which have both an intensity
a force operates along a straight line drawn from and a direction) played no decisive part in his
the first object (or event) to the other, (c) The theorizing. His main concepts were obviously
direction in which a force operates is either that of meant to be scalars. There is no particular spatial
an attraction or of a repulsion, of a reduction or of direction in a habit strength, none in a reaction po-
an increase of the given distance, (d) The formula tential, and none even in what he called a drive
by which the intensity of a force between two ob- state. Hence, the core of modern physics as de-
jects is given contains two terms which refer to the veloped by Faraday and Maxwell had no influence
sizes of a decisive property (for instance, an elec- on his system. For this reason, and also because
tric charge) in one object and in the other. It is he refused to consider motivation as an experienced
always the product of these two terms on which, vector, he could not discover that the operations of
according to the formula, the intensity of the force motivation appear to be isomorphic with those of
depends. Consequently, a reduction of the crucial fields or forces in the brain.
term on one side can be compensated for by an in- But, if motivation is to be interpreted in this
crease of the term on the other side. fashion, certain assumptions often made by Behav-
734 THE AMERICAN PSYCHOLOGIST
iorists may no longer be acceptable. Take the con- the subject might just learn more about the charac-
cept of learned drives. As I understand this term, teristics of the given object itself than he knew in
it means that learning can attach a drive state to the beginning; and the characteristics revealed by
a great variety of stimuli which, as such, are neu- this learning might be such that now the same ob-
tral facts. Now, so long as a drive is not regarded ject fits a need. It seems to me that all these possi-
as a vector, this seems indeed quite possible. But, bilities ought to be considered before we accept the
if the drive in Hull's sense is replaced by a motiva- thesis that motivations in the present sense can be
tional force which operates between a subject and attached to actually neutral objects. Incidentally,
some perceptual fact, no arbitrary connections of similar changes of objects may also be responsible
this kind can be established. For, now motivation for the developments which Gordon Allport once
becomes the experienced counterpart of a force in regarded as evidence of "functional autonomy."
the brain, and this force depends entirely upon the . You will ask me whether my suggestions lead to
relation between conditions in the subject and the any consequences in actual research. Most surely,
characteristics of the perceived object. There can they do. But, since I have lived so long in America,
be no such force if the object is, and remains, a and have therefore gradually become a most cau-
neutral object. Forces only operate between ob- tious scientist, I am now preparing myself for the
jects which have the right properties. Any ex- study of motivation by investigating, first of all,
ample of a force in nature illustrates this fact. the action of dynamic vectors in simpler fields, such
How, then, are the observations to be explained as cognition and perception. It is a most interest-
which are now interpreted as a learning of drives? ing occupation to compare motivational action with
After all, some learning must be involved when an dynamic events in those other parts of psychology.
originally neutral object gradually begins to at- When you do so, everything looks different, not
tract or repel a subject. From the present point of only in perception but also in certain forms of
view, only one explanation is possible. Supposing learning. Specific work? There is, and will be,
that the subject's need does not vary, learning must more of it than I alone can possibly manage. Con-
change the characteristics of the object, and thus sequently, I need help. And where do I expect to
transform it into an adequate motivation object. find this help? I will tell you where.
One instance would be what Tolman calls a sign The Behaviorist's premises, we remember, lead to
Gestalt; in other words, the neutral object would certain expectations and experiments. What I have
become the signal for the appearance of something just said invites us to proceed in another direction.
else which is a proper motivational object. This I suggest that, in this situation, we forget about
expected object would now be the object of the mo- schools. The Behaviorist is convinced that his
tivation. Or also, when a neutral object is often functional concepts are those which we all ought
accompanied by facts which are natural motiva- to use. The Gestalt psychologist, who deals with
tional objects, the characteristics of such facts may a greater variety of both phenomenal and physical
gradually "creep into" the very appearance of the concepts, expects more from work based on such
formerly neutral object and thus make it a proper premises. Both parties feel that their procedures
motivational object. Years ago, comparative psy- are scientifically sound. Why should we fight?
chologists in England stressed the importance of
Many experiments done by Behaviorists seem to me
such processes, to which they gave the name "as-
to be very good experiments. May I now ask the
similation." They regarded assimilation as a par-
Behaviorists to regard the use of some phenomenal
ticularly effective form of an association. And is
it not true that, as a consequence of learning, a facts, and also of field physics, as perfectly permis-
coffin looks forbidding or sinister? I also know sible? If we were to agree on these points, we
somebody to whom a bottle covered with dust and could, I am sure, do excellent work together. It
just brought up from the cellar looks most attractive. would be an extraordinary experience—and good
As a further and particularly simple possibility, for psychology.