See other formats

Full text of "Gestalt Psychology A Survey Of Facts And Principles"

. 

Keep Your Card in This Pocket 

Books will be issued only on. presentation of prcp* 
Hhrary oardft. 

Unloes labeled otherwise, books may be tfriainad 

lor two weeks. Borrowers finding books raarkftd, dte 
faoad or mutilated are expected to report at 

library desk; otherwise the last borrower will bo held 
responsible for all Imperfection!* discovered. 

The card holdftr to mtponsible lot all books drawn 
on tills card* 

Penalty for ovr-dud books 2o a day plua ol 
notices. 

Lost cards and change ol rwidenoa must be im- 
ported promptily, 

Public Library 

Kansas City, Mo. 





V'-' 



c 5 f977 



I < 



W&T3W 



APR 1,2 1993 



21 



X ->** 



P S Y O H O L O G Y S K R 1 ES 

ALBERT T, Pom;NUKKtiKK, EDITOR 



ABNORMAL PSYCHOLOGY ITS CONCEPTS AND THEORIES 
II. L. Hot.UN<*\voRTH, Ph,D., Professor of Psyehology, Harnim 

College, Columbia University. 

CONTEMPORARY SCHOOLS OF PSYCHOLOGY. ROIURT S. WCKKI 
WORTH, Ph.D., Sc.D,, Professor of P,\yehology Columbia University 

INTRODUCTION TO PHYSIOLOGICAL PSYCHOLOGY. GHAY 
DON L&VERNE FRRKMAM, PhJX, ncpunmviu {' P?*ychnU>KV. North 
western University. 

INTRODUCTION TO COMPARATIVE PSYCHOLOGY, ^ C\ J 
WARDEN, Ph.D. AH^intant Pr<tVf*xr <ii P,vi*hIoj',>% Columbia Uni 
vermty; T. N, JENKINS, Ph.D., Assistant Prolt*sN>r of PNvrhloja;j 
New York UniWrNity; JUH! L. IL WAHKI-, Ph.O*, Rrsrarch Amtt 
elate. Department of Xmhj>% Pomona C>U^<\ 

GENETIC PSYCHOLOGY. ADAM HAYMOW <rn,ur,\Ni>, PhJX, Pro 
f<ssr of Psychology, Northwestern University* 

PSYCHOLOC5Y AN KMPIRU'AL STUDY OK BEHAVIOR Fn 
KRICK II. LUND* Ph.D., Professor of Psychology, Temple 



THE PvSYCHOLCXJY OF DRESS. RwAT f . HimwvR, Ph.D., li 
struetor in 



GESTALT PSYCHOLO(5Y A SURVEY OF FA( 1 TH AND PRllN 
CIPLKS. (iKORtJK W. HARTMANN, Ph.D., Professor of Pnycholog; 

Pennsylvania State College, 

In Pwp&r&lfon 

COMPARATIVE PSYCHOLOGY. C ; J, W/IRI>EH Ph.D., 
Professor of I^ychology, t^ulisinbLi Uttivrrnhy ; T N, 
Ph,D., Aswistam" ProfesHor of P?*yehiiltmv f New York 
and L. II. WARNKR, Ph.D., Rencareh AjiMjehttCt De^artmeiU 
Zoology, Pomona College* 



G 
PSYCHOLOGY 

A SURVEY OF FACTS AND PRINCIPLES 



By 

GEORGE W. HARTMANN, Pn.D. 

PROFESSOR OF PSYCHOLOGY, THE PENNSYLVANIA STATE COLLEGE, 
SOMETIME SOCIAL SCIENCE RESEARCH COUNCIL FELLOW AT THE 

UNIVERSITY OF BEKJLtN 



PSYCHOLOGY SKRIKS 
ALBERT T* POFFKNHKRCZKR, PH.!)., Editor 

Prtftt*ifior of Pfiychulogy, Cuiuntbiiii tlniverRity 



Till'. 'RONALD TRKSS COMPANY 
NICW YORK 



Copyright, I935 
THE KONAU> I*RKSS C 



All Riyhts Reserved 

The text of this imblicatitm or any part 

thereof may not hi* ^I'pnwhiml in any 

itKUUHH* wli!its<H*v<f wttiukut fx'i'ttii'. *tnn ih 

writituf from the fnihltslu't 1 . 



TO 

ROBERT S. WOODWORTH 



PREFACE 

Within the last decade American psychological circles have 
witnessed a recession of interest in the hehaviorist movement 
and a steady growth of critical investigations centering about 
the doctrine of Gestalt. Although historically the two theo- 
retical positions developed contemporaneously, a number of 
influences conspired to delay the vogue of configurationisin. 
First -of all, as a product of German thought, it was couched in 
an alien tongue with all the difficulties of translation and inter- 
pretation which that implies, particularly in a highly technical 
field. Second, the war and its consequences interposed a brief 
but huge obstacle to professional contacts. TJiijjd, Gestalt 
psychology is pitched in a rather high intellectual key and con- 
sequently offers, more resistance to popularization than be- 
haviorism, whose positivistie and materialistic features represent 
an extension of, a, familiar tradition to. all phases, of mental life. 
And last^the leaders of the movement have been so .busy pur- 
suing an ambit iotts research program, proselytizing, and engag- 
ing in polemics that a really adequate synthetic treatment from 
their own bands constitutes one of the biggest gaps hi Contem- 
porary psychological literature* 

(Herman scientific writing is proverbially famous for its ob- 
scurity and there is much in the sheer phraseology of Gestalt 
which repels one by its vagueness. Precise definitions are con- 
spicuous by their absence. Nevertheless, the difficulties which 
even well-trained psychologists with different experimental 
backgrounds experience in comprehending the basic contribu- 
tions of configurationism lie largely in the intentional and vio- 
lent break which it makes with the past. Tt is like some sudden 
forward leap disturbing the continuity of scientific development 
The revolution which has occurred in physics within the last 
quarter-century presents an analogous situation, for few lay- 
men, and perhaps few physicists, are fully at home in the new 
realm of relativity. 



vi PREFACE 

This volume is an attempt to firing together in convenient 
form all the material necessary for a more than superficial un- 
derstanding of the subject with which it; deals. It is intended 
to be read with profit by any one who has had a first course in 
elementary psychology* Essentially, I have aimed to give a 
sympathetic picture of the Gestalt system from the standpoint of 
a mm-conflgurationist, although [ must confess that an examina- 
tion of the evidence has left me wore favorably disposed toward 
the theory than I had originally anticipated. My major task 
has been that of an expositor and interpreter and only sec- 
ondarily that o f a critic. I lad my equipment and resources been 
adequate, I should have preferred to make this work a complete 
reference handbook, but external limitations compel me to rest 
satisfied with a less pretentious, but 1 hope equally serviceable, 
summary and orientation. 

Many passages derived from important scattered sources ap- 
pear here in English for the first time; and while 1 realize 
that the manner in which they have been woven together 
leaves much to be desired, the busy reader and student should 
find here the gist of the most significant documents in this 
field, which should give a background that will make pos- 
sible an intelligent reading of the new material constantly being 
prepared and issued under Gestalt influence. Economy and 
pithiness of expression have been sought throughout, and in an 
effort to attain the maximum of compactness, liberal use has 
been made of quotations, both direct and condensed, More 
sources than actually appear in the references have been ex- 
amined with an eye to inclusion in some way or other* As im- 
plied above, the necessity for brevity has always been regarded, 
otherwise the volume would have inevitably expanded into a 
digest of psychology in general for it is very easy to 'cross the 
loose boundaries of the Gestalt domain. The experimental mass 
may seem a bit unwieldy but in emphasising the laboratory 
products of research motivated by the Gestalt principle 1 am 
stressing what I am persuaded will prove the most permanent 
contribution of all. Minor inaccuracies and inconsistencies 
probably still remain, but a genuine effort has been made in re- 
duce them to a minimum. 



PREFACE vii 

I am indebted to the Social Science Research Council for 
the fellowship award which made it possible for me to spend 
the academic year 1930-1931 in the very citadel of Gestalt 
psychology, participating in characteristic investigations and 
profiting from the personal contacts and library facilities then 
available at the Psychologisches Institut of the University of 
Berlin. 

Permission to quote text material and to reproduce figures 
has been generously granted by the numerous publishers, authors 
and editors whose works are referred to in the present volume, 

GEORGE W. HARTMANNT 

State College, Pa. 
April 25, 1935 



CONTENTS 

Part I Historical 

CHAPTER 1 p AGE 

THE ANTECEDENTS AND DEVELOPMENT OF THE DOCTRINE . . 3 

Wertheirner's experiment: the problem of origin. Mach and 
Ehrenfels. The contribution of Wilhelm Dilthey. Other German 
theorists of the nineties. Stout's position. The work of William 
James. Dewey and the biological trend. Contributions of David 
Katz and Karl Bithler. Rubin and experimental phenomenology. 

Part II Theoretical 

CHAPTER 2 

THE PHYSICAL BASIS OF THE GESTALT THEORY 31 

Kohler's naturalism. Gestalt and relativity. 

CHAPTER 3 

THE PHYSIOLOGICAL FOUNDATIONS 53 

Lashley's cerebral studies. Driesch's ernbryological experiments. 
Weiss* transplantation work. Marina's eye-muscle surgery. Coghill 
and Minkowski on the origin of reflexes. 

CHAPTER 4 
THE PHILOSOPHICAL FOUNDATIONS 63 

Wertheirner's anti-synthetic position. Lewin on scientific method. 
Wheeler's organismic laws. 

CHAPTER 5 

VARIETIES OF GESTALT THEORY 78 

, Krueger and the Leipzig school. Sander. The hunger studies of 
David Katz. Different attempts to solve the part-whole problem. 



x CONTENTS 

PAGE 
Part III Empirical 

CHAPTER 6 
PHENOMENA OF VISUAL PERCEPTION 93 

Sensation and perception: fundamental laws. Dependence of form 
and space qualities upon the configuration. Dependence of brightness 
and color qualities upon the configuration. Movement as conditioned 
by the gestalt. Constancy effects in vision. 

CHAPTER 7 

STUDIES IN AUDITION AND THE SKIN SENSES 132 

Experiments in the perception of sound. The skin senses. 

CHAPTER 8 
THE UNITY OF THE SENSES 141 

Hornbostel. Further evidence of intersensory relations. Werner's 
sensation-stages. 

CHAPTER 9 

MEMORY PROCESSES 152 

Wulf s experiment and its American extensions. Inadequacy of as- 
sociationism. 

CHAPTER 10 

LEARNING: DATA AND INTERPRETATIONS 159 

The anthropoid studies. The nature of mental development. Ogden's 
new laws of learning. 

CHAPTER 11 

THINKING AND REASONING 179 

Primitive thought. Formal logic. Closure. Relation versus Gestalt. 

CHAPTER 12 

RESEARCHES ON INSIGHT 187 

Terminology. Varieties of insight. Patterns and learning. Thorn- 
dike and belongingness. Humor. 

CHAPTER 13 

ACTION, EMOTION, AND WILL 202 

Lewm's theory of psychic tensions. Value and meaning. Reward 
and punishment. Contributions of Lewin's pupils. 



CONTENTS X i 

PAGE 

Part IV Practical 

CHAPTER 14 
MENTAL PATHOLOGY 241 

The work of Gelb and Goldstein and their associates. Contributions 
to psychiatry. Speech deficiency. 

CHAPTER 15 

INDUSTRIAL AND PERSONNEL SUGGESTIONS 252 

Business possibilities. Estimating personality. 

CHAPTER 16 
APPLICATION TO EDUCATIONAL PROBLEMS 258 

Changes in policy and teaching method. Intelligence testing. Theory 
of experimentation. Kinship with progressive education. Athletics, 
Aesthetics. Ethics. 

Part V Critical 

CHAPTER 17 

CRITICISMS OF GESTALT THEORY BY OTHER SCHOOLS OF 
THOUGHT 277 

The Muller-Kohler debate. Rignano-Kohler controversy. Scheerer's 
judgment. Brunswik's interpretation. Petermann's analysis. Hel- 
son's verdict. Woodworth and the critique of the eclectics. The 
configurationist's defense. 

CHAPTER 18 
CONCLUSION : THE PRESENT STATUS OF THE GESTALT SCHOOL 300 

A NOTE ON BIBLIOGRAPHIES 302 

BIOGRAPHIES 303 

CHRONOLOGY OF SIGNIFICANT DATES AND EVENTS IN THE 
DEVELOPMENT OF GESTALT PSYCHOLOGY 305 

GLOSSARY 37 

INDEX 3 I S 



ILLUSTRATIONS 

FIGURE 

1. A Simple Device for Showing the Phi- Phenomenon 4 

2. Schema Indicating Wertheimer's Method of Disproving the Eye- 

Movement Explanation of the Phi- Phenomenon 5 

3. Rubin's Vase Figure 24 

4. Another Reversible Experience (Iron-Cross Figure) 25 

5. The Dependence of the "Quality" of the Contour upon the Nature 

of the Area Embracing It 28 

6. Optic Nerve Pathways Illustrative of the Electromotive Forces in 

Figure-Ground Percepts 45 

7. Sander's Parallelogram Illusion 85 

8. One of Wertheimer's Dot Patterns ' 95 

9. Illustrations of Wertheimer's Principle of Innigkcit or "Intimacy" 97 

10. Diagram Exemplifying the Principle of u Closure" 98 

11. Sketch Indicating the Changes Effected by Different Positions of 

the Component Parts of Patterns 98 

12. Differing Knowledge of the Cursive Script of the Ancient Lan- 

guages Conditions our Interpretation of this Datum 99 

13. How Internal Organization Outweighs Effects Traceable to Past 

Experience 99 

14. Each Circle Symbolizes a Clay Pot 103 

15. Materials Employed in the Training Series with Crows to Test their 

Ability to "Transfer" Habits of Brightness Discrimination . . 103 

16. How Distinctive Figures Become "Lost" when Imbedded in a Dif- 

ferent Perceptual Setting 105 

17. The Third Dimension as a Product of Gestalt Factors 106 

18. One of Hornbostel's Reversible Solids 108 

19. A Blind-Spot Card Modelled after Scripture's 114 

20. Benary's Figures Demonstrating the Influence of "Form" upon 

Brightness Properties 120 

21. The Central Stripe is the Brightest Single Region in the Figure . . 121 

22. Schema Illustrating the Influence of "Form" upon Color Quality . 123 

23. Koffka's and Wittmann's Figures Modified from Linke's Originals . 125 

24. Ternus 1 Method of Demonstrating Phenomenal Identity Despite 

Objective Difference 126 

25. Wulf s Memory Figures 153 

26. A Simple Illustration of the Meaning of Closure 167 

27. A Sample of "Wholeness" in Perceptual Experience 170 

28. Problems 182 

29. Lewin's Method of Depicting the Action of Field Forces in Conduct 21 T 

30. Same as Figure 29 but Complicated by the Introduction of ,a Barrier 211 

xii 



ILLUSTRATIONS xiii 

FIGURE PAGE 

31. Same as Figure 30 but with a New Vector toward the Goal Cir- 

cumventing the Obstacle 211 

32. Introduction of a Negative Vector 212 

33. Consequences of Introducing Disciplinary Measures 212 

34. A Conflict Situation Provoked by Two Equally Attractive Goals . 213 

35. An Ambivalent Conflict Situation 213 

36. How the Child Tries to Go> out of the Field in Which He is Caught 214 

37. Tension at its Climax 215 

38. The Reward Situation 217 

39. Same as Figure 38 with a Barrier Added to Insure the Performance 

of the Task if the Reward is to be Gained 217 

40. Combination of Reward and Punishment 217 

41. Pseudofoveal Vision of Fuchs's Patient 245 

42. Rupp's Beehive Pattern 252 

43. One of Street's More Difficult Gestalt Completion Tests .... 266 

44. Brunswik's Survey of Gestalt Phraseology 285 



TABLES 

TABLES PAGE 

I. Influence of the Manner in Which Visual Learning and Recall are 

Made upon the Frequency of Error 101 

II. Subjective Differences in Equating Light Discs for Size When One 

is Projected Forward and the Other Above 112 

III. Judgments of Relative Sound Intensities of Paired Stimuli with 

Varying Times Elapsing between the Members of the Pairs 135 

IV. Number of Times Certain Figures were Reported in Response to 

a Given Pattern Stimulus 138 

V. Mean Reaction-Times of Subjects in Performing Tasks Involv- 
ing Paired Associates 157 

VI. Nature of Changes in Character of Tasks Attempted after Ex- 
periences of Victory or Defeat 231 

VII. Relative Memory for "Real" and "Irreal" Items of Experience . 236 



GESTALT PSYCHOLOGY 



PART I HISTORICAL 



CHAPTER 1 

THE ANTECEDENTS AND DEVELOPMENT OF THE 

DOCTRINE 

Wertheimer's Experiment : The Problem of Origin 

That type of scientific research in the field of psychology 
which is identified by the name of "Gestalt" seems to have been 
first definitely introduced by_Max Werthejmer in a famous ex- 
perimental paper l on the "Perception of Apparent Movement" 
which appeared in 19.12. This hundred-page study (to which 
frequent and detailed reference will be made during the course 
of this volume) is a meticulous and thorough investigation of a 
specific laboratory problem, but apart from the consequences to 
which it led, it contains little to distinguish it from countless 
other researches so representative of German scholarship. Even 
its main finding that an object seems to move from one position 
v to another when it is merely presented twice in two different 
places with an appropriate short time interval between both ex- 
posures was well known to earlier specialists in sensory physi- 
ology and is now familiar to many amateur attendants at motion 
picture performances* Evidently the fact with which Wert- 
heimer's monograph dealt cannot be responsible for its impor- 
tance ; it is rather the new explanation of the phenomenon which 
he offered which lends significance to his work. 

What was this interpretation ? It Is not easy to give a simple 
answer without anticipating much of the material which must 
come later. Briefly, Wertheimer demonstrated that the various 

1 "Experlrnentelle Studien uber das Sehen von Bewegtmg," Zcitschrift filr Psychologic, 
vol. 60; also reprinted in Drei Abhandlungen sur Gcstalttheorlc, Erlangen, 1925. 



4 GESTALT PSYCHOLOGY 

conventional theories offered at the time were inadequate to 
explain the occurrence of the event under all conditions, and that 
a wholly new interpretation was necessary. Traditional notions 
of retinal and brain action during visual perception were rejected 
and a preferred substitute sketched (but not fully developed!) 
by assuming the existence of cerebral "cross-processes" as the 
correlate of the psychic relations which seemed to be just as 
"original" in the experience as the fundaments upon which 
they were previously supposed to depend. The attachment of 



u 
II 

IS 

II 






ii 
\\ 
n 
it 




>b 


. _......! 






T-2 V-2 



Figure I. A Simple Device for Showing the Phi-Phenomcnou 

When two light bulbs are illuminated in rapid succession the resulting shadows cast 
l)y the steel rod appear to jump back and forth, 

clarity and richer meaning to this statement hinges upon the 
development of the concept in the following pages. 

Wertheimer's method of investigating the movement phe- 
nomenon can be readily reproduced with the aid of two electric 
bulbs and a metal upright. 2 Place two lights along the edge of 
a table separated by a few feet ; between them and the wall of 
the room stand a thin rod. If one now rhythmically switches the 
two lamps on and off, the shadow cast by the rod will move back 
and forth in lively style between two positions, provided the 
optimal time interval (about 60 milliseconds) is used. If the time 

2 See Higginson's article on "A Simple Class Demonstration of Apparent Movement" 
in J, Bxfrcr. Psychol., 1927, jo, 67-68. The essentials of this experience may be seen m 
the flashing of fireflies on' a warm summer evening 1 , in the sparks of the fireplace log-, and 
in the change of traffic lights when the red ball seems to be converted into a green one. 



ANTECEDENTS AND DEVELOPMENT 5 

between the two light exposures is too short (say 30 millisec- 
onds), the two shadows of the rod will appear simultaneously; 
if the interval is too long (in the neighborhood of 200 millisec- 
onds or one-fifth of a second) mere "quiet" succession is noted. 
'Under favorable conditions, apparent movement possesses all 
'the phenomenal characteristics of objective motion. 

Naturally, no one who has ever observed this event disputes 
its subjective nature: the real conflict appears when opposing 
theoretical accounts are examined. One plausible view held by 
Wundt maintained that kinaesthetic sensations produced by the 
"jump" movement of the eyeballs were the determining factors. 



Figure 2. Schema Indicating Wertheimer's Method of Disproving the Eye- 
Movement Explanation of the Phi-Phenomenon 

With constant fixation the simultaneous but opposed displacement of two lines can 
be observed. 

However, Wertheimer obtained the phenomenon when he 
arranged his tachistoscopic exposures so that the total presen- 
tation time for the first object plus the time-interval plus the 
exposure period for the second object did not exceed one-tenth 
of a second ; since the minimal time for eyeball reactions is some- 
what greater (around 130 milliseconds), this explanation must 
be excluded. Furthermore, the movement readily occurs even 
with rigid fixation of a definite point in the visual field. An 
even more decisive elimination of the eye-movement hypothesis 
was accomplished by the simultaneous production of two antag- 
onistic apparent movements, so that ai appeared to shift to bi 



6 GESTALT PSYCHOLOGY 

and a 2 to b 2 , as in Figure 2. Obviously the eyeball cannot move 
in two opposed directions at the same time, with the result that 
the explanation of the event must be sought in some other 
mechanism. 

Marbe's theory that after-images of the first object ac- 
counted for the apparent shift was equally well excluded by the 
following procedure : Before the apparent movement was pro- 
duced, a strong after-image was created by gazing at a small, 
bright cross or a lamp filament. This was now projected upon 
a definite spot within the exposition field on the screen. Again, 
the phenomenon was witnessed despite the stationary holding 
of the after-image. Wertheimer proposed calling this concrete 
observation of apparent motion the phi-phenomenon, the $ 
standing for whatever occurred between a, the first exposure, 
and b y the second. Under special conditions, "pure" phi was 
obtainable; i.e., the observer saw neither a nor b nor the move- 
ment, but simply "something in motion/' This is a genuine 
"dynamic" occurrence which on neither a priori nor other 
grounds requires to be traced to a "static" base. The curious 
reversal of revolution in a spoked wheel often noted while 
watching a moving automobile may thus be explained. 

Wertheimer's physiological hypothesis is built upon the 
existence of central "diagonal functions" (Qu\crfitnktionen in 
the original) or "between-processes." Whenever a central 
locus in the brain is excited a concentric neural spread of a cer- 
tain magnitude occurs around ify If two such spots are aroused, 
two excitation rings are formed which predispose the areas they 
embrace to further excitation. If now, point a is stimulated, 
and shortly thereafter an adjoining point b, some kind of physi- 
ological "short-circuit" occurs between a and b, and a specific 
excitation occurs over the intervening distance. If the con- 
centric overflow from a is at its maximum and similar excita- 
tion rings now come from 6, the direction of the neural process 
is determined by the fact that a was first there. The nearer the 
two points a and &, the more favorable are the conditions for the 
arousal of the phi-process, which in itself is an extensive spe- 
cific whole. 



ANTECEDENTS AND DEVELOPMENT 7 

Since Wertheimer claims that as far as the personal experi- 
ence goes there is no difference whatsoever between the percep- 
tion of real and illusory movements, he is implicitly proposing 
that wherever two identical phenomena are found, it is neces- 
sary to assume that the corresponding brain-processes are iden- 
tical. The Talbot-Plateau law, which comes to mind readily in 
this connection, would then have to be interpreted as follows : 
A uniform grey disc and one constructed out of any variety of 
sectors which produce an equivalent visual experience when ro- 
tated must arouse the same brain-process. 3 

The main reason why Wertheimer may be considered the 
founder of Gestalt psychology rather than many other eminent 
predecessors and contemporaries whose general attitudes and 
whose specific claims were often amazingly like his own is 
that he, or the small group of brilliant young Dozenten who at- 
tached themselves to his standard at the very beginning, gave a 
militant or reformist turn to the movement. 4 He and his dis- 
ciples Kohler and Koffka were his subjects in the above ex- 
periment became the ultra-radicals of the day, abandoning 
practically all the axioms and postulates of established psychol- 
ogy with the exception of the naturalistic standpoint and the 
appeal to experiment ; whereas most psychologists, even though 
they shared many of the views and furnished much of the sup- 
porting material for the opposition's most trenchant criticisms, 
protested against such extremism and clung to the "grand tra- 
dition/' The non-Gestalt or anti-Gestalt theorists were moti- 
ped by a desire to patch and repair the imperfections implicit 
in the conceptual structure of the older experimental psychol- 
ogy; the school of energetic disciples which centered around 
Wertheimer, on the other hand, was more interested in making 

3 This thesis is developed with a high level of technical perfection in the later experi- 
ments of Ccrmak and KoKTka. See their article on "Untersuchungen iiber Bewegungs-und 
Verschmelzungsplianomene," PsychoL Porsch,, 1922, i, 66-130. 

4 The breadth of application of the idea of Gestalt soon made it overflow the rather 
narrow boundaries of psychology. ^ This implicit _ universality seems to have been felt by 
its adherents almost from the beginning and is in part accountable for the boastful and^,, 
superior air with which it has been expounded. That the conngurationists do not think' 
meanly of their discovery is apparent from Koffka's assertion, "The terrn^ Gestalt is a/ 
short name for a category of thought comparable to other general categories like^substance,! 
causality, function. But Gestalt may be considered more than simply an addition to pre-J 
existing conceptual principles; its generality is so great that one tnay^ask whether causality! 
itself or substance does not fall legitimately under it." See his article on Gestalt in the) 
Encyclopedia of tlie Social Sciences. 



8 GESTALT PSYCHOLOGY 

an absolutely new start. So faithfully have the configuration- 
ists adhered to this initial program that a deficient sense of his- 
torical continuity is occasionally painfully evident in their pub- 
lished works, which seem to imply that scientific psychology did 
not exist before 1912 ! 

It is an intricate and unrewarding task for an examiner of 
documentary evidence to reach a decision on the debatable ques- 
tion of priority, especially, as in this case, where the methodo- 
logical viewpoint and procedure is more significant than the 
concrete and immediate experimental accomplishment. 5 The 
situation has a strange and illuminating parallel in the rise of 
behaviorism, an attitude now internationally designated by this 
American neologism. Watson undoubtedly is the father of the 
behavior ist movement because in 1911 he vigorously cham- 
pioned the cause and compelled even the most unsympathetic 
psychologists to attend to it. His program was too audacious 
to be disregarded or even quietly assimilated. But one could 
wage a pretty war of words concerning the intellectual ancestry 
of Watson's ideas. Disregarding the direct influence of the 
functional philosophy of the Chicago school ultimately derived 
from William James, or even the broader effects of an en- 
lightened materialism, there was a strong "anti-consciousness" 
trend among German biologists of the 'nineties (Bethe, Beer, 
von Uexkiill, and in some respects, Loeb.) The physiological 
and neurological beliefs of Pavlov and Bekhterev had begun to 
permeate foreign thought in the first decade of this century. 
Even in English writings the atmosphere had been altered by 
the new emphasis given to the word "behavior" by McDougall, 
Max Meyer, and others. So far as the paper record of dates 
goes, one could easily justify a charge of "unoriginality" 
against both Watson and Wertheimer but the same device 
would deny that Darwin initiated the modern theory of evolu- 
tion. From the historical point of view, all that leaders of any 

B As early as 1894, Exner had written: "The total impression aroused by a picture 
which moves across the retina is made up of excitations from a great many functionally 
dissimilar fibers. That we should, nevertheless, have a unitary experience, in which the 
part-sensations are unnoticed, is caused by what I shall term the principle of central con- 
fluence." What difference there may be between the "short-circuit" of Wertheimer and 
the "confluence principle" of Exner is still uncertain. See Entwurf zu riwer physiologischcn 
Erklarung dcr psychischen ErschriMtngen t Vienna, 1894. 



ANTECEDENTS AND DEVELOPMENT 9 

great intellectual change do is integrate the scattered ideas al- 
ready existent on any scientific problem, adding impetus and 
direction to the inarticulate or uncoordinated trends. 6 The 
supporting evidence for this view is presented in sections below. 

Mach and Ehrenfels 

The man who first made the problem of Gestalt a controver- 
sial issue in modern psychology was Ehrenfels, an Austrian 
philosopher, who wrote a discussion on "Gestaltqualitaten" in 
1890 for an obscure and relatively inaccessible journal. 7 
Ehrenfels 7 point of departure was Mach's notable "Analysis 
of Sensations" which first appeared in 1886 and passed through 
many editions during the life of the author. 8 Mach was inter- 
ested in a question of descriptive psychology : what are intrin- 
sically those mental presentations which we call "spatial forms" 
(like a square upon paper) or "melodies" ? Are they a mere com- 
position of elements, or are they something new, accompanying 
the aggregation, but nevertheless distinguishable therefrom? 
It is evident from the phraseology of his inquiry and the whole 
tenor of his discussion that Mach looked upon a gestalt not as 
a mere composite of elements but instead as something (in con- 
tradistinction to the elements upon which it rests) novel and to 

The writer's point of view on this topic is better expressed by Boring's claim that 
"The progress of thought LS gradual, and the enunciation of a *new' crucial principle iti 
science is never more than an" event that follows naturally -upon its antecedents and leads 
presently to unforeseen consequences." Of. the rest of his clever critical article on "Gestalt 
Psychology and the Gestalt Movement," American J, Psychol. f 1930, 42, 308315. If. 
however, a saltatory theory of intellectxtal progress be correct and the Gestalt idea of 
insight applied to human hisUu-y would seem to support it then the judgment of the 
chronicler would need to be different, 

7 Christian JRhrenfelH, "Ueber Gestaltqtialitaten," Viertclfahrschrift fur wisfensckaft- 
HcJie Philosophic, 1890. Also reprinted with some later comments in a curious post-war 
monograph entitled, Das rrimsahlcngcscta cntwickeltr ttnd dargcstellt auf Grund* der 
Gestaltthcpric, Leipzig, 1922. There is little point in tracing: the faint hints of a doctrine 
championing the primacy of the whole throughput the entire history of European thought. 
To do so in this connection would merely be a display of uneconomical scholarly effort. Our 



ably the first historical record of reflection upon this theme is found in the remark of the 
Chinese sage Lao-Tse (B.C. 600) in his Tao-te-Kitiff, 39th saying, "The sum of the parts 
is not the whole!" Tn this volume, however, we are concerned with the immediate and not 
the remote origin of the Gestalt concept. 



tree 

context, S tf>wevor" I w<>'"'icarn "tfiaT'thi^ naive" impression* must be succeeded by scientific 
analysis if the experience (and the thing experienced) is really to be understood. . The dis- 
tinction between a pre-analytical and post-analytical datum appears to be of importance here 
and in all other Gestalt discussion. 



ID GESTALT PSYCHOLOGY 

a certain extent independent. Descriptively, he was a sensa- 
tionalist of the most extreme sort, emphasizing the elements and 
neglecting the superstructure. Biihler, in fact, claims that lie 
was "gestalt-blind." There is plenty of ambiguity in Mach's 
position and he probably would have leaned more toward the 
interpretation later made by the Graz school (see page 43) than 
that adopted by the adherents of Wertheimer ; nevertheless, he 
had touched ever so lightly the problem of organisation in men- 
tal life, an issue which continues to dog the footsteps of the 
theorist and experimentalist alike. 

Ehrenfels saw that this point was of paramount importance 
and proceeded to develop it. Like many Germans with a strong 
musical background, he was fascinated by the patterns displayed 
in tonal sequences, and offered a most persuasive account of 
one's auditory experiences. "In order to apprehend a melody, 
it is not enough to have an impression of the momentarily sound- 
ing tone in consciousness, but where the tone is not the first 
one it is necessary to have at least a few of the preceding 
tones simultaneously presented in memory" (page 1 1 of Ehren- 
fels' article). Isolated reception of the individual sensation 
certainly does not occur in this situation. 

He then raises the question whether the experience is more 
than the sum of the separate "local determinants 7 * or whether in 
the total presentation more can be found than appears in the com- 
ponent sensations. With what now appears unnecessary aca- 
demic sublety, Ehrenfels asks us to conceive the difference be- 
tween separate tonal sensations brought together in a single con- 
sciousness on the one hand, and distributed among n conscious 
units on the other; and decides introspect ively, of course 
that "more" is present in the first instance than in the latter. 

Better, or at any rate more comprehensible, evidence is of- 
fered by another classic illustration. Consider a measure of any 
simple melody with which you are familiar. If it be played in 
a given key, identical notes will recur with a certain frequency. 
Now change the key in which the melody is played. It will then 
contain not a single one of the tones out of which it was con- 
structed in the first case ; nevertheless, the similarity is immedi- 
ately recognized without reflection by any one with a minimum 



ANTECEDENTS AND DEVELOPMENT u 

of musical sensitivity. Return next to our original key and 
play the melody again and follow this by another repetition, 
retaining the same rhythm, but changing the tonal sequence so 
that exactly the same number of identical tones appear as before 
but in different order; i.e., if the original sequence was e g f a g 
g f e c e d, make tip another permutation which like it contains 
three e'$, three g's, two fs, one a, one c, and one d. With the 
exception of the preserved rhythm, no one will notice any simi- 
larity to the original melody, unless an artificial analysis leads 
to a comparison and counting of individual notes (Ehrenfels, 
page 1 6). 

* x '""'Elirenf els concludes that the resemblance between spatial and 
tonal patterns rests upon something other than a similarity of 
their accompanying elements. Th^y^ota^ 
be different entjtjes^,.!!].^^^]:^ sums of their parts. Added evi- 
dence is offered by the fact that a^person wlierT asked to repro- 
duce a melody in other than the original pitch (which, because 
"of the differences among voices, he usually must do), does not 
reproduce the sum of the original single impressions, but a 
wholly different complex, which has the sole resembling attri- 
bute that its members stand in an analogous connection to each 
other as do those of the previously experienced complex. In 
other words, the "Gestaltqualitat" (= form quality) or whole 
has been reproduced : the elements or parts have not. 9 

Beyond the rather banal distinction between temporal and 
non-temporal form-qualities the latter being considered sim- 
pler that is all of pertinence in Ehrenfels' original article. But 
it was enough to set many wheels a-going. In an appendix 
which was issued long after the first publication, the author 
tried to deal with some of the problems which had grown out of 
the initial discussion, particularly the perennial one concerned 
with the distinction of "relation" and "form-quality." He 
maintains that the melody can be heard and the square seen, but 
not the similarity or difference of two tones or of two spatial 
points, i.e., the former (Gcstaltqualitat) can be directly sensed 

Some writers have remarked that the phrase, "He can't see the forest for the trees/ 7 
is a popular expression for the failure of the form quality to appear. 

i 10 Kohler has made a terminological clarification which should be of service from this 
\point on. "In the German language at least since the time of Goethe, and especially ia 



12 GESTALT PSYCHOLOGY 

or perceived, while the latter (=the relation) lacks this self- 
regulatory character and cannot occur without our own contri- 
bution or without the peculiar activity of the process of compari- 
son. This involved logical analysis became even more compli- 
cated in the hands of some fellow epistemologists. 

In accordance with his philosophic temperament, Ehrenfels 
tried to exhaust the broader implications of the idea which he 
had developed. Whoever has reached the conviction that all 
combinations of psychic elements yield something unique will 
inevitably attach to them (i.e., the "combinations") a much 
greater importance than if they are simply seen as displacements 
of eternally recurring constituents. These psychic combina- 
tions never repeat themselves with complete exactness a prin- 
ciple which finds its best expression in the field of aesthetics. 
Just as some colors are more colorful than others, so some pat- 
terns are more "figured" than others. A heap of sand or a 
lump of earth has less form than a tulip or a swallow. This is 
what Ehrenfels terms "height" or "degree" of Gestalt a fea- 
ture which increases with the product of the constituents, their 
unity, and the multiplicity of parts. In fact, it is suggested that 
what we call beauty is none other than "degree" of Gestalt! If 
it is rather difficult for us to accompany von Ehrenfels into the 
higher realms of speculative fancy, it should be even plainer 
that we must part company with him as soon as he maintains 
that the cosmological principle of the universe is the bringing 
forth of new patterns. It is not our concern to test the validity 
of these deductive outcomes of preoccupation with the Gestalt 
doctrine; it is enough to show that much of the ground later 
to be traversed by the experimentalist had already been sur- 
veyed from the theorist's armchair. 11 

his own papers on natural science the noun *gestalt' lias two meanings: besides the 
connotation of 'shape* or 'form* as a property of things, it has the meaning of a concrete 
individual and characteristic entity, existing as something detached and having a shape or 
form, as one of its attributes. Following this tradition, in Gestalt theory, the word 'gestalt* 
means any segregated whole, and the consideration of Gestaltqualitaten has become a more 
special^side of the gestalt problem, the prevailing idea being that the same general type of 
dynamical process which leads to the formation and segregation of extended wholes will 

, also explain their specific properties." Gestalt Psychology, New York, 1029. 

In this work, Gestalt (capitalized as in the original (jferman) will be used in contexts 
where the theory or viewpoint of an existing school of writers is involved, and the inter- 

I nationalized lower-case form, gestalt, for which there is already good precedent, will be 

, employed in discussions of the concrete pattern phenomena so labeled. 

{ 31 In, 1891 appeared Husserl's "Philosophy of Arithmetic" in which he suggested the 
use of the term "figural moment" to designate the "peculiar unitary analogies among vari- 
ous sense-qualities.* This is the same as Ehrenfels' Gcstaltqtutlitat and appears to have 
been independently developed. 



ANTECEDENTS AND DEVELOPMENT 13 

The Contribution o Wilhelm Dilthey 

From its very inception, the idea of Gestalt was most hos- 
pitably received by those "tender-minded" individuals who saw 
in it the inevitable reaction of the human spirit to its own dis- 
section as practiced by the "new" experimental psychology of 
the day. The dominant school of Wundt with its "brass instru- 
ment" technique and physiological bias exemplified a trend to- 
ward minute analysis of mental activities, which even many 
friendly critics felt was forfeiting the substance of mind for 
the thin shadow of acceptable scientific form. Psychology In 
its early decades undoubtedly fed upon the crumbs which fell 
from the richly laden tables of physics and biology, and those 
who held a higher conception of its mission demanded that it 
cease playing the role of an auxiliary science. One man who 
\was destined to play a prominent part in German intellectual 
Plife Wilhelm Dilthey set his energies to the task of saving 
f-psychology from its own barren futility in a remarkable paper 12 
which appeared in 1894, wherein he presented his Ideas on the 
^difference between a "descriptive" and an "analytical" type of 
^psychology ; a distinction now often represented by the con- 
trast between the terms "understanding" and "explanation." 
What these designations imply may be gathered from the dis- 
cussion below. 

Dilthey postulated a fundamental difference between the 
spirit and method of the "natural sciences" and the contrasted 
^"Geisteswissenschaftcn" an idiomatic compound approxi- 
mately rendered by an older nineteenth-century label, "mental 
O^and moral sciences." In fact, the German term seems to have 
-arisen as a translation of this phrase in John Stuart Mill's writ- 
gings. It includes not only that which we have recently learned 
kto embrace in the description "social science" but also Hnguis- 
-tics, the humanities, and the philosophical disciplines* From 
our contemporary vantage point, it Is plain that Dilthey was 
battling with the dual nature of psychology a dualism which is 
conveniently but probably not permanently expressed by the 
saying that psychology serves as the apex of the biological 

l2 "TcIcen liber cine beschreibende und zerffliedernde Psychologic," Gesammelte 
Schriftcn, V, i, pp. 139-240, Leipzig, 1924. This is a reprint of the original 1894 article. , 



I 4 GESTALT PSYCHOLOGY 

sciences and the foundation of the social sciences ; or, as Weiss 
has condensed it, by the term biosocial. 

Basically, Dilthey questioned the legitimacy of applying the 
procedures and hypotheses of the anorganic sciences to the prob- 
lems of mental life. It was his contention that the causal inter- 
connections employed by the natural sciences are derived in char- 
acter, while in the Geisteswissenschaften they are original, i.e., 
comprised within the datum. "By dcscriftizfc psychology I 
understand the representation of the components and connec- 
tions of each mature human mind as they are united in a single 
tie, which is not added by reflection or deduction, but is directly 
experienced. This psychology is a description and analysis of a 
relation which is as primitively and directly given as life itself" 
(Dilthey, p. 152). This new position was opposed to the older 
scheme of explanatory psychology by which all mental phe- 
nomena were constructed from a limited number of definitely 
determined elements a point of view brilliantly championed 
by James Mill, Spencer, Herbart, Fechner, Helmholtz and 
others. The explanations of psychic activity offered by this ap- 
proach are founded solely upon the intellectual processes ; under- 
standing, however, is possible only through the cooperation of 
all the dispositional powers (Gemutskraftc) in the act of appre- 
hension (Dilthey, p. 172). 

Dilthey maintains in a passage which startles one by its pre- 
monition of the thought later crystallized in the catchword slo- 
gan, "The whole is more than the sum of its parts" that all 
psychic processes are characterized by the fact that the apprehen- 
sion of the "total" is a condition precedent to the adequate in- 
terpretation of the "item" The full import of this suggestion is 
not developed at length but the following citation has a pro- 
grammatic ring, "In psychology all functional connections in 
experience are intrinsically given. Our knowledge of indi- 
vidual facts is simply a dismemberment of this union. Herein 
is manifested a firm structure, immediately and objectively pres- 
ent (page 173, Dilthey). It is in this spirit that the standpoint 
of "atomism" is rejected, and Faust's mockery of Wagner's at- 
tempt at a chemical synthesis of the homunculus is quoted as a 



ANTECEDENTS AND DEVELOPMENT 15 

fitting commentary upon the efforts of contemporary sys- 
tematists. 

Such a position, of course, ran counter to the enthusiastic 
positivism of the science of the fin de siecle and it is small won- 
der that Ebbinghaus sharply attacked this view. But a minority 
definitely felt the force of Dilthey 's reasoning and saw that his 
strictures did not necessarily mean the end of all psychology. 
He had merely tried to create a "useful" psychology to serve 
as the basis of the moral sciences in the same way that mathe- 
matics affords a valuable and indispensable foundation to the 
natural sciences. To be sure, Dilthey took a wholly different 
path and insisted that neither introspection nor laboratory ex- 
periments could reveal the nature of mind: that, he felt, was 
best sought by an examination of a man's history. 15 It is plain 
from the whole tenor of his writings that Dilthey needed an 
ancillary science of human nature just as the earlier utilitarians 
and the classical economists elaborated a crude one incidentally 
to their main concern, and just as contemporary students of 
social life find psychiatric views more serviceable than the 
orthodoxies of the experimentalists. /Nevertheless, he left a 
ferment behind and the following highly theoretical passage 
later proved to be richly charged with experimental possibilities : 
/^The psychic life process is in all cases an original unity from 
its simplest to its highest forms. Mental life does not grow 
together from parts ; it does not build itself up out of elements; 
it is not a composite, nor a result of cooperating atoms of sensa- 
tion and feeling : it is primitively and always a comprehensive 
unity. Out of this unity psychic functions differentiate them- 
selves, maintaining, however, their original connections. This 
fact, which is expressed on the highest plane by the unity of con- 
sciousness and the unity of the personality, distinguishes mental 
life completely from the entire corporeal worldT (Dilthey, p. 
211 )/ We shall have ample occasion to meet this recurring 
theme as it presents itself with numerous variations in the suc- 
ceeding decades. 




stellation of aw individual. 



16 GESTALT PSYCHOLOGY 

'* Other German Theorists of the Nineties 

There is good reason for characterizing the early 'nineties as 
a period when the totality concept fascinated those psychologists 
who could not rest satisfied with the dominant sensationalism 
of the day. People wanted to study the organism in "long-sec- 
tion" as well as ^cross-section." Even such diverse views as 
those held by Freud in his thesis of the meaningful determina- 
tion of every psychic act and the doctrine of "creative synthe- 
sis" as formulated by Wundt 14 share in common the endeavor 
to overcome the purely aggregative character of mental life 
which experimentalism had apparently demonstrated. Sander 
(vide infra) in his comments upon this interesting coincidence 
suggests that there may be something uniquely German about 
the attitude, since it can be traced (under many disguises, of 
course) back to the national mystics of the middle ages when 
the concept of totality as a description of mental states first 
emerged. The dynamics of Leibnitz and Herbart's "appercep- 
tion" were definite moves in this direction, albeit over the mis- 
taken road of associationism. Certain it is that many German 
writers with all varieties of metaphysical bias were quickly at- 
tracted to the doctrine. Meinong, 15 Hofler, Husserl, Witasek, 
Cornelius, 16 Martius, Schumann, Stern, Avenarius, 17 to mention 

1A Ideas such as these appear frequently in his Physiological Psychology, especially in 
the section devoted to a discussion of his principle of creative resultants (vol. 3, 6th ed., 
PP- 7S57S6*) : ''All actual psychic processes are compounds;" "The product which results 
from any collection of elements is more than the mere sum of those elements;" "A clang is 
more than the sum of its constituent tones." That Wundt mea^t this to be taken in a 
functional and not just descriptive sense is indicated by his citation of binocular vision as 
a case of a complex perception in which "certain elements have lost their independence." 
Apparently even, the foremost champion of modern clementarism had caught a glimpse of 
complexities which lay outside the bounds of his systematic beliefs, 

15 Meinong is the hyper-logician of the Graz school whose main contribution seems 
to have been a terminological one. That "squareness" which Ehrenfels termed a Gcstalt- 
qttalitdt is held to be related to the four sides demanded as genus to species, but Meinong 
prefers the phrase "funded" or "consolidated content'* (fundicrtcr Inlialt) to designate the 
new entity; of necessity, the various part-contents which merge to produce this are called 
the fusing or unifying contents (fundicrendc Inhaltc"). The derivation of the idea from 
the jargon of public finance is plain. As in other hypotheses concerning mental synthesis, 
he presupposes something which can be synthesized, Meinong, like most of the Grazers, is 
not easy reading. Cf. his most strictly psychological paper in the Zsch. f. Psychol,, 1891, 
II, 245 & 

16 Hans Cornelius approached the problem by working down from above in opposition 
to Meinong's procedure. He starts with the complex which he terms the fusion. To him 
the original mental structure is a total which is gradually split into sections by means of 
analytic attention. _ Before we know anything well, it is but a huge total impression. This 
massive whole disintegrates up^on examination, as a result of which something appears to 
be lost. It is strange that this insight failed to lead to anything more specific, save by 
proxy in the person of his pupil, Felix Krueger. Cf . the interesting papers in the 
Vicrteljahrsschrift f. wissen. Pktlos., i%ga, m i6 f 404$:. and 1893, 17, 30$;, The best single 
survey of this entire group of "old-timers" is still to be found in Madison Bentley's "Psy- 
chology of Mental Arrangement/' Amer. J. PsychoL, 1902, ij?, 260-293. 

17 It has been claimed that what configurationists mean by gestalten is only a special 



ANTECEDENTS AND DEVELOPMENT 17 

only a few of the prominent writers at the turn of the century 
manipulated this theme in diverse ways, giving a logical or 
experimental turn to the doctrine as their temperaments dic- 
tated. The tachistoscopic and eye movement studies of Erd- 
mann and Dodge (1898) in connection with reading phenomena 
demonstrated the presence of "higher" perceptual units, but 
most colleagues interpreted them as synthetic products rather 
than as primary realities. 

.Stout's Position 

The English-speaking world did not lack important figures 
whose critical thought had led them to similar views. Ward, 
and especially his disciple, Stout, had discerned, even if they did 
not loudly proclaim, the reality and significance of the " whole- 
ness" factor in perception. In the third chapter of his first 
volume on "Analytical Psychology" (1896), concerning the 
apprehension of form, Stout raises three fundamental questions, 
all of which were patently influenced by the contemporary Ger- 
man debate: 

1. Can the form of combination remain the same or relatively 

the same, while the constituents vary ? 

2. Is it possible to apprehend all the components of a whole 

without apprehending their mode of connection ? 

3. Can the whole be apprehended without apprehension of 

the parts ? 

To the first two inquiries he returns an affirmative answer, 
adding the following significant comment, "We do not wish to 
imply that the same elements may enter into different combina- 
tions without themselves undergoing modification. An element 
which is apprehended first as part of one whole, and then as part 
of another, is presented in two different points of view, and so 
far suffers transformation" (page 71 of Stout, italics mine). 
The importance of this recognition for later theoretical develop- 
ment was probably diminished by his negative reply to the third 

case of what Avenarius called "characters" in contrast to elements in his biological 

psychology. 



18 GESTALT PSYCHOLOGY 

item, in which he alleges that where there are no parts there can 
be no combination of parts either in fact or for thought, and that 
only in Wonderland can the grin be divorced from the cat. This 
type of reasoning illustrates beautifully the limitations of the 
older British school, for while unimpeachable on logical grounds, 
Stout's conclusion does not square with the plain results of 
observation. Existential reality is again confused with per- 
ceptual presence. One may know a person's face well without 
discerning the individual lines and curves of his cheeks, 18 

"' 7 The Work of William James 

The late Miss Calkins 19 has performed a special service by 
showing that pioneer honors for preparing the background of 
configurationism must be awarded to that arch-eclectic, William 
James. Let us examine some of his more notable dicta to see 
how this position is justified. 

In his Principles (1890), James fought a bitter and steady 
fight against the current psychic atomism, or, as he christened 
it, the "mind-stuff" theory. He protested against the obscure 
assumption that our mental states are composite in structure, 
made up of smaller states conjoined. Against this view, he 
delivered his most telling blows in the famous chapter on "the 
stream of thought." Almost at the very first sentence we read, 
"No one ever had a simple sensation by itself" (James, page 
224) . Like Heraclitus of old, James was profoundly impressed 
by the dynamic and changing character of personal awareness 
as the following excerpt shows : 

It is often convenient to formulate the mental facts in an atomistic 
sort of way, and to treat the higher states of consciousness as if they 
were all built out of unchanging simple ideas. It is convenient often to 
treat curves as if they were composed of small straight lines, and elec- 
tricity and nerve force as if they were fluids. But in the one case as in 
the other we must never forget that we are talking symbolically, and that 




ana jraraccisus, wiucn is prouaDiy causca uy me ract mat tney arc* now. lour-syJuaDie 
names with similar distribution of accent, as well as sharing the property of being rare 
and "classical." 

19 Critical comments on the "Gestalt-Theorie," Psychol. Review, 1926, jj, 135-158. 



ANTECEDENTS AND DEVELOPMENT 19 

there is nothing in nature to answer to our words- A permanently exist- 
ing "idea" or " Vorstclhtng" which makes its appearance before the foot- 
lights of consciousness at periodical intervals, is as mythical an entity as 
the jack of spades. (I, p. 236.) 

His insistence upon the importance of "transitive" states of 
consciousness and of the reality of "feelings of relation" is a 
note which constantly reappears. No Gestalt writer has satirized 
sensationalism more keenly as numerous passages testify. "The 
traditional ( !) psychology talks like one who should say a river 
consists of nothing but pailsf ul, spoonsful, quartpotsful, barrels- 
ful, and other moulded forms of water. Even were the pails 
and the pots all actually standing in the stream, still between them 
the free water would continue to flow" (I, page 255), "What- 
ever things are thought in relation are thought from the outset 
in a unity, in a single pulse of subjectivity, a single psychosis, 
feeling, or state of mind" (I, p. 228). And in a footnote he 
preaches the doctrine of the primacy of the whole in the follow- 
ing terms, "In a sense a soap-bubble has parts ; it is a sum of 
juxtaposed spherical triangles. But these triangles are not sepa- 
rate realities. Touch the bubble and the triangles are no more. 
Dismiss the thought and out go its parts. You can no more 
make a new thought out of 'ideas' that have once served you 
than you can make a new bubble out of old triangles. Each 
bubble, each thought, is a fresh organic unity, sui generis'' (I, p. 
279). If any further evidence be necessary to demonstrate that 
James was thinking very definitely along lines which were ulti- 
mately to assume clarity in the Gestalt theory, it is certainly 
conveyed in this positive statement, "All brain-processes are 
such as give rise to what we may call figured consciousness" 
(II, p. 82). 

Dewey and the Biological Trend 

A few years later one of James* famous disciples, John 
Dewey, expressed the discontent of the period in an historic 
paper on the reflex arc concept. 20 The big problem of that 



of Chicago Press, 1903) make thinking in terms of natural wholes the criterion of sound 
speculation. 



20 GESTALT PSYCHOLOGY 

period was to explain how a movement became adjusted to a 
particular sensation which released it, and in Dewey's view the 
two were part of the same unit, viz., the reflex circuit. He wrote, 
"The reflex arc idea is defective in that it assumes sensory 
stimulus and motor response as distinct psychical existences, 
while in reality they are always inside a coordination and have 
their significance purely from the part played in maintaining or 
reconstituting the coordination; and (secondly) in assuming 
that the quale of experience which precedes the 'motor' phase 
and that which succeeds it are two different states, instead of 
the last being always the first reconstituted, the motor phase 
coming in only for the sake of such mediation. . . . There is 
just a change in the system of tensions. . . . What the sensa- 
tion will be in particular at a given time will depend entirely 
upon the way in which an activity is being used. It has no fixed 
quality of its own. The search for the stimulus is the search 
for exact conditions of action; that is, for the state of things 
which decides how a beginning coordination should be com- 
pleted." 

It should not be assumed from the fact that most of these 
writers possessed a heavy philosophical bent that the concept 
of organic unity or wholeness was a trivial speculative concern 
of unscientifically-minded metaphysicians, 21 for biologists in this 
period as well as psychologists had begun to feel uncomfortable 
in the presence of phenomena which did not fit the conventional 
scheme. In 1894, for example, cytological research had advanced 
far enough so that E. B. Wilson 22 could say, "I will point out one 
all-important point ; that the cell cannot be regarded as an iso- 
lated and independent unit. The only real unity is that of the 
organism, and as long as its cells remain in continuity they are 
to be regarded, not as morphological individuals, but as spe- 
cialized centers of action into which the living body resolves 
itself, and by means of which the physiological division of labor 
is effected." Conceivably this is the intellectual atmosphere out 

21 Not that this feature is wholly absent. General Smuts' recent ambitious volume on 
Holism and Evolution (New York, 1926) is robbed of potential influence by the failure of 
the idea to carry with it the necessary support of sufficient empirical data. His book 
repeats Hegel's methodological error on a smaller scale. 

22 The Mosaic Theory of Development (Woods Hole Biological Lectures) Boston, 1894. 
The nee-Aristotelian viesv that life is systemic organization is anticipated by Aristotle's 
remark that a dead finger is not a true finger because it has lost its organic relation. 



ANTECEDENTS AND DEVELOPMENT 21 

of which behaviorism with its slogan of the organism-as-a- 
whole developed, although It is curious that the logical implica- 
tions of Wilson's position which were so clear with respect to 
morphological units were not discerned with functional ones, 

Contributions o David Katz and Karl Biihler 

With all this evidence at hand indicating that many of the 
-more fundamental doctrines of the contemporary Gestalt school 
had been "anticipated" a generation or more ago, why didn't the 
new movement become conscious of itself much earlier ? Frankly, 
this is one of the riddles in the history of psychology. 23 One can 
only hazard a few guesses in explanation. Perhaps psychologists 
were still too prone to take their methodological cues from classi- 
cal physics and had to wait until the advent of the quantum 
theory had rocked this complacency ; possibly the great but much 
misunderstood work of the Wiirzburg school 24 In the first decade 
of this century was an indispensable preliminary to any new 
departure in the science; most likely, the professional experi- 
menters needed the intervening years In which to adjust their 
laboratory practices to the new ideas born far from the work- 
shop. 

Allotting then to the proponents of Gestaltqualitat the role of 
John the Baptist, what event shall we consider to mark the first 
definite appearance of the new faith? We have decided above 
In favor of Wertheimer's Hdbttitationsschrift of 1912; largely 
because this became the Scripture of a personal following, but 
partly because the only other possible contenders, Katz, Biihler, 
and Krueger, have repudiated the later developments of the 
theory. Without attempting for the present the intricate task 
of distinguishing among the several varieties of Gestalt for to 
paraphrase a famous political utterance, we are all Gestaltists 

now it will be profitable to examine the contributions of other 

distinguished psychologists to the movement. 



Gesta! 

the dramatic consequences of the po: 



is a common way of dismissing the Pretensions of confiffitratonisni See a typ cal article 
by Commins, "Some Early Holistic Psychologists," /. of Philos., }$& *$> *f~W' l{ 

a*Ktilpe's attacks upon associationism vie in thoroughness with those of Wertheimer. 
Sec his Outlin-cs of Psychology (Titchener's translation), 1909, p. i9. 



22 GESTALT PSYCHOLOGY 

In 1909, Katz offered a most exhaustive example of the new 
phenomenological type of analysis in his monograph, Erschei- 
nungsweisen dcr Farbcn (modes of appearance of colors). 
The main accomplishment of this comprehensive study with its 
variety of new experiments, observations, and viewpoints was 
the final demonstration of the absence of a one-to-one corre- 
spondence between the experience of color and the external 
stimulus. This fact as such had been known since Hering's 
elegant visual work on contrast, adaptation, pupillary action, and 
the influence of peripheral factors, but Katz's further descrip- 
tions compelled a new orientation toward the problem, par- 
ticularly since he added a distinct set of perceptual laws and a 
novel terminology. For example, one of his generalizations on 
field magnitude states that under non-normal illumination, the 
color quality of objects approaches the appearance under normal 
lighting conditions in proportion to the increase in the area of, 
the visual field affected by the non-normal illumination. For 
instance, if one views the objects on a wall through a colored 
or smoked glass and either recedes in person from the wall, hold- 
ing the glass at a constant distance from the eye, or if one main- 
tains a steady position and brings the glass nearer to the eye, one 
finds that an increasing portion of the visual plane becomes 
affected by the non-normal lighting. In both cases, an adjust- 
ment to the illumination occurs; e.g., a white piece of paper is 
seen to become whiter as the "field magnitude" grows. 

Such a central adaptation to illuminating conditions occurs 
only in those colors which are the hues of definite things, i.e., 
which can be apprehended as qualities of the objects, as in the 
case of a "yellow banana.' 7 These "substantial" colors Katz 
called "surface colors" (Oberflachenfarbcn) and distinguished 
them from "plane" or "film colors" (Flachenfarben) , which may 
be illustrated by the color of the blue sky, most after-images, 
the hues which one sees in a spectral apparatus, or which one 
may discern when the eye is illuminated through the closed lid. 
These last do not appear as the colors of definite objects, but are 
generally vaguely localized and have a less compact structure 
than the "surface colors." Such finely executed experiments 
gave the death-blow to the "constancy hypothesis," which was 



ANTECEDENTS AND DEVELOPMENT 23 

cleverly satirized by Kohler 2 * a few years later as having long 
been bolstered up by appeal to auxiliary hypotheses such as 
unconscious judgment or inference, attention, experience, etc. 

Meanwhile, Btthler, the assistant and intellectual executor of 
Kiilpe, had been experimenting directly in the field of form per- 
ception. The idea of the dominance of perceptual wholes over 
their constituent parts first began to take concrete shape in his 
volume on Gcstaltiuahrnchniiing (1913), A sample experiment 
will best illustrate the nature of Biihler's contributions: He 
found that discrimination of linear magnitude was better when 
the lengths to be compared were parts of rectangular figures 
than when they were merely isolated. 26 The orthodox determi- 
nation of psychophysical limens was seen to be markedly de- 
pendent upon the configuration employed during the measuring 
process. 

Rubin and Experimental Phenomenology 

But the one work outside their own circle which gave the 
initial Gestalt coterie more aid and comfort than any other was 
the product of the Danish psychologist Rubin, a protege of 
Hoffding at Copenhagen. Oddly enough, the study was per- 
formed at Gottingen where experimental phenomenology ap- 
pears to have been well established about 1910 under the direc- 
tion of G, E. Mtiller, one of the most ardent champions of the 
legitimacy of the sensory reduction of all mental processes. 
Since its findings have played such a prominent part in the de- 
velopment of Gestaltthcoric, it will be rewarding to examine 
them in some detail. 

Rubin's original problem 27 concerned the fixation and recog- 
nition of figures, but in developing this research, he had to in- 

2t > "Ueber tmbemcrkte Empfinclungen uncl Urteilstauschungen," Zsch. f. PsychoL, 
1913, 66, 51-80. The constancy hypothesis assumed that the relation between local stimulus 
and sensation observed under one set of conditions held equally well under all circum- 
stances, provided that the state of the sense-organ was unaltered. 

20 This does not mean that spatial thresholds of parts arc always greater than those 
of wholes. If a circle is split so that two semi-circles with an intervening white space 



part ' in any cone: -. 

ically, however, a part is more than a whole of smaller range. 

27 Visucll wdhrg,cnoinincnc Figurcn, Copenhagen and Berlin, 1921, This is the edition 
commonly referred to. The original Danish monograph, which is a closed book to most 
scholars, appeared in 1915. 



24 GESTALT PSYCHOLOGY 

quire into the differences in the phenomenology of figure and 
ground. He had early noticed that if one sketched a square by 
means of four lines, one tended to describe it as composed of 
four elements ; but if instead it were produced by complete in- 
cision of a piece of paper, one tended to describe it as a single 
figural surface, i.e., a unit or whole with parts. He examined 
critically his own and others' experiences in the presence of 
meaningless black contours on a white background, and distin- 
guished between the "included" or positive field and the "includ- 
ing" or negative field : simplicity dictated calling the former 
"figure" and the latter "ground." 




Figure 3. Rubin's Vase Figure 

The black and white surfaces function alternately as figure and ground, 

In one of his first experiments along this line, Rubin showed 
that these two aspects of the total visual field behaved differently 
in memory; e.g.,; there was a pronounced tendency (as repre- 
sented by a correlation coefficient of .83) for the subjects to per- 
ceive as figure in a later test-series the same datum which had 
been fixated as figure originally a fact which he labelled the 
"fi-gural after-effect." This effect persisted even where the ab- 
solute magnitudes of the figures had been altered. Under spe- 
cial instructional conditions it appeared that if the field which 
had been experienced as figure (or ground) at the moment of 
fixation appeared as ground (or figure) in the later test it was 



ANTECEDENTS AND DEVELOPMENT 25 

not recognized as such. We can acquaint ourselves readily 
with this discovery if we look critically at the areas surrounding 
what are normally perceived as "figures" in a familiar rug; how 
often is amazement the result of identifying these new and odd 
contours ! 

The significance of this distinction for the theory of percep- 
tion can hardly be overestimated. Helmholtz (Physiological 
Optics, vol. 3, p. 163, 1910 ed.) had called attention to the fact 
that the Zollner or herring-bone illusion disappeared if he re- 




Figure 4. Another Reversible Experience 
Either a black "iron cross" or a white propeller can be perceived. 

acted to the white strips as snow-covered branches on a black 
ground rather than making the normal reaction of seeing the 
black strips as the objects fixated. But no one followed this 
suggestive hint until Rubin brought out those distinctions which 
are best observed in ambiguous figures (cf . Figure 3 ) . Normally 
one sees a plain vase ; it is only after a period of fixation that the 
profiles of two faces spring forth. What was once ground be- 
comes figure and vice versa. In Figure 4, the black and white^ 
crosses have a similar reciprocal relation. The important feaA 
ture to note here is that the surfaces emerge as a whole and nfy 



26 GESTALT PSYCHOLOGY 

piecemeal." 6 The ground (as one would expect from the desig- 
nation), is generally localized behind the figure, is less structured 
or differentiated, less penetrating, less independent, less mean- 
ingful, and in a sense, less "real" and lively than the figure. 
The figure has "things-character and the ground "stuff"-char- 
acter, a thing being considered as stuff plus form, as in the case 
of a water drop which is essentially water and a representa- 
tive spherical contour. 

A very interesting experiment probably inspired by Katz's 
procedure (vide supra) was performed as follows: Upon one 
part of a white sector of the black- white cross pattern (Figure 
4), a faint shadow was cast. The subjects were asked to alter- 
nate voluntarily between the two figured patterns and to tell in 
which instance the shadow was more pronounced. Invariably 
the observers reported a more marked shadow whdii the black 
cross was seen as figure. Evidently the central factors which 
produce an approximation of the brightness of the shadowed 
part of a white field to the brightness of the remaining field 
are more effective when the shaded region is a part of the field 
which is experienced as figure than when it is a part of the field 
which appears as ground. The important fact to note is that 
whichever field acts as figure for the moment functions as an in- 
tegrated wmit in its effect upon the events. Colors, e.g., which 
belong to the figure in a percept are more impressive and better 
retained than the same ones located in the "ground." Certain 
types of aphasia, again, may merely be extreme instances of 
the reversal of the normal figure-ground relationships. What 
a hideous abnormality it would be to see the ground of the Sis- 
tine Madonna 20 as figure ! While it is true that generally the 
area responded to as ground appears on the peripheral and less 
sensitive portion of the retina, this fact alone does not suffice to 
give an adequate causal interpretation of Rubin's phenomena. 

28 "Tn some cases sensory organization seems to change without any influence being 
exerted upon it from without, simply because processes which remain the same for some 
time in the same part of the nervous system tend to alter conditions in that part and to 
Mock their own path. We know that the same thing occurs in elcclrolytical cells in^which 
the current polarizes the electrodes and thereby creates forces opposed to itself." Kohler, 
Gcstalt Psychology, New York, 1929, p. 185. 

20 Tt has long been recognized in portraiture that the field or setting in which the 
figure occurs is an Important determinant of the artistic value of the^ whole. 

An unpublished Aussage experiment conducted by the writer with four pictures grave 
an average error of 55% for the "figure*' items and 73% for the "ground" items, revealing 
a definite memorial advantage in favor of the former. 



ANTECEDENTS AND DEVELOPMENT 27 

The basic rule which emerges from these researches is that 
a field cannot be experienced simultaneously as figure and as 
ground. Another fundamental rule expressing the probability 
that one field will be seen as figure rather than ground is phrased 
by Rubin as follows: When two homogeneous, differently- 
colored fields exist in which one is definitely larger than the other 
and encloses it, the predominant probability is that the smaller 
enclosed field will be apprehended as figure (Rubin, p. 79). 
There is also a strong inclination to view as figure the lower part 
of the field or at least that region which stretches from below 
upwards. Moreover, as in the case of the double-cross pattern, 
it is easier to hold as figure that cross whose sectors stand ver- 
tically and horizontally, i.e., more erect. The law of precision 
or Pragnanz (not yet formulated in Rubin's day) is here beau- 
tifully illustrated. Certain colors, too, appear to outweigh 
others; e.g., if blue and green systems comprise the cross pat- 
tern, the blue system tends to dominate as figure. The size of 
the sector system is likewise influential. 

Rubin draws some interesting episternological conclusions 
from his investigation. The fact that the contour exercises its 
major formative influence on the enclosed rather than the en- 
closing field is a principle of fundamental importance in delimit- 
ing the boundaries of objects and thereby facilitating our knowl- 
edge of "things." 30 It is unsatisfactory and superfluous to ex- 
plain the customary limits of objects to the action of attention. 
All that is attributed to the attentive function is better expressed 
by the figure-ground relation. "If the field appears as figure, 
the form of the field has a greater degree of clearness ; but a 
lesser degree of clarity obtains where the field is apprehended 
as 'ground' " (Rubin, p. 99). 

It would be surprising if the contour which separates the 
figure from the ground did not arouse considerable interest con- 
cerning its properties. That something unique characterizes it 
is clear from Meissner's "paradoxical" observation that if the 
hand be immersed in quicksilver no pressure sensations appear 
within the skin area submerged, but only on the borderline be- 

30 In the letter E one sees three black horizontal projections to the right, but not two 
white ones to the left. The intervening white belongs to the ground and not to the figure. 
In the letter O the white within is phenomenally "livelier" than the white outside. 



28 GESTALT PSYCHOLOGY 

tween the free and the immersed regions. Rubin's further 
analysis shows that the contour is not a separate entity but rather 
determined by the nature of the figured surface to which it 
"belongs" ; e.g., in Figure 5, if the contour "belongs" to one 
field, it is concave, if to the other, it is convex. 31 From the con- 




Figure 5. The Dependence of the "Quality" of the Contour upon the Nature 

of the Area Embracing It 
The border line is convex if part of the black and concave if part of the white. 

tour to the line is a natural step and here, too, it was easy to dem- 
onstrate that the "character" of a line changed with the plane 
figure to which it "belonged." 

It would be a grave error to assume from Rubin's illustra- 

31 Compare this fact with another familiar observation: If one pours info a test-tube 
water, mercury, phosphorus, aniline, hexane, and gallium, six distinct liquid layers are 
formed. This is a stable system; if the cylinder is shaken, the layers will again separate. 
To the layman it is astonishing that six distinct fluids can be found, each of. which furnishes 
so hostile an environment to the molecules of the other five that they remain largely 
segregated. 



ANTECEDENTS AND DEVELOPMENT 29 

tions that the figure-ground relationship holds only for vision. 
An experiment by Wever and Truman 32 shows that it operates 
in audition as well and is probably a general characteristic of 
perceptual experience. Normally, sounds are heard against a 
background of relative silence but it is perfectly possible to de- 
termine the course of the auditory threshold in the presence of a 
tonal background provided by an organ pipe or any similar 
sound source. At first one has difficulty in hearing the figural- 
tone at all unless it is of high intensity, but an adjustment to 
the groundal-tone soon occurs and the figural-tone can be dis- 
criminated almost as readily as under other conditions. Cor- 
related with this change of threshold is a change in the ob- 
server's experience, a change which is best characterized as the 
emergence of one tone as figure and the recession of the other to 
become the ground of the perception. 

Rubin's work represents a type of experimental observation 
in which the analysis of sensory components is eliminated in 
favor of the more direct description of one's experiences. The 
attempt to build a scientific psychology upon the detailed pro- 
tocols of trained introspectors is abandoned in favor of reports 
which aim to convey the immediate "appearance" or impres- 
sion. It is, therefore, a sample of that experimental phenomen- 
ology which seerns to have originated in the suggestions con- 
tained in the writings of the distinguished contemporary phil- 
osopher, Edmund Husserl. Husserl's own writings vie in ob- 
scurity with those of Kant and Avenarius, and a lengthy discus- 
sion of his neo-Cartesian logic and metaphysics would be out of 
place here, save that historically he unquestionably contributed 
something to the intellectual background from which configura- 
tionism emerged. 

Gurwitsch 33 has given what are probably the most useful 
hints of the relation between Gestalt theory and phenomenology. 

32 /. Exp, PsychoL, 1928, j/, 98-112. A study now in progress at the Berlin labora- 
tory; employs the groundal tone alone as the effective stimulus, simply raising or lowering 
its intensity at periodic intervals, whereby it becomes functionally a figural tone set off 
from the preceding" and succeeding 1 vibrations. In this way the difference limen under 
special conditions can be determined, whereas the absolute threshold was the main concern 
of Wever and Truman. Simple reaction time, too, ^ varies inversely as the degree of dif- 
ferentiation between the figure and ground in the stimulus-situation. 

381. Altho 
object havin^ 
Gestalt writers. 



30 GESTALT PSYCHOLOGY 

Methodologically, both approach the psychic realm from the 
descriptive angle: objects in the customary meaning of the 
word vanish, leaving only noemata^ i.e., things which are known 
the world as it really is, is excluded in the account and only the 
world as it appears to be at the moment is considered. In their 
interpretations of cognition and epistemology, Gestalt and phe- 
nomenology rest upon a common base. What the latter calls 
the thenia or thing thought of is, of course, the "figure/' and its 
thematic field similarly corresponds to the "ground." We can- 
not think of an inkwell without logically implying its surround- 
ings any more than we can perceive a figure without a milieu of 
some kind. Just as an inkwell upon an office desk is different 
from one upon a piano keyboard, so a thought or idea in one set- 
ting is felt as fitting and in another as inappropriate. This 
method of phenomenology is substituted for the analytical pro- 
cedure, and the wholes or gestalten are taken as they come, in 
their own right, and described, not in terms of any arbitrary set 
of elements, but in any manner that seems adequate to the prac- 
tical situation. Whatever introspection is employed is not a 
perception and report of the self, but a perception of experiences 
affecting* the organism. Throughout the methodological spirit 
of Gestalt one senses a tacit agreement with Bergson's claim that 
"the relation between the 'phenomenon' and the 'thing' is not 
that of appearance to reality, but merely that of the part to the 
whole" (Matter and Memory, Eng. trans., Allen & Unwin, 
1919), 



PART II THEORETICAL 



CHAPTER 2 

THE PHYSICAL BASIS OF THE GESTALT THEORY 
Kohler's Naturalism 

Gestalt psychology came of age in 1921 with the appoint- 
ment of Wolfgang Kohler to the chair of philosophy at the Uni- 
versity of Berlin, vacated by the retirement of Stumpf, under 
whom he took his doctorate. This post carried with it the direc- 
torship of the Psychological Institute with its well-equipped lab- 
oratory and other facilities offering ample opportunity for the 
"new school" to justify the scientific promises it held forth. 
The one thing which more than any other contributed to Kohler's 
election as ordinarius was undoubtedly the timely publication of 
a study in natural philosophy which most friendly critics con- 
sider his opus, vis., the book with the forbidding title, Die Phy- 
sischen Gestalten in Ruhe und im stationdren Zitstand 1 ( Static 
and stationary physical configurations). Since this volume has 
become one of the outstanding charters of the new movement 
as measured by the frequency of citation from both friend and 
f oe it is imperative that its contents be examined for whatever 
they may yield to a better understanding of this distinctive ap- 
proach to the problems of psychology. 

The first item which catches the eye as one turns the title 
page is the laconic dedication of the work to Carl Stumpf . Since 
most of the original Gestalt coterie had been pupils of Stumpf, 
it is interesting to speculate upon his part in fostering the move- 

1 Braunschweig, 1920; also Erlangen, 1924- My citations are based upon the second 
German printing. It is a misfortune that this book never appeared in an English version, 
whereas many lesser ones have. . , . 

I am indebted to my colleague, Dr. John G. Aston, Professor of Physical Chemistry, 
for a critical perusal of this difficult chapter. If any errors remain they are due solely to 
my defective interpretation of the original. 

3* 



32 GESTALT PSYCHOLOGY 

nieiit As far as the written record goes there is little to indicate 
that lie was in any way responsible for the later drift of events. 
His own researches were largely confined to analyses of tonal 
experiences and his systematic papers are transitional hybrids 
between what one may summarily designate as Wundtianism. 
and the phenomenology of the South-west German theorists. 
Stunipf has himself disclaimed any direct indoctrination, and 
whatever influence he may have exercised in the initial develop- 
ment of the Gestalt view is too vague and personal to be iden- 
tified. 

An unusual note is struck as soon as one observes that there 
are two prefaces : one for philosophers and biologists and an- 
other for physicists. In the former, Kohler begins with the 
statement that the "definite impression" of a visual figure, the 
"specific character" of a musical motif, and the "meaning" of 
an intelligible sentence are surely more than the sum of the re- 
spective colored points, tonal sensations and individual word- 
meanings. Such structures have been named gestalten and 
thereby subsumed under a concept which appears to be very 
fruitful scientifically. If psychology were the first or only em- 
pirical science, this new concept could advance by wholly dif- 
ferent means from those which it is expected to employ, because 
the prevailing criteria of scientific thinking are those which have 
been developed by other sciences with an advantage of several 
centuries on their side. Chemistry, especially in its law of com- 
pounds, would seem to offer some helpful clues, but the entire 
basis of chemical combination is too shifting at present to aid 
in the determination of fundamental psychological categories. 

The appeal to the biologist appears in the similarity of the 
physical and mental facts of organismic unity and adaptability, 
In his neurological and physiological hypotheses, Wertheimer 
had considered the brain action as a configured total process ; if 
now, it is possible to demonstrate the existence of physical ges- 
talten, a legitimate hope arises that the central brain processes 
are merely special cases thereof. 2 

2 Kohler does not mention the point here, but it is quite obvious that the logic of 
Gestalt is opposed to the usual interpretation of Mendelian factors in inheritance, even 
though this be the proudest quantitative area of modern genetics. Genes, chromosomes, 
and determiners are mixed together and an organism results. ^ Bxit if the part is altered 
by an alteration, of the whole this position certainly needs modification.. 



PHYSICAL BASIS OF GESTALT THEORY 33 

In the physicists' foreword, the Initial idea states that physical 
systems reach a given state such as equilibrium only when a con- 
dition for the entire system is fulfilled. For the system as a 
whole, the available work energy 3 must be a minimum, the 
entropy a maximum ; and the scalar or vector magnitudes, whose 
grouping constitutes the system, do not assume definite amounts 
and positions, but through their joint organization relative to 
each other, produce a lasting structure (Kohler, p. xvi). Tke 
simplest and perhaps ideal illustration is found in a closed electric 
circuit in which the current at any one point is determined by 
the_ nditions a ^ ec ^j n g a jj O ^ QT p O i n ts. On the other hand, a 

group of mutually isolated neighboring circuits comprise a phys- 
ical complex of independent single systems. In the first case, 
the structure of the system presents an objective unity no matter 
how extended in space the system may be, since at no point does 
the local condition preserve itself unaltered ; it is only in the total 
structure' that it is maintained. In the second case, however, the 
parts of the physical complex are a pure summative multiplicity, 
depending upon arbitrary human thought as to whether they 
will be combined or not (Kohler, p. xvii). 

If, as the foregoing has rendered likely, there are gestalten 
in visual space perception, basically nothing more is required to 
explain them than that we acknowledge that the optical sector 
of the nervous system possesses the properties of a physical 
system (Kohler, p. xix). A consideration of many striking 
so-called visual "illusions" lends plausibility to this view. How 
any object will be seen depends not only upon the particular 
stimulus affecting a limited receptor region but also upon the 
remaining stimuli which influence other regions. 

The keynote having been struck in his dual introduction, 
Kohler proceeds to the derivation of the first physical Gestalt 
factor. We observe gestalten only when the spatially extended 
perceptual field is filled in a non-homogeneous manner. A form 
must touch our body surface upon some restricted area, the pres- 
sure must be marked off against uninfluenced or more weakly 

3 It hardly seems necessary to remind the reader that in modern theory "energy is 
riot a physical thing, but rather what we would call a property of a system as a whole," as 
Bridgman remarks in his Logic of Modern Physics, New York, 1928. 



34 GESTALT PSYCHOLOGY 

stimulated skin, before one can speak of a tactile configuration, 
and it is only when color differences appear in the visual field 
that it is possible to see patterns. The implication is (although 
Kohler does not say so directly) that the contrast or opposition 
resulting from the "figure-ground" phenomenon is fundamental 
to all experience. 

Having granted this postulate, Kohler next asks us to con- 
sider the sensory surfaces as part of the central nervous system, 
an admission easy to make, since in many respects embryologi- 
cally, chemically, and structurally this certainly holds for the 
retina. The receptor surface, then, is simply part of .the somatic 
field like the nervous system. Under normal life conditions the 
somatic field will be differentially excited because stimulation 
falls unevenly upon the sensory areas. The neural reactions 
which ensue are distinguishable into five distinct types of proc- 
esses: (i) Complete equilibrium or rest; (2) Stationary con- 
ditions; (3) Quasi-stationary processes ; (4) Periodic-station- 
ary events; and (5) Dynamic sequences. All these varieties 
correspond to elementary distinctions in physics, but psycholo- 
gists have yet to make them part of their thinking in practice. 
The first type is most simply represented by a body lying upon 
a table, whose gravitation is compensated by the elastic counter- 
stress of the table-top. Were this same body to be falling in a 
vacuum, a dynamic process (class 5) would be involved. It 
could be roughly characterized as a process in which no phase 
of its activity showed signs of equilibrium. The fourth type 
is exemplified by the resonance of a membrane responding to a 
definitely and repeatedly given vibration. A stationary process 
(class 2} occurs whenever a system passes continually through 
the same procedure without altering any of its systematic prop- 
erties. The best example of this is the coursing of a liquid in 
pipes : if as much flows away from a given point as flows toward 
it, the complex does not change its attributes. Constant electric 
currents also belong under this head. If the changes in the con- 
ditions to which the system is subject are slowly made, so that 
the action of the specific dynamic factors is unnoticeable, a quasi- 
stationary process results (class j) (see discussion of this point 
in Kohler, pp. 4-6). 



PHYSICAL BASIS OF GESTALT THEORY 35 

Kohler's first use of this classification is in his hypothesis that 
the excitation consequent upon definitely prolonged stimulation 
is a chemical reaction which may be dealt with as a stationary 
condition, or at least as a quasi-stationary process. These proc- 
esses are isothermic (in the thermodynamic sense), a fact which 
permits the application of the law of mass action to them. Lean- 
ing at this point rather heavily upon Nernst's 'Theoretical 
Chemistry/' Kohler reaches the conclusion that a stationary re- 
action may be characterized as one in which the concentration 
(i.e., mass per unit volume) of all participating substances in 
the mixture is preserved constant during the process ; the abso- 
lute mass as such is secondary or irrelevant. More fully stated : 
The excitations of somatic fields under constant external condi- 
tions are quasi-stationary chemical reactions in dilute solutions 
in which ions participate. Consequently, the condition of excita- 
tion is always adequately determined by the concentrations of 
the reacting molecular substances including the ions (Kohler, 

P- 13)- 

The meaning of all this for psychology lies in the fact that 

a purely homogeneously aroused field (visual or otherwise} 
cannot develop anything other than a uniform stationary chemi- 
cal reaction. If, however, the sensory surface be inhomogene- 
ously but steadily stimulated, unequal stationary reactions occur 
in different parts of the neural field involved. If, in addition 
(as in the case in the nervous system), ions participate in the 
reaction, and if in the regions of varying reaction the ionic con- 
centration is not exactly equal (i.e., the concentration of positive 
ions is not equal to that of the negative ions), there arises at 
once along the entire boundary curve between two distinct areas 
a leap of electrostatic potential (Kohler, p. 16). Thus, a new 
and startling interpretation is given to the phenomenon of the 
contour as brought out in Rubin's figures (vide s^tpra) ; the 
same interpretation, of course, holds for tactual contours like 
points or geometrical outlines applied to the skin. Between the 
subordinate parts of homogeneous fields, no differences of 
osmotic pressure or ionic concentration arise; hence, jumps in 
potential do not take place since the arousal of these electromo- 
tive forces is dependent upon excitation differences in non- 



36 GESTALT PSYCHOLOGY 

homogeneous fields. The biological significance is found in the 
fact that "excited" substance becomes negatively charged in 
comparison with its unaffected environment a phenomenon 
which many have felt is the surest criterion of life itself (Kohler, 
p. 22). Applied to the figure-ground situation, this suggests 
that the figure is negatively loaded and the ground presumably is 
positive. In general, wherever a gestalt is perceived, leaps of 
potential occur along the constituent borders within the optical 
field. 

Having laid this naturalistic foundation, Kohler proceeds to 
ask: Are the properties of the parts adequate to explain the 
properties of the whole? This question may be answered in the 
affirmative when strictly summative or additive situations exist 
as in the case of weight, where the weight of the total is undoubt- 
edly identical with the sum of the weights of the separate items. 
But in the case of the potential difference presented above, we 
have a "functional" difference which appears only at the moment 
that the two electrical charges are brought into physical connec- 
tion. 4 

It is literally false physically for one to say that the potential 
difference of the system arises from the potentials of the parts : 
'on the contrary, the reversed statement would be appropriate. 
The word "difference" has normally misled one into thinking of 
the potential difference as an algebraic resultant, whereas it is 
actually a primary attribute of the total system (Kohler, p. 30). 
If, for example, we seek to bring the total system to a higher 
or lower, negative or positive "electrostatic level," the absolute 
potential magnitudes of both solutions may be altered at will, 
but the leap of potential between them remains always the same 
(Kohler, p. 31 ). Before the union of the parts, there exists no 
electromotive force, and the electrostatic potentials of the parts 
previous to the physical communication have nothing to do with 
the potential leap in the whole ; if the system be divided into 
parts again, the electromotive force vanishes. Such a system is 
a physical gestalt and satisfies both criteria of Ehrenfels for 

4 "To be the electropositive side of such a physical system is no less a gestalt property 
in a definite electrochemical whole than to he the dark side is a gestalt property in a 
sensory pair." Kohler, Gestalt Psychology, New York, 1929, p. 219, 



PHYSICAL BASIS OF GESTALT THEORY 37 

phenomenal configurations, viz., that ( i ) the whole is more than 
the separate experiences added together and that (2) the relative 
displacement of the items does not affect the essential identity of 
the whole this last being more commonly known as the criterion 
of transposability. The mathematical theory of Nernst shows 
not only that the electromotive force between two solutions can 
remain the same with utterly different absolute concentrations, 
but that this is always true in the case of equally proportionate 
concentrations on both sides. Transposition in the physical 
field occurs under exactly the same conditions as those which 
produce gestalten in psychology; i.e., through the maintenance 
of the proportions between the absolute data, as in the case of 
the "squarish character" of a square two inches on a side pre- 
served in the case of a square two feet on the side. Had psy- 
chology taken its cues more from physics and less from geometry 
it would have seen the full import of these hints long ago. 

Since the opposition of summation or addition versus system 
or gestalt 5 is prominent in the preceding discussion, Kohler 
finds it necessary to describe in some detail the difference be- 
tween the two concepts. In sociological theory, much is made of 
the distinction between an aggregation and an organisation and 
it is curious that Kohler fails to make use of the attractive illus- 
trations which this field provides. Perhaps, since he had a very 
special audience in mind as indicated by the two prefaces he 
deliberately avoided such attractive but semi-speculative ex- 
amples. Instead, simple cases of the differential effects of ar- 
rangement uporj. the essential character of the component items 
are presented. Six coins, for instance, may be laid upon a table 
so that they form the corners of a regular hexagon but this 
set-up does not alter their natures in any noticeable way ; if half 
of them are removed the other three remain in exactly the same 
order which they had when all six were therd. Similar facts 
lie at the bottom of the laws of the conservation of mass, of 

5 Erich Becher, in his critique of this position, holds that system and gestalt cannot 
be equated on all occasions. "Four-sideclness" is a true gestalt quality but color differences 
are founded upon systems or complexes. A shadow, again, is a physical gestalt but its 
parts are not causally coherent like the parts of electrical structures. See his review, 
**W. Kohlers physikalisclie Theorie dor physiologischen Vorgange," etc., ZscJi. f. PsychoL, 
1921, 8?, I-44- 



38 GESTALT PSYCHOLOGY 

electricity, and of energy, where the simple totalling of arith- 
metical quantities is decisive. A lack of reciprocal influence may 
be found in other natural distributions.' 5 Three stones, one in 
Australia, another in Africa, and a third in the United States, 
constitute such a nonfunctional grouping, since each can be re- 
moved or displaced without in any way disturbing the others 
(Kohler, p. 47). Philosophers, of course, have argued that 
every element in the universe has some kind of connection with 
every other, but in most instances such hypothetical connections 
have a vanishing effect upon observed happenings. We would 
be much surprised if in shifting the furniture of a room the 
table responded with a change of shape or a shift in position 
to the displacement of the chair ! - 

However, one must distinguish sharply between these in- 
stances of the arrangement of physical objects or "things" as 
such and the non-aclditive groupings which occur in physical 
systems. } Where thermodynamic laws arc applied to gases (as 
in Gibbs* phase rule) the principle governing the system pre- 
scribes what shall take place in the parts. The electric charge 
on the surface of a conductor, too, is a structure without genuine 
parts. If the real distinction between these two types is ob- 
scured, it is because our traditional logical criteria have neglected 
the structural feature. In the case of the electric charge as it 
is spread over the conducting surface, it would be more appro- 
priate to term that which for purposes of analysis one would 
normally tend to call a part a "structural moment." A "mo- 
ment," then, is that which is borne by the remaining structure 
and in turn supports it, the total structure itself being a unitas 
multiplex. The structure of an electrically charged hemisphere 
is strikingly different from the structure upon the half of a full 
sphere, since the hemisphere has sharp edges and it makes a big 
difference for the structure whether the hemisphere is solid or 
hollow ; but hollow or solid spheres condition the same electrical 
structure. This structure, be it noted, is also independent of 
the absolute amount of the charge. If the charge be increased 

If I gras the leg of a chair and pull, the entire chair follows but the air particles 
do not. There is a "dependence" in one case which is absent in the other. It is a meta- 
physical riddle why we see objects directly rather than the air-waves which mediate them. 
Cf. Fritz Ileider's original essay on "Ding und Medium" in Symposion, 1927, 2, 109157. 



PHYSICAL BASIS OF GESTALT THEORY 39 

or diminished, the total structure goes through a quick dynamic 
process or readjustment, i.e., a change of "level" occurs 
throughout. Kohler thus draws the highly important Inference 
that electrostatic charges arc transposable, and, therefore, sat- 
isfy the second Elirenfels criterion for gestalten. 

One of the postulates of Maxwell's electromagnetic theory 
Is that a structure cannot be fully determined without reference 
to Its "field."; This may be taken to mean that the "energy" 
present in a structure is configured. Science in the centuries be- 
fore Faraday behaved as though the limits of a body were de- 
termined by its elastic and chemical effectiveness a view still 
held by the scientifically nai've but there is no reason why the 
boundaries of an object cannot be equally well-defined by the 
extent of its electrostatic or magnetic influence. 7 Although a 
definite revolution in one's thinking is requisite in order to feel 
at home with this new viewpoint, its scientific "respectability" 
is beyond question. Kohler quotes with obvious approval a 
neglected passage in Maxwell's 8 great work : 

We are accustomed to consider the universe as made up of parts, and 
mathematicians usually begin by considering a single particle, and then 
conceiving its relation to another particle, and so on. This has generally 
been supposed the most natural method. To conceive of a particle, how- 
ever, requires a process of abstraction, since all our perceptions are re-^ 
lated to extended bodies, so that the idea of the all that is in our conscious- 
ness at a given instant is perhaps as primitive an idea as that of any 
individual thing. Hence, there may be a mathematical method in which 
we proceed from the whole to the parts instead of from the parts to the 
whole. (Vol. II, pp, 176-179.) 

Another more practical testimony to the reality of physical 
gestalten is found in the peculiar difficulty of all precise measure- 
ment. 9 It is only the novice who has not learned that the act of 

7 Wertheimer has also declared in an informal public lecture that "the effects of a 
thing belong to its innermost nature." 

8 In. the preface to his first edition, Clerk Maxwell remarks with approval, Faraday s 
methods resembled those in which we begin with the whole and arrive at ^the parts by 
analysis, while the ordinary mathematical methods were founded on the principle of begin- 
ning with the parts and building up the whole by synthesis." Cf. A Treatise on Electricity 
and Magnetism, 3rd ed,, Oxford, 1892, The skeptic will probably be tempted to express 
the hope that this formal resemblance of the two approaches really implies a genuine 

The methodology of tapping the field in order to establish the value of the charge or 
potential at the separate points with the aid of a probe has characteristic difficulties. The 
probe hegins to distort the field .of the conductor being investigated, and thus also its 
charge structure, as soon as it is introduced into the field. 



40 GESTALT PSYCHOLOGY 

observation itself affects the character of the event observed. 
Many persons have assumed this to be the unique disadvantage 
of the introspective technique in psychology, forgetful of the 
fact that the "inner behavior" of the observer is influential in 
all scientific observation : the difference, in other words, is only 
a matter of degree and not of kind. The ideals of "isolation" of 
the datum to be investigated and "control of all essential condi- 
tions" in general scientific method have probably conduced to a 
neglect of existing physical gestalten which are just as numer- 
ous as "things" because of technical difficulties in the attack. 
All the numerous "corrections" which have had to be attached 
to the measuring procedures in every field of science are silent 
witnesses to the reality of configurational influences. 10 

Since simple arithmetical or algebraic operations give us a 
wholly incorrect picture of the nature of the physical gestalt, 
Kohler turns to another branch of mathematics for guidance 
and finds it in the tlie,9J.y-o. integral calculus and the concept of 
functional equations ; in fact, he reaches the significant conclu- 
sion that gestalten exist wherever the integral calculus can be 
applied and vice versa. This implies that all cognized relations 
/must ultimately be quantitative in character. Furthermore, all 
cases in nature where equilibria of some sort occur are found 
upon closer examination to possess the properties of gestalten. 
Specific instances which require the Laplace equation for solu- 
tion, and, therefore, show that the differential calculus is also ap- 
plicable to configurations, are as follows : the magnetization of 
bodies without residual magnetism when influenced by perma- 
nent magnets; the equilibrium of a membrane stretched with 
uniform tension over an even plane; the stationary heat cur- 
rents in plates or three-dimensional bodies; the stationary dif- 
fusion stream of a dissolved substance in a solvent; the sta- 
tionary electric current in tri-dimensional bodies; and many 
problems in hydrodynamics, such as the stationary movement 

m ao The mformerl reader will undoubtedly know that this is only one of many views 
which physicists and psychologists have recently begun to share in common. Eddington, 
in discussing the consequences of the FitzGerald contraction, i.e., the fact that a rod 
moving- in the direction of its long axis is shorter than when it is in motion along the short 
axis, states: "Absolute distance, not relative to some special frame, is meaningless." Such 
a rock-bottom physical magnitude as length is a dependent function of the gestalt of which 
it is a part. Cf. The Nature of the Physical World, New York, 1929. 



PHYSICAL BASIS OF GESTALT THEORY 41 

of a fluid in an extended field when a rigid body of prescribed 
form is fixed in this field as an obstacle (Kohler, p. 1 19). Two 
famous questions in mathematical theory the related problems 
of Dirichlet and Neumann also refer directly to physical ges- 
talten. Most of these problems are rarely investigated in ex- 
perimental physics and are almost exclusively questions for 
mathematical physics to solve. Nevertheless, an experimental 
attack is by no means excluded one has simply to determine, 
if, in the case of a constant conditioning form or topography, 
an arbitrary local alteration of the extended structure can be 
introduced without causing the total system to react with a 
complete mass displacement throughout. These clearcut cases 
in which a change in the part immediately affects the whole are 
known as "strong" gestalten; if, however, the energy of the 
system is unordered (as in set-ups for studying the conserva- 
tion of energy), the presence of a "weak" gestalt is indicated. 
Even in these limiting cases, it is incorrect, or at least unwise, 
to speak of a "part" of a system; one remains closer to the 
heart of things if the term '^art^is, understood to meaiij 'struc- 
tural moment in an organization-." 

Resistance to the Gestalt brand of thinking is encountered 
because most Western European scientists, trained as they are 
in a certain variety of analytical research, are prone to see in 
it evidence of contamination with the vagaries of intuitive 
mysticism. Against such erroneous identification Kohler and 
his~confreres strongly protest. One view of the physical world 
would be that nature consists of purely mechanical links (Und- 
verbindungcn "and-connections" or plus-summations) be- 
tween independent parts, whose additive total constitutes re- 
ality. Another opposed view would be that nature contains no 
independent parts, but that alljconditi.ons j> cLnd eyents-a-re real 
only in a to^Luxmf^r^al -relation, all "parts" being products 
of abstraction. Kohler maintains'lTTaFTlTe first proposition is 
wholly false, but that even the second hinders rather than fur- 
thers an understanding of the Gestalt principle. As expressed 
above, it is less a result of empirical research than the emotion- 
ally-toned utterance of a human being weary of strenuous 



42 GESTALT PSYCHOLOGY 

thought and readily satisfied with a romantic-philosophic con- 
struction. The "union of all with all" is probably correct in a 
purely formal sense, for the observer of the disintegration of 
a radioactive preparation cannot demonstrate that some infini- 
tesimal influence from the fixed stars is not operative ; but such 
an attitude would contribute little to the understanding of the 
specific characteristics of limited types of processes (Kohler, 
p. 156). 

Not everything in the world can be regarded as relevant to 
a given system, for if that were so there could be no scientifically 
manageable isolated systems. If things were infinitely complex 
and all aspects of their interrelations were equally relevant to 
all purposes, nothing could be understood. Their very "togeth- 
erness" would have to be sensed by a kind of dumb mystic rap- 
ture. Thejgcistenca-of a ^scientific body of knowledge is itself 
evidence that natural systems can be found which are neutrally 
isolated from other systems. (Sidney Hook.) 

Logically, the existence of indefinitely vanishing effects can- 
not be denied, but the "cosmic" and "particular" quantities ap- 
pear to be related as the magnitudes i and io 10 . This is one 
among many points where Gestalt theory and the theory of rela- 
tivity come into contact. Nevertheless, the greatest enemy of 
the fruitful and concrete consequences of the configurationist 
principle in psychology would be the thesis that the "immedi- 
ately given" is a total state of consciousness as such. The older 
psychologists appreciated this dangerous cloudiness, but unfor- 
tunately leaped to the other extreme in their policy of punctate 
sensations. Both evils are avoided as soon as one recognizes 
that the laws of , science an the tem a f systems, i.e., "structures 
of finite extent" a generalization applicable to both physics 
and psychology. 

Another enemy of the true Gestalt position Kohler sees in 
the so-called production theory of the Graz school of psycholo- 
gists, a theory which viewed the elements as "funding contents" 
and the configuration itself as a "funded content." The gestalt 
in their eyes is "produced" by a strictly psychic process; the 
physical world as such possesses no gestalten. 11 Kohler views 
this theory as an unprofitable compromise which is still at heart 



PHYSICAL BASIS OF GESTALT THEORY 43 

attached to purely geometrical thinking, and rejects it as a 
sheer psychologisni, to which his opponents, one may readily 
suppose, would retort that his own position is tainted with 
"physicalism"! 

Having made this long excursion into the anorganic realm, 
Kohler returns to consider the organism in a chapter prophetic- 
ally labelled with a citation from Goethe, "Denn was innen, das 
ist aussen." (For that which is internal is also external.) The 
sensory processes underlying perception have in the past been 
wrongly assumed to arise from independent local peripheral 
stimuli, each insulated from the rest in the somatic field, and 
individually meaningless, acquiring significance only through 
the action of higher mental operations, such as association, at- 
tention, judgment, etc., upon them. To be sure, physiological 
contrast theories based upon the assumption of physical systems 
in the receptor surface (Hering) existed, but the resistance 
which they encountered showed how dominant the opposite 
ideology was. Nothing seemed clearer and more reliable than 




Brain level of cognition of shape 

Brain level of sensation (cortical ends of sensory nerves 

Peripheral end of sensory nerves O 

GRAZ VIEW 

Cortical ends of sensory nerves O > O > O r < > O * 

Peripheral ends of sensory nerves 

WERTHE1MER VIEW 



t t t t t 

O O O O 

GRAZ VIEW 
o > o > o > o *> o 

t t t t t 

o o o o o 



Some irreverent Yankees liave called this the difference between the one-steppers and 
the two-steppers! Nevertheless, the Graz writers are not to be lightly dismissed, par- 
ticularly since they represent an older school of conngurationisrn. They maintain that the 
gestalt is by no means tmivocally determined, either as to form or as to structure. Thus, 
four dots arranged in the form of a square may be seen as two horizontal lines, two vertical 
lines, or two oblique lines at right angles to each other. Benussi, the ablest experimentalist 
who followed Meinong's theoretical position, was, therefore, not altogether unjustified in 
holding that the determinant in equivocal ca'ses must be some psychical factor which unites 
sensory contents into patterns. 



44 GESTALT PSYCHOLOGY 

a correspondence between experience and reality accounted for 
on the basis of pure point geometry. 

Nevertheless, our visual world reveals an order of a much 
more striking and immediate kind. A homogeneous field is 
seen as uniform without "points" and consequently without 
"relations" between them, and where gestalten in the narrow 
sense arise they are seen as compact closed structures lifted up 
from the rest of the field. When Kohler asserts that the phe- 
nomenal field has non- or trans-geometrical properties, he 
simply means that the sensory response is conditioned by, but 
does not exactly correspond to, the stimulating situation. In- 
stead, the physiological process in the optical system has the 
same general properties possessed by physical spatial gestalten. 
Although this process may be extended in time, it acts as a 
complete objective unity. The optic-somatic sector is a physical 
system in more than just an analogous sense and the correlations 
between it and the phenomenal field are built upon the common 
ground of identity, i.e. /the likeness between the "two worlds" 
is a structural one. 12 A similar equivalence for the thought 
processes is also claimed. 18 This factual resemblance is not just 
a material one; the specific configurational features are them- 
selves preserved (Kohler, p. 193). Existing consciousness, 
then, is directly related to the real structural properties of psy- 
chophysical events and not arbitrarily linked to elements, which, 
by their very nature, are incomparable with it. 14 Physiological 
psychology, therefore, has been at a loss to explain how vibra- 
tions or displacements of brain molecules could produce the 
actual content of our mental life. But that is because the wrong 
physical concepts have been brought to bear upon the problem : 

12 Cf. an utterance of Wertheimer in one of Ins university lectures, "The material of 
sensation is indeed different from the material of the physical world, but the structure is 
the same." The thorough-going rationalism of Gestalt could hardly be better attested, for 
it is the very essence of rationalism to postulate symmetry, if not identity, between thought 
and reality. 

_ 13 The aid afforded by the device of Euler's circles in solving syllogistic problems is 
an illuminating case in point. Kohler's thesis is that perceptual and thought patterns are 
involved in a network of physical and physiological conditions which rrmst always be taken 
into account. _ By following this clue, the configurationists hope to. bridge the gap between 
psychology biology, and physics. An animal is an elaborate "machine" only if the molecule 
is a very simple "animate" individual. 

M Von Frey, whose early researches in cutaneous sensitivity were a bulwark to- the 
clementarists, later acknowledged the strength of Kohler's arguments. He says, "Progress 
lies in the conviction that the somatic processes corresponding to the psychical gestalten 
must have a similar structure." Cf. his paper, "ITeber Wandlungen der Kmpfindungen 
bei formal verscbiedener Remmg einer Art von Sinnesnerven," Psychol. l<orsch,, 19-3, 
3, 209-218. 



PHYSICAL BASIS OF GESTALT THEORY 45 

the most significant aspects of consciousness have least to do 
with this. If, instead, the essential structural properties were 
investigated, a more promising "lead" would present itself. 15 
This should not be taken to imply the existence of a double con- 
sciousness with a purely summative sensory multiplicity plus a 
configured consciousness over and above this. Such a view 
would play into the hands of the "production-theory" which has 
already been rejected. 

There are many facts in visual perception, especially the 
phenomena of color induction, which are difficult to reconcile 
with the older assumption that simple stationary streams course 
between the retina and the central areas. The difficulty is in- 




Figure 6. Optic Nerve Pathways Illustrative of the Electromotive Farces in 
Figure- Ground Percepts. (After Kohler) 

creased by the knowledge that electromotive, forces are devel- 
oped during retinal functioning. Now every electrical displace- 
ment produces a magnetic field within and about itself. This 
must aIsoTKb'13 for the nervous system ; consequently, whenever 
the paths of two currents run adjacent to each other, the two 
streams must each cross the magnetic field of the other. A 
rough illustration of how this may operate is offered by the 
following situation : 

Assume that we have before us upon a homogeneous gray 
ground a single white figure of limited size and simple form 

15 In a recent experiment which has aroused a good deal of acrimonious debate, Wever 
and Bray placed electrodes on the exposed auditory nerve of a cat and led off the nerve 
action currents through an amplifying system to a detecting apparatus. Sounds applied to 
the ear of the animal produced neural effects which corresponded to the frequency of these 
sounds. If confirmed, this means that identity and not similarity of wave number is 



involved, oee tnei 
1930, 13, 373-387. 



46 GESTALT PSYCHOLOGY 

such as a circle. Electromotive forces arise first, at points a and 
a' (see Figure 6), i. e., on the border between the unevenly stim- 
ulated retinal regions; and second, between the excited areas, 
and the optic pathways (Kohler, p. 206). The same electrical 
displacement must in general ensue throughout the homogeneous 
surroundings of the circle as in the "retinal circular field" itself, 
and since the figure extends over a relatively smaller retinal sur- 
face than its differently-toned "environment/' the average cur- 
rent density within its circular field is far greater than in the 
neighboring field. This proportion in energy density must be 
retained in the neural tracts as far as the central areas ; 16 in so 
far as the configured process is electrical in nature, 17 the "figure" 
corresponds to a much more lively state of affairs than the broad 
ground. The energy in the figure or gestalt in the narrow sense 
is closely packed, while in the ground it is thinly spread out. 
Had the situation been reversed with a grey circle of the same 
magnitude in a white environment, the significance of the elec- 
trical current would also have been reversed, but the region of 
greater electrical density would still have been the "figure." 
Consequently, the area which will be seen as the gestalt is not 
dependent upon the absolute brightness or coloration, which is 
distinctly a secondary factor. Indeed, the color-quaUtyjofjvisual 
surfaces is itself subject to configurational influences. 

A further ingenious step is taken with the application of the 
Gestalt principle to the historic Weber-Fechner law. In an 
earlier section, it was noticed that the leap of^potentiaHs^ inde- 
pendent of the absolute ionic concentrations, but dependent upon 
their -ratio to each other; e.g., in determining differential bright- 
ness threshol9s"~with two neighboring fields, one may triple the 
concentration on both sides and obtain entirely altered speeds 
of reaction, but the electromotive force jj^ajwa&_fhe same in 
other words, transposition in Ehrenfels* sense has occurred. 

18 This assumption, may well run counter to the general properties of nerve. There 
is some reason for believing that conductivity is much lower in the longitudinal section than 
in the cross-section. One investigator estimates this relationship to be about 1:5. See 
Hermann, "Ueber die Fahigkeit des Weissen Lichtes, die Wirkungen fertiger Lichtreize 
zu schwachen/' Zsch. f. SinnesphysioL, 1913, 47. 

1T Becher has indicated that Kohler's exclusive reliance upon an assumed electrical 
character of the nerve impulse is more favorable to the Gestalt hypothesis than an admis- 
sion of its chemical character (which is also real enough) would be. The latter feature 
can still be better understood if the transformations occur within isolated pathways. 



PHYSICAL BASIS OF GESTALT THEORY 47 

The electrochemical formula for the dependence of the jump 
in potential upon the basic variables is 

C" 
$ i ~ $2 = const, log -pr 

which is none other than the relation expressed by the familiar 
Weber-Fechner law that the intensity of sensory response in- 
creases in arithmetical series as the stimuli increase in geomet- 
rical series. This striking formal correspondence is probably 
more than accidental. The ill repute into which this generaliza- 
tion has fallen a fashion probably set by James is unfortu- 
nate, since it is probably the classic example of a great scientific 
generalization holding only within definite limiting conditions. 
All empirically-derived laws, when pushed to their upper and 
lower limits, reach zones where their validity becomes uncertain. 

The problem of mental and physical parallelism with which 
Fechner wrestled receives another extension in Kohler's treat- 
ment of symmetrical figures. In a circle, we have the simplest 
and most perfect example of symmetry. It is alleged that the 
phenomenal symmetry of the seen circle permits us to conclude 
that the psychophysical event in the region of the "figure" must 
likewise be symmetrical. However, this implies a functional 
resemblance between the two and not a geometrical one. The 
only possible way unless we postulate pre-established har- 
mony by which a geometrical symmetry in the stimulus pat- 
tern can give rise to a symmetrical form-perception is via a 
psychological process which itself possesses dynamic and func- 
tional symmetry. Lotze's 'local sign" theory with its auxiliary 
physiological index may, therefore, be dismissed as superfluous. 

This is a parallelism of a wholly different sort from that 
regnant a half-century ago, which postulated an exact relation 
between brain processes and states of consciousness with utter 
incomparability of both. But the universe is not as devoid of 
meaning as that and the parallelism of Gestalt insists that the 
highest psychic activities are literally, concretely, and in every 
case specifically related through their configurational properties 
to the somatic structures 18 (Kohler, p. 234). The dependence 

18 This is possible because physiologically a gestalt consists of the afferent, central, and 
efferent portions of the nervous system functioning as a unit: there are^no isolated paths, 
no single point excitations connected by strands; the form of the excitation and its spread 
are here of importance. 



4 g GESTALT PSYCHOLOGY 

of many forms of human thinking upon spatial gestalten has 
often been noted. 19 

On a simpler level, the kinship of tactile and visual gestalten 
a kinship not to be explained by associationist construc- 
tionsand the re-discovery of certain "optical illusions" (e.g., 
apparent movement) upon the skin furnish interesting support- 
ing evidence. It is evident that a doctrine which many persons 
felt was a disguised 'Vitalism" is merely positivism of a subtler 
kind. Driesch and other vitalists overlook the universality of 
that type of organization just demonstrated as existing in an- 
organic structures. 

1 The concluding chapter of this closely written work lays the 
groundwork for the law of Pragnans a term which may best 
be translated by "precision/ 3 although "eidotropy," a word 
earlier used by G. E. Muller to designate the tendency of an 
image to become "typical/ 3 would have some advantages. This 
theme is developed by noting that a physical gestalt results from 
a grouping of certain forces which operate upon a given form 
and only cease to transform it when the form has become stable. 
In all processes which terminate in such an end-state, the diffu- 
sion takes a ejection which tends toward a minimum of struc- 
tural energy. The question is raised, "How do gestalten look 
which correspond to smallest possible energy values?'' One of 
the well-known laws of electrodynamic displacement states that 
two currents which originally form an angle together tend to 
run parallel and in the same direction. Numerous electromag- 
netic illustrations could be adduced to show that these transfor- 
mations occur in the direction of heightened uniformity, sim- 
plicity, and symmetry. Another favorite instance, worthy of 
recording because of its dramatic persuasiveness, seems to in- 
dicate that the principle of Pragnanz represents a tendency 
present in all natural events. If one produces a soap film upon 

M Koliler insists strenuously upon the special type of parallelism which this implies. 
"When, some one experiences that flash by which a new idea or the solution of a problem 
comes to him, he will suddenly interrupt his walking or abruptly strike his head. Here 
both his inner experience and his outer aspect will exhibit the same interruption of conti- 
nuity.^ For similar reasons the activity of a man as it is seen by us will often show an 
organization which corresponds to the organization of his doing and planning, as experi- 
enced subjectively." Gestalt Psychology, New York, 1939, pp. 



PHYSICAL BASIS OF GESTALT THEORY 49 

a wire frame and then carefully places a si lall amorphous closed 
loop of thread upon it, nothing occurs at first to either film or 
thread; but if one pricks the membrane inside the outline of 
the loop, the thread immediately assumes a circular shape. 
Bursting the inner film subjects the thread to the exclusive 
control of the surface tension forces of the outer film which 
seek to enclose the greatest possible surface within the smallest 
possible space. 20 Small wonder that Kohler quotes with enthu- 
siastic approval a saying of P. Curie, "Asymmetry produces the 
course of nature. The absence of symmetrical elements is 
necessary for the occurrence of every natural event." 

That a strong tendency toward simplicity and regularity of 
gestalten exists can be demonstrated by a great many disparate 
facts in psychology. 21 At the Gottingen Congress of Experi- 
mental Psychology in 1914, Gelb reported that if three light 
points were briefly flashed successively in a straight line, they 
were seen as a symmetrical group, even when the two intervals 
had an objective ratio of 10 to 30 cms. This pressure toward 
symmetry of gestalt was especially marked if the time intervals 
from points I to 2 and 2 to 3 were equal or appeared to be equal. 
Moreover, a neat observation of Bourdon indicates that if three, 
four, five, or six points are arranged evenly about a center, phe- 
nomenally a triangle, square, pentagon, or hexagon result ; eight 
points, however, no longer compulsively result in an octagon, 
but yield instead a circle. Goethe in his color researches had 
long ago noticed that the after-images of square originals quickly 
became blunt at the corners and terminated in smaller rounded 
images, evidently because the circle is a more "pragnant" or 
precise gestalt than the rectangle. 22 

With this theme the most profound single volume in the field 
of Gestalt psychology is brought to a close. Even if metaphysi- 
cians succeed in the usually easy task of demolishing Kohler's 

20 Macli in his "Science of Mechanics" was the first to popularize this interesting 
observation. Cf. page 386 of the English translation (1893). 

21 It is curious that there are only six unsymmetrical large letters in the entire Latin 
alphabet F, G, J, L, P, Q, R. Similarly, the animals which Polonius saw in the cloud 
may be interpreted as an effort to "structurate" an amorphous mass. 

22 It is interesting to note how much of our language has meaning only with reference 
to the description of gestalten and not elements. "Straight" is significant only with respect 
to an implied structure; similarly with "sharp," "round," "simple," "symmetrical," and 
many other adjectives. 



50 GESTALT PSYCHOLOGY 

philosophical conclusions, the extraordinary value of the work 
in stimulating new forms of investigation remains unaltered. 23 

Gestalt and Relativity 

When the cessation of the war permitted German scholars to 
visit foreign shores again, they brought with them two scientific 
theories which a future historian may well consider the out- 
standing intellectual triumphs of the first quarter of this century 
the theory of Einstein and the principle of Gestalt. As befits 
the dignity of an elder brother the physical version of the new 
mode of thinking about nature won a popular hearing some 
years before the psychological approach was expounded, but 
when it did arrive keen minds everywhere recognized the kin- 
ship between the two. It would be false to assert that con- 
figurationism is simply a diluted application of the concepts of 
relativity to mental life. Since the two arose together within 
the same span of years ( 1905-1920) and shared the same milieu 
(German university institute life), there is much reason for 
suspecting that the contact was a one-way affair, for psycholo- 
gists ever since Weber's day have been taking many hints from 
the physicists without the compliment being returned. We do 
know that Einstein and Kohler were both students of Planck 
and it is altogether likely that they absorbed related ideas from 
a common speculative atmosphere. 

Wherein does the similarity of the two views lie? Ppmar- 
ily in the opposition of both systems to the summativc.and ad- 
ditive treatment of data. Einstein's declaration that the theo- 
rem of the addition of velocities employed in classical mechanics 
can no longer be maintained is matched by Wertheimer's asser- 
tion that the component members of a psychic structure are 
really interdependent, each of them being what it is by virtue 
of its place in the whole ! Similarly, the former's claim that the 
sole dynamical datum in two or more bodies moving relatively 
to each other is matched by the configurationist's denial of the 
reality of a single isolated sensation and insistence upon the 

33 Many persons are inclined to think that Kohler *s physical data are purely gratuitous, 
but he himself would object to viewing them as mere analogies. 



PHYSICAL BASIS OF GESTALT THEORY 

^^ Humphrey 2 * has shown 



these relations more clearly than any one else as the following 
quotation indicates: 

The teaching of Einstein is that absolute rest and motion are meaning- 
less for physical science and that motion can signify only the changing 
position of bodies relatively to each other. Thus, to take the simplest 
possible case, all that we can observe is a change in distance in the two 
bodies in mutual relation. We have then a dynamic .whole process, with 
respect to which it would be distorting the facts to claim, for the purpose 
of calculation or description that either of the reference bodies is unique 
as compared with each other. . . . 

Between these two theories, that of the German physicist and of the 
German psychologist, there stands out an immediate parallel. Each 
strikes at the discreteness of the cosmos, at the local autonomy respec- 
tively of the physical and the psychic. Each begins" with" aliegative, dis- 
tinctive attack^ sensations are not independent, velocities and spaces 
and times are not absolute. True, the notion of cosmic interdependence 
was not new. Thus, while we have Newton claiming the gravitational 
interdependence of physical matter, we find also a school of philosophers 
declaring an associational interdependence of psychic material, and 
indeed the parallel between these two modes of thought was brought out 
by Hume in a famous passage (Treatise.. on .Human Nature, Pt. I, sec. 
i), "Here is a kind of attraction which in the mental field will be found 
to have as extraordinary effects as in the natural." 

/ Reiser 25 has recently shown in even more detail the common 
ground shared by physical relativity and psychical relativity. 
In dealing with solutions with different ionic concentrations the 
answer to the question of whether a given solution is "electro- 
positive" or * 'electronegative' 1 depends on the relation of this 
given solution to the other solutions with which it is being com- 
pared. This is matched in the phenomenon of simultaneous 
colorjrontrast whefFt'he chromatic properties of the perceived 
pattern are not "absolute" but are "relative' 5 to the total situa- 
tion^ In some phenomena it is often impossible to disentangle 
one form of relativity from another. The Doppler effect, e.g., 
occurs in astronomical observations with a displacement of the 
Fraunhofer lines toward the red or the violet ends of the spec- 
trum, depending upon whether a star is receding or approaching 

2 * "The Theory of Einstein and the Gestalt Psychology: a Parallel," Amer. J. PsychvL, 
I9 2 4, 35, 353-359- 

25 "Physical Relativity and Psychical Relativity," Psychol. Rev., 1930, 37 f 257-263. 



52 GESTALT PSYCHOLOGY 

the observer, but the same relation also obtains in sound trans- 
mission. "We are all familiar with the experience of hearing 
the pitch of a bell at a railroad crossing change as the train 
we are in approaches, overtakes, and passes the ringing bell. 
When the train is approaching the crossing the source of the 
sound is moving (relatively) toward us. Accordingly, we over- 
take a greater number of sound waves per unit interval of time, 
and hence, the pitch appears to be higher ; as we recede from the 
bell fewer waves strike our ear per unit interval of time and the 
pitch of the sound appears to be lower. A variation in the rela- 
tive velocities of the observer and the thing observed produces 
the same effect as an objective variation of the stimulus when 
the observer is at rest with respect to the source of the stimulus."^ 
Whether this remarkable datum is to be treated by the physi- 
cist or the psychologist separately or in conjunction is a matter of 
indifference as long as one recognizes the community of law 
and standpoint involved in both approaches. 



CHAPTER 3 
THE PHYSIOLOGICAL FOUNDATIONS 

The physiological data in support of the Gestalt hypothesis 
are drawn largely fronj the newer researches in the field of 
cerebralJ0caliaa-tion, the 'genetic ^development of reflexes, and 
that division of biology which is concerned with ' 'regulation" as 
expressed by the facts of coordination in organic functions, re- 
generation of members, etc. It is not at all difficult to see why 
the configurationists should be deeply interested in the implica- 
tions of such phenomena, since in all these instances the question 
of the dominance of the organized whole over its members comes 
to the fore. Traditionally, neurologists have accepted special 
anatomical topography as the only explanation for order a 
restricted machine concept with monotonous elementary cur- 
rents conducted compulsively from a point of stimulation to a 
point of reaction, for which Gestalt wishes to substitute the 
principle of dynamical self -distribution. 

Aglance at the history of cerebral localization since the days 
of Gall and the phrenological controversy shows a peculiar 
cyclical oscillation of opinion between the view that the brain 
is a complex mechanism subdivided into definite areas with re- 
stricted functions (Broca, Munk, Wernicke, Ferrier, Sherring- 
ton) and the opposing position that the cortex at least operates 
as a unit (Flourens, Goltz^JLoeb, Franz, Lashley, Head) . The 
problem to-day remains one of the biggest issues" m science. 
The orthodox physiological psychology (perhaps best repre- 
sented by Ladd and Woodworth and Herrick's manual) upon 
which most contemporary American psychologists have been 
reared, sponsored a fairly precise sensori-motor localization, 
although acknowledging that the greater part of the cerebrum 
consisted of "silent areas" whose role was largely conjectural 
Neurology has always been influenced by psychological systems 
witness the name ''association fibers" and the "neo-phre- 

53 



54 GESTALT PSYCHOLOGY 

nology" of a generation ago was perfectly adapted to the piece- 
meal description of mental life then current. The early animal 
work of Franz, indicating that habits were lost after extirpation 
of certain regions but that they could be re-learned after recov- 
ery from the insult, first threw doubt upon the adequacy of the 
rigid doctrine of "centers' 3 as typified by Broca's area. ^ The 
resulting theories of substitute function and equipotentiality, 
and the clinical successes of re-education in aphasia emphasize 
the limitations of the strict specificity approach, and fit in nicely 
with the Gestalt theory which conceives the nervous system as 
highly permeable and without isolated segments functioning 
independently of the rest. 

Lashley's Cerebral Studies 

A pretty illustration of the inevitable drift of events may be 
found in following the sequence of Lashley's publications in 
this sphere of research. When he began his investigations about 
1916, his approach was that of an uncompromising behaviorist, 
fully persuaded that the conditioned response and the reflex arc 
would provide an adequate account of the adaptive conduct of 
organisms. Barely a decade later, the logic of his own findings 
had forced him into the opposite camp. In the volume x which 
best expresses this change of heart, the conclusions offer definite 
aid and comfort to none but the Gestalt theorist and to Spear- 
man's "g" factor hypothesis. What was the evidence which 
compelled this new allegiance? 

By means of a thermocautery, varying amounts and different 
parts of the cortex in a large group of rats were destroyed and 
their learning records made after recovery from the operation 
(10-30 days) compared with those of normal animals in such 
tasks as maze-running, brightness discrimination, etc. Compar- 
ing the records for the two groups according to errors, time, 
and number of trials, Lashley found that in general the cerebral 
lesions were attended by an increase in the amount of practice 
necessary to solve the problems, but that the degree of deteriora- 
tion in learning ability and retentvueness was proportional to 

1 Brain Mechanisms and Intelligence, University of Chicago Press, 1929. 



THE PHYSIOLOGICAL FOUNDATIONS 55 

the amount of brain tissue injured and independent of the area 
of the cortex affected. "Equal injuries in different cortical areas 
produce equal amounts of retardation. . . . The magnitude of 
the injury is important; the locias is not" (Lashley, pp. 58-60), 
The gap between the normal and operated rats varied with the 
nature of the problem to be learned; in the case of brightness 
discrimination, the operated animals were even superior to the 
normals ! Previous study had indicated another peculiarity of 
the visual function. "If the animals are first trained and then 
subjected to operation, the habit is abolished by any extensive 
lesion within the occipital third of the cerebrum, and the amount 
of practice necessary for re-learning is closely proportional to 
the extent of the injury. . . . Destruction of any other than 
the occipital region does not affect the retention of the habit. 
If the destruction is made before training, it has no effect upon 
the later formation of the brightness habit, even though the 
entire occipital third of the cortex is extirpated. Thus, the habit, 
once formed, is definitely localized, in the sense that it is de- 
pendent upon a definite part of the cortex for performance. 
Its initial formation, on the contrary, is not conditioned by the 
presence of any part of the cortex; the learning process is not 
localized, in the foregoing sense" (Lashley, p. 86). 

Another startling experiment showed that even the kinaes- 
thetic cues which Watson had earlier apparently demonstrated 
to be the basis of the rats' maze-learning were not only neg- 
ligible but conceivably irrelevant factors in explaining the per- 
formance. Since Lashley's language is so precise on this point, 
one can do no better than cite the appropriate passage, "The 
ability to perform the maze habit was unaffected by complete 
transsection of the fasciculus gracilis and fasciculus cuneatus at 
the third cervical level, although marked disturbances of kinaes- 
thetic sensitivity resulted. Section of all ascending tracts in the 
lateral columns of the cord likewise produced no disturbance in 
maze-running. After such lesions the animals ran the maze 
accurately in darkness when all other directive sensory cues 
seemed ruled out. . . . All of this evidence opposes the view 
that reduced learning ability for the maze or the loss of the maze 
habit after operation is the result of sensory defect as such. 



56 GESTALT PSYCHOLOGY 

The habits are not lost after destruction of the peripheral sensory 
mechanisms, and this lack of effect of sense privation is not due 
to shift in the use of other sensory directive cues. The habit 
is lost after destruction of the visual cortex in animals which 
were blind during learning, which indicates that the loss is due 
to some other than the sensory function of this area" (Lashley, 
pp. 114-115). 

As if this conclusion alone were not enough to shock the 
schema of a conservative neurology, Lashley proceeds to dem- 
onstrate the indifference of effect where the efferent pathways 
are interrupted' as well as the afferent. To quote again, "In one 
set of animals, the pyramidal tracts were completely cut; in 
another, the rubrospinal ; and, in the final series of cases, all 
possible descending tracts were interrupted. All these animals 
ran the maze in retention tests without errors. They showed 
marked incoordmations, greater difficulty in negotiating the 
turns of the maze than did any of the cases with cerebral lesions, 
but absolutely no disturbance of orientation in the maze. This 
shows the ability of the rat to perform the maze habit after 
destruction of any of the organized descending tracts of the 
cord" (Lashley, p. 115). 

It does not require more than an elementary knowledge of 
general psychology to see that such findings, if verified, pro- 
nounce the death knell of the standard theories of the reflex 
arc, synaptic changes, etc. Lashley, himself, is fully aware of 
these implications, although his phraseology occasionally be- 
trays a reluctance to admit the utter incompatibility of these 
two positions. Notice how the following passage supports 
Koffka's earlier contention, even to the extent of employing a 
similar illustration : "With the eye fixed and a pattern moved 
across ^the field of the macula, the same reaction (e.g., naming 
the object) may be elicited at any one of a thousand points, no 
two of which involve excitation of exactly the same retinal cells. 
To say that a specific habit has been formed for each of the pos- 
sible positions is preposterous for the pattern may be one never 
before experienced. The alternative is that the response is deter- 
mined by the proportions of the pattern and within the limits 
of visual acuity, is independent of the particular cells excited." 



THE PHYS- .,,' OUNDATIONS 57 

This means that, not o fl ' , but also in the central projec- 

tion on the cortex, there . : . of stimulation such that the 

same cells are rarely, if , 1 by the same stimulus, yet a 

constant reaction is proc tty of the visual cortex must 

resemble that of one of . -in which a pattern of letters 

passes rapidly across a st< , amps. The structural pattern 

is fixed, but the functio '. . ' .-. over it without limitation to 

specific elements. . . . 

Neurologically, these ' ;. ' i be in the nature of ratios of 

excitation, patterns witl ^ axcci auatomical substratum, since the 
sensory and motor elements of a situation may change fundamentally 
without altering its logical significance. We seem forced to the conclu- 
sion that a final common path may be somehow sensitized to a pattern of 
excitation so that it will respond to this pattern in whatsoever part of 
the nervous tissue it may occur. In the simplest cases, the relationships 
forming the basis of reaction seem expressible as ratios of intensity of 
excitation ; in others, as ratios of spatial extent or temporal distribution. 
The relationships involved in insight are more difficult to analyze, but 
there is in some instances sufficient similarity to cases of sensory dis- 
crimination to suggest that the basic mechanism must be fundamentally 
the same. 

The problem of reaction to ratio thus seems to underlie all phases of 
behavior to such a degree that we might be justified in saying that the 
unit of neural organization is not the reflex arc or the system of recip- 
rocal innervation but is the mechanism, whatever be its nature, by which 
reaction to a ratio is produced (pp. 158-159-160-161). 

Did ever avowed behaviorist speak the dialect of Gestalt more 
plainly? 

A brief excursus into the field of educational psychology will 
help to indicate some of the practical influences which a belated 
but inevitable recognition of the material presented above must 
exert. The familiar Thorndiklan position (known to teachers 
throughout the length and breadth of the American continent) 
maintains that if learning is restricted to particular synapses, 
there can be no influence of training upon other activities than 
those actually practiced ; any improvement in unpracticed func- 
tions must be the result of nervous connections which they have 
in common with the practiced activities. 2 The rejection of the 

2 Although transfer effects may be small they are real enough. Arabic numbers traced 
with the finger-tip upon any part of the body are easily identified, but it is hard to imagine 
what the "common elements" responsible for this overlapping can be. Even if the subject 
must convert the tactile-motor impressions into a visual image, how has he learned to do 
this? 



58 GESTALT PSYCHOLOGY 

older doctrines of "formal discipline" and "transfer of train- 
ing," leading indirectly to a great variety of curricular changes, 
seems to have been based as much upon the plausibility of such 
reasoning as upon any convincing experimental evidence. How- 
ever and here is where one experiences most vividly the con- 
sequences of an ever-changing body of scientific facts it is 
very doubtful if the same neurons or synapses are involved even 
in two similar reactions to the same stimulus. "Our data seem 
to prove that the structural elements are relatively unimportant 
for integration and that the common elements must be some sort 
of dynamic patterns, determined by the relations or ratios among 
the parts of the system and not by the specific neurons activated. 
If this be true, we cannot, on the basis of our present knowledge 
of the nervous system, set any limit to the kinds or amount of 
transfer possible or to the sort of relations which may be directly 
recognized" (Lashley, p. 173). 

But these animal experiments of Lashley are not the only 
ones which lend a most convincing touch to the correctness of 
the Gestalt hypothesis in the field of physiology and neurology. 
The manner in which grammatical forms function in speech is 
hard to reconcile with the classic assumption of confined path- 
ways. One of the great riddles of syntactical expression, so far 
as the neurologist is concerned, is that once we learn a new word, 
we use it in correct grammatical relations in limitless combina- 
tions with other words, without having to form new associa- 
tions for each setting. How is this possible? In his Presiden- 
tial address 3 before the Ninth International Congress of Psy- 
chology the full import of which has not as yet been appre- 
ciated Lashley offers a tentative explanation in terms which 
fit neatly into a configurationist context : 

Unity of action seems to be more deeply rooted than even structural 
organization. In working with animals and human beings, I have been 
more and more impressed by the absence of the chaotic behavior which 
we might expect from the extent and irregular form of the lesions. 
There may be great losses of sensory or of motor capacities, amnesias, 
emotional deterioration, dementia but the residual behavior is still 
carried out in an orderly fashion. It may be grotesque, a caricature of 

^Karl Lashley, "Basic Neural Mechanisms in Behavior/* PsychoL Review, 1930, 



THE PHYSIOLOGICAL FOUNDATIONS 59 

normal behavior, but it is not unorganized. . . . Even dementia is not 
wholly unintelligent. It involves reduction in the range of comprehen- 
sion, in the complexity of the relations which may be perceived, but what 
falls within the patient's range is still dealt with in an orderly and 
intelligible fashion. . . . 

Such phenomena suggest that the nervous system is capable of a self- 
regulation which gives a coherent logical character to its functioning, 
no matter how its anatomical constituents may be disturbed. If we could 
slice off the cerebral cortex, turn it about, and replace it hindside before, 
getting a random connection of the severed fibers, what would be the 
consequences for behavior? From current theories, we could predict 
only chaos. From the point of view which I am suggesting we might 
expect to find very little disturbance of behavior. Our subject might 
have to be re-educated, perhaps not even this, for we do not know the 
locus or character of habit organization but in the course of his re-edu- 
cation he might well show a normal capacity for apprehending relation- 
ships and for the rational manipulation of his world of experience. 

This may sound like a plunge into mysticism, but an example from 
another field will show that such selfregulation is a normal property of 
living things. Wilson and later Child have crushed the tissues of 
sponges and hydroids, sifted the cells through sieves of bolting cloth and 
observed their later behavior. The cells are at first suspended inde- 
pendently in the water, but may be brought into aggregates by settling 
or centrifuging. Starting as flat sheets, they round up into spherical 
masses and begin differentiation. Embryonic stages may be simulated 
and eventually adult individuals with characteristic structures, mouth, 
hypostome, tentacles, and stalk in normal relative positions are produced. 
In spite of the abnormal conditions to which it is subjected, the formless 
mass of cells assumes the structure characteristic of the species. Of 
course, many abnormal forms appear, but even these follow the charac- 
teristic scheme of organization (pp. 19-20). 

Driesch's Embryological Experiments 

The famous growth studies of Driesch bear directly upon 
the problem of regulation as interpreted by the Gestaltists, save 
that they are far from adopting the familiar conclusions which 
the distinguished vitalist has drawn. In one of his experiments 
with the larvae of a sea urchin, he destroyed one cell among 
four double cells. According to the "machine" theory of ajrjigid 
celj^uet, one would expect that if the creature developed further 
a "cripple" would result, because of the lack of the correspond- 
ing part. Actually a complete organism appeared whose total 
size was one-fourth less than normal. These results have been 



60 GESTALT PSYCHOLOGY 

confirmed by Roux in similar studies made in a later stage of 
development. The phenomenon evidently appears because of 
the division of another cell which assumes the function of the 
one removed. To Driesch, this is evidence that the yyhole serves 
as the natural "cntelechy" whereas Gestaltists prefer to" think 
in terms of Pragnanz, a principle of lesser metaphysical shading 
because of its more recent promulgation. 4 

Weiss' Transplantation Work 

In one of those unusual investigations which Americans have 
been led to consider typical of the Viennese school of biologists, 
Paul Weiss amputated complete mature limbs of salamanders 
and transplanted the extremity by placing it beside a normal one. 
In the many instances in which perfect healing ensued, the trans- 
planted limb hung loosely at the side of the body during- the first 
few weeks as a result of the severing o the nerve-fibers. After 
an interval of uselessness, the transplanted limb slowly recovered 
the power of motion and finally was functionally normal, its 
action in no way being distinguishable from that of an intact 
member, J At this stage, the normal and the transplanted extrem- 
ity invariably moved together like a pair of well-coordinated 
structures. 

In all cases where the supernumerary limb received a part 
of its reinnervation from the limb level oi the spinal cord it 
finally "came to perform precisely as did the host limb. Weiss 
refers to this as the homologous function of the supernumerary 
and host. When an arm was inserted in the region of the host's 
leg, it performed homologously with the host limb, i.e., by bend- 
ing at the elbows when the host bent its knee, bending the wrist 
with the host's ankle flexions, etc. This effect held irrespective 
of the point of attachment, the location, or the orientation of 
the grafted member. In one experiment, the tipper arm was 
removed and the elbow inserted where the shoulder had been. 
The hand nevertheless continued to function (after recupera- 
tion) as a hand, never as an elbow, whose place relative to the 

BerKt, 8 ^ (Kritischc Theorie der FormWdung, 

IJw ?5c 8f rt U e * a n hea Tr y ' * tl i Q - G stalt direction. Both have been influential in moulding 
f S ^ttjrw, according to his own admission. (It may be worth noting that a century 
ago, Von Baer compared the growth of an individual with a melody ) ceniury 



THE PHYSIOLOGICAL FOUNDATIONS 61 

body it now occupied. Certainly neither re-established direct 
connection nor "learning" in the usual sense account for these 
phenomena-. - 

Obviously, some kind of connection with the central nervous 
system of the new animal must have been established. Weiss 
offers a theory of nervous resonance 5 to explain this extraordi- 
nary fact, according to which each muscle does not react to every 
efferent discharge, but only to excitation of a quite definite form 
characteristic for it (isochronisms). He believes that the 
muscles, end-plates* and possibly the motor fibers (provided 
the latter acquire specificity under the dominance of the par- 
ticular muscle) must themselves possess properties of selectivity 
which enable them to respond to the appropriate action currents 
and not to others which reach them. His evidence indicates a 
dynamic coordinated interaction ^between (i) a central "broad- 
casting", system, (2) transmitting motor nerves, and (3) a 
motor receiving system, wEtfcE we can understand only in terms 
of the properties of all three. Weiss agrees with Kohler's 
hypothesis to the extent that he rejects any point-to-point co- 
ordination between periphery and center and discards the notion 
of constant geometrical-anatomical connection in favor of a 
more dynamic distribution process. 

Marina's Eye-Muscle Surgery 

A rare study which the Gestalt theorists delight to cite as 
confirmation of their viewpoint, but which seriously needs com- 
petent verification, is the novel experiment of A. Marina on the 
effects of interchanging the eye muscles of the ape. Marina 
published in 1905 and 1910 the results of his visual operations 
in which he interchanged first the medial rectus and lateral rec- 
tus^ and later substituted the sugcbtHrccttir'f6r - 'fhe lateral 
rectus. This meant in the first instance that the eye was moved 
outward by the previously inward-moving muscle, and vice 
versa; while in the second instance, the muscle moving the eye 

Best described in an article on "Erregungsresonanz tmd Erregungsspecifmtat" in 
Ergebmsse dcr Biologie, 1928, .?, 1-151. The neurological explanation of Gestalt phe- 
nomena must be sought alonsr such lines as here inrliratprl nr ; iiii/*+;n*i ifk T ,-..*" 



tv ^ u.^A.~que r 3 

_ , ._, , - v j - ~ww .,,_..*>. Tia,y be found 

* ^*s*ng Conceptions in Physiological Psychology" by H. R. De Silva and W. D. 
Ellis, /. Gen, PsychoL, 1934, u, 145-159.) 



62 GESTALT PSYCHOLOGY 

outward was eliminated and its place taken by a lifting muscle. 
Having accomplished this curious rearrangement of the natural 
alignment, the animal ought to have exhibited a most extraordi- 
nary type of nystagmus on the principle that a definite impulse 
is conducted from the brain center through a fixed pathway to 
each muscle. Instead, as soon as the wound had healed, the 
voluntary and automatic sideward movements of the eye \yere 
executed in normal fashion ! Certainly, as the author himself 
suggests, such an outcome argues very strongly for the indif- 
ference or lack of pre-determination of function in the con- 
duction pathways. 

Coghill and Minkowski on the Origin of Reflexes 

Coghill is another experimenter whose work is a powerful 
aid to the configurationist cause. In his embryological study of 
the common salamander, he found that the earliest movements 
are made by the most anterior part of the muscle system of the 
trunk. As the external gills, fore limbs, and hind limbs develop, 
their first movements are performed only as the trunk moves. 
According to Coghill, this method of "expanding" the behavior 
pattern is applicable to all vertebrates, and he has assembled 
some valuable testimony concerning the origin of reflexes in 
man. The iiwiss neurologist, Minko^ski, has similarly shown 
that with human fetuses only a few centimeters in length, the 
reflexes never occur as discrete acts but are always allied with 
other movements. The pattern of movement of the toes as 
a group "individuates 3 * out of the larger pattern of movement 
of the leg, foot, and toes; the characteristic Babinski sign is 
a relatively late development. There can be little question that 
such an entity as the "simple reflex" never occurs in the life 
of the individual, a point recognized by Sherrington, for in 
1905 he had called the simple reflex a "pure abstract concep- 
tion." This apparently had been forgotten by many of the less 
critical, so that now new material is required to show that com- 
plexity of behavior is not derived by the progressive integration 
of more and more originally discrete units. 6 

6 "The Early Development of Behavior in Amblystoma and Man," Archives of Neu- 
rology and Psychiatry, 1929, 21, 989-1009. A less specialized but more complete account of 
the implications of these facts is found in Coghill's Anatomy of Behavior, Cambridge, 1929. 



CHAPTER 4 

THE PHILOSOPHICAL FOUNDATIONS 
Wertheimer's Anti-Synthetic Position 

In less systematic but in far more dramatic manner than 
Kohler, Wertheimer has laid down the broad outlines of Gestalt 
theory. In a semi-popular address, 1 he has characterized the 
basic doctrine of configurationism as follows : There exist 
natural circumstances 2 in which what happens in the total is 
not conditioned by the nature of the parts or their mode of 
combination, but on the contrary, what occurs in any part of 
this whole is determined by the inner structural laws of this 
entirety. As he told his audience, all of Gestalt theory is em- 
braced in this formula neither more, nor less. 

If one asks how this differs from Ehrenfels' position, the 
answer is readily given. If a melody consisting of six tones is 
reproduced and recognized when six utterly different notes are 
played, what has happened ? The six elements are surely pres- 
ent, but in addition thereto one must presuppose a seventh, the 
so-called Gestaltqualitat. It is this seventh datum which makes 
possible the recognition of the melody. So runs Ehrenfels' 
argument. But contemporary configurationism takes a much 
more radical view. According to it, the flesh and blood of a 
single tone is dependent upon its role in the melody. When b 
is a predecessor to c, it is something altogether different from 
b as the tonic. This dynamic position is understood only when 
one appreciates the presence of total-tendencies in a broader 
field of energy. 

The Gestalt conception is not just limited to the provincial 
domain of psychology, but involves a characteristic approach 

1 "Ueber Gestalttheorie," Symposion, I, 1927, 39-60. 

2 Note that this phrase does not claim universal applicability. However, the jhmplica- 
tion at least of most Gestalt language is that the principle possesses cosmic dimensions and 
is true without exception. 

63 



64 GESTALT PSYCHOLOGY 

to all science and all knowledge. Suppose I look at the universe 
from the standpoint of a "collective" theory and ask, "How 
must a world appear within which there can be no science, no 
understanding, no penetration into and comprehension of inner 
relations ?" The answer is very simple. Such a situation re- 
sults whenever we have a multiplicity of purely disparate sec- 
tions. Again, how must the world be constituted, or how 
must the manifold it presents be conceived in order that science 
of the piecemeal variety may be pursued? For this one needs 
nothing more than repeated meaningless connections, which is 
all one has to postulate in order to carry on traditional logic, 
mathematics and science. This is the view most people still 
hold of physics. But actual physical enterprise, rightly consid- 
ered, shows the world in a different light. It is more like a 
Beethoven symphony, wherein we have the possibility of sensing 
the whole from one of its parts, which permits some premonition 
of the structural principle of the total, and in which the funda- 
mental laws are not laws of the parts, but properties of the 
entire event. 

There are two doctrines implicit in the older psychology to 
which the Gestalt position is flatly opposed. These are ( i ) the 
mosaic or "bundle" hypothesis and (2) the associationist thesis. 
The first holds that all complex mental states rest upon a collec- 
tion of juxtaposed simpler contents in much the same way that 
a sand heap consists of a multitude of separate grains of sand. 
The second proposition fundamentally states that if a given 
mental content a has occurred together (in spatio-temporal con- 
tiguity) with another content &, a tendency for a to draw b 
after it is thereby established. Both views are essentially alike, 
the first thesis supplying the bricks and the second the mortar 
of the psychic structure. Should it be said that no serious 
psychologist since James Mill has held these ideas in their raw 
form, Wertheimer would retort that they are implicit in most 
attacks on concrete problems made by contemporary research, 
regardless of the occasional modifications (which do not basic- 
ally alter the situation, for they, too, are handled "analytically") 
introduced by the use of "simultaneous attention," "imageless 
thoughts," "determining tendencies," or "meaning functions." 



THE PHILOSOPHICAL FOUNDATIONS 65 

Now while it is true that behavior occasionally exhibits phe- 
nomena which approach this summative type of response, it is 
far from typical of mental events, being only a limiting case 
and occurring under rare and very special conditions. On the 
contrary, all reality (=the "given") is structured, formed, or 
patterned with highly concrete total attributes and invariably 
obeys the law that total processes define what shall take place 
in subordinate areas. 3 Consequently, fidelity to the facts de- 
mands that one speak of a stimulus * 'constellation' ' and never 
of a stimulus, of an "action-total" and not of a movement. The 
Gestalt concept restores meaning to our conduct by its recog- 
nition of the primacy of restricted "wholes" and the derivative 
and dependent nature of all part-processes. 

JL|vojaIbn Scientific Method 

Following the heavy philosophical penchant of German psy- 
chological literature, Lewin has elaborated in an impressively 
original paper 4 his^belief_tha,t .the. studyjpJE behaviorJs^a^sitig 
fromJthe.Aristotdian stage, into the Galilean phase,, and that at 
present we are witnessing a reconstruction of fundamental 
views akin to that which accompanied the birth of modern 
physics in the seventeenth century. Aristotelianism in psy- 
chology means_among Bother* things, (i) the persistence of 
valuational attitudes so well represented in the outworn distinc- 
tion between normal and pathological phenomena. We still 
speak of "childish" errors, of drill, and "improvement" in the 
same anthropomorphic sense with which early botany consid- 
ered weeds and parasites. (2) The Aristotelian standpoint 
which is usually but not necessarily exhibited in the actual writ- 
ings of the Stagirite tends to identify the individual with the 
accidental. Ordinary observers exhibit this bias to perfection 
when watching some unique bit of child behavior. Do all or 
at least most youngsters do that? If the answer is negative, 




the free-association method! 



tne tree-association metnoai 

4 "Der Uebergang yon der aristotelischen zur galileischen Denkweise in Biologic ttnd 
Psychologic," Erkenntnis, 1930, i, 421466. Dr. Donald Adams has made a faithful 
translation of this paper for the 7. General PsychoL, i93* 5, 141177- 



66 GESTALT PSYCHOLOGY 

the silent assumption is that the action lacks scientific interest. 
(3) The historical Aristotle did restrict the reign of natural 
law to regularly recurrent or oft-recurring events, with the 
result that law and individual were considered opposites, since 
law was limited to those cases in which something common was 
repeated. The general or the abstract had to be discerned in 
the particular and the concrete. This emphasis upon the fre- 
quency feature has led to a one-sided use of mathematics in 
psychology with almost exclusive devotion to statistical methods 
and "unreal" averages, and a consequent neglect of procedures 
for apprehending the full reality of the single situation. 5 ^ It has 
also nourished the belief that quality and quantity are -ttrecon- 
cilably opposed in psychology, a situation similar to that which 
the followers of Galileo had to meet when it was objected : How 
can one comprehend under one rule of motion such utterly dif- 
ferent events as the movement of the stars, the drift of leaves 
with the wind, the descent of the rolling stone, and the flight 
of birds ? One still encounters a subtle version of this attitude 
in the assertion that a law is simply an extreme instance of 
regularity and probability ; whereas, a true Galileian must main- 
tain the universal validity of psychological laws. 6 

The contrast between the two systems of thinking is most 
decisive in the different explanations of movement. In Aris- 
totelian dynamics, the manner and direction of the physical vec- 

5 In another article, Lewin writes, "The statistical method is usually compelled to 
define its groups on the basis, not of purely psychical characteristics, but of more or less 
extrinsic ones (such as number of siblings), so that particular cases having quite different 
or even opposed psychological structures may be included in the same group. . . . But the 
very relation that is decisive for the investigation of dynamics viz., that of the position 
of the actual individual child in the actual concrete total situation is thereby 'abstracted/ 
An inference from the concepts of the average child and of the average situation are ab- 
stractions that have no utility whatever for the investigation of dynamics. The laws of 
falling bodies in physics cannot be discovered by taking the average of actual falling move- 
ments, say of leaves, stones, and other objects, but only by proceeding from so-called 'pure 
cases.' " Essay on "Environmental Forces in Child Behavior and Development" in 
Murchison's Handbook of Child Psychology, 1931, p. 95. For a pleasant satire of sta- 
tistical quantification, ^see Waller, "Scientific Method and Insight," Amer. J. Social, 1934, 
40, 285297. The entire essay is Lewinian in spirit. 

6 Gestaltists object to the usual view of causation as an inevitable chain-like sequence 
of cause and effect, and prefer to predict consequences in terms of general "conditions" 



to real functional wholes akin to melodies. ^Scientific law must state, under the Gestalt 
category, the intrinsic properties of such spatial and temporal wholes." See his article on 
Gestalt in the Encyclopaedia of the Social Sciences. In Lewin's language, a natural law 
is then essentially an action-total. Only such facts can affect each other which have a 
place in the same field. The cause of the events is the relationship between the parts of the 
situation as dynamical facts' and a complete characterization of these dynamical facts would 
be a complete analysis. 



THE PHILOSOPHICAL FOUNDATIONS 67 

tor is completely determined in advance by the nature of the 
object in question ; in modern physics, however, the occurrence 
of physical vectors always rests upon a conjunction of several 
physical factors, especially upon some relation of the object to 
its surroundings (cf. the "total situation" in psychology). If, 
therfras seems generally acknowledged, the dynamics of an 
event is not only dependent upon the object but also primarily 
conditioned by the situation, then it is indeed absurd to assume 
that general laws can be revealed by ruling out the influences of 
the different circumstances as much as possible. 

The reader need hardly be told now that Lewin, in presenting 
the distinction between Aristotelian and Galileian approaches to 
scientific problems, is really interested in magnifying via the 
principle of Pragnans! the fundamental opposition in method- 
ology between traditional nineteenth-century psychology and 
the spirit of configurationism. The conflict is not based solely 
nor even primarily upon questions of fact or content, such as 
the role of kinsesthetic images in the thought process, but upon 
a basic cleavage in the general principles governing experi- 
mental technique and interpretation. A little table may assist 
in picturing the character of some of the more significant con- 
trasts : 

ARISTOTLE GALILEO 

1. The Regular and Uniform 1 ^ ^ kw 1 

The Frequent j I subject to law 

The Individual is a matter of 

chance -^ 

2. Criteria of Natural Law are 1 " niformit y not required 

J frequency 

3. That which is common to his- a manifestation of an "accident," i.e., 
torically and geographically the essence of the only historically 
given instances is occasion; an ex- conditioned 

pression of the na- 
ture of the thing 

An American disciple of Lewjn's, J. F. Brown, 7 has prepared 
a more extended parallel between the two modes of thinking 
under the general caption of "class" versus "field" theory. The 

7 "Freud and the Scientific Method," Phil, of ScL, 1934, L 323-337- 



68 GESTALT PSYCHOLOGY 

central idea of a "field" theory (of which post-Copernican as- 
tronomy and physics and contemporary Gestalt psychology are 
examples) is that the behavior of an object is determined-by the 
field structure or spatio-temporal configuration of the energy 
within the field. The central idea of a "class" theory (pure 
illustrations are difficult to find but differential and abnormal 
psychology are saturated with this viewpoint) is that member- 
ship in a certain class determines the behavior of an object. A 
summary in terms of ten sharply opposed characteristics of the 
two approaches is reproduced below : 

CRITERIA FOR CLASS THEORY CRITERIA FOR FIELD THEORY 

1. The behavior of objects is deter- i. The behavior of objects is deter- 
mined by the class to which they mined by the structure of the field 
belong. of which they are a part. 

2. The force directing behavior 2. The force directing behavior 
shows the properties of an en- shows the properties of a vector, 
telechy. 

3. There is local determination. 3. There is no local determination. 

4. The concepts used in "class'* 4. The concepts used in "field" 
theory are primarily substantial.^ theory are mainly functional. ^ 

5. The method of scientific analysis 5. The method of scientific analysis 
is largely structural. is primarily relatioTial. 

6. The analysis 1s*ln terms of his- 6. The analysis is in terms of ahis- 
torically and geographically con- torical-typical laws, 

ditioned regularities. 

7. The method is primarily em- 7. The method is hypothetical-<fe- 
pirical. duct'we. 

8. The analysis allows dichotomies. 8. The analysis allows no dicho- 

tomies. 

9. Class theory tends to use valua- 9. Field theory insists on non-vcdua,- 
tive concepts. twe concepts. 

10. Class theory attempts to answer a ia Field theory attempts to answer a 
metaphysical "why?" ' scientific "how?" 

In another disquisition 8 on the problem of method, Lewin 
attacks the customary theory of induction, viz., that the uni- 
versal necessity of law rests upon a leap "from many cases to all 
cases," There are two kinds of uniformity which must be 
distinguished: (i) those dependent upon historico-geograph- 
ical connections and (2) those involving conditional-genetic 
factors. The former is illustrated by the sequence of omnibus 

8 "Gesetz tmd Experiment in der Psychologic, " Symposwn, 1927, i, 375421. 



THE PHILOSOPHICAL FOUNDATIONS 69 

lines passing a given corner : if one watches long enough, one 
may notice that a No. 12 car passing in one direction is followed 
within a short time by a No. 14 going in the opposite direction. 
This sequence, however, is not a true causal connection, but the 
result of a time-schedule and the plan of the city streets. The 
latter is represented by the step from "example" to "type," the 
type remaining invariant with reference to spatio-temporal in- 
dices, just as a plant grown at different altitudes. Gold as a 
"type" is not identical with the aggregate of all the gold in the 
world. Similarly, the relation between individual things or 
structures and thing-types is paralleled by the relation between 
individual "happenings" and occurrence-types. 

Lewin holds that a natural law is none other than the descrip- 
tion of a definite conditionals-genetic occurrence-type. To say 
that an event obeys a cerfain law is to say nothing more than 
that it belongs to a specific type of occurrence. For instance, 
the meaning back of the various gas formulae in physics is simply 
that events a and b are necessarily dependent moments of a 
unitary happening. In order to determine whether an event 
belongs in the ^tancg^gtogr^flbi^ or the conditional-genetic 
category, one must have recourse to experiment. That is why 
a single experiment is able to disprove a rule which one has 
tended to accept as a consequence of thousandfold repeated 
experience in daily life. The major difficulty lies in the selec- 
tion of crucial examples, but once found, the penetrating analysis 
of a single experiment is better than the accumulation of statis- 
tical cases. 9 To use terms of a biological origin, the genotypical 
experiment leads to causal explanation in terms of laws, while 
phenotypical observation is limited to a here-and-now descrip- 
tion of characteristics, many of whicETaif fc irrelevant to the 
event. Explaining an event consists in adequately describing 
the underlying genotype and seeing if the phenotype (experience 
or data) may be precisely ordered to it. Laws are descriptions 
of genotypes. 




situation. 



;o GESTALT PSYCHOLOGY 

Wheeler's Organismic Laws 

The one American writer who has offered the most system- 
atic and encyclopaedic exposition of Gestalt psychology has been 
Raymond Wheeler. In a series of four volumes 10 (in which 
he has always been the senior author) published in rapid suc- 
cession between the years 1929 and 1932, he has tried to give 
a comprehensive picture of the development, and to demonstrate 
its teachableness by reducing the widely scattered interpretative 
literature to the confines of a compact textbook. Not only has 
he gone farther than his German predecessors in this highly 
valuable pedagogical contribution, but he has attempted to make 
the Gestalt edifice "hang together" through the consistent ap- 
plication to each of its divisions of eight organismic laws of his 
own formulation. Although not original with him, their ex- 
tensive use as fundamental explanatory principles in psycho- 
logical science was first made through his efforts. At bottom, 
the conceptions are Kohler's, but broadened and freely employed 
with an enthusiasm which shows how readily the disciple's ardor 
may exceed that of his prophet. 

Like Kohler, he upholds the identity of physical and mental 
laws and rejects the ancient problem of materialism versus vital- 
ism as the artificial dichotomy of dualistic ignorance. He traces 
it to the difficulty the Greeks encountered in applying the struc- 
tural and functional points of views with apparently opposed 
conclusions. The one is a question of composition and the other 
a matter of activity. This grand problem of antiquity became 
one of the persistent traditions of European philosophy, a heri- 
tage which might have been avoided had the originators appre- 
ciated that "structure and function are merely aspects of each 
other ; the same thing is envisaged first from a static and then 
from a dynamic point of view. Structure is the form of an 
activity, a something moving, or a something that changes ; func- 
tion is the activity of the form, the moving or changing of the 
thing. Structure and function have opposite but mutually de- 
pendent meanings" (Laws of Human Nature, p. 9). This is 

10 The books referred to are, in order: (i) The Science of Psychology, 1929; (2) Read- 
ings in Psychology, 1930; (3) The Laws of Human Nature, 1932 (Appleton, New York); 
(4) Principles of Mental Development, 1932 all but (3) published by Crowd!, New York. 



THE PHILOSOPHICAL FOUNDATIONS 71 

the larger whole into which both mind and matter fit; J?hyslcs 
and psychjodQgjLJn^ead of being -diagietrica^lj^a^_aTe funda- 
mentally akin in their efforts tojind^J^^ 

This strucfrure-f unctioix antinomy, has exaggerated* the com- 
mon-sense disparity between the physical and the psychical 
Wheeler offers a brief illustration to demonstrate the unreality 
of this logical distinction : "Hold up your little finger and wiggle 
it. You can cut your finger, for the finger is structure. Where 
is the movement ? In the finger ? Cut it open and see. No, the 
movement is not inside. The movement is not in the finger, nor 
is it outside of it ; it is of the finger ; it is something the finger 
does. Similarly, an idea is not in or outside of the brain. It is 
something the brain does. These ABC's of dialectics are use- 
ful, in spite of the fact that they sound rather foolish ; for they 
show how there are just as many phenomena of the so-called 
physical world that cannot be cut and burned as there are of 
the mental. These phenomena are processes, events. Science 
investigates events that have no claim to material versus mental 
substantiality" (op. dt., p. 23). Those who protest that mental 
events are after all unique are exaggerating the qualitative dis- 
tinctions which admittedly exist, but forget that these variations 
do not require a different set of laws. Oil and water are decid- 
edly unique fluids, but both obey hydrodynamic principles. 

The mechanistic 11 bias of modern psychology was a one-sided 
and well-intentioned endeavor to avoid the tangles of an impos- 
sible dualism, but it suffered from the atomistic logic and ele- 
mentarism which dominated Western thought in the nineteenth 
century. Its efforts to explain mental organization by means 
of creative synthesis out of a chaos of unrelated primordial 
parts proved abortive because the relations had to be "forced" 

u Gestalt theorists are in part responsible for the misunderstanding of their anti- 
mechanistic position because of their neglect of the overtones of this phrase. Wheeler ap- 
parently is guilty of the same tactical error here. To the ordinary laboratory scientist, 
"mechanism" implies a belief in immanent law and order in the natural world. However, 
mechanism is only one variety of "naturalism" which configurationists, of course, adopt. 
It is mechanism of the atomistic variety against which the Gestalt writers protest and the 
postulates which they use in their arguments are just as "naturalistic" or "non-spiritual" 
as those held by their opponents. Intrinsically, the realism of Gestalt considers mind, or 
rather mental processes, not as something outside of nature, but as just such natural events 
as any others. They are links in the chains of reactions produced by an organism in an 
environment, and cannot legitimately be isolated from this context. The monism of Gestalt 
is implied in the assertion that one and the same reality are expressed in two different 
conceptual systems the psychological and the physical. The gestalt is the tcrtium com- 
parationis which permits one to equate bodily with mental events since the same configura- 
tion is found in both. 



72 GESTALT PSYCHOLOGY 

upon them from without Actually, there are no disconnected 
pre-existing items : the elements are always related. The fun- 
damental conception which promises to displace the "psychic 
atom" is the "potential," which implicitly has no existence out- 
side of a field of stresses and strains. Unity is no longer to be 
derived from discord but is present from the very first. If this 
be granted, the soul or thought (entities which have previously 
been supposed to create this unity) become superfluous as sep- 
arate explanatory concepts. 

Since Gestalt has taught psychology to speak the language 
of twentieth-century physics, it is an asset to know" the helpful 
features of that idiom. The basic term is the complex organic 
whole. All measurable and observable phenomena in nature are 
differentiations of a pre-existing organic system, whether that 
be a gravitational field or a human being. From this fact, Hei- 
senberg, the brilliant young Leipzig theorist, has drawn his con- 
troversial but appealing "principle of uncertainty," which 
maintains that accuracy of prediction is proportional to the 
knowledge of the totality within which an event is occurring. 
The more isolated the part the less certain the prophecy of its 
behavior. For this reason, the orthodox assumption that psy- 
chological phenomena must ultimately be reduced to physiologi- 
cal terms and these in turn to physico-chemical properties is 
grossly misleading, for such "parts" can never explain the 
whole. All things and events are subject to the primary fact of 
relativity and interdependence. Actually, the universal laws of 
nature are no more peculiarly physical than they are psychologi- 
cal. Man and the cosmos are governed by the same rules and 
not just the same kind of rules. The laws of perception are the 
laws of learning and emotion. These are the laws of energy 
and dynamics, neutral concepts which all the sciences or any 
human enterprise may share. These general principles termed 
by Wheeler his eight organismic laws are as readily applicable 
to problems of conscious behavior as to conventional physical 
and biological affairs, and are defined as follows : 

i. The Law of Field Properties: Any item of reality is in 
its own right an integrated whole that is more than the sum of 



THE PHILOSOPHICAL FOUNDATIONS 73 

its parts. Note that this principle is not just a psychological 
curio, but is claimed to be a universal truth. The earth has a 
density-gradient from its surface to its center which no sep- 
arate portion possesses. Water has properties which are not 
characteristic of either hydrogen or oxygen. A step on a stair- 
way is merely a board in the lumber pile. Metabolic fields of 
force or physiological gradients are likewise attributes of the 
entire animal body and not of its cells and organs as such. 
Meaning is a field property of any perceptual experience; a 
"fractional" meaning is a fiction. 

2. The Law of Derived Properties: Parts derive their 
properties from the whole. A stone, e.g., has weight and yet 
that weight has no independent existence. The weight is merely 
a relation between the body and the gravitational system which 
contains it. As Lashley and other students have shown, the 
function of a certain cerebral region hinges upon what is oc- 
curring throughout the entire neural system. 

3. The Law of Determined Action: The whole conditions 
the activities of its parts. This principle states the Gestalt posi- 
tion on causation or determination, which, as a logical concept 
holds only between a whole and its parts, and never from part 
to part. Exact prediction of an event is possible only when the 
larger framework within which it occurs is known. This is why 
the behavior of electrons has recently been explained by sub- 
atomic researches in terms of statistical probabilities because the 
structure within which the members would conduct themselves 
"lawfully" has been destroyed. Early transplantation of tissue 
from the head end to the tail results in the growth of a tail and 
not a head because the developmental future of each cell is con- 
ditioned by the total organism. The phenomena of polarization 
and neurobiotaxis illustrate the same point. 12 

4. The Law of Indimduation: Parts emerge from wholes 
through a process of differentiation or individuation, as every 
user of a tachistoscope knows. This is the configurationist key 
to the evolution of species and the development of the person. 

12 Perhaps this is the way in which the somewhat figurative expression, "the end deter- 
mines the means" becomes neurologically understandable. The end is the position of 
equilibrium finally attained; the means are the events which will take place until equilibrium 
is attained. 



74 GESTALT PSYCHOLOGY 

Out of a physiological field or ground there arise segregated 
patterns or figures an idea strangely reminiscent of Spencer's 
famous definition of evolution and Lloyd Morgan's emergent 
philosophy, and an interesting support to the notion of corre- 
spondence between organic and inorganic evolution. 13 

5. The Law of Field Genesis: Wholes evolve as wholes. 
This is largely a corollary to the preceding law. The teeth, heart, 
and brain do not change their size or texture in isolation, but 
the embracing organism undergoes an expansion or new struc- 
turation. Sheer physiological growth is "a function of the 
organism-as-a- whole ; it is the progressive internal differentia- 
tion of a single protoplasmic individual." 

6. The Law of Least Action: Energy interchange takes 
place through the shortest spatio-temporal interval. Air cur- 
rents move from regions of high atmospheric pressure to those 
of low; electricity "flows" from high to low potential; nerve 
fibers grow maximally at those points where metabolic rates are 
highest ; and art is dominated by symmetry and harmony. This 
law was first publicly enunciated by Maupertuis in the eigh- 
teenth century and later put into serviceable form by Euler. 

7. The Law of Maximum Work: Where the balance of a 
system is disturbed, all the available energy is employed in re- 
storing equilibrium. The all-or-none character of nerve fibre 
conduction and the "self-preservation" efforts of the organism 
are illustrations of this principle. 

8. The Law of Configuration: One isolated, discrete event 
can never interact with another because things of this character 
are non-existent. The present determines the future in no more 
real sense than the future governs the past ! Temporal acts are 
unified in the same way that spatial patterns are a conclusion 
implicit in the phi-phenomenon. 

These eight fundamental organismic laws Wheeler applies to 
all the chapter headings of the standard textbooks. They are 
assumed to be adequate to account for all psychological phe- 

33 In evolution, the accidental emergence of forms, In the Darwinian sense of selection, 
Is supplanted by the creative dynamic pressure of an organic system. Kohler has expressed 
this idea in "Gestaltprobleme und Anfange einer Gestalttheorie," Jahresbcricht it. d, gcs. 
PhysioL -it. exp. Pharmakologie } 1925, 3, 512539. 



THE PHILOSOPHICAL FOUNDATIONS 



75 



nomena and consequently other explanatory principles such as 
attention, association, experience, and memory-traces are re- 
jected as tautologous. To explain learning in terms of exercise, 
or by the use or disuse of neural arcs, is gratuitous. "Drop a 
ball a million times and it will not fall more easily the last time 
than the first. . . . An electric current will travel just as well 
through a switch the first time as it will the thousandth. Repe- 
tition makes no difference to the switch nor to the current." No 
one benefits by experience unless he has enough insight into that 
experience to profit from it. A child or an imbecile could read, 
or better, gaze at, the text of Newton's "Principia" for a year 
at a time and not gain any more than a few simple percepts. The 
experience hypothesis is unable to explain how the first learning 
process in the life of the learner occurs when prior to it there has 
been no experience. If the initial act cannot be explained on 
this basis, neither can the tenth nor the thousandth. Instead, 
learning is a growth in insight derived from the prpjpjeriy^timed 
environmental stimulation of a maturing organism. If the 
inner maturation has not gone far enough to enable the learner 
to master the situation, learning, of course, fails to take place. 
That is why concentrated learning is so inferior to distributed 
learning the "rest" or recess periods are as important as the 
practice sessions ! 

Wheeler's theory of recall is interesting because of its re- 
semblance to Hollingworth's account of learning as a "reduction 
of cues," the fundamental difference lying in the former's ap- 
peal to * 'closure" or Pragnanz and the latter 's associationist ver- 
sion of redintegration. The configurationist explanation is as 
follows : The same organization of potentials in the nervous 
system and the same stimulus pattern that were present in the 
original observation function in recall, "except that (i) matura- 
tion has taken place meanwhile, and (2) there is only a partial 
duplication of the original stimulus pattern. Suppose you climb 
a mountain on your vacation. When you return you are able 
to recall the scenery. You are doing nothing more than to see 
the scenery with part of the stimulus pattern missing, the 
scenery itself. When you were actually looking at the moun- 
tain you were responding to many more stimuli than the 



76 GESTALT PSYCHOLOGY 

mountain itself. There were your companions, your camping 
equipment and your conversations with each other, even with 
yourself. You return home, bringing these stimuli with you, 
including language. Circumstances, then, construct a stimulus 
pattern partly duplicating the original ; the response partly dupli- 
cates the original, and we designate it a recall. If the repeated 
stimulus-pattern were as complete as the original, the perception 
would have been equally as complete. You would then be re- 
observing the mountain, for, by assumption, it would be in front 
of you again. Errors in the recall are to be explained in terms 
of maturation on the one hand, and removal from stimulus- 
patterns on the other " (Laws of Human Nature, pp. 168-169), 

One feature of Wheeler's personal version of the implica- 
tions of configurationism, which should appeal to those who be- 
lieve that the coming century belongs to the social sciences, is 
his insistence that an individual (or any "scientific object") can- 
not be defined adequately in terms of himself alone. The logical 
starting-point for all psychology should be social psychology, 
for it is out of the social matrix that individual behavior 
emerges. The most complex situation is in a certain sense also 
the most primitive. The human being must first be thought of 
as a "member-of-a-group," and only later under narrowly re- 
stricted conditions can he be studied as an isolated biological 
individual. The group has been shown by anthropologists to be 
the true evolutionary unit, and no theory of personality or the 
self (or of any other mental feature) can be correct which dis- 
regards the formative influence of powerful social forces. "A 
person could no more acquire a personality In the absence of the 
social group than a man could become a soldier without being 
in an army." Human nature itself is a pattern or group phe- 
nomenon, as most of the early sociologists insisted it was. 

Concerning the problem of intelligent behavior, many non- 
Gestalt critics have found themselves questioning the dogma of 
the "constancy of the I. Q." Without denying the figures which 
seem to support this view, Wheeler maintains that this assertion 
violates a basic scientific principle ; viz., that any event is a prod- 
uct of its conditions. The score which a person makes on a men- 
tal test is partly, but not exclusively, conditioned by his inheri- 



THE PHILOSOPHICAL FOUNDATIONS 77 

tance ; social status, physical constitution, and emotional block- 
ing are also significant factors. When the intelligence quotient 
appears to remain relatively stable, it is because the environ- 
ment and these other influences have undergone no marked 
change. Capacities being potentials of some kind are as depend- 
ent upon the environment as upon the race. Within limits, then, 
the intelligence of the individual is under the control of the 
social group certainly a more hopeful view to an enlightened 
educational system than the notion that it is exclusively a result 
of native endowment. The efforts of nature-nurture theorists 
to measure the percentage contributions from each source are 
absurd because there would be no achievement-product at all 
without the constant use and simultaneous cooperation of both. 



CHAPTER 5 
VARIETIES OF GESTALT THEORY 

In the very nature of things, it is impossible for a universal 
concept such as the relation of the whole to its parts to receive 
the same systematic interpretation at the hands of different 
thinkers, despite their common acceptance of certain major pos- 
tulates. The psychological circles of German-speaking coun- 
tries to-day are virtually unanimous in affirming the merits of 
the "totality" mode of thinking, but they differ violently in the 
consequences to be .drawn therefrom. The leading Gestalt 
school (by which we may mean nothing more than the "best- 
known") is the Wertheimer-Kohler-Koffka group, for brevity's 
sake often identified as the Berlin school from the name of the 
laboratory which has given rise to most of its positive contribu- 
tions. It is curious, too, that this group has preserved the 
greatest degree of internal unity and presented a closed front 
to the "opposition/' in spite of a rather large number of adher- 
ents. Since it is almost proverbial that there are as many psy- 
chologies as psychologists, some observers have commented on 
the distinction to be made between the Gestalt theory and the 
Gestalt movement the former being an intellectual construc- 
tion which most scientists can envision with dispassionate de- 
tachment, while the latter is a band of aggressive personalities 
engaged in propagating the faith. Although this book will be 
mainly devoted to the narrower version of configurationism, 
the demands of perspective make it necessary to examine a few 
alternative interpretations. 

The picture is complicated by the presence of a number of 
contenders, each asserting the honor of priority in announcing 
the main outlines of the new theory. To state the matter vul- 
garly, many individuals appear to be dominated by the "I-said- 
it-myself -first" complex, which may lend animation to the en- 

78 



VARIETIES OF GESTALT THEORY 79 

suing polemics, but certainly adds confusion to both the be- 
fuddled reader and the distressed historian who must weigh the 
rival claims. Just as Binet complained with some justice that 
the famous Wurzburg school of imageless thought could with 
equal fairness have been termed the Paris school (one might 
also add a New York school in the person of Woodworth), so 
configurationism might but for the accidents of emphasis and 
opportunity have been associated with the schools of Leip- 
zig, Vienna, 1 Marburg, 2 Rostock, 3 and Hamburg, 4 to mention 
only a few of the more prominent varieties. Each offers an at- 
tractive and stimulating variation from the main theme, and it 
would be profitable to inspect all of them with some care, but to 
avoid complicating the story unduly, only the point of view of 
a single competing group will be presented. 

Krueger and the Leipzig School 

The major novelty in Felix Krueger's position 5 (since 1917 
Wundt's successor at Leipzig) is the strategic role in mental life 
allotted to feeling ; but as with every other system-builder, one 
must look to the entire framework of ideas before this division 
can be appreciated. The heavy philosophical and valuational un- 
dertone in Kraeger's schema has repelled many of the "tough- 
minded/' yet basically it involves little more than an insistence 
that experimental and non-experimental methods must operate 
conjointly to advance the realm of psychological discovery, a 
view championed by many authorities, particularly in the fields 
of applied and abnormal psychology. He holds that the associ- 

1 Buhler, Brunswik, and the child psychologists centering about them. The lack of 
rigorous experimental productions from the Vienna laboratory in the last decade is prob- 
ably accountable for its failure to shine more prominently in the debate. 

2 Jaensch and his numerous disciples. His voluminous writings defy digestion, but 
the most representative of the "wholeness" approach is "Ueber den Aufbau der Wahrneh- 
mungsxvelt und ihre Struktur in Jugendalter," Zsch. f. Psycho!., 1923, gg. Spranger has 
acknowledged its influence over him. Pre-occupation with his extraordinary eidetic dis- 
coveries has probably led to a neglect of his configurationist views by others. 

3 Katz, of course. A provincial university, outside the main current of international 
traffic in scholars, explains this particular case of obscurity. 

4 William Stern, Werner, Muchow, and others. The influence of Stern's peculiar 
version of "personalism" has been limited to his own immediate circle. Werner's clever 
experimental program is an extremely promising development. A little late in arriving, 
his ideas have not been publicized by foreign lecture-tours to the extent that the Berlin 
groups have. The latter nave been conspicuously successful in retailing^ their wares, and 
their success has led to much domestic criticism, the motivation of which is not always 
certain. 

r > "Ueber psychische Ganzheit," Nette Psychologische Studien, Vol. i, Beck, Munchen, 
1926, 1-121,, 



So GESTALT PSYCHOLOGY 

ationist systems of Ziehen and G. E. Miiller are in process of 
dissolution, despite the efforts to preserve the mechanistic play 
of ideas or "residues" by means of attention in its many guises, 
apperception, determining tendencies, etc. Instead, there has 
developed a growing recognition of the social and genetic condi- 
tioning of all mental events. Like Wertheimer, he deprecates, 
the extreme analytic approach to scientific problems, but pre- 
serves a better balance and saves himself from misunderstand- 
ing by citing with approval one of Goethe's utterances, "Analy- 
sis and synthesis are both as necessary to the thinking spirit as. 
inspiration and expiration to the organism." 

To a certain extent, Krueger disclaims any marked original- 
ity for his view and points in a sober nationalistic vein to its 
kinship with the ideas shared by a long line of distinguished 
'German teachers. The dynamics of Leibniz (his monads cor- 
respond to the "segregated wholes" of contemporary theory) 
and his conflict with the dogmatic rationalism of the Cartesians 
and the sensualism of the British empiricists began a tradition 
of broad inclusiveness which was steadily developed by Herder, 
Tetens, Kant (in his emphasis on the creative activity of the 
subject in perception) and Hegel (with his familiar triad and 
stress upon "becoming"). The fullest sweep of the totality 
problem is seen in the efforts of such metaphysicians to systema- 
tize into comprehensive unities all cosmic processes and exist- 
ences, the most plausible means of effecting this integration 
being the evolutionary principle. 

It must not be supposed that this standpoint is vague or ir- 
relevant to psychological research, although these terms prob- 
ably describe the average American psychologist's reaction to 
such a position. One must regret that to the strict experimen- 
talist, philosophy has come to mean futile verbal disputation of 
uninteresting private opinions tinged with irresponsible emo- 
tions, when, as a matter of fact, his own attitude reflects but one 
among many possible intellectual adjustments. For the repre- 
sentative German scientist, however, the Weltanschauung pre- 
scribes his methodology and serves as a heuristic device in all in- 
vestigations. To be sure, it has the same effect in the work of 



VARIETIES OF GESTALT THEORY gi 

others, the only difference being that they usually seem to be un- 
aware of it ! 

It appears that Krueger was led to 'his present theories on the 
basis of his older acoustic studies. These observations clearly 
Indicated that consonance and dissonance of all types were 
Komplexqualit'dten ; i.e., they did not occur with single tones but 
were inevitably aspects of a chord. Every one can appreciate 
the simple truth of this contention, itself implicit in Ehrenfels 5 
earlier illustrations. 

Krueger acknowledges his indebtedness to his teacher, Hans 
Cornelius, for the suggestion that the feelings are Gestalt quali- 
ties in Ehrenfels' sense, although the latter rejected this particu- 
lar identification. On the basis of the best introspection avail- 
able, Krueger holds that the feelings are functionally to be 
reckoned among the complex-qualities as the specific total col- 
oration of every experienced whole and as the mode in which the 
unitariness of psychic processes is directly and originally made 
conscious (Krueger, p. 26). 

Physiological and biological evidence is freely appealed to. 
The effect of a local stimulus is not a local response; on the con- 
trary, an irradiation throughout the total organism ensues. 
Neurologists appreciate the general influence of localized trau- 
mata and surgeons realize that their work not only affects the 
regional organ, but the whole personality. It is a commonplace 
among orthopedists that their patients are not only handicapped 
directly by their deformity or distortion of movement, but that 
a general internal deviation from the norm usually appears. 
Without going very far into either physiology or pathology, one 
can discern the operation of holistic principles in organic as well 
as psychic occasions. Structures and total performances appear 
in both regions of living activity, but an experienced whole can 
occur only within the mental world; it can be known more 
completely than all other whole forms because we understand 
it from within. 

Although it is possible to study the peripheral data of psy- 
chology with little or no sociogenetic postulates, the atomism 
which this involves collapses utterly when brought to bear upon 



82 GESTALT PSYCHOLOGY 

the cultural behaviour of man, especially as studied by the nor- 
mative disciplines. Dilthey was one of the first to see this 
clearly and he made a fine pioneering endeavor to overcome this 
limitation by a liberal and flexible use of his favorite expression 
"structure" in connection with the more complex aspects of con- 
duct. However, it remained for Driesch to be the first to erect 
a thorough-going system of philosophy upon the idea of the 
whole. His classic demonstration of organic regulation in the 
maintenance of form in the embryo is a beautiful experiment, 
but the assertion that every genuine totality in the world is 
"produced" by spiritual factors is a pure metaphysical hunch. 
Krueger pleads for a sharper distinction between real and ideal 
wholes to save us from such errors. Any part of a present 
experience approaches the character of a feeling the more of 
the perception it comprises and the less it is "membered" or dis- 
tinguished from the rest. Feelings are qualities of total im- 
mediately-given wholes each feeling being a diffused whole, 
preserved at every level of growth, and either entirely or largely 
without members. The course of both mental and cultural 
development moves steadily from a primitive totality to a less 
primitive, that is, more "structural" whole. Krueger finds the 
Berlin school of configurationism grossly neglectful of the im- 
portance of the feelings, a charge which had considerable justi- 
fication before 1925, but which the researches of Lewin and his 
pupils since then amply refute. Kohler's brain-theory and his 
interpretation of structure-functions are anathema to Krueger 
who sees in them a revival of a meaningless physicalism. His 
anti-rationalist bias is clearly presented in his treatment of 
Spearman's account of spatial shapes, the main weakness of 
which he finds to be an overintellectualistic brand of "relation- 
ism." The "g" factor problem, however, upon which the two 
men labored in conjunction some years ago 6 is held to support 
his conception of a stable and enduring total structure underly- 
ing personality. 

One sees, then, how insistent Krueger is that the unitary 
character of all experience is earliest^ most strongly and most 

fi Krueger and Spearman, Zsch. f. Psych&L, 1906, 



VARIETIES OF GESTALT THEORY 83 

richly expressed in the domain of feeling. However, feelings, 
with which the emotions are grouped, are not usually configura- 
tions, although timbre and rhythm are such, and Krueger goes 
to great length in distinguishing between Gestalt (pattern) and 
Ganzheit (totality), the former being a species of the latter 
genus and marked by a preservation of the "membership traits" 
of the component parts. Gestalt phenomena are present only 
where experienced wholes occur ; but this proposition cannot be 
reversed. 7 

The climax of Krueger's structuralism (the meaning of 
which is worlds removed from the connotation which Titchener 
gave to the term in America) is found in his interpretation of 
values. These most influential forms of psychic structuration 
rest upon instinctive bases and are the dominant forces in men- 
tal life. They constitute the core of personality and character 
and are the determinants of all typical and essential individual 
differences. Even the remote but fundamental field of episte- 
mology profits from the suggestion that meaning or significance 
bears the traits of specific wholes, particularly the features of 
''belonging together/' 

In this system it is clear that the notion of feeling becomes 
much broader than the simple opposition of pleasantness-un- 
pleasantness. 8 On the other hand, many familiar doctrines are 
retained even though submitted to a new interpretation; e.g., 
"an emotional complex loses in the intensity and plasticity of its 
emotional character to the degree that it becomes analyzed, so 
that its parts become relatively separated, or that the partial 
moments in it come out clearly as such" (Krueger, p. 62). This, 
of course, is simply a rephrasing of the elementary truth that an 
emotional state tends to be dissipated by attention to it as such, 
One sees, too, how some of the difficulties resulting from the 
James-Lange theory can be understood in a new light. The now 

7 "Der Strukturbegriff in der Psychologic," Bcricht ubcr den VIII, Kong. f. exp. 
Psychologic, Leipzig, 1923, 31-56. Jena, Fischer, 1924. 

8 This aspect of his doctrine has been most conveniently presented by Krueger in his 
tribution to the Wittenberg Symposium on the Feelings and Emotions, Clark University 



Press, 



3S, 1928, 58-88. 

A similar position had been outlined earlier by Claparede in an article published in 
the Archives de Psychologic, 1908, 7, p. 195. The emotion is nothing^ other than the ^con- 
sciousness of a "form" of multiple organic impressions, i.e., the emotion is the conscious- 
ness of a global attitude of the organism. This confused, primitive, and general perception 
of the whole, Claparede called "syncretic perception." 



84 GESTALT PSYCHOLOGY 

generally admitted evidence indicates that the organic and vis- 
ceral changes accompanying many distinct emotions are sur- 
prisingly alike if not identical (cf. for instance, the glandular 
accompaniments of fear and rage). Apparently the same inter- 
nal elements have a different function depending upon the nature 
of the total mental state or even the total objective situation to 
which they refer or within which they operate. 

Since there is a grave danger that even repetition with vari- 
ation will not convey the precise outlines of Krueger's unfa- 
miliar theory, it may be wise to let him state his position in his 
own language. "Never are the differentiable parts or sides of 
real experience as isolated from one another as the parts of physi- 
cal substance, i.e., its molecules or its atoms. All things which 
we can differentiate there, by comparison, always grip into one 
another and around one another in the greatest elaboration. And 
every time it is, without exception, imbedded within a total- 
whole, by which it is penetrated and more or less completely 
enclosed. Feelings are the qualities of experience of this total- 
whole. . . . The total- whole of experience always has a specific, 
immediately observable quality which changes in a particu- 
lar, continuous way. Such qualities of the total-whole are 
the different kinds of pleasantness and unpleasantness, excite- 
ment, tension, relaxation; and many other manifold tintings, 
shadings and forms of flight of total experience cannot be lim- 
ited by number, and, until some future time, cannot be com- 
pletely classified" (Krueger, pp. 67-69). 

Some persons, repelled by the disorder which appears to re- 
sult from the abandonment of the traditional categories, will 
wonder if this view does not eliminate the standard opposition 
between the emotional and the non-emotional. Krueger would 
willingly admit this, but believes a closer approximation to 
reality results therefrom. "It is certainly a fact," he says, "that 
feelings (e.g., of excitement without an object, of excitement 
resembling fury, of purely moody excitement) , always pass over 
into qualities of more circumscribed and, primarily, of less or- 
ganized par ^-complexes, e.g., into the consciousness of that 
about which I become excited, of that for which I hope, of that 
which I seek or of which I am afraid; and conversely, it is a 



VARIETIES OF GESTALT THEORY 85 

fact that the one set of events is, moreover, qualitatively related 
to the other. The conception of the feeling-/^ is necessary to 
designate those phenomenological similarities and transforma- 
tions. . . . Everywhere isolated sensations, perceptions, rela- 
tions, also memories, clear ideas, decided volitions in brief, 
all experience-organization split off only after some time from 
the diffuse tendencies of emotion, and secondly, they always re- 
main functionally dominated by them" (Krueger, pp. 70-72). 

Sander 

This genetic primacy of feeling over perception and thought 
is buttressed by numerous investigations from the Leipzig lab- 
oratory, especially those coming from the hands of Volkelt and 
Sander (now at Giessen). The former has shown that the 




B F c 

Figure 7. Sander's Parallelogram Illusion 
Which of the two diagonals is the longer? Verify your impression by measurement. 

younger the child the more overwhelming are the evidences of 
perception according to larger wholes, indicating that the analyt- 
ical trend of adult observation is a derived tendency. Sander 9 
has also constructed his new parallelogram illusion (see Figure 
7) on this general principle and finds that the coefficient of illu- 
sory effect drops with increasing age. 

Here two objectively equal lengths, AF and FD, are seen as 
unequal because they belong to two parallelograms of different 
size. The "apparent" size of an important member is domi- 

F. Sander, "Experimentelle Ergebnisse der Gestaltpsychologie," X, Kong. /. exp. 
Psycho!., Jena, 1928. 



86 GESTALT PSYCHOLOGY 

nated by the "actual" size of the whole and consequently shifts 
in that direction. If, however, the equilateral triangle AFD is 
"lifted" out of the complex, the two sides tend toward equiva- 
lence. 10 

Since the Leipzig school is decidedly more historically minded 
than the Berlin group, one need not be surprised to see them 
searching the records of older studies for support of their brand 
of configurationism. The have examined the , handiwork 
of the naive painters of the German Renaissance and find 
the law of Pragnanz operative in the unusual regularity of 
the sketched thunderbolt among one of these "primitives," 
thus _r~l_n_n n - While the exaggeration here is ex- 
treme, we know that the customary representation of lightning 
shows nothing of the highly irregular gnarled branching which 
the photograph reveals. This same drift toward symmetry 
appears in innumerable cases. Incited by Goethe's interest in 
entoptic pictures (easily evoked by rubbing one's eyes in the 
dark), Purkinje sketched some of them and found them un- 
usually regular ; in fact, they approached snowflake patterns in 
appearance. One is tempted to apply Kohler's concept of physi- 
cal configurations to the actual falling snow with its hexagonal 
flakes, but that would lead us too far afield from our own al- 
luring quest. 

Less spectacular but still impressive laboratory studies con- 
firm this natural struggle for proportion. Kirschmann observed 
that when polygons were arranged to stimulate the peripheral 
areas of the retina, they were seen as circles, and it is well known 
that very brief tachistoscopic exposures of complicated plane 
figures result in the reproduction of the latter as circular objects. 
Figures apprehended in twilight or at a great distance exhibit a 

10 Note: The isolation is facilitated by mentally erecting a perpendicular at F. 

In general, a change in one part or aspect of a configuration can be made only^at the 
cost of or change in another. Thus two objects of unlike size but equal weights are judged 
to differ in weight, the smaller being the heavier (size-weight illusion). Here what the 
larger gains in size it loses in weight; what the smaller loses in size it gains in weight, phe- 
nomenally. Shortening the temporal interval between stimuli (visual, auditory or tactual) 
results in shortening the experienced spatial interval; brighter objects are judged to weigh 
lessjthan darker (DeCamp) ; objects that leave the ground at a higher speed are judged to 
be lighter than those leaving at a slower rate. As Kelson says, "In all these cases the 
experienced change (usually called the 'illusion') carries all the ear-marks of reality and 
can be explained only by the assumption of a unitary underlying process in which the 
various aspects exert real and mutual effects upon each other." See his useful summariza- 
tion entitled, "The Fundamental Propositions of Gestalt Psychology," Psy. Rev., 19 33. 
40 t 13-32. 



VARIETIES OF GESTALT THEORY 87 

kindred uniformity; e.g., in measuring visual acuity, one can 
take the letter K (which lacks all rounded features) and if held 
well beyond the point of distinct vision, it will be traced as a 
small ball or dot, provided, of course, the subject is kept in 
ignorance of the real nature of the stimulus. 

The Hunger Studies o David Katz 

In a few well-chosen studies on the appetite of barnyard 
fowl, Katz neatly demonstrated the dependence of an animal's 
"instinctive" and motor behavior upon systematic or configur- 
ational influences. He was able to show that the functioning 
of the hunger drive does not rest on a purely physiological basis, 
but that it is surprisingly independent of the physical state of 
the organism and depends to a high degree on outer circum- 
stances. 

If one places a hen before a pile of 100 grams of wheat, she 
will eat on the average 50 grams, and leave 50 grams. The hen 
in the same hunger condition eats from 33 to 54 grams more 
when confronted with a 200 gram pile. In general, the amount 
of food consumed increased with the amount of food presented, 
a fact applicable to all hens and all grains. 

In another arrangement, a hen was allowed to remain before 
a food pile until fully satisfied. The remaining food was then 
removed with a brush and immediately replaced. When this 
was done the hen invariably began to eat again. When it ceased 
to eat the experiment was repeated once again with a similar 
result. There were hens with which the operation could be suc- 
cessfully repeated eight or more times. Many hens ate as much 
as 67% more. Katz triumphantly remarks that we have here a 
method in applied animal psychology which might profitably be 
used for fattening hens of other poultry ! 

The size of the single grains appear to exert an astonishing 
influence upon the quantity of food eaten. In a similar state of 
hunger, a hen ate two or three times as much in weight from a 
pile of whole rice as from one of cracked rice. The grains of the 
latter were only one-quarter as large as the former. Katz ex- 
plains the difference in terms of the painful vibrations in the 



88 GESTALT PSYCHOLOGY 

lien's beak caused by excessive pecking, a view supported by the 
fact that hens peck only about 40 grams of food from a hard 
wooden surface but twice as much from soft felt. When con- 
fronted with a choice, the animals always preferred the soft 
surface. 

Some clever observations on the social behavior of hens 
are relevant to this discussion. A hen No. I is allowed to peck 
until motionless. Then a second hungry hen is brought in be- 
fore the same heap. Instantly she begins to peck. Then the 
behavior of the first hen differs according- to whether she has 
previously tyrannized over the other or has been subdued by her. 
In the former case, she tries to hinder the second hen from 
pecking the food, and if she fails immediately begins to peck 
again herself ; in the latter case, she immediately resumes eating. 
Under the social influence of the second hen, hen No. i eats 
60% or more food than when alone. This amount is still fur- 
ther increased if three hens instead of one are brought in, which 
shows that up to a certain limit the hunger behavior of the 
first hen is influenced by the number of hens brought in. 

The reverse process was also followed. Three hens are put 
before a large hill of grain and satiated. If now a fourth hen is 
brought into this circle, she immediately begins to peck, and the 
other three remain almost motionless; they peck very little. 
They form, to a certain extent, a unity of behavior, no longer 
susceptible to the influence of the isolated fourth hen, who in 
turn is not noticeably affected by the "trinity." 

The human analogues to this sort of behavior are numerous 
and convincing. Most people realize that they eat more when 
the table is loaded with an abundance of food, and every one 
recognizes the potency of social influence. The bachelor who 
goes to the hotel to dine because he wants to eat with other people 
gives^ evidence of the fact that in isolation his appetite is in- 
sufficient. Perhaps the added weight which most married 
couples exhibit is not solely attributable to maturity or the effect 
of wedded contentment social facilitation may be the real 
cause. Katz suggests that the best way of improving a child's 
appetite is to let him eat with the other children. He claims that 
in earlier times lack of appetite was seldom heard of and that it 



VARIETIES OF GESTALT THEORY 89 

has resulted largely from the reduction in the number of chil- 
dren in the family. 11 

Different Attempts to Solve the Part- Whole Problem 

Both the strength and weakness of the Leipzig school have 
resided in its pronounced eclectic leanings which have kept it in 
close touch with other research centers without involving the 
abandonment of its characteristic viewpoint. One of its younger 
members, Burkhardt/ 2 has examined the major German efforts 
to deal with the problem of totality, which for purposes of sum- 
mary, may well be reproduced here. He finds that all represent 
attempts to solve the two following basic questions : 

1. How does psychic "wholeness" arise? 

2. In what does this totality exist ? 

Three different answers have been given to the first problem : 

1. The production theory 

2. The physical theory 

3. The psychical theory 

The first view is historically the oldest and was shared by 
such diverse individuals as the founders of the Graz school, 
Schumann and G. E. Miiller. Meinong and most of his fol-l 
lowers held that the intellect or "intelligence" was responsible 1 
for the integrations observed in experience; Schumann simplyf 
adopted the general view that the gap between summation and 
"wholeness" was bridged by some synthetic function, and Wita- 
sek and Miiller boldly identified this with attention. However, 
Miiller's admission 13 of the existence of "involuntary collective 
apprehension" is tantamount to a confession of the failure of 
this principle to operate satisfactorily as a universal explanation. 

n The material for this section is largely derived from "The Vibratory Sense and other 
Lectures" in the University of Maine Studies, Second Series, No. 14, 1930, Orono. 

Yoshioka later discovered that both wild grey rats and tame albino rats, when pre- 
sented with an equal number of "large" and "small" sunflower seeds, tended in any given 
time to eat a greater number of large seeds than of small ones. See his "Size Preference 
of Wild and Albino Rats," /. Genet. PsychoL, 1930, 37, 159-162, 427-430. 

^Zum Problem der Gansheit, "Ein Beitrag zur Theorie^des Psychischen" (disserta- 
tion), Leipzig, 1925. The issue at stake may well vie in significance with, the historic 
problem of the three bodies. 

13 Komplextheorie und Gestalttheorie, Gottingen, 1923, p. 101. 



go GESTALT PSYCHOLOGY 

The second type of explanation is best represented by 
Wertheimer's pupils, Kohler, and the Berlin school in general. 
It is a form of psychophysical monism with its weakest point in 
the alleged physical parallelism. An appreciation of the quali- 
tative difference between causally effective units like extended 
physico-chemical wholes and the intensive psychic totals makes 
difficult the acceptance of this view. Driesch, 14 the great vitalist, 
considers Kohler an opponent because of the latter's rejection of 
the autonomy of the living, i.e., the Berlin group is committed 
to pan-vitalism and not vitalism as a distinctly limited explana- 
tory concept. Kohler's physical gestalten prove only the unity 
of his systems and not their wholeness; how frightened, e.g., 
we would be if a broken Leyden jar were suddenly to transform 
itself into two proportionately exact smaller jars ! However, 
no one has ever denied that everything is dynamically joined to 
everything else in inanimate nature and that unities of effect 
thus obtain. A Mid- Victorian physicist had already main- 
tained, "If I strike upon the table Sirius trembles." But it is not 
true as the Berliners allege that a change in a system composed 
of n parts changes its essence so that a total of n I parts is 
utterly different from it; e.g. (ironically!), a dog still remains 
a dog even though I cut off some of his parts, such as his hairs. 

The third view, represented by Driesch and Krueger, is truly 
dualistic and a psychology with a soul There is a qualitative 
peculiarity about mental totalities which is not found elsewhere 
in nature a fact which in itself makes of psychology an auton- 
omous research discipline. This original uniqueness of psy- 
chic wholes constitutes the starting point for all investigations 
concerning them. 

Turning to the second of the problems with which we began 
this section, we find that here, too, there have been three cor- 
respondingly different answers : 

1. The theory of relations 

2. The theory of structure-functions 

3. The theory of qualities 

TT -! "? hysiscb Gestalten j und Organismen," Annalcn dcr Philos 102 < < t-ii- and 
'Kntisches zur Ganzheitslehre," same journal and volume, pages f 381-304?' 



VARIETIES OF GESTALT THEORY 



91 



Ehrenf els, Buhler, and Spearman are all inclined to adopt the 
first position which would make logical constructs of all com- 
plexes and gestalten. The inadequacy of this approach is most 
conspicuously revealed in feeling and will in which relations as 
ordinarily understood play almost no part. Kohler's account in 
terms of structure-function suffers from its objectivistic 
orientation and cannot satisfactorily explain in naturalistic 
terms why the same cliff side can be successively apprehended 
as a chaotic complex, as an animal figure, and as a bit of scenery. 
The third view is the right one in Burkhardt's opinion, and con- 
sists in the realization that wholes are qualitatively irreducibles ; 
they are given existents with which we must work. 



PART III EMPIRICAL 



CHAPTER 6 
PHENOMENA OF VISUAL PERCEPTION 

The limitations of the human mind make necessary the classi- 
ficatory rubrics which segregate the divisions of every science. 
Despite the obvious logical service which they render, particu- 
larly to the novice in the field, there is always the grave and per- 
sistent danger of reification which accompanies their use. The 
topics of heat and light illustrate this plainly in physics, while 
it is but recently in psychology that we have begun to question 
the substantive character of memory, will, and other traditional 
chapter headings. 1 With the breakdown of the standard nine- 
teenth-century forms of pigeon-holing data, it has become no- 
torious that the same item can appear in the most unexpected 
of places, depending solely upon the whim of the writer. Many 
have assumed that this question of "labels" is a matter of indif- 
ference, but any one with a minimum of pedagogical insight 
knows this to be false. 

This problem of the categories is especially pertinent to the 
present discussion because configurationism is a bold attempt to 
sweep away these lines of cleavage within the subject-matter 
and to substitute a more unified picture by means of such basic 
concepts as figure-ground, whole-part, pattern, insight, and ten- 
sion, which are designed to offer a comprehensive interpretation 
of all the organism's activities. Conceivably one could center 
the experimental data around each of these themes, but since 
they are fundamentally interrelated, it would be difficult to do 
justice to any one of them without undue overlapping of the 
others which, while psychologically inevitable, is logically in- 

1 Cf. on this point, the second chapter of Woodworth's revised Psychology (New York, 
1934), in which he suggests the use of verbal forms memorizing, "willing, etc. instead 
of nouns. 

93 



94 GESTALT PSYCHOLOGY 

convenient. Although the same objection arises with any alter- 
native treatment, it appears to be less pronounced with the re- 
vised functional groupings of recent texts. This method of 
presentation lias been adopted here, not only because it mini- 
mizes the air of "queerness" which might otherwise result, but 
because it permits a readier appraisal of the genuine contribu- 
tions of Gestalt to the orthodox concerns of experimental psy- 
chology. 

Sensation and Perception: Fundamental Laws 

For the Gestalt psychologist the subtle distinction between a 
"pure" sensory experience and a percept is dissolved by the 
recognition that they both involve perceptual processes which 
differ only in complexity, just as a chemical element and a com- 
pound are both "matter." A so-called sensation of brightness, 
e.g., is always attached to something : we do not see redness dis- 
embodied and independent but only a red object. The attempt 
of the Titchenerians to reduce all mental states to little psychic 
units, devoid of meaning and merely "existent," is said to in- 
volve the grossest sort of "stimulus error" because of its im- 
possibility. For one who wishes to think clearly in this tangled 
field it is distressing to encounter at the start the vexed problem 
of meaning, so heavily tarred with the metaphysical brush. To 
wave the question aside may seem like a gesture of contempt 
for anything even faintly tinctured with philosophy, a procedure 
which is too often only a veiled retreat from an uncomfortable 
problem. Perhaps it is best for psychologists to err on the side 
of positivism, for while it may lead to wrong views at least it 
rarely encourages vague ones ! 

The classic account of configurationism's stand on the mat- 
ter is undoubtedly Wertheimer's brilliant essay 2 in on-e of the 
early volumes of the Forschung. With magnificent directness 
he asks us to consider the following situation : 

I look out of my window and see a house, some trees, the sky. On 
theoretical grounds I could say: There are 327 brightnesses and color- 

2 "Untersucbungen zur Lehre von der Gestalt/* Psychol. Forsch., 1923, 4, 301350. 
Some interesting new illustrations appear In his *'Zu dem Problem der Unterscheidung von 
Einzelinhalt und Teil," Zsch. f. Psychol., 1933, 129, 353-357. 



PHENOMENA OF VISUAL PERCEPTION 95 

tones. Do I observe 327 ? No. Sky, house, trees, and the experiencing 
of the 327 Items as such no one can realize. If in this odd reckoning we 
assume house as 120, trees 90, and sky 117, then I have this "together- 
ness/ 7 this "separateness," and not 127 and 100 and 100. Or suppose I 
hear a melody of 17 tones with an accompaniment of 32 tones. I hear 
the melody and accompaniment, not simply 49, or at least certainly not 
normally and ad libitum 20 plus 29. (p. 301.) 

Although Wertheimer's cryptic and choppy style is not the 
kind of writing most in keeping with his exaltation of integrated 
patterns, it is clear that he is denying the reality of haphazard 
perceptual combinations which psychic atomism makes possible 
when pushed to its logical extreme. That there must be fixed 
laws governing the order found in our visual world is illustrated 
with numerous little drawings. He is trying to answer the 
following general question : Suppose stimuli a b c d e . . . . 
are conjointly effective; what are the principles responsible for 
the typical appearance of such a constellation as the grouping 
or division a b c / d e instead of a b / c d e (Wertheimer, p. 
302) ? In the example below (Figure 8) the diagonally-placed 
dots are seen as units, but not the other diagonal passing through 
abc. Why? 



9 O C 

Figure 8. One of Wertheimer's dot patterns 

With a little critical examination, even the untrained reader 
will hit upon likely and partial explanation. The formation of 
a dot-group with small distances is the one resulting naturally ; 
the point-set involving larger spaces either does not occur at all, 
or when it does arise, is more difficult, artificial, uncertain, and 
labile. Tentatively one may say the unification takes place 
(ceteris paribus) according to the principle of least distance 
(= the factor of nearness) . 3 This rule holds true for more than 

3 Schroff, working with Heidelberg school children, found that the great majority of 
them obeyed this factor of nearness in indicating what patterns "belonged together. " The 
principle worked most powerfully with younger children, suggesting that with experience 
other factors beside nearness operate to produce "natural" groupings. See "Ueber 
Gestaltaufiassung bei Kindern im Alter von 6 bis 14 Jahren/' Psychol. Forsch., 192$* 
II, 235-266. 



96 GESTALT PSYCHOLOGY 

just optical and spatial data as may be seen if one taps out the 
following rhythm : 

e & & o 

12 34 56 

I -2 and 3-4 are organically united but 2-3 are not. 

Consider now a somewhat more elaborate situation as repre- 
sented below : 

o o o o 
o . o o o 
o o o o 

O O O 

o o o o 

If one is asked to describe what is first discerned, the usual 
answer is that the verticals are readily seen but the horizontals 
are not ! This leads to a second principle with this wording : 
If several stimuli be conjointly operative, there is a tendency for 
a pattern composed of like stimuli to appear (= factor of like- 
ness). This maxim, too, is not restricted to the optical field as 
may be observed from any tonal sequence with similar em- 
phases : 

Ill tif e tc. 

These two simple principles can be made the subject of further 
study when they function in the same or opposed directions. 
Which, for example, is the stronger when counteracting near- 
ness or similarity ? But there are other supplementary factors 
operative in many special situations with a less static character. 
Suppose in the following series c d e i k I are simultaneously 
thrust upward above the original line of writing : this common 
movement evokes a new gestalt, for instead of seeing a b c / d e f 
. . . one observes ab/cde/fgh/ikl. 

9 * O 

A A A A A A 

a b c d e f ghi klm 

Wertheimer calls this third influence the law of "common 
fate," or "joint destiny/' which can also function counter to the 
factor of nearness as in this case (Wertheimer, p. 316). 



PHENOMENA OF VISUAL PERCEPTION 97 

A fourth factor is that of objective set 4 (or Einstellung) 
which explains why a stimulus constellation or series C appears 
differently organized if series A and B have preceded it than if 
G, F, E came before. It is the familiar agency of anticipation 
or expectancy which helps determine the structure of our per- 
cepts. 

Another interesting principle is exemplified in Figure 9. In 
the first drawing we spontaneously see "a vertical standing 
upon a horizontal line' 7 or A C / B, despite the fact that geo- 
metrically all points in C are nearer to B than to A. In the 
second sketch one sees just a long straight line with diagonal 
"hooks." The third figure is seen as A C / B, and in the last 




A C ^ A C 

(a) (c) (d) } 

Figure 9. Illustrations of Wertheimer's Principle of Innlgkeit or "Intimacy" 

Certain lines are more readily seen as prolongations of some parts than of others. 

it is obvious what parts are to be considered as continuations 
of others. Continuous lines or dot-constellations are inter- 
changeable and produce identical effects. Certain items which 
belong together are seen as a whole; those which do not, are 
differently segregated. Lines which go all the way through or 
groups in one direction are preferred to others. Wertheimer's 
idiomatic German is all but untranslatable at this point; he 
speaks of "immanent necessities" and "intrinsic togetherness" 
or, more generally, the principle of the "good gestalt" 5 
(Wertheimer, p. 324). 

Still another factor is operative in this situation (Figure 10), 
which Wertheimer calls: "closure" (Geschlossenheit) or com- 
pactness. It may be formulated thus : Where A, B, C, D, are 

4 Later more commonly known as "gestalt-disposition." This factor is a weak point 
in the system as Petermann has indicated. 

5 This seems to be considered the most fundamental "law" of all by its adherents. As 
a theoretical maid-of-all-work, it is called upon as the explanatory factor in an extraordi- 
nary variety of circumstances. 



9 8 



GESTALT PSYCHOLOGY 



present and A B / C D yield two enclosed or completed processes 
and A C / B D open and unclosed ones, then A B / C D is 
preferred in the perceptual response (Wertheiraer, p. 325). 

The factor of position is also effective, which one might 
expect on the general ground that field forces alter with it, 
Note carefully the left and right arrangement in Figure n. 




Figure 10. Diagram Exemplifying the Principle of "Closure" 
Although contiguous, ab forms one clear-cut outline, bd another equally distinct. 

In the one case, two hexagons with a minimum of common 
area are easily perceived; but in the other instance, one typically 
observes a single long hexagon with a little lozenge in the center, 
despite the fact that this little rectangle is formed by a slight 
overlapping of the bases of the two smaller constituent hexa- 








Figure ii. 



Sketch Indicating the Changes Effected by Different Positions of 
the Component Parts of Patterns 



gons. The total field in the a set-up" on the right is more unitary 
and homogeneous than its partner. Position, then, seems to be 
capable of destroying as well as producing configurations. 

The principle of experience or habit, which many behavior- 
ists would insist is the factor underlying all the preceding laws, 
is acknowledged to be influential, but to play a strictly subordi- 



PHENOMENA OF VISUAL PERCEPTION 99 

nate role. For instance, it is previous custom which prompts us 
to respond to 314 cm. as a b c / d e and not as 31/4 cm. or 
314 c/m. An even happier illustration is found in this datum 
(Figure 12) which a Hellenist would see as a dual pattern 
(cursive sigma and gamma) while a Latinist would observe a 
unitary organization (an embellished V). 

/7j**-*f JL-^ Figure I2 - Differing Knowledge of the Cursive 
^-*S V s^~"^ Script of the Ancient Languages Conditions our 
yr Interpretation of this Datum. (After Wert- 

A heimer) 

While thus admitting the presence of earlier repetitions or 
custom as a partial explanation of certain common reactions or 
individual differences, Wertheimer explicitly denies that all of 
the above principles are matters of prior "experience." To ac- 
count for the preference for right angles and circles in these 
situations, some would say that is simply due to the omnipres- 
ence of these forms in the visual world as in house, room, and 
furniture dimensions. However, even assuming that right 
angles occur more often in nature than others which is very 
unlikely as a casual inspection of the twigs of any tree will show 
subjectively, we see right angles only in a frontal-parallel 



Figure 13. How Internal Organization Outweighs Effects 

Traceable to Past Experience 

No one normally sees the familiar 3 or 4, or the E and S con- 
tained within the pattern. 



position of the eyes. Similarly, the circles of our environment 
are overwhelmingly distorted ellipses when projected upon the 
retina. The untenability of the "empirical" theory is almost 
dramatically exhibited in the case of a novel figure (Figure 13) 
where field forces compel one to see something odd and not the 
familiar 3 + 4 + E + S, all of which are parts of the 
constellation. 




ioo GESTALT PSYCHOLOGY 

Although the major part of this work was done between 
1911 and 1914 and hence could not have been directly influ- 
enced by Rubin's significant monograph, nevertheless, it is plain 
that the figure-ground concept is basic in Wertheimer's anal- 
yses. There is not a single instance cited in the foregoing para- 
graphs which does not support the view that the nature of the 
perceived figure is dependent upon the character of the sur- 
rounding field, which may, and as a matter of fact usually does, 
comprise not only a simple "ground" but also other "figures." 
The direction of influence, however, does not only converge 
from the field upon the figure, but may diverge from that pat- 
tern and also affect the quality of the ground, thus establishing 
a condition of equilibrium characterized by reciprocity. 

The burden of Wertheimer's study is that we do not react 
in uniform or constant ways to specific stimuli as though they 
were always insulated, but that the nature of the setting in 
which they are found determines the mode of organismic re- 
sponse. In psychology, the right formula is, Constellation of 
stimuli >- Organisation ^-Reaction to results of organiza- 
tion, rather than the usual S >- R type. The organism is not 
barren functionally, it is not a box containing conductors each 
with a separate function; it responds to a situation, first, by 
dynamical events peculiar to it as a system and, then, by be- 
havior which depends upon the results of that dynamical or- 
ganization and order. This is the cardinal principle of all 
Gestalt psychology : the whole is something other than the sum 
of its parts it is genetically and functionally prior to them. 6 
Such is the Leitmotiv which runs through the impressive 
mass of experimental literature reported in the Psychologische 
Forschung and liberally scattered through other periodicals. 
Since a general theory becomes convincing to the extent that 
it offers satisfactory interpretations and predictions in diverse 
fields, it will be desirable to follow its application in the various 
situations to be described below. 

Methodology is often pushed into the foreground with the 
advent of new theoretical positions as psychologists have re- 

6 Most persons can retain a melody, fewer recognize a definite interval, and still fewer 
the individual tones, as shown by the rarity of absolute pitch. 



PHENOMENA OF VISUAL PERCEPTION IQI 

cently seen in the debates concerning the adequacy of behavior- 
istic technique or the psychoanalysts' "talking cures." Few 
writers have seriously questioned the legitimacy of the pro- 
cedure used by the configurationist school, since in the main 
it conforms to the naturalistic and experimental tradition of 
modern and scientific positivism. However,, since there is some 
reason for believing that the way in which facts are discovered 
is often as important as the facts themselves, researches which 
are primarily directed toward specifying the differential effects 
of various methods upon the data obtained have a significance 
all their own. 

A study by Katona 7 falls into this category. He was con- 
cerned with an examination of the consequences of successive 
and simultaneous comparison of stimuli. In one division of 
his work, he exposed six nonsense drawings rapidly one after 
another like a pack of cards, or presented them all at once by 
withdrawing a screen placed before six adjacent figures. In 
the critical test series, three of the old pictures and three new 
were mingled, and the subject indicated which were familiar. 
To remove the objection that the manner in which the testing 
was done was influential, a successive learning exposure was 
measured later by the simultaneous procedure and vice versa, 
with results from three subjects as indicated in Table I. 

TABLE I. INFLUENCE OF THE MANNER IN WHICH VISUAL LEARNING 
AND RECALL ARE MADE UPON THE FREQUENCY OF ERROR. (AFTER 

KATONA) 



Manner of 
Impression 


Test 
Method 


Average Number 
of Errors 


Successive 
Simultaneous 


Simultaneous 
Successive 


5-33 
11.13 



The inferiority of the simultaneous-successive sequence to 
its converse is explained in terms of the whole-part contrast, 
simultaneous comparison being favorable to the formation of 

7 "Experimentelle Untersuchungen iiber simultane und sukzessive Gesichtswahrneh- 
tnungen," PsychoL Forsch., 1926, 7, 226-256. 



102 GESTALT PSYCHOLOGY 

a total but unfavorable to isolated units, while the opposite situa- 
tion holds for the successive impression. As far back as 1846, 
E. H. Weber had announced that two simultaneous touch sensa- 
tions cannot be distinguished as well as two successive ones, 
because of the mutual disturbance they exercise upon each other. 
Stumpf s stated in connection with tonal experiences that ' 'suc- 
cessive sensations constitute a mere sum, but simultaneous ones 
form a whole/' Contiguity In time, the fundamental law of 
association, would seem to be the non-Gestalt phrase which 
best accounts for the peculiarities of simultaneous sensitivity. 
To configurationism, comparison is no longer a new act super- 
vening upon the given sensation. The question how the two 
sensations are compared no longer exists, because the two sen- 
sations themselves do not exist. All that is found is an un- 
divided, articulated whole. 

Dependence of Form and Space Qualities upon the 
Configuration 

Some of the more powerful supports of the correctness of 
the configurationist's thesis are found in the facts of animal, 
child, and primitive behavior. The reason is simple : if among 
relatively undeveloped mentalities one finds phenomena which 
are more in accord with the Gestalt hypothesis than with any 
alternative explanation, the arguments in its favor are strength- 
ened. Whatever laws may govern human conduct presumably 
are best studied under less complex conditions than those affect- 
ing the "normal adult European white male." The restriction in 
variability which more primitive circumstances impose makes 
the investigator much more certain of the generality of his 
findings. 

Consequently, especial interest attaches to a novel study of 
crows' responses to figural elements made by Mathilde Hertz. 9 
She arranged a large number of inverted brown clay flower pots 
about five centimeters in diameter in various patterns; under 

8 Tonpsychologie, 1890, I/, 64. 

Daughter of the distinguished physicist Her researches appear an 
e Untersuchungen an Eichelhaher," Zsch. f. ver S l.PhyM. 



PHENOMENA OF VISUAL PERCEPTION 103 

one of the pots within sight of the confined bird, a piece of 
food was placed. Most of the items were, therefore, "negative" 
and only one "positive." When the bird was released, the 
combined memory and perceptual difficulties confronting him 
were similar to those which we experience in trying to keep track 
of a particular individual in a swarm of sea-gulls. In this ex- 

O oo 00r , o o oo o4 b o 

O O O O O O O o OO O 

o 04 "o 




Figure 14. Each Circle Symbolizes a Clay Pot 
The plus sign marks the one containing the objective. 

periment the crows had no difficulty with pots distinguished as 
to form, size, or color from the rest of the complex situation, but 
made errors when position was the sole clue, unless that location 
happened to be pragnant or in some way set off from the confu- 
sion stimuli, as in the examples in Figure 14. Any type of cen- 



Figure 15. Materials Employed in the Training Series with Crows to Test 
their Ability to "Transfer" Habits of Brightness Discrimination. (After 

Hertz) 

The birds responded correctly in critical situations when other figures than circles 
were substituted as "inner fields. 

tral cluster with the goal placed peripherally to the main mass 
succeeds with a minimum of error. 

This pronounced form-sense of the crow is even more spec- 
tacularly demonstrated in a supplementary account of figure- 
ground transfer. The birds were trained to react positively to 
all objects with a light figure on a dark ground, and to dis- 
regard the items having a dark figure on a light field as in 
Figure 15. 



104 GESTALT PSYCHOLOGY 

So well was the appropriate figure-ground relationship ap- 
preciated that the crows had no difficulty in responding success- 
fully according to the comparative brightness difference when 
crosses, squares, stars, and other "figures" were used in place 
of the circles shown in the figure. 

An impressive rejoinder to the common "empirical" explana- 
tion of Gestalt phenomena is provided by Kurt Gottschaldt 10 in 
an experimental procedure which is all but Euclidean in char- 
acter. Normally, most psychologists shrug their shoulders 
when the nativistic-empiristic controversy is mentioned, because 
it appears so futile and impossible to set up the requisite condi- 
tions early enough in the life of the developing organism which 
would be adequate to test either hypothesis. However, Gott- 
schaldt's highly original attack consisted in submitting to his 
subjects several hundred times a simple figure for overlearning 
and then presenting a novel context in which the original was 
"imbedded" to see if it could be spontaneously identified. For 
instance, the a-figures in Figure 16 were presented a varying 
number of times and then the corresponding &-figures were ex- 
posed, the observer being required to describe what he first saw. 
Only in rare instances did the a-figure "spring out" of its 6- 
setting. To one who has pinned his faith in the explanation of 
visual organization by experience, it is startling to encounter 
Gottschaldt's evidence that when the a-figure forms a dependent 
part of a larger whole, the fact that it has been presented more 
than 500 times as an independent total does not noticeably in- 
crease the chances of its being seen as such in the new setting. 

Since it is doubtful if the casual experience of ordinary life 
ever gives one as many repetitions of the stimuli commonly in- 

10 "XJeber den Einfluss der Erfahrung auf die Wahrnehmung von Figuren," Psychol 
Forsch. f 1926, 8, 261-317, and 1929, 12, 1-87. Koffka had earlier called attention to the 
fact that the experience explanation was unable to account for the apparent "translation" 
ol a red square into a yellow circle, which is a familiar stroboscopic effect. No observer 
has ever witnessed such an event and yet it can be obtained by presenting one object to the 
left eye and another to the right. Koffka holds that empiricism and nativism are both 
false, or better, that the opposition between them is resolved by this new way of viewing 
sensory impressions i.e., as dynamic resultants of the complex immediately-given situa- 
tion, bee his article under a title similar to Gottschaldt's in Die Naturwissenschaften 
1919* 7, 597-005. J ' 

At Heidelberg, Strauss had made a similar study, emphasizing, however, the ability 
of the subject to isolate a constant pattern in a succession of varying forms. She found, 
W/?^' ^*r Vu* C ? n only be se ^ n as a sub-whole in a larger constellation where 
Werthettner s Gestalt factors are operative. Cf. "Untersuchungen uber das Erloschen und 
Heraussprmgen von Gestalten," Psychol. Forsck, 1927, 10, 57-83. ^nobcncu una 



PHENOMENA OF VISUAL PERCEPTION 



105; 



volved in laboratory studies as those deliberately secured here, 
and since even directions to search for the original figure did 
not noticeably aid the subject in discovering it, Gottschaldt con- 
cludes that the total configuration must have completely altered 
the functions of the parts. As soon as certain contours of the 
a-figure serve to outline some strange sub-wholes in the fr-figure, 
they lose their original character and acquire wholly new func- 
tional properties. 11 






Figure 16. How Distinctive Figures Become "Lost" when Imbedded in a 

Different Perceptual Setting. (After Gottschaldt) 
The "a" figures are all contained in their corresponding "b" patterns. 

The problem of depth perception was cleverly attacked by 
Kopf ermann 12 with the guidance which earlier configurationist 
experiments provided. She was concerned primarily with the 
question, "How do two-dimensional sketches succeed in giving 
an impression of spatiality?" The common theory holds that 
the way in which attention is divided on the basis of prior ex- 
perience is the determining agent : certain parts of the figure are 
interpreted as "front" and others as "rear." However, care- 
ful observation shows that in the transition from a plane to a 
spatial apprehension the changes which occur are not restricted 
to depth values and "attensity" ; there also occur characteristic 
structural transformations, re-groupings within the figure, clo- 

n An American critic has objected that the connotation of "experience" as used in 
Gottschaldt's research has been unwarrantedly limited to frequency of repetition. Cf. 
M. G. Moore, "Gestalt vs. Experience," Amer. /. PsychoL, 19:30, 42, 453-455- 

12 "Psychologische Untersuchungen liber die Wirkung zweidimensionaler Darstelhmg 
korperlicher Gebilde," Psychol. Forsch., 1930, 13, 289-364. 



io6 



GESTALT PSYCHOLOGY 



sure effects, etc. How is this process of reversal to be described 
phenomenally ? 

If one looks at Figure I7a, one will probably see first an "un- 
even rectangle with diagonals" followed suddenly by a "tetra- 
hedron.'* * Geometrically, the transition is accomplished by the 
following shift : At first, the "composition" present is (a, &, c, 
$} 5 (X /)? ( 9 3 ^)j but after the conversion into the third- 
dimension, one sees [a, d y (e + /)] ; [a, b, (g + A)] ; \_d, c, 
(h + #)] ; [b> c, (e + /)]. It is psychologically impossible 
to maintain the former "togetherness" and realize a solid, and 
the converse is equally true. This perception of the parts in 
a definite role is the essence of uaiderstanding a figure. The 
individual parts need not be observed but their union must be, 
as, e.g., when we see a complicated crystal and react to it 
spatially. 






(a) (W (c) (d) 

Figure 17. The Third Dimension as a Product of Gestalt Factors 

Whether (a) is perceived as a plane or as a solid hinges upon the manner in which 
the parts are subjectively combined. The "depth" character in (b) is due to the double 
function of the contour a as opposed to the unitary function of contour b. Conversely, 
the "plane" quality of (c) results from viewing two transparent solids in perspective. 
In (d) the tilt of the transverse bar obviously determines the different spatial effects of 
the two crosses. 

Kopf ermann believes that the boundary- functions of various 
linear forms which fluctuate with the setting in which they are 
found play an important part in determining whether a two- 
dimensional or three-dimensional figure shall be seen. For 
example, a circle drawn on a white surface will normally be 
seen as a plane, but if a square be drawn around it, as in Figure 
I7b, so that it becomes a figure within an enclosing area, the 
double function of its contours stand out more sharply and the 
entire figure acquires a "depth" property. Contour b has only 
a limiting function toward within, but contour a limits the circle 
centripetally and the square centrifugally. 

Spatiality would seem, then, to be a product of field forces. 
This view was experimentally tested by means of a light-proof 



PHENOMENA OF VISUAL PERCEPTION 107 

"tunnel" box in which glass plates were arranged one behind 
the other, each containing a different section of some drawing. 
Figure ijc, for instance, was produced by two separate but 
identical sketches of a cube, held some distance apart. When 
placed in a straight-line vision, most observers saw this as a 
symbolic or formal "web" instead of a doubled cube, indicating 
that the objectively existing differences in depth were overcome 
by such figural factors as the nature of the "composition," the 
contour- functions, and the relation of the parts to the whole. 
The influence of mere position within the field is also important 
as one may see from the non-spatial effect created by the sym- 
metrical cross in Figure ijd and the definite spatial character of 
its asymmetrical neighbor. 

The "wholeness" of our visual percepts is curiously illus- 
trated by some quaint observations of Hornbostel on optic 
inversion, from which he concludes that convexity and con- 
cavity are "total" properties. He constructed some wire models 
of solids of the type sketched in Figure 18, and held them be- 
fore a mirror in various positions. When such a figure is turned 
in different planes, it "tumbles" from one apparent shape to 
another like the familiar figure-ground reversals, save that 
with a tri-dimensional object more possibilities are opened; 
Figure 18, e.g., can be apprehended in at least four ways. When 
the inversions occur, they come as a single complete unified 
movement: one part does not "tip," then another, and so on. 
These quasi-solid inversions are neither illusory nor ideational 
"constructs," but things which under special conditions we 
perceive. 

The phenomenal properties of a "convex" percept are typ- 
ically distinct from a "concave" one. A convex affair is closed, 

13 Contrast this explanation with Washburn's motor account: "In the ambiguous cube, 
whether a cube resting- on the ground or one suspended in air is perceived may depend 
upon whether one fixates one point or another; the point fixated seems to be nearer, i.e., 
suggests a shorter reaching movement, because in looking at an object we generally look 
at its nearest point, which, of course, carries with it that surface of the cube to which it 
belongs. Further, if a record is kept of the duration of the phases of the cube, we usually 
find that it is seen longer as a cube resting on the ground than as a hanging cube; this may 
be explained by the fact that we have a stronger tendency, due to habit, towards sitting 
down on cubes than towards looking up at them." "Gestalt Psychology and Motor Psy- 
chology,** Amer. /. PsychoL, 1926, 37, 516-520. Her entire criticism of cpnfigurationism 
is based upon a preference for an explanation in terms of our muscular activity in connec- 
tion with objects previously experienced. Wk et k er one * s entitled to maintain this position 
after Gottschaldt's and Kopfermann's experiments is an open question. 



io8 



GESTALT PSYCHOLOGY 



excludes the observer, projects forward, cannot be penetrated 
visually because of its opacity or manipulated because of its 
impenetrability. Visual "objects" are convex. A concave item, 
on the other hand, is open, embraces the observer, permits visual 
and manual exploration, and possesses the characteristics of an 
empty background. ''Spaces" are concave. The entire process 
of inversion involves making a convex item concave and vice 
versa, although for some reason it is normally harder to invert 
a convex object than a concave one. 14 




Wire Framework 



Metal Handle 



Figure 18. One of Hornbostel's Reversible Solids 



When held before a mirror and properly rotated, the image alternately produces a 
convex or concave impression. For details, see the description in the text. 

A considerable number of American psychologists have been 
interested in taking experimental hints from the continental 
Gestalt group and by repeating, refining, and checking their 
technical procedures have contributed to advance our under- 
standing of configurationist theory and data. Wever, 15 e.g., 
took Rubin's "nonsense figures" and by a series of tachisto- 
scopic exposures of varying lengths studied the characteristic 
changes in appearance which occurred as a result of the different 
presentation-times. The most primitive figure-ground experi- 

^"Ueber optische Inversion," PsychoL Forsch., 1922, i, 130-156. The familiar 
clinical reading disability of confusing "saw," "was/* "b," "d," etc., may be more than 
just a case of poor directional habits ! 

15 "Figure and Ground in the Visual Perception of Form," Amer. J. PsychoL, 1927, 
38, 194-226; "Attention and Clearness in the Perception of Figure and Ground," ibid,, 
1928, 40, 51-74- 



PHENOMENA OF VISUAL PERCEPTION 109 

ence apparently arises when the exposure lasts about ten milli- 
seconds. The indispensable characteristic of such a perception 
appears to be a certain type of heterogeneity in the visual field : 
the field is so differentiated that two separate, intrinsically homo- 
geneous regions are seen. As one would expect, with this 
feature of heterogeneity is always correlated a difference in 
brightness between the two fields. A primitive bounding mech- 
anism is thus established. The limen for a continuous or com- 
pletely enclosed contour is somewhat higher around 13-14 
milliseconds. Localization effects next emerge, and as Rubin 
observed, the figure seems to protrude, but the ground is re- 
cessive. In general, as the time of the presentation is increased 
the "goodness" of the figure-ground experience is enhanced. 
These time-relations easily make a dull recital, but they do show 
plainly what are the essential and the less significant qualities 
of the figural experience. One important finding of Wever's 
is his "discovery" that clearness is the presence of something 
that can be reported upon; since clearness is introspectively the 
heart of attention, its dependence upon the figure-ground oppo- 
sition is of high theoretical importance as Rubin and the Gestalt 
school have maintained. 

Contours acquire a special significance in Gestalt theory, be- 
cause, whether sharp or weak, they serve to mark a figure off 
from its ground and thus establish the "natural" wholes with 
which configurationism prefers to deal If the contour were 
not present to differentiate the two fields, no figure would be 
present, and no meaning or significance would attach to the 
presentation. The physico-chemical forces which operate 
within the contour are very distinct from those outside, and 
the perception of the resulting figures presumably rests upon 
an emergence or a central physiological delimitation of a spe- 
cific area within a field. However, in our present state of 
knowledge these underlying dynamic agents cannot be identified 
except in terms of the grosser designations of nearness, simi- 
larity, etc., as indicated by Wertheimer. Nevertheless, some 
comparison of their relative strength can be made. Liebmann 16 

1( ? M Ueber das Verhalten farbiger Formen bei Helligkeitsgleichheit von Figur ttnd 
Grand," Psychol. Forsch., 1927, 9, 300-353, 



no GESTALT PSYCHOLOGY 

reports marked phenomenal differences in the appearance of 
colored objects with and without brightness equivalence. Ap- 
parently, when variations in brightness are excluded, colored 
figures exhibit a lack of solidity, firm outlines, and "thing" 
character. Based upon threshold determinations for both grey 
and colored fields, she found that brightness differences are far 
more influential than color distinctions in producing the funda- 
mental perceptual situation of a figure upon a ground. 

A curious incidental finding is worthy of record. If an 
observer is placed at a distance so that the accompanying figures 
a and b appear united without a gap, and then 
gradually allowed to approach the stimuli, he 
must come nearer in order to discern the little 
break in the plus sign than to distinguish as 
separate the two complete figures, despite the 
objective equality of the spatial interval. 
The reader will probably be reminded of the 
"closure 5 ' explanation 1T as pertinent here and 
in view of Gottschaldt's evidence on the un- 
tenability of the experience hypothesis it must 
be admitted as a plausible alternative. 
What has since come to be known as the "Liebmann effect" 
has been subjected to an exhaustive analysis by Koffka and Har- 
rower. 18 As stated above, this name refers to the observation 
that mere color difference without the aid of a concomitant 
difference in brightness has surprisingly little organizing power. 
Colored figures on a neutral or colored ground lose their defini- 
tion, become blurred and utterly diffuse if the brightness of 
figure and ground are made equal. Evidently pure color differ- 
ence is not the main factor in segregating an area from its 
surrounding area. 

Koffka and Harrower devised a photoprojector for varying 
the color and brightness of various figure-ground combinations. 
They were primarily interested in noting the effect of the pres- 
ence or absence of thin black or white contours upon the prop- 

17 Note in this connection how the assumed "canals" of Mars may be as much a psy- 
chological reality as an astronomical. The Landolt ring in optometry is also affected by 
this interpretation. 

18 "Color and Organization," PsychoL ForscJi., 1931, 15, 145192 and 193-273. 



PHENOMENA OF VISUAL PERCEPTION m 

erties of the contained color. They found the non-contour 
figure more highly saturated than the contour figure, the color 
of the contour figure being hard, glossy and condensed, that of 
the non-contour figure soft and spongy. A contour, whether 
black or white, makes the color within the figure darker. Since 
both black and white contours have the same effect of darkening 
the color of the enclosed figure, this cannot be produced either 
by contrast or by assimilation, but must be due to the higher 
degree of segregation which the contour figure possesses with 
regard to the non-contour ones. It appears as though the energy 
consumed in producing the better organization were supplied 
by a loss in brightness. 

Regardless of the contour influence, red and yellow have 
relatively high articulatory power, i.e., they are "hard" colors, 
whereas green and yellow are low in articulatory power, i.e., 
they are "soft" colors. Throughout these observations the 
general Gestalt thesis appears to be upheld, viz., that the color 
in any part of the field depends upon the formal or figural char- 
acteristics of this region, and conversely the articulation of 
the field into different units depends upon the differences of 
brightness and color within its boundaries. 

Perhaps one of the finest examples of the integrating possi- 
bilities of Gestalt theory is a study executed by Erna Schur 19 
under Kohler's direction on the ancient problem of the moon 
illusion and its relation to the constancy of visual magnitude. 
This issue has baffled some of the world's keenest intellects, 
largely because the appropriate experimental insight was lack- 
ing. Aristotle supposed that atmospheric conditions near the 
horizon were responsible for the differences in the optical size 
of the heavenly bodies, but photography and exact physical 
measurements do not yield variations large enough to account 
for the extraordinary change in appearance at the zenith. The 
moon really casts a larger retinal picture at the zenith than at 
the horizon, since the observer is nearer by the distance of the 
earth's radius. With the discarding of explanations based upon 
refraction have appeared many built around the so-called errors 

19 "Mondtatisdmng 1 und Sehgrossenkonstanz," PsychoJ. Forscli., 1926, ? t 44-80. 



H2 GESTALT PSYCHOLOGY 

of judgment; e. g., the moon near the horizon is said to look 
larger because of the greater contrast with "neighboring" earthly 
objects in the same line of sight. However, when one looks 
upward in a canyon where terrestrial objects are also visually 
"near" the zenith, the moon reveals no such equivalent increase 
in size, since the apparent ratio of the average lunar diameter 
in culmination and at the horizon is about I to 2.5. 

Acting upon some earlier abortive investigations of Stroo- 
bant, Zoth, 20 and Guttmann, Schur prepared to reproduce the 
essence of the celestial phenomenon on a terrestrial scale. A 
simple apparatus was designed capable of throwing a circular 
disk of light upon a homogeneous dark background from a 
central floor position to the ceiling and directly forward upon 
one of the walls. The distances compared ranged all the way 
from a 3-meter room to a 32-meter height in a large theatre and 
Zeppelin hangar. Adjustments in the size of the comparison 
disc were made until it was judged subjectively equal to the 
standard. Typical data for one observer are given below : 

TABLE II. SUBJECTIVE DIFFERENCES IN EQUATING LIGHT Discs FOR 
SIZE WHEN ONE is PROJECTED FORWARD AND THE OTHER ABOVE. 

(SCHUR) 





Actual distance (in meters) , 




4.80 


6 




22 






^ 






o D 




f 


Overhead diameter (in cm.) , 


. . 6.8 


IO.Q 


13.6 


^0 


^0 




Mean diameter of the lower circle. . . , 


.. =5.8 


Q.I 


10. 1 


17 


27 


( 















From the corresponding curve one can compute that after 
a distance of 100 meters the magnitude of the illusory effect re- 
mains constant. A control test was instituted by reversing the 
direction of the head and eye movements, a requirement which 
was met by having the observer stand upon an elevated plat- 
form. In this case, when the diameter of the disc below was 
18 cm. at a distance of 3.60 meters, the circle straight ahead had 
to be about 20.5 cm., or 14% larger, before a judgment of 
equality was reached. 

20 This writer interpreted the "illusion" as a result of over-compensated convergence 
tendencies of the eyeballs. 



PHENOMENA OF VISUAL PERCEPTION 113 

Gauss and Zoth had Independently pointed out the probable 
influence of differences in the degree of convergence of the eye 
musculature upon the presence of the illusion, and it is more 
than accidental that the effect increases up to 100 meters, the 
customary limits of convergence, although the curves for the 
two are far from running parallel. That the direction of the 
visual axes is significant is clear from the fact that the negative 
after-image of the setting sun is much smaller at the zenith than 
at any point on the horizon. This change is all the more re- 
markable when one remembers that the size of the retinal region 
affected is unaltered. Moreover, that this is not a solar peculiar- 
ity is easily demonstrated by using an analogous white paper 
circle on a black background where the same results are obtained. 

The dependence of the phenomenon upon the position of the 
observer was neatly elicited under circumstances which exag- 
gerated the role of the extrinsic eye muscles. The subject 
lay flat upon his back and compared a "moon" upon the ceiling at 
a distance of 5.20 meters with one equidistant on the wall be- 
hind him ! Since the essential conditions of the illusion were 
maintained, it was anticipated that the object on the wall would 
appear smaller because it corresponded to the normal zenith 
position of the moon to an erect observer. Actually, the ex- 
pected diminution occurred but it was relatively small (5- 
13%), probably because of complicating influences exerted by 
the static organs and tonic tensions of the neck musculature. At 
any rate, the fertile character of the Gestalt hypothesis in lead- 
ing to new experimental proposals is amply demonstrated by 
this study. 

One of the most positive results of Gestalt experimentation 
has been the complete refutation of the conventional theories of 
the "blind spot." Since the blind spot contains no rods and 
cones it is easy to show that if an isolated object falls within 
that area of the retina it Is not seen. Every undergraduate is 
familiar with the simple and impressive demonstration sup- 
porting this view. Nevertheless, it is well to remember that 
this blindness is relative, for while it is true that one does not 
see clearly the figure at that point, the background, no matter 



GESTALT PSYCHOLOGY 

how homogeneous it may be, is preserved intact. Field-forces 
operating according to the principle of closure give the Gestalt 
version of the phenomenon. Moreover, Tschermak somewhere 
reports that if a colored patch, say blue, is projected upon the 
blind spot, the subject perceives a ring-like halo of yellow, indi- 
cating that contrast effects which are fundamentally dual in 
character are not suspended at this point and that light sensi- 
tivity of some kind must be preserved. 

Stern 21 showed that the perception of motion is still possible 
when the path traversed by the "object" must pass through the 




Figure 19. A Blind-Spot Card Modelled after Scripture's 

Close the 1-eft eye and fixate the cross at a distance of about six inches. If the eye 
movements are carefully controlled, the white circle disappears entirely and ,- - " 
straight Ime border between the black and grey areas takes its place 



. a pei 



he eye 

xfectly 



blind spot. Using a modification of Wertheimer's technique 
for exhibiting apparent movement, she found that the phi-phe- 
nomenon occurred if two light points were successively exposed 
on the upper and lower border of the blind spots ; in fact, move- 
ment was actually perceived more readily and distinctly than 
upon the rest of the retina ! Even more spectacular was hei 
discovery that motion was present if one or both of the neces- 
sary^ light points was exposed within the "blind" area, despite 
inability to see each point when exposed singly. It is knowr 
that a straight line or circular circumference passing throng* 

^w/i-? mUIlg V n Bewegun ^ en In der Gegend des ttinden Flecks," Psyehol 



PHENOMENA OF VISUAL PERCEPTION 115 

the blind spot is not seen as "broken" a fact for which the 
configurationist can easily account by the photo-electric gra- 
dients responsible for gestalt completion and the law of 
Pragnans. The reader can easily determine this for himself by 
projecting the white circle in the Figure 19 upon his blind spot 
when the eye is fixated upon the cross. Although the white 
circle "disappears," the boundary line between the black and 
grey surfaces appears straight and the "ground" figures show 
no evidence of indentation. A small isolated form vanishes 
when cast upon the blind spot but a larger figure is preserved 
intact by a completional or totalizing process released by the 
rest of the field. 

It should be emphasized that the completion observed is not 
at all "mental" (as Poppelreuter has maintained) but takes place 
wholly within the sensory sector. The difference can be made 
clear by the following trick : Stand a friend against a wall sur- 
face in such a way that his head falls within the blind spot. In 
spite of our experience that every person has a head, the head 
is not seen! Commonly, a rounded light impression is pre- 
served, but the facial details vanish. Why? Presumably be- 
cause a head is in itself a natural sub-total, whereas sections of 
a stick or simple circular figure obey the dynamics of closure 
under these circumstances. 

Kelson's extensive investigation 22 has also shown that the 
optic disk he prefers this term to the misleading label of 
"blind-spot" can be illuminated while the rest of the retinal 
field is dark, and that under these conditions after-images, con- 
trast effects, adaptation, and many other color phenomena char- 
acteristic of normal vision will be found in the blind-spot, only 
in somewhat different degree. 

Dependence o Brightness and Color Qualities upon the 
Configuration 

One of Kohler's classic experiments 23 was designed to re- 
veal the principle of relativity in animal conduct. The idea was 

22 "The Effects of Direct Stimulation of the Blind Spot," Amer. J. PsychoL, 1929, 

23 "Optische f Untersuchungen am Schimpansen und am Hausliuhn," Abhandluneen d 
A. Prcuss. Akad. d. Wisscnschaften, 1915, Phys.-math. Kl., Nr. 3. 



Ii6 GESTALT PSYCHOLOGY 

very simple. A number of hens were trained to eat kernels 
of corn from the darker of two grey paper surfaces, A and B, 
and to avoid the lighter. As in all good objective studies, the grey 
grounds were alternately placed on the right or the left during 
the pecking process, so as to avoid the position habits which ani- 
mals readily establish. The training was continued long enough 
(400 to 600 trials) to ensure a certain degree of over-learning 
as measured by a succession of errorless choices. 

Suppose now the perceptual field is changed. If A in the 
above situation represents a light grey and B a middle grey, and 
a new dark grey ground C is paired off with B in place of A, 
from which surface will the hens now peck ? According to the 
theory of restricted bonds, one would expect the chick to go to 
B, since that is the identical surface which was presented in the 
first series and consequently has the advantage of familiarity. 
However, in about 70% of the critical series the hens pecked 
from the new grey paper C. 24 Why? This utterly unforeseen 
result makes one question the validity of the original basis of 
prediction. But on Gestalt principles, the outcome is very 
simply explained. The animals had not responded to a par- 
ticular brightness value 25 but to a relation the "darker of the 
two !" When this interpretation is checked in a new training 
series by positive drill upon the lighter of a pair instead of the 
darker, similar behavior occurs and the configurational account 
is confirmed. Have we not here a species of transfer which 
satisfies neither Thorndike's theory of identical elements, nor 
Judd's theory of generalization, nor any other of the classical 
transfer hypotheses? 

This has been a hard nut for opponents to crack. The sta- 
tistical reliability of the differences between the choices has been 
questioned and some studies have reached opposite conclusions. 26 

24 The same type of experiment had already been performed by H. C. Bingham in 
America. He trained chicks to select the larger of two circles and then presented them 
with circles reduced in size from the original, but with the size ratios maintained. Those 
animals who had been taught to pick a 6-cm. circle and to reject a 4-cm. one would select 
the latter when it was given in company with a 3 -cm. circle. Evidently, relative rather 
than absolute magnitude had been discriminated. See "Size and Form Perception in 
Gallus domesticits," J. Animal Behavior, 1913, $, 65113. This entire study has unfortu- 
nately been neglected and deserves to be much better known than it is. 

25 In his experiments, Pavlov found that a dog can be trained to a tone as a food signal 
with a coordination of the reaction to the absolute quality so striking and definite that no 
reaction occurred even with tones deviating only to a very small extent. 

20 See Howard Taylor, "A Study of Configuration Learning," /. Comp. PsychoL, 
1932, 13, 19-26. Taylor asks: If four hens in the critical series selected the neutral paper 



PHENOMENA OF VISUAL PERCEPTION u 7 

Still the vast amount of controversy which has arisen around 
this special problem is one indication of its fundamental signifi- 
cance, not only for psychological research, but for the me- 
chanics of social control, especially in education. Kohler's little 
supplementary experiment with a three-year-old child shows the 
"human" implications of his major study. A situation was ar- 
ranged in which the brighter of two colored boxes always con- 
tained candy. Errorless responses were established after 45 
trials, following which a new and brighter box was presented in 
the company of the original positive stimulus, replacing the 
darker negative box of the training series. Under these changed 
conditions analogous to those confronting the hens the child 
always selected the new bright container, rather than the one 
from which it had previously secured the candy. Presumably 
the native and primitive perceptual responses as they appear in 
both racial and individual development are more in harmony 
with a view which stresses totals and wholes than with one which 
maintains the primacy of elements. 27 

Koffka 28 has made many fine observations of out-of-the-way 
visual phenomena, but none which exceed in vividness his ex- 
periments pointing to a configurationist theory of marginal con- 
trast. If one rotates a Massgn disc with its familiar concen- 
tric grey rings, it is possible to isolate a single ring so that one 
can detect the contrast haze on both of its borders. If now a 
screen is held over the disc in such a way that it covers one of 
the marginal contrast bands of the ring attended to, the other 
forthwith disappears ! The contrast apparently exists only be- 

59 times and the original positive paper only^ out of a total of 85 trials, is it not possible 



resolved in the interests of both fact and theory. Perhaps there are gross individual differ- 
ences among chicks, some behaving "analytically" and some "synthetically." Certainly 
children react in these two distinct ways. 

- ^ "Kleine Mitteilungen aus dem psychologischen Institut der Universitat Giessen." 
Psychol. Forsch., 1922, 2, 145-153. 

Y ?" u ber Feldbegrenzung und Felderfiillung," Psychol. Forsch., 1922, 4, 176-203. 
In 1878, Eugen Fick had reported some observations on "field forces" in perception. If a 
pin-hole is made in a cardboard and the latter laid over a bit of colored paper in bright 
illumination, it is impossible to identify the color at 6.5 meters. If now 15 added pin-holes 
(making a square of 16, four on a side) are inserted far enough apart to be differentiated, 
the color is quickly recognized. Cf. Pftiigers Archiv., 17, 152-153. 



Il8 GESTALT PSYCHOLOGY 

cause of a brightness gradient, and as soon as one pole of the 
potential is destroyed the other simultaneously vanishes. This 
finding led Koffka to the generalization that marginal contrast 
can occur only in a field which is itself imbedded between a 
brighter and a darker. Elimination of this necessary gradient 
means the elimination of the effect 

A good way of appreciating some of the successes and posi- 
tive contributions wrought by Gestalt theory is to follow the 
changes it has introduced into the very citadel of orthodox ex- 
perimentalism psychophysics. This division of the science, 
which has seemed hopelessly barren to every younger genera- 
tion since James' f ulrninations, is, nevertheless, in its techniques 
of precision an imposing paradigm of methodology. Conse- 
quently, the uses which configurationism, the newest speculative 
position, makes of the psychophysical approach, perhaps the old- 
est quantitative procedure peculiar to psychology, cannot fail to 
be of interest. 

The exact determination of absolute and differential thresh- 
olds in the separate sense modalities has been one of the major 
items of business assigned to the psychophysicist. Gelb and 
Granit 29 raised the original question whether the lower limen of 
color vision for the primary hues was influenced by the presence 
of the stimulus to be discriminated upon an area which could be 
seen either as figure or ground. Previous studies had con- 
sidered only the effect of the brightness value of the field, but 
these investigators believed it made a difference if the field was 
a closed entity like a ring or a homogeneous field. Their sub- 
jects looked with one eye through a blackened tube at a grey 
Maltese cross (similar to Rubin's pattern; see Figure 4) out- 
lined so as to yield a readily reversible combination. A sublim- 
inal spot of red light projected upon it was gradually increased 
in luminosity until distinguished as such. In all instances, the 
color threshold was higher when the light was cast upon the 
phenomenal figure than upon the ground ; qualitatively, too, the 
color appeared suddenly within the figure, but emerged as from 



R <r\c \T TT w *^ vnBa^jL/ucua auuucr n LUC iieiu is perceived as ground rat 

figure. Cf. Neue Verschmelzungsprobleme," PsychoL Forsch., 1923, 3, 319-396. 



PHENOMENA OF VISUAL PERCEPTION 119 

a depth in the ground. As a control upon this observation, a 
red-white reversible cross was employed with a clear patch of 
red light upon the white region. In this case, the spot paled 
noticeably when the white cross was isolated as figure, but as 
soon as the red cross was thus apprehended its intensity was re- 
stored ! Fluctuations in attention do not offer a plausible explana- 
tion, since in both figure and ground the projections of the light 
patch lay equidistant from the fixation point, the center of the 
cross. The interpretation which Gelb and Granit give is that 
the inner field or figure becomes functionally a "ground" when 
the light spot appears upon its surface. Its resistance to this 
fundamental change in its perceptual character is greater than 
that offered by the original ground ; hence, a higher light inten- 
sity or threshold is required before it succumbs to the differen- 
tiation process. 

The influence of form, even in the narrower sense of the 
word, is well illustrated "in some fine observations of Benary, 30 
who carried out the suggestions contained in one of Wert- 
heimer's lecture-demonstrations : A plain black cross rests upon 
a white field. Of two objectively equal grey triangles, one (Ti) 
is placed upon the surface of the cross itself, the other (T 2 ) 
in a corner enclosed by two arms of the cross so that geometric- 
ally more black but less white adjoins it than in the case of TI. 
According to summative theories of brightness contrast, the 
magnitude of the contrast effect is dependent upon the amount 
of black or white in the immediate neighborhood, on which basis 
T 2 should appear lighter than Ti. Actually TI appears brighter 
than T 2 (see Figure 2oa). 

It is characteristic of this presentation that even psychologic- 
ally-trained observers fall into the error of maintaining that 
since Ti lies within the black region it must have more black 
surrounding it until a little geometrical reflection shows that 
there is considerably more white in its environment than in T 2 's. 
The extent of the black area contiguous with the triangle in the 
corner is substantially greater than the black area adjoining the 
other triangle. This phenomenon is not an artifact of a peculiar 

30 "Beobachtungen zu einem Experiment iiber Helligkeitskontrast," Psychot. Forsch., 
1924, 5, 131-142. 



120 GESTALT PSYCHOLOGY 

example, since analogous effects are obtained with a large I and 
H. Moreover, reversal of the direction of contrast by means of 
white figures on black ground produced a corresponding result : 
TI now appeared darker than T 2 . 

In order to quantify his data, Benary inserted his critical 
grey fields of 180 black and 180 white in black figures on 
white ground in the cross-triangle comparison (Figure 2oa 
matched with Figure 2ob). 

Seven subjects all saw T x as brighter than T 2 ; the brightness- 
difference, Ti-T 2 , required on the average an additional 3O-4O 
black in T I to compensate or neutralize the figural stresses pro- 
ducing It. 





(a) 



Figure 20. Benary's Figures Demonstrating the Influence of "Form" upon 
Brightness Properties 

The Tj homogeneous grey patches above are all seen as brighter than the correspond- 
ing To sections, presumably because the first type clash with the "closure" tendencies of 
the figures upon which they are laid. 

In the experiment above, T x always belonged to the figure, 
but To did not. Is it just this relation to the figure as distinct 
from the ground which is responsible for the occurrence of the 
contrast phenomena described? In this neat configuration 
(Figure 200), the grey wedge in the black, T 1? corresponds to 
the critical field in the preceding triangle, (Figure 2ob), while 
the wedge upon the white, T 2 , corresponds to the critical field in 
the cross. Normal observation makes T x seem brighter than 
T 2 . It is possible for experienced observers to respond simul- 
taneously to this configuration as two equivalent figures, one 
black and one white, with the included critical fields. But even 
under these conditions of apprehension, TI is brighter than T 2 , 
suggesting that the brightness gradient is independent of the 
surface to which it "belongs," and that the division of the criti- 
cal fields upon the figure and ground is not the condition sine 



PHENOMENA OF VISUAL PERCEPTION 121 

qua non of the effect. Nevertheless, inclusion within a gestalt 
seems to be a more potent cause of brightness contrast than the 
kind, mass, and proximity of the inducing surface. 31 

Figure 21 prepared by Matthaei 32 is useful for many pur- 
poses, but especially good as a clear-cut instance of non-summa- 
tive brightness contrast. The completely enclosed upper part of 
the "R" appears greyer than the white area above the letters 
which are exposed on three sides. Summatively, the upper field 
of the "R" should look brighter, but the central stripe is more 
strongly elevated, being a characteristically new structure. Even 
in those cases where the borders of the 
central stripe lie within the interior of a T> ^ ** I"O1^ 

1 ,, ,1 1C yj J -11 **-* **~> M.JLJLJLJ A- VJLA 3 

letter as in the o and more particularly 

in the second "n," the stripe as such is Figure 21. The Central 

1 1 , rr Stripe is the Brightest 

sharply set off. Sin * le Region in th * ^ 

A similar condition prevails in an ure - (After Matthaei) 
interesting experiment of Helene 

Frank's 3S dealing with the effect of the pattern of the pro- 
jection surface upon the behavior of after-images. If an 
after-image of a circle is cast upon a screen containing a draw- 
ing in perspective of a long corridor, the disc appears larger if 
projected upon the apparently remoter end and definitely smaller 
at the subjectively nearer end. A more dramatic effect is se- 
cured when the after-image of a plus ( + ) sign is projected 
orthogonally upon Rubin's cross pattern (cf. Figure 4, which 
appears alternately as an iron cross or a propeller) ; when the 
iron cross emerges as "figure," the after-image of the plus sign 

31 Mikesell and Bentley, in an elaborate repetition and more accurate extension of 
Benary's technique, confirmed his finding that the traditional laws of contrast need to be 
supplemented by other factors. They contend, however, that he confused configurations 
with geometrical or symbolic designs, and that the changed quality of the incorporated gray 
in mutilated patterns does not wholly support the Gestalt version. See "Configuration and 
Brightness Contrast," /. Exper. PsychoL, 1930, 13, 1-23. 

Jenkins has also shown that central and not peripheral factors are responsible for 
the changes of relative brightness of grey patches included in designs. "Inherence" must 
be added to Pragnanz as one of the fundamental laws of perceptual dynamics. Cf. "The 
Perception of Plane Designs," ibid., pp. 24-46. 

For a complete confirmation of Benary's work, see Atwater, "The Effect of Form on 
Color Contrast," 7. Gen, PsychoL, 1933* 8, 472-478. 

32 "Das Gestaltproblem," Munchen, 1929, p. 60. Originally published in Ergebnisse 
der Physiologic, 1929, 29, 1-82. 

83 "Ueber die Beeinflussung von ISTachbildern durch die Gestalteigenschafteti der 
Projektionsflache," PsychoL Forsch., 1923, 4, 33-37. About the same time, Rotschild had 
found that slight irregularities in the contours of the original were eliminated in the after- 
image in accordance with the law of Pr'dqnans. This is of some importance for a theory 
of memory as will be seen later. Cf. "IJeber den Einfluss der Gestalt auf das negative 
Nachbild ruhender visueller Figuren," Archiv. /. Ophthalmol., 1923, uz, 1-28. 





122 GESTALT PSYCHOLOGY 

is clearly observed, but as soon as the propeller becomes focal it 
all but disappears; however, with the restoration of the "iron- 
cross attitude" the plus image becomes clear again. 

A check upon this unique observation is found in the results 
obtained with the after-image of an eight-pointed star. 
When the iron cross is isolated only the plus ( + ) rays 
are clearly seen while the cross arms ( X ) appear dimly 
in the background. 

Colored after-sensations follows the same rule, as 
was shown by using the hollow cross herein as background. 
The after-image of this original was thrown 
upon it so that two alternating percepts were 
possible one, a multiplication symbol and the 
other, two adjacent angles. When the "times'* 
sign emerges as "figure" retinal color mixture 
apparently occurs, but in the double-angle re- 
sponse the definite bi-coloration 
is preserved. In general, these illustrations sub- 
stantiatethe claim that the visibility or clearness 
of the after-image is in part dependent upon 
Its position within either figure or ground. 34 

In a series of experiments by Fuchs 35 some highly convincing 
evidence of the influence of formally-perceived combinations 
upon apparent color was accumulated. A typical illustration of 
this effect is afforded by the nine-spot figure (Figure 22) , which 
permits one to see either a yellow plus sign ( + ) or a blue-green 
cross ( X ) - The middle yellow-green circle can belong to either 
pattern. It is possible for experienced observers to see the mid- 
circle as belonging to the yellow plus-sign, in which case it be- 
comes yellow, too, despite the simultaneous apprehension of the 
blue-green gestalt. Conversely, should it belong to the cross- 

34 Another favorite illustration in this : Place a i-cm. thick gray ring about 8 cm. in 
diameter symmetrically upon a half -blue, half-yellow surface. One can see the ring as a 
whole or as two half-rings, but in the first case it seems homogeneously gray and in the 
second case, dark and blue upon the yellow ground, and bright and yellowish upon the 
blue; but it is impossible to see two equally grey half -rings. Here, too, the structure 
determines its own color. A thread separating diametrically the halves facilitates the ^ effect. 

An arrangement traceable to Benussi consists in placing twenty grey discs in a circular 
pattern upon an equally apportioned red and green field. With appropriate cognizance of 
the circle gestalt the contrast effect entirely disappears. 

35 "Experimentelle Untersuchungen uber die Aenderung von Farben unter dem 
Einfluss von Gestalten," Zsch. f. PsychoL, 1923, 92 f 249325. 




PHENOMENA OF VISUAL PERCEPTION 123 

sign, the central disc becomes bluish-green like the other circles 
of that figure, although the yellow "square" is perceived at the 
same time. This double isolation of two figures, only one of 
which "contains" the middle circle is a neat procedure which dis- 
poses of the possibility that attention and not the configured per- 
cept is the cause of an adaptation in hue. Wundt had made a 
similar observation long before, but his interpretative bias led 
him to offer an explanation in terms of "associative effects" in- 
stead of a more vivid psychophysical dynamism. 

o KEY 

Blue-Green 
Yellow .-.. 

Yellow-Green- 
O Q 

Figure 22. Schema Illustrating the Influence of "Form" upon Color Quality 

The middle circle looks yellow if perceived as part of a plus-sign comprised mainly 
of the other yellow discs and more greenish if seen as part of an X primarily determined 
by the corner items. 

Kohler himself has reported a simple observation which 
strengthens the view that such fundamental qualities of visual 
objects as brightness and color are properties of an extended 
field and depend upon more than local uniform stimulation. If 
we take a glass of water in which soap is dissolved, the aspect of 
such a liquid is defined by such adjectives as dim or turbid. Now 
if one looks at it through a little hole made in a piece of card- 
board, one will see the hole filled with a certain grey hue (pos- 
sibly a bit bluish or reddish), but the quality of "dimness" or 
"turbidness" will have disappeared. Why? Fundamentally, 
because the properties of parts are not the properties of wholes. 

Movement as Conditioned by the Gestalt 

Duncker's dissertation gives some of the best testimony in 
favor of the inevitability of a Gestalt explanation of perceptual 



124 GESTALT PSYCHOLOGY 

experience, since it deals with phenomena with which every lay- 
man is familiar. There is a certain type of "induced" or appar- 
ent and illusory movement which occurs only as the result of 
simultaneously present and objectively "real" motion. It ap- 
pears in its commonest form with two neighboring trains in a 
railroad station one's own car seems to be moving when 
actually the adjoining one is in motion. Significantly enough, 
this illusion is all the more pronounced the less one sees of the 
stable or resting environment (= ground?) of the station. This 
explains the impressive effect of a reversal of direction when 
one's own coach is overtaken and passed by another train at 
night on an open stretch, since the normal "field" upon which 
motion is projected is absent. 

There are other experiences which show that this phenome- 
non of relative or contrast movement demands at least two 
stimuli one objectively at rest and the other in real motion. 36 
If one stands upon a bridge and focuses upon a point in the 
stream, one seems ultimately to be moving with the bridge in a 
direction opposed to the flow of the water; again, an experi- 
ence which is the more compulsive the less one sees of the firmly- 
anchored shores peripherally. The "wandering" moon when 
the clouds sweep by, the "floating" island in a rapidly moving 
stream, and the "falling" tower in a shifting fog are other per- 
ceptual effects conditioned by similar configurations. In all 
these cases, the common feature appears to be that phenomenal 
movement is essentially displacement in a natural system of re- 
lations. 

One of Duncker's experiments revealed an important rule 
with respect to the production of induced movement, viz., the 
induction is solely a one-way affair from the larger to the 
smaller object and not reversible. A small light point was pro- 
jected perpendicularly upon a cardboard surface attached to a 
moving band and fixated at a distance of one meter. When the 
carton moved, the objectively resting point also moved in an 
opposite direction and with an apparent velocity directly propor- 
tional to that of a real movement. This law was checked by 

39 "Ueber induzierte Bewegung," Psychol. Forsch., 1929, 12, 180-239. 




PHENOMENA OF VISUAL PERCEPTION 125 

an arrangement in which a point in a rectangle 
could be either moved while the rectangle was 
fixed, or kept at rest while the rectangle moved. 
If the actually resting object was fixated in either 
instance, the enclosed object appeared to have the 
stronger movement, i.e., the figure is "livelier" 
than the ground. Oddly enough, an induced motion of the 
observer himself occurs if the distance from the fixation point 
is decreased to 30 cm. ; seemingly the body of the observer be- 
comes an annex to the induced motion of the point. 

The appeal to "field forces" as the major determinants of or- 
ganized wholes is cleverly illustrated in Koff ka's persuasive ref- 




Figure 23. Koffka's and Wittmann's Figures Modified from Linke's 

Originals 

If the black ball is rapidly projected in different positions within the arcs, it 
appears to slide back and forth in contact with the circular rim. 

utation 37 of Linke's empirical account of stroboscopic illusions. 
Linke prepared four identical semi-circles with the concave 
side turned upward; on the inner side of each and in contact 
with the rim a small black point was drawn in the first semi- 
circle toward the left, in the third toward the right, and in the 
other two below in the center. If one exposes the figures in the 
appropriate order, a definite impression of a ball rolling within 
the perisphere results. However, if the points are presented 
without the semi-circles (which apparently serve as a runway) 
the rolling movement disappears and the points simply hop 
from one position to another. Linke considered this evidence 
of the assimilating effect of past influences upon present experi- 
ences, since we have observed moving objects slide back and 

37 "Kleine Mitteilungen aus dem psychologischen Institut der Universitat Giessen," 
Psychol. Forsch., 1922, 2, 144-155. 



126 GESTALT PSYCHOLOGY 

forth within vessels. Koffka nicely disposes of this explanation 
by simply inverting the figures so that the concave sides of the 
semi-circles are turned downward. Oddly enough, the point 
still clings to the periphery during its apparent movement under 
these conditions and describes a rolling motion along an arc 
rather than a jump across a chord ! It slides up and down as 
though on the ceiling of a vault in a manner which defies all 
prior experience. Wittmann supports Koffka by his claim that 
a similar effect is indifferently obtained in other positions such 
as indicated by Figure 23. The most satisfactory explanation 
seems to be that the arcs by their "passive" presence exert some 



Figure 24. Ternus' Method of > Demonstrating Phenomenal Identity Despite 
Objective Difference 

Dots represent light points first exposed, circles those immediately following. Under 
tfcese conditions the new points appear to be the old ones slightly displaced. 

direct influence upon the adjacent curve of motion which is inde- 
pendent of individual experience. 

A pupil of Wertheimer's, Ternus, 38 performed a number of 
simple but original experiments on the problem of phenomenal 
identity. This is a matter of some epistemological import since 
it is logically rather disturbing that two different points of light 
successively exposed should appear like the same point moving. 
By means of point patterns made by thrusting pin-holes into 
cardboard and a simple lever device which could be pushed in 
and out or up and down, thus exposing or screening certain 
points, Ternus was able to demonstrate that phenomenal identity 
is preserved by the whole as such and not just by relations be- 
tween the parts. For example, in Figure 24 the dots indicate 

38 "Experimentelle Untersuchung iiber phanomenale Identitat/' Psych&L Forsch 
1920, 7j 81136. *' 



PHENOMENA OF VISUAL PERCEPTION 127 

the first pattern briefly exposed and the circles the second one. 
If one moves the screening lever slightly, some of the original 
points (represented by dots) disappear and others (typified by 
the circles) appear. The subject, however, does not see some 
points go and others come; instead, he sees a general upward 
movement of the entire triangle. If old parts vanish and new 
ones assume their relative position, despite actual displacement, 
the integrity of the whole is preserved. A dynamic gestalt in- 
volving movement must be accounted for in terms of total field- 
forces just as much as a static gestalt (i.e., stable geometrical 
forms and figures). 

A wealth of minor problems has been found in the experi- 
mental treatment of movement phenomena ever since Wer- 
theimer scratched open a new vein. Lindemann 39 is responsible 
for a novel attack on the phenomenon which Kenkel in 1913 had 
termed the gamma-movement. This refers to a slight intrinsic 
apparent motion which appears in an object when tachistoscopic- 
ally exposed. The evident expansion of an electric light bulb 
when the current is suddenly switched on is a special case of this 
effect; similarly, when the object disappears there is a definite 
contraction observable with the movement of the contours di- 
rected toward the center. 

Working with very short exposures, Lindemann found that 
various geometrical figures presented in different ways behaved 
differently. In the case of a circle, the gamma movement was 
most pronounced along the horizontal direction, and the same 
held for an ellipse whether resting upon its long or short axis. 
A square resting upon one of its sides as a base also showed its 
maximum motion laterally, but when standing upon a point all 
four corners moved energetically in and out ! To Lindemann, 
these observations suggest that the experiences of form and 
movement must be intimately related, a conclusion which is 
supported by the fact that an irregular aggregate of dots fails 
to show the phenomenon plainly until patterns of some kind 
emerge. The gamma movement is an accompaniment of the 

89 "Experinientelle Untersuchungen fiber das Entstelien und Vergehen von Gestalten," 
Psychol. Forsch., 1922, 2, 5-60. Cf. also the confirmation of Feinberg in "Experimentelle 
Untersuchungen liber die Wahrnehrnting 1 im Gebiete des blinden Flecks," ibid., pp. 16-43. 



128 GESTALT PSYCHOLOGY 

formation of a gestalt. The stationary resting figure is merely 
an end-state of a dynamic process, but it is all that concerns the 
organism practically. Biologically, this neglect of the prelim- 
inary function is of service because a creature's orientation in 
the visual world would be much disturbed by the unceasing dila- 
tion and contraction of the objects in his view ! 

Constancy Effects in Vision 

Gestalt writers have had more than average interest in the 
phenomena of perceptual "constancy' 1 because their theory 
seems to offer the most reasonable account of these unusual ef- 
fects. Perceptual constancy refers to the fact the relative size, 
color, and brightness or other qualities of an object seem to be 
maintained under conditions where it must have different phys- 
ical properties. Hering long ago was puzzled to know what 
mechanism was responsible for the fact that snow and coal in 
full sunlight and at dusk preserved their apparent blackness and 
whiteness, even though the absolute brightness of snow at twi- 
light was less than that reflected by coal at high noon ! One can 
place a grey surface beside a deep black one in such a way that 
one can see the grey as brighter (despite its presence in diffuse 
dim roomlight) than the black surface, which, when locally 
illuminated by a powerful lamp must necessarily be brighter, as 
can be shown by looking at a small section of both surfaces 
through a darkened tube. An explanation in terms of experi- 
ence, such as suggested by the hypothesis of "memory-colors" 
clashes with our knowledge that the attributes of an inner field 
are functions of the qualities of the surroundings. The sus- 
picion is not unwarranted that cases of { "constancy" represent 
the determination of part-properties by a larger total constella- 
tion, i.e., they are conditioned by the cooperation of the environ- 
ment with the object. 

Frank 40 studied the problem of visual constancy of magni- 
tude in young children, who are desirable subjects for this type 
of investigation, because the empirist holds that this phenome- 
non is a matter of acquisition. Kohler had shown in his Tenerif e 

40 "Untersuchtmg fiber Sehgrossenkonstanz bei Kindern," Psychol. Forsch., 1926, 7, 
137-145- 



PHENOMENA OF VISUAL PERCEPTION 129 

monographs that young chimpanzees behave in accordance with 
constancy principles, and Frank simply adapted his technique 
to her subjects. Thirty children, ranging from eleven months 
to seven years of age, were trained to go to the larger or smaller 
(depending upon the group) of two boxes: correct choices 
contained chocolate, fruit, or a toy hidden beneath. In the 
critical test-situations which followed, the boxes were so placed 
that the remote larger box cast a much smaller image upon the 
retina than the nearer small one. In some instances, the retinal 
image of the larger box had only i % of the areal size of the 
smaller. Nevertheless, 23 out of 30 children made no errors on 
this basis at all and four made but one. Even the little young- 
sters, who were unable to walk, crawled a good distance to the 
objectively correct box. With six of the children, a transposi- 
tion experiment modelled after Kohler's work on brightness dis- 
crimination In hens was undertaken. The subjects were trained 
to go to the larger of two boxes, and in the crucial test a still 
larger one was introduced. In all instances the results were 
positive, the choices taking place according to the principle of 
relative size. 

Voigt 41 found that the accuracy of target-aiming is condi- 
tioned by a number of special configurational factors. He had 
his subjects shoot at a milk-glass plate target with a "light pis- 
tol," i.e., an instrument for projecting a circular beam of light 
upon a distant goal. The subjects aimed at the objective under 
the direct guidance of the motor system without optical control 
since no fine sighting was permitted. Contrary to ordinary ex- 
pectations, the angle of error diminished with distance : thus, at 
2 meters the mean angular error was 170'; at 4 meters, 158'; 
at 6 meters, 109'; and at 8 meters, 101'. 

The same decrease in angular error occurred when retinal 
constancy of the target was achieved, i.e., a target 15 cm. in 
diameter at 3 meters gave a mean error of 104' while its retinal 
equivalent of 1 50 cm. at 30 meters showed an error of only 49'. 
With the same objective target size at these distances, the errors 
were 105' and 60' respectively, values directly comparable in 

41 "Ueber die Ridittrngsprazision einer FemBandlttag," Psychol. Forsck., 1931, 16* 
70-113. 



130 GESTALT PSYCHOLOGY 

magnitude. Evidently, this factor could not be the cause of the 
reduction. 

Believing the crux of the situation to lie in some alteration in 
the optical structure of the field of action, Voigt set a box di- 
rectly adjacent to the target on either the right or left edges. 
The resulting shots were displaced toward the side with which 
the box was in contact. The errors now clustered around a new 
center of gravity. Voigt explains his paradoxical reduction in 
relative error with increasing distance in terms of phenomenal 
accentuation of the distance element. In the case of the nearby 
target, the internal diameter of the target is emphasized in the 
subject's experience; with the far target, on the other hand, the 
gap between the marksman and the target is more prominent. 
The more sharply and strongly this gap is perceived, the greater 
the accuracy of shooting. 

Another related problem was studied by Brown. 42 If I ob- 
serve a specific objective speed at a distance of two and four 
meters respectively, then as far as the retinal projections of the 
moving contours are concerned, unequal retinal extents are tra- 
versed in equal times. The objective distance covered in one 
second is twice as great in the first case as in the second, so that 
the corresponding "retinal velocity" must also be in the ratio of 
two to one. Nevertheless, no such discrepancy is perceived (as 
we all know from watching automobile traffic) and it would 
seem as though there were a constancy of velocity analogous to 
the constancy of visual magnitude, of which it may, as a matter 
of fact, be but a special case. 

In Brown's experiment on this topic, he had his subjects ad- 
just the speed of a comparison motor driving an endless paper- 
band with a dot and other simple geometrical figures with that 
of a similar standard : the two moving fields were successively 
compared by being placed at right angles to each other. De- 
spite the fact that the distance of the comparison field from the 
observer varied from three to ten meters, its adjusted velocity 
remained approximately constant, i.e., the ratio of the two 
speeds at different distances approached unity. More important 



gesehene Geschwindigkeiten," Psycho!. Forsch., 1928, io t 84-106; also a 
series of three English papers on the same theme in the same journal, 1931, 14, 183-268. 



PHENOMENA OF VISUAL PERCEPTION 131 

than this was Brown's discovery that if a moving field in a 
homogeneous surrounding field is transposed in its linear rela- 
tions as 1 :2 the stimulus velocity must be transposed by a like 
amount in order that the phenomenal velocity in both cases be 
identical; e.g., if the "constant" figures are twice as large as the 
"comparison" ones and if the former are also twice as far 
away from the observer as the latter, the constant speed of the 
first divided by the mean of the adjusted speeds of the second 
approaches two (1.92 in the report). This indicates that phe- 
nomenal velocities (or movements) are determined in a dynam- 
ical field, the essential nature of which cannot be described as a 
sum of independent local events. They correspond to dependent 
events in the functional whole. "Therefore, the whole structure 
of the excited field, not the excitation present at any given point 
within the field, must be considered in order that one under- 
stand the physiology of the visual perception of velocity." We 
can now appreciate why there is little difference between saying, 
"It moves quickly" and "It takes a short time to get there." 
Whatever structural change alters phenomenal velocity affects 
phenomenal time by a corresponding amount. The relativity of 
Einstein and the configurationism of Wertheimer converge at 
this point. 

The uniqueness of the Gestalt explanation of all types of 
"constancy" is clearly seen in its treatment of the brightness 
variety. Is it really an illusion if a white tablecloth remains 
phenomenally bright under yellow illumination? Does our 
judgment play a trick upon us in this and similar cases ? Hardly ; 
for it would be much more of an illusion if we saw the white 
tablecloth by lamplight as yellow, or the man at the other end 
of the room as a dwarf ! Pragnanz affects shape, size, and sur- 
face attributes so that the most significant aspects of an object 
are preserved if the field conditions are not altered too severely; 
i.e., a "thing" is invariant or transposable just like a melody! 



CHAPTER 7 
STUDIES IN AUDITION AND THE SKIN SENSES 

Experiments in the Perception o Sound 

Exhaustive investigations in the field of sound had been con- 
tinued long enough in Stumpf 's day to establish a tradition of 
specialization for the Berlin laboratory, and it was only natural 
that when the Gestalt theorists captured this stronghold that 
they should turn some of their energies in this direction. One 
of the first fruits of this union of a new theory and excellent, 
if old, material equipment was a paper by Eberhardt 1 on the 
phenomenal pitch and intensity of partial tones. Using Lewin's 
mirror-membrane apparatus for measuring tonal intensities and 
resonators for "hearing out" partials, she sought to determine 
whether a partial picked out of a clang could be compared in 
intensity with a pure individual tone of the same character. 
When successive pairs were thus presented and judged under 
conditions of objectively equal strength, in no case did the in- 
tensity of the resonator partial exceed 90% of its intensity when 
isolated, and it often dropped as low as I % \ Apparently the 
sound context makes any single tone lose in intensity to an 
extraordinary degree. In addition, Eberhardt was able to make 
the important demonstration that a partial tone which was too 
feeble to be heard through the resonator was nevertheless able 
to influence the timbre of the clang. Any summative theory 
was bound to be weakened by the discovery of these incon- 
testable facts. 

Onoshima 2 performed a distinctive acoustic experiment con- 
cerning the influence of temporally adjoining tonal groups upon 
the judgment of intensity steps. Using an electrically-controlled 

1 "Ueber die pMnomenale Hohe und Starke von Teiltonen," Psychol. Forsch., 1022, 
*, 346-367. 

~ "Ueber die Abhangigkeit akttstlcher Intensitatsschritte von eineni tmfassenden 
Tonverband," Psychol. Forsch,, 1928, u t 267-289. 

132 



AUDITION AND THE SKIN SENSES " 133 

hammer, he produced sounds of nine degrees of intensity and 
presented them to his subjects in symmetrical double pairs sepa- 
rated by a pause, thus fore pair main fair. Two curious phe- 
nomena were observed. If the interval between the fore pair 
and the main pair was a step in the same direction (up or down) 
as that represented by the two tones in the fore pair or only 
weakly reversed, then an adaptation to this trend occurred 
within the two tones of the main pair, i.e., they also seemed to 
move up or down the intensity scale ; but if the interval between 
the two pairs was sharply in the counter-direction to that indi- 
cated by the gradient between the two members of the first pair, 
then an opposed movement of the intensities of the main pair 
occurred. The first effect, however, was the more readily ob- 
tained with symmetrical pairs, irrespective of the times elapsing 
within or between the pairs; but the second effect was com- 
moner with unsymmetrical time relations. This principle, like 
so many others of the Gestalt school, is a general sensory prop- 
erty and not peculiar to any modality, since Onoshima obtained 
analogous results with successive weight pairs in a supplemen- 
tary study. 

Kohler 3 has offered an interesting new version of the process 
of successive comparison and the time error which usually ac- 
companies it. 4 The familiar phenomenon to which this refers 
was first noted in the early days of Wundt's laboratory in con- 
nection with sound intensities and pressures. According to these 
results, a slight difference in mass between two successively lifted 
weights is more readily discriminated if one first "lifts" the 
objectively lighter one and then the heavier (ascending step) 
than in the reversed case (descending order). Borak 5 found 

3 "Zur Theorie des Sukzessiwergleichs und der Zeitfehler," Psychol. Forsch., 1923, 
4, 115-175. Lauenstein has^ extended this theory by showing that the same potential dif- 
ference found between two immediately adjoining- fields is not only the basis for the per- 
ception of contours but also necessary for the impression of relations. See his "Ansatz zu 
einer physiologischen Theorie des Vergleichs^ und der Zeitfehler," ibid., 1933* 17, 130-177- 

* The time error normally occurs only in studies dealing with difference limens. It 
is perhaps significant that alcohol lowers me absolute and raises the differential threshold, 
a fact which suggests a functional distinction between the two structures the one, a figure 
against a ground, and the other, a part against another part of a figure. See Specht, Arch, 
f. d. Ges. Psychol., 1907, 9, 180-295. 

5 "Ueber die Empfindlichkeit fur Gewichtsunterschiede bei abnehmender Reizstarke," 
Psychol. Forsch., 1922, 374-389. The main outlines of Kohler's theory are already implied 
among the possible explanations of the phenomena considered in Borak's paper. Any ex- 
perience appears to the percipient as either a continuation of some previous one, on the 
same level, or as a rise or fall from a level, or as a change in figure within a given ^level. 
The reactions are not to "simple" stimuli, but rather to situations which represent step- 



I 3 4 GESTALT PSYCHOLOGY 

that for any given pair the comparison in the ascending order 
always yielded more correct judgments (and consequently, a 
lower limen) than the other arrangement, and that even small 
descending steps were often interpreted not only as equal but 
as ascending! If this asymmetry has general neural significance 
and is not due to peripheral peculiarities, then Kohler argues 
the phenomenon should be found in all successive perceptions of 
intensities. He rejects a plausible interpretation which claims 
that all successive comparison is really a simultaneous compari- 
son between a present second impression and a memory image 
of the first response. This explanation in terms of transitional 
experiences Kohler maintains is as inadequate as accounting for 
the essence of noon because it is then that the clock strikes 
twelve. 

Neither is he willing to rest content with the task of descrip- 
tion at the expense of explanatory efforts as certain pragmatic 
specialists would have us do. While some may protest against 
the spinning of intricate brain theories ad libitum because they 
are referred to a realm where facts elude us, nevertheless it is 
impossible to preserve all our hypotheses on a purely psychologi- 
cal level. The neural speculations which result from a search 
for the raison d'etre of experimental findings are not really 
fantastic constructions, but if properly developed lead to definite 
consequences which can be tested by further investigation. Or- 
dinarily, psychological theories have been so timid and indefinite 
that like the planks of a political platform one can read well- 
nigh everything into them ! 

Kohler's specific brain theory is built upon the biological fact 
that all sensory stimulation makes the excited area of the nervous 
system electronegative with respect to its environment. During 
this process positive hydrogen ions are liberated. Now it is 
probable that the stronger stimulus frees the ions in a greater 
concentration, producing an increase in the acid reaction and 
potential difference as contrasted with that of the environment. 

wise" or differential aspects fundamentally related to figure and ground. Lauenstein (see 
footnote 3 above) found a negative time error when his paired stimuli were presented 
against a dark ground (in Hornbostel's sense) and a positive time error if light ground 
was used (relative to the brightness of the "figures"). This fact is in accordance with 
Eauenstein's theory that the "trace" of he first figural stimulus gradually accommodates 
Itself to the nature of the ground, hence the interesting reversal of effect which can be 
demonstrated with both light and sound. 



AUDITION AND THE SKIN SENSES 135 

This simply means that the effects of the stimulus outlast its 
actual duration. If the phenomenon of the "time error " is 
caused by the dynamic transition from a slowly sinking "still 
picture" of the first excitation (reminiscent of Herbart's "resi- 
dues" but devoid of their mechanical connotation) to a fresh 
second excitation, with the result that the gradient "upward" 
is stronger than the one "downward," then the time error must 
increase in proportion to the length of the time-interval as long 
as the "still picture" continues to sink. In other words, the 
relation of the judgments "higher" and "lower" is a function 
of the time-interval between the stimuli, a position which is con- 
firmed by the following table with objectively equal auditory 
pairs (Kohler, page 152) : 

TABLE III. JUDGMENTS OF RELATIVE SOUND INTENSITIES OF PAIRED 

STIMULI WITH VARYING TIMES ELAPSING BETWEEN THE MEMBERS OF 

THE PAIRS. EIGHT SUBJECTS. (AFTER KOHLER.) 

[Italicized figures in parentheses represent data obtained from 12 subjects giving ten 
judgments each in response to repeated pipe-organ notes (from unpublished study by 

G. W. Hartmann).] 

Reports Intervals 







i* 


'2(2) 


3(4) 


4 1 / 


W) 


6(8) seconds 


Rising 
Equal 


(second-stronger) . . 


i 
8 


(43) 


7(62) 

sC?*) 


*3 

=1 


(71) 
(24) 


7(l6) 


Falling 


(second- weaker) . . 


15 


(*5) 


12(26) 


6 


(25) 


2(16) 


Total 




24 


(120) 


24(1,20) 


24 


(120) 


2 A, ( 120} 



















Evidently the longer the gap between the two members to be 
compared, the more "intense" the second one appears to become, 
a fact known as the "negative" time error. Why? Presum- 
ably because the first stimulus changes in some way the electro- 
chemical "level" of the nervous system so that the second stimu- 
lus is judged with reference to a wholly different "plane" of 
sensitivity a plane which is constantly changing its altitude. 
Figuratively, the second stimulus leaps from a stool of varying 
height provided by the first stimulus, which when it appeared 
had to jump unaided from the floor. 

Any perceptual hypothesis in which the time element becomes 
a central factor inevitably becomes a theory of memory as well. 



136 GESTALT PSYCHOLOGY 

Kohler saw this clearly in the paper under discussion but re- 
frained from any immediate experimental extension of his posi- 
tion. Because of its peculiarly unifying character the point is 
important enough to be restated in Wheeler's 6 precise condensa- 
tion : 

When an external stimulus impinges upon the nervous system, the 
existing state of the system changes until it is in equilibrium with the 
force that is acting upon the sense organ. Let another stimulus affect 
another part of the nervous system and the first approach to an equili- 
brium will be altered by the change going on toward equilibrium with 
the second stimulus, and vice versa. In this way, the responses to the 
two stimuli are interdependent and the final outcome depends upon the 
dynamic relation between the two levels of equilibrium. Second, it is 
supposed that the approach to equilibrium which commences upon stimu- 
lation so alters the concentration of reacting substances in the brain 
that a process of adjustment must continue after the stimulus is re- 
moved. This process of adjustment may determine for a considerable 
time the direction of changes elicited by subsequent stimuli. But this is 
after the "conscious" process associated with the first stimulus has dis- 
appeared; accordingly, the continuing brain process is called a "non- 
process condition/ 5 because it is not associated with conscious processes. 
The ''non-process condition 17 is Kohler's substitute for the older concept 
of trace. 

Theoretical attempts to appraise the respective merits of the 
Gestalt position and that of behaviorism of the conditioned re- 
sponse type have been numerous, but relatively few psychologists 
have put both to the sort of ingenious test which Humphrey 7 
devised. His point of departure was the correct assumption that 
if Wertheimer's account be valid, then a musical note heard in 
isolation is a different thing from the same note heard in a 
musical melody. The technique consisted in drilling nine sub- 
jects to react with a withdrawal of the hand to a given note by 
means of an electric shock simultaneously applied. After "con- 
ditioning" had been established, the latency of the reaction was 
measured and found to be about 400 milliseconds, an interval 
about equal to the time required for a so-called "discriminative 

6 The Science of Psychology, New York, 1929, p. 276. 

7 "The Effect of Sequences of Indifferent Stimuli on a Reaction of the Conditioned 
Response Type," /. Abn. & Soc. PsychoL, 1927, 22, 194212. 

In connection with Humphrey's point, it may be worth recalling that St. Catherine of 
Genoa at one time of her life vomited all food except that taken as part of the Eucharist 1 



AUDITION AND THE SKIN SENSES 137 

response." Humphrey's significant discovery was that when 
presented in isolation, practically perfect responses were ob- 
tained with a given note; but when played in a melody it was in 
no case followed by a response. He states, "We may then fairly 
claim to have proved that the inclusion of a tone within a musical 
unity, such as a melody, scale, or arpeggio (Home, Sweet Home) 
may destroy the response habitually following the same note 
in isolation." 

A defender of the conditioned response principle might reply, 
however, that Pavlov could easily explain this in terms of the 
inhibiting effect caused by companion stimuli. But Humphrey 
showed that these companion stimuli must have a very special 
kind of organization before they can exert any inhibiting influ- 
ence, since a supplementary test demonstrated that a non-musical 
sequence failed to inhibit the response, while a musical one 
definitely did so. 

An even more decisive objection to the usual neurological 
interpretation of the conditioning process has been made by 
Gibson and his associates. 8 They wished to see if a conditioned 
response was really limited to a connection between a fixed and 
definite type of stimulation and a specific muscle movement or 
whether there is any transfer or spread of effect to other regions. 
Using the familiar shock-and-buzzer combination for condition- 
ing adult humans, they found that in 62% of the cases the 
establishment of a conditioned response of the right middle 
finger was accompanied by the formation of a similar condi- 
tioned response of the corresponding finger of the other hand. 
With several subjects the conditioned response readily trans- 
ferred to the index finger of the same hand. Apparently a more 
inclusive process than that ordinarily alleged to occur takes place 
when a narrow S-R "bond" is established. The hypothesis 
seems justified that the conditioned withdrawal to shock in- 
volves, or perhaps is in itself, a generalized habit of avoiding or 
withdrawing when the buzzer is heard, which may be evoked 
from another part of the body than that in which the response 
was learned. 

8 "Bilateral Transfer of the Conditioned Response In the Human Subject," J. Exp. 
PsychcL, 1932, 15, 416-421. 



138 GESTALT PSYCHOLOGY 

The Skin Senses 

Passing from the auditory field to the cutaneous, we see 
again how Gestalt principles have revivified this standard type 
of investigation. Rosenbloom 9 stimulated ten subjects on the 
same region of the palm five times with the following seven fig- 
ures circle, square, triangle, open circle, open square, right 
angle, open triangle. The instrument applied followed the prin- 
ciple of the cookie-cutter, with pressure effects equalized by a 
one-pound weight placed on the top of each figure. An identifi- 
cation of the character of the tactual stimulus was demanded. 
Table IV shows that the number of correct reports was greatest 
with the triangle and circle and least with the open square. It 
is interesting to observe that those "figures" which are psycho- 
logically (not geometrically !) the simplest are also the ones least 
subject to confusion with others. The closed patterns are more 
easily perceived than the open ones ; the square seems to be the 
weakest of the closed figures. A closure effect is noticeable, 
since the "open" figures are more often perceived as "closed" 
than the converse, with the open triangle showing the strongest 
tendency. 

TABLE IV. NUMBER OF TIMES CERTAIN FIGURES WERE REPORTED IN 
RESPONSE TO A GIVEN PATTERN STIMULUS (ROSENBLOOM). 

RESPONSES 

STIMULI Open Open Right Open 

Circle Square Triangle Circle Square Triangle Triangle 



Circle 


40 


o 





4 





O 


o 


Square 


o 


25 


3 


o 


2 


o 


o 


Triangle 








40 


o 


O 


O 


I 


Open Circle 





I 


O 


27 


I 


o 


o 


Open Square 
Right Triangle .. 
Open Triangle . . . 


o 





8 





4 
6 
32 


4 

2 
I 


20 

O 


I 

37 




I 



3 



For decades psychologists have been reared in the belief that 
the reduction of cutaneous sensitivity from a generalized sense 
of "touch" to the four discrete properties of pain, cold, warmth, 

Q "Configurational Perception of Tactual Stimuli," Amer. J. PsychoL, 1929, 41, 87-90. 



AUDITION AND THE SKIN SENSES 139 

and contact constituted one of the most conspicuous triumphs 
of the analytical method. The existence of a mosaic of tiny 
sensory areas in the skin has been unquestioned since the early 
work of Blix, Goldscheider, and von Frey. Nevertheless, care- 
ful researches made during the last decade have thrown doubt 
upon this convenient and plausible dogma of the punctate dis- 
tribution of separate receptor systems. No one denies the 
anatomical existence of special nerve fibers but their functional 
discreteness has recently been made a matter of debate. 

Waterston, 10 a British clinician, seems to have been the first 
to strengthen a latent suspicion that the assignment of limited 
functions to certain cell-bodies and corpuscles was purely hypo- 
thetical. He believes that the epithelium as a whole is the 
receptor organ for touch, for when a film of it is preserved in an 
experimental area the exposed surface is dry and a touch expe- 
rience can be elicited from it, but when more of the epithelium 
is removed and the surface is raw and moist, no tactile sensation 
can be elicited, and only pain is excited by the stimulus. Espe- 
cially significant is his discovery that the functions of given spots 
fluctuate considerably. 11 To quote: 

If the point of a fine metal probe cooled to the requisite degree is 
brought into contact with a cold spot, and the touch repeated again and 
again at intervals of about a second, at first the contact produces a 
clear and distinct sensation of cold. After a few contacts, however, 
the sensation of cold is no longer excited by the stimulus, but that of 
touch only. The cold sensation may disappear after two or three con- 
tacts, or only after some eleven or twelve, and it may disappear and then 
reappear once or twice and finally disappear. If now the stimulus point 
be moved slightly to one or other side, the sensation o cold is at once 
set up from a spot from which before no such sensation was elicited. 

It is too early to reach any conclusion as to the degree to 
which these facts support the Gestalt contention, but the pos- 
sibility of field conditions regulating the response of a given 
region has been much increased thereby. In fact, a novel experi- 
ment by Madlung 12 ends all doubt as to the reality of such 

10 Reports of the St. Andrews Institute for Clinical Research, Vol. 2, Oxford, 1924, 
123-132. 

11 Dallenbach's painstaking and exhaustive experiments on the temperature spots and 
end-organs shows that the localization varies with the method of "mapping" employed. 
See Amer. J. Psychol., 1927, 393, 402-427. 

12 "Ueber anschauliche und funktioneUe Nachbarschaft von Tastemdrucken, Psychol. 
Forsch,, 1934, 19, 193-236. 



140 GESTALT PSYCHOLOGY 

"field" effects. Von Frey had found that two adjacent contact 
spots are experienced as closer together when stimulated simul- 
taneously than when touched separately. More particularly, he 
had observed that an intense second stimulus tended to "suck 
up" a weaker predecessor, i.e., it exerted something like a gravi- 
tational pull toward itself. In this schema, ; J distance 2-3 
may be experienced as smaller than distance 1-2. 

Von Frey carried on his investigations with but one arm sur- 
face. Madlung used both arms laid parallel so that the fingers 
of one were near the elbow of the other ; this gives two possible 
positions, dependent upon the relative location of the right or 
left arm to each other. His contact apparatus made it possible 
to strike two points or one arm surface and a third on the other. 
Surprisingly enough, Von Frey's attraction effect was demon- 
strated to take place from one member to another "through the 
air/' The interspace between the two members is just as "rear* 
a part of the gravitational field as the surfaces of the members 
themselves. Madlung showed that the attraction diminishes 
with the physical and phenomenal distance of the stimuli ; where 
the physical or (anatomical) and phenomenal fields diverge, the 
latter is decisive in determining the character of the experience. 



CHAPTER 8 
THE UNITY OF THE SENSES 

It must be clear now that one of the most interesting conse- 
quences of the Gestalt position has been the renewal of experi- 
mental activity in the ancient and honorable fields of sensation 
and perception. Probably most laboratory specialists who be- 
lieve that a "brass instrument' ' psychology is the only kind worth 
developing appreciate the revivifying influence of configura- 
tionism which has swept through this division like a gust of 
fresh wind. Recently, the Gestalt theorists have been surprised 
to discover that the logic of their own thinking demanded an 
extension of the viewpoint to the problem of intersensory in- 
fluences. If local action within any receptor area is undoubt- 
edly affected by processes taking place in the total field, then 
it would appear plausible that the course of a sensory event in 
any modality would be conditioned by the character of the trans- 
formations simultaneously occurring in all the remaining recep- 
tors. Although the number of investigations motivated by this 
theme are still meagre, the prediction is not unwarranted that 
experiments of the intersensory type may grow with the same 
extraordinary rapidity that characterized the researches in mental 
tests during the last two decades. Certainly the present tendency 
of configurational experimentation seems to be toward extend- 
ing configurational principles to the various sense modalities, 
unifying as many different fields as possible, and reducing ap- 
parently diverse phenomena, like movement and fusion, spatial 
extent and movement, form and quality, to more simple under- 
lying configurational processes. 

Hornbostel 

Hornbostel is the German writer who has most clearly dis- 
cerned the fundamental import of this question. In a delightful 
article on "the unity of the senses," he has expounded this 

141 



142 GESTALT PSYCHOLOGY 

doctrine in the charming literary manner of an intuitive philoso- 
pher of the romantic era. 1 The tenor of his paper is apt to 
repel the a he-men" among contemporary psychologists, the very 
people who owe their own hard-headed techniques to another 
rhapsodic mystic, Fechner. One beautiful thing about the 
proponents of Gestalt theories is that their own thought processes 
exemplify perfectly the principle which they proclaim, viz., that 
all mental development proceeds from an original vague undif- 
ferentiated total (^mysticism?) to well-defined and sharply- 
bounded patterns (the "clear and simple ideas" of Locke). 
With this in mind we may listen more sympathetically to Horn- 
hostel's argument. 

His first appeal is to philology which suggests that sensory 
functions have not been fully differentiated (at least conceptu- 
ally) among primitive groups. Although we all know the sad 
fate which befell Grant Allen's hypothesis that the Homeric 
Greeks were color-blind because of the deficient color terminol- 
ogy of the early epics, still it is interesting to note that there is 
a negro tribe which has a special word for "seeing," but only 
a common term to designate "hearing, tasting, smelling, touch- 
ing." In French, sentir means to smell, touch, or feel in gen- 
eral. In modern German and English, "brightness" is a recog- 
nized attribute of both lights and sounds. When this relation- 
ship is mentioned, many persons maintain that sheer analogy 
has led to a transfer of meaning from the visual to the auditory 
realm, but as a matter of fact in Middle High German the sound 
connotation represented by hell (bright) was the exclusive one! 

Hornbostel's personal contribution to this problem consists 
in the development of intersensory equations upon which his 
claim for the existence of a common supra-sensory factor of 
"brightness" is based. 2 This is the characteristic shared by 
most high-pitched tones, "loud" colors, penetrating but pleasant 
odors, sharp "pointed" tactile stimuli as opposed to dull, blunt 
surfaces, etc. The strongest evidence in favor of its existence 




in Alexander and Marburg's Handbitch der Neurologie des Ohres, Vienna, 1923, 419-464. 

Is there any significance in the fact that most people pronounce "up" with a rising 

inflection and "down" with a falling (apart from the different vowel qualities of the words) ? 



THE UNITY OF THE SENSES 

is the curious triangular equation easily obtained from large 
groups of subjects. If one produces upon the color wheel greys 
resulting from five different proportions of black and white and 
presents with each a definite tone, there is considerable agree- 
ment with the interesting exception of color-blind observers 
that one and only one of the greys has a "brightness value" 
most like the sound. 3 Unanimity, of course, does not occur 
but even the most confirmed skeptic is surprised to find the 
" voting'' far more uniform than chance could ever allow. The 
really startling demonstration, however, appears when one finds 
that a given odor can similarly be equated with a specific grey, 
and that this can be indirectly checked by equating the odor with 
the tone originally used! Things equal to the same thing (in 
some respect) must be equal to each other (in the same respect). 
The situation may be clarified by this table and equation, in 
which the subscript designates the brightness feature : 

SENSE BRIGHT DARK 

Vibratory smooth rough 

Pressure hard soft 

Contact pointed blunt 

Kinesthetic light heavy 

Temperature cold warm 

Pain "sharp" "dull 1 ; 

Organic hunger satiation 

Gustatory sweet ? bitter 

Olfactory flowery? burnt? 

Visual "whitish" "blackish" 

Auditory "high" "low" 

Grey X b 
(41 White -|- 319 Black) 

/ \ 

Odor Y b = Tone Z b 
(Benzol) (220 d.v.) 

Consequently, there must be something sensory in nature 
which is not restricted to a single receptor. What has been 
popularly dubbed the "hearing" of the skin points in the same 
direction : with the fingertips one can detect which of two vibra- 

s Among the Berlin cognoscenti the story goes that when Hornbostel first had the idea 
m embryo, he telephoned his friend, the late Musikwissenschaftler, Otto Abraham, and 
inquired what note on the piano corresponded to the odor of violet When this extraordi- 
nary question was answered in the way that Hornbostel expected if the theory were sound, 
lie was almost beside himself with joy. 



144 GESTALT PSYCHOLOGY 

tlon frequencies is the ''brighter/' even if separated by a full 
"tone" ; similarly, an octave dichord is sensed as consonant by 
the skin in contrast to a seventh. To Hornbostel, this means 
that brightness is the most primitive general sensory quality, 
all other specific attributes being later differentiations. The 
uniqueness of the eye is color, of the ear, musical quality 
hence, being relatively new biological acquisitions they are most 
readily lost, as all clinical experience testifies. 

It is with this special meaning in mind that one may legiti- 
mately speak of a transcendental sense-perception. There is no 
easy way to illustrate this phenomenon in the absence of the 
appropriate technical terms, but Hornbostel claims that he has 
distinct conscious states, akin to moods, but more tangible, 
which are equally well represented by a day in the Black Forest, 
a picture by Schwind, the works of Moricke, or the seventy- 
third movement of Wolf's "Foot Journey. 7 ' The unity of the 
senses thereby becomes the fundamental reason for the unity 
of the arts. Perhaps Wagner dimly discerned this truth. At 
any rate, no one can allege that James stands alone in his en- 
deavor to rehabilitate the "vague" and restore it to its proper 
psychological rights ! Historically, it seems that the Viennese 
otologist, Urbantschitsch, made the first relevant observations 
in this field. James (Principles, II, p. 29), speaking of his 
work, says : 

All our sense-organs influence each other's sensations. The hue of 
patches of color so distant as not to be recognized was immediately per- 
ceived by Urbantschitsch' s patients when a tuning-fork was sounded 
close to the ear. . . . The acuity of vision was increased, so that letters 
too far off to be read could be read when the tuning-fork was heard . . . 
sounds which were on the limits of audibility became audible when lights 
of various colors were exhibited to the eye. Smell, taste, touch, sense 
of temperature, etc., were all found to fluctuate when lights were seen 
and sounds were heard. 

Possibly Weber's observation that a thaler laid on the skin 
of the forehead feels heavier when cold than warm represents 
an even earlier recognition of the existence of "convergent'' 
stimulation. Indeed, the existence of this entire problem is a 
good example of the need for re-discovering old facts because 



THE UNITY OF THE SENSES 145 

at the time of the original researches the appropriate theoretical 
background for interpreting and assimilating them was missing. 

If what we "see" is literally influenced by what we are hear- 
ing, touching, smelling, etc., it must be equally obvious that what 
we "hear" is similarly conditioned by what we see. A number 
of observations support this view. In 1669, the distinguished 
Copenhagen anatomist, Thomasius Bartholinus, announced that 
partially deaf individuals could hear better in the light than in 
the dark ; improved hearing occurred even by the dim light of 
a candle. Otologists have long noted that the auditory percep- 
tion of hard-of-hearing patients fluctuates with the weather it 
tends to be better on bright warm days and poorer on raw and 
cloudy days. More recently, the Moscow biophysicist, Lazaref f , 
has popularized the following simple demonstration: Strike 
a piano key with the pedal depressed and while the tone lasts, 
silently switch an electric light bulb on and off; one hears in the 
same rhythm the swelling and fading of the tonal intensity ! If 
the eyes are closed or covered, the effect fails to occur. The 
discrimination of both pitch and intensity differences also under- 
goes a slight but consistent improvement under conditions of 
brilliant total illumination. 

On the basis of these and other similar studies, the follow- 
ing tentative generalization has emerged : A simultaneous aux- 
iliary heteromodal stimulus augments the main stimulus when 
the former is in the background, but it acts in an inhibiting way 
when it is focal. As figure, the auxiliary stimulus diminishes 
sensitivity in accordance with conventional expectations of the 
results of distraction, but it increases it as ground (when the 
main stimulus assumes the role of "figure"). The generaliza- 
tion readily harmonizes with Heymans' law (known since 
1899), if we revise the latter to read: A more intense sensation 
(^figure) exerts an inhibitory action on a less intense one 
( =ground) , the effect being proportional to the strength of the 
inhibitor. 

Further Evidence of Intersensory Relations 

The experimental possibilities of this far-reaching doctrine 
have only slowly begun to be appreciated by the rank-and-file of 



146 GESTALT PSYCHOLOGY 

psychologists. Goldstein and Rosenthal-Veit, 4 who appear to 
have been among the first to follow out its implications in audi- 
tory localization under multi-sensory stimulation, found that 
a sound can be displaced toward the right from a subjective 
center by an eye movement toward the left, even though the 
head be kept stationary. More recently, the Russian scientist 
Kravkov 5 has unearthed new evidence of inter-sensory relation- 
ships which were formerly thought to be restricted to the oral 
cavity with its union of olfactory and gustatory elements. Using 
a telephone tied up with a low frequency cathode tube generator, 
Kravkov found that a tone of 2,100 vibrations per second in- 
creased the visual acuity of nine out of ten subjects tested with 
black figures on a white ground. This phenomenon was con- 
firmed by the present writer for both high and low auxiliary 
sound stimuli, but even more astounding was the fact that 
simultaneous olfactory, tactual, and pain stimuli appeared to 
have equivalent effects in temporarily improving visual acuity. 6 
Zietz, 7 in a paper on the reciprocal influences of tonal and 
color experiences reports that an auxiliary sound exerted a 
powerful influence upon the course of an after-image, causing 
it to flicker when the tone "vibrated" intermittently, and ex- 
tinguishing the image when it ceased. When the auxiliary tone 
had a vibration number of 200, the color became darker, warmer, 
softer, duller ; the contours also tended to vanish. 8 With a tone 
of 550, the same color became brighter, colder, clearer, harder; 
and the contours were sharper. Occasionally, under the influence 
of a high tone (1,100 vibrations) a rounded after-image took 
on a squarish form ! As far as the reversal of effect was con- 
cerned, it appeared that a tone sounded in a lighted room was 
judged higher than one acting In the dark. 

***Ueber akustische Legalisation und deren Beeinflussbarkeit dutch andere Sin- 
nesreizeS'PsycJwL Forsch., 1926, 8, 318-335. 

5 * < TJeber die Abhangigkeit der Sehscharfe vom Schallreiz," Arch. f. Ophthalmologie, 
1930, 124, 334-338. 

6 G. W. Hartmann, "Changes in Visual Acuity through Simultaneous Stimulation of 
Other Sense Organs," /. Exp. PsychoL, 1933, 16, 393-407; also "The Facilitating Effect 
of Strong General Illumination upon the Discrimination of Pitch and Intensity Differ- 
ences," ibid., 1934, 17, 813-822. 

7 "Gegenseitige Beeinmissui 



1931, I21 f 257-356. 

sveral yea 



;ung von Farb-und Tonerlebnissen/* Zeitsch. f. PsychoL, 



8 Several years earlier, Noll had called attention to the exquisite obedience which after- 
images yield to Gestalt laws;, e.g., the blurred contour of an after-image as it fades away 
is very easily drawn magnet-like within the sharply outlined boundaries of a simple gestalt 
should the image be projected somewhat eccentrically thereto. Cf. the rest of his inter- 
esting article on "Versuche iiber Nachbilder," PsychoL Forsch., 1926, 8 t 327. 



THE UNITY OF THE SENSES 147 

Von Schiller, 9 working in a similar situation involving "het- 
eromodal" stimulation, found that "pulsation" or "coarseness" 
may be considered along with Hornbostel's "brightness" as an- 
other universal intersensory moment. Using the threshold of 
fusion with black-white discs as his "base/' he found that beats 
from an auditory source strengthen and emphasize the flicker 
effect, while dichords in firsts and fifths (as well as the simulta- 
neous finger "feeling' 7 or exploration of smooth cardboard) 
minimize it. The prominence of the flicker increased as the 
beats rose from 4 to 12 per second, but at 20 a better fusion 
occurred. Although von Schiller finds ample evidence for mu- 
tual influences, the conditions most favorable for "inter sensory 
induction/ 7 as we may term the general phenomenon, are pro- 
vided when the "induced" experience is a sense-perception with 
an objective reference and the "inducing" event something 
affecting the entire subjective state of the organism; the 
former, moreover, will have a figure-character and the latter a 
ground-character where optimal results are obtained. 

A study of Usnadze 10 on the psychological basis for con- 
ferring certain names upon certain objects fits in very nicely 
with HornbosteFs view of a common sensory element in both 
auditory and visual experience. This author exposed six mean- 
ingless drawings to his subjects for about five seconds each with 
instructions to avoid any associations with familiar objects, 
and then had them select the most appropriate nonsense sound 
for each picture from a suggested list of 42 syllables. Each 
sound had an equal chance of being chosen, but as a matter of 
fact only 29 or 68% were selected, the remaining 32% simply 
not being considered. The degree of agreement among the 
subjects with respect to the matching of sound and picture was 
far greater than sheer probability would ever permit. Accord- 
ing to Usnadze, this implies that there must be some equivalence 
of impression among the two members to be thus united. The 
modified onomatopoetic theory of language formation of which 

9 "Das optische Verschmelzen in seiner Abhanglgkeit von hetercmodaler Reiztmg," 
Zeitsch. f. Psychol., 1932, 125, 24.9288. Schiller also found that motor performances 
were affected by sensory field conditions. In pouring water from one receptacle to another, 
i 6% less spilling occurred when high tones were given than when lower ones accompanied 
the task. See his **Wirkung des Umfeldes auf motorische Leistungen," Zsch. f. Psyckol., 
1934, 13.2, 83103. 

10 "Ein experimenteller Beitrag zum Problem der psychologischen Grvmdlagen cler 
Namengebungj PsychoL Forsch., 1924, 5, 2443. 



I 4 8 GESTALT PSYCHOLOGY 

this gives a hint is of more than passing interest : all vocal and 
verbal complexes possess a specific Gestaltqualitdt which fits 
them to become the designation of those objective visual forms 
which they most closely resemble in organization and emotional 
congraency. Conversely, if a given sound is to be pictorially 
represented, certain forms are recognized as fitting and others 
are not. 

Certain hitherto incomprehensible facts now find an under- 
standable niche in this portion of the Gestalt structure. Travis u 
noted that auditory acuity seemed to be lowered during the per- 
formance of certain mental tasks, such as solving problems, 
reading and memorizing a result which was bound to puzzle 
those who had to work with the prevailing theories of attention 
and distraction. Laird 12 and his pupils reported that subjects 
estimate the quality of four identical commercial products 
higher if a faint agreeable odor of which the observers remain 
ignorant is fused with the other sensory impressions necessary 
for inspection. 

Benussi 13 had noted that if three light points in a dark room 
were arranged in a straight line and the points successively illu- 
minated, the apparent magnitudes of the two distances involved 
become functions of the corresponding time-intervals, i.e., the 
shorter the interval between the two exposures, the shorter the 
phenomenal distance. Working with telephonic stimuli, Kester 14 
was able to secure a similar kind of "sound movement," analo- 
gous to Wertheimer's phi-phenomenon in that two sounds suc- 
cessively emitted from different places produced the effect of 
the same sound jumping from one place to another. Here, too, 
the shorter the time elapsing between the sounds, the shorter 
the apparent distance through which they "moved." This rela- 
tivity and interdependence 15 of our percepts can be made plain 

31 "Changes in Auditory Acuity during the Performance of Certain Mental Tasks," 
Anier. J. PsychoL, 1926, 37, 139-142. 

13 "How the Consumer Estimates Quality by Subconscious Sensory Impressions, 3 " 
/. Applied PsychoL, 1932, 16, 241246. 

13 Psychologic dcr Zeitauffassung, 1913. 

14 "Ueber Localisation und Bewegungserscheinungen bei Gerauschpaaren," PsychoL 
Forsch., 1926, 8, 75113. 

15 "We of this generation are involved in Mid-Victorian physical thought, which sepa- 
rates time from space and thus tends to think of a situation as primarily a spatial complex; 
whereas modern physics makes no fundamental difference between the space and the time 
coordinates." George Humphrey, "Learning and the Living System," PsychoL Rev., 
1930, 37, 497-Sio. 



THE UNITY OF THE SENSES 149 

to any one by the following simple demonstration : Have a friend 
mark off three equidistant points on your bare forearm. Then, 
while you keep your gaze averted, have him tap each point in 
succession with the tip of his index finger, alternating a rapid 
and a slow movement, while you try to estimate which distance 
in each pair was greater. The Benussi phenomenon is thus 
easily observed on the cutaneous level, for the slow motion 
yields an impression of greater distances, whereas the same gap 
quickly traversed seems shorter. With naive subjects, the varia- 
tions in estimates are more pronounced. Much of this is com- 
mon observation, but the advantage of a new theory is that 
isolated facts may now be intelligibly interconnected. 

The numerous experimental possibilities of this single idea 
which inevitably suggest themselves to the most amateurish 
beginner are a tribute to the fruitfulness of the underlying 
theory. 

Mogensen and English 16 performed an odd little experi- 
ment on the apparent warmth of colors which seems to fall 
within the Hornbostel schema. They made what some un- 
sympathetic observers might consider an almost laughable at- 
tempt to see whether conventionally "warm" colors like red and 
yellow would lower the threshold for heat sensitivity; con- 
versely, "cool" colors like blue and green were expected to raise 
it. Using a slightly modified method of paired comparison, 
they presented six saturated colors wrapped smoothly about two 
slide rheostats which were kept at a constant temperature of 
42 C. The subjects simultaneously grasped both cylinders for 
one second while looking at them and making judgments of 
relative warmth. Strangely enough, the highest "warmth 
values" were ascribed to green and blue, while red had a sur- 
prisingly low position. The authors reached the conclusion 
that the apparent w T armth of colors is insufficiently intrinsic to 
enter into the total configuration in such a way as to modify the 
judgments of tactual warmth. 

To this attractive paper with its curious conclusion, Metz- 
ger, 17 a former assistant of Kohler, took violent exception. He 

18 See Amcr. J. Psychol., 1926, 37, 427-428. 

17 "Certain Implications in the Concept of Gestalt," Amer. J. PsychoL, 1928, 40, 
162-166. 



1 5 o GESTALT PSYCHOLOGY 

claimed that the experimenters made a technical and termino- 
logical error in supposing that the attributes of a whole could 
be found by adding the attributes of its parts. The visual and 
tactual warmth of the papers would have been in configuration 
in any case (in such a relative situation and proportion of extent 
in space or time as can be stated in geometrical terms, such as 
being close together, one surrounded by the other, and the 
like). But visual and tactual warmth are in a gestalt 1S only if 
a greater or lesser mutual modification is possible. To make a 
successful study out of the idea under discussion, it would be 
necessary to take the entire field into consideration by placing 
the observer within the center of a room and then seeing if a 
shift in color will balance a temperature increment or decre- 
ment. To date, no one has followed up this interesting sugges- 
tion. 

Werner's Sensation-Stages 

Werner, 10 a pupil and colleague of Stern at Hamburg, and 
consequently committed to a personalistic version of the Gestalt 
problem, is just as strongly convinced of the scientific and meta- 
physical significance 20 of the "unity of the senses" as Horn- 
bostel is. In a less rhapsodic but equally persuasive vein, he 
has directed a series of experiments during the last decade which 
point definitely to synaesthesia of all forms as the original type 
of sensory functioning and support the view that the restriction 
of stimulation to a single modality is simply a limiting case. 
For instance, if one projects motion picture film so slowly that 
no motion can be seen, movement appears at once if one simul- 
taneously presents corresponding acoustic rhythms ; apparently, 
the dynamics of the bodily sensation-sphere, which was too 
weak to be propagated by purely optical stimulation, was sup- 
ported by the dynamic consequences of acoustic excitation. 

18 It is doubtful if most English-speaking psychologists make this distinction or even 
consider it a desirable one to introduce. The words are used interchangeably in this 
volume. Both configuration and gestalt have come to include not only the spatial configura- 
tion, but also any and all factors which may change the experience, such as the time be- 
tween the stimuli, the spatial distance, the intensity, and whatever else affects the phe- 
nomenal result; in other words, if the terms are not to be identified with the total situation, 
at least they mean very much the same as the total effective situation. 

19 "Das Problem der Empfindens und die Methoden seiner experimentellen Prufung," 
Zsch. f. PsychoL, 1930, 114, 1 52-166. 

^Herder, the^ great German writer of the eighteenth century, evidently held the same 
view on semi-mystical grounds. 



THE UNITY OF THE SENSES 151 

Apparent visual movement of the phi-phenomenon type either 
ceases or is completely deranged when presented in conjunction 
with irregular, unrhythmical beats from a sound source. Just 
as the constancy-hypothesis fails within any sensory area, so it 
fails to apply even to the isolated modality as a whole. It is the 
entire organism which responds and not just one of its struc- 
tures. 

Werner claims that all forms of sensation pass through four 
stages. In the first stage, when a tone is struck upon the piano, 
the sound is localized externally and has a purely objective char- 
acter; in the case of a color, this stage is represented by the 
surface hue of a familiar article. In the second stage, the tone 
seems to lose its m'oorings and becomes a property of the total 
surrounding space; the visual analogies are the atmospheric 
colors of Katz. During the third stage, a significant transfor- 
mation occurs and the object and the self are no longer distin- 
guishable both tone and color have entered into the percipient. 
He is the sounding vessel and the observer himself is filled with 
color. In the fourth stage, the ego itself recedes and there 
remains only the experience of a condition of bodily sensibility. 

The report is also made that with certain observers the re- 
verse sequence can be demonstrated, thus confirming the dis- 
creteness of these stages. A person experiencing blue, e.g., in 
all its totality finds himself in a general physical condition which 
is noticeably different from that involved in reacting to any 
other color. If one presents tachistoscopically certain words, 
such as "warm/' the bodily sensation of something warm is 
occasionally experienced before the word is optically read or 
articulated, indicating that total organismic responses are pos- 
sible even on the symbolic level. Premonitory stages before 
final clear perception occurs usually class the stimulus-word in 
its proper "species" or grammatical category; e.g., even though 
the subjects have an unclear view of the word, they "feel" it to 
be an adjective, verb, etc. 



CHAPTER 9 

MEMORY PROCESSES 

Wulfs Experiment and Its American Extensions 

The most characteristic memory experiment of the Gestalt 
school comes from the hands of Wulf / His point of departure 
was G. E. Miiller's "principle of convergence/' which simply 
stated that with decreasing clearness memory images tend to 
lose their sharp outlines. The experimental procedure consisted 
in presenting samples of relatively simple line drawings to the 
subjects and requesting re-drawn reproductions after various 
intervals. These reproductions were then compared with the 
original specimens to see if the special figural peculiarities 
were preserved, weakened, or strengthened; the corresponding 
processes producing these effects may be called in order con- 
servation, levelling r , and precision (occasionally also termed 
normalization or accentuation), these last three all being mani- 
festations of the general principle of Pr'dgnanz, or "essen- 
tiality," 

Contrary to Miiller, Wulf finds that changes in the direction 
of mdefiniteness are not the only shifts undergone by the visual 
image. For example, the original saw-tooth pattern in Figure 
25 is changed during the interval so that sharper points in the 
angles result ; similarly, the two arcs are made more regular and 
uniform by being construed as equivalent parts of concentric 
circles. The "normalizing" process seems to act in such a way 
that a familiar pattern ultimately forces its own shape upon the 
original so that a different reproduction is given. 2 The common 
feature in all cases is a change in the direction of a "better 

1 "Ueber die Veranderungr von Vorstellungen," Psychol. Farsch., 1922, i, 333-373. 

2 Meyer had earlier noted that a figure often approached in appearance the auxiliary 
memory aid more than the actually seen original; see Zeitschr. f. Psychol., 1913. The 
after-image studies of Frank and Rotschild are also of pertinence here Csee the chapter on 
visual phenomena). 

152 



MEMORY PROCESSES 153 

gestalt" In perception, that which is phenomenally observed 
is subject to the general law of Pragnans, which holds that every 
gestalt tends to become as good as possible, i.e., definite, pro- 
nounced forms are preferred in both nature and experience. In 
memory, the "engram" is still under the influence of this prin- 
ciple and any transformations it suffers must be accounted for 
in such terms. 

ORIGINALS REPRODUCTIONS 

/\/vv\/\ AAAAA 




Figure 25. Wulf s Memory Figures 

The items at the left represent the initial impressions, those at the right the delayed 
reproductions. Note the more marked "angularity" and "curvilinearity" in the latter 
drawing. 



From this study emerges the generalization that all configura- 
tions change through accentuation or leveling of parts. Thus, 
in the field of sensation, visual contrast is a configurative change 
through membering of parts, while adaptation is a change in- 
volving leveling of parts. 

Wulf 's work has been provocative enough to call forth four 
different efforts to confirm his major thesis, viz., those of Gib- 
son, Allport, Perkins, and Carmichael and his pupils. All of 
these experimenters appear to have been impressed by Wulf 's 
remark that in some instances his subjects spontaneously identi- 
fied the presented forms with different phenomenal objects. 
This identification involves the linguistic naming of the objects, 
and this naming appears to have been important in the produc- 
tion of the reproduced forms. Hence, it is little surprise that 



154 GESTALT PSYCHOLOGY 

Gibson 3 finds language association adequate to explain Wulf s 
results and his own, i.e., the nature of a change found in the 
reproduction depends upon the manner in which the figure was 
apprehended, e.g., as "irregular 55 or "Gothic 55 in appearance. 
This verbal label would suffice to produce changes in the engram 
without an appeal to the uncertain principle of Pragnanz. 

Allport 4 made an elaborate investigation of the reproduc- 
tions of two drawings (a simple Greek key pattern and a trun- 
cated pyramid) secured immediately, after two weeks, and after 
four months, from 350 pre-pubertal school children. He found 
definite evidence of the "self-distributing 55 nature of the deterio- 
ration since the change tends to persist in the same direction in 
which it starts. A reduction in size with time also occurs, but 
the shrinkage is confined to a certain range in the absolute size 
of the stimulus. Above all, "there can be no escape from the 
conclusion that one of the properties of the trace is for it to 
retain or to acquire symmetry. 55 While Allport is just as dis- 
satisfied as Gibson with the vagueness of the so-called law of 
Pragnanz, he does incline to accept the principle of the dynamical 
properties of the engram or trace as the most fruitful hypothesis 
for dealing with the decay of the memory image. 

On the grounds of parsimony (-or at least tradition) it should per- 
haps, wherever possible, be conceded that perceptual set, associative 
context, various Influences of habit, etc., modify and change the nature 
of the image and the reproductions. But in the present investigation 
there seem to be several findings that offer grave difficulty to any asso- 
ciational explanation : the persistence of change in the direction in which 
it starts, the spontaneous commencement of a change after a lapse of 
time, the equalizing of strips and the changing of the rectangle to a 
square, the preservation of the old, or the achievement of a new, 
symmetry in the position of the loops, the retention of an impression of 
two, the frequency of displacements, the continuous shrinkage of the 
image, the rigidity of the rectilinear form, the continuity or cohesion in 
each of the figures, the relative lack of elaboration, and the tendency 
toward simplification. Still another feature, too subjective to sustain 
much weight, and yet important, is the fact that the recognizability of 
a design often remained even though every portion of it was incor- 
rectly reproduced. (Op. dt. 3 p. 147.) 



133-148. 



3 "The Reproduction of Visually Perceived Forms," J. Exp. PsychoL, 1929, 12, 1-39. 
* "Change and Decay in the Visual Memory Image," Brit. J. PsyctioL, 1930, 21, 



MEMORY PROCESSES 155 

An attractive variation of this theme is involved in a neat 
experiment performed by Carmichael 5 and his colleagues. 
They presented the same visual figure tachistoscopically to all 
their subjects, but before the exposure one list of names was 
read to one group and another list to a second group. A small 
control group was used to whom the forms were presented with- 
out the assignment of any name. In about three-fourths of 
those reproductions which deviated most from the original pres- 
entation, the figure appeared more like the object represented 
by the concomitantly given word, irrespective of the group con- 
cerned. While the authors prefer a description of the outcome 
in terms of a "dynamically considered process of association" 
it is clear that the following passage is not far removed from a 
configurational account and could readily be assimilated to it : 

It seems that if a subject has just heard, e. g\, the word "eyeglass/* 
certain processes in his organism have been started that initiate other 
processes which are possible because of the past experience of that 
individual with eye glasses as words and as objects. If, while these 
processes are in progress, a figure of two visual circles connected by a 
line is presented to the subject, this figure may later be reproduced in a 
different manner than if the processes present in the individual at the 
time of the same visual presentation had been evoked by the word 
"dumbbell." In other words, without recourse to any elaborate theory, 
one who wishes to make an empirical statement of fact may say: If a 
verbal stimulus form and a visual stimulus form are presented to a 
subject in certain temporal relationships, the processes in question may 
be modified, or rather a new total process may result, which is in cer- 
tain respects unlike either of the previous sets of processes. On subse- 
quent arousal by any "part" stimulus the "reproduction" is thus a com- 
plexly determined total, and not either of its component processes. 
(Op cit., p. 86.) 

Perkins' experiment 6 is an enthusiastic acceptance of Wulf 's 
main thesis. He presented to 150 subjects five simple non- 
symmetrical geometrical figures, such as a disjointed "B" or "8." 
A change toward bilateral symmetry or proportionate relations 
between parts was the most conspicious transformation found 
in periods of recall extending over fifty days. The greatest 
change toward standardization and simplification occurred 



s An Experimental Study of the Effect of Language on the Reproduction of Visually 

933, 44, 473-490. 



Perceived Form," J. Exp* PsychoL, 1932, 15, 73-86. 

6 "Symmetry in Visual Recall, Amer. /. PsychoL, i 



156 GESTALT PSYCHOLOGY 

within the first ten days following the impression, but the drift 
toward maximum "balance" was still noticeable even after seven 
weeks. 

Inadequacy o Associationism 

At the bottom of all standard experiments and Interpreta- 
tions dealing with memory lies the concept of association. This 
doctrine, to be sure, has been protean in its manifestations, but 
even the most radical version has clung to contiguity in expe- 
rience as an explanatory guide. In a long and unusually de- 
tailed monograph, Lewin 7 has assembled the evidence which 
constitutes the Gestalt refutation of this principle. He repeated 
Ach's famous experiment on the "associative equivalent of the 
will," in which the latter had his subjects overlearn certain paired 
associates such as miv-jek, and then later presented them with 
the first member (miv), with instructions to find a rhyming 
syllable such as tvu. This task was relatively easy when only a 
few repetitions were involved, but practically impossible with a 
hundred repetitions when failures to obey directions and re- 
lapses toward the old associates (i.e., the second members) in- 
variably occurred. The will to rhyme, then, had a strength 
equal to 75 repetitions if below that there were no failures and 
beyond it no successes. These various intentions to perform 
certain tasks, such as rhyming, reversing, etc., Ach called "de- 
termining tendencies." 

Lewin challenged this conception. Historically, the deter- 
mining tendency was an outgrowth of Kiilpe's interest in image- 
less thought, the admission of which, as G. E. Miiller pointed 
out, did not involve a rejection of associationism, because this 
mechanism applies only to behavior under given conditions. 
Lewin, however, considers the difference between imaginal and 
imageless thought an unimportant one, since there are many 
fluid transitional states between them, and views the "determin- 
ing tendency" as a superfluous effort to patch a crumbling edifice 
like the Ptolemaic epicyclical theories. 

In one of his sub-experiments on this topic, Lewin secured 
270 repetitions of a nonsense syllable list so that it was learned 

7 "Das Problem der Willensmesstmg trad das Grundgesetz der Assozlation," Psycho!. 
Forsch-, 1922, 1 3 191302; and 1922, 2 t 65140. 



MEMORY PROCESSES 157 

to the saturation point within a period of fifteen days. In the 
later test series, a heterogeneous activity was introduced, con- 
sisting in reversing the syllable (mut-tum) and pronouncing it* 
In order to compare the resulting inhibiting effect, 8 a neutral 
series was also presented : these neutral syllables were not brand 
new, because unfamiliarity would probably have had a slow- 
Ing-up effect, but they had been learned previously, except that 
on this occasion they appeared in a different sequence. Table V 
gives the comparison in terms of average reaction-times in 
milliseconds of each subject. 

TABLE V. MEAN REACTION-TIMES OF SUBJECTS IN PERFORMING 
TASKS INVOLVING PAIRED ASSOCIATES (LEWIN). 

Neutral Reversed 

Syllables Syllables Difference 





581 


602 


21 


AM 




6^2 


^0 


MV 


4Q 


71 




n 


, 16 


IS 













Clearly, the preceding 270 associations failed completely to 
inhibit the "intent to reverse/' as measured by the speed or accu- 
racy of the responses. The inadequacy of the associationist 
explanation of reproduction was even more dramatically revealed 
by a comparison of the reaction times for three series, the first 
of which involved rhyming repeated with an X frequency, the 

second the same task, but with one-half ( J that amount of 

practice, and the third an alternation on the odd and even items 
of different acts. In all instances, the obtained differences were 
of vanishing magnitude, indicating that the activities were on 
the same footing so far as ease of reproduction or execution 
was concerned. 



8 The neurology of inhibition remains a profound puzzle. Configurationists seemingly 
dislike the concept, for Dahl, who found that recognition of nonsense syllables was con- 
sistently 10% better if a sleeping rather than a waking interval intervened, rejects an ex- 



papei, 

1928, n, 290-301. 



158 GESTALT PSYCHOLOGY 

Apparently the syllables simply serve as activating stimuli 
for a specific "readiness for performance." No reproduction 
occurs without the attititdinal direction being one of reproduc- 
ing. The task itself is the important thing and the fact that 
certain items have been connected with one task in the past has 
little or no effect upon what will take place when they are joined 
with another task later on. The "task" may be considered as 
the larger "action-total" to which all part-performances are sub- 
sidiary, and there is reason for believing that genetically the 
whole performance is not the product of a chain-like union of 
minor acts, but that the mastery of the total appears earlier than 
the control of separate divisions. That this general adjustment 
of the organism must be more influential than the effects of 
serial activity is clear from Kiihn's finding 9 that attentively 
reading the same syllable list without the intention of learning 
yielded no mastery of the material, and Hornbostel's confession 
that he had reached for cigarettes in a definite vest pocket 1 50,- 
ooo times, but when he wore another vest in which the tobacco 
pocket was in a different place, he made perhaps one failure at 
the start but no more ! Some larger organization must deter- 
mine what shall happen to individual combinations previously 
established. As soon as this is appreciated, one's faith in asso- 
ciationism is dead. 

The major defect of associationism was its implicit assump- 
tion that two things or processes can interact quite independently 
of their properties, an assumption which is contradicted by the 
data of every other natural science. In chemistry the intensity 
of the reaction or the failure of it to occur is determined by the 
relation of the properties of the atoms in any given case. In 
electricity the occurrence of attraction or repulsion depends upon 
the nature of the electric charges. And, of course, in astronomy, 
acceleration is conditioned by the masses of two or more bodies 
moving interdependently. 

9 "Ueber Einpragen durch Lesen und Rezitieren," Zeitschrift f. PsychoL, 1914, 68, 
396 ff. There are plenty of non-experimental observations which support this view. ' For 
example, familiar objects are certainly thoroughly associated with their names, yet when 
we walk along a street and attend successively to many well-known things, we are far 
from reproducing their names. 



CHAPTER 10 
LEARNING: DATA AND INTERPRETATIONS 

Despite the extraordinary fruitfulness of Gestalt theories in 
many areas of psychological interest, they do not constitute at 
present a well-rounded and finished system of ideas. Common 
characteristics are found, to be sure, in the numerous explana- 
tions of varied mental phenomena, but the interpretative prin- 
ciples can hardly be considered fixed. This is probably inevitable 
in a young and growing movement and perhaps even desirable 
from the standpoint of rapid scientific advance, but it is a cir- 
cumstance which helps account for the many ad hoc and purely 
suggestive views of configurationism. All fields of psychology 
to which the new school has directed its efforts exhibit this 
situation, but none to a less satisfactory extent at least from 
the standpoint of the impartial critic than the domain of learn- 
ing! Nevertheless, the learning experiments with both animal 
and human subjects which the Gestalt position has inspired are 
usually unique and interesting on the theoretical side even to 
the most unconvinced skeptic, and a consideration of the more 
prominent and representative of these studies must now be 
undertaken. 

The Anthropoid Studies 

While marooned in Tenerife as a result of the World War, 
Kohler was able to make some very important observations of 
chimpanzee behavior at the Anthropoid Station established 
there by the Prussian Academy of Sciences. Disbelieving the 
low estimates of animal intelligence which the work of Thorn- 
dike and the American school in general had encouraged, he 
prepared different types of problems for solution which have 
now come to be known as Umweg or detour situations. These 
experiments were first reported in a monograph of the Berlin 

159 



160 GESTALT PSYCHOLOGY 

Academy 1 for the year 1917, and are all characterized by a 
situation set tip by the experimenter in which the direct path 
to the objective is blocked but a roundabout way left open, so 
that the animal must often go further away from his goal in 
order ultimately to attain it. 2 Sometimes this involved merely 
a detour around an obstacle, 3 or the use of implements such as 
sticks, ropes, etc., but in every instance all the general conditions 
needed for the solution were completely visible. All the experi- 
ments were qualitative in form, but executed with such high 
ingenuity that important evidence against the standard trial- 
and-error formula was established. Perhaps the best-known and 
most illuminating of all these observations is described in the 
following excerpt : 

Are two sticks ever combined so as to become technically useful? 
. . . Sultan's sticks are two hollow, but firm, bamboo rods, such as the 
animals often use for pulling along fruit. The one is so much smaller 
than the other that it can be pushed in at either end quite easily. Be- 
yond the bars lies the objective (a banana), just so far away that the 
animal cannot reach it with either rod. They are about the same length. 
Nevertheless, he takes great pains to try to reach it with one stick or 
the other, even pushing his right shoulder through the bars (of the 
cage). When everything proves futile, Sultan commits a "bad error." 
... He pulls a box from the back of the room towards the bars ; true, 
he pushes it away again at once as it is useless. . . . Immediately after- 
wards, he does something which, although practically useless, must be 
counted among the "good errors" : he pushes one of the sticks out as 
far as it will go, then takes the second, and with it pokes the first one 
cautiously towards the objective, pushing it carefully from the nearer 
end and thus slowly urging it towards the fruit. This does not always 
succeed, but if he has got pretty close in this way, he takes even greater 
precaution ; he pushes very gently, watches the movements of the stick 
that is lying on the ground, and actually touches the objective with its 
tip. Thus, all of a sudden, for the first time, the contact "animal- 

* "Intelligenzprufung an Anthropoiden," Abhandlungen d. k. breuss Akad d. 
W%ssenschaften,^9'L7, I. Also in modified book form as Intelligenzpriifungen an Men- 
schenaff<en, Berhn, 1921. The English translation entitled The Mentality of Apes appeared 
in *92S, from which my own citations are taken. 

3 The difficulties some children encounter in learning to sit down is a case in point. 
In order to do this the child must abandon his customary straightforward orientation and 
turn his bade upon the objective before he can attain it! (Lewin) 

3 Adams has made this fact the crux of his interesting "restatement of the problem 
of learning. If there is no obstruction in the field learning cannot occur because the 
creature s need is satisfied with maximal potential parsimony at once. By "parsimony" is 
meant the belief that, animals approach a needed object or state of affairs by the shortest 
route or most economical means functionally possible. This is presumably a feature of all 
living behavior, for Adams suggests that some interesting Umweg experiments might be 
made with potato sprouts in cellars. See Brit. J. PsychoL, 1931, 22, 150-178. 



LEARNING: DATA AND INTERPRETATIONS 161 

objective" has been established, and Sultan visibly feels (we humans 
can sympathize) a certain satisfaction in having even so much power 
over the fruit that he can touch and slightly move it by pushing the 
stick. The proceeding is repeated; when the animal has pushed the 
stick on the ground so far out that he cannot possibly get it back by 
himself, it is given back to him. But although, in trying to steer it 
cautiously, he puts the stick in his hand exactly to the cut (i. e., the 
opening) of the stick on the ground, and although one might think that 
doing so would suggest the possibility of pushing one stick into the 
other, there is no indication whatever of such a practically valuable 
solution. Finally, the observer gives the animal some help by putting 
one ringer into the opening of one stick under the animal's nose (with- 
out pointing to the other stick at all). This has no effect; Sultan, as 
before, pushes one stick with the other towards the objective, and as 
this pseudo-solution does not satisfy him any longer, he abandons his 
efforts altogether, and does not even pick up the sticks when they are 
both again thrown through the bars to him. The experiment has lasted 
over an hour, and is stopped for the present, as it seems hopeless, car- 
ried out like this. As we intend to take it up again after a while, Sultan 
is left in possession of his sticks ; the keeper is left there to watch him. 
Keeper's report : Sultan first of all squats indifferently on the box, 
which has been left standing a little back from the railings; then he 
gets up, picks up the two sticks, sits down again on the box and plays 
carelessly with them. While doing this, it happens that he finds him- 
self holding one rod in either hand in such a way that they lie in a 
straight line ; he pushes the thinner one a little way into the opening of 
the thicker, jumps up and is already on the run towards the railings, 
to which he has up to now half turned his back, and begins to draw a 
banana towards him with the double stick. I call the master: mean- 
while, one of the animal's rods has fallen out of the other, as he has 
pushed one of them only a little way into the other ; whereupon he con- 
nects them again. 

Sticks were necessary implements to the apes in these experi- 
ments, since they were universally employed to pull in fruit. 
If sticks were unavailable, substitute "tools" were readily used. 
Thus, one ape pushed a blanket between the bars of the cage, 
flapped at the fruit a few feet away, and succeeded in beating 
it toward her. If the banana rolled on to the tip of the blanket, 
the procedure was instantly altered, and the blanket with the 
banana was drawn very gently toward the bars. Is this to be 
explained as learning through trial-and-error ? Hardly, since 
the delicate adaptation occurred the first time it was needed. 
When the chimpanzee learned to use a stick in bringing in food, 



i62 GESTALT PSYCHOLOGY 

he came also to appreciate its functional or instrumental value 
in relation to his goals. Under certain conditions, a piece of 
wire, a blanket, an old straw hat, etc., may possess the same 
properties as tools and be used as substitutes. Climbing upon 
the shoulders of a human when a box was absent in order to 
reach an object out of convenient distance is another illustration 
of the chimpanzee's direct perception of significant spatial 
relations. A definite "transfer of training" appears to take 
place, since the essential relational features perceived in the 
first instance are present in later cases, even though the parts 
are all new. 

On the basis of this and numerous other brilliant concrete 
observations, Kohler maintains that the processes of thinking 
and learning have been very much misinterpreted. According 
to the prevailing trial-and-error hypothesis, the succession of 
movements which leads to a solution is "as accidental as the 
winning numbers in a roulette/' The parts of the finished act 
are put together in a purely external way without any "neces- 
sary" connection between them. To this Kohler opposes his 
version of learning by insight. "We can, from our own experi- 
ence, distinguish sharply between the kind of conduct which 
from the very beginning arises out of a consideration of the 
characteristics of a situation, and one that does not. Only in 
the former cases do we speak of insight, and only that behavior 
of animals definitely appears to us intelligent which takes ac- 
count from the beginning of the lie of the land, and proceeds 
to deal with it in a smooth, continuous course. Hence follows 
this characteristic : to set up as the criterion of insight the ap- 
pearance of a complete solution with reference to the whole 
lay-out of the field'' (Kohler, p. 198). 

All the actions of animals in solving problems appear as 
unitary wholes and result directly from" a visual survey of the 
situation. This conclusion is flatly opposed to the "chance" 
theory which maintains that the first smooth and correct execu- 
tion of a relatively complex act is always the result of frequent 
repetition of the "component parts" which are put together 
through sheer "accident." According to this view, the apes' 
ability to use packing boxes as footstools required more or less 



LEARNING: DATA AND INTERPRETATIONS 163 

extensive preliminary "practice." To this Kohler retorts that 
the ape would not even have made the first move toward the 
box unless it appreciated (i.e., had "insight") its possible func- 
tion in helping reach a banana suspended overhead. Even the 
simple "imitation" of the child requires some understanding of 
the necessary inner relations of the action imitated before it 
can be done, and is far from being a purely fortuitous and 
mechanical affair. 

Along with this novel interpretation of animal behavior, the 
configurationists introduced a different conception of the kind 
of experiment appropriate to infrahuman creatures. Accord- 
ing to Kohler's remarks, "One must know and, if necessary, 
establish by preliminary observation, within which limits of 
difficulty and in which functions the chimpanzee can possibly 
show insight ; negative or confused results from complicated and 
accidentally-chosen test-material, have obviously no bearing 
upon the fundamental question, and, in general, the experimenter 
should recognize that every intelligence test is a test, not only 
of the creature examined, but also of the experimenter him- 
self" (op. tit., p. 275). The implication is that if one's the- 
oretical positions lead to the construction of problem-situations 
which can only be solved by "chance" movements, then of course 
one draws out of the experiment just exactly what was put 
into it. 

The daring originality of the Gestalt position is often best 
seen in many incidental suggestions concerning fruitful possi- 
bilities of investigation in fields which the "hard-boiled" and 
"tough-minded" American psychologist of the behavior ist strain 
usually considers unworthy of a rigorous laboratory science. 
The following observation illustrates some of these curious and 
occasional "chips from a German workshop" : One day Kohler 
approached the stockade within which the apes were confined 
and suddenly pulled over his head and face a cardboard copy of 
the mask of a primitive Cingalese plague demon ; instantly every 
chimpanzee fled in horror from this appalling object. 4 Why? 
Surely not just because everything new and unknown appears 

4 The use of a military gas mask is said to have a similar effect upon domestic animals. 



1 64 GESTALT PSYCHOLOGY 

terrible to these creatures, since the changes in the clothing of 
the attendants never produced such fear. Kohler does not hesi- 
tate to draw upon all the imaginative possibilities which the 
Gestalt hypothesis makes available, and inquires, "Is it not an 
admissible hypothesis that certain shapes and outlines of things 
have in themselves the quality of weirdness and fright fulness 
(not because any special mechanism in us enables them to pro- 
duce it), but because, granted our general nature and psyche, 
some shapes inevitably have the character of the terrible, others 
grace, or clumsiness, or energy, or decidedness" (op. tit., p. 
335) ? That is, they do not acquire this fearful aspect through 
experience or association with another originally terrible object, 
but they possess it by virtue of their intrinsic constitution. 5 Not 
easy to demonstrate, indeed, but perfectly plausible in the light 
of configurationism. 

The Nature of Mental Development 

A more extensive application of the insight version of learn- 
ing has been made by Koffka in his Growth of the Mind. 6 He is 
especially concerned here with the manner in which modifica- 
tions in a child's behavior occur during the course of its develop- 
ment. The perennial problem of heredity and environment has 
been unjustly preempted by the rival positions of nativism and 
empiricism, according to which the major phenomena of mental 
life are either inborn functions or products of experience. 
Koffka, instead, is inclined to adopt as a more satisfactory de- 
scription of the problem Stern's convergence theory, according 
to which every capacity is the result of the constant cooperation 
of both inner and outer conditions of development. 

This position is nicely developed in an analysis of the eye 
reflexes of the infant by means of an illustration which has now 
become widespread. " Assume that the gaze of a child is first 

5 An obscure passage In one of Wertheimer's lesser known papers points in the same 
direction.^ **In the dance lie grace and joy. [If we think in terms of the usual body-soul 
hypothesis] does that mean that we really have on the one side a sum of physical muscle 
and limb movements, and on the other something psychically conscious? No" . . . There 



cast in the same mould or pattern even though their substances are distinct! 
6 New York, 1925. Reprinted with corrections in 1931. 



LEARNING: DATA AND INTERPRETATIONS 165 

of all directed straight ahead upon a point A. 
There appears now in the same plane a point 
of light at B on the right. The eyes will then ^ ^ 
move so that this point falls upon the fovea. If ^ 
now another point of light B 1? is introduced 
vertically above B, the eyes will move upward 
and fixate it. Let us assume that the eyes are ^ ^ 
again directed upon A, after which a point AI T^ Tf 
is flashed vertically above it. In passing from 
A to Ai the same retinal position will be effected 
which received B x when the gaze was first directed upon B, 
Again, there is an upward movement of the eyes to effect the 
fixation of Ai ; but although in this case A 2 stimulates the same 
retinal point which in the case of the first retinal movement from 
B to B was stimulated by the point B 1? yet the two movements 
are not at all the same, because the movement from A to Ai and 
that from B to B! require different innervations of the eye 
muscles. What is shown in this special case may be stated in 
general terms as follows : the innervations which the eye muscles 
undergo in movements of fixation are determined, not only by 
the position of the retinal points which arouse the movement, 
but also by the pre-existing position of the eyes. It, therefore, 
follows that every sensory fibre must possess not merely one 
connection with the motor nerves, but as many as may be re- 
quired for all possible positions of the eyes" (Koffka, p. 77). 
It is at this point that Koffka develops his "teleological" 
theory of the reflexes as opposed to the older and possibly simpler 
connectionist view. The empiristic standpoint accounts for the 
learning of optical fixation and coordination in terms of the 
establishment of specific connections between neurons, but in 
the illustration above we have a case of the same end (a periph- 
eral point made foveal) accomplished by two different means 
(movements A A x and B BI). The empirical position 
would thus needlessly demand an infinite number of neural 
connections before ordinary seeing could take place properly 
and it is clear that an infinite number of such separate acts are 
neither learned nor practiced. Without espousing a nativistic 
approach which would explain everything in advance by main- 



166 GESTALT PSYCHOLOGY 

taining that the entire complex mechanism comes ready-made 
Koffka does insist that this event, like all other natural occur- 
rences is fully defined by all the conditions determining it, and 
heredity is but one of these influences. The following radical 
pronouncement gives his explanation of the phenomenon : 

We can no longer assume that the sensory function serves merely to 
release the motor function without involving any inner or material con- 
nection between the two. Instead, the hypothesis is advanced that the 
specific pattern of the seen-object itself regulates the movements of the 
eye. From this It follows at once that the optical sensorium and 
motorium cannot be regarded as two independent pieces of apparatus, 7 
since for many types of performance they constitute a unitary organ a 
physical system within which separate organic parts may react upon 
other parts. Accordingly, what happens at one point in the organism 
is never independent of, or without its influence upon, what is taking 
place at any other point in the organism (p. 80). 

Clearly, if the concept of the reflex is thus seriously modified, 
a similar alteration is bound to affect the field of instinct. 
Gestalt theory inevitably takes exception to the "chain-reflex" 
explanation because it fails to account adequately for the signifi- 
cant characteristic of instinctive acts "persistency with varied 
effort." If the Instinct-apparatus is conceived as a system of 
predetermined pathways, then the fine adaptation of movements 
to a definite end or goal such as one observes even in the 
sucking of the infant becomes incomprehensible. Moreover, 
"when we consider a typical instinctive action as it appears in 
the natural course of an animal's life, the impression is not at 
all that of a summation of part-activities which have in them- 
selves nothing to do with one another. On the contrary, an 
instinctive activity takes a uniform course; it is a continuous 
movement ; it does not appear as a multiplicity of separate move- 
ments, but as one articulate whole embracing an end as well as 
a beginning. Every member of this activity seems to be deter- 
mined, not only by its position with reference to what has gone 
before, but also with reference to all the members of the com- 
pleted act, especially to the last phase which leads to the re- 
sult" (Koffka, pp. 98-99) . To speak figuratively, the instinctive 

7 Cf. tliis utterance with Dewey's position in 1896 (vide supra"). 



LEARNING: DATA AND INTERPRETATIONS 167 

activity Is a melody and not just a succession of tones; it is a 
reaction adjusted to its stimulus and not just released' by it. 8 
So long as the instinctive activity is incomplete, every new situa- 
tion created by the animal is to it a transitional situation. Every 
member of the instinctive activity is determined not only by its 
position with reference to what has gone before, but with refer- 
ence to all the members of the complete act and particularly, the 
last. "Native" behavior of this sort is much more like volition 
than like a concatenation of pure reflexes since it reveals the 
same forward direction that is characteristic of voluntary 
action, 9 

If it be objected that instinctive activity is not characterized 
by cognitive awareness of the end, Koffka would reply that this 



Figure ?6. A Simple Illustration of the 
Meaning of Closure. (After Koffka) 

The nature of the situation practically "compels" 
one to see a completed triangle at the apex. 




is only approximately true. As the goal draws near, apprecia- 
tion of the end-situation is enhanced a fact which exhibits the 
phenomenon of closure as represented by the open triangle below. 
This figure displays "non-closure" (like the transitional stages 
of an instinctive series) and yet indicates with a relatively high 
degree of certainty the direction in which closure is to be effected. 

At this point, an interesting conclusion is reached. It is well- 
known that reflexes and instincts differ largely in degree of com- 
plexity, and even classifications made on this score tend to be 

, .. \ TIle necessity for unitary organization in instinctive action was nicely shown by 
rrunbaum. If a dead fly is brought into contact with a spider's body the spinning reaction 
nes not occur, but it does appear if the fly is fastened to a vibrating wire. A tuning fork 
es the spider to wrap itself wildly around the tine as long- as it vibrates. If the web 
brated and no obiect found, no sniimine- takes -nTae** A-nnar*nfKr rt^. trihr--.-^^ ->,* 



G 
does 



sponse 



Forsch., 1929, o, 275-399. 

It is significant, too, that if a spider's web is slightly injured, it is invariably repaired, 
But if the destruction is extensive, the web is abandoned. 

9 Wundt, for different reasons, however, held a similar view. 



168 GESTALT PSYCHOLOGY 

uncertain. Instead, however, of looking upon the reflexive 
mechanism as the simplest and fundamental fact of behavior, 
upon which other more elaborate psychic structures are built, 
the configurationist reverses the process. Rather than trying to 
explain instincts in terms of reflexes, as Spencer and many 
latter-day behaviorists have done, he would explain .reflexes as 
derived instincts. The whole is prior to the part. .^Genetically, 
the pure reflex Is merely a product of instinctive fixation having 
become differentiated and segregated from a larger and more 
massive pattern of movement. 10 

Mental growth cannot be based, then, upon an assumption of 
Inherited neural bonds, which permits only the possibility of a 
re-combination of old elements. On the other hand, if we al- 
low for the initial existence of primitive figure-ground con- 
figurations, the possibilities for novel expansion and elabora- 
tion are infinite. The perceptual world of the new-born child 
is not the "blooming, buzzing conf usion" of which James spoke, 
but plainly organized with qualities upon a ground. The most 
primitive conscious phenomena are figural as Illustrated by a 
luminous point set off from a uniform field. The child recog- 
nizes its mother's face as early as the second month, and at six 
months it responds differently to a "friendly" face (i.e., a total 
aspect) than It does to an "angry" face a difference which 
obliges us to conclude that "friendly" and "angry" faces are 
phenomenal facts 11 to the infant, and not mere distributions of 
light and shade. There is nothing in the reactions of the infant 
to indicate that It builds up its acquaintance with "objects" by 
uniting an original chaos of sensations. The fact that most of 
his definitions are in terms of use shows how hard it is for a 

10 Some excellent evidence of the correctness of this view has been obtained by Irwin 
in his continuous observation of four neonates during the first ten days. What "he calls 
mass activity appears before local movements and seems to be more primitive than the 
individual reflexes. The infant cries at first only as the whole body moves, etc. See the 
detailed observations in "The Amount and Nature of Activities of New-Born Infants," 
Genet. Psychol. Monog., 1930, 8, 1-92. 

These facts may help to interpret the polymorphous perverse sexuality which Freud 
attributes to the child. On Gestalt grounds the original sex urge would be diffuse, and the 
successive stages of narcissism, homosexuality, and heterosexuality merely representative 
of new differentiations or transformations of an evolving pattern. 

11 It may be of some incidental interest to report a much-needed observation by 
Stratton. Contrary to common view, he found that there is no correlation between the 
presence of a red nag and anger in cattle. The anger reaction is aroused only when the 
total situation is established when the flag is waved in an annoying manner, when the 
waver jumps about, making peculiar vocal sounds and stirring into motion the dust about 
him. See his article on "The Color Red and the Anger of Cattle/' Psychol. Rev. f 1924, 
3Q f p. 321. 



LEARNING: DATA AND INTERPRETATIONS 169 

child to form a concept of a thing outside of Its "setting." His 
number ideas, too, originally refer to groups and not to arith- 
metic units. Mental life, even at the start, is always more or 
less well-organized. 

All these interpretations are necessary preliminaries to the 
doctrine of insight, which is a natural outgrowth of them. 
The characteristic Gestalt ideology respecting the learning proc- 
ess is of especial significance in the light of its tremendous 
educational possibilities. Its point of departure is the question : 
How does the first performance of any new act come about? 
If the structure of an initial achievement can be understood, 
the problem will have been attacked at its most central point. 

The trial-and-error hypothesis simply assumes that a large 
number of random movements are made from those already 
present in the learner's repertoire : the selection of the correct 
responses occurs by the gradual elimination of the useless move- 
ments, whereupon the useful ones fall into their proper serial 
order. The pleasure or pain accompanying these reactions 
serves to "stamp in" the "right" behavior or "stamp out" the 
wrong. Despite the admitted simplicity of this account, it fails 
utterly when the hypothesis is stated in terms which admit of 
an experimental test. Mere casual reflection upon the alleged 
role of frequency and repetition shows that the learner never 
duplicates exactly the same movement, but only the same general 
kind of behavior. A number of authors have observed that a 
cat which has once freed itself from a puzzle box by pulling a 
string with its foot, may upon another occasion pull the same 
string with its teeth, although the latter activity has never been 
exercised. The supplementary law of effect in turn suffers 
from the defect that the reward or punishment often occurs 
after a whole series of movements, some appropriate and some 
inappropriate, have been executed, so that it is difficult to see 
how any retroactive influence would help the learner to know 
which unit should be retained and which discarded. 

In place of all this, configurationism would substitute the 
admitted fact that improvement In efficiency goes hand in hand 
with an increased insight into the nature of the task. If this 



170 GESTALT PSYCHOLOGY 

insightful moment can be understood, then the problem of ac- 
counting for the first correct performance of any act is solved. 
Kohler's animal experiments are invariably brought in to illumi- 
nate this point. If insight be the name given to the perceptual 
reorganization of the visual field confronting the learner, then 
an alteration in the character or "meaning" of certain objects 
is the necessary condition of mastery. What initially was just 
an indifferent stick or plaything for the chimpanzee becomes, 
tinder the pressure of goal-seeking forces, an instrument for 
fetching fruit out of arm's reach. One of the principles affect- 
ing the formation of such optical units is spatial separation 
(Wertheimer, vide supra), and it is easy to comprehend why 
this factor might render this process difficult or at least delay it, 
because an isolated thing can spring into a complex (or be 
assimilated within a larger gestalt) more readily when it can 



Figure 27. A Sample of "Wholeness" in Perceptual 
Experience 

The most natural and primitive designation of this figure is 
"angle"; only upon later analysis do its constituents a hori- 
zontal and a vertical line spring forth. 



be viewed simultaneously with the complex than when it is 
spatially remote from it. In Kohler's words, "the chance of a 
stick becoming a tool is a function of a geometrical constella- 
tion." What the learner eventually suceeds in doing is to make 
an irrelevant subject relevant to the situation, which is some- 
thing quite different from an external connection between a 
certain stick in the field of perception, and a certain sequence 
of movements. On the contrary, an inadequate configuration 
must be broken up into a more adequate one. Chance, in the 
mathematical sense, may produce conditions favorable to such 
a transformation, but does not in itself explain the solution. 

The reason that insight is possible is because things "hang 
together" in nature anyhow. This bearing of one datum, event, 
or experience upon another is a fundamental property of the 
phenomenal world and psychology is not the only science which 
must deal with this fact. When two colored surfaces are 



LEARNING: DATA AND INTERPRETATIONS 171 

placed side by side, it is incorrect to maintain that nothing is 
given in this pair except one color here and another there, just 
as it would be incorrect to describe the figure above as one 
vertical and one horizontal line. What we actually see here 
is an angle. Similarly in the pair of colors what we notice is 
a combination, a configuration, for which we require no other 
transitional, secondary, derived, or superimposed experience. 

Insight, then, takes the place of practice or repetition as the 
key word in a configurationist picture of learning. This does 
not deny that the frequency with which acts are performed may 
affect the course of progress in any performance. Instead, 
however, of serving to strengthen bonds, the chief function of 
repetition according to Koff ka is to prepare the ground for the 
construction of an appropriate figure. In motor acts, e.g., a 
"movement-melody" must be composed, but it may be "played" 
well or poorly. After the configuration has once been con- 
structed, repetition serves to make the behavior appreciably 
firmer and easier but not before. 

This assumption seems to agree better than any other with the 
known facts. We know, for instance, that in a purely habitual achieve- 
ment like that of mechanically learning a series of nonsense-syllables, 12 
a "collective apprehension" is requisite, in which the several members 
are bound together in a uniform whole. Usually this construction of 
a unity occurs in the form of rhythmical groups, but in general what 
we mean to say is that in order to be learned the material must first 
receive some kind of figure, every facilitation in the construction of 
which is a facilitation of learning. . . . All learning requires the 
arousal of configured patterns. ... It follows that repetitions without 
the achievement of a configuration remain ineffective whenever they 
are not positively harmful. In the broadest sense, practice means the 
formation of a figure, rather than the strengthening of bonds of con- 
nection (op. cit., pp. 233-234). 

The problem of learning is ineluctably joined to the question 
of the memory functions. In the associationist interpretation, 
memory is usually accounted for in terms of the re-activation 
of previously used paths or of "traces" left behind by earlier 

12 The ''autochthonous organization" of ordinary visual perception is absent _ here. 
Most experiences are better self -organized when the subject encounters them, but in the 
case of the usual syllable lists, the members are so artificially homogeneous that the 
individual must create his own group structures. Symbolically, the resulting unity is 
better represented by (ABC) than by the conventional sequence A-B-C. 



172 



GESTALT PSYCHOLOGY 



impressions. This hypothesis was neatly disposed of a gene- 
ration ago by von Kries and Becher. Their argument ran as 
follows : Having once seen a figure we are still able to recognize 
it after its position, magnitude, and color have been so greatly 
altered that different neural pathways must now be involved. 
The whole process must take place in quite a different manner 
than it did before ; indeed, no object is ever twice reflected in 
the eye in exactly the same way. If the same fixed paths and 
identical local "traces" are necessary to recognition, it is im- 
possible to see how we ever recognize anything at all, since only 
in exceptional cases would the same avenues be traversed. The 
theory is killed by its own specificity. Nevertheless, it needs 
but a slight modification in order to be made acceptable to 
Gestalt. If the phenomena A, B, and C have been present once 
or oftener as members of a configuration, and if one of these 
reappears with its "membership-character/ 7 it will have a ten- 
dency to supplement itself more or less definitely and completely 
with the remainder of the total pattern. 13 The only ambiguity 
lies in the phrase "membership-character/ 7 which, however, may 
be taken to mean the simultaneous presence of the relevant 
Aufgdbe, determining tendency, or mental set, so that the whole 
to which the part refers is at least roughly indicated. 

One of the doctrines of associationism which plagued the 
more rigid mechanists of an older era was the principle of re- 
production by similarity. According to it an idea X can repro- 
duce another idea Y without any previous existential connection 
established by contact, provided that X and Y are sufficiently 
similar. The thought of thunder not only occurs with light- 
ning, but also whenever we hear any other dull noise or when 
a mountain, cross, or forest is recalled. To a poet, the drooping 
and snow-laden branches of certain evergreens may suggest the 
curious roof of a pagoda. Now this fact, if admitted and 
there is ample experimental evidence in its favor really lies 
outside the bounds of strict associationism, because a purely 
external tie is replaced by an "internal" connection. The name 

33 Note that "membership-character" is the key to the explanation. Shepard and 
Fogelsonger had earlier demonstrated in a memory experiment that the association will 
not work effectively when one attends to the syllables in a different combination from 
that jri which they were learned. See their valuable "Studies in Association and In- 
hibition," Psychol. Rev. f 1913, 20, 290-311. 



LEARNING: DATA AND INTERPRETATIONS 173 

"association by similarity" has served to obscure the fact that 
an entirely new principle of explanation had been introduced 
into mental life, although psychologists since the days of Hart- 
ley and Mill have felt uncomfortable in its presence, as indicated 
by their efforts to reduce this type to sheer contiguity. Koffka 
holds that this principle of reproduction by similarity and the 
fact of recognition are related, and thus avoids the difficulties 
of the fixed pathway hypothesis by considering both as special 
cases tinder a more general law. 

Ogden's New Laws o Learning 

An American psychologist, Dean Ogden of Cornell, who 
prepared the English translation of Koffka's Growth of the 
Mind, has offered in his own volume on Psychology and Edu- 
cation^ an interesting development of the same trend of 
thought. Like the originators of the Gestalt doctrine, upon 
whom he leans heavily for interpretative support, he holds the 
view that the whole organism, by virtue of its very organization, 
is so built that it must behave "pattern- wise/' This conception 
leads to an annihilation of the old cleavage between a stimulus 
on one hand impinging upon a receptor and a response produced 
by an effector; in its place appears the methodological notion 
that behavior is not a response to a situation, but a situation-re- 
sponse, i.e., a total action-unit. An organism and its environ- 
ment are one. 15 The basic facts for psychology (or any other 
science for that matter) are "not isolated, self-sufficient entities, 
but . . . dynamic patterns, the very existence of which involves 
form." Facts of any kind simply are not independent items, 
each existing in its own right quite apart from everything else. 

In the field of native conduct, this approach is illustrated by 
the primacy of instinct, the larger whole, over reflex, the part. 
This sounds incomprehensible to many but is a conclusion but- 
tressed by the neurological researches of Child, Herrick, Cog- 
hill, and others. Child 16 writes, "Summing up, the reflex, 
strictly speaking, is a specialized behavior-pattern depending on 
the presence of certain morphological mechanisms; but it is 

14 New York, 1926. A revision in conjunction with Freeman appeared in 1932. 
:3 Cf. J. S. Haldane, Organism and Environment, 1917, 93-99. 
311 Physiological Foundations of Behavior, New York, 1924, p. 235. 



I 7 4 GESTALT PSYCHOLOGY 

physiologically a development from the primary organismic 
behavior-mechanism, the excitation-gradient." And a gradi- 
ent, of course, always involves a reciprocal dynamic relation 
between at least two areas of high or low potential (or metabolic 
activity), respectively. Since the reflex itself is genetically a 
derived mode of reaction, the futility of attempting to derive 
intelligence and the higher mental faculties in general from 
reflexes, habits, or any other form of fixed or determinate be- 
havior must be obvious to all, as Herrick has correctly 
observed. 17 

Just as the special reflexes differentiate from a broader 
matrix of behavior, so those fixated responses which we call 
"habits" emerge from total-responses, whose intrinsic whole- 
ness has a voluntary character. The volitional process is a com- 
prehensive adjustment of the entire organism to all phases of the 
situation confronting it, and consequently is more variable and 
less stereotyped than a segmental habit which is merely an iso- 
lated behavior-fragment. 

This principle of differentiation is of considerable suggestive 
value to social psychology with its customary hostility to the 
group or "institutional fallacy/' as the following excerpt shows : 

There are two conceptions of society. According to one of them a 
society consists of members which are "integrals" of the group. Ac- 
cording to the other conception society consists of separate "integers," 
each a self-sufficient individual. In the first conception the group- 
whole is more significant than the individuals that compose it ; whereas 
in the second the group, being constituted of individuals, is less sig- 
nificant than its members. In choosing the first of these two concep- 
tions we maintain that a social group determines its members, each of 
whom will possess whatever degree of individuality the particular 
group makes possible. We, therefore, deny the opposite view that 



distinct categories^ behavior. ^ Their apparent difference with reference to the magnitude 
of the pattern is simply a function of growth." "The Genetic Interrelation of Instinctive 
Behavior and Reflexes," Psychol. Rev., 1930, 57, 264-266. 




LEARNING: DATA AND INTERPRETATIONS 175 

society accrues to individuals in congregation, and is, in effect, simply 
the sum of several distinct entities. Social integration is not a summa- 
tion of "integers," instead, the members of a social group are the 
''integrals" of a whole. In other words, priority attaches to the group, 
not to its members. 18 (op. cit., p. 90.) 

By implication, of course, personality and biological uniqueness 
are derivatives of social life, since, like the formation of a figure 
upon a ground, individuality accrues to any body or any thing 
possessed of a degree of articulateness sufficient to define it and 
set it off from its surroundings. Contrary to the usual view, 
sociology and anthropology may rightly consider themselves 
more fundamental disciplines than psychology. 

The remembering activities of the organism may similarly 
be assumed to represent differentiations from the original uni- 
tary "eidetic" phenomena of childhood. It is well-known that 
Jaensch and the Marburg school have assembled interesting evi- 
dence pointing to the eidetic image as either the bridge between 
impressions and ideas (in Hume's sense) or the matrix from 
which both emerged. What we call the perception of external 
or objective reality is a relatively late development in individual 
life like the awareness of self, both being produced by the estab- 
lishment of stable configurations in experience. 

With these premises, Ogden proceeds to the assertion that 
the conditions under which a phenomenal revival of memory 
occurs can be stated in three laws : ( i ) Perseveration, (2) Asso- 
ciation, and (3) Practice. The law of perseveration simply 
holds that "any perceived configuration, the articulateness of 
which is marked, will persist for a brief period of time in a 
state of sub-arousal after the perception is over" (op. cit., p. 
182). Few persons will find difficulty with this statement, 
since it is a pure description of a familiar fact. 

The second law of association is obviously not the historical 
version, reading as follows : "Whenever a number of more or 
less discrete perceptions (or ideas) enter into a configuration, 

18 The hyper-reality of the group finds expression in ancestor-worship and numerous 
primitive practices. The dead, though absent, are treated as integral parts of the social 
organization. Cf. the observations of D. Westermann, "Tod und Leben bei der Kpelle in 
Liberia," Psychot. Forsch., 1922, I, 59-65. 

The psychology of leadership, too, is illuminated by the fact that parts having a poor 
position within a whole may be made more impressive by change of position or function 
within the whole. 



176 GESTALT PSYCHOLOGY 

they become joined by virtue of their membership in a whole; 
the members are thereafter held together, not by the external 
agency of an associative 'glue/ but by the transformation which 
they have undergone in losing something of their individuality 
and becoming the members of a single pattern." The pattern 
and not the individual contents constitutes the memorial re- 
siduum. Ogden submits the following homely illustration as 
a sample of the operation of this principle : 

When I meet Jones and his wife, I am not just encountering two 
individuals, but I am meeting- a "married couple." Later when I meet 
Jones alone, I recall his wife because I perceive him as one of a pair. 
The pattern is supplemented or "filled in" and practically becomes what 
it was upon the first encounter by virtue of the recollection of the miss- 
ing member. If, however, the "context" in which I meet Jones later 
is such that I react to him as an "individual" only, or as a member of 
some other configuration, such as a team, society, etc., to which he also 
belongs, then I do not recall his wife. 

Thus restated and limited, the law of association is derived 
from the more general law of configuration that controls all 
phenomena and all behavior. "Instead of likening association 
to a string of beads, each bead recalling the next in sequence, 
we remember first the string, and then supply the beads as best 
we can." First comes the purview, then the details ; or, if one 
wishes, the entire figure must first arise, after which the parts 
or members fall into place. 

Ogden's third law of practice defines the service of practice 
as "that of furnishing conditions favorable to a closer articula- 
tion and consequent fixation of whatever would otherwise be 
loose and inarticulate in form" (Ogden, p. 193). Frequency 
of repetition under certain favorable conditions facilitates this 
process, since it is all but impossible to repeat any sensori-motor 
coordination without promoting its articulation. To that extent 
even the dullest of drill may be helpful. To the familiar correc- 
tion that "practice makes perfect" only when the right activity 
is rehearsed, Gestalt would add, "only when it favors articula- 
tion into a whole." 

A recognition of the correctness of this picture of the memory 
process was for a long time delayed by the prestige accruing to 



LEARNING: DATA AND INTERPRETATIONS 177 

the standard nonsense-syllable technique which apparently had 
demonstrated that meaningless material could be "linked to- 
gether' 7 purely through bonds formed by contiguity. Never- 
theless, these artificial units can and do enter into a variety of 
configurations of order, rhythm, appearance, and sound. In- 
deed, Ogden alleges that "it is only when the series becomes a 
unified whole, with a certain articulatory progression, or with 
beat, measure, and melody, that the series can be memorized at 
all" (Ogden, p. 210). A pattern must be formed which sug- 
gests its own completion, a position which is strengthened by the 
acknowledged existence of remote forward and retrograde asso- 
ciations. 

The bitter controversy begun by the Wiirzburgers concerning 
the function of imagery in mental life is nicely illuminated by 
an observation which apparently places the Gestalt theorist in the 
ranks of the "imageless thoughters." It is known that different 
minds attain the same ends in reasoning by methods, which, 
when analyzed as to their sensory qualities, vary markedly. 
'The important thing is always the figure; how this shall be 
clothed, sensorially and imaginally, is more or less a matter of 
taste in the selection of draperies which happen to be available," 
i.e., since the dynamic gestalt is transposable, the material vehicle 
by which it is borne is irrelevant. 

Few adherents of configurationism have tried to give a more 
systematic portrayal of learning from that standpoint than has 
Ogden. Its chief characteristic is the close link established 
between it and the ancient category of perception, whereas previ- 
ously the motor aspect of learning had received almost exclusive 
attention. According to the new version, learning begins with 
"the particularization of an objective," or, in more familiar 
language, the establishment of a goal. These objectives arise 
out of a general ground of experience in accordance with the 
same laws which govern the formation of any pattern of re- 
sponse. The end, it seems, corresponds to an area of low 
resistance upon which all the energies of the organism readily 
converge. That is why a motor act is best discharged by con- 
centrating upon the desired objective, and why the same act 



178 GESTALT PSYCHOLOGY 

suffers in smoothness of execution when attention is directed 
toward the parts. All learning depends upon this fundamental 
process of particularization whereby a limited coherence of 
forces is established, but since there are various conditions of 
perception, so also there are different ways of learning. There 
are four of them: differentiation, assimilation, gradation, and 
re-definition, each of which will demand some elaboration. 

Learning by differentiation and assimilation corresponds to 
the analytic and synthetic methods respectively. In analysis of 
this type, an old whole or pattern disintegrates during perception 
and a new one takes its place or becomes subordinated to the 
original ; similarly, the synthesis referred to here is not a com- 
pounding of behavior elements but the integration of a larger 
whole. Gradation is a form of learning involving the percep- 
tion of a direction of change as in the responses of Kohler's 
hens to his paired greys. This ability to respond to a gradient 
or scale (such as brightness-darkness, big-little, left-right) is the 
true basis of all genuine transfer of training, and it is probably 
more than accidental that intellectual discrimination is also based 
less upon the absolute data of sense than upon the pattern of 
arrangement. Re-definition, finally, is illustrated by the vary- 
ing reactions made to ambiguous figures : nothing is added or 
subtracted in making these transformations, but the same lines 
and angles enter into different patterns through the re-orienta- 
tion of the figure. Presumably shifts in viewpoint on contro- 
versial topics as a result of prolonged reflection may be accounted 
for by this mechanism. 



CHAPTER 11 
THINKING AND REASONING 

Primitive Thought 

Anthropological and ethnological data concerning the thought 
processes of primitive men have attracted Gestalt authors be- 
cause of the opportunity thus afforded to trace the emergence 
of differentiated patterns from cruder forms. They claim that 
the customary biases with which European investigators attack 
the problems of tribal life are serious obstacles to a proper 
understanding of the phenomena. 1 Our modern categories 
number, causes, abstract concepts have evolved as a result of 
the biological and social relations of a restricted historical de- 
velopment, and to apply these "prejudices" to simpler societies 
is often a barrier to the discovery of the true situation. 

To us, number is a property of groups of objects which is 
unrelated to their material aspects, but its integration with 
reality is much closer among primitives. For instance, I horse 
-f- i horse = 2 horses ; i man + i man = 2 men ; but i man 
+ i horse a rider ! This unification is a consequence of the 
existence of original natural groups from which separate sub- 
groups eventually split off. 2 Six is really something other than 
3 + 3. The "counting" of animals must have this rudimentary 
quality of form-perception, for if one kitten in a litter of four is 
removed, the mother behaves as though she knew one were miss- 
ing. The gestalt resulting from four small creatures is obvi- 

1 A recent paper by Ruth Benedict indicates that anthropology has "begun to be affected 
by Gestalt views. The strict diffusionists of the last decades operated solely with detached 




i practice observed in many ...__..___ _ 

bed when a child is born is also based upon the primacy of the biological group. When 
any one is ill the medicine-man must treat the entire family ! Reasoning 1 itself appears to 
be the cerebral counterpart of the general biological differentiation of part patterns within 
larger wholes, 

179 



iSo GESTALT PSYCHOLOGY 

ously different from that produced by three. This dependence 
of thought upon natural units is still observed in the modern 
adult, for we do not consider as biologically meaningful many 
divisions which are mathematically irreproachable, such as a 
half-pot, 1/26 of a penny, 1/5 of a goose, 1/4 of an Armenian. 

Time estimates similarly indicate that the temporal aspect 
was and is a dependent property of some other comprehensive 
unit. Yesterday to-day tomorrow constitute a single time 
interval for the child, and even now there are Bavarians who 
estimate distance in terms of the number of pipes of tobacco 
which can be smoked on the way ! Our own ability to tell time 
correctly from a modernistic clock whose dial lacks the con- 
ventional Roman or Arabic figures is a more refined case of 
the primitive imbedding of the time factor in a spatial pattern. 

The language of many tribes reveals this same imperfect 
isolation of the abstract "elements." 3 That Indian, who in the 
process of interpreting a speech lesson was asked to translate 
"The white man killed six bears to-day" and refused on the 
ground that it was impossible for a white man to kill that many 
bears in so short a time, illustrates how symbolism is significant 
to the primitive mind only when an agreement with his total 
experience is involved. 4 For much the same reason the fallacy 
of post hoc ergo propter hoc is so deeply rooted in the mind, not 
only of primitive man, but of the man-in-the-street to-day. 

Formal Logic 

Wertheimer has made an interesting contribution to the 
theory of the thought processes involved in deductive reasoning 
by showing that it obeys Gestalt principles. Logicians have 
long quarreled over the standard syllogism, especially over the 
mode Barbara all men are mortal; Caius is a man; therefore, 
Caius is mortal some arguing that this dull reasoning is a 
petitio, a mere recapitulation of something already known. 




aroi 

vSlker' 
which 

proze: 




THINKING AND REASONING 181 

Wertheimer, however, holds with the scholastics that one can 
often penetrate ahead and increase one's knowledge through 
the use of this device. For example, a citizen who does not 
know in what fiscal district he lives comes to the city hall for 
information. The official says : 

You live in X street 

X street belongs to tax precinct 426 (reading from his record) 

Therefore, . . . 

A genuine transformation has occurred here, because the 
notion one has of a thing is not only enriched, but altered, im- 
proved, and deepened. Or suppose I am looking for a slip of 
paper belonging to assignment A and am unable to locate it. 
I know that I have burnt all manuscript material pertaining to 
assignment B. Suddenly I realize that the A sheet which I am 
searching for also belongs to project B ; therefore, . . . i.e., 
the desired slip appears in two configurations, once as part of 
A and again as part of B. The heart of the performance lies 
in changing the A slip into a B slip. This is exactly what hap- 
pens when an ape breaks off a branch of a tree and converts it 
into a stick for reaching an objective. Again, is 1,000,008 di- 
visible by 9, or not? Obscurity and confusion dominate until 
one sees that the major quantity is 999,999 plus 9, whereupon 
the matter is plain. In fact, the entire domain of reasoning 
bears testimony to the rule that the phenomenal pattern de- 
termines the reaction pattern ; e.g., in his ape experiments, Kohler 
found that a rope wound around a beam presented insuperable 
obstacles to the animals ; what appears to us as a clear, articulate 
structure evidently appears to these animals as a tangled, intri- 
cate structure. 

A demonstration of Wertheimer 's is another excellent ex- 
ample of the concrete projection of the operations of thought 
into the perceptual field. In the following problem the area 
of the square tangent to the circle is to be found ; all that is given 
is the length of the radius R. The "gap" is : What help can 
the radius be in finding the area of the square? The solution 
is delayed until the radius "displaces" itself to a position parallel 
to the base line, where one no longer perceives it as part of the 



1 82 



GESTALT PSYCHOLOGY 



circle but as part of the square; i.e., as a half-side of the square. 
The answer is then read off directly: A = (2R) 2 . 

The same attack is illustrated in this problem : Given a square 
of nine dots (Figure 28b) draw four lines so that each dot will 
be passed through at least once. The pencil is not to be taken 
from the paper and no lines are to be retraced. This task is 
difficult because a person continues his attempts within the area 
of the square. As soon as it is suggested that one could go 
outside the area of the square, the solution is easily found. 





B 



Figure 28. Problems. (From Scheerer) 



(A) Find the area of the square when nothing is given but the radius of its in- 
scribed circle. Hint: visualize R in a position parallel to either the horizontal or verti- 
cal sides of the square. (B) Without lifting the pencil from the paper or retracing any 
lines, draw four straight lines so that each dot will be passed through at least once. 
Hint: the lines may extend beyond the outlines of the square. 



Again, from six matches construct four equilateral triangles 
having a side equal in length to a match. This is also difficult 
because one tries to place them in one plane. If one asks, "Why 
limit yourself to two dimensions ?" a new way of looking at the 
problem (=: direction) is seen and the solution is likely to 
follow. 

Most logicians, unfortunately, have acted on the assumption 
that all cognitive operations must be executed with a knife (sub- 
tractive abstraction) or with a sack (the "class*' concept) ; in- 
stead, even in the barest of syllogisms, certain definite moments 
may be pushed into the foreground or new combinations (not 
aggregations) ma 7 emerge. It would seem that thought can 
tease something new out of existing propositions by identifying 



THINKING AND REASONING 183 

the members of two different configurations (the premises) as 
belonging to a common new configuration (the conclusion.) 

As one might expect, any revision of the logical processes 
leads to a new view of the nature of truth. The familiar "cor- 
respondence" theory of truth holds that any proposition is "true" 
if it can be matched with a fact or item of experience with which 
it agrees. Wertheimer 5 considers this literally a "partial" and 
one-sided approach. It is atomistic logic operating again at 
the very foundations of thought. 

For example, a thing may be true in the piecemeal sense, 
and false, indeed a lie, as a part in its whole. We must dis- 
tinguish the object as piece la], the object as part of its whole 



a 
abc 



the object as part of another 



a 

amn 



To illustrate his 



doctrine of the "two-fold truth," Wertheimer calls attention 
to the case of a man who persuaded another person to steal 
something, and yet who could honestly deny that he stole it, 
even though his responsibility is plain to any one fully acquainted 
with the situation. The number of false conclusions which 
can be encouraged by varying the inflection is another instance 
of this point. 

These two kinds of truth may deal with the same data, but 
one (the "whole" truth?) goes to the heart of the matter, while 
the other may remain external viz., when the set of facts is 
not merely a sum. That is why statistics lie when they fail to 
portray the entire picture. Similarly, an engineer discovers in 
his measurements that he has a straight line. Measurements 
taken in other subsections show that he is dealing not with a 
straight line, but with the asymptote of an hyperbola and his 
equation alters accordingly. Again, on the moral level, it is 
plain that it is in the total conduct of men rather than in their 
statements that truth or falsehood lives. 

Wertheimer does not state this explicitly, but it seems plain 
the Gestalt position strikes directly at the most fundamental 
rule of logic, viz., the law of identity. This solidly-established 
convention maintains that what is true of a group is also true of 
each individual component of the group. This law as it stands 

5 Presented in Ms first publication in English, Social Research, 1934, I, 135146. 



1 84 GESTALT PSYCHOLOGY 

appears to forbid thought to contradict itself. Nevertheless, if 
group or "systemic" realities are different modes of existence 
from "elemental" ones, this ancient dogma will have to undergo 
appropriate modification if it is to be preserved as a guide to 
knowledge. 

Closure 

The closure concept is obviously a godsend to any theory of 
the reasoning process since it justifies one in saying that a felt 
"difficulty" with which, according to Dewey, all thinking be- 
gins is a partial pattern indicating more or less adequately the 
way it is to be completed. The function of reasoning or insight 
is simply to bridge the exposed gap. 6 Thorndike claims that 
the original thinker is one who is dissatisfied with old solutions, 
with the result that he makes an effort to reach interpretations 
which are more agreeable, but the configurationist would hold 
that the imperfectly formed neural patterns create tensions or 
gradients leading inevitably to their own completion. A prob- 
lem presents itself as an open gestalt which "yearns" for solution, 
and it is the business of thought to find the solution by trans- 
forming the open gestalt into a closed one. 7 From this point of 
view, an intelligent person is one who is able to transpose one 
situation into another and to reconstruct his perceptual field. 

The uniqueness of the Gestalt version of thought is, like its 
other contributions, best understood in the light of the associa- 
tionist accounts which it hopes to supplant. According to the 
constellation-theory of Selz, which is the climax of associa- 
tionist speculation, a cross-section of consciousness at any 
moment reveals a chaotic mass of competing reproductive ten- 
dencies. To explain the cosmos of our orderly thinking we 
need to know how the presentational series reaches the goals 
which it establishes for itself. Consequently, associationism 

8 Gestalt theory maintains that understanding and inventio 




that the proper closure's are mostrfadiS -made ' ^e patterns m sucn a way 
J Some excellent illustrations of this method may be found in O. L. Reiser, "The 
Logic of Gestalt Psychology/; Psychol. Rev., 1931, 38, 359-368. 



. . 

bndse and no soal; what can be l 



THINKING AND REASONING 185 

postulates the presence of a relatively stable mental content from 
which other ideas or definite reproductive tendencies proceed 
and from which a selection is made. Koffka, 8 however, holds 
that the introduction of this conception of goal or direction is 
really an anti-associationist admission. To an idea I wish to 
find its opposite. The goal-presentation must be an idea that 
the concept sought is the "opposite." But what kind of an idea 
is this and how is this joined with the idea ultimately sought? 
By virtue of what associations is it able to inhibit or facilitate 
the numerous reproductive-tendencies at hand? This steady 
move toward a defined objective constitutes a difficulty which 
the associationist theory is unable to solve, because it makes 
thought essentially a memory- function. 

The Gestalt solution is quite different. "Opposite-pair" is 
a gestalt and like all other gestalten it evokes a "gestalt-disposi- 
tion" with an important property. 9 If a child has comprehended 
that "big" is the opposite of "little," then a gestalt disposition 
does not only exist for this and similar conceptual pairs but the 
configuration "opposite" in general has been acquired just as 
the child which has learned to distinguish yellow and green 
finds it much easier to master the distinction between red and 
blue. What makes a task a task, a question a question ? Noth- 
ing other than incomplete thought-structures, thought gestalten 
with gaps, and from the configuration as a whole proceed strong 
tendencies to close these gaps. 

The difference between the two accounts has been so well 
expressed by Helson 10 that one can do no better than quote his 
pertinent paragraph in its entirety here: 

The logical analysis of the associationist is responsible for the plight 
he gets into ; for, having begun with elements having" no relation to one 
another, he must by some means connect them. The configurationists 
answer the question of thinking by asking how thought progresses from 
the beginning to the end of a problem. The historic antitheses of sense 

3 Cf. his section on "Psychologic" in Lehrbuch der Philosophie, edL Dessoir, Berlin, 
1925, 493-603. 




The second question would "nave been correctly answered but for Its predecessor, 
which established a latent Einstettung. 

w "The Psychology of Gestalt," Arner, J. PsychoL, 1926, o7, p. 54. 



186 GESTALT PSYCHOLOGY 

and intellect, sensation and thought, do not exist in configurationai 
psychology; hence there is no need to build one on the other or to 

derive one from the other For thought to progress along 

logical lines it must proceed within a structure, and this structure 
is very often furnished in perception. The relation of objects to one 
another are given in phenomenal configurations won in perception. A 
change in the object of perception may mean a change in the whole 
course of thinking, since rational insight demands the perception of 
the structure of the whole. Fruitful thinking, consists therefore, in 
approximating, either in perception or in thought, to the sacldicher 
Ziisammenhang (= functional relation) of configurationai structures. 
Configurationai connections are not the results of frequent associations 
or blind juxtapositions, but are inner bonds imposed by the demands 
of the total structure. Once this structure is seen, the parts must fol- 
low one another in a certain way. 

Relation versus Gestalt 

To many persons the notion of gestalt is wrongly identified 
with that of "relation." Koffka tries to remove this difficulty 
by showing that relation is the species of which gestalt is the 
genus. 11 The relation-seeing attitude cannot be identified with 
the "structure-seeing" attitude because these are two entirely 
different perceptions and involve wholly different phenomenal 
patterns ; the perception of a structure is prior to and independent 
of relational judgments. For instance, the impression of a 
rectangle or of any gestalt is not sufficient in itself for the appre- 
hension of relation. If I see a rectangle, I see a regularly closed 
figure, but I do not see the equivalence of the two vertical or 
horizontal boundary lines. The presence of such a gestalt is not 
yet the basis of the judgment : a = 6. However, that does not 
mean that a perceptual foundation for this judgment is absent. 
If I try to describe the phenomenal condition underlying the ver- 
dict, "Distances a and b are equally long," I find that the rectan- 
gular character is there but somewhat altered, so that the sides 
a and b are now the main part of the total and in tension with 
respect to each other, while the two other sides merely serve as 
the transition between them, the character of the transition- 
divergent or convergent determining the relation. 

ctiltv^tl^? -^ genus ; s P e 9 es "relation" Is itself a gestalt! The common-sense diffi- 
ISum^lfSSS? f g f ? m etmS S f r J ela . tl . ons . lies in determining which, among an 
orTcS^oncepl ^ of \ssence ^ S ' *" 7 dCCI81VC m ***** * what i4 is ' fcf ' the his ' 



CHAPTER 12 
RESEARCHES ON INSIGHT 

To an outsider the greatest strength and weakness of the 
Gestalt position appears in the application of its central notion of 
insight or organized response to the problems of thought and 
learning. According to it, all learning occurs via insight ; where 
there is no insight, there is no learning. We have already seen 
evidence of the positive power of this concept to illuminate older 
observations and to advance new types of research. Neverthe- 
less, it remains the one feature of the configurationist program 
which has been least acceptable to the orthodox. Investigators 
who have reluctantly acknowledged the reality of field- forces in 
perception and who appreciate the derived character of the 
simple reflex bitterly attack the insight explanation as unscien- 
tific in the extreme, even where it means little more than trans- 
posability of the general properties of response from one 
problem to another. Why? 

Terminology 

Part of the difficulty has undoubtedly been a matter of termi- 
nology. In English "insight" is an acceptable popular and loose 
equivalent of understanding, comprehension, or occasionally, 
intuition. The colloquial "hunch" is probably a low-level form 
of insight in its stricter use. On the other hand, insight has also 
been employed in abnormal and social psychology to designate 
the personality trait involving awareness of one's condition or 
status. Supplementary to these uses (and perhaps embracing 
them all) is the more technical version, which strangely enough 
was not introduced by the Gestaltists as many suppose, but by 
Woodworth and Ruger * between 1908 and 1911. True they 
did not identify it with configurational learning, but emphasized 

* The Psychology of Efficiency, New York, 1910. 

187 



1 88 " GESTALT PSYCHOLOGY 

the presence of sudden solutions in problem-solving, an excep- 
tional fact which Thorndike had noted in his early animal experi- 
ments and which he had suggested might be explained by the 
presence of "a totally different mental function, namely, free 
ideas." 2 

Another limitation has been the failure to reach an acceptable 
set of criteria s by which the phenomenon of insight could be 
identified. 4 This has obviously been needed in animal studies 
and is equally desirable in describing the conduct of human sub- 
jects. The best single effort to meet this requirement has been 
made by Yerkes 5 in the following list : 

In acts which by us are performed with insight or understanding of 
relations of means to ends, we are familiar with certain characteristics 
which are important, If not also differential. These features are: (i) 
Survey, inspection, or persistent examination of problematic situations. 
(2) Hesitation, pause, attitude of concentrated attention. (3) Trial of 
more or less adequate mode of response. (4) In case initial mode of 
response proves inadequate, trial of some other mode of response, the 
transition from one method to the other being sharp and often sudden. 
(5) Persistent or frequently recurrent attention to the object or goal 
and motivation thereby. (6) Appearance of critical point at which 
the organism suddenly, directly, and definitely performs required adap- 
tive act (7) Ready repetition of adaptive response after once per- 
formed. (8) Notable ability to discover and attend to the essential 
aspect^or relation in the problematic situation and to neglect, relatively, 
variations in non-essentials. 

The definition of insight which would be most in accord with 
the Berlin school of Gestalt states it to be the phenomenal 
correlate of the "dosing" of a configuration. This is merely 
another way of saying that when an experience is patterned so 



" 1 ^ 6 ^ 1 , 3 ^ 6 ? f Monkeys," Psyckol. Rev. Monog., 1901, 3, No. 5, p. 16. How much 
Tl !^ e ' 1CaI 7 Ieanm f s 9* ** observer affect his interpretation is seen in contTa?t?nS 
Thorndike s early work Wl th the elaborate repetition of the "cat in the puzzle-box^ problem 



_ j", <5 S ^ I * 6 ^ xperfmentaI Studies of Adaptive Behavior in Cats," Comp. Psy. Monographs, 
articl/on ^ThTVn^f VC ^ 5 u ?, sio . n of ,^ is P^f the reader is referred to the writer's 




"*"'. ";-'-* . *iBui. o,ua i-wc v^untexi or uestait," Amer. J. Ps'vchol 1012 ^i e-rfi-c-?^ 
5 TA^ Grefli ^^ej, New Haven, 1929, p. 156. 



RESEARCHES ON INSIGHT ^9 

that certain aspects of the pattern are felt directly to depend upon 
other aspects, then the term insight may correctly be applied. 
Insight is the how and why of a situation, an understanding of 
the innermost nature of the field, the stresses not merely appre- 
hended but comprehended. A highly important question deal- 
ing with the adequacy of the resulting configuration arises from 
this claim. Alpert 6 applied a variation of Kohler's stick and 
box-building devices to children aged 19-49 months with con- 
firmatory results, but added the important modification of partial 
and gradual insight. One is tempted to believe that where in- 
sight is not immediate, or at least sudden, it has lost its essential 
character, and one may also suspect that where partial and grad- 
ual insights are admitted they represent either what would ordi- 
narily be termed sheer trial-and-success or at best transition 
stages between that and insight per se. The main difficulty of 
the trial-and-error-hypothesis how the correct solution could 
ever be hit upon by chance is only partly removed by the insight 
view which means little more than "appropriateness" without 
explaining wherein the appropriateness consists other than the 
pragmatic test provided by the solution. 

In the same connection the work of Kreezer and Dallenbach 7 
is of interest in showing that the relation of opposites is learned 
by varying percentages of children between five and seven and 
one-half years after but one or two illustrative examples. A 
sort of all-or-none law seemed to function on this level as the 
presence of insight was a sine qua non for mastering the task. 
Those who learned, learned by insight, and those to whom the 
insight did not come, did not learn at all. The only danger in 
such a bald statement of the fact is that it suggests that the solu- 
tion itself becomes the criterion of insight a conclusion which 
several critics have not hesitated to draw. 

Zeininger s has noticed a type of thinking in young children 
which he characterizes as "If -then-thinking." It consists in a 



6 The Solving of Problem-Situations by Pre-School Children, New York, 1928. 

7 "Learning the Relation of Opposition," Amer. J. PsychoL, 1929, 41, 432441. Dal- 
lenbach had earlier made some rough investigations of this problem with his own and 
Ogden's children. See "A Note on tie Immediacy of Understanding a Relation," PsychoL 
Forsch., 1926, I, 268-269. 

s "Magische Geisteshaltung im Kindesalter und ihre Bedeutung fur die religiose 
Entwicklung," Zsch. f. angew. PsychoL, 1929, Beiheft 47. Huang adopts this view with 
fly few reservations in his "Children's Explanations of Strange Phenomena," 



surprisingly tew r 
PsychoL Forsch., i 



931, 14, 63-182. 



190 GESTALT PSYCHOLOGY 

mere statement of what one has to do in order to realize certain 
ends, without insight- into the situation. Many adults, too, 
know that if they wish to attain a desired result they must press 
a given button, but they are wholly ignorant concerning the 
"why" of the machine's behavior. One may also be compelled 
to distinguish two l levels" or kinds of insights one involved 
in the internal solution of the problem itself and the other present 
as a sort of perception of the most promising way of approaching 
the task. 

Varieties o Insight 

Ideally, a demonstration of the presence of insight requires 
the solution of problems in such a way that errors abound up to 
the point at which the "principle" is discerned after which no 
further mistakes appear, i.e., 100% failure precedes the critical 
moment followed by 100% success. In an experiment con- 
ducted by the writer 9 ten types of problems of this sort were 
constructed. All of them were ludicrously easy as soon as the 
major common relations had been established by the subject. 
One series consisted of a reversed analogies test as follows : 

1. steel : hard burden : light 

2. stomach: middle head: bottom 

3. sky : blue snow : 

4. Socrates: wise Hercules: 

The subject was shown the first line, told that light was the 
correct word because of a certain principle which the experi- 
menter was consistently using, and then asked to insert the right 
term in the second line. If, as usually happened, he failed to 
respond properly after an arbitrary time limit had elapsed, the 
answer was given to him and he was directed to proceed further. 
Some subjects, of course, needed more illustrations than others 
before they saw the point of the arrangement, but all sooner or 
later realized that a word opposite to what one would normally 
give was correct ; thus, black and weak are right answers in lines 

9 The full report may be found tinder the title, "Insight versus Trial-and-Error in the 
Solution of Problems," Am&r, /. PsyclioL, 1933, 45, 663-677. If insight be taken to mean 
that no problem-solving activity is ever planless but always involves a search for some 
pattern which can be applied, little reason for quarreling remains. 



RESEARCHES ON INSIGHT igi 

three and four respectively. In all cases introspective records 10 
were used to throw light upon the mode of solution. 

The main feature observed in every situation and confirmed 
by the subjects' reports was the frequency with which varying 
hypotheses were developed and rejected before the desired in- 
sight made its appearance. 11 A more or less extensive period 
of experimentation was evidently necessary before any of the 
problems were solved. During this preliminary struggle some- 
thing very much like trial-and-error 12 learning unquestionably 
occurred, otherwise why should so many distinct possibilities of 
action have arisen only to be discarded ? and the time relations 
involved confirmed a classification of the varieties of insight 
made earlier by Alpert : 

1. Solution with immediate insight (i.e., first "idea" correct) 

2. Solution with gradual insight (= built up by accretion) 

a. Partial insight (= "good" errors) 

b. Complete insight 

3. Solution with sudden insight (sharp transition occurring 

after perception of crucial factor) 

Another study by the writer 13 on this theme may be referred 
to here to indicate that Gestalt theorists do not restrict the use 
of insight to the "rational" aspects of man's conduct. It is as 

10 Most phenomenological descriptions of the moment of insight tend to be disappoint- 
ingly meager and barren. The process seems never to be as amenable to observation as the 
product. Benussi mentions his ability to produce the insightful experience in hypnotized 
subjects without requiring any conscious efforts at problem-solving. This notion of an 
autonomous or transcendental insight in which the function is divorced from both the 
normal content and situation is puzzling in the extreme, but it may account for the divine 
illumination reported by many mystics, both medieval and modern. See his little-known 
study, "Zur experimentellen Grundlegung hypno-suggestiver Methoden psychischer Ana- 
lyse/* PschoL Forsch., 1927, p, 197-274. 

11 This same point was independently made by Patton, whose subjects in a more limited 
perceptual task always worked on the basis of some theory or principle. The present 
writer holds that this systematic sampling by hypotheses and "guesses" (rather than dis- 
jointed movements) is all that the trial-and-error position maintains. Why should the last 
trial which happens to be correct be dignified with the special label of "insight" ? f To 
demonstrate that no behavior occurs without^ a goal is not proof of the presence of either 
insight or trial-and-error since the same objective can often be obtained by both means. 
See "An Experimental Study of the Emergence of Insightful Behavior," Psychol. Mono., 
1933, 44, 98-124. From the Gestalt standpoint, what seems to be trial-and-error learning 
is simply behavior whose "maturation" is incomplete, and hence the responses are gross and 
inadequate. Unfortunately, there is a lot of ambiguity and mystery about the maturation 
process. 

13 This would seem to dispose of Wheeler and Perkins* claim that to deny insight to 
animal conduct is to commit the "fallacy of the double standard" (Principles of Mental 
Development, p. "85). While there may be some legitimate doubt ^as to the relevancy of the 
human standard when imposed upon animals through the experimentally-controlled situa- 
tion, there is simply no other standard which can be followed in observing adult behavior, 
especially if one combines objective records and introspective testimony. 

13 Cf. the original, "Configurational Factors in the Understanding of Actions," /. Gen- 
eral Psychol., 1932, 7, 438-452. 



192 GESTALT PSYCHOLOGY 

Immanent a factor in our crudest social perceptions as in the 
more severe forms of thought and learning, characterizing static 
visual figures as well as dynamic action-totals. Our comprehen- 
sion of another's behavior is alleged to be dependent upon this 
condition, e.g., an identical act such as wiping a tear from the 
eye has one meaning when it occurs in the spectator of a stage 
tragedy and another on a windy road. No isolated act can be 
comprehended by itself. 14 To understand is equivalent to tak- 
ing into consideration the total situations. 

Kohler's remark that one must never cease to wonder at the 
obvious and commonplace details of experience is a useful intro- 
duction to the experiment under discussion. In outline it was 
absurdly simple, consisting only of a number of acts partially 
completed by the performer in the presence of an observer who 
was then asked to tell what the experimenter intended doing. 
The identical words and actions were carried out in two slightly 
different settings in which the physical articles of the room were 
the only items displaced, added, or removed. 

Example: Upon the work-table were the following objects: 
Left (of the performer), a small torn colored map of Europe; 
right, a fair-sized piece of stiff yellow paper, upon which small 
paper strips had been pasted to form geometrical figures one a 
complete triangle, the other a hexagon with two sides missing. 
In the middle of the table were scissors, ruler, a tube of paste, 
and white paper of the same kind as the paper strips described! 

The subject, who was seated opposite the experimenter, ob- 
served the following action which he was prepared to interpret : 
The experimenter soliloquized, "The length is about 14 cm" and 
slowly cut out two equal paper strips, whose dimensions corre- 
sponded to the strips forming the figures (but also to the length 
of the rent in the map). After the experimenter had smeared 
the surface of one strip with some paste, he broke off the action. 

Under the conditions just described the subjects in every 
single case declared that the intention was to complete the hexa- 
gon, whereas in a parallel experiment, in which this unfinished 
design was removed, every observer maintained that the purpose 



f^?P^ Untersuchnngen 



RESEARCHES ON INSIGHT I93 

was plainly to repair the map. The vectors or field-forces in one 
case pointed clearly to a certain inevitable outcome while the goal 
in the second seemed equally obvious. Two distinct perceptual 
organizations resulted from two slightly altered objective con- 
figurations in which the role of every remaining part had been 
changed by the removal of one. There is some reason for be- 
lieving that the situations here portrayed may represent the most 
primitive and general stage of the phenomenon of insight. 15 

Patterns and Learning 

Guilford's study on the role of form in learning 16 is an excel- 
lent demonstration of the familiar fact that the perception of the 
system and organization of the material to be mastered is a 
facilitating factor of the first order. Nonsense syllables are 
notoriously more difficult to learn than a random list of numbers 
or meaningful disconnected words and these in turn are far more 
difficult than items with a plan or pattern uniting them. The 
way in which the "form" emerges is of interest for the light it 
throws upon the development of insight. To illustrate, let the 
reader endeavor to memorize the two lists here given : 

force thick 
Korans wall 
heaven it 
ye tea 
sago , of 
are myrrh 
feathers seize 
brat knot 
four strain 
than debt 
this drop 
count path 
anent Asa 
anew gent 
neigh tell 
shun reign 

13 According to Erismann, the presence of insight is the major trait of the psychical as 
opposed to the physical. His claim, however, that insightful knowledge can never be false 
is rejected by configurationists as a confusion of the epistemological and psychological 
standpoints. See Die Eigenart dss Geistigen, Leipzig, 1924. 

19 /. Exp. PsychoL, 1927, io } 415423. For a more extended demonstration of the 
thesis that grouping is a necessary phenomenon in learning, see F. H. Lumley, "An Inves- 
tigation of the Responses Made in Learning a Multiple Choice Maze," PsychoL Mono., 
193 1 42 No. l8 9- 



194 GESTALT PSYCHOLOGY 

The first oral reading or so rarely suggests the scheme in- 
volved in the series, but there is a definite evolution through 
quick stages until the order comes with a flash. As soon as the 
Gettysburg speech and Portia's plea are seen to apply, the learn- 
er's difficulties are over. James, of course, had earlier referred 
to the same phenomenon in his clever illustration, pas de lieu 
Rhone que nous. 

There is an extraordinary incident in the childhood life of 
Helen Keller, 17 who was born without the power to use her 
major sense organs, which is pertinent to any discussion of 
insight, for it shows how this distinguished woman became 
aware of the existence of language and discovered that every- 
thing has a name. She had been taught certain ideas and sym- 
bols by means of tactile stimulation but even at the age of seven 
her vocabulary was very meagre, the words employed designat- 
ing only total situations and not specific individual objects. On 
one occasion, the teacher took the girl to the pump and while 
Helen was pumping water into a container, she spelled into the 
child's hand the word w-a-t-e-r. As the cold water touched her 
skin, Helen suddenly recognized that the manually spelled letters 
had the definite meaning of water. She had never before under- 
stood that a word had reference to a distinct and limited item of 
experience. Excitedly she pointed to the ground while its name 
was similarly impressed upon her hand, a process which contin- 
ued with other familiar objects, so that within a few hours she 
had enormously increased her vocabulary and understanding of 
common concrete things. 

An experiment by Lewis 18 on the configural response in the 
chick is an excellent case of learning of the non-specific type. 
He trained his animals to select the correct light of three abso- 
lutely different intensities (bright, moderate, or dim) in varying 
positions with reference to each other, and then "stepped" the 
illumination up or down so that all the original intensities were 
altered. Almost perfect transfer was found to prevail when 
the total situations were changed, indicating clearly that the 

17 See Ber appealing Story of My Life, New York, 1920, pp. 315-316. The entire auto- 
biography is a mine of excellent psychological illustrations. 
^Journal of Experimental PsychoL, 1930, 13, 6175. 



RESEARCHES ON INSIGHT 195 

chicks saw the lights in their mutual brightness relations; e.g., 
the chicks who were trained to select the medium light must have 
been able to do so only because they perceived it in connection 
with a brighter and a dimmer. 19 Most of the learning was 
sudden and the obvious initial delays before choice argue against 
any simple mechanical view of learning. In many trials, the 
chick, although hungry, would remain motionless upon being 
placed in the apparatus and would run to a compartment only 
after surveying the situation. The following observation is of 
interest because of the theoretical reconstruction it demands : 
"The box furnished us with a convincing bit of evidence against 
the theory of identical elements. A light of moderate intensity 
was placed in the box and on succeeding trials only the lights in 
the semi-circle were changed. Thus the box-light, remaining 
constant, would assume all possible relationships to the total pat- 
tern, now the brightest, now medium, and now dimmest ; but as 
it was of the same absolute nature both in position and in inten- 
sity, it was an identical element of all identical elements. When 
the light was correct or incorrect the chicks chose or deserted it, 
as the case might be, without the slightest compunction ; the light 
changed its identity with the changing conditions related to it." 



A Canadian psychologist, George Humphrey, 20 without com- 
mitting himself to a thorough-going Gestalt schema, has never- 
theless lent credence to the insight doctrine by showing that all 
learning or modifications of the organism must obey systemic 
laws involving more than purely local effects. Going to bio- 
physics for his inspiration, he has reasoned that there are certain 
properties of systems in general which, if the phrase "biological 
system" is anything more than a pious metaphor, should be 
found to belong not less to living than to non-living complexes. 
Le Chatelier's rule, originating in physical chemistry, would also 
appear applicable to the problem of habituation in animals: "If 

19 DeCamp, in an unpublished experiment performed at the Pennsylvania State Col- 
lege, finds that rats can learn to go to the first of two entrances successively illuminated or 
even to the middle of three, lighted one after another, provided the correct entrance is illu- 
minated a little longer than the incorrect. When one considers how many combinations of 
order the three entrances permit, the generality of the number pattern response is striking. 

20 "Le Chatelier's Rule, and the Problem of Habit and Dehabituation in Helix 
albolabrif," Psychot. Forsch., 1930, 13, 113-127; also **A Note on the Applicability of Le 
Chatelier's Rule to Biological Systems," ibid., pp. 3^5-3 6 7- 



196 GESTALT PSYCHOLOGY 

a change occurs in one of the factors determining a condition of 
equilibrium, the equilibrium shifts in such a way as to tend to 
annul the effect of the change/' While it is true that there is a 
difference between chemical and physiological equilibrium in 
that the latter must call upon outside energy to maintain itself, 
intra-organic changes of a compensatory character take place 
after every stimulation with the result that a fresh equilibrium 
or new stationary condition consistent with the new external or 
internal environment is brought about. Although Humphrey 
does not say this directly, it is clear that if no learning can occur 
without a rearrangement of neural patterns 21 we have precisely 
that sort of reconstruction of conduct which the insight hypothe- 
sis demands. 

Thorndike and Belongingness 

The one author who has had to bear the brunt of the Gestal- 
tists' attacks because he seems to them the stoutest champion of 
the trial-and-error version of learning is Thorndike. Much of 
their criticism has an element of unfairness born from limited 
acquaintance with only a fraction of his writings, for he has 
always stressed the importance of the situation as contrasted 
with the individual stimulus. However, it is true that Thorn- 
dike has exhibited to perfection the analytical trend which is 
anathema to the configurationists in his emphasis upon the 
reality of bonds and connections, but this opposition is far from 
being as naive neurologically as many opponents seem to 
believe. 

One fact which is not yet sufficiently appreciated is that 
Thorndike himself 22 is responsible for some of the strongest 
evidence against the crude repetition or frequency theory of 
learning as advanced by Watson. As every critic realizes, the 
law of use never intended to explain more than the strengthening 
of an association without accounting in the least for the original 

21 The "tumble" and "flop" of colloquial speech and slang express the same process 
more picturesquely, and at the same time suggest the interesting kinship with reversible 
figures.^ 

It is significant that the transition from waking life to sleep and vice versa takes place 
suddenly. If the physiological processes of the cortex overstep the threshold with respect 
to quantity, intensity, or complexity, by ever so little, consciousness is present. 

Human Learning, New York, 1931. See also Brown and Feder, "Thorndike's 
Theory of Learning as Gestalt Psychology," Psy. Bull., 1934, 31, 426-437. 



RESEARCHES ON INSIGHT 197 

formation. From the beginning he has maintained the greater 
significance of the Law of Effect and in recent years has made a 
certain concession to the Gestalt claim by admitting a new prin- 
ciple of "belongingness." 23 This is simply illustrated by read- 
ing repeatedly to an audience such sentences as 

Norman Foster and his mother bought much. 
Alice Hanson and her teacher came yesterday. 

If one then asks, "What word came after 'Norman Foster and 
his mother' ?" the correct answer comes much more often than 
if one asks for the word following "much." In Thorndike's 
opinion, the temporal contiguity of one item with the w r ord fol- 
lowing it, the mere sequence without belonging, does little or 
nothing to the connection. In our example, the last word of the 
first sentence did not arouse the first word of the next because it 
belonged to a different thought unit. Repetition of a "connec- 
tion" in the sense of the mere sequence of the two things in time 
has very little pow r er as a cause of learning. Belonging is neces- 
sary. To quote a striking instance : 

If a man simply experiences A and B in succession repeatedly with- 
out any sense that B after A is right and proper, or even that B belongs 
with A, and without himself producing B when he suffers A, the in- 
fluence upon the man is very, very slight. You practically always raise 
the body and bend it back after tying your shoes, and so have the sensa- 
tions of bending the body back as a sequent to those of tying your shoes. 
You have done this from say 10,000 to 40,000 times (according to 
your respective ages and predilections about changing your shoes often) 
but the experience of tying your shoes has probably never called to 
mind any sensation, image, or idea of the backward body-bend in one 
person in 1,000. Mere sequence with no fitness or belonging has done 
little or nothing, (p. 19.) 

On the other hand, Thorndike holds all the more insistently 
to his basic Law of Effect, claiming that any action which fos- 
ters the life processes of the neurons is thereby "stamped in" or 
learned and anything which has disagreeable consequences is 
eliminated. In an analysis of this concept, Cason 24 endeavors 
to show its untenability by pointing to the fact that numerous 

23 His less clearly-defined principle of identifiability winch is now first specified is very 
similar to his opponents' notion of degree of structuration or "goodness" of gestalt, op. cit. f 
pp. 8790. 

24 "Criticism of the Laws of Exercise and Effect," PsychoL Review, 1924, 31, 397-417- 



198 GESTALT PSYCHOLOGY 

associations, as in reading, are formed without the accompani- 
ment of any feelings, either pleasant or unpleasant. In con- 
sidering the question of the elimination of useless movements, 
it is pointed out that the rat in the maze cannot distinguish be- 
tween the "successful" and the "unsuccessful" turns at the time 
he makes them, and, of course, a human subject in similar situa- 
tions has the same difficulty. "One turn looks like the other to 
the rat at the time he makes it,- and it is difficult to understand 
how the successful turn could have been 'stamped in' and the 
itnsuccessfiil turn 'stamped out' at the time they were made." 
Moreover, it is highly improbable that readiness or unreadiness 
of certain specific reaction-arcs to conduct can influence the sat- 
isfaction of the whole organism, for whatever entities pleasure 
and pain may ultimately be proved to be, they indubitably affect 
the total personality and not just a limited function of it. 

Humor 

The Gestalt position readily provides an additional hypothesis 
to the long list of theories of humor by calling into action its 
handy concept of insight. The sudden laughter which accom- 
panies "seeing the point" is an obvious feature of kinship be- 
tween the two. According to Maier, 25 who has developed this 
relation most explicitly, humor is essentially like insight and dif- 
fers from reasoning or understanding only in so far as it con- 
tains the elements of the ridiculous. A pun is the clearest 
example of humor created by displacing an item from one setting 
to another. Kant's intellectualistic version of laughter is the 
nearest of the classical interpretations to the configurationisfs 
schema, and Maier's account may be considered as a moderniza- 
tion of the Kantian theory of incongruity. He says : 

The thought-configuration which makes for a humorous experience 
must (i) be unprepared for; (2) appear suddenly and bring- with it a 
change in the meaning of its elements; (3) be made up of elements 
which are experienced entirely objectively (no emotional factors can 

t 2S A "Gestalt Theory of Humor," Brit. J. PsychoL, 1932, 23, 69-74. Conceptually the 
main difficulty of this highly plausible theory is that while the ridiculous is always the 
incongruous, the converse cannot be held. Hence, the numerous provisos attached. 

Perhaps Emerson caught a glimpse of this notion in his*"ssay on the Comic," for he 
writes, "Separate any ^particular object from the connection of things, and contemplate it 
alone, standing there in absolute nature, it becomes at once comic." 



RESEARCHES ON INSIGHT 199 

be part of the configuration) ; (4) contain as its elements the facts ap- 
pearing in the story, and these facts must be harmonized, explained, 
and unified; and (5) have the characteristics of the ridiculous in that its 
harmony and logic apply only to its' own elements. 

In Harrower's study, 26 the Gestalt view of this problem is 
supported by an appeal to experimental data. She assumes that 
one can find structured wholes in non-perceptual content, and 
that such wholes possess the same or similar functional proper- 
ties as the corresponding perceptual structures. In her judg- 
ment a joke is just such a non-perceptual unit. A joke, told 
among other casual items, stands out as a unity with its own 
particular flavor : it has unmistakable boundary lines as to its 
beginning and end. Further, no one who has been told two 
jokes, and is then asked to make a natural division in his flow of 
consciousness over that period, would take half of one joke as a 
unit, over and against one and a half jokes as the other. 

Harrower's method was a derivative of the one earlier em- 
ployed by Flach, 27 Arnheim (footnote 5, page 254) and Usnadze 
(footnote 10, page 147). One of Flach's subjects, on being 
asked to express Bach's music, Gothic architecture and a "com- 
promise" by means of drawings, had sketched the accompanying 
diagrams in the order given : 



/wwx 



Harrower's technique was to read a joke to her subjects and 
show how it might be diagrammed ; the diagram was not meant 
as a pictorial representation of any detail or part of the joke, 
but was an expression of the distribution of the meaning as a 
whole. Then a number of jokes were read and a page of mis- 
cellaneous symbolic drawings placed before the subject with 
instructions to select that drawing which seemed to approximate 
a given joke. It is to be understood, of course, that it is not the 
geometrical figure, drawn in pencil on paper, that can bear any 

26 "Organization in Higher Mental Processes," Psychol. ForscJi., i933> *7> 5 6-120. 
- 7 "Ueber syrnbolische Schemata Im produktiven Denkprozess," Arch. /.a. ges. Psy- 
chol., 1925, 52, 438 ff. 




200 GESTALT PSYCHOLOGY 

resemblance to the joke, but it is the subject's "experience," aris- 
ing from the diagram, which may have similar properties to the 
experience resulting from listening to the joke. 

This illustration is typical of the preliminary procedure used : 
The old pawnbroker was on his death bed, his entire family had 
gathered there in the little back bedroom of the pawn-shop. At 
last the old man spoke : 

s/ 

Is Momma here? > 

Yes, Poppa. :' 

Is Rifkahere? 

Yes, Poppa. 

Is Max here ? 

Yes, Poppa. 

Is de whole family here ? 

Yes, Poppa. 

Den who in heaven's name is out front minding de shop ? 

The regular increasing steps and the final outburst will be 
seen as the properties intended to be conveyed by the diagram. 
When the companion diagram was interspersed among a number 
of confusion stimuli, a correct choice was made in roughly 73 % 
of the cases, a value obviously well beyond chance. 

To study the closure effect, Harrower read aloud fifteen 
unfinished jokes to twenty-two subjects who were told to put 
down on paper any completion that occurred to them. When 
these spontaneous completions were analyzed, the most outstand- 
ing characteristic was the fact that there may be two answers 
of an exactly opposite meaning which are given as the closure 
of the structure. Here is a concrete example : 

The Prisoner in Court: But your worship, I wasn't going 50 

miles an hour, nor 40, nor even 30 ... 
The Judge: 

For the judge's answer two very different statements were 
obtained the one continues the descrescendo already indicated 
in the unfinished situation, viz. : 

"Well, you'll be going backwards soon." 
The other type of answer appears in : 

"No, I suppose you were going 70 miles an hour/' 



RESEARCHES ON INSIGHT 201 

Diagrammatically, the first alternative is the simpler since it 
merely continues the "direction" of thought already established 
by the interrupted statement ; but the second alternative is not 
arbitrary because the simple reversal of "movement" is also 
implicit in the situation. The two possibilities are organized 
with respect to some common center. 

Harrower obviously is committed to the view that meaning 
and organization are closely related. The following examples 
are offered to show the possibility of changing from sense to 
nonsense by a re-distribution of emphasis: 

That that is is that that is not is not is not that so that's so. 

It was and I said not so. 

Time flies you cannot they alight at such irregular intervals. 

The first of these sentences is generally read by the na'ive 
observer as : 

That that, is is, that that, is not, is not, is not, that so, that's so. 

and he pronounces it utter nonsense (or a passage from Gertrude 
Stein !). The correct and meaningful structure reads : 

That that is, is ; that that is not, is not. Is not that so ? That's 
so. 

The second is almost invariably read as : 
It was, and I said, "not so." 

which, though not sheer nonsense, is at least not a completely 
meaningful sentence. The correct structuring yields the 
meaning : 

"It was 'and/ I said, not 'so'." 

In the third sentence we get : 

Time (emphasized) flies, you (emphasized) cannot, they alight 
at such irregular intervals. 

whereas meaning demands : 

Time (verb) flies (noun), you cannot, they alight at such irregu- 
lar intervals. 

Clearly, the organizing power of punctuation and intonation 
is the decisive factor in these cases. 



CHAPTER 13 
ACTION, EMOTION, AND WILL 

Any system of psychology which pretends to be valid for all 
manifestations of mental life must demonstrate its claims with 
approximately equal effectiveness in every field whose problems 
can be investigated by way of the experimental method. It is a 
commonplace among many observers of the rapid growth of 
Gestalt psychology that in spite of a broad range of researches 
in the animal and abnormal regions of the subject most of the 
positive contributions center about the old some would say 
hackneyed theme of perception. To some extent this tend- 
ency has been inevitable since the revolutionary character of any 
new doctrine can best be exhibited in the extension and re-inter- 
pretation of old problems, just as the Einstein theory acquired 
significance because of its novel mode of dealing with such 
classic problems as velocity, time and space coordinates, etc. 
Moreover, the defenders of the Gestalt position assert that while 
configurationist principles are most easily derived from an exam- 
ination of perceptual behavior their studies in this area should 
never be considered as narrow researches in perception as such. 
For them, the principle is the important thing, and the particular 
type of organismic activity in which it is revealed is a secondary 
concern. Nevertheless, toward the end of the 'twenties, a defi- 
nite increase in experimental productivity in the fields of the 
emotions and the will occurred, partly as an answer to the 
implied suggestion of neglect of the more complex activities, 
and partly to prevent the topic from being pre-empted by the 
dominant Geistesivisscnschaft school. 

Lewin's Theory o Psychic Tensions 

The leader in this distinct but fundamentally related division 
of Gestalt psychology is Kurt Lewin. Like most of his group, 

202 



ACTION, EMOTION, AND WILL 203 

he was a pupil of Stumpf at Berlin. His academic career was 
interrupted by service as captain in the German army during the 
war, during which he maintained sufficient composure to prepare 
a unique article on the figure-ground phenomena present in the 
camouflaged scenery of trench warfare ! Lewin's early papers 
indicate his growing interest in the problem of the relation of the 
will to the memory functions, but his theoretical position was not 
mature until the publication in 1926 of a brilliant programmatic 
essay entitled, " Purpose, will and need; With a preliminary 
account of psychic forces and energies and the structure of the 
mind." 2 To other than German eyes much of his discussion is 
irrelevant to the main problem of the title, but since many of 
these digressions are illuminating and interesting in their own 
right they will be included in the following exposition. 

For Lewin, the difficulties in the way of psychological re- 
search result less often from obstacles of an experimental and 
technical nature than from a failure to develop the appropriate 
theories. Primarily the psychologist must be guided by his 
theory for without it all his investigation^ activity is blind and 
meaningless. On the other hand, the experimentalist must 
prove the correctness of his hypothesis by an appeal to a psychic 
event occurring in a perfectly concrete situation and at a definite 
moment of time within a certain individual and a prescribed 
environment. However, every step forward depends first upon 
progress in the sphere of theory. In both theorizing and experi- 
menting the tension between the striving after comprehensive 
views and the grasping of concrete events in all their variety may 
be considered the basic phenomenon of the scientific life. The" 
omnipresence of this tension allows us to say that experiment 
and theory are the poles of a single dynamic whole. 

It is Lewin's opinion that most experimental inadequacies are 
due either to errors in theory or to failures to carry the theory 
through in a sufficiently radical manner, i.e., to select extreme 
instances in which the theory must apply if at all. This last 
defect is the commoner and more serious one. Certain basic 
assumptions of science, too, must never be surrendered in the 
apparent obstacles, such as one's faith in the reign of law 



1 "Vorsatz, Wille nnd Bediirfnis," Psychol. Forsch., 1926, 7, 330-385. 



204 GESTALT PSYCHOLOGY 

among mental events or the infinite possibilities of experimenta- 
tion. The first postulate has often been limited quantitatively 
or qualitatively by good authorities ; e.g., rejected for the higher 
mental processes or excluded during headache ; and even uncom- 
promising natural scientists denied the possibility of an ex- 
perimental psychology of will on the ground that one cannot 
artificially evoke vital decisions, since the investigator cannot 
interfere with the occupational and family life of the subject. To 
this Lewin retorts that one could just as well have held this objec- 
tion against the early studies of electrical phenomena, on the 
ground that it was impossible to produce a genuine thunderstorm 
with the weak replicas of the laboratory ! This is a wholly false 
conception of the nature of experiment which is not at all de- 
signed to recreate the natural world any more than art aims at 
photographic fidelity to the objective original. In fact, it is not 
the rare or the extraordinary which presents the valuable or even 
severe challenges to research, but rather the commonplace and 
the banal. 

The first clear-cut indication of Lewin's affinities with con- 
figurationism appear in his conception of an "action-total, 3 ' 
which he proposes to substitute for the microscopic reductions of 
emotions to blends of feeling-tone and sensation-masses, or the 
equivalent treatment of choice and other volitional phenomena. 
An action-total occurs in a field (note the kinship with the figure- 
ground contrast in perception) and typically has a beginning and 
end, which serve as "contours" of the event. The termination 
is marked by such unit-actions as making a period, sighing in 
relief, or emphasizing dropping of the finished task, all of which 
^rves to isolate it as a relatively distinct whole from its temporal 
"field." The reality of these action-totals may be seen from 
the differential effect of "imbedding" the same action in a differ- 
ent setting; e.g., the man who is too tired to talk before an 
audience may use the same number and kind of words in a con- 
versational group with surprising ease. Similar cases may be 
found in the old pedagogical tricks for getting a child to do 
something which it does not like (such as eating a certain food, 
taking medicine), or for arousing interest in certain things. 



ACTION, EMOTION, AND WILL 205 

The all important topic of the causes of mental events is* 
approached by employing the notion of psychic energies in place 
of the theories which associationism has built upon an adhesion 
basis. These energies or tensions result from some want or 
need, consequently a "tense 33 psychic system is the condition 
precedent to all mental activity, regardless of what course it 
runs, or the magnitude of the energy involved. The unfortu-* 
nate tendency to thrust into the foreground an objection based 
upon pure intensity is a caricature of the method of inductive 
science. Just as little as one investigates the laws of falling 
bodies with volcanic eruptions or tiles which have been blown 
from the roof by the wind, or the principles of hydraulics with 
brooks and rivers, so likewise is it futile to consider as scientifi- 
cally meaningful only those psychological experiments which 
are simple replicas of reality. The practicality of an experiment 
does not rest upon such a foundation but upon the successful 
production of those dynamic structural systems whose laws are 
to be investigated. Neural connections can never be the* 
"causes" of mental events. In order that anything conjoined " 
can move and give rise to a happening or process, kinetic energy 
must be released a fact which holds even for purely mechanical 
systems. Hence, for every mental occurrence one must inquire : , 
Where do the causative energies originate? However, this 
change of accent does not deny that associative couplings exist 
or that they are important when they do occur as in verbal 
groups. It is simply denied that association as such can ever 
be a motor of mental activity. 

Lewin's use of familiar scientific terms like energy, tension, 
and system does not imply a reduction of mental phenomena to 
the data of physics, but simply results from the general require- 
ments of a dynamic logic whether employed in economics or 
elsewhere. The forces of which he speaks occur in the psychic 
field and not in the physical. To clarify their nature one needs 
to examine the circumstances of their operation. At the very ! 
start, it is obvious from many facts that the amount of energy 
involved in a given psychic process does not flow from the 
momentary perception. The Umweg or indirect and round- 
about means, which even an animal or young child uses to reach 



206 GESTALT PSYCHOLOGY 

a goal, is a case in point, since the objective may be literally 
hidden or obscured during much of the solving process. What 
the perception of an object or event does to the organism is best 
stated in terms of four propositions : 

1. It may produce a definite tension in a psychic system which 
immediately evokes a wish or intention which was previously 
non-existent. 

2. An already existing state of tension, traceable to some 
need or purpose or an interrupted activity, attaches itself to a 
distinct object or occurrence in such a manner that this tension 
regains the dominance over the motor system. Such objects 
possess what one may term a "demand or compulsive character" 
or valence (Aufforderungscharakter), i.e., certain things call 
upon us to perform specific acts with them : steps lure the two- 
year old to go up and down them ; doors stimulate the opening 
or closing act ; a dog's furry coat evokes a desire to stroke him 
etc. 2 

3. These "demand characters" function as field forces in the 
sense that they influence, i.e., guide or steer other psychic proc- 
esses, especially the motor factors. 

4. Certain actions, partially initiated by "demand characters" 
lead to satiation processes or the attainment of desired goals, 
after which a condition of equilibrium of the basic system on a 
lower level of tension ensues. 

, Stresses always remain within the perceptual field and, there- 
fore, the most frequent way of transforming this organization 
is through action. If one sees an "attractive" object, one is 
actually attracted by it and tends to approach it, i.e., the organi- 
zation of the psychophysical field contains a pull which is relieved 
by the movement of the body. Thus, action is directly adapted 
to experience because it is regulated by it in a perfectly natural 
way. 




PI 

pat; 




ACTION, EMOTION, AND WILL 207 

Successful operation with these concepts is subject to at least 
one major caution one must not work with vague totals but 
always with specific concrete sub-totals. When this is done, 
mind becomes divided into natural structures such as layers, 
spheres, and complexes ; however, on all such occasions the pri- 
mary problem is to determine where wholes are present and 
where they are absent. When found, action-wholes exist co- 
extensively with the presence of some sort of tension. 3 This 
tension confers upon these totals whatever internal unity they 
possess and they are not held together by any associative "ce- 
ment" as the older theories would have it. Lewin' s favorite 
illustration of the significance of the former, and the futility of 
the latter, explanation runs as follows : Suppose one leaves home 
in the morning with a letter in one's coat-pocket which one plans 
to drop into the nearest mail-box. As soon as the first mail-box 
comes into sight, the "demand-character" which it possesses for 
the organism (by virtue of the existing tension set up by the 
mailing purpose) leads to the proper insertion of the letter. 
Thereafter, one may pass a dozen letter-boxes without experi- 
encing any impulse to repeat the act. On associationist princi- 
ples, having performed the act once should make one more likely 
to do it again. As a matter of fact, the release of tension seems 
to be the principal determinant of the action since it is obvious 
that one is really less likely to repeat the act upon seeing a second 
mail-box because the necessary inner condition is now absent. 
Such "determining tendencies"- had already been recognized by 
Ach as a necessary supplement to simpler associationism, but 
Lewin goes one step further in tracing their origin to a pur- 
posive act, itself not reducible to any forms of association. 

In this system of ideas, the volitional life develops from more 
primary needs. The obvious parallelism between the operation 
of a genuine want and the after-effects of an intention or pur- 
pose justifies Lewin in speaking of the latter as qmsi-netds. 
These quasi-needs are far more numerous and flexible than the 
original and primitive needs and substitute satisfactions arise 
much more frequently in conjunction with them. The young- 

8 It is interesting that the words which we use to express inner conditions, such as 
"tense/* "oppressed," "upset/* etc., are largely drawn from descriptions of external 
objects. 



208 GESTALT PSYCHOLOGY 

ster, e.g., who dares not announce the departure of the train can 
at least mimic the "All aboard !" cry of the station officials. Or 
a boy is attracted to a puppy, wishes to pet it, but is uncertain 
and afraid. An older girl takes the puppy in her lap and the 
same conflict ensues ; suddenly, the boy turns around and strokes 
the girl's head instead of the dog's ! 

In general, one may say that instinctive acts 4 (needs) are 
those which function directly in accordance with the field forces 
as typified by the presence of natural "demand-characters" 
whereas "willed 5 ' acts (quasi-needs) are those which require a 
purpose or intention in advance. Most, if not all, laboratory 
studies fall in this class because of the necessity of inducing the 
subject to agree to obey directions. It is of more than pass- 
ing interest to notice that phenomenally, the understanding of 
instructions can hardly be distinguished from the intent to 
execute them ! 

A major pre-supposition of Lewin's investigations is the 
necessity of surrendering finally any attempt to deal with ff the 
will" as a unitary psychic datum or area of research to be con- 
trasted with, and sharply delimited from, "intelligence/' "in- 
stinct," "memory," etc. Actually the term "will" embraces 
many very distinct kinds of problems such as decision, intention, 
self-control, separation from the environment, concentration, 
persistence, unified or inharmonious organization of volitional 
goals, and many others. Lest, however, the word sink into 
the disrepute which has befallen the designation "instinct," he 
proposes to substitute the conception of "need" to cover both 
problems a step which compels him to make his peace with the 
reigning theories of native behavior. Typically, the theory of 
"drives" views them as permanent urges which confer upon the 
organism a tendency to respond in certain directions. How- 
ever, Lewin maintains that the effectiveness of these drives 

4 Lewin's position is different from McDougall's on this point. Instead of champion- 
ing the reality of a limited number of inner "drives," he merely holds that the organism 
has definite possibilities of reacting when certain conditions are present. We attribute 
different capacities to iron it can fall, be electrified, etc. without assuming the presence 
of mysterious impulses in these directions. For Lewin there is no independent source of 
energy inherent in a purpose any more than there is weight in a rock or in a table. In- 
stead, purpose is a field property or a structured field of energy. A high energy potential 
exists in a field in reference to a low. The goal and the line of resolution are both simul- 
taneously conditioned by the immediate situation. 



ACTION, EMOTION, AND WILL 209 

hinges upon the presence of acute states of tension, to explain 
which the notion of "need" is more serviceable. 5 

Value and Meaning 

Psychologically, any discussion of human "values" is impos- 
sible without reference to the affections. 6 Some percepts pos- 
sess "initial" character, i.e., that which is perceived is "bad" ; 
others have an "end" character, i.e., the perceived item is good. 
It is these dynamic features which we normally call "values." 
These affective qualities of perceptual structures are just as fun- 
damental as any, and it is an error to consider them as purely 
subjective creations which we "project" into other objects. The 
upward urge of a Gothic cathedral is a property of the cathedral 
and not of my ego, much as it may provoke in the ego a similar 
urge. A thing is equally as "weird" as it is "black," in fact, 
even more so, since the first influence is stronger. "Weird" 
simply means that the thing serves as an "initial" phenomenon, 
from which one must move away rather than toward. Con- 
versely, a desirable object is one which is felt as a goal to be 
striven for, the goal having a quasi-physical meaning of closing 
a gap. For this reason, every process directed toward some end 
can be termed "valuational" or meaningful, the two being in- 
separable. Value and being, consequently, are equally real and 
objective. Ideals of all sorts are simply end-goals of dynamic 
activity, i.e., figures or situations which are maximally pragnant, 
symmetrical, and stable. With this recognition, Dilthey's cleav- 
age between an explaining and an understanding psychology dis- 
appears. Value and meaning are not creations of pure reason 
but are rooted in the nature of the world itself. It is through 
this avenue of reflection that Gestalt aims to break down the 

s Gestalt theory could be much clearer on the subject of instinct than it is. It rejects 
flatly McDougalPs classification and yet acknowledges the reality of native urges. If by 
instincts we mean the systems which the organism has at birth, configurationism certainly 
has a place for them, because the organism at any time is made up of physiological systems 
that are responsible for its conduct. These systems are functions of what they were at 
birth or as a conse'quence of maturation, and of the changes that may have been, produced 
in them by use (changes commonly referred to as learning). If this approach be correct, 
then instead of two sharply divided classes, instincts and habits, we really have a con- 
tinuum from maximally modified systems to those which have remained relatively un- 
altered. See the exceptionally intelligent discussion of this hotly-contested issue of the 
past decade "by Miss Drury, "Can Gestalt Theory Save Instinct?*' in J. General Psychol, 
1931, 5, 88-94. 

e This paragraph is based upon Koffka's discussion in the Lehrbuch der 
ed. Dessoir, Berlin, 1925, p. 600. 



210 GESTALT PSYCHOLOGY 

venerable opposition between judgments of fact and judgments 
of value, which seemed so final to our immediate forbears. 7 
The normative sciences logic, aesthetics, ethics which are 
impossible on mosaic principles thus win a sanction which theo- 
retically places them on an equal footing with the descriptive 
ones. 

In order to appreciate this conclusion, we must realize that 
from the configurational point of view, perception and action 
cannot be separated ; the one flows over into the other. Con- 
sequently, there is no break in continuity between the phenome- 
nal and behavior patterns. This is the standpoint of isomorph- 
ism, i.e., the view that the form of mental events is the same as 
physical It is the fundamental thesis of Kohler's "Physische 
Gestalten." From the Gestalt view on methodology and ex- 
perimentation, it is first necessary to understand the effect of 
the world upon the organism ; hence, an introspective report may 
throw light upon explicit behavior. Since the configurational 
concepts are not limited to static events in the nervous system, 
perception may be regarded as a stage preparatory to action. 
If an object appears "friendly," this appearance implies my re- 
action to it as well as my perception of it. 

Reward and Punishment 

In the last decade or so, Lewin and his pupils have been en- 
gaged in extending the conquests of configurationism to those 
deeper and less accessible areas of organismic conduct repre- 
sented by the topics of action, emotions, and the will. 8 A super- 
ficial view of their work is apt to leave the impression that the 
characteristic procedures which Wertheimer and K5hler have 
so successfully applied to the fields of perception and learning 
are abandoned in favor of a novel and less convincing approach 
in the more complex functions. As a matter of fact, the ideol- 
ogy is preserved essentially unaltered but merely extended in- 
geniously to newer enterprises. 



Ki- ^ tu . m > " AIles , wa s ist Bat Sinn" (All that exists has meaning), was 

probably influential m establishing this view of Gestalt, 

mmnc'^Sl i; igina !| the series of about two dozen papers which Lewin has edited with his 
pupils s .aid M ; termed Unteratchnng** sur Handling s-und Affektpsychologie. Presumably 
this union indicates a belief in the essential identity of the problems of action and affection 



ACTION, EMOTION, AND WILL 



211 



Since Lewin implies that the psychologist's task is exhausted 
as soon as he has fully described the dynamic processes occur- 
ring in every concrete individual situation, it is only inevitable 
that he should depend largely upon pictorial diagrams to clarify 



B 



A 



D 



V 







(-'9) 



(30) 




Figure 29. Lewin's Method of Depicting the Action of Field Forces in 

Conduct 

The topology of the situation is: A child (C) is attracted by a doll (the plus indi- 
cating positive pull; the resulting tension is represented by the vector (arrow Va). 

Figure 30. Same as Figure 29 but Complicated by the Introduction of a 

Barrier (B) 
The doll e.g., may be within the child's vision but out of its reach. 

Figure 31. Same as Figure 30 but with a New Vector toward the Goal 
Circumventing the Obstacle 

When a direct attack upon a problem, becomes impossible, a roundabout solution 
(detour or "Umweg") is undertaken. 

his position. Perhaps the best introduction to the specific 
uniqueness of his thought is found in a recent monograph 9 on 
reward and punishment, which presents in simple and clear form 

9 Die psyckotogische Situation bei Lohn und Strafe, Leipzig, 1931- -For a sympa- 
thetic criticism of Lewin's concept of vectors, see an article with that title by E. C. Tolman 
in/. General PsychoL, 1932, 7, 3~is; Lewin's answer entitled, "Vectors, Cognitive Proc- 
esses, and Mr. Tolman's Criticism," appears in ibid., 1933, 8, 318-345. 



212 



GESTALT PSYCHOLOGY 



his matured views through the instrumentality of a qualitative 
mathematical topology. 

. The theme of this work is that the actual behavior of the 
organism depends in every case both upon the individual's char- 



T 




V 



V 







Vr 



O) (33) 

Figure 32. Introduction of a Negative Vector 

An unpleasant task (T with a minus sign) repels the child, as shown by the reverse 
direction of the arrow. 

Figure 33. Consequences of Introducing Disciplinary Measures 
The negative "valence" of the threat of punishment pushes the child back in the 
direction of the task. He is now influenced by two vectors. 

acteristics and upon the momentary structure of the existing 
situation. If we wish to portray what occurs when a child is 
led to perform a simple "interesting" act, such as playing with 
its doll, then Figure 29 results. The plus sign indicates that the 
doll has a positive appeal (Aufforderungscharakfer) to the child, 



ACTION, EMOTION, AND WILL 



213 



and if this attraction be sufficiently strong relative to the other 
psychic forces present in the field, an action of the child in this di- 
rection ensues (represented by the vector a). If, however, 
physical or social obstacles are placed in the way, a psychologi- 
cal barrier is established as in Figure 30. This alters the posi- 
tional relationship between the child and the objective with the 
result that a new vector in the direction of the goal appears the 
prototype of all Umwege or detours. Behavior of this kind has 






Figure 34. A Conflict Situation Provoked by Two Equally Attractive Goals 
This is the classical dilemma of Buridan's ass. 

a "purposive" cast and is dominated by a "natural teleology." 
Where a disagreeable task is imposed, such as performing a 
home assignment, a vector of rejection or repulsion becomes 
operative. 

The application of reward and punishment, however, alters 
the total-situation by changing the "valence" of the goals or 
introducing a new vector. This means that in some way a field 




Figure 35. An Ambivalent Conflict Situation 
The mental state of "temptation" with both positive and negative vectors operative. 

force of sufficient strength to counteract it must be set up in 
opposition; to the vector of repulsion. Topologically, then, a 
threat of punishment evokes the situation of Figure 33. 

The "negative appeal" of the task is dynamically akin to a 
barrier since it blocks approach, and this, coupled with the pres- 
ence of another negative character the penalty creates for the 
individual a situation involving conflict. Psychologically, con- 
flict is distinguished by the presence of simultaneously antago- 



214 



GESTALT PSYCHOLOGY 



nistlc and approximately equally powerful forces. Where the 
vectors involved are both negative, Figure 33 applies; where 
they are both positive, as with Buridan's ass who starved be- 
tween two equidistant bundles of hay, Figure 34 is pertinent; 
and if the situation is ambivalent (e.g., a child wishes to stroke 
a dog but is mildly afraid, or to eat " forbidden fruit"), then 



T 



GH- 



Figure 36. How the Child Tries to Go out of the Field in Which 

He is Caught 

Torn between two negative valences he seeks escape by following the 
resultant vector (Vr). 

Figure 35 offers the best symbolism. Obviously, a conflict sit- 
uation tends toward resolution since it involves a labile equilib- 
rium ; hence, the slightest lateral displacement of C must pro- 
duce a strong resultant V r perpendicular to the direction : punish- 
ment task (Figure 36). The child then seeks to "go out of 
the field," avoiding both the unpleasant duty and the penalty. 



ACTION, EMOTION, AND WILL 



215 



Going out of the field may involve literally running away or 
hiding, or busying one's self with a more agreeable but irrele- 
vant occupation against which no one can legitimately protest. 
Such behavior is inevitable in accordance with the topology and 
the field- forces of the situation, unless special measures are taken 
to create an outer barrier which prevents flight. The function 
of the barrier is to forbid escape until either the task or the pun- 
ishment is accepted. The simplest barriers are, of course, phy- 
sical, as when one confines a youngster to his room until he obeys, 
but they are often of a sociological nature, such as the control 




Figure 37. Tension at its Climax 

The presence of the outer barrier (the heavy lines 
at B) forbids flight and the organism is completely 
encased by hostile forces. This is probably the origin 
of most prison psychoses. T and P still stand for task 
and punishment. 



exercised through the "bad black man" or "God," from whose 
omniscience one cannot flee (cf. Francis Thompson's "Hound 
of Heaven" for an impressive poetic testimony of the strength 
of such "barriers"). At a more mature level, the appeal of 
group morality becomes effective ("unladylike!" "Be a man!") 
by virtue of a hierarchy of goals and values which the individual 
personally accepts. An appeal of this sort contains one of the 
most potent of threats the danger of exclusion from the group. 
When an outer barrier is rigorously held (represented by the 
heavy lines) , "going out of the field" is impossible and a general 
heightening of tension is the consequence. (Figure 37). At 
every point in the field the child experiences an increase in pres- 
sure. A new negative vector Vb traceable to the barrier is now 
effective and tension, which is always due to the presence of op- 
posed vectors, dominates the situation (symbolized in the 



216 GESTALT PSYCHOLOGY 

drawing by the cross-hatching throughout the field). Under 
these circumstances the child may execute the command, in which 
case the vector coming from the punishment proves itself 
stronger than the counter-forces ; or he may accept the penalty 
should the reverse hold; or he may direct his energies against 
the barrier through temper tantrums, flattery, deception, or de- 
fiance, in the hope that it may be lifted through these means. 
Should the barrier be firmly maintained and a sufficiently strong 
tension throughout the entire field promises to endure, suicidal 
tendencies may arise since self-destruction appears to be the only 
way of "going out of the field/' 10 

Whenever a child is threatened with punishment through 
adult intervention there necessarily results a situation in which 
the two are opposed as enemies. The child is normally caught 
within the net of more powerful field forces exerted by the adult, 
from which escape is often possible only through transference 
to a "plane of irreality." This region of psychic space (typified 
by dream and fantasy) is characterized dynamically by an unus- 
ual fluidity due to the absence of rigid barriers and the shifting 
nature of the boundaries between the self and the environment. 
If the plane of reality contradicts too sharply the wishes and 
needs of the child, substitute satisfaction will be found on the 
unreal level of play or daydream. This transition is simply an 
internal variety of "going out of the field." Revenge fancies 
and "conquering hero" episodes enable the actual weak field 
force of the defeated child to enlarge into the dominating in- 
fluence on the level of imagination. 11 

The concrete resemblance of these situations to those treated 
by the psychoanalysts, especially Adler, is plain, but the Gestalt 
interpretation, as usual, is wholly different. It takes the fol- 
lowing form : The solution of intellectual tasks admittedly al- 
ways requires a certain tension, a vector in the direction of the 
problem. Now conflict situations with general tension, of the 

10 The suicide of the Russian commander when he realized that his army had been 
trapped in the Masurian marches by Hindenburg deserves to become a classic illustration 
of this outcome. 

11 Lewin illustrates this process at great length from an incident in the boyhood recol- 
lections of Tolstoi. His tutor, St. Jerome, had locked him in a lumber room for some 
boyish prank. The story appears in Tolstoi's Complete Works, Vol. i, "Boyhood," Chs. 
1415, translated by Wiener, Boston, 1904. 



ACTION, EMOTION, AND WILL 



217 



kind just described, are particularly unfavorable to rational so- 
lutions. The essence of problem-solving consists in a transfor- 
mation of the structural relations of the situation so that a new 
organization or arrangement emerges. A prerequisite to the 
occurrence of such a shifting demands that the individual be able 
to survey the field as a whole. This decisive reconstruction of 




R 











(39) 



J 




(.40) 



Figure 38. The Reward Situation 
The reward (R) normally can be approached only by discharging the task. 

Figure 39. Same as Figure 38 with a Barrier (heavy border at B) Added 
to Insure the Performance of the Task if the Reward is to be Gained 

Note that in this situation the child stands outside the barrier rather than within 
it as in the punishment situation symbolized by Figure 37. 

Figure 40. Combination of Reward and Punishment 

The conflict situation here_ portrayed is typical of "examination fever" and is 
probably the cause of much pupil-teacher antagonism. 

/ 

the field postulates the capacity to stand "above" the situation, to 
have a minimum amount of personal detachment, for only then 
does one see the general inter-connections and not just the spe- 
cific details. However, if the child is in a conflict situation in- 
volving strong tension, it is bound to become submerged within 
or "under the situation, deprived of the possibility of survey 
and perspective a circumstance decidedly unfavorable for quiet 
intellectual solutions." 



218 GESTALT PSYCHOLOGY 

In conditions involving reward rather than punishment, the 
subjective state gives less evidence of friction. Here it is the 
intention to create a positive vector of sufficient intensity to 
overcome the negative repulsion of the disagreeable task ; hence, 
it is placed in the same direction but "behind" the problem (Fig- 
ure 38). Functionally, the unpleasant requirement constitutes 
a barrier lying between the person and his goal (R). Since 
there is always the danger that the child will make a circuitous 
path to his goal and thereby avoid the problem, a barrier must 
be erected so that access to the reward can occur via the task 
only (Figure 39). The main diff