PSYCHOLOGY LIBRARY EDITIONS:

PERCEPTION

Volume 21

THE PERCEPTION OF
CAUSALITY
THE PERCEPTION OF
CAUSALITY

ALBERT MICHOTTE
First published in English in 1963 by Methuen & Co. Ltd
This edition first published in 2017
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa
business
© 1963 Methuen & Co. Ltd
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or
registered trademarks, and are used only for identification and
explanation without intent to infringe.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library

ISBN: 978-1-138-68824-7 (Set)

ISBN: 978-1-315-22895-2 (Set) (ebk)
ISBN: 978-1-138-69839-0 (Volume 21) (hbk)
ISBN: 978-1-138-69842-0 (Volume 21) (pbk)
ISBN: 978-1-315-51905-0 (Volume 21) (ebk)

Publisher’s Note
The publisher has gone to great lengths to ensure the quality of this
reprint but points out that some imperfections in the original copies
may be apparent.
Disclaimer
The publisher has made every effort to trace copyright holders and
would welcome correspondence from those they have been unable to
trace.
The Perception of Causality

A. MICHOTTE
Emeritus Professor of Psychology
University of Louvain

LONDON
METHUEN & CO L TD
36 ESSEX STREET WC2
 Originally published under the title
La Perception de la Causalite
by the Publications Universitaires de Louvain
Foreword by R. C. Oldfield, M.A.(Cantab.),
Professor of Psychology, University of Oxford
Commentary by T. R. Miles, M.A.(Oxon),
Lecturer in Psychology, University College, Bangor
English translation by T. R. Miles, M.A.(Oxon)
and Elaine Miles, M.A.(Oxon)
This edition first published in 1963

Printed in Great Britain

by Hazell Watson & Viney Ltd
Aylesbury and Slough
Catalogue No. 2/6424/10
 Foreword
Visitors to Professor Michotte's laboratory - since the last war alone
they must have been numbered by the hundred - get the best oppor-
tunity there can be of grasping the essence of his work. Perhaps not all
visits to laboratories are as profitable as some visitors seem to think.
But anyone fortunate enough to go to Louvain finds himself welcomed
by a family which has all manner of exciting things going on. He arrives
a guest but soon becomes a participant. Unfortunately not everybody
can get to Louvain. Some who have not been able to, have heard, if
briefly, about Professor Michotte's experiments from his lectures at
international meetings, and through his visits to other laboratories.
Even so there must be many whose acquaintance is at second-hand. For
these, as for those already acquainted who wish to study his work in
more detail and at greater leisure, the published account is indispensable.
It must be admitted at once that this is not an easy text, and to some
people may have seemed the more forbidding through being written
in French. For this reason people in psychological laboratories almost
everywhere ought to be extremely grateful to Tim and Elaine Miles, who
undertook the very difficult and arduous task of producing an accurate
and scholarly translation. To do so needed a mastery of special topics
ranging from the calibration of rotating apparatus to seventeenth-century
philosophy. Their qualifications - Mr Miles is a psychologist whose
competence in philosophy is at a professional level: Mrs Miles' uuiver-
sity studies were in languages - and their conscientiousness are a
guarantee that this version is not the product of hasty enthusiasm and
incomplete grasp of substance and context, but will stand the test of
time.
Michotte's approach to psychological problems is an individual one,
and readers to whom it is new may perhaps be helped by a brief sketch
of the sequence of influences and interests which led him to undertake
the experimental study of the perception of causality which is the sub-
ject of this book.1 Mter a distinguished career as a student, first in
1 I have drawn largely on Professor MICHOTTE' s Autobiographie (Editions
Nauwelaerts, Louvain, 1954), a somewhat amplified French version of the
contribution which appeared in A History of Psychology in Autobiography, Vol.
Ill, Clark University Press, 1952. But I am specially grateful to Professor Mich-
otte, who has himself amplified certain points in private correspondence.
vi FOREWORD
philosophy, then in biology - his first research publications were in the
micro-anatomy of the nervous system - he paid an extended visit to
Wundt's laboratory. Perusal of some of Binet's writings, however,
suggested to him that the work of Wundt, 'though admirable, was in-
sufficient', and 1907-8 found him at Wiirzburg. Returning to Louvain,
he continued his task of building up the psychological laboratory, in
which he was much encouraged and effectively supported by Cardinal
Mercier. But he was able to combine this with sustained experiment on
memory, especially that form of it in which logical structure plays a part.
This phase came to an end with the First World War. Not long after the
end of it, he tells us, appreciation of the cogency of at least some aspects
of the behaviourist argument, together with criticisms levelled by
Wundt, Titchener, and others at the Wiirzburg School, disrupted his
confidence in an approach which leaned as heavily and essentially on
introspection as did that of Kiilpe's associates. From then on his experi-
mental work came to rely on subjects' introspective reports only as
regards the presence or absence of an impression or phenomenon. Even
this resource he eschewed in the work he carried out during the nine-
teen twenties. In his studies of movement during this time his outstand-
ing flair for mechanical design, which later so largely contributed to the
success of his work on causality, first found extensive scope. Careful
recording of the movements made by a limb in a simple task, both before
and after practice, allowed Michotte to demonstrate the tendency for
motor patterns to take on progressively simpler and more organised
forms as skill develops, and in several investigations he emphasised a
number of similarities to the Gestalten which were at the same time
being studied in the field of perception. Mter the International Congress
at Oxford in 1923, where he had become acquainted with Kohler and
Koffka, Michotte had, he tells us, become vividly interested in Gestalt
principles. He soon became convinced that these principles apply not
only to static forms, to motor patterns and to simple instances of apparent
movement such as the phi phenomenon, but also to many instances of
complex perceived movement. He was led to meditate, for example,
upon the nature of what we see when we watch a man operating a
machine or using a tool, and concluded that in such a case the complex
of objects and movements forms a whole having specific properties of its
own in the same way as, for example, a square is something specifically .
different from its four sides.
By 1940 Michotte had become anxious to obtain direct experimental
 FOREWORD vii
evidence for a conclusion which hitherto he had been able to support
only by appeal to general Gestalt principles. Choosing a very simple
action, that of one thing hitting another and setting it in motion, he
devised the remarkably elegant method of presenting this situation to
the subject by the rotation of a disc, bearing appropriately painted
coloured stripes, behind a slit. He thereby achieved effective and flex-
ible control of the time, position, and velocity parameters, and in the
facilities it offers for innumerable experiments the apparatus deserves
to rank among the classics of psychological instrumentation. The very
fact, too, that in a sense the objects which appear to move and to collide
are not real objects at all adds a further touch of piquant interest - an
interest which perhaps has not yet been exhausted.
The reader must be left to enjoy the story of how, in a series of
experiments perhaps unmatched in the history of psychology for their
simple forcefulness and sustained collected march, Michotte uncovered
the characteristics and conditions of a variety of types of perceptual
event which in the less analytic mood of everyday life we lump together
under the heading of 'causal phenomena'. His cardinal finding, namely
that the impression of causality is dependent on specific and narrowly
limited spatio-temporal features of the event observed, is clearly of
fundamental importance. We do not, according to Michotte, see one
billiard ball cause another to move either because we intuitively appre-
hend a fact of nature, or because past experience leads us to see the
event in this fashion, but because the spatio-temporal organisation is
such that it directly unleashes this impression in us. Alter the relevant
variables by a small but measurable amount and the impression disap-
pears. But as the story develops it cannot be denied that difficulties
intrude. And the conclusion to which it is brought by this book, first
published in I946, was, as anyone who knows Professor Michotte would
expect, but the starting-point for exploration offresh ideas aroused by it.
To both the difficulties and the new findings Michotte attaches great
importance. If I add some comment about them, I do so in the hope
that, inadequate as it is bound to be, it will not be misleading or wholly
unhelpful.
A number of very interesting considerations arise when we ask
whether the instances in which causality is perceived in the encounters of
Michotte's coloured forms are co-extensive with the naturally occurring
cases in which material bodies move in similar ways. Are there cases of
naturally occurring causal events which cannot be imitated in the labor-
P.C.-I*
viii FOREWORD
atory by corresponding selection of parameters? Are there instances, on
the other hand, where an artificial event in the laboratory gives rise to
an impression of causal action while a similar concatenation of moving
bodies in the real world does not? In fact, both these contingencies arise.
Michotte calls the former negative, and the latter paradoxical cases. By
way of clarifying this situation he introduces an idea, which has other
applications as well, which some people may find as difficult as it is
interesting. This is the concept of ampliation of the movement (Chapter"
XIV). When, for instance, in the Launching Effect there is an impression
of causality in the way object A sets object B in motion, this, Michotte
holds, is dependent on the condition that, phenomenally, A's motion is
extended on to and partaken of by B, quite apart from the de facto dis-
placement which B suffers as a result of the collision. This extension of
motion from the dominant on to the secondary object, and its appear-
ance as a quality of the latter, is what Michotte means by ampliation.
And he is able to distinguish in terms of the position, directions, times
and velocities those cases in which it can be manifest from those in
which it cannot. He is thus enabled to give an account of the negative
and paradoxical cases and to clarify, in terms of the concept, that 'pro-
ductive' aspect of causal events which has often been emphasised by those
concerned with their phenomenology. But the concept of ampliation
itself ought to give rise to some interesting discussion, for its status is
unquestionably a peculiar one. On the one hand, ampliation does not
seem to be directly apparent as such in experience, but seems rather to
be an underlying condition that an experience should possess a certain
character. On the other hand, if it is not immediately given in experience,
it is difficult to see from where it derives status and explanatory power.
My own suggestion about this difficulty (though whether it would
appeal to Professor Michotte I do not know) would be that the status
of the concept of ampliation is somewhat akin to that of 'good continu-
ation' or of 'common fate' in relation to static forms. 'Good continuation'
is not, in one sense, given immediately in experience, but we can discern
it when its character is conveyed to us by someone else. And according
to its presence or absence in a particular figure we are enabled to predict
that this figure will have this or that other property, such, for instance,
as prominence in a complex perceptual field or resistance to mnemic
decay. The concept of ampliation seems to me to have much the same
kind of status. It would be out of place to argue here the general case for
and against concepts of this kind. But there would seem to be scope for
 FOREWORD ix
further exploration, not only of the conditions in which ampliation is
and is not manifest, but also of what, to me at any rate, is an outstanding
paradox, namely that impressions of causality in natural events do not
always tally as regards their constitutive conditions with artificial ones in
the laboratory. Each, as perceived, has only phenomenal status, and it is
difficult to understand why ampliation is not equally and correspond-
ingly a constituent of both.
Another fascinating and challenging conception, which has arisen
particularly in the work undertaken by Michotte and his associates
since this book was completed, is that of amodal perception. 2 Pursuing his
experimental analysis of different types of mechanical phenomena, he
tried to produce the impression of one object compressing itself by mov-
ing up against another - stationary - one. When the leading edge of the
moving object reaches the stationary, its trailing edge continues to
move, and it thus grows shorter along the line of its motion. To Mich-
otte's considerable surprise there was no impression of compression at
all. Instead, there was an absolutely compelling impression that the
moving body retains its original size and moves behind the stationary
one as if it were passing behind a screen. An effect quite bizarre is, in-
deed, produced if the stationary object is omitted from the experiment,
while the first goes through the same evolution as before. In this case a
slit is seen apparently to open in the background, through which the
moving object passes and disappears. A third striking instance of this
kind is the 'Tunnel Effect', in which an object, disappearing behind one
edge of a stationary object, reappears from the other. If the time interval
is longer than would be appropriate to the object's steady passage
through the tunnel, there is a most striking impression that the object
halts inside. This kind of perception, in which the total situation pre-
sented includes features which lack any specific sensory determination,
and to which in the nature of the case there can be no corresponding
physical stimulus, Michotte calls amodal. Gaps or 'holes' in the modal
sensory pattern acquire, so to speak, a status equal to that of actual
sensations modally determined. The situation is not quite like that of
the phi phenomenon, or even the completion of a figure part of which
falls in a blind part of the visual field, for in both these cases the com-

2 Vide e.g. A. MICHOTTE, and L. BURKE, Proc. XlIIth Int. Congo Psychol.,
1951, 179-80. L. BURKE, Quart.J. Exp. Psycho!., 1952,4, 121-38. A. GLYNN,
Quart. J. Exp. Psychol., 1954, 6, 125-39. A. C. SAMPAIO, La translation des
objets comme facteur de leur permanence phenomenale, Louvain, 1943.
X FOREWORD
pletion appears in the guise of a modally determined, namely visual,
element. Some people may prefer to avoid reference to anything so
apparently self-contradictory as amodal perception, and to couch their
analysis of the phenomenal fact in terms of the known constructive
tendencies of perceptual function. But Michotte's formulation has at
least the merit of bringing clearly to our notice the degree and manner
in which much of our everyday awareness of objects as permanent and
real is dependent upon elements to which no physical stimulus corre-
sponds at the time. This fact, whatever its proper provisional interpre-
tation, deserves far more experimental study than it has yet been given.
Concern with these problems has latterly led Michotte and his stu-
dents to a more general enquiry into the conditions in which things
appear as real, 3 and to ask what features of a stimulus pattern need to be
present in order that a seen object shall have this character. In one very
striking series of experiments, for instance, outline perspective drawings
of solid figures, regarded from an appropriate angle, take on an appear-
ance of reality in the specific sense that they no longer present themselves
as drawn upon a paper surface, but stand out in three-dimensional
space. The lines which form them are 'liberated', so to speak, from their
attachment to the surface, which is what normally causes them to appear
as a two-dimensional drawing. By manipulation of conditions such as the
microstructure and the relative contrast which affect such liberation,
the degree of apparent reality can be effectively controlled. These and
other experiments open up a field which should surely keep even
Professor Michotte busy for many a year yet.
Taken all together, this work constitutes a massive, bold, and exciting
exploratory attack on the perceptual mechanisms by which we become
aware of and react to the external world in the way we do. It is a world
ef substances and things which, phenomenally at any rate, retain their
permanence, their reality, their living or inanimate character, and which
act causally on one another. Each of these aspects Michotte has brought
under experimental study, and on everyone he has shed light. But to me,
at least, the inspiration to be derived from his work does not stem only
from its results, but to a major extent from the quality of its manner.
As to this, let me end with his own words:

• Vide e.g. A. MICHOTTE, Bull. Acad. Roy. Belg., 1948, 34, 268-88, ibid.,
1960, 46, i-xv. A. MICHOTTE, Arch. Psicol. Neurol. Psichiat., 1957,18,2°3-13.
A. MICHOTTE, Rev. into filmol., 1948, 1,249-61. M. R. PHEMlSTER, Quart. J.
Exp. Psychol., 1951, 3, 1-18.
 FOREWORD xi
'I have always tried to gather data as well-founded as possible, that
could provide a basis for general syntheses which, it seems to me,
ought to be reserved for the future, and perhaps for minds more so
inclined than mine.'
R. C. OLD FIELD
Institute of Experimental Psychology,
Oxford University
 Preface to the English Edition
When the question arose of having La Perception de la Causalite trans-
lated into English, I found myself confronted with the same problem as
occurred over the second French edition of 1954. Was it desirable to
keep the 1946 text in its original form, or ought it to be re-drafted so as
to take account both of new material discovered since the original date
and of the way in which my own ideas had evolved? This would have
necessitated an enormous amount of work and would have prevented
me from devoting to other investigations which continue to arouse my
enthusiasm such time as I might perhaps hope at my age to have at my
disposal.
For this reason I decided, rightly or wrongly, not to modify the
original text, but to add an extra chapter at the end (Appendix II), the
main purpose of which was to state the theoretical issues which seemed
to me of most importance. At the start (pp. 304-313) there is some de-
tailed information about the methodological principles which I have
followed in my research. Then comes the complete statement of the
theory of perceived causality according to my present views, along with
the arguments, based on a whole series of findings, for making changes
which both simplify it and enlarge its scope (pp. 313-369).
No one is more aware than I am that this solution is only partially
satisfactory, since it did not enable me to describe -let alone evaluate-
a number of investigations which, although adding to our knowledge of
perceived causality, did not have any direct relevance on the theoretical
side.! In addition this procedure makes matters more difficult for the
reader; and for this reason I should like to make a few suggestions to
those interested which I think may make their task easier.
Before they read the main book it seems to me desirable that they
should first read Appendix II, section 1 (,General Considerations on
Method'). This will give them a better understanding of the precise
scope of the work.
Secondly, it will be helpful for them to know at the outset that the

'On pp. 4I6-420 there is a complete list of all the research which, so far as I
know, has been undertaken to date. Some of this material was published when
my own book was already in the press, and in these cases it was therefore not
possible to take the results into consideration.
xiv PREFACE TO THE ENGLISH EDITION
chief modifications introduced into my earlier theory are those which
concern 'polarisation of the movements'. New observations have shown
that, contrary to my original view, this phenomenon, interesting as it is
in its own right, does not play the part in perceived causality which I
formerly attributed to it. In addition, although the fundamentals of my
theory have not changed, I have come to recognise that the factor of
'phenomenal permanence', to which there is allusion in several places
in the original text, is in fact much more important for the theory than
are the phenomena of polarisation, as will be seen in Appendix 11,
sections 2 and 3.
This being so, the theoretical views put forward in Chapter IV and
in part of Chapter VIII (pp. 130-133) on polarisation of the movements
have lost much of their interest as far as causality is concerned. The
same does not of course hold of the experimental facts reported in these
chapters, but they require to be reinterpreted. This is true in particular
of the form taken by the 'radius of action' of the passive o1:ject as
described and considered on pp. 53 seq.; the new account of this
phenomenon is given on p. 338. With regard to the experiments on
camouflaging the causal structures (pp. 72 seq.), these retain their
demonstrative validity since they in fact relate to a point which is
fundamental for the formulation both of the new theory and of the old
(see the argument on pp. 346-347).
I cannot end without expressing sincere thanks to all those whose
work has contributed to the publishing of the English translation of my
book. First of all I am very grateful to Tim and Elaine Miles, who
originally thought of the idea and who courageously undertook to trans-
late a book full of nuances which could not easily be expressed in a
language different from that of the author. They have done the job
with remarkable competence and with a most commendable thorough-
ness. Next I must thank myoId and close friend, R. C. Oldfield, who
by using his distinguished reputation in support of the project enabled
it to be carried through, and who kindly undertook to read over the
English text before publication. Finally I must express my gratitude
to Methuen and Co. for all the trouble which they have taken in con-
nexion with both printing and editing.
Louvain, 1 October, 1962 A. MICHOTTE
 Translators' Glossary
The meaning of most of Michotte's technical terms is clear from the
context. In a few cases, however, particularly those where the French
expressions have no straightforward English equivalent, readers may
find it helpful to be given some further explanation. This is done in the
glossary which follows. Page-numbers indicate where the word or
phrase first occurs, and the original French expression is given in brack-
ets.
AMPLIATION OF THE MOVEMENT: ('l'ampliation du mouvement')
(p. I43). This refers to the extension of a movement from one object
on to a second, in such a way that it remains the movement of the
first object while bringing about the 'displacement' (change in posi-
tion) of the second. For a discussion of the difference between
'movement' and 'displacement' see p. I34.
ENTRAINING, THE ENTRAINING EFFECT: (,entrainement, l'effet
entrainement') (p. 21). This refers to the impression of one object
joining another and carrying or pushing it along. Sir Frederic
Bartlett has spoken of 'pushing with follow up'. Since 'entraine-
ment' is used by Michotte as a technical term whose meaning is
clear from the context, it seemed to us preferable to introduce the
unfamilar word 'entraining' rather than try to find a more familiar
word which would necessarily be somewhat misleading. The Oxford
English Dictionary does in fact give a sense of 'entrain' fairly near
the required one, viz. 'to draw away with ... oneself'.
IMPACT-WRICH-LAUNCHES: ('choc-qui-lance') (p. 65). This phrase
implies that the impact itself has the character of being active; it is
the impact which actually does the launching. (For an explanation
of the word 'launch' see below). Elsewhere the word 'choc' is trans-
lated either as 'impact', as in 'the impact-noise experiments' ('les
experiences sur le choc-bruit') (pp. 235 seq.), or in some cases as
'blow'.
KINEMA TIC: ('cinetique') (p. 4). A contrast is frequently made in
the book between static Forms (Gestalten) and kinematic ones, i.e.
those involving movement.
xvi TRANSLATORS' GLOSSARY
LAUNCHING, THE LAUNCHING EFFECT: ('lancement, l'effet lance-
ment') (pp. 20-21). This refers to the impression of one object
'bumping into' another and setting it in motion. The French word
'lancer' has regularly been translated as 'launch'.
LAUNCHING-IN-FLIGHT: ('lancement au vol') (p. 66). This refers
to the impression which occurs when two objects are in motion and
when the first comes up and sets the second in motion yet again.
PHENOMENALISM: ('phenomenisme') (p. 13). The word is used in
this book in connexion with the views of Piaget on the child's
awareness of causal links. According to Piaget, causal links can be
established for the child between any two separate events even in the
absence of any rational knowledge connecting them. Thus if a child
says that a pebble sinks 'because it is white', the causal relationship
would be of this 'phenomenalist' character.1 At an early age, how-
ever, the child in forming his conception of causality also takes
account of his own feelings of effort or 'efficacy'. Hence arises Piaget's
view that the child's awareness of causal relations involves both
'phenomenalism' and 'efficacy'. (Compare p. 13 ofthis book.)
PHENOMENOLOGICAL: ('phenomenologique') (Appendix Il, P.304).
Phenomenology in Michotte's sense is the study of the way in
which things appear to the subjects in perception, as indicated by
their verbal descriptions.
PRIMARY FACT: ('fait primitif') (p. II). The French phrase is taken
from Maine de Biran, and refer~ to the basic impression of effort,
which he claims to be the source of all ideas of cause and force.
TRACE-MAKING, THE TRACE-MAKING EFFECT: ('tra<;age, l'effet
tra<;age') (Appendix I, p. 289). This refers to the making of a trail,
stain, or mark by an object. Either the object appears to cause the
trace ('tra<;age causal', Causal Trace Effect), or the trace appears
to emanate from the object ('tra<;age-ecoulement', Emanation Trace
Effect, p. 297).
1 For further discussion see J. Piaget, The Child's Conception of Causality
(tr. Marjorie Gabain), Kegan Paul, London, 1930, especially p. 253. In the
present work we have decided in favour of 'phenomenalism' as a translation of
'phenomenisme' in preference to Gabain's 'phenomenism', since 'pheno-
menalism', as used in British philosophy, has at least some of the required
associations. The translation 'phenomenalism' is also that adopted by Margaret
Cook in The Child's Construction of Reality, Kegan Paul, London, 1955.
 TRANSLATORS' GLOSSARY xvii
TRIGGERING, THE TRIGGERING EFFECT: ('d6cienchement, l'effet
d6cienchement') (p. 57). This refers to the impression that one
object has 'touched off' the movement of another, the second object
then appearing to move of its own accord.
All footnotes are Michotte's own except those which refer the reader
to the glossary; these have been marked with an asterisk. Words and
phrases to be found in the glossary are indicated as such only on the
occasion when they are first used.
The translators would like to express particular gratitude to Pro-
fessor R. C. Oldfield for reading the English text before its publication
and for making a large number of helpful comments and suggestions.
We are also very grateful to Professor Michotte himself for elucidating
many points and for his co-operation and sympathetic encouragement.
T.R.M.
Bangor, I960 E.M.
 Contents
FOREWORD by Professor R. C. Oldfield page v
PREFACE TO THE ENGLISH EDITION xiii
TRANSLATORS' GLOSSARY XV

AUTHOR'S NOTE xxiii

Introduction
I The Problem 3
r. Historical Survey 3
2. Causality and Activity 18

11 The Experimental Apparatus 27

1. The Disc Method 27
2. The Projection Method 34

Mechanical Causality
PART ONE-THE LAUNCHING EFFECT

III The Segregative Influence of the Objects 43

IV The Polarising Influence of the Objects 53

I. The Radius of Action in the Launching Effect 53
2. The Radius of Action in the Approach and
Withdrawal Effects 58
3. The Launching Effect and the Approach and
Withdrawal Effects 63
4. The Inversion of Polarity in the Launching Effect 68
i The case of the Tunnel Effect 68
ii Launching-in-flight 69
iii Camouflage experiments 72

V The Phenomenal Aspect of the Objects 82

xx CONTENTS
SUMMARY NO. I. RESUME OF CHAPTERS Ill, IV,
AND V page 87
VI Spatio-Temporal Integration 91

I. Temporal Unity 91
2. Spatial Unity 99
i Spatial contiguity 99
ii The relative orientation of the movements 101
iii The localisation of the movements in the same plane 103

VII The Speeds and the Hierarchisation of the

Aiovements 106

I.The Common Speed of the Objects and the Causal

Impression 107
2. The Relative Speeds and the Causal Impression 108
3. The Relative Speeds and Integration II3
4. The Hierarchy of the Movements and the Launching
and Triggering Effects II9

SUMMARY NO. 2. RESUME OF CHAPTERS VI AND VII 124

VIII The Launching Effect seen as a Whole 128

I. The Launching Effect 128

2. The Triggering Effect 144

PART TWO-THE ENTRAINING EFFECT

IX The Structural Organisation of the Entraining

Effect 149
I. The Entraining Effect and the Launching Effect 149
2. The Entraining Effect and the Transport Effect ISO
3. The Relative Speeds and the Entraining Effect 158
4. The Traction Effect 160

SUMMARY NO. 3. RESUME OF CHAPTER IX 162

X Launching-by-Expulsion 165

XI Propulsion 172

XII Animal Locomotion 183

 CONTENTS Ul

XIII Tactile-Kinaesthetic Perception of Mechanical

Causality page 201

GENERAL CONCLUSIONS

XIV Ampliation of the Movement 21 7

Qualitative Causality
XV Movement of One Object linked with Qualitative
Change in Another 231

XVI The Linking of Qualitative Changes in Two Objects 242

SUMMARY NO. 4. RESUME OF CHAPTERS XV AND XVI 251

The Origin of the Idea of Causality

XVII Critical Reflections on Different Theories 255
I. Hume's Theory 255
2. Maine de Biran's Theory 266
3. The Work of Piaget 275
4. The Apparent 'Sources' of the Emotions 282

Appendices
I A Special Case of Propulsion: the Trace-making
Effect 289
I. The Causal Trace Effect 289
2. The Emanation Trace Effect 297

11 Theory of Phenomenal Causality. New Perspectives 304

I. General Considerations on Method 304

2. Theory of Perceived Mechanical Causality 313
i The Distinction between Movement and Displacement 313
ii Theory of the Entraining Effect 317
iii Theory of the Launching Effect 344
xxii CONTENTS
3. General Conclusion and Discussion page 348
4. 'Cause', 'Condition', and 'Dependence' 35 8

Commentary
(by T. R. Miles)
LINTRODUCTORY 373
n. SUMMARY AND NOTES 374
In. TWO CRITICAL ESSAYS 402

1. Phenomenology and Scientific Method 402

2. Michotte's Experiments and the Views of Hume 410

References
GENERAL BIBLIOGRAPHY OF WORKS ON PERCEIVED
CAUSALITY 416
INDEX OF AUTHORS 421
INDEX OF SUBJECTS 422
 Author's Note
The results of the research described in this book seem difficult to
reconcile with the many different theories put forward by psychologists
and philosophers as to the origin of the idea of causality and its applica-
tion to the data of experience. For this reason I have been at pains to
ensure that the conclusions reached should have as sound and broad a
basis as possible. The experiments on which these conclusions depend
have therefore had to be very numerous.
In order to follow the general development of our ideas, however, and
to understand the theoretical views which we have been led to adopt
with regard to the perception of causality, it is not necessary to be
familiar with all these experiments. It should be possible, without any
disadvantage, to omit certain chapters which are particularly full of tech-
nical details, and to read instead the series of 'summaries' in which I
have given a resume of their contents. This has been done mainly for
the benefit of those who are not very familiar with laboratory work in
psychology, but who are nevertheless interested in the problem with
which this book is concerned.
One final point. Since I wished throughout to insist on the purely
experimental character of the work, I have deliberately avoided all
discussion of the philosophical significance of our results, reserving this
question for a later study. I have limited myself to a brief statement of
my position in relation to the classic views of Hume and Maine de
Biran, which are, of course, the starting-point for present-day philoso-
phical thinking on the subject of causality; and my discussion has dealt
only with the question whether these theories find confirmation or
justification in the experimental data.
INTRODUCTION
 CHAPTER I

The Problem
I. HISTORICAL SURVEY
At a conference held during the Ninth International Congress of
Psychology at Yale, U.S.A., in 1929, and in some lectures given at the
College de France in 1937, I ventured to suggest that laboratory re-
search on human perception should concern itself more than it had done
in the past with action. Among the numerous problems which arise in
this connexion, I mentioned in particular that of the perception of
mechanical actions - the actions exerted by one body on another. It is
this problem which forms the principal subject-matter of the present
work.
Since then I have undertaken extensive experimental research on
the subject. Some of the results, and a number of theoretical conclusions
to which the research led, have already been published as a preliminary
work. This was written in a non-technical style, and was intended for
readers of a philosophical journal.1 The present work contains not only
the technical details which are of interest to specialists in psychology,
but also descriptions of many new experiments, and an account of such
developments in the theory as we have been able to make as a result.
As has often been said before, perception is simply one phase of the
total process of action, and its biological role is to initiate and direct
the behaviour of men and animals. It not only provides material for their
contemplation, but invites them to action, and allows them to adjust this
action to the world in which they live.
The phenomenal world does not consist of a simple juxtaposition of
'detached pieces', but of a group of things which act upon each other
and in relation to each other. Thus the regulation of conduct requires a
knowledge of what things do or can do and what living creatures (and
ourselves in particular) can do with them.
We need to know that things can be moved, e.g. by pushing them,
causing them to slide, lifting them, or turning them over, by hurling,
breaking, bending or folding them, by leaning on them, and so on. We
1 A. MICHOTTE, La causalite physique est-elle une donnee phenomenale?
Tijdschrift voor Philosophie, Ill, 2, 1941, 290-328.
4 INTRODUCTION
need to know, too, that certain gestures, certain looks, or certain words
can attract or repel other men and animals, or modify their conduct in
some other way.
Similarly, it is necessary to understand the influence that things
exert on people - hurting us when they bump into us, pricking or cutting
us, resisting our efforts, confronting us with shapes that are easy or
difficult to handle, and so on.
Again, we need to know that one object goes up to another or with-
draws from it, that one person pursues another or hides from him, that
people lock up objects in drawers or chests, pour wine into glasses, and
so on.
These examples are taken from situations which occur regularly in
everyday life, and they are so ordinary that they do not seem at first
sight to raise any special problems.
This, however, is not the case. Although these events all have a spatial
and a kinematic* aspect, the most important feature about them is that
they imply functional relations between objects. These relations are
largely outside the range of the many investigations that have been car-
ried out on the subject of space-perception and perception of movement.
Even in a case as simple as that of our last example, do we not see the
wine come out of the bottle and run into the glass? That is something
quite different from a simple change of position in space.
These functional relations, then, constitute the essential fabric of
the phenomenal world; they must be considered as a highly important
factor in the adaptation of activities to their environment. They are
important, also, in enabling an external observer to understand the
human and animal conduct which he sees, and in this connexion they
play a major part in social psychology.2 It is these relations which give
the things around us their significance, since it is by coming to know
what things do that we learn what they are. What they are for us is
much more than their shape, their size, and their colour; it is above all
what they are capable of doing, or what can be done by means of them.
Here is a story which illustrates this point. One of my children, when
aged about three or four, once asked me, 'What are pictures hung on
* Cinetique. See Glossary.
2 Some writers have particularly stressed the part played by functional
relations in animal psychology, e.g. w. KOHLER in The Mentality of Apes
(tr. E. Winter), Kegan Paul, London, 1927 (also in Pelican Books), and E. C.
TOLMAN, both elsewhere and especially in his book Purposive Behaviour in
Animals and Men, University of California Press, Berkeley, 1932.
 THE PROBLEM 5
walls for?', and when I explained, he replied, 'Then pictures aren't
anything!' This is a good illustration of the extent to which, for us, the
essence of things consists in what they are able to do.
Among the functional relations which give things their significance,
the causal relations which unite them clearly play a very considerable
part. These, however, are not the only ones that have to be considered.
Spatial relations can fill this role, for instance. Thus a child will for a
long time think of a hat as an object 'which one puts on one's head', and
a box as an object 'inside which one puts other things' without thinking
of their functions of protection and conservation.
All this seems so obvious that nobody, with the exception perhaps of
some extreme behaviourists, would wish to dispute it. Moreover, it is
only necessary to read the works of psychologists who describe and
study behaviour to realise the major part played by functional relations,
whatever theoretical views these psychologists hold about them. 3 The
study of these relations, however, has found only a very small place in
the work of psychologists of the experimental school. The only problems
which have to any extent aroused discussion concerning relations are
(i) the theoretical problem of the knowledge of relations in general, and
(ii) a more empirical problem, that of the origin offunctional relations.
According to the prevalent view, these were considered to be the pro-
duct of a secondary elaboration of what is given by the senses. The result
of this was to give them a significance which by themselves they did not
possess, but which was in some way superimposed on them.
On this account of the matter it is clear that the question of functional
relations has to pass to a second level in the psychology of perception, or
at any rate must be reserved for further study, the fundamental problem
on this view, the experimental problem par excellence, being that con-
cerned with what is actually given.
The revolution brought about in the problems of perception by the
new approach of the Gestalt psychologists was of a kind to raise the
question of functional relations in quite a different way. In spite of this,
the research done up to now by the representatives of this school has
scarcely touched the problem. This is understandable, since the first
task confronting them was obviously that of re-examining and recasting
3 The wealth of material makes references almost superfluous; but among the
most distinguished works on this subject recently published in French may be
mentioned those of P. J ANE T, particularly Les debuts de l'intelligence, Paris,
1935, and L'intelligence avant le langage, Flammarion, Paris, 1936, and the
numerous writings of Piaget on child psychology.
6 INTRODUCTION
the information already obtained. As a result, the great majority of these
works have been devoted to the perception of shape and movement as
such.
In the present work I have limited myself to the question of the
relation of causality; but it goes without saying that other functional
relations can be studied along the same lines and by comparable methods.
We have come across a number of these in the course of our experi-
ments, and other research, now in process of being carried out, has been
devoted to them.

Having defined the purpose of our work in this way, we must now
examine the solutions which have been proposed or assumed by
psychologists as to the origin of the idea of causality.
It is well known that the question of how this idea is acquired has had
an important place in modem philosophy. This is not at all surprising
in view of the extreme importance of the question in the natural sciences,
in psychology, and in philosophy in general.
Now the problem with which philosophers such as Malebranche,
Hume, and Kant were dealing was essentially an epistemological one.
They were concerned to discover what could justify the characters of
necessity and universality in causal relations, and the work of the empiri-
cists was primarily intended to show that these characters could not be
derived directly from the data of experience. 4 If matters had rested there,
there would have been nothing in their views to cause us any special
concern. But Hume has gone further. He has expressly asserted that in
perceptual experience we have no direct impression of the influence
exerted by one physical event on another. This assertion has been so
widely accepted that it can still be regarded today as an almost universal
assumption; and it is to be found in very different contexts. For example,
the following lines come from the pen of one of our leading present-
day thinkers, namely Durkheim:
'In the first place, it is evident and recognised by all that it [the
notion of power] could not be furnished to us by external experience.
Our senses only enable us to perceive phenomena which co-exist
or which follow one another, but nothing perceived by them could give
4 Thus, according to Malebranche, CA true cause is a cause between which
and its effect the mind perceives a necessary connexion.' Quoted from
L. BRUNSCHVICG, L'experience humaine et la causalite physique, Alcan, Paris,
1922 , p. 7·
 THE PROBLEM 7
us the idea of this determining and compelling action which is charac-
teristic of what we call a power or force. They can touch only realised
and known conditions, each separate from the other; the internal
process uniting these conditions escapes them. Nothing that we
learn could possibly suggest to us the idea of what an influence or
efficaciousness is.'5
It is very interesting to contrast this quotation with similar passages
from Hume. Indeed the Durkheim passage almost seems like a trans-
lation of Hume - a point which shows clearly the persistence with which
the ideas of the British philosopher have been maintained.
By way of illustration, here are some characteristic sentences from
the Enquiry.6
'It appears that, in single instances of the operation of bodies we
never can, by our utmost scrutiny, discover anything but one event
following another.•.. So that, upon the whole, there appears not,
throughout all nature, anyone instance of connexion, which is
conceivable by us. All events seem entirely loose and separate. One
event follows another; but we never can observe any tye between
them. They seem conjoined, but never connected.'7
The idea of causality (in the sense of necessary connexion) is derived,
as we know, according to Hume, from the regularity in the succession
of phenomena; it is based entirely on anticipation, on the expectation
that when one event occurs, another event, which ordinarily follows it,
will do so again.
'Mter a repetition of similar instances, the mind is carried by habit,
upon the appearance of one event, to expect its usual attendant, and
SE. DURKHEIM, The Elementary Forms a/the Religious Life (tr. J. W. Swain),
AlIen and Unwin, London, 1915, pp. 363-4.
6 D. HU ME, An Enquiry concerning Human Understanding, section VII, part ii.
7 It seems, to judge by a number of passages, that Malebranche had more
perspicacity than Hume on this point, e.g.:
'When I see one ball bump into another, my eyes tell me, or seem to tell
me, that this first ball is really the cause of the movement which it transmits to
the other.'
'Nor does he [Aristotle] doubt that a ball which bumps into another ball
has the power to set it in motion. This is how it appears visually and that is
sufficient for this philosopher; for he nearly always follows the evidence of
the senses, and rarely that of reason; whether this evidence is intelligible or
not is a question which does not very much trouble him.' Quoted from L.
BRUNSCHVICG, op. cit., pp. 6 and 7.
P.C.-2
8 INTRODUCTION
to believe that it will exist. This connexion, therefore, which we feel
in the mind, this customary transition of the imagination from one
object to its usual attendant, is the sentiment or impression, from
which we form the idea of power or necessary connexion' (ibid.).

Now, of course, it is not just the prestige of its author which has won
for Hume's thesis the favour which it has found even with psychologists.
From the evidence, the conclusion seemed inevitable that we never can,
even by our 'utmost scrutiny', discover anything in the unfolding of
natural events except their simple succession.
Indeed the truth of Hume's claim is obvious, so long as our observa-
tion of the facts is objective and analytical, as is the practice in the physi-
cal sciences. We are then trying to understand what is 'really' happening
in 'the external world'; and, to do this, we examine separately the differ-
ent parts of the object or the different stages of the event which we want
to understand. That is what takes place, for example, when we try, like
Hume, to understand what happens (in the physical world) when two
billiard balls collide. Analytical observation clearly allows us to recognise
only a succession of movements.
This type of observation has also been the one used by psychologists.
As was only to be expected, after imposing itself on the natural sciences
it spread to every discipline with scientific claims. As a result, nearly
two centuries had to pass before it was realised that a wrong course was
being followed, and that even though this mode of observation was the
most suitable to give an accurate account of physical facts, it had the
result of splitting the phenomenal world into pieces and making the
most interesting psychological facts disappear. In the case which we
are considering, it resulted in the suppression of the 'tye' of which Hume
spoke, and whose presence Malebranche had so well recognised, as the
quotation in note 7 shows. More details on this point will be found in
Chapter VIII of this book.
Moreover Hume's view seemed all the more plausible since for a long
time the world of sensation was regarded as a replica of the world of
stimuli. As a result, it necessarily seemed impossible that anything should
be perceived unless there was something corresponding to it in the
sphere of stimulation. This ruled out the possibility of perceiving the
'influence' exerted by the impact of one billiard ball on the movement of
another.
In view of this, it is scarcely surprising that psychologists did not
 THE PROBLEM 9
take the trouble to verify by systematic research whether Hume's claims
were correct. It seemed definitely established that, from the point of
view of perception, there were no other problems connected with physi-
cal causality except those of the distribution of objects in space and the
perception of movement, change, succession, and simultaneity. Thus
although writers for the most part did not even mention the question,
we may assume that the very great majority of them would willingly
have subscribed, as far as external experience is concerned, to this
statement of Ziehen's :
'The idea of the relation of casuality is an empirical element that
always appears when two successive ideas are very closely associated.'B
On the other hand, however, all psychologists were forced to take
note of our natural conviction that we are the masters of our own actions,
that we can produce at will the movements of our limbs, direct the course
of our thought, etc., and, in short, that voluntary actions are initiated
by the subject, i.e. by the 'self' which is their cause.
Hume, following Malebranche here, expressly mentions this belief
and rejects it as fallacious:

' ... Our idea of power is not copied from any sentiment or con-
sciousness of power within ourselves, when we give rise to animal
motion, or apply our limbs to their proper use and office. That their
motion follows the command of the will is a matter of common ex-
perience, like other natural events. But the power or energy by which
this is effected, like that in other natural events, is unknown and in-
conceivable' (Enquiry, section VII, part i).

Hume admitted that the impression of effort against resistance can

enter into the popular inaccurate idea of power or cause; but he denies
that it has anything to do with power in the strict sense, which implies
necessity.
'It must, however, be confessed, that the animal nisus, which we
experience, though it can afford no accurate precise idea of power,
enters very much into the vulgar, inaccurate idea, which is formed of
it' (ibid., footnote).

8 TH. ZIEHEN, Introduction to Physiological Psychology (tr. C. C. van Liew

and Otto W. Beyer), London and New York, 2nd ed., 1895, p. 296.
10 INTRODUCTION
'Internal' experience, then, is on the same footing as external ex-
perience, as far as causality is concerned.
This view of Hume's, to some extent made precise and amplified, has
also been taken up by the associationist psychologists. While belief in
the causality of the self is only an illusion according to these writers,
there are none the less two phenomena which explain such a belief;
the first is our ability to foresee a result before it actually takes place,
the ;econd the presence of a feeling of 'activity'. This theory has been
brilliantly expounded in a classic work by Miinsterberg. 9
Starting from a radical sensationalism, he asks himself in what way
sensations must be combined to produce the feeling of inner freedom,
of active will.
One essential, he says, is a preliminary picturing to oneself of the
result. In addition, however, we need to note the part played by organic
sensations, and, in particular, what he calls the 'feeling of innervation' .
(This is his term for the kinaesthetic memory of previous movements.)
'On all occasions, however, when we become conscious, during the
act ofwill, of our own inner activity, there exists at that moment a lively
feeling of innervation; it is precisely in this that the feeling of inner
activity consists, and the strength of the exertion of the will is a direct
expression of the intensity of the feeling of innervation' (p. 72).
Thirty years later the theme has scarcely changed, as is shown by the
following passages, taken once more from Ziehen:
'We have yet to discuss the question as to how we come to regard
the idea of our ego as the cause of our actions .... It is obvious(ly) ...
because of its very frequent appearance in the series of ideas pre-
ceding each action. It is almost always represented several times
among the ideas preceding the final movement' (op. cit., p. 296).
In another passage he writes:
'This combination of motor sensations often gives our thought the
character of attentiveness and an appearance of volition and self-
activity which in fact do not belong to it at all' (ibid., p. 244).
Other psychologists, however, have continued to assert the existence
in internal experience of a specific feeling of activity, characteristic of
voluntary intervention and intimately linked with the self. This thesis
9 H. MUNSTERBERG, Die Tf7illenshandlung, Freiburg, Mohr, r888.
 THE PROBLEM II
has, of course, in fairly recent times, found experimental confirmation
in the work of Ach, Michotte and Priim, and others.IO
This opinion, in its extreme form, is found in the old and well-known
theory developed by Maine de Biran. According to him, we have a direct
experience of our own causality. This experience constituted, for Biran,
the primary Jact* which provided the foundation for all psychology
and all philosophy. It consisted
'in a simple fundamental relation that cannot be resolved in pheno-
menal terms, in which cause and effect, the subject and the act of
performing, are indivisibly united in the same feeling or the same
perception of effort (nisus). This effort has as its organ the muscles
which obey the dictates of the will. It is from this original impression
of effort that all ideas of force and cause derive.'ll
Voluntary movement, then, is an act immediately experienced; it
has as its culmination muscular sensations, and the self is its cause:
'It is in this change in the muscles surely, that effort shows itself
internally. Effort is simply the soul's particular power in action ....
Let us say, then ..• that the same internal feeling which reveals to
the soul its own particular effort identical with this act of will, re-
veals at the same time the organic modifications produced by the
effort, with its character of product or effect relative to its cause.'12
It is from this experience alone that the idea of causality is derived:
'A being who had never made an effort would not in fact have any
idea of power, nor, as a result, any idea of efficient cause. He would
see one movement succeed another, e.g. one billiard ball bump into
another and push it along; but he would be unable to conceive, or
apply to this sequence of movements, the idea of efficient cause or
acting force, which we regard as necessary if the series is to begin and
continue.'13
* Fait primitij. See Glossary.
10 N. ACH, Uber den Willensakt und das Temperament, QueUe und Meyer,
Leipzig, 1910, p. 240. A. MICHOTTE and E. PRUM, Etude experimentale sur le
choix voluntaire, Arch. de Psychol., X, 1910, p. 194.
11 Oeuvres choisies de Maine de Biran, Editions Montaigne, Aubier, Paris, 1942,

P· 165.
12 Quoted from G. MADINIER, Conscience et mouvement, Alean, Paris, 1938,
pp. 169 seq.
13 Quoted from L. BRUNSCHVICG, op. cit., p. 34.
12 INTRODUCTION
This theory of Biran's has been revived and made part of contempor-
ary psychological thought as a result of the work of Piaget on the origin
of the idea of causality in children.14
'The fact that the idea of force owes its existence to inner ex-
perience seems to be beyond dispute. To Maine de Biran belongs the
merit of having stressed this origin' (J. P I AGE T, The Child's Concep-
tion of Causality, 1930, p. 126).
Piaget, however, has an altogether new theory as to how this idea of
causality, derived from internal experience, is applied to 'things' and to
external experience. Biran postulates an 'initial induction which trans-
fers the causality of the self to the not-self'; he, nevertheless, hestitates
to ascribe to it the status of a true induction in the logical sense of the
term. 1S Others have seen it as a projection, what Lipps calls 'Einfiihlung',
i.e. empathy. In Piaget's view, of course, the prilnitive world of the child
is undifferentiated; it does not contain 'things' and a 'self'separated
from each other, as would be necessary if something belonging to the
self were to be applied to bodies. Internal experience, therefore, is
constantly 'jumbled up', as it were, with the so-called 'external' data
of experience. In the presence of this undifferentiated whole
'Everything happens as though the child began by attributing forces
to all outside bodies, and as though he only ended by finding in bim-
self the "I" that was the cause of his own force' (The Child's Concep-
tion of Causality, 1930, p. 128).
It is in the course of the child's general mental development that the
separation is made:
'Force becomes gradually withdrawn from external objects and
confined within the ego' (The Child's Conception of Causality, 1930,
P·13 2 ).
'Assilnilation and accommodation separate out to form increasingly
complex systems ... To the extent that this occurs, the causal nucleus
- personal activity - is broken down into a series of centres by the

14 For Piaget's ideas see particularly the following works: De quelques formes
primitives de causalite chez l'enfant, Annee psychologique, Alcan, XXVlme
annee, 1925; The Child's Conception of Causality (tr. Marjorie Gabain), Kegan
Paul, London, 1930; The Child's Construction of Reality (tr. Margaret Cook),
Kegan Paul, London, 1955, Chapter Ill, 'The Development of Causality'.
15 M. F. MAINE DE BIRAN, Oeuvres choisies, p. 248.
 THE PROBLEM 13
progressive objectification of causality' (The Child's Construction of
Reality, 1955, p. 316).
Piaget's views are neatly expressed in the following formula:
'The idea of force ... is the result of internal experience, but not
of an experience which is felt as internal from the first' (The Child's
Conception of Causality, 1930, p. 130).
To sum up, what is essential from our point of view is his assertion
that the idea of causality in the child is linked with 'efficacy', i.e. with
'feeling a bond between the desire and the result obtained' (De
quelques formes primitives de causalite chez l'enfant, 1925, p. 63).
This idea of causality involves a mixture of 'phenomenalism'* and
'efficacy'. In other words, the child is aware of particular events or
phenomena (whatever they may be) between which links can be estab-
lished, and is also aware of an 'efficacy' which gives these causal links
their dynamic character. But in the case of the objective world, this
early idea of causality fades and gradually disappears, to be replaced by
the mechanical and rational conception of causality.I6
Another view comparable with that of Biran is the sociological theory
outlined by Durkheim. 17
To Durkheim, no less than to Biran and Piaget, it is clear that the
idea of causality cannot emerge from external experience, and that we
must try to find its origin in internal experience. (See the quotation
on pp. 6-7.) Nevertheless, he does not agree that Biran's theory pro-
vides the answer. This is because, in Durkheim's view, effort, or
voluntary movement, is essentially a personal and incommunicable
experience, and cannot therefore account for the impersonal and
communicable character of power. On the other hand, the feeling of
pressure exerted on the individual by the community satisfies all the
conditions necessary to constitute a prototype of causal experience. This
pressure does in fact come from outside (and is therefore impersonal);
* Phinomenisme. See Glossary.
16 Piaget's work, as everyone knows, has been outstandingly successful in
giving rise to a considerable quantity of most interesting experimental research
on the way in which causal explanations evolve during the mental development
of the child. These studies, however, and the discussions resulting from them
do not really provide material that can be used to solve the problems dealt with
in this book which are of a different kind, and no special purpose would there-
fore be served in quoting them here. (For further discussion of Piaget's views
scc Chapter XVII.)
17 E. DURKHEIM, op. cit., pp. 365 seq.
14 INTRODUCTION
yet, on the other hand, it makes itself felt in our inner life. It intervenes
in the form of a constraint; individuals say that they 'feel' it 'when it
is acting upon their wills, to inhibit certain movements or command
others'.
Admittedly we have here only the beginnings of a theory, as Durk-
heim himself says; and it may be doubted, too, whether it really bears
upon the problem of causality. Indeed, it appears to be concerned with
motivation rather than with causality; and it seems as if the imperative
character of community influence is perhaps better fitted to explain the
idea of obligation or duty than the ideas of necessity and production.
Quite apart from the question of voluntary action, the world of in-
ternal experience presents us with other data which are interesting in
connexion with the problem of causality. They are summed up in the
statement, made time and time again by writers concerned with intro-
spective psychology, that emotions, needs, and tendencies are directly
linked with the events which give rise to them, or with other events which
result from them. Here there is not just a simple succession of indepen-
dent phenomena, but an intrinsic bond which observers, rightly or
wrongly, often describe in terms suggesting a causal connexion. 18
In the interesting chapter which he devotes to 'insight' in his
Gestalt Psychology, K6hler devotes many pages to stressing this point,
and cites an impressive set of examples indicating links of this sort. 19
These examples are extremely instructive. Indeed we can appreciate
how difficult it is to determine, or rather to indicate in a satisfactory
way, the character of the link in cases of this kind. I had myself made a
similar observation in the course of Phelan's experiments, in which I
took part as a subject (and that is my reason for referring to this work
rather than to that of others). K6hler too, in the course of his exposition,
makes use, no doubt intentionally, of a great variety of expressions;
he uses the general term direct determination to refer comprehensively
to a whole series of nuances. We have the impression e.g. that this
depends on, arises out of, follows from that, that we do this because of
that, that this develops from, is based upon, is related to, springs from,
refers to, is the outcome of that, we see the how and the why, etc.
The wealth of such expressions clearly shows that, even if there is no
doubt as to the existence of a link which we 'directly experience', the
18 See, for example, in this connexion, G. PHELAN, Feeling Experience and its
Modalities, Uystpruyst, Louvain, 1925, pp. 249 seq.
19 w. KOHLER, Gestalt Psychology, Liveright, New York, 1947, pp. 349 seq.
 THE PROBLEM IS
phenomenal aspect of this link is ill-defined, or very variable, and
there is no question of a characteristic impression of true causality,
at any rate not one common to them all. These cases are no less
interesting for that, and we shall have to return to them at the eDd of this
book.
The thesis which I put forward at the Yale conference was in direct
disagreement with all theories of the traditional kind which we have just
been considering, since I expressed the opinion that certain physical
events give an immediate causal impression, and that one can 'see'
an object act on another object, produce in it certain changes, and
modify it in one way or another. 20 I quoted various examples in this
connexion, e.g. that of a hammer driving a nail into a plank, and that of
a knife cutting a slice of bread. The question that arises is this: when we
observe these operations, is our perception limited to the impression
of two movements spatially and temporally co-ordinated, such as the
advance of the knife and the cutting of the bread? Or rather do we directly
perceive the action as such - do we see the knife actually cut the bread?
The answer does not seem to me to admit of any doubt.
Since that time some progress has been made on this question, at
least in the sense that a number of psychologists have made a close
study of it.
Koffka expressly mentions, in his textbook of psychology, that on
Gestalt principles it is perfectly conceivable that one should have a
specific impression of causality.21
About the same time Duncker was led to consider the problem from a
more specialised angle. 22 His research, it is true, did not bear directly
on this point, but rather on the question of problem solving. But since
solving problems frequently involved the discovery and use of causal
relations, the author was compelled to try to give some account of the
relevant psychological ideas.
Duncker rightly emphadses that, where there are causal relations, the
20 I prefer to use the phrase 'causal impression' rather than 'perception of
causality' which is the title of this book. Although in my opinion they are
equivalent, I think, perhaps, that the words 'causal impression', which I shall
use regularly in what follows, bring out more clearly me idea of an immediate
datum, of something directly 'lived' (cf. the German word 'Erlebnis'). Thus
'causal impression' could be exactly translated as 'Verursachunserlebnis'.
21 K. KOFFKA, Principles of Gestalt Psychology, Harcourt Brace & Co., New
York, 1935, pp. 378 seq.
22 K. DUNCKER, On Problem Solving (tt. L. S. Lees), Psychol. Monog.,
1945, 58, no. 5.
P.C.-2*
 16 INTRODUCTION
two events, effect and cause, are not always so 'detached' or separate
from each other as Hume had indicated. In many cases there is a certain
degree of 'Einsichtlichkeit' - the quality of being accessible to insight;
in other words, the link between the two is at least partially open to view.
First, this link can be seen in the spatial coincidence of the two
events; the spatial position of the cause is shown by the spatial position
of the effect. Thus when the problem involves bringing into a cage a
banana which is outside, as in the classic experiments of K6hler on apes,
the location of the banana indicates the point in space where the cause
must intervene; the position of the cause is not just any position, but
is marked by the position of the effect.
Secondly, the link is to be seen in the temporal coincidence of cause
and effect. Thus, according to Duncker, when a gust of wind blows a
door shut, and at the same time an electric light happens to go on at the
other end of a corridor, the impression of a causal relation is forced upon
us, so decisive is the factor of temporal coincidence.
Thirdly, the link can be seen in the direct correspondence often found
between effect and cause, both as regards form and matter. When one
billiard ball strikes another, it conveys to it a movement which, by and
large, is analogous to its own and in the same direction. The tracks left
by an object, e.g. the marks made by the paws of an animal, have the
same shape as the object which imprints them; the humidity of the rain
passes into the humidity of the pavements which it wets, and the same
holds in other cases. In short, the characteristic shape, appearance,
direction, etc., often pass from cause to effect in a way that can be im-
mediately intuited. Causa aequat effectum. If that has been neglected by
Hume and his supporters, it is because they chiefly had in mind cases
where this correspondence does not exist - cases of simple releasing, or
cases in which movement from cause to effect is concealed. The co-
incidences and correspondences which we have just been considering
cannot always be immediately intuited, but when they can the search
for causes is made very much easier.
In fact, according to Duncker, the chief point is this:

'We generally perceive as "cause" of an event, of a singularity,

another singularity which coincides spatially and above all tempor-
arily with the first. This in its turn results as "intersection" of two
uniform developments or "world-lines".' (Weltlinien) - i.e. lines
extending in space and time (p. 67).
 THE PROBLEM 17
It is essential, then, to the 'cause' event that it consists of an 'en-
counter'; thus the coming on of an electric light coincides with the
encounter of two world-lines, that of the switch and that of the arm in
motion, and appears to be effect of the two of them; and in the same
way, the wetting of the pavement coincides with the encounter between
the rain and the path.
Metzger has devoted a page of his recent psychology book - a work
of profound scholarship - to the question of phenomenal causality.23 He
too is convinced that one can have an immediate impression of causality
in external experience, and he stresses the importance of the structural
principle of proximity in space and time for producing this impression.
He adopts a general formula similar to that of Duncker:
'Thus there are two breaks in the general continuity. The one, the
cause, is the coming together of two entities previously separate and
unconnected; the other, the effect, is the genesis or passing away of
some entity, or a change in it as regards any quality, state, or mode of
behaviour.'
He particularly stresses the point that phenomenal causality implies
the passage from one object to another of a process or a property of one
of them; and he has above all the merit of expressing this idea in very
precise terms.
'Qualities of the cause reappear in the effect; in other words, at
bottom, nothing new comes into existence; rather there is sinlply the
transfer of something already existing to a new bearer. . . . In the most
obvious case, the colliding of two billiard balls, the two entities which
come together merely exchange certain qualities .... This happens in
such a way that, while there is discontinuity as regards the individu-
ality of the two entities, the 'world-lines' of their respective qualities
or states, as they transfer from one bearer to the other, remain con-
tinuous.'
These assertions seem extremely bold; for even if daily experience
often allows us to recognise a similarity between effect and cause, the
transfer, the leap of one and the same property from one object to the
other is surely not visible. Indeed, if it were, we should have difficulty
in understanding how Hume's thesis came to be so popular.
Yet, as will be seen later, the ideas expressed in this last quotation
from Metzger have some affinities with the theory which has emerged
.8 w. METZGER, Psychologie, Steinkopff, Leipzig, 1941, pp. I20 seq.
 18 INTRODUCTION
as the result of my own research, at least in the sphere of mechanical
causality.
2. CAUSALITY AND ACTIVITY
SO much for the historical background to our problem. The next task is
to give an account of our own research.
The first requirement was obviously to try to produce experimentally
some typical causal impressions, and to determine by tests the conditions
in which they occur. As might be expected, we began by examining the
classic case of one object striking another. From our very first trials
we were able to establish a fact which is important both theoretically
and from the point of view of practical experimentation, viz. that the
causal impression was not necessarily dependent on the use of 'real',
solid objects. It can be produced perfectly clearly by using objects
which are simple coloured shapes without apparent thickness, or even
images projected on a screen; and this is possible even when the ob-
servers know perfectly well what is going on. Our task was thus greatly
simplified, and it became possible for us to carry out experiments of
many different sorts. We were able to vary in a systematic way such
things as the colour, size, and shape of the objects, the speed and direc-
tion of their movements, the amplitude of their paths, the temporal
interval between the 'action' and the 'reaction', and so on.
Next we contrived to make the causal impression appear and dis-
appear at will; and we were able to compare directly the cases where it
occurs with those where it was absent. The study of these cases, as we
shall find later, brings to light the operation oflaws closely related to the
Gestalt laws. It is thus possible to link causal impressions with other
perceptual phenomena that are already known. We can also exclude
categorically any attempt aimed at reducing this impression to a 'pro-
jection' of our own power into things, or alternatively to a secondary
'interpretation' based ~n past experience and acquired knowledge.
When we had disposed of these problems we were confronted with a
second task, that of 'understanding' the phenomenon, of making a theory
about it, of seeking to find out why such and such conditions were
necessary for its production, and why it possessed such and such pro-
perties. This constituted a counter-proof, such as would provide a
definite demonstration of the original and primary character of the
causal impression.
The method which we systematically used for this purpose was that
of genetic analysis. This consisted essentially in simplifying in various