CD
117883
3?
r
Keep Your Card in This Pocket
Books will be issued only on presentation of proper
library cards.
Unless labeled otherwise, books may be retained
for two weeks. Borrowers finding books marked, de-
faced or mutilated are expected to report same at
library desk; otherwise the last borrower will be held
responsible for all imperfections discovered.
The card holder is responsible for all books drawn
on this card.
Penalty for over-due books 2c a day plus cost of
notices.
Lost cards and change of residence must be re-
ported promptly.
Public Library
Kansas City, Mo.
International Library of Jf sycroalogy : " -
Philosophy and Scientific Method
The Language
and Thought of the Child
;;: : \: c. K. OGDKN, M.A.
I* I ."**./ I (Magdalene College, Cambridge)
PHILOSOPHICAL ^STUDIES . . . . by G. E. MOORE, Litt.D.
THE Misus^pfoMiND by KARIN STEPHEN
CONFLICt ^I2>;^ REAM ' ' ' ty W ' IL R ' RlVER S, K.R.S.
PSYCHOLOGY* AN*D POLITICS . . . by W. II. R. RlVERb, F.R.S,
MEDICINE, MAGIC AND RELIGION . by W. H. R. RIVERS, F.R.S.
TRACTATUS LOGICO-PHILOSOPHICUS , . by L. WITTGENSTEIN
THE MEASUREMENT OF EMOTION . . by W. WHATKLY SMITH
PSYCHOLOGICAL TYPES . . . . by C. G. JUNG, M.I)., LL.D.
SCIENTIFIC METHOD by A. D. RITCHIE
SCIENTIFIC THOUGHT by C, I). BROAD, Litt D.
MIND AND ITS PLACE IN NATURE . . by C, P. BROAD. Litt.D.
THE MEANING OF MEANING, by C. K. OGDKN and I. A. RICHARDS
CHARACTER AND THE UNCONSCIOUS . . by ]. H, VAN DER HOOP
INDIVIDUAL PSYCHOLOGY by ALFRED ADLER
CHANCE, LOVE AND LOGIC by C. S. PEIRCE
SPECULATIONS (Preface by Jacob Epstein) . . by T. K. HULME
THE PSYCHOLOGY OF REASONING . . by KUGKNIO RIGNANO
BIOLOGICAL MEMORY by KUGENIO RIGNANO
THE PHILOSOPHY OF Music . . . . by W. POLE, F.R.S.,
THE PHILOSOPHY OF ' As IF ' . . . . by H. VAIHINGER
THE NATURE OF LAUGHTER . . . . by]. C. GREGORY
THE NATURE OF INTELLIGENCE .... by L. L. THUKSTONE
TELEPATHY AND CLAIRVOYANCE .... by R. TLSCHNER
THE GROWTH OF THE MIND by K. KOFPKA
THE MENTALITY OF APES by W. KOHLER
PSYCHOLOGY OF RELIGIOUS MYSTICISM . . . by J, H. LKUBA
THE PSYCHOLOGY OF A MUSICAL PRODIGY . . by G. REVESZ
PRINCIPLES OF LITERARY CRITICISM . . -by L A. RICHARDS
METAPHYSICAL FOUNDATIONS OF SCIENCE by K. A. BURTT, Ph.D.
COLOUR-BLINDNESS by M. COLLINS, Ph.D.
PHYSIQUE AND CHARACTER by ERNST KKETSCHMER
PSYCHOLOGY OF EMOTION ... by ], T. MACCURDY, M.D.
PROBLEMS OF PERSONALITY : . . in honour of MORTON PRINCE
PSYCHE by E. ROHDB
PSYCHOLOGY OF TIME by M. STURT
THE HISTORY OF MATERIALISM ... . by F, A. LANGK
EMOTION AND INSANITY by S. THALBITXKR
PERSONALITY by R. G. GORDON, M.D.
EDUCATIONAL PSYCHOLOGY by CHARI.KS Fox
LANGUAGE AND THOUGHT OF THE CHILD . . , by J, PIAOET
COMPARATIVE PHILOSOPHY . , . by P. MASSON-OURSKL
IN PREPARATION
THE LAWS OF FEELING by F. PAULHAN
CONVERSION by S. PK SANCTIS
THOUGHT AND THE BRAIN by H. PI&RON
SEX AND REPRESSION IN SAVAGE SOCIETY . by B. MALINOWKKI, D.Sc.
THE ANALYSIS OF MATTER . . by BBRTRAND RUSSELL, F.R.S.
PSYCHOLOGY AND ETHNOLOGY . . by W. H. R. RIVRKS, F.H.S.
STATISTICAL METHOD IN ECONOMICS , by p. SARGANT FLORENCE
THE PRIMITIVE MIND by F. RAWN, Ph.D.
COLOUR-HARMONY . by JAMK WOOD
THE THEORY OF HEARING . . . . by H. HARTRIIXSB, D.Sc,
SUPERNORMAL PHYSICAL PHENOMENA , . by K, J. DINGWALL
THEORETICAL BIOLOGY by J. VON UKXKOLL
THE INTEGRATIVE ACTION OF THE MlND by K. MlLLER
PLATO'S THEORY OF KNOWLEDGE . . . by F. M. CORN FORD
PRINCIPLES OF PSYCHOPATHOLOGY , by WM. BROWN, M.1X, D.Sc.
THEORY OF MEDICAL DIAGNOSIS . by F, G. CKOOKKHANK, M.D.
LANGUAGE AS SYMBOL AND AS EXPRESSION , . by K, SAPIR
A HISTORY OF ETHICAL THEORY. . . by M. GINSBERG, D.Lit.
THE PHILOSOPHY OF LAW by A. L. GoonHART
PSYCHOLOGY OF MUSICAL GENIUS . . by G, REVESZ
MODERN THEORIES OF PERCEPTION , . . by W. J. H. SPROTT
SCOPE AND VALUE OF ECONOMIC THEORY . by BARBARA WOOTTON
MATHEMATICS FOR PHILOSOPHERS . , by G. H. HARDY, K.R.S,
THE PHILOSOPHY OF THE UNCONSCIOUS , by E. VON HARTMANN
THE PSYCHOLOGY OF MYTHS . . by G. FXLIOT SMITH, F.R.S.
THE PSYCHOLOGY OF Music ... by EDWARD j. DENT
PSYCHOLOGY OF PRIMITIVE PEOPLES . bv B. MALINOWSKT. D.Sc.
The
Language and Thought
of the Child
By
JEAN PIAGET
Professor at the University of Neuchatel
and at the Institut J. J. Rousseau, Geneva
Preiace by
PROFESSOR E. CLAPAREDE
NEW YORK
HARCOURT, BRACE & COMPANY, INC.
LONDON: KEGAN PAUL, TRENCH, TRUBNER & CO., LTD.
1926
Translated by
MARJORIK WARDEN
CONTENTS
PAGE
PREFACE ix
FOREWORD . . . *. . . . xix
CHAPTER I
THE FUNCTIONS OF LANGUAGE IN TWO
CHILDREN OF SIX I
I. The material ....... 5
I. An example of the talk taken down, 6 2. The functions
of child language classified, 9 3. Repetition (echolalia),
ii 4. Monologue, 13 5. Collective monologue, 18
6. Adapted information, 19 7. Criticism and derision, 26
8. Commands, requests, threats, 27 9. Questions and
answers, 28.
II. Conclusions 34
10. The measure of ego-centrism, 34 n. Conclusion, 37
12. Results and hypotheses, 43.
CHAPTER II
TYPES AND STAGES IN THE CONVERSATION OF
CHILDREN BETWEEN THE AGES OF FOUR
AND SEVEN SO
i. Check of the coefficient of ego-cenmsm, 51 2. Types
of conversation between children, 52 3. Stage I :
Collective monologue, 56 4. Stage HA, First type:
Association with the action of others, 58 5. Stage
UA, Second type : Collaboration in action or in non-
vi CONTENTS
abstract thought, 60 7. Stage I IB, First type:
Quarrelling, 65 8. Stage IIu, Second type; Primi-
tive argument, 68 9, Stage IIlB : Genuine argument,
7010, Conclusions, 73.
CHAPTER III
UNDERSTANDING AND VERBAL EXPLANATION
IJEWEEN CHILDREN OF THE SAME AGE
BETWEEN THE YEARS OF SIX AND EIGHT. 76
i. The method of experiment, 792. Parcelling out the
material, 86 3. Numerical results, 944, Ego-centrism
in the explanations given by one child to another, 99
5. The ideas of order and cause in the expositions given
by the explainers, 1076. The factors of understanding,
119 7. Conclusion. The question of stages and the
effort towards objectivity in the accounts given by children
to one another, 124.
CHAPTER IV
SOME PECULIARITIES OF VERBAL UNDER-
STANDING IN THE CHILD BETWEEN THE
AGES OF NINE AND ELEVEN 127
7. Verbal syncretism, 131 2, Syncretism of reasoning,
136 3. The need for justification at any price, 1454,
Syncretism of understanding", 1505. Conclusion, 157.
CHAPTER V
THE QUESTIONS OF A CHILD OF SIX . . 1 62
I. f< Whys" 164
I. Principal types of "whys," 166 2. "Whys of causa
explanation." Introduction and classification by material
171 3. Structure of the "whys of explanation/' 1804.
Whys of motivation/* 1885. a Whys of justification, 1 '
191 -6, Conclusions, 197.
LUJNIJiJNIb vn
II. Questions not expressed under the form " why " . 199
7. Classification of Del's questions not expressed under
the form, "why," 199 8. Questions of causal explanation,
202 9. Questions of reality and history, 207 10. Ques-
tions about human actions and questions about rules, 214
ii. Questions of classification and calculation, 216.
III. Conclusions 217
12. Statistical results, 217 13. The decline of precaus-
ality, 223 14. Conclusion. Categories of thought or
logical functions in the child of seven, 227.
APPENDIX 239
INDEX 245
PREFACE
THE importance of this remarkable work deserves to be
doubly emphasized, for its novelty consists both in the
results obtained and in the method by which they have
been reached.
How does the child think? How does he speak?
What are the characteristics of his judgment and of
his reasoning? For half a century the answer has
been sought to these questions which are those which
we meet with at the very threshold of child psychology.
If philosophers and biologists have bent their interest
upon the soul of the child, it is because of the initial
surprise they experienced at his logic and speech. In
proof of this, we need only recall the words of Taine,
of Darwin and of Egger, which are among the first
recorded in the science of child logic.
I cannot give a list here of all the works that have
appeared since that period those of Preyer and of
Sully, of P. Lombroso and of Ament, of Binet, of
Stern, of Cramaussel and many others in order to
estimate the contribution made to the subject by Jean
Piaget. I shall try, however, to state roughly and
approximately the special characteristics of his work,
so as to indicate wherein consists the novelty of the
studies which this volume inaugurates.
Up till now such enquiries as have been made into
the intelligence and language of the child have been
in the main analytic. All the different forms which
reasoning and abstraction, the acquisition and forma-
x PREFACE
tion of words and phrases may take in the child have
been described, and a detailed and certainly serviceable
catalogue has been made of the mistakes, errors and
confusions of this undeveloped mentality, and of
the accidents and deformities of the language which
expresses it.
But this labour does not seem to have taught the
psychologist exactly what he wanted to know, viz. why
the child thinks and expresses himself in a certain
manner; why his curiosity is so easily satisfied with
any answer one may give or which he may give himself
(testifying to that riimportequismc which Binet con-
sidered one of the chief characteristics of imbecile
mentality) ; why he affirms and believes things so
manifestly contrary to fact ; whence comes his peculiar
verbalism ; and how and by what steps this incoherence
is gradually superseded by the logic of adult thought
In a word, contemporary research has stated the pro-
blem clearly but has failed to give us the key for its
solution. To the psychologist the mind of the child
still gives an impression of appalling chaos. As M.
Cramaussel once truly remarked: "Thought in a
child is like a net-work of tangled threads, which may
break at any moment if one tries to disentangle them."
Explanations have naturally been forthcoming to
account for such striking facts. The weakness, the
debility of the child's brain have been cited. But, I
would ask, does this tell us anything? The blame has
also been laid on the insufficiency of the experience
acquired, on the lack of skill of the senses, on a too
limited contingent of associations, on the accidents of
imitation. . . True as they may be, however,
these statements lead us nowhere.
After all, the error has been, if I am not mistaken,
that in examining child thought we have applied to
PREFACE
XI
it the mould and pattern of the adult mind ; we have
considered it from the point of view of the logician
rather than of the psychologist. This method, excel-
lent perhaps for establishing our first inventory, has
yielded all it can yield, and ends only in a blind alley.
It enables us to straighten out the skein but it does
not teach us how to disentangle the threads.
M. Jean Piaget's studies offer us a completely new
version of the child's mind.
M. Piaget is lucky enough to be still young. He
was initiated into psychology at a time when the
superficial associationism which had more or less intoxi-
cated his seniors thirty or forty years back was deacl
and buried, and when vistas full of promise were open-
ing out before our science. For James, Flournoy and
Dewey it was the dynamic and pragmatic tendency that
counted ; for Freud, psycho-analysis ; for Durkheim
(no matter whether his doctrine was sound or not)
the recognition of the role played by social life in
the formation of the individual mind ; for Hall, Groos,
Binet and the rest, genetic psychology propped up
by a biological conception of the child.
By a stroke of genius, M. Piaget having assimil-
ated these new theories, or rather having extracted the
good from each, has made them all converge on to
an interpretation of the child's mentality. He has
kindled a light which will help to disperse much of
the obscurity which formerly baffled the student of
child logic.
Do you know the little problem which consists in
making four equal triangles with six matches? At
first, while one thinks of it on the flat, it appears, and
quite rightly so, to be insoluble; but as soon as one
thinks of solving it in three-dimensional space, the
difficulty vanishes. I hope I do not misrepresent M.
Xll
PREFACE
Piaget's ideas in using this simple and somewhat crude
example to illustrate the nature of his contributions
to child psychology.
Up till now we were as helpless in face of the problem
presented by child mentality as before a puzzle from
which several important pieces were missing, whilst
other pieces seemed to have been borrowed from another
game and were impossible to place. Now M. Piaget
relieves our embarrassment by showing us that this
problem of childish thought does not consist of one
puzzle, but of at least two. In possession of this key we
no longer try to arrange on the flat, pieces which in
order to fit together require a space of three dimensions.
Our author shows us in fact that the child's mind is
woven on two different looms, which are as it were
placed one above the other. By far the most important
during the first years is the work accomplished on the
lower plane. This is the work done by the child him-
self, which attracts to him pell-mell and crystallizes
round his wants all that is likely to satisfy these wants.
It is the plane of subjectivity, of desires, games, and
whims, of the Lustprinzip as Freud would say. The
upper plane, on the contrary, is built up little by little
by social environment, which presses more and more
upon the child as time goes on. It is the plane of
objectivity, speech, and logical ideas, in a word the
plane of reality. As soon as one overloads it, it bends,
creaks and collapses, and the elements of which it is
composed fall on to the lower plane, and become mixed
up with those that properly belong there. Other pieces
remain half-way, suspended between Heaven and Earth.
One can imagine that an observer whose point of view
was such that he did not observe this duality of planes,
and supposed the whole transition to be taking place
on one plane, would have an impression of extreme
PREFACE
xiu
confusion. Because each of these planes has a logic of
Its own which protests loudly at being coupled with that
of the other. And M. Piaget, in suggesting to us with
confirmatory proofs that thought in the child is inter-
mediate between autistic thinking and the logical
thought processes of the adult, gives us a general
perspective of child mentality which will singularly
facilitate the interpretation of its various functions.
This new conception to which M. Piaget leads us,
whether in tacit or explicit opposition to current
opinion, could be stated (though always in very schematic
and summary fashion) in yet another set of terms.
Whereas, if I am not mistaken, the problem of child
mentality has been thought of as one of quantity,
M, Piaget has restated it as a problem of quality.
Formerly, any progress made in the child's intelligence
was regarded as the result of a certain number of ad-
ditions and subtractions, such as increase of new
experience and elimination of certain errors all of
them phenomena which it was the business of science
to explain. Now, this progress is seen to depend first
and foremost upon the fact that this intelligence under-
goes a gradual change of character. If the child mind
so often appears opaque to adult observation, it is not
so much because there are elements added to or wanting
in it, not so much because it is full of holes and ex-
crescences as because it belongs to a different kind of
thought autistic or symbolic thought, which the adult
has long since left behind him or suppressed.
The method which in M. Piaget's hands has proved
to be so prolific is also one of great originality. Its
author has christened it "the clinical method." It is,
in fact, that method of observation, which consists in
letting the child talk and in noticing the manner in
which his thought unfolds itself. The novelty consists
XIV
PREFACE
in not being content simply to record the answers given
by the child to the questions which have been put to
him, but letting him talk of his own accord. "If we
follow up each of the child's answers, and then, allowing
him to take the lead, induce him to talk more and more
freely, we shall gradually establish for every department
of intelligence a method of clinical analysis analogous
to that which has been adopted by psychiatrists as a
means of diagnosis." 1
This clinical method, therefore, which is also an art,
the art of questioning, does not confine itself to super-
ficial observations, but aims at capturing what is hidden
behind the immediate appearance of things. It analyses
down to its ultimate constituents the least little remark
made by the young subjects. It does not give up the
struggle when the child gives incomprehensible or
contradictory answers, but only follows closer in chase
of the ever-receding thought, drives it from cover,
pursues and tracks it down till it can seize it, dissect it
and lay bare the secret of its composition.
But in order to bear fruit this method required to be
completed by a judicious elaboration of the documents
which it had served to collect And this is where M.
Piaget's qualities as a naturalist have intervened. All
his readers will be impressed by the care with which
he has set out his material, by the way in which he
classifies different types of conversation, different types
of questions, different types of explanations ; and they
will admire the suggestive use to which he puts this
classification. For M. Piaget is a first-class biologist
Before going in for psychology, he had already made
his name in a special branch of the zoology of molluscs,
As early as 1912 (he was then only fifteen) he published
studies on the molluscs of the Neueh&tel Jura, A little
i Arch, de Psycho!, % XVIIL, p. 276.
PREFACE xv
later, he wrote a monograph on the molluscs of the
Valais and Leman districts. The subject of his doctor's
thesis in 1918 was the Distribution of the different
varieties of molluscs in the Valaisian Alps.
It must not be supposed, however, that in collecting
psychological material in the place of snails, and in
ordering and labelling it with so much care, M. Piaget
has simply turned from one hobby to another. Far
from it. His observations are not made for the pleasure
of making them. Even in the days when he was collect-
ing shells on the dry slopes of the Valais mountains, his
only object was to discover whether there was any
relation between the shape of those little animals and
the altitude at which they live, between variation and
adaptation. Still more is this so in his psychological
work. His only aim in collecting, recording, and
cataloguing all these different types of behaviour is
to see the assembled materials in a clearer light, to
facilitate the task of comparing and affiliating them
one to another. Our author has a special talent for
letting the material speak for itself, or rather for hearing
it speak for itself. What strikes one in this first book
of his is the natural way in which the general ideas
have been suggested by the facts ; the latter have not
been forced to fit ready-made hypotheses.
It is in this sense that the book before us may be
said to be the work of a naturalist. And this is all the
more remarkable considering that M. Piaget is among
the best informed of men on all philosophical questions.
He knows every nook and cranny and is familiar with
every pitfall of the old logic the logic of the text-
books ; he shares the hopes of the new logic, and is
acquainted with the delicate problems of epistemology.
But this thorough mastery of other spheres of know-
ledge, far from luring him into doubtful speculation,
XVI
PREFACE
has on the contrary enabled him to draw the line very
clearly between psychology and philosophy, and to
remain rigorously on the side of the first. His work
is purely scientific.
If then M. Piaget takes us so far into the fundamental
structure of the intelligence of the child, is it not because
the questions he set himself in the first instance were
questions of function? The writer of these lines may
perhaps be allowed to emphasize this idea of which
he is a particular advocate. The functional question
fertilizes the structural question, and states the problem
better than any other way. It alone gives full signifi-
cance to the details of mechanism, because it sees them
in relation to the whole machine. So it may well be
because he began with the questions: " Why does the
child talk? What are the functions of language?" that
M. Piaget has been led to such fertile observations and
conclusions.
But we would never have done if we once began
to point out all that is new and suggestive in this
book. Why should we? The reader will discover it
for himself by the perusal of its pages. I only wish
in conclusion to address to my colleague a word of
thanks in the name of the Institut f. f. Rousseau.
When we opened this Institute in 1912, it was hoped
that the two main pillars upon which we intended to
build the edifice the scientific study of the child and
the training of teachers would not remain isolated, but
be spanned and mutually reinforced by many a con-
necting arch. But the cares of organization, the un-
expected developments of an undertaking which receives
fresh impetus and grows faster than one had calculated,
the requirements of daily teaching, to say nothing of
the disturbances caused by the war all these have
prevented our scientific investigations from proceeding
PREFACE xvii
as we would have wished. The Institut Rousseau has,
it is true, given birth to some remarkable works, such
as V Instinct combatif by M. Bovet, the director, such as
the patient investigations made by Mile Descoeudres
on child language ; our students too have often col-
laborated in research, and have constantly taken part
in experiments. It is only since M. Piaget's arrival,
however, that a union closer than could ever have been
hoped for has been achieved between the most rigorous
scientific research and an initiation of the students to
the psychology of the child.
Having therefore witnessed for two years the con-
summate skill with which my colleague has utilized and
directed his developing powers in tracking down the
quarry with which he has presented us in so masterly
a fashion, I feel it both a privilege and a duty to express
rny sincere admiration for his work.
E. CLAPAREDE.
FOREWORD
THESE Studies in child logic are the outcome of research
work carried out in collaboration with others at the
Institut Rousseau during the school year of 1921-1922,
and of a course of lectures on child thought given
at the Geneva School of Science, which were based
upon the material collected during the same year.
This means that the essays before us are first and fore-
most a collection of facts and documents, and that the
bond between the various chapters is not that of
systematic exposition, but of unity of method applied
to a diversity of material.
Child logic is a subject of infinite complexity,
bristling with problems at every point problems of
functional and structural psychology, problems of logic
and even of epistemology. It is no easy matter to
hold fast to the thread of consistency throughout this
labyrinth, and to achieve a systematic exclusion of
all problems not connected with psychology. If we
try too soon to give a deductive exposition of experi-
mental results, there is always the risk of succumbing
to preconceived ideas, to the easy analogies suggested
by the history of science and the psychology of primitive
peoples, or worse still to the prejudices of the logical
or epistemological system to which we unwittingly sub-
scribe, try though we may to maintain a purely
xix
xx FOREWORD
psychological attitude of mind. The logic of the text-
books and the naive realism of common sense are both
in this respect fatal to any sane psychology of cognition,
and the more so since in trying to avoid the one we
are often thrown back upon the other.
For all these reasons I have abstained on principle
from giving too systematic an account of our material,
and a posteriori from making any generalizations out-
side the sphere of child psychology. All I have
attempted has been to follow step by step the facts as
given in experiments. We know well enough that
experiment is always influenced by the hypothesis
which occasioned it, but I have for the time being
confined myself strictly to the discussion of facts.
Moreover, for teachers and all those whose work calls
for an exact knowledge of the child's mind, facts take
precedence over theory. I am convinced that the mark
of theoretical fertility in a science is its capacity for
practical application. This book is therefore addressed
to teachers as much as to specialists in child psychology,
and the writer will be only too pleased if the results
he has accumulated are of service to the art of teaching,
and if in return his own thesis finds practical confirma-
tion in this way. He is convinced in this connexion
that what he tries to prove in this work concerning
the ego-centrism of child thought and the part played
by social life in the development of reason, must admit
of pedagogic application. If he personally has not
attempted straightaway to establish these consequences,
it is because he prefers to let professional opinion have
the first say.
Specialists in child logic will not, I hope, take me
to task for the disjointed character of this book, which
FOREWORD xxi
Is, as I have said, simply a study of the facts of the
case. I hope in a few years' time to produce a work
dealing with child thought as a whole, in which I
shall again take up the principal features of child
logic, and state their relation to the biological factors
of adaptation (imitation and assimilation). This is
the subject which was dealt with in the lectures above
referred to. Before publishing these in systematic form
it will be necessary to give as minute and exhaustive
a catalogue as possible of the material on which their
conclusions are based. The present volume is the
first of this series. I hope to follow it up with
another, which will be entitled : Judgment and Reason in
the Child. Together, these two will go to make up
" Studies in Child Logic." In a second work I shall
undertake to analyse causality and the function of reality
in the child. Then only shall we be in a position to
formulate a synthesis. If it were attempted any sooner,
any such synthesis would be constantly interrupted by
an exposition of the evidence, which in its turn would
tend to be distorted in the process.
One last word in acknowledgment of my debt to
those, without whose teaching this work could never
have been undertaken. M. Clapar&de and M. Bovet of
Geneva have consistently helped me by referring every-
thing to the point of view of function and to that of
instinct two points of view without which one passes
over the deepest springs of activity in the child. Dr
Simon of Paris introduced me to the tradition of Binet.
M. Janet, whose influence will often be traced in these
pages, familiarized me with a psychology of conduct
which offers a happy combination of genetic methods
and clinical analysis. I have also been deeply in-
xxii FOREWORD
fluenced by the social psychology of M. C. Blondel and
Professor J, M. Baldwin. It will likewise be apparent
how much I owe to psycho-analysis, which in my
opinion has revolutionized the psychology of primitive
thought. Finally, I need hardly recall Flournoy's
contribution to French psychology by his fusion of
the results of psycho-analysis with those of ordinary
psychology.
Outside the sphere of psychology I owe much to
authorities who have not been quoted or not suf-
ficiently quoted because of my desire to exclude all
but strictly paedological questions. The classic works
of M. Levy-Bruhl, for instance, have been a perpetual
source of inspiration. But it has been impossible in
this book to define my attitude towards sociological
explanations. The reason for this is very easy to
understand. Child logic and the logic of primitive
races are' far too much alike in some respects, and
far too different in others to justify us in discussing
so delicate a parallel in connexion with the scanty
evidence with which I propose to deal. I shall there-
fore keep this discussion for a later date. In logic,
in the history and philosophy of the sciences, and
in the theory of knowledge all spheres of knowledge
connected more closely than one would think with
the development of logic in the child I am indebted
to the historico-critical method of rny teacher, M.
Arnold Reymond, and to the standard works of M. E.
Meyerson and M. Brunschvicg. * Among the writings
of the latter, Les Etapes de la Philosophic math&tnatique
and more recently Uexplrience humaine et la causalite
physique have exercised a very decisive influence.
Finally, the teaching of M. Lalande and his work
FOREWORD xxiii
on the part played by the convergence of minds in
the formation of logical norms, have supplied a valuable
touchstone in our researches upon the ego-centrism of
the child,
JEAN PIAGET.
GENEVA* INSTITUT ROUSSEAU.
CHAPTER I
THE FUNCTIONS OF LANGUAGE IN
TWO CHILDREN OF SIX 1
THE question which we shall attempt to answer in this
book may be stated as follows : What are the needs
which a child tends to satisfy when he talks? This
problem is, strictly speaking, neither linguistic nor
logical ; it belongs to functional psychology, but it
should serve nevertheless as a fitting prelude to any
study of child logic.
At first sight the question may strike one as curious,
for with the child, as with us, language would seem to
enable the individual to communicate his thoughts to
others. But the matter is not so simple. In the first
place, the adult conveys different modes of thought by
means of speech. At times, his language serves only
to assert, words state objective facts, they convey in-
formation, and are closely bound up with cognition.
"The weather is changing for the worse/' "Bodies
fall to the ground." At times, on the other hand,
language expresses commands or desires, and serves
to criticize or to threaten, in a word to arouse feelings
and provoke action " Let's go,'* " How horrible ! " etc.
If we knew approximately in the case of each individual
the proportion of one type of speech to another, we
should be in possession of psychological data of great
interest. But another point arises. Is it certain that even
adults always use language to communicate thoughts?
To say nothing of internal speech, a large number of
1 With the collaboration of Mile Germaine Guex and of Mile Hilda de
Meyenburg.
2 STUDIES IN CHILD LOGIC
people, whether from the working classes or the more
absent-minded of the intelligentsia, are in the habit
of talking to themselves, of keeping up an audible
soliloquy. This phenomenon points perhaps to a pre-
paration for social language. The solitary talker in-
vokes imaginary listeners, just as the child invokes
imaginary playfellows. This is perhaps an example of
that return shock of social habits which has been de-
scribed by Baldwin; the individual repeats in relation to
himself a form of behaviour which he originally adopted
only in relation to others. In this case he would talk to
himself in order to make himself work, simply because
he has formed the habit of talking to others in order
to work on them. Whichever explanation is adopted,
it would seem that language has been side-tracked from
its supposed function, for in talking to himself, the
individual experiences sufficient pleasure and excite-
ment to divert him from the desire to communicate his
thoughts to other people. Finally, if the function of
language were merely to l communicate,' the pheno-
menon of verbalism would hardly admit of explanation.
How could words, confined as they are by usage to
certain precise meanings (precise, because their object is
to be understood), eventually come to veil the confusion
of thought, even to create obscurity by the multiplication
of verbal entities, and actually to prevent thought from
being communicable? This is not the place to raise
the vexed question of the relation between thought and
language, but we may note in passing that the very
existence of such questions shows how complex are
the functions of language, and how futile the attempt
to reduce them all to one that of communicating
thought.
i The functional problem therefore exists for the adult
| How much more urgently will it present itself in the
I case of defective persons, primitive races and young
children. Janet, Freud, Ferenczi, Jones, Spielrein, etc.,
have brought forward various theories on the language
THE FUNCTIONS OF LANGUAGE 3
of savages, imbeciles, and young children, all of which are
of the utmost significance for an investigation such as we
propose to make of the child mind from the age of six.
M. Janet, for example, considers that the earliest
words are derived from cries with which animals and
even savages accompany their action threats, cries of
anger in the fight, etc. In the earliest forms of social
activity, for instance, the cry uttered by the chief as he
enters into battle becomes the signal to attack. Hence
the earliest words of all, which are words of command.
Thus the word, originally bound up with the act of
which it is an element, at a later stage suffices alone
to release the act 1 The psycho-analysts have given
an analogous explanation of word magic. The word,
they say, having originally formed part of the act, is
able to evoke all the concrete emotional contents of
the act. Love cries, for instance, which lead up to
the sexual act are obviously among the most primitive
words ,* henceforward these and all other words alluding
to the act retain a definite emotional charge. Such
facts as these explain the very wide-spread tendency
of primitive thought to look upon the names of persons
and objects, and upon the designation of events as
pregnant with the qualities of these objects and events.
Hence the belief that It is possible to work upon them
by the mere evocation of words, the word being no
longer a mere label, but a formidable reality partaking
of the nature of the named object. 2 Mme Spielrein 3
has endeavoured to find the same phenomena in an
analysis of the very earliest stages of child language.
She has tried to prove that the baby syllables, mama^
uttered in so many tongues to call the mother, are
formed by labial sounds which indicate nothing more
1 British Joum. of Psych. (Med. Sect.), Vol. I, Part 2, 1921, p. 151.
2 See Jones, E, * f A Linguistic Factor in English Characterology," Intern.
Journal of Psycho- Anal., Vol. I, Part 3, p. 256 (see quotations from Ferenczi
and Freud, p. 257).
3 See Intei n. Zeitschrift f. PsychoanaL, Vol. VI, p. 401 (a report of the
proceedings of the Psycho -analytical Conference at the Hague).
4 STUDIES IN CHILD LOGIC
than a prolongation of the act of sucking. & Mama '
would therefore be a cry of desire, and then a command
given to the only being capable of satisfying this desire.
But on the other hand, the mere cry of * mama ' has In it
a soothing element ; in so far as it is the continuation of
the act of sucking, it produces a kind of hallucinatory
satisfaction. ' Command and immediate satisfaction are
in this case therefore almost indistinguishable, and so
intermingled are these two factors that one cannot tell
when the word is being used as a real command and
when it is playing its almost magical role.
Meumann and Stern have shown that the earliest
substantives of child language are very far from denot-
ing concepts, but rather express commands or desires ;
and there are strong reasons for presuming that
primitive child language fulfils far more complicated
functions than would at first appear to be the case.
Even when due allowance is made for these theories
in all their details, the fact remains that many ex-
pressions which for us have a purely conceptual
meaning, retain for many years in the child mind a
significance that is not only affective but also well-
nigh magical, or at least connected with peculiar modes
of behaviour which should be studied for themselves
and quite apart from adult mentality.
It may therefore be of interest to state the functional
problem in connexion with older children, and this is
what we intend to do as an introduction to the study of
child logic, since logic and language are obviously inter- '
dependent. We may not find any traces of ' primitive '
phenomena. At any rate, we shall be very far removed
from the common-sense view that the child makes use
of language to communicate his thoughts.
We need not apologize for the introductory character
of the questions dealt with in this work. We have
simply thrown out certain feelers. We have aimed
first and foremost at creating a method which could be
applied to fresh observations and lead to a comparison
THE FUNCTIONS OF LANGUAGE 5
of results. This method, which it was our only object
to obtain, has already enabled us to establish certain
facts. But as we have only worked on two children of
six years old, and as we have taken down their talk
in its entirety, it is true only for a month and during
certain hours of the day, we advance our conclusions
provisionally, pending their confirmation in the later
chapters of the book.
I. THE MATERIAL
The method we have adopted is as follows. Two
of us followed each a child (a boy) for about a month at
the morning class at the Maiso?i des Petits de flnstitut
Rousseau, taking down in minute detail and in its
context everything that was said by the child. In the
class where our two subjects were observed the scholars
draw or make whatever they like ; they model and play
at games of arithmetic and reading, etc. These activities
take place in complete freedom ; no check is put upon
any desire that may manifest itself to talk or play
together ; no intervention takes place unless it is asked
for. The children work individually or in groups, as
they choose ; the groups are formed and then break up
again without any interference on the part of the adult ;
the children go from one room to another (modelling
room, drawing room, etc.) just as they please without
being asked to do any continuous work so long as they
do not themselves feel any desire for it. In short,
these school-rooms supply a first-class field of observa-
tion for everything connected with the study of the
social life and of the language of childhood. 1
We must anticipate at once any objection that may
be advanced on the plea that since these children were
used as subjects they were not observed in natural
conditions. In the first place, the children, when they
1 Our grateful thanks are due to the ladies in charge of the Maison des
Petits, Miles Audemars and Lafandel, who gave us full freedom to work in
their classes.
6 STUDIES IN CHILD LOGIC
are in the play-room with their friends, talk just as
much as they would at home, since they are allowed to
talk all day long at school, and do not feel censured or
constrained in any way whatsoever. In the second
place, they do not talk any more at school than they
would at home, since observation shows that up to a
certain age, varying between 5 and yf, children gener-
ally prefer to work individually rather than in groups
even of two. Moreover, as we have taken down in its
entirety the context of our two subjects' conversations,
especially when it was addressed to an adult, it will be
quite easy to eliminate from our statistics all that is
not spontaneous talk on the part of the children, z..,
all that may have been said in answer to questions that
were put to them.
Once the material was collected, we utilized it as
follows. We began by numbering all the subjects'
sentences. As a rule the child speaks in short sen-
tences interspersed with long silences or with the talk
of other children. Each sentence is numbered separ-
ately. Where the talk is a little prolonged, the reader
must not be afraid of reckoning several consecutive
sentences to one number, so long as to each sentence
containing a definite idea only one number is affixed.
In such cases, which are rare enough, the division is
necessarily arbitrary, but this is of no importance for
statistics dealing with hundreds of sentences.
Once the talk has been portioned out into numbered
sentences, we endeavour to classify these into elementary
functional categories. It is this method of classification
which we are now about to study.
i. AN EXAMPLE OF THE TALK TAKEN DOWN. Let
us first of all give one complete example of the docu-
ments collected in this way, and let us examine it in
all its complexity :
23. Pie (to Ez who is drawing a tram-car with carriages
in tow): B r it>t the trams that are hooked on behind don't have
any flags. (No answer.)
THE FUNCTIONS OF LANGUAGE 7
24. (Talking about his tram). They don't have any
carriages hooked on . . . (He was addressing no one
in particular. No one answers him.)
25. (To Bea), 'Tsa tram that hasn't got no carriages.
(No answer.)
26. (To Hei), This tram hasn't got no carriages, Hei,
look, it isn't red, d'you see . . . (No answer.)
27. (Lev says out loud, ' A funny gentleman J from a
certain distance, and without addressing himself to Pie
or to anyone else). Pie : A funny gentleman ! (Goes on
drawing his tram.)
28. Fm leaving the tram white.
29. (Ez who is drawing next to him says, f Fm doing
it yellow '), No, you mustn't do it all yellow*
30. I'm doing the stair-case, look. (Bea answers, ' I
can't come this afternoon, I've got a Eurhythmic class/)
31. What did you say? (Bea repeats the same
sentence.)
32. What did you say ? (Bea does not answer. She
has forgotten what she said, and gives Ro a push.)
33. (To Bea), Leave him alone.
34. (Mile B. asks Ez if he would like to come with
her), Come here Ez, it isn't finished. 34 bis. Please
teacher, Ez hasn't finished.
35. (Without addressing himself to anyone,) I'm
doing some black stones* . . .
36. (Id), Pretty . . . these stones.
37. (To Ez), Better than you, eh ? (No answer. Ez
had not heard the previous remark.)
We have chosen this example from Pie (6J years)
because it is taken during the most sociable activity of
which this child is capable : he is drawing at the same
table as his bosom friend, Ez, and is talking to him
the whole time. It would therefore be natural in a
case of this kind if the sole function of speech were to
communicate thought. But let us examine the matter
a little more closely. It will be seen that from the
social point of view the significance of these sentences
or fragments of sentences is extremely varied. When
Pie says: " They don't have . . . etc." (24), or " Fm
doing . . . etc." (35) he is not speaking to anyone.
He is thinking aloud over his own drawing, just as
8 STUDIES IN CHILD LOGIC
people of the working classes mutter to themselves over
their work. Here, then, is $ first category which should
be singled out, and which in future we shall designate
as monologue. When Pie says to Hei or to Bea : " 'T'sa
tram . . . etc." (25) or a This tram . . . etc. " (26) he
seems on this occasion to want to make himself under-
stood ; but on closer examination it will be seen that he
cares very little who is listening to him (he turns from Bea
to Hel to say exactly the same thing) and, furthermore,
that he does not care whether the person he addresses
has really heard him or not. He believes that someone
is listening to him ; that is all he wants. Similarly, when
Bea gives him an answer devoid of any connexion with
what he has just been saying (30), it is obvious that he
does not seek to understand his friend's observation nor
to make his own remark any clearer. Each one sticks
to his own idea and is perfectly satisfied (30-32). The
audience is there simply as a stimulus. Pie talks about
himself just as he does when he soliloquizes, but
with the added pleasure of feeling himself an object of
interest to other people. Here then is a new category
which we shall call the collective monologue. It is to be
distinguished from the preceding category and also from
those in which thoughts are actually exchanged or
information given. This last case constitutes a separate
category which we shall call adapted information, and to
which -we can relegate sentences 23 and 34 6. In this
case the child talks, not at random, but to specified
persons, and with the object of making them listen and
understand. In addition to these practical and objective
forms of information, we can distinguish others of a
more subjective character consisting of commands (33),
expressions of derision or criticism, or assertions of
personal superiority, etc. (37). Finally, we may
distinguish mere senseless repetitions, questions and
answers.
Let us now establish the criteria of these various
categories.
THE FUNCTIONS OF LANGUAGE 9
2. THE FUNCTIONS OF CHILD LANGUAGE CLASSI-
FIED. The talk of our two subjects may be divided
into two large groups the ego-centric and the socialized.
When a child utters phrases belonging to the first
group, he does not bother to know to whom he is
speaking nor whether he is being listened to. He talks
either for himself or for the pleasure of associating
anyone who happens to be there with the activity of
the moment. This talk is ego-centric, partly because
the child speaks only about himself, but chiefly because
he does not attempt to place himself at the point of
view of his hearer. Anyone who happens to be there
will serve as an audience. The child asks for no more
than an apparent interest, though he has the illusion
(except perhaps in pure soliloquy if even then) of being
heard and understood. He feels no desire to in-
fluence his hearer nor to tell him anything ; not unlike
a certain type of drawing - room conversation where
every one talks about himself and no one listens.
Ego-centric speech may be divided into three categories :
i Repetition (echolalid) : We shall deal only with the
repetition of words and syllables. The child repeats
them for the pleasure of talking, with no thought of
talking to anyone, nor even at times of saying words
that will make sense. This is a remnant of baby prattle,
obviously devoid of any social character.
, 2 Monologue : The child talks to himself as though
he were thinking aloud. He does not address anyone.
3 Dual or collective monologue: The contradiction
contained in the phrase recalls the paradox of those
conversations between children which we were discuss-
ing, where an outsider is always associated with the
action or thought of the moment, but is expected neither
to attend nor to understand. The point of view of the
other person is never taken into account ; his presence
serves only as a stimulus.
In Socialized speech we can distinguish :
4 Adapted information; Here the child really ex-
io STUDIES IN CHILD LOGIC
changes his thoughts with others, either by telling his
hearer something that will interest him and influence
his actions, or by an actual interchange of ideas by
argument or even by collaboration in pursuit of a
common aim.
Adapted information takes place when the child
adopts the point of view of his hearer, and when the
latter is not chosen at random. Collective monologues,
on the other hand, take place when the child talks
only about himself, regardless of his hearers' point of
view, and very often without making sure whether he
is being attended to or understood. We shall examine
this criterion in more detail later on.
5 Criticism : This group includes all remarks made
about the work or behaviour of others, but having the
same character as adapted information ; in other words,
remarks specified in relation to a given audience. But
these are more affective than intellectual, /.., they assert
the superiority of the self and depreciate others. One
might be tempted in view of this to place this group
among the ego-centric categories. But ' egO-centric J
is to be taken in an intellectual, not in an ethical sense,
and there can be no doubt that in the cases under con-
sideration one child acts upon another in a way that
may give rise to arguments, quarrels, and emulation,
whereas the utterances' of the collective monologue are
without any effect upon the person to whom they are
addressed. The shades of distinction, moreover, between
adapted information and criticism are often extremely
subtle and can only be established by the context.
6 Commands, requests and threats: In all of these there
is definite interaction between one child and another.
7 Questions: Most questions asked by children
among themselves call for an answer and can therefore
be classed as socialized speech, with certain reservations
to which we shall draw attention later on.
8 Answers: By these are meant answers to real
questions (with interrogation mark) and to commands.
THE FUNCTIONS OF LANGUAGE n
They are not to be compared to those answers given
in the course of conversation (categ. 4), to remarks which
are not questions but belong to " information."
These, then, are the eight fundamental categories
of speech. It goes without saying that this classifica-
tion, like any other, is open to the charge of artificiality.
What is more important, however, is that it should
stand the test of practical application, i.e., that any
reader who has made himself familiar with our criteria
should place the same phrases more or less in the same
categories. Four people have been engaged in classi-
fying the material in hand, including that which is
dealt with in the next chapter, and the results of their
respective enquiries were found to coincide within 2 or
3 per cent.
Let us now return to one of these categories in order
to establish the constants of our statistical results.
3. REPETITION (ECHOLALIA). Everyone knows
how, in the first years of his life, a child loves to repeat
the words he hears, to imitate syllables and sounds, even
those of which he hardly understands the meaning. It
is not easy to define the function of this imitation in a
single formula. From the point of view of behaviour,
imitation is, according to Claparede, an ideomotor
adaptation by means of which the child reproduces and
then simulates the movements and ideas of those around
him. But from the point of view of personality and from
the social point of view, imitation would seem to be, as
Janet and Baldwin maintain, a confusion between the
I and the not-I, between the activity of one's own
body and that of other people's bodies. At his most
imitative stage, the child mimics with his whole being,
identifying himself with his model. But this game,
though it seems to imply an essentially social attitude,
really indicates one that is essentially ego-centric. The
copied movements arid behaviour have nothing in them
to interest the child, there is no adaptation of the
I to anyone else; there is a confusion by which the
12 STUDIES IN CHILD LOGIC
child does not know that he is imitating, but plays his
game as though it were his own creation. This is why
children up to the age of 6 or 7, when they have had
something explained to them and are asked to do it
immediately afterwards, invariably imagine that they
have discovered by themselves what in reality they are
only repeating from a model. In such cases imitation
is completely unconscious, as we have often had occasion
to observe.
This mental disposition constitutes a fringe on the
child's activity, which persists throughout different ages,
changing in contents but always identical in function.
At the ages of our two children, many of the remarks
collected partake of the nature of pure repetition or
echolalia. The part played by this echolalia is simply
that of a game ; the child enjoys repeating the words
for their own sake, for the pleasure they give him,
without any external adaptation and without an
audience. Here are a few typical examples :
(Mile E. teaches My the word 6 celluloid') Lev,
busy with his drawing at another table: " Lulo'id. . .
le le loid . . ." etc.
(Before an aquarium Pie stands outside the group
and takes no interest in what is being shown. Some-
body says the word ' triton J ). Pie: "Triton . . .
triton" Lev (after hearing the clock strike coucou ') :
" Coucou . . . coucou."
These pure repetitions, rare enough at the age of
Pie and Lev, have no interest for us. Their sudden
appearance in the midst of ordinary conversation is
more illuminating.
Jac says to Ez : " Look, Ez, your pants are showing."
Pie, who is in another part of the room immediately
repeats : " Look, my pants are showing, and my shirt , too."
Now there is not a word of truth in all this. It is
simply the joy of repeating for its own sake that makes
Pie talk in this way, i.e.) the pleasure of using words
THE FUNCTIONS OF LANGUAGE 13
?;">
not for the sake of adapting oneself to the conversation,
but for the sake of playing with them.
We have seen on page 7 the example of Pie hearing
Lev say: "A funny gentleman," and repeating this
remark for his own* amusement although he is busy
drawing a tram-car (27). This shows how little repeti-
tion distracts Pie from his class-work. (Ez. says: "I
want to ride on the train up there "), Pie: "/ want to
ride on the train up there."
There is no need to multiply examples. The process
is always the same. The children are occupied with
drawing or playing ; they all talk intermittently with-
out listening very much to each other ; but words
thrown out are caught on the bounce, like balls.
Sometimes they are repeated as they are, like the
remarks of the present category, sometimes they set
in action those dual monologues of which we shall
speak later on.
The frequency of repetition is about 2% and i%
for Pie and Lev respectively. If the talk be divided
into sections of 100 sentences, then in each hundred
will be found repetitions in the proportion of i%, 4%,
o%, 5% 3%, etc.
4. MONOLOGUE. Janet and the psycho-analysts
have shown us how close in their opinion is the bond
which originally connected word and action, words
oeing so packed with concrete significance that the
mere fact of uttering them, even without any reference
to action, could be looked upon as the factor in initiating
the action in question.
Now, independently of the question of origins, it is
a matter of common observation that for the child words
are much nearer to action and movement than for us.
This leads us to two results which are of considerable
importance in the study of child language in general
and of the monologue in particular. i The child is
impelled, even when he is alone, to speak as he acts,
to accompany his movements with a play of shouts and
i 4 STUDIES IN CHILD LOGIC
words. True, there are silences, and very curious ones
at that, when children work together as in the Maison
des Petits. But, alongside of these silences, how many
a soliloquy must take place when a child is alone in a
room, or when children speak without addressing them-
selves to anyone, 2 If the child talks even when he is
alone as an accompaniment to his action, he can reverse
the process and use words to bring about what the
action of itself is powerless to do. Hence the habit of
romancing or inventing, which consists in creating
reality by words and magical language, in working on
things by means of words alone, apart from any contact
either with them or with persons.
These two varieties belong to the same category, that
of the monologue. It is worth noting that the mono-
logue still plays an important part between the ages of
6 and 7. At this age the child soliloquizes even
in the society of other children, as in the class-rooms
where our work has been carried on. We have some-
times seen as many as ten children seated at separate
tables or in groups of two or three, each talking to
himself without taking any notice of his neighbour.
Here are a few examples of simple monologue (the
first variety) where the child simply accompanies his
action with sentences spoken aloud.
Lev sits down at his table alone : " / tvant to do that
drawing^ there . . . / want to draw something \ I do. I
shall need a big piece of paper to do that"
Lev knocks over a game : l ' There ! everything's fallen
down"
Lev has just finished his drawing : "Now I want to
do something else"
Lev is a little fellow who is very much wrapped up
in himself. He is always telling every one else what he
is doing at the moment. In his case, therefore, mono-
logue tends in the direction of collective monologue,
where every one talks about himself without listening to
the others. All the same, when he is alone he goes
THE FUNCTIONS OF LANGUAGE 15
on announcing what he is going to do, with no other
audience than himself. It is in these circumstances that
we have the true monologue.
In the case of Pie, the monologue is rarer, but more
true to type ; the child will often talk with the sole aim
of marking the rhythm of his action, without exhibiting a
shade of self-satisfaction in the process. Here is one of
Pie's conversations with context, where monologue is
interspersed with other forms of talk :
53. Pie takes his arithmetic copy-book and turns the
pages : "1, 2 . . . 3, 4, 5, 6, 7 . . . 8 . . . <?, 8, 8, 8 and
8 ... 9. Number 9, number 9, number 9 (singing) 7
want number 9. (This is the number he is going to
represent by a drawing).
54. (Looking at Bea who is standing by the counting-
frame but without speaking to him) : Now Tm going to
do 9> 9, I'm doing 9, I'm doing 9. (He draws).
55. (Mile. L. passes by his table without saying any-
thing). Look, teacher j 9, 9, 9 . . . number 9.
56. (He goes to the frame to see what colour to choose
for his number so that it should correspond to the gth
row in the frame). Pink chalk^ it will have to be 9. (He
sings).
57. (To Ez as he passes) : Pm doing 9, I am (Ez)
What are you going to do? Little rounds.
58. (Accident to the pencil) Ozv, ow !
59. Now Pve got to 9."
The whole of this monologue has no further aim than
to accompany the action as it takes place. There are
only two diversions. Pie would like to inform someone
about his plans (sentences 55 and 57). But in spite of
this the monologue runs on uninterrupted as though
Pie were alone in the room. Speech in this case
functions only as a stimulus, and in nowise as a means
of communication. Pie no doubt enjoys the feeling of
being in a room full of people, but if he were alone, his
remarks would be substantially the same.
At the same time it is obvious that this stimulus
contains a certain danger. Although in some cases it
16 STUDIES IN CHILD LOGIC
accelerates action, it also runs the risk of supplanting
it. "When the distance between two points has to
be traversed, a man can actually walk it with his legs*
but he can also stand still and shout: 'On, on ! . . :'
like an opera singer." 1 Hence the second variety of
child soliloquy where speech serves not so much to
accompany and accelerate action as to replace it by an
illusory satisfaction. To this last group belong certain
cases of word magic ; but these, frequent as they are,
occur only in the strictest solitude. 2 What is more
usual is that the child takes so much pleasure in solilo-
quizing that he forgets his activity and does nothing
but talk. The word then becomes a command .to the
external world. Here is an example of pure and of
collective monologue (cf. next chapter) where the child
gradually works himself up into issuing a command
to physical objects and to animals :
" Now then, it's coming (a tortoise). Ifs coming, ifs
coming^ its coming. Get out of the way, Da, ifs coming,
ifs coming, ifs coming. . . . Come along, tortoise /"
A little later, after having watched the aquarium,
soliloquizing all the time : " Oh, isn't it (a salamander)
surprised at the great big giant (a fish)," he exclaims,
" Salamander, you must eat up the fishes / "
In short we have here the mechanism of solitary
games, where, after thinking out his action aloud, the
child, under the influence of verbal excitement as much
as of any voluntary illusion, comes to command both
animate and inanimate beings.
In conclusion, the general characteristic of mono-
logues of this category is that the words have no
social function. In such cases speech does not com-
municate the thoughts of the speaker, it serves to
accompany, to reinforce, or to supplement his action.
It may be said that this is simply a side-tracking of
the original function of language, and that the child
1 P. Janet, loc. cit., p. 150.
2 These cases will be dealt with elsewhere.
THE FUNCTIONS OF LANGUAGE 17
commands himself and external things just as he has
learned to command and speak to others. There can
be no doubt that without originally imitating others
and without the desire to call his parents and to
influence them, the child would probably never learn
to talk ; in a sense, then, the monologue is due only to
a return shock of words acquired in relation to other
people. It should be remembered, however, that
throughout the^time when he is learning to speak, the
child is constantly the victim of a confusion between
his own point of view and that of other people. For
one thing, he does not know that he is imitating.
For another, he talks as much to himself as to others,
as much for the pleasure of prattling or of perpetuating
some past state of being as for the sake of giving orders.
It is therefore impossible to say that the monologue
is either prior to or later than the more socialized forms
of language ; both spring from that undifferentiated
state where cries and words accompany action, and
then tend to prolong it ; and both react one upon the
other at the very outset of their development.
But as we pass from early childhood to the adult
stage, we shall naturally see the gradual disappearance
of the monologue, for it is a primitive and infantile
function of language. It is remarkable in this con-
nexion that in the cases of Pie and Lev this form
should still constitute about 5% and 15% respectively
of their total conversation. This percentage is con-
siderable when the conditions in which the material
was collected are taken into account. The difference
in the percentages, however, corresponds to a marked
difference in temperament, Pie being of a more practical
disposition than Lev, better adapted to reality and
therefore to the society of other children. When he
speaks, it is therefore generally in order to make him-
self heard. It is true, as we saw, that when Pie does
talk to himself his monologue is on the whole more
genuine than Lev's, but Pie does not produce in such
B
i8 STUDIES IN CHILD LOGIC
abundance those rather self-satisfied remarks in which
a child is continually announcing his plans to himself,
and which are the obvious sign of a certain imaginative
exuberance.
5. COLLECTIVE MONOLOGUE. This form is the
most social of the ego-centric varieties of child language,
since to the pleasure of talking it adds that of solilo-
quizing before others and of interesting, or thinking
to interest, them in one's own action and one's own
thoughts. But as we have already pointed out, the
child who acts in this manner does not succeed in
making his audience listen, because, as a matter of fact,
he is not really addressing himself to it. He is not
speaking to anyone. He talks aloud to himself in
front of others. This way of behaving reappears in
certain men and women of a puerile disposition (certain
hysterical subjects, if hysteria be described as the sur-
vival of infantile characteristics) who are in the habit
of thinking aloud as though they were talking to
themselves, but are also conscious of their audience.
Suppress the slightly theatrical element in this attitude,
and you have the equivalent of the collective monologue
in normal children.
The examples of i should now be re-read if we
wish to realize how socially ineffectual is this form of
language, z.., how little impression it makes upon the
person spoken to. Pie makes the same remark to
two different persons (25 and 26), and is in nowise
astonished when he is neither listened to nor answered
by either of them. Later on he asks Bea twice, " What
did you say?" (31 and 32), but without listening to
her. He busies himself with his own idea and his
drawing, and talks only about himself.
Here are a few more examples which show how
little a child is concerned with speaking to anyone in
particular, or even with making himself heard :
Mile L. tells a group of children that owls cannot
see by day. Lev : " Well> I know quite well that it carit"
THE FUNCTIONS OF LANGUAGE 19
Lev (at a table where a group is at work): cc I've
already done ' moon ' so I'll have to change it"
Lev picks up some barley-sugar crumbs: "/ say,
Pve got a lovely pile of eye-glasses"
Lev : " / say, I've got a gun to kill him with. I say, I
am the captain on horseback. I say, I've got a horse and
a gun as well."
The opening phrase, "I say, I" which occurs in
most of these sentences is significant. Every one is
supposed to be listening. This is what distinguishes
this type of remark from pure monologue. But with
regard to its contents it is the exact equivalent of the
monologue. The child is simply thinking out his actions
aloud, with no desire to give anyone any information
about it,
We shall find in the next chapter examples of collective
monologues no longer isolated or chosen from the talk
of two children only, but taken down verbatim from
all-round conversations. This particular category need
not therefore occupy us any longer.
The collective monologue represents about 23% of
Lev's and 30% of Pie's entire conversation. But we
have seen that it is harder to distinguish the pure
from the collective monologue in Lev's case than in
Pie's. Taking therefore the two types of monologue
together, we may say that with Lev they represent
38%, and with Pie 35% of the subject's sum of conversa-
tion.
6. ADAPTED INFORMATION. The criterion of
adapted information, as opposed to the pseudo-informa-
tion contained in the collective monologue, is that it is
successful. The child actually makes his hearer listen,
and contrives to influence him, i.e., to tell him some-
thing. This time the child speaks from the point of
view of his audience. The function of language is
no longer merely to excite the speaker to action, but
actually to communicate his thoughts to other people.
These criteria, however, are difficult of application, and
20 STUDIES IN CHILD LOGIC
we shall try to discover some that admit of greater
precision.
It is adapted information, moreover, that gives rise
to dialogue. The dialogues of children deserve to be
made the object of a special and very searching investi-
gation, for it is probably through the habit of arguing
that, as Janet and Baldwin have insisted, we first
become conscious of the rules of logic and the forms
of deductive reasoning. We shall therefore attempt in
the next chapter to give a rough outline of the different
stages of conversation as it takes place between children.
In the meantime we shall content ourselves with ex-
amining adapted information (whether it takes place in
dialogue or not) in relation to the main body of talk
indulged in by our two subjects, and with noting how
small is the part played by this form of language in
comparison to the ego-centric forms and those socialized
forms of speech such as commands, threats, criticisms,
etc., which are not connected with mere statement of
fact.
The form in which adapted information first presents
itself to us, is that of simple information. Here are a
few clear examples :
Lev is helping Geo to play Lotto : " / think that goes
here.''" Geo points to a duplicate card. Lev: " If you
lose one> there will still be one left" Then : " You^ve got
three of the same" or : " You all see what you have to do"
Mile R. calls Ar ' Roger/ Pie: "He isn't called
Roger."
Such remarks as these are clearly very different from
dual monologues. The child's object is definitely to
convey something to his hearer. It is from the latter's
point of view that the subject speaks, and no longer
from his own. Henceforward the child lays claim to
be understood, and presses his claim if he does not
gain his point ; whereas in the collective monologue
words were thrown out at random, and it little mattered
where they fell.
THE FUNCTIONS OF LANGUAGE 21
In adapted information the child can naturally talk
about himself as about any other subject of conversa-
tion. All that is needed is that his remarks should be
6 adapted ' as in the following examples :
Ez and Pie : " I shall have one to-morrow (a season-
ticket on the tramway) I shall have mine this afternoon"
Ez and Pie are building a church with bricks :
" We could do that with parallels too. I want to put the
parallels on^
We are now in a position to define more closely
the distinction between the collective monologue and
adapted information. The collective monologue takes
place whenever the child talks about himself, except
in those cases where he does so during collaboration
with his hearer (as in the example just given of the
church building game), and except in cases of dialogue.
Dialogue, in our view, occurs when the child who has
been spoken to in a proposition, answers by talking
about something that was treated of in this proposition
(as in the example of the tramway season-ticket), and
does not start off on some cock-and-bull story as so
often happens in collective monologue. 1
In conclusion, as soon as the child informs his hearer
about anything but himself, or as soon as in speaking
of himself, he enters into collaboration or simply, into
dialogue with his hearer, there is adapted information.
So long as the child talks about himself without collab-
orating with his audience or without evoking a dialogue,
there is only collective monologue.
These definitions and the inability of collective mono-
logue to draw others into the speakers sphere of action
render it all the more remarkable that with Pie and
Lev adapted information numbers only half as many
remarks as collective monologue. Before establishing'
the exact proportion we must find out what sort of
things our two subjects tell each other, and what they
1 For such cock-and-bull stories, see p. 7, sentence 30.
22 STUDIES IN CHILD LOGIC
argue about on those rare occasions when we can talk
of arguments taking place between children.
On the first point we may note the complete absence
between the children of anything in the nature of
explanation, if by this word we mean causal explana-
tion, i.e.> an answer of the form " for such a reason"
to the question "why?" All the observed cases of
information which might be thought to resemble ex-
planation are statements of fact or descriptions, and are
free from any desire to explain the causes of phenomena.
Here are examples of information which simply state
or describe :
Lev and Pie: "Thafs 420." "It isn't 10 o'clock."
"A roof doesn't look like that" (talking of a drawing).
" This is a village , a great big village," etc.
Even when they talk about natural phenomena, the
information they give each other never touches on
causality.
Lev : " Thunder rolls No, it doesn't roll It's water
No it doesn't roll What is thunder? Thunder is . . ."
(He doesn't go on.)
This absence of causal explanations is remarkable,
especially in the case of machines, motors, bicycles,
etc., which the subjects occasionally discuss, but always
from what we may call the factual point of view.
Lev: "It's on the same rail. Funny sort of cart, a
motor cart A bicycle for two men"
Now each of these children taken separately is able
to explain the mechanism of a bicycle. Pie does so
imperfectly, but Lev does so quite well. Each has a
number of ideas on mechanics, but they never discuss
them together. Causal relations remain unexpressed
and are thought about only by the individual, probably
because, to the child mind they are represented by
images rather than by words. Only the underlying
factual element finds expression
THE FUNCTIONS OF LANGUAGE 23
This peculiarity comes out very clearly when children
collaborate in a game.
Here for instance are Pie and Ez occupied in drawing
a house together. Pie : " You must have a little button
there for the light, a little button for the light . . . Now
Pm doing the 'lectric light . . . There are two 'lectric
lights. Look we'll have two 'lectric lights. These are all
squares of*lectric lights"
We shall have occasion in later chapters to confirm
our hypothesis that the causal * why ' hardly enters
into child conversation. We shall see, particularly in
Chapter III, that the explanations elicited from one
child by another between the ages of 6 and 8 are
for the most part imperfectly understood in so far as they
seem to express any sort of causal relation. Questions
of causality are therefore confined to conversations
between children and adults, or to those between younger
and older children. Which is the same thing as saying
that most of these questions are kept hidden away by
the child in the fastness of his intimate and unformulated
thought.
Here are those of the remarks exchanged by Lev and
Pie which approach most nearly to causal explanation.
It will be seen that they are almost entirely descriptive :
Lev : " We ought to have a little water. This green
paint is so very hard, most awfully hard " . . . "In card-
boar d> don't you know ? You dorft know how to> but it is
rather difficult for you , it is for every one! 1
Childish arguments, it is curious to note, present
exactly the same features. Just as our two subjects
never communicate their thoughts on the why and
wherefore of phenomena, so in arguing they never
support their statements with the < because J and < since '
of logic. For them, with two exceptions only, arguing
consists simply in a clash of affirmations, without any
attempt at logical justification. It belongs to the type
24 STUDIES IN CHILD LOGIC
which we shall denote as " primitive argument " in our
essay in the following chapter on the different stages of
child conversation, and which we shall characterize by
just this lack of motivation.
The example given on page 22 (the argument between
Lev and a child of the same age about thunder) proves
this very clearly. Here are three more examples, the
first two quite definite, the third of a more intermediate
character.
Ez to Pie: " You're going to marry me Pie: No, I
won't marry you Oh yes you'll marry me No Yes
ptr "
* CLO.
" Look how lovely my 6 is going to be Lev : Yes it's
a 6 but really and truly ifs a g No, it's a 6, Nought
You said nought, and ifs not true, ifs a 9. Really it is
No Yes It was done like that already Oh no, thafs
a lie. You silly "
Lev looks to see what Hei is doing: " Two moons
No, two suns. Suns aren't like that, with a mouth.
They're like this, suns up there They're round Yes
they're quite round, but they haven't got eyes and a mouth.
Yes they have, they can see No they can't. Ifs only
God who can see"
In the first two examples the argument is simply a
clash of contrary affirmations, without mutual conces-
sions and without motivation. The last is more complex.
When Lev says "Ifs only God who can see ..." or
" They are like this", he does seem at the first glance to
be justifying his remarks, to be doing something more
than merely stating facts. But there is no explicit
justification, no attempt to demonstrate. Hei asserts
and Lev denies. Hei makes no effort to give any
reasons for believing that the sun has eyes, he does
not say that he has seen pictures which have led him
to such an idea, etc. Lev for his part does not attempt
to get at Hei's point of view, and gives no explicit
reason for defending his own. In the main then there
is still only a clash of assertions, different enough from
THE FUNCTIONS OF LANGUAGE 25
the two following little arguments, one of which, by the
way, takes place between a child and an adult.
These indeed are the only examples we have found
where the child tries to prove his assertions. They
should be carefully examined, considering how seldom
the fact occurs before the age of 7 or 8.
Lev talking to Mile G. : " You've been eating paint
No, I haven't, which? White paint No Oh> yes you
have 'cos theris some on your mouth"
The reader will note the correct use made of * be-
cause' at the age of 6J. In the three lists of complete
vocabularies given by Mile Descoeudres 1 * because* is
used by the seven-year-old but not by the five-year-old
child.
Here is another instance, again of Lev: " That is
420 But it's not the number of the house Why not?
The number of the house is on the door."
Note here the use of 'why' in the sense of "for
what reason" (cf. Chapter V). The reader will see
how superior these two arguments are to the preceding
examples.
We can draw the following conclusions from these
various facts :
i Adapted information, together with most of the
questions and answers which we shall examine later,
constitute the only categories of child language whose
function, in contrast to the divers functions of the
ego-centric categories, is to communicate intellectual
processes.
2 The frequency of adapted information is only
13% for Lev and 14% for Pie, a remarkable fact, and
one which shows how little the intellectual enquiry
of a child can be said to be social. These figures
are all the more striking when we remember that
1 A. Descoeudres, "Le developpement de 1' enfant de deux a sept ans,"
Coll. Actual. Fed., 1922, p. 190.
26 STUDIES IN CHILD LOGIC
collective monologue constitutes respectively 23% and
30% of the sum of the remarks made by the same
subjects.
3 These informations conveyed from one child to
another are factual in the sense that they do not point
to any causal relations, even when they deal with the
material used by the children in their work and with
the numerous objects, natural or artificial, which they
like to draw or build (animals, stars, motor-cars,
bicycles, etc.).
4 The arguments between the two children are, with
two exceptions only, of a low type, inasmuch as they
consist merely of a clash of contrary assertions without
any explicit demonstration.
7. CRITICISM AND DERISION. If we set aside ques-
tions and answers, the socialized language of the child
in its non-intellectual aspect may be divided into two
easily distinguishable categories : on the one hand com-
mands, on the other criticism and derision. There is
nothing peculiar about these categories in children ;
only their percentage is interesting.
Here are a few examples of criticisms, taunts, Schaden-
freude, etc., which at the first glance one might be
tempted to place under information and dialogue, but
which it will perhaps be found useful to class apart.
Their function is not to convey thoughts, but to satisfy
non-intellectual instincts such as pugnacity, pride,
emulation, etc. :
Lev : < * You're not putting it in the middle " (a plate on
the table). " That's not fair? "Pooh! that's no good!'
" We made that house, it isnt theirs? " Thafs not like an
owl Look, Pie, what he's done'' " Well, I know that he
can't? "Ifs much prettier than ours." a Tve got a much
bigger pencil than you'' " Well, I'm the strongest all the
same," etc.
All these remarks have this in common with adapted
information that they are addressed to a specified person
whom they influence, rouse to emulation and provoke
THE FUNCTIONS OF LANGUAGE 27
to retort and even to quarrelling. This is what obliges
us to class as socialized language such remarks as those
towards the end, beginning : " Well, I," which in other
respects resemble collective monologue. What, on the
other hand, distinguishes these phrases from information
proper, is that with the child even apparently objective
criticisms contain judgments of value which retain a
strongly subjective flavour. They are not mere state-
ments of fact. They contain elements of derision, of
combativeness, and of the desire to assert personal
superiority. They therefore justify the creation of a
separate category.
The percentage of this group is low : 3% for Lev,
and 7% for Pie. This may be a question of individual
types, and if this category is too weakly represented in
subsequent research, we may have to assimilate it to one
of the preceding ones.
8. COMMANDS, REQUESTS, THREATS. Why is the
ratio of adapted information so low in comparison to
that of the ego-centric forms of speech, particularly in
comparison to collective monologue? The reason is
quite simple. I The child does not in the first instance
communicate with his fellow-beings in order to share
thoughts and reflexions ; he does so in order to play.
The result is that the part played by intellectual inter-
change is reduced to the strictly necessary minimum.
The rest of language will only assist action, and will
consist of commands, etc.
Commands and threats, then, like criticisms, deserve
a category to themselves. They are, moreover, very
easy to recognize :
Lev (outside a shop) : " Mustrit come in here without
paying* I shall tell G/" (if you come). " Come here
Mr Passport" " Give me the blue one" " You must -make
a flag" " Come along, Ro. Look . . . you shall be the
cart" etc.
Pie : " E%) come and see the salamander" " Get out of
the way, I shan't be able to see" etc. (About a roof) : " No,
take it away, take it away ''cos I want to put on mine" etc.
28 STUDIES IN CHILD LOGIC
We need not labour the point. The only distinction
calling for delicate discrimination is that between requests
which tend imperceptably to become commands, and
questions which contain an implicit request. All requests
which are not expressed in interrogative form we shall
agree to call ' entreaties/ and shall include in the present
category ; while for interrogative requests a place will
be reserved in our next category. Here are some
examples of entreaty :
Lev : " The yellow paint \ please" " / should like some
water? etc.
Pie : " The india-rubber, teacher , / want the india-rubber"
Under requests, on the other hand, we shall classify
such sentences as: " Ez, do you mind helping me?"
"May I look at it?" etc. This distinction is certainly
artificial. But between an interrogative request and a
question bearing on immediate action there are many
intermediate types. And since it is desirable to dis-
tinguish between questions and commands, we must not
be afraid of facing the artificiality of our classification.
So long as we are agreed upon the conventions adopted,
and do not take the statistics too literally, the rest need
not detain us. It is not, moreover, the ratio of commands
to orders that will be of most use to us, but the ratio of
the bulk of socialized language to the bulk of ego-centric
language. It is easy enough to agree upon these funda-
mental distinctions.
The percentage of the present category is 10%
for Lev and 15% for Pie. Dialogue and inform-
ation were for the same subjects respectively, 12%
and 14%.
9. QUESTIONS AND ANSWERS. A preliminary diffi-
culty presents itself in connexion with these two
categories which we propose to treat of together : do
they both belong to socialized language? As far as
answers are concerned, we need be in no doubt. Indeed,
we shall describe as an ( answer ' the adapted words used
THE FUNCTIONS OF LANGUAGE 29
by the person spoken to, after he has heard and under-
stood a question. For instance :
"What colour is that? (Lev) Brownish yellow"
" What are you doing, Lev? The boat" etc.
To answers we shall assimilate refusals and accept-
ances, which are answers given not to questions of fact
but to commands and requests :
"Will you give it back to me? (the ticket). No y I
don't need it. Pm in the boat" (Lev).
These two groups, which together constitute answers,
obviously belong to socialized language. If we place
them in a separate category instead of assimilating
them to adapted information, it is chiefly because
answers do not belong to the spontaneous speech of the
child. It would be sufficient for his neighbours to
interrupt him and for adults to question him all the
time, to raise a child's socialized language to a much
higher percentage. We shall therefore eliminate
answers from our calculations in the following para-
graph. All remarks provoked by adults will thus be
done away with. Answers, moreover, constitute only
1 8% of Lev's language and 14% of Pie's.
The psychological contents of answers are highly
interesting, and would alone suffice to render the
category distinct from information. It is of course
closely connected to the contents of the question, and
we shall therefore deal simultaneously with the two
problems.
And the questions which children ask one another
do they too belong to socialized language? Curiously
enough the point is one that can be raised, for many
remarks are made by children in an interrogative form
without being in any way questions addressed to any-
one. The proof of this is that the child does not listen
to the answer, and does not even expect it. He supplies
it himself. This happens frequently between the ages
30 STUDIES IN CHILD LOGIC
of 3 and 5. At the age of our two subjects it is
rarer. When such pseudo-questions do occur, we have
classed them as monologue or information (e.g. " Please*
teacher is half right ? Yes, look " Lev). For the present
we shall therefore deal only with questions proper.
Questions make up 17% of Lev's language and
13% of Pie's. Their importance is therefore equal
and even superior to that of information, and since
a question is a spontaneous search for information,
we shall now be able to check the accuracy of our
assertions concerning this last category. Two of its
characteristics were particularly striking : the absence
of intellectual intercourse among the children on the
subject of causality, and the absence of proof and
logical justification in their discussions. If we jump
to the conclusion that children keep such thoughts
to themselves and do not socialize them, we may be
met with the counter assertion that children simply do
not have such thoughts, in which case there would be
no question of their socializing them ! This is partly
the case as regards logical demonstration. With
regard to causal explanation, however and by this
we mean not only the appeal to mechanical causality
such as is made only after the ages of 7 or 8 (see
Chapter V, 3), but also the appeal to final, or as we
shall call it, to pre-causality, z.., that which is invoked
in the child's * whys ' between the ages of 3 and 7
to 8 as regards this type of explanation, then, there
are two things to be noted. In the first place, the children
of the Maison des Petits deal in their drawings and free
compositions with animals, physical objects (stars, sky,
rain, etc.), with machines and manufactured objects
(trains, motors, boats, houses, bicycles, etc.). These
might therefore give rise to questions of origin and
causality. In the second place, * whys' play an
important part in all questions asked of grown-ups
by children under 7 (cf. p. 284 where out of three
groups of 250 spontaneous questions we noted respec-
THE FUNCTIONS OF LANGUAGE 31
tively 91, 53 and 41 * whys'). Now among these
'whys' a large number are "whys of explanation,"
meaning "for what reason" or "for what object."
Explanation supplies about 18% of the subject-matter
dealt with in the questions of the child of 6 or 7,
such as we shall study it in Chapter V. If, there-
fore, there are few questions of explanation in the
talk of our two present subjects, this is strongly in
favour of the interpretation we have given of information
and dialogue between children in general. Intellectual
intercourse between children is still factual or descriptive,
/.&, little concerned with causality, which remains the
subject of conversation between children and adults or
of the child's own solitary reflexion.
The facts seem to bear this out. Only 3 out of Pie's
173 questions are 'whys.' Out of Lev's 224 questions
only 10 are ' whys '. Of these, only two ' whys ' of Lev's
are "whys of explanation." 1
" Why has he turned round?" (a stuffed owl which
Lev believes to be alive), and " Why has he turned round
a little?" (the same).
The rest are ' whys ' not of causal but of psychological
explanation, ' intentions ' as we shall call them, 2 which
is quite another matter :
"Why did he say : 'Hullo Lev* ?" "Why was Rey
crying" ? " " Why has he gone away ? " etc.
In addition to these we have one " logical why " from
Lev, that which we dealt with in connexion with the
discussion on page 25. It is clear how rarely children
ask each other 'why?', and how little such questions
have to do with causality.
Thus out of the 224 questions asked by Lev and the
173 asked by Pie only two are about explanation, and
those two both come from Lev. All the rest can be
divided as follows. First of all, we have 141 questions
1 For the definition of this term, see Chapter V. 2 Id.
32 STUDIES IN CHILD LOGIC
of Lev's and 78 of Pie's about children's activities as
such, about " actions and intentions" : l
Lev: "And my scissors, can you see them?" "Are
we going to play at Indians ?" " Pm working, are you ? "
" / didn't hurt you, did I?" "Do you know that gentle-
man?" "How shall I paint the house?" "How does
this go?" (d ball in the counting-frame).
Pie : "Are you coming this afternoon, Bea ?" "I say,
have you finished yet ? " etc.
This enormous numerical difference between the
questions bearing upon children's activity as such, and
those dealing with causal explanation is very remark-
able. 5j[t proves how individualistic the child of 6
still shows himself to be in his intellectual activity,
and how restricted in consequence is the interchange
of ideas that takes place between children.!
A second category of questions, made up of 27 of
Lev's and 41 of Pie's, deals with facts and events, time
and place (questions of ( reality ' treated of in Chapter V).
Facts : "Is your drum closed? " "Is there some paper,
too ?" "Are there snails in there ? " (Pie.)
Place: "Where is the blue, Ez?" "Where is she?"
(the tortoise).
Time: "Please teacher, is it late?" "How old are
you ? " (Pie.)
It will be seen that these questions do not touch upon
causality, but are^all about matters of fact Questions
of place predominate in this category, 29 for Pie and
13 for Lev.
Another numerous category (51 for Pie, 48 for Lev)
is made up of questions purely concerning matters of
fact, questions of nomenclature, classification and
evaluation.
Nomenclature : " What does * behind ' mean ?" " What
is he called? " (a cook) (Lev).
Classification: "What ever is that?" "Is that
yellow ? " (Lev).
Evaluation: " Is it pretty ?" (Lev, Pie).
1 See Chapter V.
THE FUNCTIONS OF LANGUAGE
33
We may add a few questions about number (5 by
Lev, i by Pie) :
"Isn't all that enough for zfr.$o ?" " And how much
for n?" (Lev).
Finally, mention should be made of two questions
by Pie and one by Lev about rules (writing, etc.).
" You put it on this side, dorit you?" (the figure 3)
(Lev).
The following table completely summarizes the ques-
tions asked by Lev and Pie, including their < whys/
Questions of causal explanation .
Questions of ! Facts and events '
Actions and intentions
Rules
Questions of / Nomenclature
Classification! Classification and evaluation
Number
TOTAL
LEV
PIE
2
2
141
o
o
4i
78
7
7
13
8
4
29
I
...
2
7
48
5
o
51
7
224
173
We shall not dwell upon the criteria of the different
categories nor upon their functional interest ; these
problems form the subject-matter of a later chapter on
" A child's questions " (Chapter V). It will be enough
if we conclude from this table that questions from one
child to another (questions from children to adults play
only a negligible part in this group), bear first and
foremost upon actual psychological activity (actions
and intentions). Otherwise, when they concern objects
34 STUDIES IN CHILD LOGIC
and not persons, they bear upon the factual aspect of
reality, and not upon causal relations. These con-
clusions are markedly different from the results supplied
by Del (Chapter V: Questions of a child to an adult).
Before drawing any conclusions, however, from the
difference between questions from child to child and
questions from child to adult, we should have to solve
a big preliminary problem : how far do the questions
which Lev and Pie ask adults out of school hours
resemble those of Del (whys of explanation, etc.) ? At
the first glance, Del, although he has worked like the
others during school hours, seems to approximate much
more closely to what we know of the ordinary question- -
ing child of 6. But Lev and Pie are perhaps special
types, more prone to statement and less to explanation.
All we can do, therefore, is to extend the work of
research as carried out in this chapter and in Chapter V.
II. CONCLUSIONS
Having defined, so far as was possible the various
categories of the language used by our two children, it
now remains for us to see whether it is not possible to
establish certain numerical constants from the material
before us. We wish to emphasize at the very outset
the artificial character of such abstractions. The number
of unclassifiable remarks, indeed, weighs heavily in the
statistics. In any case, a perusal of the list of Lev's
first 50 remarks, which we shall give as an example for
those who wish to make use of our method, should give
a fair idea of the degree of objectivity belonging to our
classification. 1 But these difficulties are immaterial.
If among our results some are definitely more constant
than others, then we shall feel justified in attributing
to these a certain objective value.
10. THE MEASURE OF EGO-CENTRISM. Among the
data we have obtained there is one, incidentally of the
1 See Appendix.
THE FUNCTIONS OF LANGUAGE 35
greatest interest for the 'study of child logic, which
seems to supply the necessary guarantee of objectivity :
we mean the proportion of ego-centric language to
the sum of the child's spontaneous conversation.
Ego-centric language is, as we have seen, the group
made up by the first three of the categories we have
enumerated repetition, monologue and collective mono-
logue. All three have this in common that they con-
sist of remarks that are not addressed to anyone, or
not to anyone in particular, and that they evoke no
reaction adapted to them on the part of anyone to
whom they may chance to be addressed. Spontaneous
language is therefore made up of the first seven cate-
gories, i.e., of all except answers. It is therefore the sum
total of all remarks, minus those which are made as an
answer to a question asked by an adult or a child. We
have eliminated this heading as being subject to chance
circumstances ; it is sufficient for a child to have come
in contact with many adults or with some talkative com-
panion, to undergo a marked change in the percentage
of his answers. Answers given, not to definite questions
(with interrogation mark) or commands, but in the
course of the dialogue, i.e., propositions answering to
other 'propositions, have naturally been classed under
the heading information and .dialogue, so that there is
nothing artificial about the omission of questions from
the statistics which we shall give. The child's language
minus his answers constitutes a complete whole in
which intelligence is represented at every stage of its
development.
The proportion of ego-centric to other spontaneous
forms of language is represented by the following
fractions :
= o-47for Lev, = o.43 for Pie.
(The proportion of ego-centric language to the sum
total of the subject's speech, including answers, is 39% for
36 STUDIES IN CHILD LOGIC
Lev and 37% for Pie.) The similarity of result for Lev
and Pie is a propitious sign, especially as what differ-
ence there is corresponds to a marked difference of
temperament. (Lev is certainly more ego-centric than
Pie.) But the value of the result is vouched for in yet
another way.
If we divide the 1400 remarks made by Lev during
the month in which his talk was being studied into
sections of 100 sentences, and seek to establish for each
section the ratio ^' ", the fraction will be found to
op. jL.
vary only from 0*40 to 0*57, which indicates only a
small maximum deviation. On the contrary, the mean
variation , i.e., the average of the deviations between each
value and the arithmetical average of these values, is
only 0*04, which is really very little.
If Pie's 1500 remarks are submitted to the same treat-
ment, the proportions will be found to vary between
0*31 and 0*59, with an average variation of 0*06. This
greater variability is just what we should expect from
what we know of Pie's character, which at first sight
seems more practical, better adapted than Lev's, more
inclined to collaboration (particularly with his bosom
friend Ez). But Pie every now and then indulges in
fantasies which isolate him for several hours, and
during which he soliloquizes without ceasing.
We shall see in the next chapter, moreover, that
these two coefficients do actually represent the average
for children between the ages of 7 and 8. The same
calculation based on some 1500 remarks in quite another
class-room yielded the result of 0*45 (a. v. = 0*05).
This constancy in the proportion of ego-centric
language is the more remarkable in view of the fact
that we have found nothing of the kind in connexion
with the other coefficients which we have sought to
establish. We have, it is true, determined the pro-
portion of socialized factual language (information and
questions) to socialized non-factual language (criticism,
THE FUNCTIONS OF LANGUAGE 37
commands, and requests). But this proportion fluctuates
from 072 to 2*23 with a mean variation 0*71 for Lev
(as compared with 0*04 and 0*06 as the coefficients of
ego-centrism), and between 0*43 and 2*33 with a mean
variation of 0*42 for Pie, Similarly, the relation of
ego-centric to socialized factual language yields no
coefficient of any constancy.
Of all this calculation let us bear only this in mind,
that our two subjects of 6J have each an ego-centric
language which amounts to nearly half of their total
spontaneous speech.
The following table summarizes the functions of the
language used by both these children :
Pie Lev
1 Repetition 2 i
2 Monologue 5 15
3 Collective Monologue ... 30 23
4 Adapted Information .... 14 13
5 Criticism 7 3
6 Commands 15 10
7 Requests 13 17
8 Answers 14 18
Ego-centric Language 37 39
Spontaneous Socialized language . 49 43
Sum of Socialized language . 63 ' 61
Coefficient of Ego-centrism . . 0*43 + 0*06 0*47 + 0*04
We must once more emphasize the fact that in all
these calculations the number of remarks made by
children to adults is negligible. By omitting them we
raise the coefficient of ego-centrism to about 0*02, which
is within the allowed limits of deviation. In future,
however, we shall have completely to eliminate such
remarks from our calculations, even if it means making
a separate class for them. We shall, moreover, observe
this rule in the next chapter where the coefficient of
ego-centrism has been calculated solely on the basis of
remarks made between children.
11. CONCLUSION. What are the conclusions we
can draw from these facts? It would seem that up to
38 STUDIES IN CHILD LOGIC
a certain age we may safely admit that children think
and act more ego-centrically than adults, that they share
each other's intellectual life less than we do. True,
when they are together they seem to talk to each other
a great deal more than we do about what they are
doing, but for the most part they are only talking to
themselves. We, on the contrary, keep silent far
longer about our action, but our talk is almost always
socialized.
Such assertions may seem paradoxical. In observ-
ing children between the ages of 4 and 7 at work
together in the classes of the Maison des Petits, one is
certainly struck by silences, which are, we repeat, in
no way imposed nor even suggested 'by the adults.
One would expect, not indeed the formation of working
groups, since children are slow to awake to social life,
but a hubbub caused by all the children talking at
once. This is not what happens. All the same, it
is obvious that a child between the ages of 4 and 7,
placed in the conditions of spontaneous work provided
by the educational games of the Maison des Petits^
breaks silence far oftener than does the adult at work,
and seems at first sight to be continuously communi-
cating his thoughts to those around him.
Ego-centrism must not be confused with secrecy.
Reflexion in the child does not admit of privacy.
Apart from thinking by images or autistic symbols
which cannot be directly communicated, the child up
to an age, as yet undetermined but probably some-
where about seven, is incapable of keeping to himself
the thoughts which enter his mind. He says every-
thing. He has no verbal continence. Does this mean
that he socializes his thought more than we do ? That
is the whole question, and it, is for us to see to whom
the child really speaks. It may be to others. We
think on the contrary that, as the preceding study shows,
it is first and foremost to himself, and that speech,
before it can be used to socialize thought, serves to
THE FUNCTIONS OF LANGUAGE 39
accompany and reinforce individual activity. Let us
try to examine more closely the difference between
thought which is socialized but capable of secrecy, and
infantile thought which is ego-centric but incapable of
secrecy.
The adult, even in his most personal and private
occupation, even when he is engaged on an enquiry
which is incomprehensible to his fellow-beings, thinks
socially, has continually in his mind's eye his colla-
borators or opponents, actual or eventual, at any rate
members of his own profession to whom sooner or
later he will announce the result of his labours. This
mental picture pursues him throughout his task. The
task itself is henceforth socialized at almost every stage
of its development. Invention eludes this process, but
the need for checking and demonstrating calls into
being an inner speech addressed throughout to a hypo-
thetical opponent, whom the imagination often pictures
as one of flesh and blood. When, therefore, the adult
is brought face to face with his fellow-beings, what
he announces to them is something already socially
elaborated and therefore roughly adapted to his audience,
i.e., it is comprehensible. Indeed, the further a man
has advanced in his own line of thought, the better
able is he to see things from the point of view of others
and to make himself understood by them.
The child, on the other hand, placed in the con-
ditions which we have described, seems to talk far more
than the' adult. Almost everything he does is to the
tune of remarks such as "I'm drawing a hat," "I'm
doing it better than you," etc. Child thought, there-
fore, seems more social, less capable of sustained and
solitary research. This is so only in appearance. The
child has less verbal continence simply because he
does not know what it is to keep a thing to himself.
Although he talks almost incessantly to his neighbours,
he rarely places himself at their point of view. He
speaks to them for the most part as if he were alone,
40 STUDIES IN CHILD LOGIC
and as if he were thinking aloud. He speaks, there-
fore, in a language which disregards the precise shade
of meaning in things and ignores the particular angle
from which they are viewed, and which above all is
always making assertions, even in argument, instead
of justifying them. Nothing could be harder to under-
stand than the note-books which we have filled with
the conversation of Pie and Lev. Without full com-
mentaries, taken down at the same time as the children's
remarks, they would be incomprehensible. Everything
is indicated by allusion, by pronouns and demonstrative
articles " he, she, the, mine, him, etc." which can
mean anything in turn, regardless of the demands of
clarity or even of intelligibility. (The examination of
this style must not detain us now ; it will appear again
in Chapter III in connexion with verbal explanation
between one child and another.) In a word, the child
hardly ever even asks himself whether he has been
understood. For him, that goes without saying, for
he does not think about others when he talks. He
utters a " collective monologue." His language only
begins to resemble that of adults when he is directly
interested in making himself understood ; when he
gives orders or asks questions. To put it quite simply,
we may say that the adult thinks socially, even when
he is alone, and that the child under 7 thinks ego-
centrically, even in the society of others.
What is the reason for this? It is, in our opinion,
twofold. It is due, in the first place, to the absence
of any sustained social intercourse between the children
of less than 7 or 8, and in the second place to the
fact that the language used in the fundamental activity
of the child play is one of gestures, movement
and mimicry as much as of words. (There is, as
we have said, no real social life between children of
less than 7 or 8 years, x The type of children's society
represented in a classroom of the Maison des Petits
is obviously of a fragmentary character, in which con-
THE FUNCTIONS OF LANGUAGE 41
sequently there is neither division of work, centraliza-
tion of effort, nor unity of conversation. We may go
further, and say that it is a society in which, strictly
speaking, individual and social life are not differentiated.
An adult is at once far more highly individualized and
far more highly socialized than a child forming part of
such a society. He is more individualized, since he
can work in private without perpetually announcing
what he is doing, and without imitating his neighbours.
He is more socialized for the reasons which have just
given. The child is neither individualized, since he
cannot keep a single thought secret, and since every-
thing done by one member of the group is repeated
through a sort of imitative repercussion by almost
every other member, nor is he socialized, since this
imitation is not accompanied by what may properly
be called an interchange of thought, about half the
remarks made by children being ego-centric in character.
If, as Baldwin and Janet maintain, imitation is accom-
panied by a sort of confusion between one's own
action and that of others, then we may find in this
fragmentary type of society based on imitation some
sort of explanation of/ the paradoxical character of the
conversation of children who, while they are continu-
ally announcing their doings, yet talk only for them-
selves, without listening to anyone else. !
Social life at the Maison des Petits passes, according
to the observations of Miles Audemars and Lafendel,
through three stages. |Up till the age of about 5, the
child almost always works alone % From 5 to about
7j, little groups of two are formed, like that of Pie and
Ez (cf. the remarks taken down under the heading
"adapted information.") These groups are transitory
and irregular. Finally, between 7 and 8 the desire
manifests itself to work with others. I Now it is in
our opinion just at this age that ego-centric talk loses
some of its importance, and it is at this age, as we
shall see in the next chapter, that we shall place the
42 STUDIES IN CHILD LOGIC
higher stages of conversation properly so-called as it
takes place between children, jit is also at this age,
(cf. Chapter III) that children! begin to understand
each other in spoken explanations, as opposed to
explanations in which gestures play as important a
part as words. ;
A simple w^y of verifying these hypotheses is to
re-examine children between 7 and 8 whose ego-centrism
at an earlier stage has been ascertained. This is the
task which Mile. Berguer undertook with Lev. She
took down under the same conditions as previously
some 600 remarks made by Lev at the age of 7 and
a few months. The co-efficient of ego-centricism was
reduced to o'2j. 1
These stages of social development naturally concern
only the child's intellectual activity (drawings, con-
structive games, arithmetic, etc.). It goes without
saying that in outdoor games the problem is a com-
pletely different one ; but these games touch only on
a tiny portion of the thought and language of the
child.
If language in the child of about 6| is still so far
from being socialized, and if the part played in it by
the ego-centric forms is so considerable in comparison
to information and dialogue etc., the reason for this
lies in the fact thal| childish language includes two
distinct varieties, on4 made up of gestures, movements,
mimicry etc., which accompany or even completely
supplant the use of words, and the other consisting
solely of the spoken word./ Now, gesture cannot
express everything. Intellectual processes, therefore,
will remain ego-centric, whereas commands etc., all
the language that is bound up with action, with
handicraft, and especially with play, will tend to
1 We are at the moment collecting similar data from various children
between the ages of 3 and' 7, in such a way as to establish a graph of
development. These results will probably appear in the Archives de
Psychologic*
THE FUNCTIONS OF LANGUAGE 43
become more socialized. We shall come across this
essential distinction again in Chapter III. It will then
be seen that verbal understanding between children
is less adequate than between adults, but this does
not mean that in their games and in their manual
occupations they do not understand each other fairly
well ; this understanding, however, is not yet altogether
verbal.
12. RESULTS AND HYPOTHESES. Psycho-analysts
have been led to distinguish^two fundamentally different
modes of thinking: directed or intelligent thought, and
undirected or, as Bleuler proposes to call it, autistic
thought. Directed thought is conscious, i.e., it pursues
an aim which is present to the mind of the thinker ; it is
intelligent, which means that it is adapted to reality
and tries to influence it ; it admits of being true or
false (empirically or logically true), and it can be
communicated by language. Autistic thought is sub-
conscious, which means that the aims it pursues and
the problems it tries to solve are not present in con-
sciousness ; it is not adapted to reality, but creates for
itself a dream world of imagination ; it tends, not to
establish truths, but so to satisfy desires, and it remains
strictly individual and incommunicable as such by means
of language. On the contrary, it works chiefly by
images, and in order to express itself, has recourse to
indirect methods, evoking by means of symbols and
myths the feeling by which it is led. i
Here, then, are two fundamental modes of thought
which, though separated neither at their origin nor
in the course of their functioning are subject, never-
theless, to two diverging sets of logical laws. 1 Directed
thought, as it develops, is controlled more and more
by the laws of experience and of logic in the stricter
sense. Autistic thought, on the other hand, obeys a
1 There is interaction between these two modes of thought. Autism
undoubtedly calls into being and enriches many inventions which are
subsequently clarified and demonstrated by intelligence.
44 STUDIES m CHILD LOGIC
whole system of special laws (laws of symbolism and
of immediate satisfaction) which we need not elaborate
here. Let us consider, for instance, the completely
different lines of thought pursued from the point of
view of intelligence and from that of autism when we
think of such an object as, say, water.
To intelligence, water is a natural substance whose
origin we know, or whose formation we can at least
empirically observe ; its behaviour and motions are
subject to certain laws which can be studied, and it
has from the dawn of history been the object of technical
experiment (for purposes of irrigation, etc.). To the
autistic attitude, on the other hand, water is interest-
ing only in connexion with the satisfaction of organic
wants. It can be drunk. But as such, as well as simply
in virtue of its external appearance, it has come to
represent in folk and child fantasies, and in those of
adult subconsciousness, themes of a purely organic
character. It has in fact been identified with the
liquid substances which issue from the human body,
and has come, in this way, to symbolize birth itself,
as is proved by so many myths (birth of Aphrodite,
etc.), rites (baptism the symbol of a new birth), dreams l
and stories told by children. 2 Thus in the one case
thought adapts itself to water as part of the external
world, in the other, thought uses the idea of water not
in order to adapt itself to it, but in order to assimilate
it to those more or less conscious images connected
with fecundation and the idea of birth.
Now these two forms of thought, whose characteris-
tics diverge so profoundly, differ chiefly as to their origin,
the one being socialized and guided by the increasing
1 See Flournoy, H. <f Quelques reVes au sujet de la signification symbolique
de 1'eau et du feu." Intern. Zeitschr. f. Psychoan., Vol. VI. p. 398 (cf.
pp. 329 and 330).
2 We have published the case of Vo of a child of 9, who regards humanity
as descended from a baby who issued from a worm which came out of the sea.
Cf. Piaget, "La pensee symbolique et la pensee de Tcnfant. Arch. Psych. ,
Vol. XVIII, 1923.
THE FUNCTIONS OF LANGUAGE 45
adaptation of individuals one to another, whereas
the other remains individual and uncommunicated.
Furthermore and this is of the very first importance
for the understanding of child thought this divergence
is due in large part to the following fact. Intelligence,
just because it undergoes a gradual process of social-
ization, is enabled through the bond established by
language between thoughts ' and words to make an
increasing use'of concepts ; whereas autism, just because
it remains individual, is still tied to imagery, to organic
activity, and even to organic movements. The mere
fact, then, of telling one's thought, of telling it to others,
or of keeping silence and telling it only to oneself must
be of enormous importance to the fundamental structure
and functioning of thought in general, and of child
logic in particular. Now between autism and intelli-
gence there are many degrees, varying with their
capacity for being communicated. These intermediate
varieties must therefore be subject to a special logic,
intermediate too between the logic of autism and that
of intelligence. The chief of those intermediate forms,
i.e., the type of thought which like that exhibited by
our children seeks to adapt itself to reality, but does
not communicate itself as such, we propose to call
Ego-centric thought. This gives us the following table :
Non-communicable Communicable thought
thought
Undirected thought Autistic thought (Mythological thought)
Directed thought Ego-centric thought Communicated
intelligence
We shall quickly realize the full importance of ego-
centrism if we consider a certain familiar experience of
daily life. We are looking, say, for the solution of
some problem, when suddenly everything seems quite
clear ; we have understood, and we experience that sui
generis feeling of intellectual satisfaction. But as soon
as we try to explain to others what it is we have under-
stood, difficulties come thick and fast. These difficulties
46 STUDIES IN CHILD LOGIC
do not arise merely because of the effort of attention
needed to hold in a single grasp the links in the chain
of argument; they are attributable also to our judging
faculty itself. Conclusions which we deemed positive
no longer seem so ; between certain propositions whole
series of intermediate links are now seen to be lacking
in order to fill the gaps of which we were previously not
even conscious ; arguments which seemed convincing
because they were connected with some schema of visual
imagery or based on some sort of analogy, lose all their
potency from the moment we feel the need to appeal to
these schemas, and find that they are incommunicable ;
doubt is cast on propositions connected with judgments
of value, as soon as we realize the personal nature of
such judgments. If such, then, is the difference be-
tween personal understanding and spoken explanation,
how much more marked will be the characteristics of
personal understanding when the individual has for a
long time been bottling up his own thoughts, when
he has not even formed the habit of thinking in terms
of other people, and of communicating his thoughts to
them. We need only recall the inextricable chaos
of adolescent thought to realize the truth of this
distinction.
Ego-centric thought and intelligence therefore repre-
sent two different forms of reasoning, and we may even
say, without paradox, two different logics. By logic is
meant here the sum .of the habits which the mind adopts
in the general conaucrof its operations in the general
conduct of a game of chess, in contrast, as Poincare
says, to the special rules which govern each separate
proposition, each particular move in the game. Ego-
centric logic and communicable logic will therefore
differ less in their conclusions (except with the child
where ego-centric logic often functions) than in the way
they work. The points of divergence are as follows :
i Ego-centric logic is more intuitive, more *syn-
cretistic' than deductive, i*e n its reasoning is not made
THE FUNCTIONS OF LANGUAGE 47
explicit. The mind leaps from premise to conclusion
at a single bound, without stopping on the way. 2
Little value is attached to proving, or even checking
propositions. The vision of the whole brings about a
state of belief and a feeling of security far more rapidly
than if each step in the argument were made explicit.
3 Personal schemas of analogy are made use of, like-
wise memories of earlier reasoning, which control the
present course of reasoning without openly manifesting
their influence. 4 Visual schemas also play an im-
portant part, and can even take the place of proof in
supporting the deduction that is made. 5 Finally,
judgments of value have far more influence on ego
centric than on communicable thought.
In communicated intelligence, on the other hand, we
find i far more deduction, more of an attempt to render
explicit the relations between propositions by such ex-
pressions as therefore^ if . . . then^ etc. 2 Greater
emphasis is laid on proof. Indeed, the whole exposi-
tion is framed in view of the proof, i.e., in view of
the necessity of convincing someone else, and (as a
corollary) of convincing oneself whenever one's personal
certainty may have been shaken by the process of
deductive reasoning. ^ Schemas of analogy tend to
be eliminated, ancl to be replaced by deduction proper.
4 Visual schemas are also done away with, first as
incommunicable, and later as useless for purposes of
demonstration. 5 Finally personal judgments of value
are eliminated in favour of collective judgments of value,
these being more in keeping with ordinary reason.
If then the difference between thought that can be com-
municated and what remains of ego-centric thought in
the adult or the adolescent is such as we have described
it, how much more emphasis shall we be justified in
laying on the ego-centric nature of thought in the child.
It is chiefly in connexion with children between 3 to 7
and, to a lesser degree, with those between 7 to 1 1 that
we have endeavoured to distinguish ego-centric thought.
48 STUDIES IN CHILD LOGIC
In the child between 3 and 7 the five characteristics
which have just been enumerated actually go to make
up a kind of special logic which we shall have occasion
to mention throughout this volume and the next. Be-
tween 7 and 1 1 this ego-centric logic no longer influences
what Binet and Simon call the "perceptual intelligence"
of the child, but it is found in its entirety in his " verbal
intelligence." In the following chapters we shall study
a large number of phenomena due to ego-centrism,
which, after having influenced the perceptual intelli-
gence of children between the ages of 3 and 7, influence
their verbal intelligence between the ages of 7 and n.
We are now therefore in a position to realize that the
fact of being or of not being communicable is not an
attribute which can be added to thought from the out*
side, but is a constitutive feature of profound significance
for the shape and structure which reasoning may
assume.
The question of communicability has thus proved
itself to be one of those preliminary problems which
must be solved as an introduction to the study of child
logic. There are other such problems, all of which can
be classed under two main headings.
A. Communicability : (i) To what extent do children
of the same age think by themselves, and to what extent
do they communicate with each other? (2) Same
question as between older and younger children, (a)
of the same family ; (b) of different families. (3) Same
question as between children and parents.
B. Understanding: (i) To what extent do children of
the same age understand each other? (2) Same question
as between older and younger children (of the same
and of different families). (3) Same question as between
children and parents.
The problems of the second group will be dealt with
in a subsequent chapter. As to group A, we think
that we have supplied a partial solution to the first of its
problems. If it be granted that the first three categories
THE FUNCTIONS OF LANGUAGE 49
of child language as we have laid them down are ego-
centric, then the thought of the child of 6| Is in its
spoken manifestation ego-centric in the proportion of
44 to 47%. What is socialized by language, moreover,
belongs only to the factual categories of thought.
At this age, causality and the faculty for explanation
are still unexpressed. Does the period between 6 and 7
mark a turning point in this respect? We still lack
the material to make a sufficient number of comparisons,
but judging from what seems to be the rule at the
Maison des Petits^ we believe that the age at which the
child begins to communicate his thought (the age when
ego-centric language is 25%) is probably somewhere
between 7 and 8. This does not mean that from the
age of 7 or 8 children can immediately understand each
other we shall see later on that this is far from being
the case it simply means that from this age onwards
they try to improve upon their methods of interchang-
ing ideas and upon their mutual understanding of one
another.
CHAPTER II
TYPES AND STAGES IN THE CON-
VERSATION OF CHILDREN BE-
TWEEN THE AGES OF FOUR AND
SEVEN. 1
THIS chapter continues the preceding one and also
completes it. Our aim has simply been i to check the
statistical data obtained from observations made on Pie
and Lev, 2 to establish a certain number of types of
conversation held between children of the same age ;
these being on a wider scale than the types of simple
propositions examined in the last chapter, and capable
eventually of representing the successive stages of
childish conversation between the ages of 4 and 7,
The conclusions of Chapter I may well appear pre-
sumptuous, based as they were on observation of two
children only, i.e., of two psychological types at the
outside. The same experiment needed to be carried
out on a whole group of children, and thus reach the
greatest possible variety of psychological types. It is
such an experiment as this which will be described in
the present chapter. The subject of analysis will now
be the verbatim report of conversations held, not by
one or two specified children, but by the inmates of a
whole room, in which they move about from one place
to another and which they enter and leave at will.
What has been taken down is really the outcome of
observations made from a fixed place upon some twenty
children on the move. In the Maison des Petits, where
1 With the collaboration of Mme Valentine Jean Piaget. We also wish to
thank Mile G, Guex who helped us to collect our materials.
so
CHILDREN'S CONVERSATIONS 51
all these observations have been made, the children
between the ages of 4 and 7 occupy a whole floor of
five rooms (arithmetic room, building room, modelling
room, etc.), and they move about just as they please
from one room to another, without being compelled to
do any consecutive work. It is in one of these rooms
that the data were collected which will form the subject
of our present enquiry.
i. CHECK OF THE COEFFICIENT OF EGOCENTRISM.
One of the first results of these verbatim reports was
to show that the talk taken down could be classed in
the same categories as those which had been used for
Lev and Pie. The language of our 20 subjects, while
it reflects differences of temperament, remains the out-
come of the same functional needs. In the domineering
child there will be an increase of orders, threats, criti-
cisms, and arguments, while the more dreamily inclined
will indulge in a greater number of monologues. The
proportions will differ, but in each child all the cate-
gories will be represented. The difference will be one
of quantity not of quality.
Now in Lev and Pie, who represent fairly different
types, the coefficients of ego-centrism are very close (0*47
and 0*43). Can we infer from this that the average
coefficient between 4 and 7 will be 0*45 or somewhere
near? The calculation was made on the sum total of
the remarks made by our 20 subjects (boys and girls
differing in race and upbringing). The same procedure
was adopted as before of taking successive sections of
100 sentences each. These 100 consecutive sentences
are thus no longer the successive remarks of one child,
but the general conversation in a given room where
there are always three or four children talking together.
There is therefore every chance that the calculation
will yield objectively valid results. Now the average
coefficient of ego-centrism which was reached in this
way was of 0*45 0*05, representing the proportion of
the ego-centric categories to the total language minus
52 STUDIES IN CHILD LOGIC
answers. As the average age of the children is 6, this
is an interesting confirmation of the conclusions of the
last chapter.
2. TYPES OF CONVERSATION BETWEEN CHILDREN.
In the first chapter we established a certain number
of types of childish talk, but this was done according
to type, and not according to the stage of development
reached, i.e., without regard to the problem of the
development of these types in relation to one another,
and of conversation in general among children. This
is the problem which we must now approach. We
had, moreover, been entirely concerned with isolated
remarks, viewed of course in relation to their context,
but numbered and classified sentence by sentence. We
shall now have to find types, not of isolated, but of
general conversation, and these will be partly inde-
pendent of the earlier types, partly related to them in
a manner which we shall specify later on.
When, in the first place, can conversation properly
be said to take place between children? Whenever
to fix an arbitrary minimum three consecutive remarks
about the same subject are made by at least two inter-
locutors. Here are two of the simpler possible schemas
of conversation :
I (i) Remark by A. II (i) Remark by A,
(2) Remark by B adapted (2) Remark by B adapted
to (i). to (i).
(3) Remark by A adapted (3) Remark by C adapted
to (2). to (i) or (2).
After this, all conversation will consist of the language
which we have described as socialized. A's remarks
may be informations, criticisms, orders, or questions.
The remarks made by B and C may belong to those
same four groups, or come under answers. But, as
we have said, types of conversation will be on a wider
scale than types of remarks, and will be independent
of these. Thus informations, questions, commands,
CHILDREN'S CONVERSATIONS 53
etc., may all appear as constituents of a single con-
versational type X. The questions we have to answer
may therefore be stated as follows : i What are the
types of conversation between children? 2 Are these
types contemporaneous, or do they represent different
stages of development? 3 If they constitute stages,
what is their genesis? Are they derived from ego-
centric language ? If so, what is the process of evolution
by which a child passes from ego-centric language to
the higher types of conversation ?
Now it seems to us possible to establish certain stages
from a point which has not yet reached the level of
conversation, but is represented by the collective mono-
logue. This leads us to the following table. We offer
it with due reservation, and chiefly with the object of
ordering and schematizating our classifications.
T onologue : Conversations :
Stage I Stage I IA Stage HA Stage IIlA
(first type) (second type)
The hearer is assort- Collaboration Collaboration zt
ated with the speakers in action or non- abstract thought
action and thought ^~~ abstract thought
(without collaboration)
Jollective
Stage IlB Stage I IB Stage 1 1 IB
(first type) -> (second type) ~> Genuine argu
Quarrel < Primitive ar- ment
gttment
(Clash of contrary (Clash of un- (Clash of moti
actions) motivated asser- vated assertions)
tions)
Stage I still partakes to a certain extent of the nature
of ego-centric thought as it was described in the pre-
ceding chapter. At this first stage there is, strictly
speaking, no conversation, since each child speaks only
to himself, even when he seems to be addressing some-
one in particular. Besides, the children never speak
about the same thing. And yet this collective mono-
logue forms the starting point of childish conversation,
because it is made up of separate groups, of bundles
54 STUDIES IN CHILD LOGIC
of consecutive remarks. When a child talks in this
way, others will immediately answer by talking about
themselves, and this gives rise to a sequence of four
or five remarks, which form a conversational embryo,
without, however, transcending the stage of the collective
monologue.
Stages II and III, on the other hand, have some of
the characteristics of conversation properly so-called and
of socialized language. We have divided them into two
series, A and B, which are parallel from the genetic
point of view. (II A corresponds to I!B, and III A to
IIlB); the A series has as its origin an agreement in
action and opinion (progressive collaboration), the B
series has as its origin a disagreement, which begins
with a simple quarrel and may evolve into more or less
perfected arguments.
Stage HA can be represented in either of two con-
temporaneous types. The first type (where the speaker
associates his hearer with his own actions and thoughts)
is represented by those conversations in which the child,
although he only talks about what he is doing, associates
with it the person to whom he is talking. There is
association in the sense that every one listens to and
understands the speaker, but there is no collaboration
because each child speaks only of himself, of his own
action, or of his own thoughts.
In the second type there is collaboration in action or
in thought connected with action (non-abstract thought)
in the sense that the conversation bears upon an activity
which is shared by the talkers. The subject of conversa-
tion is thus some definite action, and not the explanation
of a past or future action. It may also happen at this
stage that some common memory is evoked, but there
is never any question of explaining this memory (e.g.
reconstructing some previously heard explanation) nor
of discussing it (looking for what is, and what is not true
in the memory or in the circumstances which complete
it). The memory recalled in common at this HA stage
CHILDREN'S CONVERSATIONS 55
serves purely as a stimulus. One evokes it just as one
tells a story, for the mere pleasure of doing so ("Do
you remember how " . . . etc).
It is not until we come to stage III A, that we meet
with abstract collaboration. By abstract we wish to
designate those mental processes in the child which are
no longer connected with the activity of the moment,
but are concerned with finding an explanation, recon-
structing a story or a memory, discussing the order of
events or the truth of a tale.
The passage from one to the other of these two stages
of the series A shows us a progressive socialization of
thought. There was no a priori reason why these three
types of conversation should represent successive stages.
Type III A might quite conceivably have appeared before
type II A or the collective monologue, or simultaneously
with them. As a matter of fact we shall see that this is
not what happens ; we shairiee that there is progression
according to age and in conformity with the table given
above. But it goes without saying that the child, as he
passes through stages HA and H!A, does not relinquish
the conversation of the earlier stages. Thus a child
who has reached stage IIlA will still indulge in occasional
monologues, etc.
Parallel with this evolution is that by which children
pass from stage I!B to IIlB, when instead of being in
agreement with one another, as in the preceding stages,
they are opposed either in opinion or in desire.
Stage HB is also present in two different types. First
of all we have the quarrel, a simple opposition of
divergent activities. Just as we have seen that in the
first type of stage HA each child, although acting in
isolation, can yet talk to the others about it and associate
himself mentally with their activities, so here also, instead
of associating himself with them, each child can criticize
and abuse the others, can assert his own superiority,
in a word, can quarrel. This type of quarrel is a clash
of assertions, which are not only statements of fact, but
56 STUDIES IN CHILD LOGIC
are connected with desires, with subjective evaluations,
with commands, and with threats. It may give rise to
argument. Thus, after having said: " give me that
No Yes No Yes, etc." the child may resort to
statement of fact. " I need that No Yes, etc." The
first dialogue belongs to quarrelling, the second to
argument. The reverse process is also possible ; argu-
ments can give rise to quarrels.
The second type of stage HB is therefore primitive
argument, i.e., argument without justification or proofs
of the assertions made. Only in the third stage, H!B,
do we come to argument proper, with motivation of
what is said.
Again, it is obvious that in the B series, once the child
has reached the stages HB and IIlB, he does not there-
fore cease to indulge in monologue or in arguments of a
primitive nature. But a child old enough to quarrel is
not necessarily capable of genuine argument.
There is between the stages II and III of the A series
and the corresponding stages of the B series no a priori
temporal connexion. But the evidence shows that
genuine argument and abstract collaboration appear at
the same age. Similarly, quarrelling and associating the
hearer with one's action are contemporaneous. They
are also contemporaneous with primitive argument and
collaboration in action. This points to a certain
parallelism.
Having established this schema, let us now examine
each stage in turn.
3 STAGE I : COLLECTIVE MONOLOGUE. The last
chapter has sufficiently familiarized us with what is
meant by "collective monologue" to enable us to deal
very briefly with it here. This stage does not belong to
conversation proper, so the criterion which we used for
classifying isolated remarks as collective monologue is
entirely valid for marking out a whole group of such
remarks. It may nevertheless be of interest to give
some fresh examples of this category, in the first place
CHILDREN'S CONVERSATIONS 57
for the sake of instancing a few cases under 5 and 6 years
old, and secondly because this preparatory stage of con-
versation is numerically by far the most important, at
any rate for children under 5.
We shall begin with examples of collective monologue
of one term only, although the remarks are all addressed
to someone.
Den (455) G 1 is talking volubly as she works. Bea
(5 ; 10) G comes into the work-room. Den : " You've got
a sweater on^ I haven't, Mummy said it wasn't cold'' Den
goes on working. Bea does not answer.
Den to Geo (6) in the building room : "I know how to,
you'll see how well I know. You don't know how. (No
answer. Den goes back to her place) I know how."
Den to Bea: "What do you want? (No answer.) /
shall want some little holes"
Ari (4 ; i) G to En (4 ; 1 1) : U W ha? s your name, my name
is Ari" No answer. Ari, without any transition, to
an adult : " She's going to let her doll drop."
These four-year-old monologues are thus entirely
similar in function to the monologues quoted in the
last chapter. They have, however, an element of
paradox owing to the use made of questions and purely
social forms of speech such as u You have put on,
you'll see, you want" which the child uses without
waiting for an answer, without even giving his com-
panion time to get in a word. Den, for instance, is
struck by Bea's jersey, but she immediately turns the
subject on to herself. " I haven't," etc. Why does
she speak to Bea? Not for the sake of telling her
anything, still less for the sake of getting an answer,
but simply as an excuse for talking. Similarly, Den's
question to Bea is purely rhetorical, it is a pseudo-
question which simply serves as an introduction to
the remark which immediately follows. The social
attitude is there only in form, not in substance. The
same thing happens between Ari and En.
Collective monologues of two terms or more are of
1 (4 j 5) O=Girl aged 4 years and 5 months.
58 STUDIES IN CHILD LOGIC
greater interest for our present purpose. Here are
some examples :
Pie (6 ; 5) : " Where could we make another tunnel?
Ah) here Eun? Eun (4; 11): Look at my pretty frock."
(The End.)
Cat (652): " Have you finished. Bur ? Bur (4 ; 1 1) :
Now it goes that way again" etc.
Talk of this kind clearly anticipates future conversa-
tion. The speaker expects an answer from his hearer.
If the two remarks together constitute only a collective
monologue, it is because the hearer is not listening.
Thus there is as yet no conversation, because the suc-
cessive terms are not adapted to one another. But
conversation is there in embryo, because the several
remarks are grouped into one bundle.
The age at which collective monologue marks a stage
of development-is between 3 and 4 to 5. The higher
forms of conversation do not on the average appear
before the age of 5, at any rate not between children
of the same age and of different families.
4. STAGE HA. FIRST TYPE: ASSOCIATION WITH
THE ACTION OF OTHERS. This type has already been
described as made up of those conversations in which
each speaker talks only about himself and from his own
point of view, but is heard and understood by each and
all. But there is still no collaboration in a common
activity. Here is an example. The children are busy
with their drawings, and each one tells the story which
his drawing illustrates. Yet at the same time they are
talking about the same subject and pay attention to
each other :
Lev (5 ; n) : "It begins with Goldy locks. Tm writing
the story of the three bears. The daddy bear is dead.
Only the daddy was too ill. Gen (5 ; 1 1) : / used to
live at Saleve. I lived in a little house and you had to
take the funicular railway to go and buy things. Geo
(6 ; o) : / can't do the bear. Li (6 ; 10) : That's not
Goldy locks. (Lev) : I havetft got curls"
CHILDREN'S CONVERSATIONS 59
This example is very clear. It is a conversation,
since they are all speaking about the same thing, the
class drawing, and yet each is talking for himself,
without any attempt at co-operation. Here is another
example :
Pie (6 ; 5) : "It was ripping yesterday (a flying demon-
stration). Jacq (5 ; 6) : There was a blue one, (an aero-
plane) there was lots of them, and then they all got into
a line. Pie : / went in a motor yesterday. And d*you
know what I saw when I was in the motor? A lot of
carts that were going past. Please teacher can I have the
india-rubber ? Jacq : / want to draw that (the aero-
planes). It will be very pretty"
The subject of conversation is one and the same,
and the dialogue has four terms. Just at first it might
seem as though some common memory were being
evoked, as in the cases of co-operation which the next
grade will show us, but we shall see from what follows
that each child is still speaking from his own point of
view. Pie talks about his motor, Jacq plans to draw
the aeroplane. They understand each other well enough,
but they do not co-operate.
Here are two more typical examples which show
very clearly that this association with the activity of
each is intermediate between collective monologue and
collaboration.
Mad (7) : " On Sunday I went to see my Granny who
lives in the i Chemin de P Escalade. ^ Geor (752): Do you
know Pierre C. ? No. I know him, he's my friend "
Rom (5 ; 9) : " D*you know what I shall have for
Christmas? Lev (5;n) and To (4; 9): No. Arm:
A bicycle with three wheels \ Lev : A tricycle. I've
got one."
The reader will notice how Geor's thoughts are
diverted by " Chemin de 1'Escalade," etc. This looks
like collective monologue. But the hearer has listened
and understood, and this stage does therefore mark
the beginnings of conversation proper, the beginnings
6o STUDIES IN CHILD LOGIC
of socialized language. But is this type of conversa-
tion only one among many others, or does it mark a
genuine stage of development ? We have seen that it
does both these things.
Given the existence of such a stage of development, it
goes without saying that no hard and fast rules can be
laid down as to its precise limits. An enormous amount
of material would be needed for any decisive statistics.
The fact remains, however, that in the material at our
disposal there are no examples of this type under 5 or
even 5 ; 6, whereas there is a large number of collective
monologues from 3 to 4 upwards.
Abstract collaboration hardly appears till about 7.
The type of conversation under discussion therefore re-
presents a definite stage of development in relation both
to collective monologue and to abstract collaboration.
But in relation to active collaboration the present
type cannot be said to occupy a position of before or
after. Collaboration in play appears from the age of
4 and 4 ; 6. Collaboration may therefore sometimes be
prior to " association with action," but the reverse is
often the case, many children collaborating in work
only after they have passed the present stage. In a
word, this type of conversation and that which follows
it are contemporaneous. They are the two possible
modes of the same stage of development.
Again, it need hardly be pointed out that if at Stage
III the child learns the use of a new type of conversa-
tion, abstract collaboration, he does not for that reason
discard the habits which he acquired at Stage II. These
different types coexist even in the adult, with the
exception of collective monologue, which is a strictly
childish form of conversation.
5 STAGE HA. SECOND TYPE: COLLABORATION
IN ACTION OR IN NON-ABSTRACT THOUGHT. In con-
versation of this type, the subject of the successive
remarks, instead of being the activities of the respective
speakers, is an activity in which they all share. The
CHILDREN'S CONVERSATIONS 61
speakers collaborate, and talk about what they are
doing. Instead of diffusion in relation to one and
the same subject as in the preceding type, there is
convergence.
Here is a typical example :
Bea (5 ; 10) G wants to draw a flag. Lev (5 ; u) :
" Do you know the one my daddy has? It isrit yours, tfs
mine. Ifs red and blue. . . . Ifs red, black and white,
thafs it yes, first red, white and first black Pve got the
right colour ; I shall take a square. No you must take two
little long things Aud now a square (shows it to Lev)
You must let me see if ifs right when you've finished"
(which she did).
This is a very good example of collaboration in
drawing. Lev advises Bea first as to colour then as
to shape, and finally checks the result. Lev, it should
be noted, knew the flag, and Bea did not. Hence the
dialogue. Now, curiously enough, all the examples
of collaboration in action under the age of 5 ; 6 or 6
are of this type, where a better informed or older child
explains an action to a younger or less informed com-
panion. It goes without saying that in reckoning the
age at which this type of conversation first appears, no
account need be taken of the age of the younger child,
so long as his part in the dialogue is not an active one.
Here are two examples. In the first the elder child
only is active.
(S J 6) to A (3 ; 9) who is drawing on the black-
board : " You want to draw something? Something But
not so long as that. You must do them like this, and then
like this, and then like this, and then some little windows,
but not so long as that" (The dialogue consists partly
of gestures).
Rog (5 ; 6) asks Ez (6 ; 4) to explain a point in an
educational game : " Was there one of these ones with the
yellow ones ? Jac (752): You musrft show him. Ez :
There are yellow ones. He's doing it all wrong. That
one's much easier. You can finish it now. Go along and
finish it"
62 STUDIES IN CHILD LOGIC
Collaboration in these cases is help given by an older
to a younger child With the very young children, i.e.,
those under 5|, collaboration between equals is first and
foremost collaboration in play. Here are two examples.
Lev (5 ;ii): "Den, Pll be the daddy and you 1 re the
mummy and Ari will be the nurse Ari (4 ; i) : Yes, and
the nurse' II take good care of the little children. Den (4 ; 5)
G : You re the daddy, Lev, you'll go out hunting, you'll go
to Germany "
Lev (5 ; 1 1) : "And then we'll play at balloons Arn
(5 ; 9) : How, balloons ? You see, we could pretend we* re
in the sky. Who'll be the sand? Ari, you'll be the sand
No, not the sand You II be the balloon, me the basket,
who* II be the sand of the balloon ? "
These conversations obviously presuppose collabora-
tion, if not in action, at least in some common game
or plan. As such, they no longer belong to the type
of " association with the action of others."
Here, finally, is a case of collaboration in evoking
a common memory. The example is unfortunately
one of two terms only, because the conversation was
interrupted by an adult.
Arn (5 ; 9) : " Ifs awfully funny at the circus when the
wheels (of the tricycle) have come off. Lev (5 ; u) : Do
you remember when the gymnastic man but who couldrft
do gymnastics, fell down. . . ."
Here, the collaboration is of thought only. In such
cases there are two boundary problems to be solved.
In the first place, there is between such dialogue as
this and that represented by the preceding type (associa-
tion with action) every degree of intermediate variety.
But in the latter type each child talks about himself
or about his personal recollections, here, on the con-
trary, the recollection is shared. This distinction is
often of great practical value. When it cannot be
applied, the two types of Stage HA can be grouped
as one. On the other hand, it is always desirable to
CHILDREN'S CONVERSATIONS 63
distinguish this collaboration in the evocation of a
common memory from abstract collaboration. For the
latter assumes in this case of a common memory that
the speakers not only evoke it together, but that they
discuss it, that they question or justify its foundation
in fact, or that they explain the why and wherefore of
events, etc. None of these characteristics is present in
the last conversation we have reported. Lev and Arm
are only trying to re-awaken in themselves one and
the same pleasant experience, without any attempt to
judge or explain the events.
In conclusion, collaboration in action or in non-
abstract thought constitutes a type contemporaneous
with the type preceding it, and these two together mark
a stage of development which extends on the average
between the ages of 5 and 7.
6. STAGE I II A. COLLABORATION IN ABSTRACT
THOUGHT. Conversations at this stage are the only
ones in which there is any real interchange of thought.
For even when they act in common, or evoke common
memories, as in the conversations of the preceding type,
children obviously have many more things in mind
than they ever say. We shall see in Chapter V that
alongside of the practical categories of thought and
of the interest he takes in his own activities, the child
shows signs long before the age of 7 of being interested
in the explanation of actions and phenomena. The
numerous i whys ' of children from 3 to 7 bear witness
to this. The conversations which we shall class under
the present type are those which bear i on the ex-
planations of things and the motives of actions, 2 on
the reality of events (" Is it true that . . . ? " " Why?
. . . ," etc.)
Now it is a curious thing, and one that confirms the
results of our investigations on Pie and Lev, that from
the twenty children under observation we obtained only
one conversation of this type, and not a very clear
one at that. This shows once more how ego-centric are
64 STUDIES IN CHILD LOGIC
the intellectual processes of the child. It also enables
us to place the beginnings of the socialization of
thought somewhere between 7 and 8. It is at about
this age, in our opinion, that conversations of this type
first make their appearance. (Probably in both girls
and boys.)
The only example obtained from our subjects is, as it
happens, a dialogue between a girl of 7 and a boy of 6.
These two children are searching together for the ex-
planation, not of a mechanism, but of an action the
absence of their teacher. The corresponding question
would therefore come under the "whys of intention and
action 1 ' (see Chapter VI) and would run : " Why has
Mile L. not yet arrived ? "
Mad (7 ; 6) : " Ok, the slow-coach ! Lev (6 ; o) : She
doesn't know it's late Well, I know what she is. And I
know where she is. She's ilL She isn't ill since she isn't
This mutual explanation is not, it must be admitted,
of a very high intellectual order ! The use of the word
' since * in the argument should, however, be noted,
though the proposition in which it occurs is of doubtful
intelligibility.
For the sake of comparison let us give an example of
this type of conversation overheard away from the
Maison des Petits between two sisters of 7 and 8. This
example contains, not only an explanation sought in
common, as in the case of Mad and Lev, but also the
mutual reconstruction of a memory. The memory is
discussed and judged, not merely recalled as in the
last stage.
Cor (7) : " Once I wrote to the rabbit that Fd like to see
him. He didn't come. Viv (8) : Daddy found the letter in
the garden. I expect he (the rabbit) had come along with
the letter, and he didn't find Cor and he went away again.
/ went into the garden he wasn't there and then I forgot
about it. He saw Cor wasn't there. He thought ' she's
forgotten 9 and then after that he went away"
CHILDREN'S CONVERSATIONS 65
Cor and Viv both believe In fairies, at least they do
so to one another in their conversations, prolonging in
this way an illusion which has lasted several months.
They have built a house for fairies in which they place
little notes in the evening. The above conversation
bears upon the outcome of one of these missives. They
are explaining a failure to each other, and criticizing the
course of events. This is enough to place this dialogue
at the stage under discussion. It is extremely curious
that we should have found no conversations of such a
simple type among the children from 3 to 7 who work
and play together at the Maison des Petits. Conversa-
tions of this type must surely occur between brothers
and sisters under 7. But this circumstance in itself
constitutes a special problem. As soon as there are
elder and younger children, their conversation expresses,
not so much an exchange of thoughts, as a special kind
of relation, for the elder child is always regarded as
omniscient, and the younger one treats this knowledge
with some of the respect which he feels for the wisdom
of his parents.
It need hardly be added that between conversations of
Stage IIlA and those of Stage IIlB (genuine argument),
there is every kind of intermediate variety.
7. STAGE HB. FIRST TYPE : QUARRELLING. We
now come to a set of developments parallel with
the preceding ones. They consist of conversations
which certainly express an interchange of thoughts
between individuals, but an interchange occasioned not
by progressive collaboration, but by divergence of
opinions and actions. It may seem idle to distinguish
two sets of developments on the basis of this difference
only, but if this classification ever comes to be applied
statistically on a large scale, the distinction may be
seen to have its importance, particularly from the
genetic point of view. It may well be through quarrel-
ling that children first come to feel the need for making
themselves understood. In any case, the study of
66 STUDIES IN CHILD LOGIC
arguing is, as the investigations of Rignano and P.
Janet have shown, of great importance for the psy-
chology of reflexion. It is therefore desirable to make
a special study of the growth of arguing in children.
We shall attempt it here, but only very schematically.
We may distinguish two stages of childish argu-
ment. The first consists in a simple clash of contrary
tendencies and opinions. This gives us two more or
less contemporaneous types the primitive quarrel and
the primitive argument. The second consists in argu-
ments in which the speakers give the motives of their
respective points of view. This stage corresponds to
collaboration in abstract thought (1 1 1 A). The first
stage corresponds to HA. Between the corresponding
stages of series A and series B there is naturally a
whole chain of intermediate links.
Here are a few examples of quarrels :
Ez (6 ; 5): "Ah I I've never had that. Pie (6 ; 5) :
You've already had it to play with Thafs for A. Well
I've never had it to play with"
Lev (6 ; o) : "I bagged this seat. . . . / shall sit
here all the same. Bea (5 ; o) G : He came here first
No, I came here first '."
Ez (653): " You wait and see what a slap r II give
you Rog (5 ; 6) : Yes you just wait Lev (5 ; 10)
(frightened): No."
Lil (6 ; 10) G: " She's nice.Ez (6 ; 5) : No. Mo
(7 ; 2) : Yes, yes, yes. (They all rise and face each
other). Ez to Mo: You'll see what a slap P II give you
at break"
Quarrelling differs from primitive arguing simply in
that it is accompanied by actions or promises of actions
(gestures or threats). It is the functional equivalent of
argument. In primitive argument the opposition is
between assertions ; here, it is between actions. Ez and
Pie quarrel over a toy. Lev and a silent opponent
defended by Bea quarrel over a seat, etc. Speech in
these quarrels simply accompanies gesture, and is not
always understood, as is shown by Lev (in the second
CHILDREN'S CONVERSATIONS 67
quarrel) who repeats what Bea has just said in the fond
belief that he is contradicting it.
In deciding upon the age of quarrelling a distinction
should be drawn between quarrels with words and those
without. The first alone are of interest to us. Now it
is a remarkable fact that children between the ages of
4 and 5, although they are extremely quarrelsome,
generally conduct their disputes without talking, and
we have found no cases of spoken quarrels in three-
termed dialogue prior to the age of s|. There even
seems to be a certain progression according to age from
the wordless quarrel through the quarrel with words,
but accompanied by actions, to the merely spoken
quarrel devoid of actions, like that between Ez and Lev
who did not slap each other in the end, but having
had their say, were content to let the matter rest at
that. But such a sequence is of course by no means
universal. It simply indicates the progress of con-
versation in the particular little society which we were
studying.
Quarrelling, in a word, is contemporaneous with the
two types of stages HA. Quarrelling and primitive
argument merge into one another through a whole
series of intermediate varieties of which we give two
examples, both being classed as quarrels :
Bea (5 ; 10) G : " You said I was a ox ! Jac (752):
NO) I said . . . silent Oh, I thought you said I was a ox"
Lev (5 ; n) : u Gen, shew me your funicular railway.
But that's not a funicular railway I Gen (6 ; o) to Pie
(6 ; 5) : He says ifs not a funicular railway. (Looking at
Pie's drawing) : Thais not pretty Pie : Gen says that
mines not pretty. He shan't see it any more. Lev to Pie :
Its very pretty "
In this last example Pie and Lev take sides against
Gen. This is something more than simply arguing.
The child is not trying to argue, but to tease or to
defend himself. The first example is more subtle. Jac
gives in at once to avoid an argument. Bea's tone at
68 STUDIES IN CHILD LOGIC
the outset, however, inclines us to class this dialogue as
a quarrel.
8. STAGE Us. SECOND TYPE : PRIMITIVE ARGUMENT.
Argument begins from the moment when the speakers
confine themselves to stating their opinions instead of
teasing, criticizing, or threatening. The distinction is
often a subtle one. We have just instanced inter-
mediate examples which we placed among quarrels.
Here is one which must be classed as an argument,
because the speakers' tone is one of statement, although
the subject-matter is one of blows.
Ez (6 54): " You wait; at the Escalade, Pll be the
strongest. Lev (5 ; n): At the Escalade, not at school.
Ez ; Everywhere Fll be the strongest."
The argument is very primitive and not quite genuine,
because there is no trace of a desire for logical justifica-
tion in the assertions of Ez and Lev. The criterion of
primitive argument is not an easy one to apply. We
must therefore try to fix the point where the attempt
is first made to justify and demonstrate the assertions
made in the course of argument. We propose the follow-
ing rule. There is demonstration (therefore genuine
argument) when the child connects his statement with
the reason which he gives for its validity by means of a
conjunction (e.g. since, because, then, etc.), and thus
makes his demonstration explicit. So long as the
justification is only implicit, and the child expresses
himself in a succession of disconnected statements, the
argument is still only primitive. This rule is purely
conventional, but it is useful because any subjective
test as to whether demonstration is present or not tends
to be still more arbitrary.
Viv (7 ; 3) G. : "My daddy is a tiger. Geo (7 ; 2) :
No, he can't be. I've seen him. My daddy is a godfather
and my mummy is a godmother."
Geo implicitly justifies his assertion : " He can't be"
with the proof; "I've seen him" But there is no
CHILDREN'S CONVERSATIONS 69
explicit connexion between the sentences. In order to
find in the second a justification of the first, it would be
necessary to follow a line of argument to which Geo
nowhere gives expression. Geo, like Viv, confines
himself to mere statement.
Similarly in the following example.
Lev (5 ; n) : " Thafs Ai\Mte (5 ; 5) G: Ifs Mie
(AY's sister). Lev : No, it's At. Ez (6 ; 4) : Its Mie,
look" (He lifts Mie's cloak and shows her dress.)
The first three terms of the argument are definitely
primitive, there is no demonstration. The fourth con-
tains an element of justification, but by means of gesture,
without any explicit reasoning. It serves the purpose
in this particular case, but the instance is none the less
primitive in character.
Justification of a statement may consist in an appeal
to one's own authority or to that of others or of one's
elders. But unless it is given in the form of reasoning,
it does not constitute an argument. Here are two
examples :
Lev (5 ; 10) : "ft isrit naughty to bury a little bird
Ari (4 ; i) : Yes, it is naughty No, no, no. Lev to Je :
It isn't naughty is it ? - Je (6 ; o) : / don't know, I don't
think so"
Ai" (3 ; 9) : " Pve got four little balls Lev (5 ; 10):
But there aren't four. You dorit know how to count. You
dorit know how much four is. Let me see . . . etc."
In all these examples, it is easy enough to recognize
primitive argument. The remarks are made as simple
statements and not as explicit reasoning. If we compare
these instances with the following example (of which
the last remark comes very near to being genuine
argument) and with the one example of genuine argu-
ment which we obtained, the difference will be seen at
once :
Lev (5 ; n): We can only give some [fish] to those
who speak English. Ez (654): We can't give her [Bea]
70 STUDIES IN CHILD LOGIC
any. I know English. Bea (5 ; 10) : No, I know English.
Lev : Then Fll give you the fish. Ez : Me too. Mad
(7 ; 6) : She doesn't know it. Lev : Yes, she does. Mad :
Ifs because she wants some that she says that"
This conversation only becomes an argument towards
the end. Mad and Lev begin by simply opposing their
respective points of view to one another. But where a
great advance is made on the previous examples is
when Mad, in order to contradict Lev, gives an ex-
planation of Bea's conduct. She therefore interprets
the adverse point of view, and justifies her own by an
explanation. Even if the other speakers are still arguing
in a primitive manner, Mad, in her last remark, has
reached the stage of genuine argument.
Primitive argument is thus, on the mental plane, the
equivalent of quarrelling on the plane of action a
simple clash of contrary opinions and desires. There
is therefore nothing surprising in the fact that these
two types of conversation should be roughly contem-
poraneous. It is true that the quarrel without words,
or at any rate without a three-termed dialogue is
prior in appearance to argument, but according to our
evidence the spoken quarrel, like the primitive argu-
ment, generally begins at about 5 or 5. Genuine
argument as represented by Stage III A does not
appear till about 7 or 7j. Before, therefore, it can be
reckoned as one among other types of argument,
primitive argument must be recognized as constituting
a definite stage in the evolution of childish conversation,
a stage, which though it has no very precise boundaries,
yet corresponds to the objective results obtained from
our statistics.
9. STAGE Ills: GENUINE ARGUMENT. The statis-
tical data are as follows. In the whole of our material
we have found only one case of genuine argument in
dialogue of more than three terms among the children
under 7. This tallies exactly with the fact that colla-
boration in abstract thought does not appear, on an
CHILDREN'S CONVERSATIONS 71
average, before the age of 7 or 7|-. Indeed, these two
different aspects, A and B, of stage III can be accounted
for by one and the same circumstance. Up to a certain
age the child keeps to himself, without socializing it,
everything that is connected in his mind with causal
explanation or logical justification. Now in order to
argue, demonstrations and logical relations etc. have
to be made explicit, all of which runs counter to the
ego-centrism of the child under 7.
Here is the only case of genuine argument which we
have obtained. The difference between it and the three
preceding examples will be seen at once. Out of the
five terms of the dialogue, three contain the word
{ because ', which in one case at least points to a logical
justification.
Pie (6 ; 5) : " Now, you shan't have it (the pencil)
because you asked for it. Hei (6 ; o) : Yes I will, because
ifs mine. Pie : 'Course it isn't yours. It belongs to every-
body, to all the children. Lev (6 ; o) : Yes, it belongs to
Mile. L. and all the children, to A'i and to My too. Pie:
It belongs to Mile. L. because she bought it, and it belongs to
all the children as well."
It is surprising that a type of conversation appar-
ently so simple should have occurred only once in the
material we have collected. The truth is that the use of
the word ' because 5 is a very delicate matter. The
logical ' because * connects, not two phenomena of
which one is cause and the other effect, but two ideas
of which one is reason and the other consequence. Now
this connexion is one which, as we shall see in Chapter
I of the second volume (examination of the sense of
conjunctions, i because' etc.) it is still very difficult for
the 7-year-old mind to make. There is therefore nothing
surprising in the scarcity of genuine arguments with
demonstrations in defence of the use of these conjunc-
tions among children under 7.
After the age of 7 or 8, however, the logical because '
and * since ' make frequent appearances in the children's
72 STUDIES IN CHILD LOGIC
conversation, thus enabling them to take part both in
genuine argument and in collaboration in abstract
thought.
Here are two examples taken at random from those
conversations of children between 7 and 8, which are
published from time to time 1 by Miles. Audemars and
Lafendel. These examples have been chosen at random
from two pages in the collection :
" Ray : (She won't be an orphan). "But she will go
to a boarding school, since she has still got her daddy."
Ray : (The chain of men is the most important of all),
' because they have worked a lot and invented a great many
things."
A logical ' since ' and a logical { because ' occur in this
example. Such verbal forms abound in the conversation
of these children, whereas they are avoided or only used
on very exceptional occasions by children under 7. It
should be noted that with Ray the ' because ' is definitely
logical, i.e., connecting two ideas or definitions, and not
psychological, z>., connecting an action with its psycho-
logical explanation.
The reasons, therefore, why genuine argument, like
collaboration in abstract thought, appears only after the
age of 7 or 7^ in the development of the child are of a very
fundamental order. Does the absence of verbal forms
expressing logical relations prevent genuine argument
from manifesting itself, or does the absence of the desire
to argue and collaborate explain the late appearance of
these verbal forms? If we admit that thought in the
child depends upon his interests and activities rather
than vice versa, then the absence of the desire to argue
and collaborate is obviously the initial factor. This is
why we have begun our study of child logic with a study
of the forms of conversation and of the functions of
language. But there is, as a matter of fact, perpetual
interaction between these two factors of evolution.
1 See 1} Educateur> Lausanne, vol. 58, pp. 312-313.
CHILDREN'S CONVERSATIONS 73
10. CONCLUSION. What conclusion are we to draw
from these facts ? Can we, in the first place, establish any
numerical results from the material on which we have
worked ? This material consists of two books containing
500 remarks each. Among these there are several
dialogues between children and adults, which have been
omitted from the following calculations. This leaves us
with about 400 remarks in each book, representing the
talk of children between the ages of 3^ and 7. There
are 31 conversations in one book and 32 in the other.
These two groups may be distributed as follows.
I II TOTAL
Stage II A ist type ... 4 6 io\
IlB ist type ... 8 3 iij
UA 2nd type 9 16
IlB 2nd type 9 6
IIlA .... i o
IIIB .... i o
Stage UA 2nd type 9 16 2 Aio
Stage IIlA .... i o i\
i/ 2
The collective monologue is naturally excluded from
statistics dealing with general conversations, and it can
be judged of only by the number of remarks which are
assigned to it. We have seen that the coefficient of
ego-centrism (collective monologue, monologue and
repetition) is 0*45 for the sum of the talk under inves-
tigation after subtraction of answers.
This result shows very clearly that genuine argument
and collaboration in abstract thought constitute a stage
of development which only intervenes after the age of 7.
This is a very useful confirmation of the conclusions
reached in the last chapter. The statistics of the remarks
made by Lev and Pie seemed to justify the conclusion
that intellectual processes (causal explanation, logical
justification) in children of less than 7 or 7^ remain ego-
centric in character. It will be remembered that in all
the talk classed as * information J we found very few cases
of causal explanation or logical justification. Mental
activity is either silent, or accompanied by monologues.
Our present results, showing as they do the rareness of
74 STUDIES IN CHILD LOGIC
genuine argument and collaboration in abstract ideas
before the age of 7, go to prove these same conclusions
by a different method.
The fact that the stage of collaboration and genuine
argument does not intervene till the age of 7 or 7| is
of the greatest importance. For it is between the ages
of 7 and 8 that we can date the appearance of a logical
stage in which the phenomenon of reflexion becomes
general ; if we agree with P. Janet in calling reflexion
the tendency to unify one's beliefs and opinions, to sys-
tematize them with the object of avoiding contradiction.
Up till the age of 7 or 8 children make no effort to
stick to one opinion on any given subject. They do
not indeed believe what is self-contradictory, but they
adopt successively opinions, which if they were com-
pared would contradict one another. They are insensible
to contradiction in this sense, that in passing from
one point of view to another they always forget the
point of view which they had first adopted. Thus in
the course of interrogation, the same children aged
from S to 7, will answer at one time that ants, flowers
and the sun are living beings, and at another time that
they are not. Others will affirm on one occasion that
rivers have been dug out by the hand of man, and
on another that they were made only by water. The
two contrary opinions are juxtaposed in the children's
minds. At one moment they adopt the one, then
forgetful of the past, and in all sincerity, they come
back to the other. This is a well-known fact in the
examination of children up to the age of 7 or 8, even
when the subjects are not deliberately inventing.
This absence of system and coherence will be examined
elsewhere (Vol. II). It will suffice in the meantime to
note that its disappearance coincides with the advent
of genuine argument. This coincidence is not. fortuitous.
If, as we said a little while ago, a correlation be admitted
to exist between a child's activity and his thought,
then it is obviously the habit of arguing which will
CHILDREN'S CONVERSATIONS 75
cause the need for Inner unity and for the systematiza-
tion of opinions to make itself felt. This it is, to
which Janet and Rignano have drawn attention in
connexion with the psychology of arguing in general.
They have shown that all reflexion is the outcome of
an internal debate in which a conclusion is reached,
just as though the individual reproduced towards
himself an attitude which he had previously adopted
towards others.^ Our research confirms this view.
It should be stated in conclusion that these studies
need to be completed by a general investigation of the
conversations of children as they are carried on apart
from work, as for instance at play in public gardens,
etc. Enough has been said, however, for the schema
which we have elaborated to serve in the studies ahead
of us. The following chapter will complete our data
by showing that if before the age of 7 or 8 children
have no conversation bearing upon logical or causal
relations, the reason is that at that age they hardly
understand one another when they approach these
questions.
CHAPTER III
UNDERSTANDING AND VERBAL EX-
PLANATION BETWEEN CHILDREN
OF THE SAME AGE BETWEEN THE
YEARS OF SIX AND SEVEN. 1
IN the preceding chapters we have tried to determine
to what extent children speak to each other and think
socially. An essential problem has been left on one
side : when children talk together, do they understand
one another ? This is the problem which we are now
to discuss.
This question is not nearly so easy to answer as the
preceding ones, and for a very simple reason. It is
quite possible to determine immediately whether children
are talking or even listening to one another, whereas it
is impossible by direct observation to be sure whether
they are understanding each other. The child has a
hundred and one ways of pretending to understand,
and often complicates things still further by pretending
not to understand, by inventing answers, for instance,
to questions which he has understood perfectly well.
These conditions therefore oblige us to proceed with
the utmost prudence ; the different questions involved
must be arranged in proper order, and only that one
approached which concerns verbal understanding.
To show the soundness of the experiments we have
instigated, let us start from observation of the child
such as has been supplied by the preceding chapters.
We have seen that in the highest types of conversation
1 With the collaboration of Mme Valentine Jean Piaget.
76
UNDERSTANDING BETWEEN CHILDREN 77
between children, i.e., collaboration and argument, two
different cases are to be distinguished, which we have
called Stage II and Stage III. The first case is con-
nected with action (collaboration in action, or primitive
argument still bound up with action and devoid of
explicit reasoning) ; the second case makes use of
abstraction. Let us call them, for the sake of brevity,
the acted case and the verbal case. In the verbal case
children collaborate or argue about a story to be re-
constructed, a memory to be appreciated, or an ex-
planation to be given (explanation of some phenomenon
or other or of the words of an adult). Now discussions
such as these take place on the verbal plane, without
actions, without the aid of any material object with
which the speakers might have been playing or working,
without even the present spectacle of the phenomena
or of the events about which they are talking. In the
acted case, on the other hand, the collaboration or
argument is accompanied by gestures, by demonstra-
tions with the finger and not with words ; it matters
little, therefore, whether the talk is intelligible or not,
since the talkers have the object under their eyes.
Hence the quaint character of much childish talk.
(That does that, and then that goes there, and It goes
like that," etc.). Were it not entirely outside the scope
of this study, the connexion should also be established
between these * acted ' conversations and the language
by gesture and mime language in movement, one
might say which is, after all, the real social language
of the child.
Now in these two cases, 'acted' conversation and
* merely spoken ' conversation, children naturally under-
stand each other in a very different manner. (The
second case, moreover, characterizes a stage which
begins only at about seven years ; it does not have its
full effect, i.e., it does not lead children to understand
each other till about the age of 8.) In ' acted 3 con-
versation one gets the impression that the children
78 STUDIES IN CHILD LOGIC
understand each other well. Hence the success of the
educational method (provided there is an adequate
supply of educational games) which consists in letting
one child explain to another, say, a certain way of
doing sums, or a certain school regulation. Thus we
owe to Mile Descceudres the knowledge that in games
of spelling (lotto, etc.), of arithmetic, and in exercises
of manual skill (threading beads, etc.) even abnormal
children collaborate very profitably, and understand
each other better than master and pupil would do.
This rule holds good even between children of the
same age, from 5 or 6 upwards, although the under-
standing between elder and younger children is on
the average higher. But all this concerns only ' acted '
conversation. As to merely spoken conversation it
may be questioned whether children really understand
each other when they use it ; and this is the problem
which we shall now attempt to solve. Let us begin
by showing the importance of the subject.
An essential part of the intellectual life of the child
takes place apart from contact either with any material
that is really within his reach or with any concrete
images. To say nothing of the ordinary schools where
from the age of 7 the child no longer manipulates a
single object, and where his thought sinks deeper and
deeper into verbalism, cases of the following kind are
of daily occurrence. A child sees a bicycle in the
street, and mentally reconstructs its mechanism. (A
Geneva boy can give this explanation on the average
from the age of 7 or 8|.) The same thing happens in
the case of motors and trains. The child, from the age
of 6 or 7, has images connected with the words ' benzine,'
' electricity/ * steam,' etc. He has others connected
with the concepts life, thought, feeling, etc., and he
has ideas on the amount of life and feeling, so to speak,
which may be accorded to animals, plants, stars, etc. . . .
He hears people talk about countries, towns, animals,
and instruments which are completely unknown to him,
UNDERSTANDING BETWEEN CHILDREN 79
but about which he reasons nevertheless. Yet another
kind of preoccupation is that concerning the amount
of truth to be attributed to dreams, stories, the fantasies
of play, etc. All these types of mental activity can
take place only on the verbal plane, and in this sense
they will always differ from those bearing upon toys
and instruments, etc., which imply manual work or at
least manipulation.
Now, as the last two chapters have shown, this verbal
activity is not social ; each child carries it on by him-
self. Each child has his own world of hypotheses and
solutions which he has never communicated to anyone,
either because of his ego-centrism, or for lack of the
means of expression which comes to the same thing,
if (as we hope to show in this chapter) language is
moulded on habits of thought. We shall even go so
far as to say in a chapter in our second volume that
a child is actually not conscious of concepts and defini-
tions which he can nevertheless handle when thinking
for himself. What then will happen when the chances
of conversation lead children to exchange their ideas
on the verbal plane? Will they understand each other
or not? This is a cardinal question in the psychology
of child thought, and will supply us with a necessary
counter-proof. If we can prove that verbal thought is
incommunicable between children we shall justify our
hypotheses concerning childish ego-centrism, and at
the same time explain some of the most characteristic
phenomena of child logic, particularly that of verbal
syncretism (cf. Chapter IV.).
i. THE METHOD OF EXPERIMENT. In order to
solve this problem we have had to undertake an
experiment which consists in making one child tell
or explain something to another. This procedure will
doubtless be criticized as being removed from every-
day life, where the child speaks spontaneously, without
being made to, and especially without having been told
what to relate or explain to his listener. We can only
8o STUDIES IN CHILD LOGIC
reply that we found no other way of solving the problem.
This method certainly has its drawbacks ; still, once
allowance has been made for the risks it incurs, it must
be granted that in some of its aspects it recalls what
happens in ordinary life ; as when a child, immediately
after hearing a story or receiving an explanation, goes
off to tell the same story or give the same explanation
to a younger brother or to a friend. The great thing
is to turn the experiment into a game, to make it
interesting. But this condition is not a very difficult
one to fulfil if the children are taken during school
hours, and consequently when they are under the spell
of the unexpected, The matter can be introduced as
an amusement or a competition: "Are you good at
telling stories? Very well then, we'll send your little
friend out of the room, and while he is standing outside
the door, well tell you a story. You must listen very
carefully. When you have listened to it all, we'll make
your friend come back, and then you will tell him the
same story. We shall see which of you is best at
telling stories. You understand? You must listen
well, and then tell the same thing ..." etc. Repeat
the instructions as often as necessary, and stress the
need for a faithful rendering, etc.
Then one of the subjects is sent out of the room, and
the set piece is read to the other. The more compli-
cated passages are repeated, everything is done to
make the subject listen, but the text is not altered.
Then one or other of the following methods is adopted :
(They have been used alternately, the one serving as
a test of the other.) Either the child who has been
waiting outside in the passage, and whom we shall
call the reproducer, is sent for, and everything that the
other child (whom we shall call the explainer] says to
him is taken down in extenso ; or else the explainer is
asked to tell us a story in the first instance, and is
then sent to tell the same story to the reproducer out
in the lobby or in the garden, *.*., in our absence, and
UNDERSTANDING BETWEEN CHILDREN 81
with the injunction to take as much time as he likes.
In both cases the story, as told by the reproducer, is
taken down verbatim. Both these methods have their
drawbacks. In the first, the story told in our presence
loses in spontaneity;. In the second, we can no longer
check matters so closely, and it may well be that the
explainer, after having told us a story perfectly well, will
take less trouble over it when he is talking to the
reproducer. There is always a certain disadvantage
in making the explainer repeat the same story twice.
We therefore dispense with this initial test in the first
method, as it is preferable to do for children between
7 and 8. Since the reproducer's degree of understand-
ing is estimated in relation to that of the explainer,
and not with reference to the original text, the fact
that the explainer may occasionally blunder is of no
importance. If, for instance, the explainer has under-
stood 8 points out of 10 and the reproducer 4 points
out of 8, the coefficient of understanding will be 0*8
( = T %-) for the explainer and 0*5 ( = !) for the reproducer.
It will not be 0-4 ( = T V) for the latter, because no
account is taken of the two points omitted by the
explainer. With children from 5 to 6, on the other
hand, one is obliged to ask for a preliminary account
of the story by the explainer, who very often has been
thinking of everything rather than of paying attention.
The results obtained by these two methods have proved
of equal value. By using them simultaneously we have
therefore a means of testing our results, which will have
to be borne in mind in our subsequent investigations.
When the experiment is over, the two children
exchange parts ; the explainer is sent out of the room
and becomes the reproducer in this second test, a new
story is told to the former reproducer who now becomes
explainer, and everything is done as in the previous
case.
After this exchange of stories, an exchange of explana-
tions was organized, bearing upon mechanical objects.
F
82 STUDIES IN CHILD LOGIC
The explainer was shown a diagram of a tap or of a
syringe (the drawing of a bicycle has also been used
occasionally), and he was given in a fixed order the
explanation of the workings of the several parts. This
unusual choice of subject was not made at random, but
in consideration of the interests of boys from 6 to 8.
The latter are often too well-informed on these subjects
for the experiment to be conclusive.
The method adopted for the explanations is as follows.
The explainer, when he has had the diagram explained
to him, takes possession of this diagram and explains it
in his turn to the reproducer. The reproducer then
gives his explanation, with the drawing still before him.
We have carried out with these methods some hundred
experiments on 30 children from 7 to 8, taken in pairs
(i.e.) 15 couples with 4 experiments per couple, say 2
explanations and 2 stories), and on 20 children from
6 to 7 (10 couples with 4 experiments per couple).
Here are the stories which were used :
I. Epaminondas is a little nigger boy and he lives in a
country where it is very hot. His mother once said to
him : * ' Go and take this shortbread cake to your granny,
but don't break it/* Ep. put the shortbread under his
arm, and when he got to his grandmother's the short-
bread was in crumbs. His granny gave him a pat of
butter to take back to his mother. This time Ep. thought
to himself: " I shall be very careful." And he put the
pat of butter on his head. The sun was shining hard,
and when he got home the butter had all melted. " You
are a silly," said his mother, " you should have put the
butter in a leaf, then it would have arrived whole."
IL Once upon a time, there was a lady who was
called Niobe, and who had 12 sons and 12 daughters.
She met a fairy who had only one son and no daughter.
Then the lady laughed at the fairy because the fairy only
had one boy. Then the fairy was very angry and fastened
the lady to a rock. The lady cried for ten years. In
the end she turned into a rock, and her tears made a
stream which still runs to-day.
UNDERSTANDING BETWEEN CHILDREN 83
III. Once upon a time, there was a castle, and in it
were a king and queen who had three sons and one
daughter. Near the castle was a wicked fairy who did
not like children. She took the king and queen's
children to the seashore, and changed them into four
beautiful white swans. As their children had not come
home, the king and queen went to look for them every-
where, and they came right down to the sea-shore.
There they saw four beautiful swans, who told them
that they were their children. The swans stayed on the
sea for a very long time, and then they went away to a
very cold country. After many years they came back
to where their castle was. There was no castle there
any longer, and their parents were dead. The swans
went into a church and they were changed into three
little old men and one little old woman.
In these three stories the events are related to one
another in the greatest variety of ways, ranging from the
most simple and natural to the most mythological. We
we now give the two mechanical explanations of which
have made use most frequently. Between the causal
relations which they imply, and the relations of events con-
tained in the preceding stories, we shall have sufficient
material for studying the way in which children under-
stand and express the whole scale of possible relations.
(i) Look, these two pictures (I and II) are drawings
of a tap.
84 STUDIES IN CHILD LOGIC
(2) This here (a) is the handle of the tap.
(3) To turn it on, look, you have to do this with
your fingers (move the finger on diagram I and
show the result on diagram II). Then it is like this
(diagram II).
(4) You see (diagram I), when the handle is turned
on like this (point to a and make horizontal movement),
then the canal (point to b, call it also the little hole,
door, or passage) is open.
(5) Then the water runs out (point to b in diagram I).
(6) It runs out because the canal is open.
(7) Look, here (diagram II), when the handle is turned
off (point to a and make a vertical movement), then the
canal (point to b ; can also be called the hole or door or
passage) is also shut.
(8) The water can't get through, you see? (point to c).
It is stopped.
(9) It can't run out, because the canal (point to b)
is closed.
The reader should note that each one of these points
has to be made for the child. It often happens, for
instance, that the subject understands, say, point (5) (the
water runs out), and thinks that the water runs out
simply because the handle of the tap has been turned,
ignoring the fact that the handle has turned round the
canal, and that this circumstance alone enables the
water to flow.
Here is a second test which we used.
(1) You see this (diagram III and IV). Do you
know what it is? It is a syringe.
(2) You know what a syringe is, don't you? It's
what you squirt water with.
(3) Do you know how it works? Look, you dip it
into the water ; that is the water, there (a).
(4) Look, there is the piston (b). When you want
the water to go up, you pull the piston.
(5) Then the water goes up, you see? (Point to the
water in c on diagram IV).
(6) It has gone up through the hole (d).
(7) It has gone up because the piston has been pulled.
That has made more room (point to ), so the water fills
the room that has been made.
UNDERSTANDING BETWEEN CHILDREN 85
(8) To squirt the water you push the piston (&).
(9) Then the water goes out (point to d).
So far, then, the method is quite simple. You read
one of the stories or explanations to the explainer, but
you must not appear to be reading, and you must talk
in the most natural manner possible. The explainer
then tells the story to the reproducer, who finally serves
it up to you again.
But this is not all. Once the reproducer's story has
been obtained and taken down in its entirety, the
explainer is taken aside for a few moments, and the
reproducer is asked a certain number of questions on
the points that have been omitted, so as to ascertain
whether he has really failed to understand them. He
may either have forgotten them or he may not know
how to express them. In order to judge of the child's
degree of understanding these factors must at all costs
be eliminated, so as to clear the ground for a more
searching investigation. If in the story of Niobe, for
example, the end is forgotten, the child is asked whether
there is nothing about a stream. Thus by means of
questions, vague at first and then more and more
precise, and with the help of that division into points
which we have just given for explanations and which
86 STUDIES IN CHILD LOGIC
we shall give in the next paragraph for stories, the
reproducer's degree of understanding can be properly
put to the test. When this has been done, the ex-
plainer is questioned in the saftie manner, to see
whether he has really understood the points which
appear doubtful.
2. PARCELLING OUT THE MATERIAL. Such experi-
ments as these will be seen to resemble on many points
the experiments made by Claparede and Borst, by
Stern, etc., on evidence. For in the manner in which
the explainer, and even more so the reproducer, distort
the story they have heard, we can see various factors
at work, such as memory of facts, logical memory, etc,,
all of which we shall call by the same name the factors
of evidence. Now it is important to eliminate these
factors in order to study understanding or the lack of
it independently of distortions of fact due to other
causes. How then are we to avoid the factors of evi-
dence, which are of no interest to us here? By the
device of parcelling out the material.
We have divided each of our set pieces into a certain
number of points, as is done in the sifting of evidence,
so as to see which of these points have been reproduced
or omitted by the subjects, and instead of choosing a
large number all bearing on questions of detail, we
have restricted ourselves to a small number of rubrics
connected solely with the understanding of the story.
In estimating the correctness of each point, moreover,
we have taken no notice in parcelling out the material
of the factors that were not essential to the understanding
of the story. Thus in the tale of Niobe the name of
Niobe plays no part whatsoever ; it is sufficient if
mention is made of "a lady" or even "a fairy."
Similarly, " 12 boys and 12 girls" can be changed
into " many children " or "3 children," etc., provided
a difference is made between the number of children
belonging to "the lady" and that of those belonging
to the fairy.
UNDERSTANDING BETWEEN CHILDREN 87
Here, in detail, are the points taken into con-
sideration :
I Niobe. (i) Once there was a lady (or a fairy, etc.).
(2) She had children (provided they outnumber these
of the other fairy). (3) She met a fairy (or a girl, etc.).
(4) This fairy had few children (or none at all, provided
their number is inferior to the first lot). (5) The lady
laughed at the fairy. (6) Because the fairy had so
few children. (7) The fairy was angry. (8) The fairy
fastened the lady (to a rock, a tree, to the shore, etc.).
(9) The lady cried. (10) She turned into a rock,
(n) Her tears made a stream. (12) Which flows to
this day.
It is obvious that each of these points except point (7),
which can easily be taken for granted, and points (9)
and (12) which are supplementary to the body of the
story, are necessary to the comprehension of the story.
It will be seen, moreover, that we are very generous in
our estimates, since any alteration of detail is tolerated.
The stories of Epaminondas and of the four swans
were parcelled out according to exactly the same
principles. 1 As for the points which we made use of in
the mechanical explanations, they have already been
given in the preceding paragraph.
Having disposed of this part of the subject, we then
proceeded to estimate the understanding of the children
1 We give in detail the points which were used, in case anyone ever
repeats our experiments with the same set pieces.
I. Epaminondas: (i) A little nigger boy. (2) A hot country. (3) His
mother sends him to take a shortbread cake. (4) Which arrives broken
(in crumbs, etc.). (5) Because he had held it under his arm. (6) His granny
gives him some butter. (7) Which arrives melted. (8) Because he put it
on his head. (9) And because it was very hot.
II. The four swans : (i) A castle. (2) A king and queen. (3) Who had
children. (4) There was a fairy. (5) She did not like children (or was
wicked, etc.). (6) She changed them into swans. (7) The parents find
their children, or the swans. (8) These go away. (9) To a cold country.
(10) They come back again, (n) The castle and the parents are no longer
there. (12) They are changed into (13) old people. (14) In a church.
We have distinguished between points (12) and (13) because it sometimes
happens that the children think the old people have appeared in the story
without realizing that they are the swans transformed.
88 STUDIES IN CHILD LOGIC
as follows. In the first place, we tried to reduce our
results to numbers and coefficients of understanding.
We are not ignorant of the objections of all sorts which
are raised against the use of measurement in psychol-
ogy. We are aware of the inaccurate and arbitrary
character of such methods of evaluation, and above all
of that element of dangerous fascination which makes
statisticians lose sight of the concrete facts which their
figures represent. But we must not, on the other hand,
judge the psychologists to be more nai've than they
really are. It is too often the reader who takes the
figures literally, whilst the psychologist moves more
slowly to his conclusions. Our figures will yield much
less than they seem to contain. We shall look to them
in this work, not so much for an exact measurement
that seems to us premature as for an aid to research
and to the practical solution of the problems. In giving
the solution of these problems we shall rely far more on
the methods of pure observation and clinical examina-
tion than upon rough numerical data. These will serve
at best to sharpen our criticism, and in this capacity
their legitimacy cannot be questioned. Let the reader
then be not too hastily shocked, but quietly wait for our
conclusions. In the meantime, let us confine ourselves
to the quest for schemas of objective evaluation, i.e.,
schemas, which, though founded on pure conventions,
admit of being put into practice by every one with the
same result.
We shall distinguish, in the first place, general under-
standing, i.e., the manner in which the reproducer has
understood the whole of the story told by the explainer,
and verbal understanding bearing upon causal or logical
relations. The latter bears upon certain points in the
stories, and will concern us later on.
Within general understanding we shall distinguish
on the one hand, between implicit understanding (i.e.,
what the child has understood without necessarily being
able to express it) and explicit understanding (i.e., what
UNDERSTANDING BETWEEN CHILDREN 89
the child reproduces spontaneously), and on the other,
between the understanding of the explainer in relation
to the adult and the understanding of the reproducer in
relation to that of the explainer.
a = What the reproducer has understood in relation to
what the explainer has understood.
^3 = What the reproducer has understood in relation to
what the explainer has expressed.
y=What the explainer has understood in relation to
what the adult has expressed.
(5= What the explainer has expressed in relation to
what he has understood.
When something is explained to the explainer, one
of three things may happen. Either he does not under-
stand, and therefore cannot repeat anything ; or else he
understands, but either cannot or will not repeat it (for
lack of the means of expression or because he thinks the
thing goes without saying and is known to his hearer,
etc.) ; or, finally, he understands and repeats correctly.
It is important to consider these thfee cases separately.
One of the chief causes of misunderstanding among
children may be due to some personal trait in the
explainer. When such a factor is present it is expedient
to make allowance for it. Here is an example of the
parcelling out of which we spoke:
Schla (6 ; 6) to Riv (6.; 6). Explanation of the draw-
ing of the tap : " You see, this way (diagram I) it is open.
The little pipe (c) finds the little pipe (b) and then the water
runs out. There (diagram II) it is shut and it can't find
the little, pipe that runs through. The water comes this way
(diagram I, c) it comes in the little pipe. It is open, and
there (II) it is shut. Look, you can't see the little pipe any
more (II) it is lying down? then the water comes this way
(c) and wouldn't find the little pipe any
If the -reader refers to the points given in the pre-
ceding paragraph, he will find what follows. Point (i)
is understood by Schla; he had told us just before
speaking to Riv that it was about a tap. But he forgets
go STUDIES IN CHILD LOGIC
to mention it to Riv, probably because for himself it
goes without saying. Point (2), the part played by the
handle is also understood. Schla had said to us :
' i There are two little bars there (a). When you turn them^
it runs out because they turn the pipe round" This ex-
planation is good. In his exposition to Riv, on the
other hand, no mention is made of the handle of the
tap. Schla contents himself with saying : "It is open"
or "zV is shut" which seems to him sufficient to recall
the movement with which one turns the handle of the
tap. Is this carelessness or forgetfulness, or does
Schla think that Riv has understood things sufficiently
clearly? We shall not discuss these points for the
moment. It will be sufficient for us to note their im-
portance in the mechanism of childish language.
Point (3) is also understood (" When you turn "). Schla
knows and tells us that it is with the fingers that the
handle of the tap is made to revolve. He does not say
so to Riv either, because it goes without saying or for
some other reason. As to the four other points, it is
obvious that they are all correctly understood and
expressed by Riv. The connexion between the fact
that "it is open" and that the water runs through the
canal b is very well indicated, as is also the movement
of the water. The opposite connexion (between the
closing of the canal, the movement of the handle and
the stoppage of the water) is also indicated.
The nine points of the explanation have been under-
stood by Schla. Even though in talking to Riv he
may not have expressed himself clearly and explicitly
throughout, still, so far as he himself is concerned,
the child has understood everything and can give us
spontaneous proof of having done so (otherwise we
would have tested him with the questions of which we
spoke in a previous paragraph). If then we calculate
the coefficient of y we get
_ Number of points understood by the explainer _ 9 __
Number of points to be understood """ """
UNDERSTANDING BETWEEN CHILDREN 91
The points not expressed to the reproducer (to Riv)
do not enter into this coefficient. They do so, on the
other hand, in the calculation of the coefficient of $
Dumber of points expressed to the reproducer _ g _ .,,
Number of points understood by the explainer 9
The significance of the coefficients y and S will now
be clear. The first gives a measure of the explainer's
understanding in relation to the experimenting adult.
The second gives a measure of the value of the explana-
tion given by the explainer to the reproducer.
Let us now see how much Riv has understood of
the explanation given by Schla. Here are Riv's words
verbatim :
Riv (6 ; 6) " Here (I, c) there* s the pipe, and then it is
opened, and then the water runs into the basin, and then
there (II, c) it is shut, so the water doesrit run any more,
then therms the little pipe (II, b) lying down, and then the
basiris full of water. The water can't run out *cos the
little pipe is there, lying down, and that stops it. "
Point (i) (the word tap) is omitted. But has Riv
understood it? We ask him "What is this all about?
A pipe Is it a tap ? No." He has therefore not under-
stood, which is hardly surprising as Schla has not told
him. Point (2) is also omitted. We show Riv the
handle (a) and ask him what it is. He does not know
a thing about it. Nor has he understood what must
be done to make the little pipe turn round (b), although
he might have guessed this through hearing Schla say,
" It is open" etc., even without understanding that the
two a's represent the handle. Thus points (3), (4) and
(7) have been missed. We now test this interpretation
by means of several questions* " What must you do
to make the little pipe lie down?" etc. All the rest,
however, has been understood.
Concerning Riv's understanding, two things have tc
be established. In the first place, there is its relation
to Schla's understanding, /.<?,, not only to what Schls
92 STUDIES IN CHILD LOGIC
has expressed, but also to what he has understood
without expressing it (a). Secondly, there is the relation
of Riv's understanding to what Schla has made quite
explicit (/3). In this connexion, points (4) and (7),
which are expressed by Schla (" There it is shut^ and
can't find the little pipe that runs through" etc.), are not
understood by Riv. Now Riv might have discovered,
even without knowing that the ds represent the handle
of a tap, that in order to close the canal b or make it
lie down, something would have to be turned, or 'shut
off.' But this relation has completely escaped his notice,
though pointed out by Schla, and emphasized with
gesture. It may be objected that Schla has not ex-
pressed this relation very clearly, but the point is that
he has expressed it in the childish manner of juxta-
position (cf. 6). Instead of saying: "It can't find
the little pipe because it is shut," Schla says : " It is
shut, and it can't find the little pipe"" This is the style
in which Riv thinks. Why should not Schla under-
stand him, since he too must surely think in the same
manner?
Riv has therefore understood 4 points out of the 6
that have been expressed and out of the 9 that have
been understood by Schla. This yields the two co-
efficients a and /?.
__ Everything understood by the reproducer _ 4 __
~" Everything understood by the explainer ~~ tf ~" 44-
R Everything understood by the reproducer _ 4 __ .,,
Everything expressed by the explainer ~ "5" ~~
Since points (4) and (7) are expressed by Schla in
the style of juxtaposition, they might be considered as
not expressed, so that coefficient /3 would be changed
to |-=roo. We shall agree, however, to look upon
juxtaposition as a means of expression until we make
a special study of it later on ( 6).
The meaning of the coefficients a and j3 is therefore
clear. Coefficient a indicates how much the explainer
UNDERSTANDING BETWEEN CHILDREN 93
has been able to convey to the reproducer. Its varia-
tions are therefore due to two factors distinct from one
another though combined in this case into a single
measure: i The fact that the explainer cannot or will
not always express himself clearly ; 2 the fact that
the reproducer does not always understand what the
explainer says, even when the latter expresses him-
self quite clearly. These two factors, the explainer's
capacity for understanding, and the reproducer's capacity
for expressing himself are indicated by coefficients S and
/3 respectively. Coefficient a therefore, which virtually
contains them both, represents in so far as the experi-
ments are not artificial, nor the method of parcelling
out arbitrary a measurement of verbal understanding
between one child and another, since it measures both
the manner in which one of the speakers makes himself
understood and the manner in which the other under-
stands. This coefficient a, moreover, is a true measure
of the understanding between child and child^ since it is
calculated in relation to what the explainer has actually
remembered and understood of the set piece, and not
in relation to what he ought to have understood. If
Schla had understood only 4 points instead of 9, a would
be % and y would be 0*44. Understanding between
child and child would be perfect, however deficient
might be that between child and adult.
The coefficient /3 is a measure of the understanding
between child and child in the restricted sense, z.*., of
the understanding of the reproducer in relation to what
the explainer has been able to express. The respective
values a and /3 must therefore not be confused, since
each has its own particular interest.
In order to show straight away what can be deduced
from such coefficients, we can say that in the case of
Schla and Riv, which we have just examined, one
child has understood the other definitely less than this
other understood us. Riv understood Schla in a pro-
portion of 0*44 (a), and Schla understood us in a
94 STUDIES IN CHILD LOGIC
proportion of roo (y =-) What is the cause of this
lack of understanding between Schla and Riv? Is it
Riv's deficient understanding, or Schla's faulty exposi-
tion ? Riv's understanding in comparison to what Schla
has expressed is 0*66 (/3=i). The value of Schla's
exposition in comparison to what he has himself under-
stood is also 0-66 (S = %). We may conclude from this,
that the non-understanding between Schla and Riv is
due as much to the deficient exposition of the one as
to the deficient understanding of the other.
The dissection of the stories follows exactly the same
method. Special kinds of understanding (causality,
etc.) will be examined later on.
3. NUMERICAL RESULTS. By parcelling out in this
way the 60 experiments made on our 30 children from
7 to 8 (all boys), we reached the following results.
Once again, however, we lay stress on the fact that
we do not consider that our problem can be solved by
figures. We have far too little confidence in the value
of our method of parcelling out, and especially in the
general value of our experiments, to come to such hasty
conclusions. Our experiments are carried out "just to
see," and are meant only as a guide to any future
research.
The figures which we shall give are therefore meant
only as a help to observation of facts and to clinical
examination. They contain, it is true, a statistical
solution of the problem. But we shall adopt this solution
only as a working hypothesis, in order to see in the
later paragraphs whether it really tallies with the clinical
evidence, and whether this tallies with the facts revealed
by everyday observation.
Having disposed of these preliminary considerations,
let us pass on to the actual figures. With regard to the
stories, understanding between children as indicated by
the coefficient a was found to be only 0*58. Now the
explainer understood us on the average quite well, since
the coefficient y reaches 0*82. The explainer's power of
UNDERSTANDING BETWEEN CHILDREN 95
exposition also proved quite good, the coefficent <S being
0*95. It is therefore the understanding of the repro-
ducer which is at fault ; ft is only 0*64.
It should be noted that if we subtract the deficit due
to the explainer (i'oo 0*95 = 0*05) from the deficit due to
the reproducer (0-64 0-05 = 0*59), we get the total deficit
(0*54). This will be of use to us later on.
With regard to explanations, understanding between
children is also greatly inferior to understanding between
explainer and adult. Here the coefficient a is 0*68 and
y 0*93. Explanations are therefore generally better
understood than stories, both between children and
between children and adults. This may be an accident,
due to the method of parcelling out (the 9 points of the
explanations are perhaps easier to remember, because
they are more comprehensive). Whether this is so or
not is of no consequence. What matters is not this value
0*68, taken by itself, but the interrelations which it implies.
The deficit on the part of reproducer and explainer is
quite different in this case from what it was in the stories.
The explainer does not express himself nearly so well ;
S is only 0*76 instead of 0*95 as in the case of the stories.
But the proportion of what the explainer has expressed
which is understood by the reproducer (ft) is 079 instead
of 0*64, as in the case of the stories. Explanations,
therefore, seem to resemble the procedure of ordinary
life much more closely than do stories. This impression
receives further confirmation from the fact that if the
share of the explainer indicated by coefficient (<5) be
added to the share of the reproducer (ft), the result is not
equal but inferior to the total deficit. 1*00076 = 0*24
and 0790*24 = o*55<o*68.
This circumstance is easy to explain. In the case
of the stories, when the explainer expresses himself
badly, the reproducer cannot supplement obscure or
forgotten passages. He has a tendency to distort even
correctly given material, and especially a tendency
not to listen to his interlocutor. This was abundantly
96 STUDIES IN CHILD LOGIC
shown in the spontaneous conversations of children
which we gave when dealing with the collective mono-
logue (e.g. when Pie says to Bea : " I am doing a stair-
case, look," and Bea answers: "I can't come this
afternoon, I've got a Eurhythmic lesson.") In the case
of mechanical explanations, on the other hand, the
reproducer has already been interested on his own
account in the handling of taps and syringes. He has
the diagrams before him, and can think about their
meaning while the explainer is talking. Thus even
if the explainer has not been listened to, or if he is
obscure and elliptical, the reproducer can reconstruct
the required explanation. This is why the complete
understanding a is better than one would expect from
the sum of the deficits indicated by the coefficients
S and /?. The relations, moreover, seem to us to exist
independently of the particular mode of parcelling out
which we have adopted.
The value of coefficient a, therefore, does not neces-
sarily imply that the absolute understanding is good.
It does not mean that the explainer is capable of
making the reproducer understand something new
and hitherto unknown to him. On the contrary, the
added deficits yield 0*56, whereas in the case of the
stories they amount only to 0*59. Roughly speaking,
then, the understanding of explanations is less good
than that of stories. If, therefore, a is better in the
case of explanations, it is because the reproducer has
identified himself more fully with what he is repro-
ducing ; and this he was enabled to do thanks to the
diagrams and to his own previous interests. Apparent
comprehension has been in this case a mutual stimulus
to individual reflexion. And this is the initial stage
of all understanding, even with adults.
The fact that the explainer's capacity for exposition
(<5) is better in the case of stories than in that of explana-
tions has nothing that need surprise us. Explanation
presupposes a certain number of verbal expressions
UNDERSTANDING BETWEEN CHILDREN 97
difficult to handle, because connected with causal re-
lations. Stories are told in a much simpler style.
These conclusions receive complete confirmation from
the results obtained between the ages of 6 and 7. We
experimented on 20 children of this age, of whom 8
were girls. 1 With these, too, understanding between
children was weaker than understanding of the adult
by the child, and in more marked proportions than
between the years of 7 and 8. Thus in the case of ex-
planations, children's understanding of each other was
a = 0-56, and of us, y = o*8o. With regard to stories,
their understanding of each other was a = 0*48, and of
us, y = o*7o. It should be noted that these coefficients
y = o*8o and 7 = 070 prove that the use of the same
explanations and stories is justified in spite of the dif-
ference of age in the subjects, since the explainer has
been able to understand us in the above proportions.
What is the cause of this relative lack of understand-
ing amongst themselves in children from 6 to 7 ? Does
the fault lie with the explainer's means of expression, or
with the reproducer's capacity for understanding? The
explainer expresses himself as well from 6 to 7 as from
7 to 8 (S = 076), and almost as well in the case of stories
(S = 0*87 as compared to 0*95). The amount understood by
the reproducer of what the explainer has duly expressed
is once again low (070 and 0*61), and, curiously enough,
in exactly the same proportions as in the cases observed
between 7 and 8. In the stories, the coefficient a is equal
to the sum of the deficits indicated by /3 and S :
1*00 0*87 = 0-13 and 0*61 o-!3 = o*48 = a.
In the case of explanations, on the other hand, the
coefficient a is greater than the sum of the deficits :
i *oo 076 = 0*24 and 070 0*24 =
1 It is certainly regrettable to mingle the sexes in an enquiry of this kind 3
but we found no appreciable differences between the boys and those eight girls
probably because of the latter's small number. Our 40 experiments between
7 and 8 may therefore be regarded as fairly homogeneous.
G
9 8 STUDIES IN CHILD LOGIC
In conclusion we obtain the following table :
Explanations. Stories.
a y p d a 7 p 5
6-7 years 0*56 0*80 0*70 0*76 0*48 070 0*6 1 0*87
7-8 years o'68 0*93 079 076 0*54 0*82 0*64 0-95
What conclusions can we draw from these figures?
We have undertaken to be cautious. Shall we assert
straight away that children understand each other less
than they understand us, at any rate in so far as verbal
understanding is concerned? This is what the experi-
ments would seem to show. But In these we took
special care to make ourselves intelligible, which is
not always done by those who talk to children. It is
true that in everyday life there is often what Stern
has called * convergence ' of the language used by
parents towards a childish style of speech. Parents
instinctively use easy expressions, of a concrete and
even animistic or anthropomorphic nature, so as to
come down to the mental level of the child. But side
by side with this there are all the manifestations of
verbalism, there is everything that the child picks up
and distorts, and there is everything that passes him by.
We need only remind our readers of the very definite
results obtained by Mile Descoeudres and M. Belot on
the lack of understanding between children and adults. 1
We shall therefore confine ourselves to the following
conclusions. In verbal intercourse it would seem that
children do not understand each other any better than
they understand us. The same phenomenon occurs
between them as between them and us : the words
spoken are not thought of from the point of view of
the person spoken to, and the latter, instead of taking
them at their face value, selects them according to his
own interest, and distorts them in favour of previously
1 A. Belot, "Les ecoliers nous compretinent ils?" Bull. Soc. Alf.
Biiiet.
A. Descoeudres, "Guerre au verbalisme," Int&rm, des Educ., 1913,
** Encore du verbalisme," ibid., 1917.
UNDERSTANDING BETWEEN CHILDREN 99
formed conceptions. Conversation between children is
therefore not sufficient at first to take the speakers out
of their ego-centrism, because each child, whether he
is trying to explain his own thoughts or to understand
those of others, is shut up in his own point of view.
This phenomenon occurs, it is true, among adults.
But these have had at least some practice in argument
or conversation, and they know their faults. They
make an effort to understand and be understood, unless
indeed distrust or anger reduces them to a childish
state, because experience has shown them the appall-
ing density of the human mind. Children have no
suspicion of all this. They think that they both under-
stand and are understood.
Such, then, is our working hypothesis. Analysis
of our material will show us what it is worth. Let the
reader not ascribe to us more than we are actually
saying. We are merely postulating that the language
of children and between children is more ego-centric
than ours. If this can be verified by analysis, it will
explain a number of logical phenomena such as verbal
syncretism, lack of interest in the detail of logical
correspondences or in the l how ' of causal relations,
and above all, incapacity for handling logical relations,
a task which always implies that one is thinking about
several points of view at the same time (Chapters IV
and V, and early Chapters of Vol. II).
4. EGO-CENTRISM IN THE EXPLANATIONS GIVEN
BY ONE CHILD TO ANOTHER. There emerges from
our statistics the paradoxical fact, common in children
from 7 to 8 and 6 to 7, that stories are less well understood
by the reproducer than are mechanical explanations,
even though they receive a better exposition on the
part of the explainer. In the case of stories the
numerical value of the exposition is 0*95 and 0*87
respectively, and the coefficient /3 0*64 and 0*61 ; for
mechanical explanations the exposition amounts to
0*80 and 0*70, and the coefficient /3 to 0*80 and 0*70.
too STUDIES IN CHILD LOGIC
This enables us to conclude that the reproducer's under-
standing is partly independent, of the exposition given
by the explainer. This exposition is one, therefore,
that is very inadequate. When we say, e.g., that its
value is 0-95, all we mean is that the points expressed
by the explainer are in a proportion of 0*95 to those
which he has understood. But the manner in which
these expressed points are connected and presented to
the hearer may be very bad. In other words, the
explainer's style may be said to present a certain
number of features which prevent it from being intel-
ligible, or at any rate from being very 'socialized.'
These are the features which we must now try to
elaborate.
The most striking aspect of explanations between one
child and another which we have had occasion to study
in the course of these experiments is constituted by
what may be called the ego-centric character of childish
style. . This feature is in full agreement with those of the
spontaneous language of children, which we described
in an earlier chapter. We must take our stand on this
agreement between the products of pure observation
and the products of experiment, for it alone will enable
us to find the significance of the latter. Now we have
seen that the child of 6 to 7 still talks to a great extent
for himself alone, without trying to gain the attention
of his hearer. Thus a portion of the child's language
is still ego-centric. When, moreover, the language
becomes socialized, the process at first only touches the
factual products of thought, z\e., in talking to each other
children avoid the use of causal and logical relations
(because, etc.), such as are used in all "genuine argu-
ment" or in " collaboration in abstract thought."
Before the age of 7 or 8 these two kinds of relations are
therefore still unexpressed, or rather, still strictly indi-
vidual. Observation shows that up till the age of about
7 or 8, the child, even when he can think of them
himself, does not spontaneously give explanations or
UNDERSTANDING BETWEEN CHILDREN 101
demonstrations to his equals, because his language is
still saturated with ego-centrism.
Now this is the very phenomenon which we met
with in our experiments. The explainer always gave
us the impression of talking to himself, without bother-
ing about the other child. Very rarely did he succeed
in placing himself at the latter's point of view. It
might be thought that this was because he was
addressing himself to the experimenter as though he
were reciting a lesson, and forgot that he had to make
his playmate understand. But spontaneous language
between children exhibits exactly the same features.
Moreover, the explainer sprinkles his exposition with
such expressions as: "You understand, you see, etc,"
which shows that he has not lost sight of the fact that
he is talking to a friend. The cause of his ego-centrism
lies much deeper. It is extremely important, and
really explains all the ego-centrism of childish thought.
If children fail to understand one another, it is . because
they think that they do understand one another. The
explainer believes from the start that the reproducer
xvill grasp everything, will almost know beforehand
all that should be known, and will interpret every
subtlety. Children are perpetually surrounded by
adults who not only know much more than they do,
but who also do everything in their power to under-
stand them, who even anticipate their thoughts and
their desires. Children, therefore, whether they work
or not, whether they express wishes or feel guilty, are
perpetually under the impression that people can read
their thoughts, and in extreme cases, can steal their
thoughts away. The same phenomenon is undoubtedly
to be found in Dementia Precox and other pathological
cases. It is obviously owing to this mentality that
children do not take the' trouble to express themselves
clearly, do not even take the trouble to talk, convinced
as they are that the other person knows as much or
more than they do, and that he will immediately under-
102 STUDIES IN CHILD LOGIC
stand what is the matter. This mentality does not
contradict ego-centric mentality. Both arise from the
same belief of the child, the belief that he is the centre
of the universe.
These habits of thought account, in the first place,
for the remarkable lack of precision in childish style.
Pronouns, personal and demonstrative adjectives, etc.?
c he, she' or 'that, the, him/ etc., are used right and
left, without any indication of what they refer to. The
other person is supposed to understand. Here is an
example :
Gio (8 years old) tells the story of Niobe in the role of
explainer: " Once upon a time there was a lady who had
twelve boys and twelve girls, and then a fairy a boy and a
girl. And then Niobe wanted to have some more sons [than
the fairy. Gio means by this that Niobe competed with
the fairy, as was told in the text. But it will be seen how
elliptical is his way of expressing it]. Then she [who?]
was angry. She [who?] fastened her [whom ?] to a stone.
He [who ?] turned into a rock, and then his tears [whose ?]
made a stream which is still running to-day."
Fom this account it looks as though Gio had under-
stood nothing. As a matter of fact he had grasped
nearly everything, and his understanding in relation to
us was y = o*9i (<S = o*8o). He knew for instance that
the fairy was angry " because she (N.) wanted to have more
children than the fairy" The pronouns distributed at
random are therefore a characteristic of the style, and
not a proof of lack of understanding. Gio knows per-
fectly well that it was the fairy who fastened N. to the
rock and not vice versa.
It is easy to foresee the results of such a style. The
reproducer, Ri (8 years old), begins by taking N. for
the fairy, and by thinking that it is N. who fastens the
lady. After being put right on this point, he reproduces
the story as follows :
There was a lady once, she had twelve boys and twelve
girls. She goes for a walk and she meets a fairy who had
UNDERSTANDING BETWEEN CHILDREN 103
a boy and a girl and who didn't want to have 12 children.
1 2 and 1 2 make 24. She didn't want to have 24 children*
She fastened N. to a stone, she became a rockj etc.
(18=072).
Another example :
Kel (8 years old) also tells the story of Niobe and
says of the fairy : " She fastened the lady to a rock. She
[who?] cried for ten years. They are still running to-day"
The word ' tears ' Is taken for granted. As It cannot be
heard that the verb is in the plural 1 this style is in-
comprehensible. It sounds as if it were the lady or
the rock that was running. We ourselves did not
understand in the first instance.
In the case of mechanical explanations, the explainer
assumes from the beginning that the 'doors,' 'pipes/
and ' bars ' are known to the reproducer, so that instead
of beginning by showing them and explaining their
uses, he speaks of them as familiar objects. Here is
an example :
Pour (7 ; 6) explains the tap to Pel (7 ; o) : " The water
can go through there [points to the large pipe in fig. i,
without designating the exact spot, the opening] because
the door [which door?] is above and below [the movable
canal b which he does not show] and then to turn it [turn
what?]jF0& must do so [makes the movement of turning
fingers but without pointing to the handles a]. There^
it [what?] can't turn round [ = the water can't get through,]
because -, the door is on the right and on the left* There^
because the water stays there^ the pipes can't get there [the
pipe is lying down. Note the inversion of the relation
indicated by the word 'because.' What ought to have
been said was: "The water stays there because the
pipes can't . . . etc."] and then the water can't run
through"
The words used by Pour, the ' door,' the ' pipes ', are
supposed to be known by Pel ; so much so, that Pour
forgets to show these objects in the diagram. And
1 The French, for tear, * larme ', is feminine and referred to by the personal
pronoun 'elle' in the singular, and Belles' in the plural. Phonetically the
two are indistinguishable. [Translator's Note.]
104 STUDIES IN CHILD LOGIC
yet Pour, as our Interrogatory proves has a fair under-
standing of the detail of the diagram (y = o-88). Only
his style is faulty. In his reproduction, consequently,
Pel talks about doors which he takes literally without
seeing where they are. " The water can't run because it is
stopped, and there are doors that stop it, they are shut, and
then the water can't run through" What is most remark-
able is that Pel manages to understand pretty well
everything, but by his own effort (a = 075). As to what
Pour has said to Pel, this remains for the latter a purely
verbal affair.
It will perhaps be objected that such phenomena are
due to the scholastic atmosphere which gives rise to
verbalism. On this hypothesis the explainer speaks,
not to make himself understood, but for the sake of
speaking, as one recites a lesson. But we have already
answered this objection by pointing out how, in their
spontaneous conversations, children express themselves
in the same vague manner, because they talk much
more for themselves than for their hearer. Take, for
example, the slip-shod use of words even in "association
with the action of each " (Chapter II, 4) where children
talk to one another spontaneously.
" The daddy bear [which one] is dead. Only [?] the
daddy [the same or another?] was too ill" " There was a
blue one [talking of aeroplanes without mentioning them.]
" 1 'want to draw that" meaning by 'that' probably a
flying race or anything else connected with it.
We find the same inaccuracy in the qualifying words,
the same method of alluding to objects supposedly
known. Here is one more example of explanation,
which was observed in the course of our experiments,
and of which the style exactly resembles that of explana-
tions given spontaneously by children.
Toe (8 years old). Fragment of the explanation of
the tap : " That and that [the two extremities of canal b]
is that and that [id. on diagram II] because there [dia-
UNDERSTANDING BETWEEN CHILDREN 105
gram I] it is for the water to run through, and that
[diagram II] you see them inside because the water can't
run out. The water is there and cannot run." Toe there-
fore points to the two extremities of a canal without
saying anything about a canal or alluding to the handle
(#), in short, without naming any of the objects which
he is discussing. Still, having got so far, he believes,
as he tells us, that his hearer (Kel, 8 years old) has
understood everything. Kel is able, indeed, to repeat
more or less the same words, but without attaching
any concrete meaning to them. We ask him before
Toe: "What was done so that the water should stop
running? It was turned round What was? The pipe
(b] [correct]. How was the pipe turned round (b) ?
.... What is this for (the handle, #)?... (He
doesn't know)." Toe is then astonished to see that Kel
has understood nothing, and he begins his explanation
all over again. But, and this is the point we wish to
emphasize, for it proved to be very usual, his second
exposition is no clearer than the first : " This is a thing
[the handle (a) which he had forgotten to mention]
like this [diagram I] this way ifs that the water can't run.
When this thing is like that [diagram II] it means that the
water can't run" Thus, even when he wants to make
things clearer to Kel, Toe forgets to tell him that it is
the handle which turns the canal or which is moved
with the fingers, etc. In a word, unless Kel guesses
and this is just what he fails to do in this particular
case the language used is unintelligible. But the main
reason for Toe talking in this way is his belief that
things go without saying, and that Kel understands
immediately.
These features of ego-centric style are still more pro-
nounced between 6 and 7, which goes to prove that they
are not scholastic habits. Between 6 and 7 the children
are still in the so-called Kindergarten classes, which
are far less coloured with verbalism than those above
them. These children, moreover, play amongst them-
selves far more than in the elementary classes. Now
the ego-centrism of their explanations is far more pro-
nounced ; which proves that this ego-centrism is due to
the general factors of language and thought which we
io6 STUDIES IN CHILD LOGIC
laid stress on in the preceding chapters dealing with
spontaneous language,
Riv (6 years old), for example, begins his explanation
of the syringe by pointing to diagram III and saying :
4 You see, there (b) is the piston [what piston ? the piston
of what?] then you pull it and it makes a squirt [concluded
too quickly]. Then it leaves room for the water [why does
it make this room ?] When you push the little piston [he no
longer points to it] it makes the water come out, it makes
a squirt, you see ? That is the bowl there (a) and then the
water."
Now Riv has completely understood (y=i). More-
over, he is very definitely addressing himself to his
hearer Schla, as is shown by the expressions "you see,
you understand," and by the interest which both the
children have evinced in the matter. It goes without
saying that Schla has not understood a thing :
Schla (6 years old) reproduces Riv's explanation :
"He told me that it was . . . something. There was
something, and then there was something where there was
water, and then the water came out. That is where the
water was. That (a) is the place where the water was, and
the water squirted the two bowls and poured into them "
(a = 0-33).
As will be seen by comparing these two texts, it is
only Riv's inaccuracy that has confused Schla. Other-
wise the explanation would have been adequate. From
Riv's last sentences the whole mechanism might have
been reconstructed. But Riv, thanks to Schla, has
taken the syringe for a tap, and consequently has under-
stood nothing of the movement of the piston.
Another example :
Met (6; 4) G, talking of Niobe : "The lady laughed
at this fairy because she [who ?] only had one boy. The lady
had twelve sons and twelve daughters. One day she [who ?]
laughed at her [at whom ?] She [who ?] was angry and she
[who?] fastened her beside a stream. She [?] cried for
fifty months^ and it made a great big stream" Impossible
to tell who fastened, and who was fastened. Met knows
perfectly well (7 = 0-83), but Her (6 ; 3) G, who is
UNDERSTANDING BETWEEN CHILDREN 107
listening, naturally understands things the wrong way
round. She thinks it is the fairy "who laughed at the
lady who had six boys and then six girls" and that it was
the fairy who was fastened, etc. (0 = 0*40).
Finally, one of the facts which point most definitely
to the ego-centric character of the explanations of
children is the large proportion of cases in which the
explainer completely forgets to name the objects which
he is explaining, as in the cases of taps and syringes.
This holds good for half of the explainers from 6 to 7,
and for one-sixth of those between 7 and 8. They
assume that their hearer will understand from the out-
set what they are talking about. Naturally, in such
cases, the reproducer gives up trying to understand,
and repeats the explanation he has received, without
attempting to assign a name to the object in question.
5. THE IDEAS OF ORDER AND CAUSE IN THE
EXPOSITIONS GIVEN BY THE EXPLAINERS. Other factors
are at work which help to render the explainer's ex-
position rather unintelligible to the reproducer. These
are an absence of order in the account given, and the
fact that causal relations are rarely expressed, but are
generally indicated by a simple juxtaposition of the
related terms. The explainer, therefore, seems not to
concern himself with the ' how ' of the events which he
presents ; at any rate, he gives only insufficient reasons
for those events. In a word, the child lays stress on
the events themselves rather than on the relations of
time (order) or cause which unite them. These factors,
moreover, are probably all connected in various degrees
with the central fact of ego-centrism.
The absence of order in the account given by the
explainer manifests itself as follows. The child knows
quite well, so far as he himself is concerned, in what
order the events of a story or the different actions of
a mechanism succeed one another ; but he attaches
no importance to this order in his exposition. This
phenomenon is due once more to the fact that the
io8 STUDIES IN CHILD LOGIC
explainer speaks more to himself than to the explainer,
or rather to the fact that the explainer is not in the
habit of expressing his thoughts to his companions,
is not in the habit of speaking socially. When an
adult narrates, he is accustomed to respect two kinds
of order: the natural order given by the facts them-
selves, and the logical or pedagogic order. Now it is
to a great extent because of our concern with clarity
and our desire to avoid misunderstanding in others
that we adults present our material in a given logical
order, which may or may not correspond with the
natural order of things. The child, therefore, who,
when he explains his thoughts, believes himself to be
immediately understood by his hearer, will take no
trouble to arrange his propositions in one order rather
than another. The natural order is assumed to be
known by the hearer, the logical order is assumed to be
useless. Here is an example :
Ler (7 ; 6) explains the tap : " Ifs a fountain. It either
runs, or it doesn't run, or it runs. When it is like that
[diagram I] it runs. And then there s the pipe \c\ that the
water goes through. A nd then, when it is lying down [b]
when you turn the tap, it doesnt run* When it is standing
upright, and then you want to turn it off, ifs lying down.
[Note the curious treatment of the temporal sub-clauses].
And then that is . . . [the basin]. And then when it is
standing upright [again the canal b] it is open and when it
is lying down it is shut?
Del (7 years old) : " That is a tap, and then you turn
it on and then the water runs into the basin, and then to find
its way, it goes through the little pipe. [Note the inversion
of these two propositions] and then there is the handle
that you turn round . . . etc."
This mode of exposition, which consists in connecting
propositions by " and then " is typical. The conjunction
"and then" indicates neither a temporal, a causal nor
a logical relation, i.e., it indicates no relation which the
explainer could use in order to link his propositions
together for the purpose of a clear deduction or demon-
UNDERSTANDING BETWEEN CHILDREN 109
stration. The term " and then " marks a purely personal
connexion between ideas, as they arise in the mind of
the explainer. Now these ideas, as the reader may see,
are incoherent from the point of view of the logical or
of the natural order of things, although each one taken
separately is correct.
Order may be absent even from the account of the
stories, but this is rarer. Here is an example :
Due (7 years old) : " Once upon a time, there were four
swans, and there was a king and a queen who lived in a
castle and had a boy and a girl. Near by, there was a
witch who did not like the children and wanted to do them,
harm. They turned into swans, and then they were on the
sea . . . etc." The swans are made to appear before the
meeting of the witch and the children, although the sequel
shows that Del is perfectly familiar with their origin.
But there is considerable difference between the ex-
position of explainers between 7 and 8 and that of those
between 6 and 7. This is one of the most important
points brought out by the parcelling out of our material,
and one which only goes to show how independent of
scholastic habits this material is. The absence of order
which we have described is more or less exceptional
between 7 and 8. Between 6 and 7, it is the rule. It
seems pretty certain, therefore, that the capacity for
arranging a story or an explanation in a definite order
is acquired some time between the ages of 7 and 8.
The question, of course, will have to be approached by
other methods, for it would be highly desirable to prove
what is at present only a hypothesis, viz., that the
capacity for order makes its appearance at the same
time as genuine argument, as collaboration in abstract
thought, (see conclusion of Chapter II) and at the same
time as incipient understanding among children (a stage
between 7 and 8 during which $ goes beyond 75 per
cent, for explanations and reaches 0*79). But there
are indices which point to such a chronology. We
know, for example, that it is at 7 that Binet and Simon
no STUDIES IN CHILD LOGIC
fixed the three messages test (to carry out three
messages in a given order). Now, before the age of 7,
children succeed in carrying out the messages, but not
in the given order. Terman, it is true, has brought the
test age down to 5, but this seems to us excessive. It
is at the lowest a test for 6-year-old children. Now it
is easier to follow a certain order in actions than in a
spoken account. This leads us back to the view that
between 7 and 7j is the age when the desire for order
in the expositions given by children first makes its
appearance.
Here, for instance, are two terms of comparison, the
story of the four swans told by a child of 7| who is
typical of that age, and the same story told by a typical
child of 6 ; 4 :
Cor (7 ; 6) : u Once upon a time, in a great big castle,
there was a king and a queen who had three sons and one
daughter. Then there was a fairy who didn't like the
children, then she brought them to the seashore, then the
children changed into swans, and then the king and queen
they looked for the children till they could find them. They
went down as far as the seashore, and then they found the
four children changed into swans. When the swans had
gone away on the sea, they went towards the castle, they
found the castle all destroyed then they went to the church,
then the three children were changed into little old men
and one little old woman" Thus the order of events
has been respected.
Met (6 ; 4) G : Once there was a fairy, there was a king
and then a queen. Then there was a castle, there was
a wicked fairy [the same] who took the children [which
children ?] and changed them into swans. She led them
to the seashore [Inversion]. The king and queen came
in, and couldn't find them. They went to the seashore
and found them. They went into a castle [the same one.
Met knows this] and changed them into little old people.
After that [ ! ] they found them. [This had already been
said. Met knows quite well that it was prior to the
transformation into old people]. It may be claimed
that this lack of order is simply due to a lack of memory.
This is certainly one of the factors at work, but not the
UNDERSTANDING BETWEEN CHILDREN in
only one. The proof of this is that when we read the
story over again to Met, she proceeded to tell it as
follows : " Once there was a king and a queen. There
were three children, one little girl and three boys. There
was a wicked fairy who had changed the children into white
swans. Their parents looked for them and found them by
the seashore. And then [!] they had been changed into
swans [reversion to what has already been said]. They
said it was their children. They had a castle [wrongly
placed]. Their parents died. They went away to a very
cold country [inversion]. They went into a little church
and they were changed into little old men and a little old
woman'"
Or again, this opening to the story of Niobe : Ce
(6 years old) : " There's a lady who was called Morel, and
then she turned into a stream . . . then [ ! ] she had ten
daughters and ten sons . . . and then after that [ ! ] the fairy
fastened her to the bank of a stream and then she cried
twenty months -, and then after that [ ! ] she crted for twenty
months and then her tears went into the stream^ and then
. . . etc."
The question may of course be raised, whether the
explainer has understood. This we have always verified
by appropriate questions. With regard to mechanical ex-
planations, the objection cannot be maintained. Logical
order is far more independent of understanding, and in
the majority of cases the child understands clearly (the
subsequent interrogatory also confirms this), but presents
his material incoherently. Here is a good example of
this incoherence on the part of an explainer who has
understood everything :
Ber (6 ; 3). You see this tap, when the handle is straight
like this [a diagram I] lying down^ you see the little pipe has
a door and the water can't get through [there is no con-
nexion between these facts. There seems to be a mistake.
As a matter of fact, Ber has passed from diagram I to
diagram II] then the water doesn't run^ the door is shut.
Then, you see here [' then ' has no meaning here. He
points to diagram I] you find the little door [b] and then
the water comes into the basin and then the two sides of the
handle [a] are like that [this has already been said] then
the water can run^ and then the pipe is like that [b in
STUDIES IN CHILD LOGIC
diagram II] then there is no little door then the water can't
find the little door. Then the water stays here [c, diagram
II]. When the tap is on [accompanying movement] there
is a little pipe, then the water can get through and the handle,
oh, well I it is lying down [a, diagram I] whereas here, the
little pipe is straight [diagram II, he calls straight what
he had called lying down in the preceding proposition]
the handle is straight [a, diagram II, this time straight
means vertical] and the little pipe [, of which he has just
spoken as being straight] is lying down"
This type of explanation is paradoxical. Ber's under-
standing is excellent (y= roo, there is a wealth of detail
and of vocabulary (e.g. the word ' wherea^ ' which gener-
ally appears somewhere about the age of 7),* but the
order is muddled to the point of unintelligibility. Even
the words (straight and lying down) are taken in a
sense which varies from one moment to another. Trie
result was that the listener Ter (6 years old) hardly
understood a thing, and was obliged to reconstruct the
whole explanation himself, in which task he did not
particularly Distinguish himself (a = 0*66).
There is no need to multiply the examples which
are all pretty much alike. We must rather seek to
establish the nature of a certain peculiarity which is
connected with this lack of order in the explanations,
we mean the fact that when a child is relating an event
or describing a phenomenon he is in no way concerned
with the 4 how ' of these happenings. For, given that
a child has simply to note facts without occupying
himself with their connexions, he is not likely to worry
himself about how these facts are produced in detail.
He is content to feel this detail, but ego-centrically,
z".., without trying to express it. If such-and-such a
condition exists, then such-and-such effect will result,
no matter how. The reason given is always incom-
plete. We shall begin by giving a few examples, and
shall afterwards try to explain this absence of interest
1 As was shown in the reports of the Maison des Petite Cf. Descoeudres,
Le D&mloppement de I* enfant de deux a sept as, p. 190*
UNDERSTANDING BETWEEN CHILDREN 113
in the ' how ' of mechanisms. Here are first of all a
few cases observed in connexion with stories :
Due (7 years old), already quoted on p. 109, relates
the transformation of the children into swans without
pointing out that it was the fairy who caused this trans-
formation. " They turned into swans " and that was the
end of it.
Maz (8 years old) also says : " There was a fairy, a
wicked fairy. They turned themselves into swans*" There
is here mere juxtaposition of the two affirmations, with
no explicit indication concerning the ' how/ Blat (8
years old) says : ' < They turned themselves into swans" etc.
In such cases as these the explainer knows perfectly
well ' how ' the transformation took place ; it was the
work of the fairy. Sometimes the reproducer under-
stands, sometimes not. In the following cases the omis-
sion of the ' how ' is more serious because the explainer
himself is not always interested in the mechanism which
he neglects to explain.
Schi (8 years old) explains the syringe : "You put the
water in there and then you pull. The water goes in there
[c], you push and then it squirts" Schi has more or less
understood (y = O'77) but he mentions neither the hole
nor the empty space left by the piston as it is drawn
up, etc. The result is that the listener understands
inadequately (a=O'55).
Gui (7 ; 6) says among other things : " The tap goes
this way^ and that prevents you making [!] the water run"
He defines neither the function of the canal nor the
effect of the handle on the rotation of the canal.
Ma (8 years old) says that the water cannot run out
of the tap ' * because it is shut, so that the water sharit get
outj because it is shutj they have turned off the tap"
In a word, all these explanations take the essential
thing (the position of the canal, b) for granted, instead
of referring to it explicitly : the explainer has understood
the ' how,' but in his opinion it goes without saying and
is of no interest.
Such vague expressions as these abound among very
H
ii 4 STUDIES IN CHILD LOGIC
young children and also among older ones. We need
not record them all. But it is interesting to note how
frequent they are, and to find out why the child troubles
himself so little with the ' how * of things, both for his
listener and for himself. This indifference concerning
the ( how ' of phenomena constitutes a well-known trait
in the spontaneous explanations of children. Why
should Schi find it quite natural that by " pulling the
piston " the water should go into the syringe, just as
though the piston made the water go up ; that by turn-
ing the tap we stop ' making' the water run, just as
though the water acted in obedience to the dictates of
the handle of the tap? This is an instance of defective
adaptation of childish thought to the details of the
mechanism. But is not this lack of adaptation perhaps
more or less directly connected with the ego-centrism
of thought? For the child, as for us, the criterion of
the value of an explanation is the satisfaction felt by
the subject when he pictures himself creating the
effect which has to be explained with means which he
now considers as causes. Now when we think for
ourselves, everything seems quite simple ; imagination
works more easily, autism is stronger, thought, in other
words, takes on new powers. Between two phenomena,
A and B, known to be connected by a causal chain
which alone will explain the 'how' in question, we
feel it unnecessary to define this relation any further.
This is because we know that we have only to look for
it to find it no matter how and because we are not
very exacting when it is a question of proving things
to ourselves. In the end or rather, from the outset
ego-centric thought ignores this question of 'how.'
When, on the other hand, we want to explain our ideas
to other people, then the difficulties begin to appear,
the need is felt for clearly defined relations, and the con-
necting links are no longer skipped, as when individual
fantasy held its sway.
We do not claim that these considerations explain
UNDERSTANDING BETWEEN CHILDREN 115
the lack of interest shown by the explainers and by
children in general in the ' how ' of phenomena. We
only believe that we have given one of the elements
in this lack of adaptation. There are other and deeper
elements, which we shall meet with again in Chapter V.
In the meantime, this one must suffice : since explainers,
as we have seen, generally speak from their own point
of view, without being able to enter into that of their
listeners, their interests remain ego-centric, and tend to
omit information about the * how J of mechanisms. The
reasons given for phenomena, are therefore generally
incomplete.
This peculiar phenomenon of " incomplete reason or
cause " is all the more interesting to observe in our
present results, because it can quite easily be produced
experimentally, and because we shall meet with it again
in a chapter of our second volume, when we shall deal
with the question of causal conjunctions. Our subjects,
moreover, present a special case of this indifference to
the 'how 1 of mechanisms, and one which we shall
meet with again (Vol. II.) ; we mean the apparent
inversion of the expression ' because.' The conjunction
i because ' seems in these cases to introduce the con-
sequence instead of the cause, as it does when correctly
used. This confusion is due simply to the fact that
the child does not bother about the l how ? which
connects the various events of which he is speaking.
Here is an example : Pour (7 ; 6) in the text of his
which we quoted in the last section, instead of saying
"the water stays there because the pipe is lying down "
or in Pour's own style, u the water stays there because
the pipes can't get there," says exactly the opposite:
"Because the water stays there the pipes can't get there"
Here is another example in which the inversion is
not of a i because ' but of a * why ' (we shall meet with
such inversions in the spontaneous language of children
in Chapter V. 2) : instead of saying, " Why does the
water run here and not run there? Because here the
tap is turned on and there it is turned off." Mart (8
n6 STUDIES IN CHILD LOGIC
years old) says: " Why is the tap turned on here and
turned off there ? [It is because] here the water is running
and there the water is not running!' This ' why ' has
all the appearance of being a "why of motivation"
( = " why was the drawing made of the tap turned off? ")
But in reality this is simply an inversion, due once more
to lack of interest in the detail of the mechanism.
These apparent inversions of cause and effect are due,
as we shall show at greater length in our second volume,
to the circumstance that ' because ' does not yet denote
an unambiguous relation of cause and effect, but some-
thing much vaguer and more undifferentiated, which
may be called the " relation of juxtaposition, 5 ' and
which can best be rendered by the word ' and/ Instead
of saying "the water stays there because the pipe is
lying down," it is of no consequence to the child whether
he says "the pipe is lying down and the water stays
there'' or "the water stays there and the pipe is lying
down." When the child replaces 'and' by * because,'
he means to denote, sometimes the relation of cause and
effect, sometimes the relation of effect and cause.
This fact is due to the important phenomenon of
juxtaposition. Juxtaposition, which really covers all the
facts enumerated in this section, is the characteristic
corresponding to that which, in connexion with drawing,
M. Luquet has called "synthetic incapacity." This it
is which renders the child incapable of making a coherent
whole out of a story or an explanation, but makes him,
on the contrary, tend to break up the whole into a series
of fragmentary and incoherent statements. These state-
ments are juxtaposed to the extent that there exists
between them neither temporal, causal, nor logical *
relations. The result is, that a collection of propositions
juxtaposed in this way lacks something more than
sequence, it lacks any sort of verbal expression denoting
a relation. These successive statements are, at the best,
connected by the term 'and.' In the child's mind this
term undoubtedly answers to a certain dynamic relation,
UNDERSTANDING BETWEEN CHILDREN 117
which might be expressed by " this goes with, 55 and
which might take on several meanings, including that
of causality. But the question is whether the child is
conscious of these different meanings, whether he would
be able to express them, and, finally, whether he
succeeds by means of this juxtaposition in making his
listener understand what he is talking about. It may
well be, on the contrary, that the feeling of relations
remains ego-centric, and therefore incommunicable and
practically unconscious. We shall see later on that
as a matter of fact the use of juxtaposition is little
understood by the reproducer. Here is an example :
Mart (8 years old) : " The handle is turned on and
then the water runs^ the little pipe is open and the water
runs. There, there is no water running, there the handle
is turned off \ and then there is no water running , and here
the water is running. There, there is no water running,
and here there is water running. "
It is obvious that there is no whole here, no synthesis,
but only a series of statements in juxtaposition. Indeed,
there is not a single ' because ' in the whole explanation,
nor a single explicit causal relation. Everything is
expressed factually ; the connexion between the handle
and the canal 6, between the position of the canal b
and the passage of the water, all this is denoted simply
by 'and 5 or 'and then.' It may be objected that we
often express ourselves in the same way. But then
we arrange our propositions in a certain order, and
above all we succeed in making ourselves understood.
Whereas, although Mart has understood everything
(y=i*oo), his listener has only understood part of the
relations (/S==o*77), We must be careful also not to
confuse the ' and 5 which marks a succession in time
as when "the fairy fastened N. and N. cried," and the
' and 5 which replaces a ' why ' and which alone is an
'and' of juxtaposition. The mere absence, moreover,
of the word 'because 5 is not sufficient to characterize
the phenomenon of juxtaposition. This absence must
n8 STUDIES IN CHILD LOGIC
be accompanied by a real incoherence in the sequence
of propositions. Here is one more example :
Ber (cf p. in): " When the handle is straight [I]
. . . the little pipe has a door and then [II] the water can't
get through" and " you'll find the little door and then the
water goes into the basin and the handle is like this" We
have in this example absence of order, absence of causal
relations between the propositions, and absence of
explicit connexions such as * because' or 'then.' It is
therefore a definite case of juxtaposition.
In a word, these remarks all lead us to the conclusion
that the child prefers factual description to causal ex-
planation. He confines himself to describing the parts
of the mechanism, to enumerating, if necessary, the
principal movements that take place, but factually, and
without bothering about the * how ' of things. It some-
times happens, moreover, that this description consists
of a series of propositions devoid of logical or temporal
order, without these propositions being connected by
explicit relations such as ' because* and 'afterwards,'
etc. In such cases as these we have 'juxtaposition.' x
It is interesting to note in our material the presence
and constancy of those features on which we have
already laid stress in connexion with the function of
child language and with the spontaneous explanations
between one child and another, studied under the
heading of 'adapted information' (Chapter I, 6).
This shows very clearly that the relative lack of
understanding between children which we are em-
phasizing here is not merely an artificial product of our
experiments, but is deeply rooted in the very nature
of childish language such as we are able to observe it
in natural conditions. We set aside the question
of language by gesture, which expresses causality in
its own way, but without special words or explicit
designations.
1 We shall return to the phenomenon of juxtaposition in greater detail in
the course of Vol II, especially in connexion with the conjunction * because.'
UNDERSTANDING BETWEEN CHILDREN 119
One of the consequences of this factual way of talking^
z>., this way that is unadapted to causality, is that the
child expresses himself better when he tells stories than
when he gives mechanical explanations. As we have
seen, the coefficient S Is always higher in the stories
than in the explanations.
6. THE FACTORS OF UNDERSTANDING. Given all
the characteristics of explanation between one -child
and another, two alternatives as results are possible.
Either owing to the fact that their characteristics are
due to a structure of thought which is common to all
children, z'.., owing to the fact that all children are
ego-centric, they will understand each other better
(being used to the same way of thinking) than they
understand us ; or, on the contrary, by reason of this
very ego-centrism, they will fail to understand each
other properly, since each one is really thinking only
for himself. The experiments have shown that from
the point of view of verbal understanding, the second
hypothesis is the more in accordance with the facts.
The time has now come for us to examine whether
this lack of understanding is to be altogether laid to
the charge of the explainer, or whether the reproducer
does not in his manner of understanding show signs of
peculiarities, which it may be worth our while to notice.
In the first place, we have seen that the main factor
in rendering the explainer obscure and elliptical is his
conviction that his listener understands from the outset,
and even knows beforehand everything that is said to
him. In this connexion it should be noted that the
listener adopts exactly the complementary attitude :
he always thinks he has understood everything. How-
ever obscure the explanation, he is always satisfied.
Only two or three times in the whole course of our
experiments has the reproducer complained of the
obscurity of the explanation that has been given him.
It may be objected that scholastic habits have helpecl
to render him so easily satisfied. But here again the
120 STUDIES IN CHILD LOGIC
objection misses the mark, because this feature is more
pronounced among very young children. It is the
reproducers of 7 or 8 who have asked of the explainer
the few questions we have had occasion to note. The
little ones, for their part, were always and immediately
satisfied. The earlier chapters showed us, moreover,
that one of the characteristics of children's conversations
is that each imagines he is understanding and listening to
the others, even when he is doing nothing of the kind.
How then are we to characterize the stage of under-
standing between children before the age of 7 or 8?
It is no paradox to say that at this level, understanding
between children occurs only in so far as there is contact
between two identical mental schemas already existing
in each child. In other words, when the explainer and
his listener have had at the time of the experiment
common preoccupations and ideas, then each word of
the explainer is understood, because it fits into a schema
already existing and well defined within the listener's
mind. In all other cases the explainer talks to the
empty air. He has not, like the adult, the art of seeking
and finding in the other's mind some basis on which
to build anew. Conversely, the reproducer has not the
art of grasping what is standing between him and the
explainer, and adapting his own previously formed ideas
to the ideas which are being presented to him. If there
are, previous to the experiment, no schemas common to
the two children, then the words spoken by the explainer
excite in the mind of the reproducer any schema which
may have been suggested by some accidental analogy
or even some simple consonance ; the reproducer then
thinks he has understood, and simply goes on thinking
without ever emerging from his ego-centric groove. 1
1 It may be of interest in this connexion to recall that M. N. Roubakine
(see Ad. Ferriere, "La psychologic bibliologique d'apres les documents et
les travaux de Nicolas Roubakine," Arch, de Psych., Vol. XVI, pp. 101-102)
came to an analogous conclusion in his studies on adult understanding in
reading. He showed that when they read each other's writings, adults of
differing mental types do not understand each other.
UNDERSTANDING BETWEEN CHILDREN 121
This is the reason why mechanical explanations are
better understood than stories, even though they are
more difficult to reproduce. The exposition, even if
it is faulty, excites analogous schemas already existing*
in the listener's mind ; so that what takes place is not
genuine understanding, but a convergence of acquired
schemas of thought. In the case of stories, this con-
vergence is not possible, and the schemas brought into
play are usually divergent.
We need not recall examples of these divergent
schemas, which were sufficiently illustrated by the
accounts of reproducers given in sections 2 and 4.
We shall simply quote one or two examples of schemas
of purely verbal origin.
After hearing one of Gio's versions, Ri (8 years old),
tells the story of Niobe as follows: " Once upon a time,
there was a lady who was called Va'ika. She had twelve
sons. A fairy only had one. Once, one day, her son made
a stain on the stone. His mother cried for five years. It
[the stain, as Ri afterwards informed us] made a rock,
and her tears made a stream which is still running to-day"
The idea of the stain (in French: 'tache') arose in
Ri's mind when Gio pronounced the words: " The son
of the fairy fastened (French : attache J ) her to the stone"
The alliteration ' tache-attache * is thus sufficient to
build up a new structure in Ri's mind, viz., that the
mother cried because of the stain which made the rock.
Now we think in sentences, not in words ; it is there-
fore not only the single word * fastened' that has been
misunderstood, but the latter part of the story which
has been completely altered.
Herb (6 years old) tells the story of the four swans,
after having heard it told by Met (see 5) : " There was
once a queen and a king and then four children, one girl
and three boys ; and then there was all the children dressed
in white. Their father and mother looked for them and
they found them by the sea shore. He said to the wicked
queen [ = the fairy]: Are those children yours? 1 * The
wicked queen said : * No, they are not yours. ' "
122 STUDIES IN CHILD LOGIC
Here again, it looks as though it were only the words
' turned into swans' that had been altered ( = dressed
in white). But there is more in it than this. The
idea of a disguise has appreciably altered the end of
the story. Instead of thinking of a metamorphosis of
children into animals who go away to a distant country,
Herb has turned the story into one of simple kidnapping.
The fairy has disguised the children in order to keep
them, and the parents have failed to find them or recog-
nize them because of the disguise.
The process of alteration is clear. Owing to one
syllable or one word being imperfectly understood a
whole schema arises in the mind of the reproducer,
which alters and obscures the rest of the story. This
schema is due to the fact that ego-centric thought
is, as we saw in Chapter I, essentially unanalytical.
The result is that it ignores isolated words and
deals with whole sentences, understanding them or
altering them as they stand without analysing them.
This phenomenon is, moreover, a very general one
in the verbal intelligence of the child, and will be
studied in the next chapter under the name of Verbal
Syncretism.
Finally, to what extent, we may ask, does the repro-
ducer understand the explainer's manner of expressing
causality? We have seen that a causal connexion is
generally replaced by a simple relation of juxtaposition.
Is this juxtaposition understood by the reproducer as
a relation of causality? This is the question we have to
answer. Here are a few results obtained from children
between 7 and 8 in connexion with the question of the
tap. Points 4, 6, 7 and 9 were taken down separately,
as being exclusively concerned with causality (4= when
the handle is horizontal the canal is open ; 6 = the water
runs because the canal is open ; 7 and 9, the contrary).
The four coefficients were calculated with reference only
to these four points. In this way a measure of the
understanding of causality is obtained, whether the ex-
UNDERSTANDING BETWEEN CHILDREN 123
plainer has expressed the causal relations by juxta-
position or not.
a = 0*48 y = o'97 /3 = o'68 = 0^52
The results are practically the same for children be-
tween 6 and 7 (a = 0-49, ,8 = 0*68).
The significance of these figures is clear. For one
thing, causality is well understood by the explainer
y = o*97 is an excellent coefficient, above the average of
the explainer's mechanical explanations, which is be-
tween 0-93 and 0-80 ; but it is badly expressed (3 = 0*52).
This last circumstance emphasizes the prevalence of the
phenomenon of juxtaposition. The result of this faulty
verbal expression is just what might be expected : the
reproducer is very unsuccessful in understanding the
explainer (a = 0*48, as compared to 0*68 for mechanical
explanations between 7 and 8, and 0*56 between 6 and 7).
Causal relations are therefore imperfectly understood by
children, whether they are expressed by juxtaposition
or not
What part exactly does the phenomenon of juxta-
position play in this imperfect understanding? In order
to solve this problem, we took down separately all
definite cases of juxtaposition in the explanation of the
tap and the syringe or in the stories, i.e., all the cases
where a causal relation is expressed simply by juxta-
position (with or without * and ') of the connected proposi-
tions. We then sought to establish in what proportion
of cases this relation of juxtaposition was understood
as a causal relation. Take, for instance, this phrase of
an explainer: u The handle is like that, and the little
pipe is closed/* In how many cases will the reproducer
understand (whether he expresses it or not is of no
consequence ; the child's degree of understanding is
always tested by supplementary questions) that the little
pipe is closed because the handle has been turned? Out
of some forty clear cases of relation by juxtaposition,
only a quarter were understood, z>., in only a quarter
124 STUDIES IN CHILD LOGIC
of the cases did the listener grasp the causal relation.
This is a crucial point. Relation by juxtaposition is
therefore an ego-centric mode of conceiving causality ;
it can never be used by the child as a means of adapted
expression.
Are these results peculiar to our particular method of
experiment, or do they correspond to something that
can be observed in the spontaneous life of the child?
It will be sufficient to recall the results of the last two
chapters to realize that this imperfect understanding
of causality among children is a perfectly spontaneous
thing; children do not talk about causality among them-
selves before the age of 7 or 8. Such explanations as
they give each other are rare and factual. The questions
they ask one another contain very few 'whys/ and
hardly any requests for causal explanation. Before
the age of 7 or 8, causality is the object of ego-
centric reflexion only. This reflexion gives rise to
the well-known questions of child to adult, but the
schemas implied by these questions or produced by
the answers of adults remain incommunicable, and there-
fore invested with all the characteristics of ego-centric
thought.
7. CONCLUSION. THE QUESTION OF STAGES AND
THE EFFORT TOWARDS OBJECTIVITY IN THE ACCOUNTS
GIVEN BY CHILDREN TO ONE ANOTHER. One last ques-
tion which can be asked in connexion with our experi-
ments is this. To what extent do children try to be
objective when they talk to one another? It should be
noted in the first place that the objectivity of thought
is closely bound up with its communicability. It is in
ego-centric thought that we give rein to our imagination.
When we think socially, we are far more obedient to
the " imperative of truth." When, therefore, does this
effort towards objectivity in explanation or story between
one child and another first make its appearance? If we
can fix this moment, we shall at the same stroke be able
to determine that critical period when understanding
UNDERSTANDING BETWEEN CHILDREN 125
between children first comes to be desired and therefore
possible.
In this connexion, our material supplies a fairly
definite answer. On the one hand, it is only from the
age of 7 or 8 that there can be any talk of genuine
understanding between children. Till then, the ego-
centric factors of verbal expression (elliptical style,
indeterminate pronouns, etc.) and of understanding
itself, as well as the derivative factors (such as lack
of order, in the accounts given, juxtaposition, etc.) are
all too important to allow of any genuine understanding
between children. Between the ages of 7 and 8 these
factors become less active, and some of them (lack of
order) even disappear. On the other hand, there exists
between children of 6 and 7 and those of 7 and 8 a
fundamental difference as regards their efforts to be
objective. This convergence of two independent pheno-
mena is certainly not fortuitous, and it has enabled us
to place the beginnings of verbal understanding between
children, approximately between the ages of 7 and 8.
We have often wondered during our experiments to
what extent the explainers, when they made their
expositions, and the reproducers, when they repeated
what they had heard, really tried to give a true account,
and to what extent they simply believed themselves to
be doing so. It often happens, for instance, that the
explainer, not having the end of his story or his ex-
planation in rnind, seems to invent the end, or at least
to alter it, as though he were making it up. It also
happens that the reproducer seems to give up the
attempt faithfully to reproduce what he has heard, and
rather than repeat what he has not understood, embarks
upon some story of his own. In this respect there is a
great difference between the two groups of children.
In the case of boys from 7 to 8 it is quite safe to say
that both explainer and reproducer try to give a faithful
account of what they have heard. They have a sense
of what is meant by the faithful rendering of a story or
126 STUDIES IN CHILD LOGIC
the truth of an explanation. When they invent, which
happens rarely, they know it, and willingly own to it.
This shows up all the more clearly because of the
marked difference which exists in this respect between
stories and mechanical explanations. Mechanical ex-
planation arouses a more lively interest ; both explainer
and reproducer try to understand it ; the results there-
fore are better. Stories arouse less interest ; the
explainer tells them with less enthusiasm ; even when
he is faithful, which is usually the case, the effort to be
objective is greater.
The younger children, on the other hand, find it far
more difficult to distinguish between romancing and a
faithful rendering. When the child has forgotten some-
thing or understood it imperfectly, he fills in the gap
by inventing in all good faith. If he is questioned on
what he has heard, he stops inventing, but left to himself
he will believe what he has made up. Romancing, or
conscious and deliberate invention, is thus connected
with an unconscious distortion of the facts by a whole
chain of intervening stages.
This distinction between our two groups of children
is one of very great importance. It proves that the
effort to understand other people and to communicate
one's thought objectively does not appear in children
before the age of about 7 or /|. It is not because the
smaller children were romancing that they failed to
understand each other in our experiments. In cases
where there was no invention the same phenomenon of
faulty understanding was observed to take place. On
the contrary, it is because he is still ego-centric and
feels no desire either to communicate with others or
to understand them that the child is able to invent as
the spirit moves him, and to make so light of the
objectivity of his utterances.
CHAPTER IV
SOME PECULIARITIES OF VERBAL
UNDERSTANDING IN THE CHILD
BETWEEN THE AGES OF NINE AND
ELEVEN. 1
WE laid stress in the first chapters on the ego-centric
nature of child thought, and we tried to point out the
importance which this phenomenon might assume in
the use of reasoning in general. We tried in particu-
lar to bring out the three following points in which
ego-centric differs from socialized thought. i It is
non-discursive, and goes straight from premises to
conclusion in a single intuitive act, without any of the
intervening steps of deduction. This happens even
when thought is expressed verbally ; whereas in the
adult only invention has this intuitive character, ex-
position being deductive in differing degrees. 2 It
makes use of schemas of imagery, and 3 of schemas
of analogy, both of which are extremely active in the
conduct of thought and yet extremely elusive because
incommunicable and arbitrary. These three features
characterize the very common phenomenon called the
syncretism of thought. This syncretism is generally
marked by a fourth characteristic to which \ve have
already drawn attention, vis n a certain measure of
belief and conviction, enabling the subject to dispense
very easily with any attempt at demonstration.
Now childish ego-centrism seems to us considerable
only up till about 7 or 8, the age at which the habits
of social thought are beginning to be formed. Up till
1 With the collaboration of Mile Alice Deslex.
127
128 STUDIES IN CHILD LOGIC
about 7f, therefore, all the child's thought, whether it
be purely verbal (verbal intelligence) or whether it bear
on direct observation (perceptive intelligence), will be
tainted with the consequences of ego-centrism, and of
syncretism in particular. After the age of 7 to 8, these
consequences of ego-centrism do not disappear im-
mediately, but remain crystallized in the most abstract
and inaccessible part of the mind, we mean the realm
of purely verbal thought. In this way, a child may
cease between the ages of 7 and ir to 12 to show any
signs of syncretism in his perceptive intelligence, i.e., in
those of his thoughts that are connected with immediate
observation (whether these are accompanied by language
or not), and yet retain very obvious traces of syncretism
in his verbal intelligence, i.e., in those of his thoughts
that are separate from immediate observation. This
syncretism, which appears only after the age of 7-8 will
be called Verbal Syncretism, and will alone concern us
in this chapter. *
It is not our intention to embark upon a compre-
hensive study of verbal syncretism, nor to make a
catalogue of the various forms which this phenomenon
assumes in children. We shall only analyse one fact
of experience which is connected with syncretism, and
which we came upon quite by chance while we were
making investigations with the object of standardizing
a test of understanding.
We occasionally make use at the Institut Rousseau
of a test of understanding which is very well suited to
the examination of schoolboys or of children from n
to 15. The subject is given a certain number of proverbs
such as : "Drunken once will get drunk again." " Little
streams make mighty rivers," etc. (10 proverbs at a time).
Then 12 sentences in no particular order, 10 of which
express severally and in a new form the same ideas
as were expressed in the proverbs. As, for example,
the sentence "It is difficult to break old habits" corre-
sponds to the proverb " Drunken once will get drunk
VERBAL UNDERSTANDING 129
again." The child is asked to read the proverbs and to
find the sentences which fit them,
Now we applied this test to children of 9, 10, and n,
and this is what happened.
In the majority of cases the children did not under-
stand the proverbs in the least ; but they thought they
had understood them, and asked for no supplementary-
explanation of their literal or hidden meaning. This is
a very common characteristic of verbalism, and as such
very interesting. It may be objected that it can be
accounted for by such scholastic habits as fear or
discipline, by false shame, by the suggestion of the
experiment. This may sometimes happen, but not in
the great majority of cases, where the child really
believes that he has understood. In these cases the
experiment only reproduces a phenomenon which is
well known in daily life. The child hears the remarks
of adults (whether they have been addressed to him or
not), and instead of interrupting to ask for explanations,
he instantly imagines that he has understood. Or else,
he tries to find out for himself, he incorporates what he
has heard to his own schemas, and straightway gives to all
the words a meaning which may be more or less constant
and precise, but is always categorical. This, however,
concerns us only indirectly; and it is not on the imperfect
understanding of proverbs that we wish to lay stress in
this chapter.
The second thing we noticed was this. The children
often found, sometimes without hesitation, sometimes
after feeling their way about a little, sentences which
corresponded with the proverbs they had failed to under-
stand, and which in the eyes of the subject really fulfilled
the condition of " meaning the same thing" as the
respective proverbs. The children therefore understood
the instructions and applied them in their own manner.
Naturally this correspondence between the proverb and
the sentence "meaning the same thing" contained ele-
ments so surprising and so absolutely incomprehensible
I
130 STUDIES IN CHILD LOGIC
that we were at first inclined to think that the children
were inventing. But this again does not concern us
directly. It is obvious that if the children imagine
they have understood the proverbs, they will have no
difficulty in rinding a corresponding sentence. The
fact that this correspondence is absurd to the logical
adult need not astonish us, and is not what we wish
to talk about. But how does this correspondence come
about? This is where we touch upon verbal syncretism.
A third point must be added to the other two. We
believe that we have shown this correspondence to be
due neither to chance nor merely to what is called
verbalism, z.e. 9 the automatic use of words devoid of
sense. On the contrary and it is only on the peculiar
nature of this correspondence that we shall lay stress
in this chapter we can discern in this activity of under-
standing and invention on the part of the child several
of those schemas of analogy, of those leaps to conclu-
sions which are the outstanding characteristics of verbal
syncretism. It is from this point of view that we deem
it useful to analyse this handful of experimental facts,
however insignificant they may seem to the reader at
first sight.
Our material has thus been collected by the most
deplorable method. But in science every opportunity
must be utilized, and it is well known that the residue
of experiments planned for a definite object is often more
interesting than the experiments themselves.
Even with these reservations we would never have
ventured to make the children look for the correspond-
ence between proverbs they did not understand and
sentences having the same meaning as these proverbs,
had it not been that each one of our subjects was able
to find the correct correspondence for at least 2 or 3
proverbs out of the total number (10, 20 or 30 according
to the experiments), and thus proved that he was able
to carry out the instructions, and to understand what a
proverb is. Once we had entered upon these investiga-
VERBAL UNDERSTANDING 131
tions, moreover, the conviction grew upon us that in
daily life the child often hears phrases, thinks he under-
stands them, and assimilates in his own way, distorting
them as much as the proverbs of which we made use.
In this connexion the phenomena of verbal syncretism
have a general bearing upon the whole verbal under-
standing of the child, and are therefore well worth
studying. We only hope that in the circumstances we
shall not be blamed for the method we have adopted.
It is not a method at all. We have simply made ex-
periments "just to see." Our results contain only
suggestions, which are meant to be taken up and tested
by other methods.
i. VERBAL SYNCRETISM. We must first say a few
words by way of introduction to the subject of syn-
cretism, independently of the circumstances in which
we happen to have observed it.
Recent research on the nature of perception, par-
ticularly in connexion with tachistocopic reading, and
with the perception of forms, has led to the view that
objects are recognized and perceived by us, not because
we have analysed them and seen them in detail, but
because of " general forms" which are as much con-
structed by ourselves as given by the elements of the
perceived object, and which may be called the schema
or the gestaltqualitdt of these objects. For example, a
word passes through the tachistocope far too rapidly for
the letters to be distinguished separately. But one or
two of these letters and the general dimensions of the
word are perceived, and that is sufficient to ensure a
correct reading, Each word, therefore, has its own
( schema.' 1
M. Claparede, in a note on the perceptions of children,
has shown that these schemas are far more important
for the child than for us, since they develop long before
the perception of detail. For example, a child of 4
1 See Mach, The Analysis of Sensations. See Biihler, Die Gssta.lt-
wakrnehmungen^ Vol. I, 1913, p. 6.
i 3 2 STUDIES IN CHILD LOGIC
who did not know his letters and could not read music
managed to recognize the different songs in a book
from one day or one month to another, simply by their
titles and from the look of the pages. For him, the
general effect of each page constituted a special schema,
whereas to us, who perceive each word or even each
letter analytically, all the pages of a book are exactly
alike. Children therefore not only perceive by means
of general schemas, but these actually supplant the
perception of detail. Thus they correspond to a sort of
confused perception, different from and prior to that
which in us is the perception of complexity or of forms.
To this childish form of perception M. Claparede has
given the name of syncretistic perceptions? using the name
chosen by Renan to denote that first "wide and com-
prehensive but obscure and inaccurate " activity of the
spirit where " no distinction is made and things are
heaped one upon the other" (Renan). Syncretistic per-
ception therefore excludes analysis, but differs from our
general schemas in that it is richer and more confused
than they are. It was the existence of this syncretism
of perception which enabled Decroly to teach children
to read by letting them recognize words before letters,
thus following the natural course of development from
syncretism to the combination of analysis and synthesis,
and not from analysis to syncretism.
This movement of thought from the whole to the part
is a very general one. It will be remembered how this
point was emphasized by M. Bergson in his criticism of
associationism. " Association" he said, "is not the funda-
mental fact ; it is by dissociation that we begin, and the
tendency of every memory to gather others around it
can be explained by a natural return of the mind to the
undivided unity of perception." 2
Students of linguistics in particular are constantly
detecting this process in language, as when they show
1 Arch, de Psych., Vol. VII (1907), p. 195.
3 H. Bergson, Matisre et Mjmoire, Paris, nth ed. (1914), p. I So*
VERBAL UNDERSTANDING 133
how the sentence is always earlier than the word, or
when, like M. Bally, they analyse the phenomenon of
' lexicalization. ' They also point out the affiliation which
we shall have occasion to deal with later (Vol. II.)
between the phenomenon of syncretism and that of
juxtaposition. M. Hugo Schuchardt has recently
pointed out that not only is the word-sentence earlier
than the word, but also that the word is derived from
the juxtaposition of two sentences, which juxtaposition
then brings about the need for co-ordination and finally
for lexicalization.
From the point of view of the psychology of language
M. Lalande has shown what is the bearing of these
linguistic considerations upon the study of thought.
He recalls the observations of M. O. F. Cook, accord-
ing to which the Golahs of Liberia do not know that
their language is made up of words. Their unit of
consciousness is the sentence. Now these sentences,
like ours, contain a certain number of words, and
Europeans who learn the language can ascribe to these
words a constant meaning. But the Golahs have never
consciously realized their existence nor the constancy of
their meaning ; just like those children who can make
a correct use of certain difficult terms in their speech,
and are yet incapable of understanding these terms
taken by themselves. M. Lalande completed these
data by an examination of the spelling of illiterate
adults. These show a tendency to run together words
that should be separate (le courier va pase ma cherami),
or to divide single words (je fini en ten beras en bien
for) 1 in complete disregard of the meaning of the units
created in this way. This, however, does not prevent
these same persons from talking perfectly good French.
In a word, the line of development of language, as
of perception, is from the whole to the part, from
1 The correct French spelling would be, " le courier va passer ma chere
amie" (the post is just coming, my dear) and, "je finis en t'embrassant bien
fort " (I end with a loving kiss). [Translator's Note].
i 3 4 STUDIES IN CHILD LOGIC
syncretism to analysis, and not vice versa. If this is so,
then we must expect to find the same phenomenon of
syncretism in the understanding of language itself. The
phenomena emphasized by M. Cook and M. Lalande
are relevant to the conscious realization of words taken
as linguistic units and of words already understood in
their relation to the rest of the sentence. What happens
then, when, for example, the child is confronted with
sentences which he does not understand, either because
of the difficulty of the thought which is expressed, or
because of the difficulty of the words which are em-
ployed ? Will he begin with analysis, and try to under-
stand words or groups of words, taken separately, or
will the child's understanding move by general schemas
which themselves will give a meaning to the particular
terms in question ? In other words, is there a syncretism
of understanding just as there is a syncretism of per-
ception and of linguistic consciousness? This chapter
will be devoted to establishing the existence of such a
syncretism and to describing some of the principal
phenomena connected with it.
This syncretism of understanding must not be con-
fused with the phenomenon which we shall meet with
again presently (Chapter V, 3), and which we have
called elsewhere syncretism of reason or of explanation. 1
This is the name given to that process by which one
proposition calls forth another, or one cause an effect,
not because of any implication which has been logically
analysed nor because of any causal relation which has
been made explicit in all its detail (analysis of the
' how '), but once again because of some general schema,
which connects the two propositions or the two re-
presentations of phenomena. This schema is given
immediately, in an indistinct and general manner, so
that the two propositions or the two phenomena are
taken in a lump and regarded as an indissoluble whole.
1 Jean Piaget, " Essai sur la multiplication logique," etc., Journ. de Psych.,
1922, pp. 244, 258 et seq.
VERBAL UNDERSTANDING 135
Example: Bea (5 years old) "the moon doesn't fall
down, because it is very high up^ because there isn't any
sun> because if is very high up." The fact of the moon
not falling down, the fact of its being very high up,
and the fact of its shining when the sun is not there
constitute one solid whole, since these features are
always perceived together. The result is that the child
explains one of these features simply by an enumeration
of the others.
The syncretism of understanding and that of reason-
ing are of course dependent on one another. We
shall see how these two forms combine in connexion
with the phenomena which will be described in this
chapter.
Finally, we wish to recall, in connexion with
syncretism, the fine work done by M. Cousinet on
the ideas of children. 1
Under the name of " immediate analogy " M. Cousinet
has described one of the phenomena closely connected
with the syncretism of perception. According to him,
children who confuse two perceptions under a single
name have not previously compared them (a child does
not explicitly compare an owl and a cat before referring
to the former as ( miaou '), but they see the compared
objects as alike before making any inference. The
analogy is therefore not mediate but immediate, because
the subject "does not compare perceptions but per-
ceives comparisons. " If, therefore, says M. Cousinet,
children perceive different things as identical, it is
because each childish representation forms an " indis-
soluble lump," in other words because their perception
is syncretistic.
M. Cousinet's argument seems to us perfectly sound ;
but we believe that there is more than immediate
analogy" in the syncretism of understanding and
reasoning which we shall presently describe. Most of
1 Roger Cousinet, "Le rdle de Fanalogie dans les representations du
monde exterieur chez les enfants," Rev. Philos., Vol. LXIV. p. 159.
136 STUDIES IN CHILD LOGIC
M. Cousinet's examples are factual, and show evidence
of syncretism of perception or of conceptual representa-
tion only. They are perceptions added to perceptions.
This of course is the form in which syncretism first
presents itself, and we have no quarrel whatsoever with
M. Cousinet's highly suggestive exposition ; what we
do think however, is that the idea of syncretism is richer
than that of " immediate analogy." As we have just
been seeing, even in ( mediate ' operations like under-
standing or reasoning we can have syncretism, i.e., the
formation of general schemas, which bind the pro-
positions together and create implications, without ever
resorting to analysis. We therefore suggest our notion
of the syncretism of thought as being more compre-
hensive than the notions of syncretism of perception
and immediate analogy, and as containing them both
in the nature of special cases.
2. SYNCRETISM OF REASONING. In Geneva our
experiments were carried out on about twenty boys of
9 and fifteen girls of the same age, and at Lavey (in
Vaud) on a similar number of subjects between 8 and
ii. We wish to remind our readers that the tests
employed 1 were originally intended to measure the
degree of understanding in children between n and 16.
The children on whom we worked were below the level
required for most of the proverbs. In order, however,
that the experiment should not be absurd, we analysed
only the answers given by children who had been able
to discover and defend the correct correspondence for
at least one or two proverbs, and had thus proved their
capacity for carrying out the instructions necessary for
the experiment The number of correct answers for
the nine-year-olds fluctuated between i (2 cases) and
23 (i case).
Let us now submit to analysis some cases of the
1 For the list of proverbs used and for the ready-reckoner see M.
Clapar&de's new book, Comment diagnostiquer hs aptitudes chez les frotiers,
Paris, Flammarion.
VERBAL UNDERSTANDING 137
syncretism of reasoning". They will prepare our mind
imperceptibly for the mechanism of the syncretism of
understanding. Syncretism of reasoning may be said
to occur in the materials which we have collected when-
ever a proverb is compared to a corresponding sentence,
not because of any logical implication contained in the
text, but because of an implication built up in the child's
imagination by means of a general schema in which the
two propositions are united. For this syncretistic im-
plication to be seen in its purest form, the proverb and
the sentence chosen by the child should both be under-
stood by him. We are then faced with the spectacle of
two propositions, which taken separately might have
been quite well understood, being actually distorted by
syncretism, which then unites them by means of a
fictitious implication. In the cases where the separate
propositions are also misunderstood, we have the ad-
ditional phenomenon of ' syncretism of understanding '
which will be studied later on. The two cases are
always more or less mixed.
Here is a case of almost pure syncretistic reasoning :
Kauf (8; 8) G. (3/xo) 1 connects the proverb: " When the
cat's away the mice can play," with the following phrase:
" Some people get very excited but never do anything."
Kauf, who would understand the meaning of each of
these sentences if they were separate, yet declares that
they mean " the same thing" " Why do these sentences
mean the same thing? Because the words are about the
same What is meant by 'some people' . . . etc? //
means that some people get very excited, but afterwards they
do nothing, they are too tired. There are some people who get
excited. Ifs like when cats run after hens or chicks. They
come and rest in the shade and go to sleep. There are lots
of people who run about a great deal^ who get too excited.
Then afterwards they are worn out y and go to bed."
The mechanism of syncretistic implication shows
here very clearly. The proverb has been understood
1 This fraction signifies : 3 correct correspondences out of 10.
I 3 8 STUDIES IN CHILD LOGIC
verbally by Kauf. According to her it means : " The
cat runs after the mice."
The symbolic or ethical meaning of the proverb is
not specified by Kauf until she has found a corre-
sponding sentence. How does this correspondence or
implication arise? By a simple fusion of the two pro-
positions in one common schema. The words: " When
the cat's away" are fused with the words: u Some
people never do anything," and take on the meaning of:
"the cat has a rest and goes to sleep." The words " the
cat runs " are brought into connexion with " some people
get very excited." The result is that the two propositions
imply one another. This implication does not come about
analytically, through reflexion upon the given text, but
syncretistically, i.e., through simple projection of the pro-
verb into the sentence by a process of immediate fusion.
Thus there is no analysis of detail, but the formation of
a general schema. Such is implication by syncretism,
which will be found in all syncretistic reasoning, and
which consists in a general fusion of two propositions.
With Kauf of course this syncretism of reasoning is
connected with syncretism of understanding. This is
how the sentence comes to be verbally distorted in
favour of a general schema which actually contradicts
the proverb. Kauf has confined herself to adding a
corollary to the proverb, which was not given in the
text; she imagines that the cat is away, "to take a
rest." The meaning of the words, however, has not
been altered, whereas the actual words of the corre-
sponding sentence have been garbled : the word ' but '
has been taken in the sense of 'and then.' Thus the
understanding of this sentence is itself syncretistic, z"..,
it is a function of the general schema, whereas the under-
standing of the proverb existed before this schema.
Here is another example :
Mat (10 ; o) G. (2/10) connects the proverb "So
often goes the jug to water, that in the end it breaks"
with the sentence " As we grow older we grow better."
VERBAL UNDERSTANDING 139
Now the proverb has been understood verbally. For
Mat it means : u You go to the water so often that the jug
cracks i you go back once again and it breaks*" The corre-
sponding sentence is explained as follows : " The older
you get the better you get and the more obedient you become
Why do these two sentences mean the same thing?
Because the jug is not so hard because it is getting old, because
the bigger you grow, the better you are and you grow old."
The syncretism here Is of reasoning only, since
neither of the propositions has been altered in favour
of a common schema. Mat's reasoning, moreover,
seems rational enough except that it seems curious to
compare a jug that gets broken to a man who grows
old. It may be claimed that this absurdity is due to the
fact that a child of 10 cannot realize that the symbolism
of a proverb is exclusively ethical. This is undoubtedly
one of the factors at work, although at this age children
realize perfectly well that all proverbs are symbolic.
But this factor alone does not explain the child's power
to connect everything with everything else by means of
general schemas, and to compare a jug to a child simply
because both grow older.
It is obvious, moreover, that this last example is far
less syncretistic than the first, and that it approximates
to the simple judgment of analogy of the adult. This
gives us two extreme cases, between which there is a
whole series of fluctuating intermediate cases. We
give one of these by the same little girl, Mat :
" White dust will ne'er come out of sack of coal"
is compared to " Those who waste their time neglect
their business." The proverb is understood verbally:
' * / thought it meant that white dust could never come out of
a sack of coal, because coal is black " Why do these two
sentences mean the same thing? People who waste their
time dotft look after their children properly. They dorit wash
them> and they become as black as coal, and no white dust
comes out. Tell me a story which means the same thing
as : ' White dust will ne'er . . . etc.' " Once upon a time
there was a coal merchant who was white. He got black
and his wife said to him : * How disgusting to have a man
140 STUDIES IN CHILD LOGIC
like that.' And so he washed, and he couldn't get white,
his wife washed him and he couldn't get white. Coal can
never get white, and so he washed his skin, and he only got
blacker because the glove [washrag] was black."
It is quite clear in a case like this that the mechanism
of the subject's reasoning cannot be explained by judg-
ments of analogy affecting the details of the propositions.
Having once read the proverb, the child is ready to
attach to it any symbolic significance which chance may
reveal in the perusal of the corresponding sentences.
All that the proverb has left in his mind is a schema,
a general image, if one prefers that of the coal which
cannot become white. This is the schema which has
been projected, whole and unanalysed, into the first of
the corresponding sentences which was fitted to receive
it (those who waste their time . . . etc). Not that
this sentence really has anything in common with the
proverb ; it can simply be imagined to do so. Now
and this is where syncretism comes in the child who
fuses two heterogeneous sentences in this way does not
realize that he is doing anything artificial ; he thinks
that the two propositions united in this way involve
one another objectively, that they imply one another.
The corresponding sentence into which the proverb has
been projected actually reacts upon the latter, and when
the child is asked to tell a story illustrating the proverb,
the story will bear witness to this interpenetration. To
reason syncretistically is therefore to create between
these two propositions relations which are not objective.
This subjectivity of reasoning explains the use of
general schemas. If the schemas are general, it is
because they are added on to the propositions and are
not derived from them analytically. Syncretism is a
" subjective synthesis," whereas objective synthesis
presupposes analysis. 1
1 The reader will now realize why the ego-centrism of child thought
brings syncretisms in its wake. Ego-centrism is the denial of the objective
attitude, and consequently of logical analysis. It therefore gives rise to
subjective synthesis.
VERBAL UNDERSTANDING 141
Here are some more examples which bring out very
clearly this super-added element, which has not been
deduced by logical analysis, but imagined by subjective
comparison.
Nove (i2;ii) (3/12) compares " Filing can turn a
stake into a needle" to " Those who waste their time
neglect their business," "because by filing, that means
the more you file it (a stake) the smaller it gets. People
who don't know what to do with their time file, and those
who neglect their business turn a stake into a needle; it
gets smaller and smaller , and you don't kuow what has
become of the stake [therefore it has been neglected]."
Peril (io;6) (7/10) identifies "The cowl does not
make the monk" with " Some people get very excited
but never do anything." Because : "Even if you make
a cowl, the monk isrit in it, the cowl cannot talk" This
comment seems more like a simple illustration which
Peril does not yet take literally. But he goes on :
' * Because people who get very excited may get excited but
they do nothing^ because the cowl does not make the monk,
people who get very excited do not make anything either -," 1
Here the identification is more serious : the empty cowl
is compared to a man who is excited ! The words " do
not make " take on a more and more concrete signific-
ance in the child's mind. " Tell us a story which
means the same thing as * The cowl does not make the
monk.' There was once a dressmaker who was making a
dress for a person, and while she was making the dress
this lady suddenly died. The dressmaker thought she could
make everything and that the dress would take the place of
everything^ but she soon saw that the dress would not do
instead of the dead lady " In this way the corresponding
sentence and the proverb gradually melt into one
another. They do so because the words "make the
monk" gradually call up the image of " getting excited
so as to represent a monk," and because the words
" never do anything " take on the meaning of "do not
succeed in replacing the monk." Thus are the two
propositions completely identified, thanks to a purely
subjective schema.
Xy (12 years old): "Whoever trusts in the help of
1 In the French the verbs * do * and * make ' are throughout represented
by the single verb * faire * [Translator's note].
142 STUDIES IN CHILD LOGIC
others risks being left without support. " " He who
sows thorns must not go unshod," because " Whoever
trusts in the support of others must have a support^ and
whoever walks on thorns must have shoes"
There is no need to multiply examples; we shall
meet with plenty more. Let us rather consider how to
interpret this syncretism of childish reasoning. Two
hypotheses are possible. The first would explain the
observed facts by invoking the process of reasoning
from analogy, by what M. Cousinet has called "mediate
analogy." In reasoning of this kind the subject argues
from the resemblance of two elements, taken respectively
from two different objects, to the inclusive similarity of
the two objects compared. In the case of the proverb
and the corresponding sentence the child would take
his stand upon an observed resemblance between two
substantives or two negations, etc., and would then
conclude that the two phrases were identical, after he
had assimilated the remaining elements one to another
term by term. The second hypothesis would explain
the facts with the aid of general schemas, of an im-
mediate syncretistic fusion of the two propositions. As
he reads the proverb, the child makes for himself a
schema in which such things as the symbolic meaning
of the proverb, the mental imagery released by the
words, the rhythm of the sentence, the position of
the words in relation to conjunctions, negations, and
punctuation, all enter as elements. All these factors
would thus give rise to a unique schema, in which
would be condensed the various concrete images called
up by reading the proverb. Then comes the search for
the corresponding phrase. By now the schema is ready
to be projected whole into the words and ideas which
present themselves. Some of these may actually resist
it, but in so far as they can tolerate the schema, its very
existence will tend to warp the subject's understanding
of the corresponding phrase even before he has finished
reading it. The corresponding phrase Is assimilated.
VERBAL UNDERSTANDING 143
one might almost say digested, by the schema of the
proverb. Once the digestion has taken place, moreover,
a return shock takes place, and the proverb is digested
in its turn by the schema of the corresponding phrase.
This is where we have evidence that the syncretism of
reasoning is a wider and more dynamic process than
the syncretism of perception which M. Cousinet has
described under the name of il immediate analogy."
The difference between these two hypotheses is often
imperceptible, for the apparition of a whole general
schema may in many cases be released by a partial
analogy. But this much can be vouched for in our
children's answers : the presence of such schemas can
always be detected. As for partial analogies, they are
sometimes the outcome of general schemas, and they
sometimes precede them. The manner in which the
child assimilates the two phrases is therefore in no way
analytical or deductive. When Kauf compares " When
the cat's away, the mice can play " to the corresponding
phrase which we mentioned, she justifies her comparison
by saying : " because the words are more or less the same"
Now the two phrases have not a word, not even a
synonym in common. " Away " is assimilated to "get
very excited," but this is an assimilation of schemas,
not an analysis of detail, because the child imagines
that the cat has gone away to rest after having been
very excited. Is it then the phrase "[the mice can]
play" which has been brought into connexion with
"get very excited?" But this comparison is only
possible on the basis of a general schema. Similarly,
when Mat wishes to justify her assimilation of "So
often goes the jug to water . . . etc." to the phrase
which we quoted, she tells us that the two propositions
have two similar words, "big and old" But it is she
who has put the word 'old' into the interpretation of
the proverb " because it [the jug] gets old" Here again,
the analogy of detail appears after that of the general
schema, or at least as its function. Even if we admit
144 STUDIES IN CHILD LOGIC
that this analogy of detail appeared first, and was the
occasion for the formation of a general schema, this
would not be sufficient to explain the nature of the
schema : " The jug becomes less hard as it grows older,
just as the man becomes wiser (or better) as he grows
older." It is obvious that the analogy of detail and
the general schema are given together, and that there
has been not inference from the part to the whole, but
immediate fusion or assimilation. Besides, we have
seen that very often no analogy of detail can account
for the syncretism in question. In Mat's other example
(" From sack of coal . . . ") there is not a single verbal
analogy between the two proposition that have been
brought together. This is equally clear in the case of
Peril. The words "do not make the monk" take on a
more and more concrete and vivid meaning as the two
phrases are fused together. It is therefore no analogy
between the words "make the monk" and the words
"never do anything" which has allowed the child to
make this assimilation, but it is the progressive assimi-
lation which has fortified the analogy.
In conclusion it must be said that there is a mutual
dependence between the formation of general schemas
and that of analogies of detail. Analogies of detail
make the formation of general schemas possible, but
are not sufficient for their formation. Conversely,
general schemas give rise to analogies of detail, but
are likewise not sufficient for their formation.
Syncretism of reasoning is, therefore, in the first
instance the assimilation of two propositions in virtue
of the fact that they have a general schema in common,
that they both, willy nilly, form part of the same whole.
A enters into the same schema as B, therefore A implies
B. This 'implication' may appear in the form of
an identification, as in the present experiments, where
the child is asked to find two phrases that mean "the
same thing." It can also take the form of implication
properly so called, or of a < because/ as in the cases
VERBAL UNDERSTANDING 145
which one of us previously published as examples of
the syncretism of reasoning. 1 In the example which
we recalled a little way back, "the moon doesn't fall
down because there is no sun, because it is very high
up/' the features " doesn't fall down," "the sun goes
out when the moon appears," and " the moon Is very
high up " form one single schema, because they char-
acterize the moon. Now it is quite enough for this
schema to be present in the child's mind for him to
say: the moon doesn't fall down because . . . etc. Here
the schema brings about a definite implication.
3. THE NEED FOR JUSTIFICATION AT ANY PRICE.
From the frequent occurrence of this pseudo-logical or
pseudo-causal * because,' we can draw the conclusion
that childish thought and ego-centric thought in general
are perpetually determined by a need for justification at
all costs. This logical or pre-logical law has a deep
significance, for it is probably owing to its existence
that the idea of chance is absent from the mentality of
the child. "Every event can be accounted for by its
surroundings," or again, "Everything Is connected
with everything else, nothing happens by chance "
such might be the tenets of this creed. The, for us,
fortuitous concurrence of two natural phenomena or of
two remarks in conversation is not due to chance; It
can be accounted for in some way which the child will
invent as best he can. With regard to reflexion on
natural phenomena, we shall see many examples of this
law in connexion with the * whys' of children (Chapter
V, 2). A large number of these questions are put as
though the child completely excluded chance from the
course of events. A few examples of the same phe-
nomenon in connexion with verbal intelligence have
been cited by one of us under the somewhat equivocal
heading of the "principle of reason." When a child
is asked the reason for something, and does not know,
he will always and at any cost invent an answer, thus
1 Article mentioned in I*
K
146 STUDIES IN CHILD LOGIC
testifying to this particular desire to establish connexions
between the most heterogeneous objects. For instance,
a child is told as a test of his powers of reasoning, to
put a slip of blue paper into a box if he finds one penny
in it, a slip of white paper if he finds two pennies, etc.
The subject puts in the white slip at random. " Why
must you put in the white slip? Because white is the
same colour as a penny " or " Because its colour shines (like
nickel) 5 ' etc. 1 This example shows that arbitrary in-
structions do not satisfy a child. The result is that he
will always find a justification for everything, which to
us is simply * given ' without any reason, which is
simply ' assumed.'
Now these facts, which we have pointed out without
explaining them, are constantly to be met with in con-
nexion with the proverbs ; the child always justifies
the most unexpected combinations. Here are a few
examples in which syncretism brings about these justi-
fications at any price :
Witt (10 years old): "Qui s'excuse s'accuse" =
" Those who are too kind-hearted have everything
taken away from them in the end" " because the other
person took something away^ and so he excused himself"
And (9 ; 6) : Same proverb = " Whoever goes to sleep
late will wake up late " because he excuses himself because
he got up late [for having got up late].
But (8 ; 10) : " Drunken once will get drunk again "
= u By pleasing some we displease others" "because
when someone is drinking ', you go and disturb him"
Hane (953): " White dust will ne'er come out of
sack of coal" = " We must work to live" "because
money is needed to buy coal"
EC (951): " By wielding his hammer a blacksmith
learns his trade " * " Men should be rewarded or punished
according to what they have done " "because if we learn
our trade properly, we are rewarded^ and if we don't^ we
are punished" and "The sheep will always be shorn"
= " By practising a thing very often we learn to do it
well" because "by much practice of shearing sheep we
de Psych. , 1922 (Vol. XIX), p. 249.
VERBAL UNDERSTANDING 147
learn to shear well enough to do others, when we have
others."
Again EC: "The flies buzzing round the horses do
not help on the coach "= u It is difficult to correct a
fault that has become a habit" "because the flies are
always settling on the horses, and they gradually get into
the habit and then afterwards it is difficult for them to cure
themselves"' And " To each man according to his
works " = " Whoever goes to sleep late will wake up
late " " because if you have something to learn you go to
bed late. His worlss, that means ours, and so we wake
up late. Our works are the things we have to do, you have
things to do and you have to go to bed late so as to know
them." This last example brings in a more definite
general schema than the preceding ones.
Xy (12 ; 11) : "To every bird his own nest is beauti-
ful " = "Insignificant causes may have terrible con-
sequences " because "a bird builds his nest very carefully r ,
whereas if you do things carelessly ', it may have terrible
consequences" and "While you are idling, the joint is
burning " = " People who are' too busy correcting the
faults of others, are not always the most blameless
themselves" because i(t when one is too busy correcting
other people's faults, one leaves the joint to burn."
The mechanism of these justifications is plain. They
are cases of syncretism, in which the general schema is
reduced to the minimum, and amounts to what M.
Cousinet has called "immediate analogy."
The faculty for justification at any price which we see
in children Is thus a consequence of syncretism. Syn-
cretism, which is the negation of analysis, calls forth
this effort by which every new perception is connected
somehow or other with what immediately precedes it.
The connexion is sometimes complex and presents itself
in the form of a general schema, in which the old fits
term by term into the new, sometimes it is simpler and
more immediate, and will supply us with cases, like those
which we are considering, of justification at any price.
It may still be objected that these justifications can be
sufficiently accounted for by the notion of immediate
analogy, without bringing in that of syncretism. We
148 STUDIES IN CHILD LOGIC
believe, on the contrary, that If It were not for his
syncretistic habits of thought, if it were not for the
fact that he can perceive nothing but general schemas
which engender this perpetual conviction that every-
thing Is connected with everything else, the child
would never produce these arbitrary justifications in
such abundance. It is obvious, moreover, from the
examples we have given that between syncretism
properly so-called and the other cases there is every
kind of intermediate variety. Here again, it is thanks
to the general schema under formation that the analogy
of detail is brought into being.
There is therefore in the childish imagination an
astonishing capacity for answering any question and
disposing of any difficulty with some unexpected reason
or hypothesis. For the child, there is no * why ' that
does not admit of an answer. A child can always say :
" I don't know" in order to get rid of you ; it is only
very late (between n and 12) that he will say : " One
cannot know." It may be suggested that these justifi-
cations are given out of amour-propre, so as not to
remain a quia, etc. But this does not explain the
wealth and unexpectedness of the hypotheses brought
forward ; these recall the exuberance of symbolist
mystics or of interpretational maniacs far rather than
the tendency shown by adults (<?."., examination candi-
dates) to hedge, when they are taken off their guard.
If anyone disputes the existence of the desire to justify
at any price, he will have to explain the affiliation
between the justifications of children and those of
interpretational maniacs in the first stages of the
disease. The peculiarities which we have mentioned
are very similar to those which Dr Dromard attributes
to interpretational maniacs 1 imaginary reasoning^ in
which every possibility becomes a probability or a
certainty; diffusion of interpretation ; i.e., that process of
1 G. Dromard. " Le de"lire d' interpretation/' Jo-urn* de Psych., Vol. VIII
(1911), p. 290 and 406.
VERBAL UNDERSTANDING 149
"linking- up, in virtue of which one interpretation is
called forth by an earlier one, and rests in its turn upon
one subsequent to it." ("I graft one thing on to
another, and in this way I gradually build up a
scaffolding of the whole, " Patient G) ; radiation, i.e., "the
fortuitous and unexpected production of innumerable
interpretations gravitating at a certain distance around
the main idea, which represents the centre and rallying
point of every part of the system " ; symbolism, or the
tendency to find in every event and every sentence a
hidden meaning of greater depth than that which is
apparent, etc. M. Dromard is therefore right in con-
cluding that " in their manner of thinking, of perceiving
and of reasoning interpretational maniacs recall some
of the essential traits of primitive and of child thought." 1
The hypothesis which attributes the exuberance of
these childish interpretations to mere invention is one
which occurs very readily to the mind in connexion
with the various facts recorded in this chapter ; we shall
show by and by what it amounts to. The objection
should also be put aside according to which the inter-
pretative character of children's justifications is due to
the fact that we experimented with proverbs. All the
phenomena with which we are dealing in this chapter
are embodied in the reflexions of children as observed
in ordinary life. Many people, for instance, have studied
the spontaneous etymology which children practise, or
their astonishing propensity for verbalism, i.e., the imag-
inative interpretation of imperfectly understood words ;
and both these phenomena show the child's facility in
satisfying his mind by means of arbitrary justifications.
Moreover, as we have repeatedly pointed out, the child
does not know at the beginning of the experiment that
the proverb has a hidden meaning. All we do is to
remind him that a proverb is "a sentence that means
something," and to ask him to find a sentence which
"means the same thing." If there were not in the
1 Ibid., p. 416.
150 STUDIES IN CHILD LOGIC
child a spontaneous tendency towards justification at
any price, and towards interpretative symbolism, he
would never, within the limitations of our experiments,
have manifested the phenomena which we have been
examining.
We can therefore conclude that the desire for justifica-
tion at any price is a universal law of verbal intelligence
in the child, and that this law itself is derived from the
syncretistic nature of childish reasoning. The fact that
for the syncretistic point of view everything is related,
everything is connected to everything else, everything
is perceived through a network of general schemas built
up of imagery, of analogies of detail and of contingent
circumstances, makes it quite natural that the idea of
the accidental or the arbitrary should not exist for the
syncretistic mentality, and that consequently a reason
should be found for everything. On the other hand,
syncretism is the outcome of childish ego-centrism, since
it is ego-centric habits of thought that induce the child
to fly from analysis and to be satisfied with general
schemas of an individual and arbitrary character. We
can now understand why it is that the justifications of
children, rooted as they are in syncretism, have the
character of subjective and even of pathological inter-
pretations due to a regression to a primitive mode of
thinking.
4. SYNCRETISM OF UNDERSTANDING. Up till now
we have had to do only with those children who more
or less understood both of the two sentences to be
compared. So far as the corresponding sentences are
concerned, there can be no doubt about this. With
regard to the proverbs, we may say that they have been
verbally understood, z>., that in reading them the child
has had a concrete idea of their meaning, missing only
their moral significance. Yet they all had the feeling
that each proverb had a symbolic sense, although we
never laid any stress on this point. Given these con-
ditions of adequate understanding, we believe that the
VERBAL UNDERSTANDING 151
phenomena which we described a little way back belong
quite definitely to syncretistic reasoning.
How does this phenomenon of syncretism originate?
So far, we have looked upon the faculty for forming
general schemas as given, and as the outcome of non-
analytical habits of thought derived from ego-centrism.
We must now examine the mechanism more closely,
and pass to the study of syncretistic understanding.
In view of the phenomena studied in the first chapters
of this book, we can take it that when a child listens to
someone else talking, his ego-centrism induces him to
believe that he understands everything, and prevents
him from understanding word for word the terms and
propositions that he hears. Thus, instead of analysing
what he hears in detail, he reasons about it as a whole.
He makes no attempt to adapt himself to the other
person, and it is this lack of adaptation which causes
him to think in general schemas. In this sense, ego-
centrism may be said to be contrary to analysis. Now,
a very easy way of studying the mechanism of the
formation of these syncretistic schemas is to note in our
experiments when one of the words in the proverb or in
the corresponding phrase is unknown to the child.
Will he be interested in this word, as would be a mind
free from ego-centrism, which tended to adapt itself
to the interlocutor's point of view, and will he try to
analyse this word before advancing another step in his
reasoning? Or will he assume that the word is known
to him, and then go on with his thinking as though no
difficulty were present at all ? We shall see that ego-
centric habits of thought are the strongest, and that
the child reasons as though he had not listened to the
interlocutor, and as though he understood everything.
The result is that the unknown word is assimilated as
a function of the general schema of the phrase or of the
two phrases. Syncretistic understanding consists pre-
cisely in this, that the whole is understood before the
parts are analysed, and that understanding of the details
152 STUDIES IN CHILD LOGIC
takes place rightly or wrongly only as a function
of the general schema. It is therefore in syncretistic
understanding that -we shall find the connecting link
between the ego-centric habits of thought already known
to us through the last three chapters and the syncretism
to which they give rise.
If we wish to get some idea of this syncretism of
understanding in the child, we need only think of the
way in which persons gifted with intuition translate a
language with which they are unfamiliar, or understand
difficult propositions in their own language. They will
often understand the general trend of a page written
in a foreign tongue, or of a page of philosophy, for
example, without understanding all the words or all
the details of the exposition. A schema of understand-
ing has been constructed, relatively correct (as appears
from the more complete understanding subsequently
obtained), but resting only on a few points which have
been spontaneously related. In these cases such a
schema precedes analytical understanding.
Now, this is the method used by the child. He lets
all the difficult words in a given phrase slip by, then he
connects the familiar words into a general schema,
which subsequently enables him to interpret the words
not originally understood. This syncretistic . method
may of course give rise to considerable mistakes, some
of which we shall presently examine ; but we believe it
to be the most economical in the long run, and one
which eventually leads the child to an accurate under-
standing of things by a gradual process of approximation
and selection.
Here is an example of this method in connexion with
our proverbs :
Vau (iojo) identifies "To each according to his
works" with " Some people get very excited, and then
never do anything," Now he does not know the word
" according." But from the first he thinks that he
understands it, and in the following manner. He
VERBAL UNDERSTANDING 153
connects these two propositions "because they are just
about the same thing" \ one sentence means: " every one
does his works ^ each has his works" and the other : * 4 Each
wants to do something but they never do anything" Except
for the antithesis, therefore, the schema is similar for
both phrases. What then is meant by ' according to ' ?
It means "Let them come. Let each one come to his
work" Half an hour later we ask Vaii to repeat the two
sentences by heart. He reproduces the first as follows :
"Each according to work" The expression " according
to " has definitely taken on the meaning of * coming ' ;
Vau reproduces the corresponding sentence as follows :
"Some people come for nothing" Why do the two
phrases mean the same thing? Because some people went,
but never did anything. There [in the proverb] they went
but they did something "
This case shows us very clearly a child ignorant of
a word, but unaware of his own ignorance. The result
is, that the unknown word is interpreted in function of
the general schema of the two phrases compared. Vau
made no attempt to analyse the details of the phrases
in so far as they were incomprehensible. He decided
that they meant the same thing, and then proceeded
to interpret the various terms in function of the general
schema which had been formed independently of the
unknown word. If, therefore, we bring this mechanism
of word understanding into relation with syncretism,
it is because in this type of understanding the mind
goes from the whole to the part exactly in the same
way as it does in primitive perception.
Here are some more examples :
Kauf (8 ; 8) G. assimilates " The sheep will always be
shorn " to the following phrase: " Small people may
be of great worth," without knowing the word * worth/
and comes to the conclusion that the latter phrase means:
"It means that they may get bigger later on" Once
again the unknown word is used as a function of the
general schema. The sheep, says Kauf, will always
be shorn, because as he grows older he grows bigger.
The two sentences therefore mean the same thing
"because sheep can get bigger as they grow older y sheep
154 STUDIES IN CHILD LOGIC
are small when they are young, when people are young they
are small, and as they grow older they get bigger. " [= They
are of great worth in the sense that has just been
expounded.]
The same corresponding sentence is assimilated by
Don (9 years old) to the proverb " The flies buzzing
round the horses do not help on the coach/' from which
Don concludes that ' worth' means "something which
is large" or " a large number of flies " Here, once more,
it is the schema that gives the meaning to the unknown
word.
We need not dwell any longer on these facts which
are those of current observation. It is they that explain
the phenomenon of verbalism. If children have so
much facility in using an unknown word, without
noticing that they do not understand it, it is not because
they think they can define it, for out of its context the
word means nothing to them. But the first time they met
with it, the whole context gave to the word a mean-
ing that sufficed, owing to the syncretistic connexion
between all the terms of the context, and owing to the
pseudo-logical justifications always ready to emerge.
We can see at the same time how the syncretism of
understanding explains that of reasoning, and bridges
the gulf between ego-centric thought and the phenomena
described in the last few sections. This is more or less
how things happen. When the child hears people talk,
he makes an effort, not so much to adapt himself and
share the point of view of the other person as to assimi-
late everything he hears to ^.^Qwtj^oint^fview and
to his own^^^tQ^^Jjx^m^tjon. An unknown word
therefore seems to him less unknown than it would if
he really tried to adapt himself to the other person.
On the contrary, the word melts into the immediate
context which the child feels he has quite sufficiently
understood. Words which are too unfamiliar never call
forth any analysis. Perception and understanding are
thus syncretized, because they are unanalysed, and
VERBAL UNDERSTANDING 155
they are unanalysed because they are unadapted. From
this syncretistic i reception, 5 so to speak (perception or
comprehension) to syncretistic reasoning there is only
a step, the step of conscious realization. Instead of
passively noting" that such-and-such a phrase "goes
with" (feeling of agreement) such-and-such another,
or that the fact of the moon not falling down u goes
with " the fact of its being very high up, the child can
ask himself, or be asked by someone * why * this is so.
He will then create implications, or discover various
justifications, which will simply render more explicit
this sense of ' agreement' which he felt in all things.
The illogical character of childish * whys,' or the absence
of the notion of chance revealed in primitive 'whys'
are therefore due to syncretism of understanding and of
perception itself, and this in its turn is due to the lack
of adaptation which accompanies ego-centrism.
We can now sum up in a few words the problem
discussed in 2 : Do syncretistic relations between pro-
positions arise out of analogies of detail, or vice versa?
This question, which we have already answered by
saying that the two are mutually dependent, can also be
stated in connexion with syncretistic understanding and
perception. Does the child understand the sentence as
a function of the words, the whole as function of the
parts, or vice versa? These questions are idle in con-
nexion with easy sentences or familiar objects, but they
become interesting as soon as there is adaptation to new
objects of thought.
With regard to perception, the subject will have to
be pursued later, but this much it seems possible to
establish. In the case quoted by M. Claparede, for
instance, the question may be raised whether the child
who recognized one page of music amongst many
others was guided (as the hypothesis of syncretism
would suggest) by the general effect of the page, or
by some particular detail (such as the ending of a line
or the portion of the page blackened with an agglomera-
156 STUDIES IN CHILD LOGIC
tion of notes). Now there is solidarity between the
general effect of the page and the separate effect of
the isolated details. If there is a general effect, it is
because over and above the vague mass which serves
as a background, certain details are picked out at
random and are specially noticed. It is because of
these distinctive details that there is a whole, and vice
versa. The proof that this is neither a sophism nor
a truism is that we adults, who are accustomed to
analyse each group of notes and each word, no longer
see either the general effect nor the outstanding details.
If the distinguishing details are no longer present, it
must be because there is no longer any general effect,
and vice versa. As soon as we half close our eyes,
however, certain groups of notes and certain words
stand out, and it is by means of them that the page
takes on, now a certain general physiognomy, now its
opposite, in a continuous rhythm of alternation. There
is the same solidarity between the distinctive details
and the general effect in children's drawings. This is
why a child, if he wishes to express the figure of a
man, is content to put down together a few details,
insignificant or essential (a head, a button, legs, a
navel, etc. at random) which we would have chosen
in quite another manner, our perception not being
syncretistic to the same degree.
It is therefore no truism to say that the general
schema and the analysis of details are mutually de-
pendent. The two elements are quite distinct, and the
one calls forth the other according to a rhythm which
it is quite easy to observe.
Exactly the same thing happens in syncretistic under-
standing. In some cases it seems as though only the
schema mattered, and that the words were only sub-
sequently understood. Here is an example :
Peril (10 ; 6) identifies " Drunken once will get drunk
again " with u Whoever goes to sleep late will wake up
late" "because there are the same words in the sentence
VERBAL UNDERSTANDING 157
before the comma^ and the words that are repeated in the
two sentences are put in the same places ; in both sentences
there is a word that is repeated." It would therefore seem
that Peril had been guided purely by the schematized
general form. But he argues from this to an Indentifi-
cation of the meaning of the words : "Because whoever
has drunk wants to go on drinking^ and whoever goes to
sleep late will also wake up late"
In other cases the child seems to be on the look-out
only for words resembling each other in sense or in sound
(' petit' and 'petites,' 'habit 1 and < habitude'); in fact
he seems to start from the understanding of particular
words. But here again the general schema is built
up just as definitely. In syncretism of understanding,
as in that of perception, there is solidarity between the
details and the general schema. One may appear before
the other, and is therefore independent of it ; but it will
call forth the other and be called forth by it by a process
of alternation which prolongs itself into an indefinitely
protracted oscillation. As this rhythm is repeated,
the details are more and more analysed, and the whole
is more and more synthesized. The result is, that to
begin with, only the largest and most distinctive details
are noticed, and only the coarsest of general schemas are
constructed. At first, therefore, the distinctive details
and the general effect are more or less mixed up to-
gether ; then analysis and synthesis develop concurrently
and at the expense of this initial syncretism.
It will now be understood why in syncretistic reason-
ing the use of explicit analogy is so inextricably inter-
twined with that of general schemas which mutually
imply one another. It is because, through a succession
of conscious realizations on the part of the subject,
syncretism of reasoning grows out of that of under-
standing and perception,
5. CONCLUSION. Our readers will perhaps be in-
clined to take the view that ego-centric thought, which
gives rise to all these phenomena of syncretism, is
158 STUDIES IN CHILD LOGIC
closer to autistic or dream thought than to logical
thought. The facts which we have been describing
are in several of their aspects related to dreaming or
day-dreaming. We mean such things as the picking
out of verbal and even punning resemblances, and,
above all, the way in which the mind is allowed to
float about at the mercy of free associations, until two
propositions are brought together which originally had
nothing in common.
This is not the place to carry out an exhaustive com-
parison between syncretism and autistic imagination.
Besides, we have already pointed out the relation which
exists between pathological interpretation and children's
justification at any price. Nevertheless, it may be
useful at this point to note that everything leads us to
consider the mechanism of syncretistic thought as
intermediate between logical thought and that process
which the psycho-analysts have rather boldly described
as the * symbolism ' of dreams. How, after all, does
autistic imagination function in dreams? Freud has
shown that two main factors contribute to the forma-
tion of the images or pseudo-concepts of dreaming and
day-dreaming condensation by which several disparate
images melt into one (as different people into one
person), and tranference by which the qualities belonging
to one object are transferred to another (as when some-
one who may bear a certain resemblance to the dreamer's
mother is conceived of as his actual mother). As we
have suggested elsewhere 1 there must be every kind of
intermediate type between these two functions and the
processes of generalization (which is a sort of condensa-
tion) and abstraction (which is a sort of transference).
Now syncretism is precisely the most important of these
intermediate links. Like the dream, it 'condenses'
objectively disparate elements into a whole. Like the
dream, it 'transfers/ in obedience to the association of
1 J. Piaget. "La psychanalyse dans ses rapports avec la psychologic de
1' enfant," Bullet. Soc. Alf. Binet^ No. 131-133, p. 56-7.
VERBAL UNDERSTANDING 159
ideas, to purely external resemblance or to punning
assonance, qualities which seem rightly to apply only
to one definite object. But this condensation and trans-
ference are not so absurd nor so deeply affective in
character as in dreams or autistic imagination. It may
therefore be assumed that they form a transition between
the pre-logical and the logical mechanisms of thought. 1
We ought certainly to be careful not to underestimate
thought carried on by means of syncretistic schemas,
for in spite of all the deviations which we have described,
it does lead the child on to a progressive adaptation.
There is nothing unintelligent about these schemas ;
they are simply too ingenuous and too facile for pur-
poses of accuracy. Sooner or later they will be sub-
mitted to a rigorous selection and to a mutual reduction,
which will sharpen them into first-rate instruments of
invention in spheres of ^thought where hypotheses are
of use. It is only at the age of the children we have
studied that this exuberance hinders adaptation, because
it is still too closely connected with autistic imagination.
These analogies between syncretism and autistic
imagination also explain why the answers which we
obtained from the children so often seemed to have been
invented. The impression must often have been created
that the children we questioned were making fun either
of us or of the test, and that the many solutions which
they discovered at will could have been exchanged for
any others that might have suggested themselves, with-
out the child being in any way put out. This would
considerably diminish the value of the facts we have
studied. This objection is one that it is always very
difficult to meet, because in questioning a child we have
1 One more proof of the analogy between syncretism and imagination has
been supplied by M. Larsson's fine work (H. Larsson, La Logique de la
potsit) transl. Philipot, Paris, Leroux, 1919). M. Larsson has shown that
artistic imagination (which is one of the forms of autism) consists first and
foremost in seeing objects not analytically as does intelligence, but syncretistic-
ally, <?,, by means of general perceptions. Art brings out the Gestaltqualitaten
of things, and the intropathic effort made by the artist consists in reviving
this primitive total perception.
160 STUDIES IN CHILD LOGIC
no criterion for ascertaining whether he is inventing or
whether he really believes what he says. On the other
hand, there are three criteria which we wish to suggest
for this kind of research, and which, if simultaneously
applied, would enable us to distinguish between in-
vention and truth. Now, in the case of the phenomena
described above, we believe there has been no invention,
and that the resemblance to invention which marks the
answers obtained was due precisely to the analogy
between syncretism and imagination, invention being
one of the forms of childish imagination. Here are the
three criteria which enable us to draw this conclusion
instead of denying all value to our results.
First criterion : Uniformity or Numerical Constancy of
the Answers. When the same experiment is carried out
on a great many different subjects, the answers obtained
either all resemble each other, or are diverse and un-
classifiable. In the first of the two cases there is far
less chance of there having been invention than in the
second. It may well be that answers which the child
gives according to his fancy or simply as a game
follow a law as regards both form and content. What
is less probable is that the same question put to 40 or
50 different children should always call forth invention,
instead of provoking sometimes an adapted, sometimes
fabricated answer. Now in the case of our proverbs,
all the answers were alike in content as well as in form.
Second criterion : The Difference of Age in the children.
Certain questions provoke romancing in children of a
given age (e.g., from 5 to 6) who do not understand
them, and therefore treat them as a game. Between
7 and 8 the same questions will be understood, and
therefore taken seriously. When all the children of
the same age answer in the same way, the question
may still be raised whether there has not been inven-
tion owing to general incomprehension. But if for
several years in succession the answers are practically
the same, then the chances of invention have been
VERBAL UNDERSTANDING 161
decreased. Now this is what our proverbs reveal.
Between 9 and n we get the same results, and the
subjects examined outside of these limits gave answers
of a similar nature. As the years increased, there was
simply a more or less sensible diminution of syncretism.
Third criterion : Finding the right Answer. It is easy
to see, once the child finds the right answer, whether
his method changes, and whether he suddenly denies
what till then he apparently believed. If so, then there
is a chance that invention has taken place. If, on the
other hand, there is continuity between the methods
which lead to error and those which lead to the right
solution, if there are imperceptible stages of develop-
ment, then the chances are against invention. Now
in the case of our proverbs, right answers and wrong
lay side by side in the child's mind, and the right
answers did not exclude the presence of syncretism
from the method by which the correspondence between
proverb and sentence was found.
We can therefore conclude that our children's answers
are not due to invention. One can never be sure, how-
ever, that in some cases the answers have not been
invented. We have even become certain of the contrary
in connexion with one or two examples which stood
out from among the others by reason of their being
more arbitrary in character. For the rest, if our material
does show any signs of invention, this is due to the
fact that syncretism, like all manifestations of the ego-
centric mentality, occupies, as will be shown later on,
a position half-way between autistic and logical thought. 1
1 For a more detailed treatment of this parallel between autistic and child
thought, see our article, "Lapensee symbolique et la pens^e de Fenfant,"
Arch, de Psych., Vol. XVIII, p. 273 (1923).
CHAPTER V
THE QUESTIONS OF A CHILD OF SIX 1
THERE is no better introduction to child logic than
the study of spontaneous questions. We have already
alluded in our analysis of the language of two children
(Chapter I) to a classification of children's questions
which would throw light on the interest taken at
successive ages in one intellectual activity or another
(causal explanation, logical justification, classification,
etc.), and we anticipated this classification in dealing
with the questions of Lev and Pie (Chapter I, 9).
The time has now come to approach the problem
systematically, and to let it serve as a transition subject
between the functional study of verbal intelligence in
the child (Chapters I-IV) and the analysis of the
peculiarities of child logic (Vol. II).
The problem which we propose to solve can be stated
as follows: What are the intellectual interests or, if
one prefers, the logical functions to which the questions
of a given child testify, and how are those interests to be
classified ? In order to solve this problem it is sufficient
to make a list extending over a certain stretch of time,
if not of all the questions asked by a child, at least of
those asked of the same person, and to classify these
questions according to the sort of answer which the
child expects to receive. But this classification is a
nicer matter than one would think, and we shall
therefore be more concerned with the creation of an
instrument of research than with its practical application.
1 With the collaboration of Mile. Liliane Veihl.
162
A CHILD'S QUESTIONS 163
The material on which we shall work consists of
1125 spontaneous questions asked of Mile Veihl over
a period of ten months by Del, a boy between 6 and 7
(6 ; 3 to 7 ; i). These questions were taken down in
the course of daily talks lasting two hours ; each talk
was a sort of lesson by conversation, but of a very
free character during which the child was allowed to
say anything he liked. These talks had begun long
before lists were made of the questions, so that the
child found himself in a perfectly natural atmosphere
from the start. Also, what is more important, he never
suspected that his questions were being noted in any
way. Mile Veihl possessed the child's full confidence,
and was among those with whom he best liked to
satisfy his curiosity. No doubt the subject-matter of
the lessons (reading, spelling, general knowledge) had
a certain influence on the questions that were asked, but
that was inevitable. The chance occurrences of walks
or games which, incidentally, played their own small
part in these interviews had just as much influence
in directing the subject's interests. The only way to
draw a line between what is occasional and what is
permanent in the curiosity of a child, is to multiply
the records in conditions as similar as possible. And
this is what we did. Finally, it need hardly be said
that we abstained as carefully from provoking questions
as from picking and choosing among those that were
asked.
Nevertheless, as this investigation was originally
destined to help only in the study of 'whys/ these
alone have been taken down in their entirety during
the first interviews in which the experimenter was at
work. For a few weeks, all other questions were only
taken down intermittently. On some days, it goes
without saying, all questions were taken down without
any exception, but on others, only the < whys' were
taken into account. Questions 201 to 450, 481 to 730,
and 744 to 993, however, represent a complete account
164 STUDIES IN CHILD LOGIC
of all the questions asked during the corresponding
periods of time. The statistical part will therefore deal
only with these three groups of 250 questions each,
or else with the ' whys/
I. ' WHYS'
Before broaching the difficult problem presented by
the types of question asked by Del, let us try, by way
of introduction, to solve a special and more limited
problem that of the different types of ( whys.'
The question of children's ' whys ' is more complex
than it at first appears to be. It is well known that the
whys which appear somewhere about the age of 3
(Stern mentions them at 2 ; 10, 3 ; i etc., Scupin at
2 ; 9, Rasmussen between 2 and 3 etc.), are extremely
numerous between this age and that of 7, and characterize
what has been called the second age of questions in the
child. The first age is characterized by questions of
place and name, the second by those of cause and time.
But its very abundance leads us to look upon the ( why J
as the maid-of-all-work among questions, as an un-
differentiated question, which in reality has several
heterogeneous meanings. Stern was right in pointing
out that the earliest 'whys' seem more affective than
intellectual in character, z'.e. y that instead of being the
sign of verbal curiosity, they rather bear, witness to a
disappointment produced by the absence of a desired
object or the non-arrival of an unexpected event. But
we have yet to ascertain how the child passes from this
affective curiosity, so to speak, to curiosity in general,
and finally to the most subtle forms of intellectual
interest such as the search for causes. Between these
two extremes there must be every shade of intervening
variety, which it should be our business to classify.
There is a certain category of childish * whys ' which
do seem, from a superficial point of view, to demand a
causal explanation for their answer. Such an one is
A CHILD'S QUESTIONS 165
this, one of the earliest questions of a boy of 3 : " Why
do the trees have leaves?" Now, if such a question
were asked by an adult, educated or otherwise, it
would imply at least two groups of answers. One
group of answers, the fmalistic, would begin with the
word 'to' ("to keep them warm, 5 ' "to breath with/'
etc.), the other, the causal or logical, would begin with
the word ' because ' ( " because they are descended from
vegetables which have leaves" or "because all veget-
ables have leaves." It is, therefore, not possible to
see at first which of these two shades of meaning is
uppermost in the child's question. There may even
be a quantity of other meanings which elude our under-
standing. The question may be merely verbal, and
indicate pure astonishment without calling for any
answer. This is often the case with the questions of
children ; they are asked of no one, and supply in effect a
roundabout way of stating something without incurring
contradiction. Very often, if one does not answer a
child immediately, he will not wait, but answer himself.
We have already (Chapter I) come across several of
these ego-centric questions, which are strictly speaking
pseudo-questions. But this will not be taken into
account in the classification which follows. However
ego-centric a question may be, it is always interesting
that it should have been expressed in the form of a
question, and the type of logical relation (causality,
finalism etc.) which it presupposes is always the same
as that which would characterize the question if it were
asked of anyone. In this connexion, the question
which we quoted admits of many more meanings for
the child than for the adult. The child may have
wanted to know, from anthropomorphism and apart
from any interest in the tree itself, "who put the leaves
on the trees." (Why have the trees got leaves ? Because
God put them there). He may have purposive or utili-
tarian ideas in relation to humanity. (Why . . . etc.
So that it should look pretty. So that people might
166 STUDIES IN CHILD LOGIC
sit in its shade, etc.), or in relation to the tree itself,
which the child would endow with more or less explicit
aims (because it likes to, etc.). In a word, a large
number of interpretations are always possible when
a child's < why ' is isolated from its context.
The lists of ' whys ' belonging to the same child, such
as we are going to discuss, will therefore by the mere
fact that they make comparisons possible, help to solve
the two following problems, which without this method
are insoluble: i What are the possible types of 'whys/
classified according to the logical type of answer which
the child expects or which he supplies him self ? 2 What
is the genealogy of these types?
i. PRINCIPAL TYPES OF 'Wnvs.* There are three
big groups of children's l whys J the ' whys J of causal
explanation (including finalistic explanation), those of
motivation, and those of justification. Inside each group
further shades of difference may be distinguished. After
a certain age (from 7 to 8 onwards) there are also the
whys of logical justification , but they hardly concern us
at the age of Del, and they can be included in the
" whys of justification " in general.
The term explanation is to be taken in the restricted
sense of causal or finalistic explanation. For the word
' explain ' carries with it two different meanings. Some-
times it signifies giving a < logical' explanation, i.e.,
connecting the unknown with the known, or giving- a
systematic exposition (explaining a lesson or a theorem).
4 Whys' referring to logical explanation (''Why is half
9, 4-5? ") are to be classified as " logical justification."
Sometimes, on the contrary, the word ( explain ' means
carrying back our thought to the causes of a pheno-
menon, these causes being efficient or final according
as we are dealing with natural phenomena or with
machines. It is in the second sense only that we shall
use the word explanation. < Whys ' of causal explana-
tion will therefore be recognizable by the fact that the
expected answer implies the idea of cause or of final
A CHILD'S QUESTIONS 167
cause. Here are some examples taken from Del : " Why
do they (bodies) always fall?" "Lightning . . . Daddy
says that it makes itself all by itself in the sky. Why ?
[does it make itself in this way]" "Why haven't they
[little goats] got any milk ? " " Why is it so heavy [a two-
franc piece]?"
Let us designate as motivation that sort of explanation
which accounts, not for a material phenomenon, as in the
last category, but for an action or a psychological state.
What the child looks for here is not, strictly speaking,
a material cause, but the purpose or the motive which
guided the action, sometimes also the psychological
cause, 'Whys 5 of motivation are innumerable, and
easy to classify: "Are you going away? Why?"
"Why do we always begin with reading ?" " Why doesn't
daddy knozv the date ? He is quite grown up?
Finally, let us designate as "whys" of justification
those which refer to some particular order, to the aim,
not of some action, but of a rule. "Why do we have
to . ... etc." These c whys * are sufficiently frequent
in the case of Del to justify the formation of a separate
category. The child's curiosity does not only attach
itself to physical objects and the actions of human
beings, it goes out systematically to all the rules that
have to be respected rules of language, of spelling,
sometimes of politeness, which puzzle the child and of
which he would like to know the why and the wherefore.
Sometimes he seeks for their origin, z>., his idea of it,
the object of the * grown-ups ' who have decided that
it should be so, sometimes he looks for their aim.
These two meanings are confused in the same question
"why .... etc.?" We have here a collection of
interests which can be united under the word * justifica-
tion ' and which differ from the simple interest in
psychological motivation. Here are examples of it,
some less obvious than others: " Why not l an' [in the
spelling of a word] ? You can't tell when it is < an' or
L en ; Why not i in' [in * Alain']? Who said it
168 STUDIES IN CHILD LOGIC
shouldn't be, the grown-ups in Paris?" " Why do people
say * strayed? does it mean lost?" " Black coffee -, why
black? All coffee is black. ..."
These, then, are the three great classes of ' whys '
which it is possible to establish straight away. But
it need hardly be said that these are ' statistical ' types,
i.e., that between them there exists every kind of inter-
mediate variety. If all the existing transitional types
could be arranged in order, and their shades of differ-
ence expressed in numbers, then the three main types
would simply represent the three crests of a graph of
frequency. Between these summits there would be the
intermediate zones. In psychology, as in zoology, we
must needs adopt a classification into species and
varieties ; even though its application is purely statis-
tical, and though an individual sample taken at
random cannot be placed for certain in one class rather
than another until its true nature and derivation have
been established by experiment and analysis.
It is obvious, for instance, that between the " causal
explanation" of physical objects called forth by the
questions of the first group, and psychological ' moti-
vation/ there are two intermediate types. Alongside
of the explanations which the child himself considers
as physical (the cloud moves because the wind, drives
it along) there are those which he looks upon as
mixed up with motivation (the river is swift because
man or God wanted it to be), and there are those which
we ourselves consider mixed (the two-franc piece is
heavy because it is in silver, or because it was made
to weigh more than a one-franc piece, etc.). Causal
explanation therefore often inclines to motivation. But
the converse also happens. In addition to the "whys
of motivations" which refer to a momentary intention
(Why are you going away?) there are those which
involve explanations of a more psychological nature,
and appeal no longer to an intention, but to cause
properly so-called (Why does Daddy not know the
A CHILD'S QUESTIONS 169
date?), which brings us back to the first type of ques-
tion. The result is, that we can give no fixed form
to the criterion used to distinguish causal explanation
from motivation. The decision in each case as to
whether the child wanted to be answered with a
causal explanation or with a motivation would be too
arbitrary. The criterion can only be practical, and will
have to adapt itself to the contents of the question.
When the question refers to physical objects (natural
phenomena, machines, manufactured objects, etc.), we
shall class it among the ' whys ' of causal explanation ;
when the question refers to human activities, we shall
class it among the < whys ' of motivation. This classi-
fication is a little arbitrary, but the convention is easy to
follow. In our opinion the attempt to define the child's
motive too closely would be still more arbitrary, for that
would be to put the purely subjective judgment of each
psychologist in the place of conventions, which may
be rigid, but which are known to be only conventions.
The distinction, between motivation and justification,
on the other hand, is even more difficult to establish
with any precision. In the main, < whys' of justifica-
tion imply the idea of rules. But this idea is far less
definite in the child than in us, so that here again we
are obliged to use a criterion bearing on the matter
rather than on the form of the question. The justifica-
tion of a rule is very closely allied to motivation, to
the search for the intention of him who knows or who
established the rule. We shall therefore say that
* whys ' of justification are those which do not bear
directly on a human activity, but on language, spell-
ing and, in certain cases to be more closely defined, on
social conventions (bad manners, prohibitions, etc.).
If we insist upon this third class of 'whys/ it is
because of the following circumstance. We have
shown throughout the last three chapters that before
7 or 8 the child is not interested in logical justification.
He asserts without proving. Children's arguments in
170 STUDIES IN CHILD LOGIC
particular consist of a simple clash of statements, with-
out any justification of the respective points of view.
The result is, that the word ' because' corresponding to
a logical demonstration (* because' connecting two ideas
of which one is the reason for the other) is rarely used
by the child, and, as we shall see in the next volume,
is even imperfectly understood by him ; in a word,
it is alien to the habits of thought of the child under
7 or 8. Now, corresponding to this group of logical
relations, to the ' because ' which unites two ideas, there
is obviously a group of t whys of justification ' whose
function is to find the logical reason for a statement, in
other words, to give a proof or to justify a definition.
For example "why is 4*5 half of 9?" This is not a
case of causal or psychological motivation, but quite
definitely one of logical reasoning. Now, if the observa-
tions made in the course of the last chapters are correct,
we shall expect and our expectation will prove to be
justified this type of ( why ' to be very rare under the
age of 7 or 8, and not to constitute a separate class.
But and this is why we wish to keep the 'whys of
justification ' in a separate category there is for the child
only a step from a rule of spelling or grammar to the
definition of a word, etc., and from the definition of a
word to a genuine u logical reason. " Everyone knows
that childish grammar is more logical than ours, and
that the etymologies spontaneously evolved by children
are perfect masterpieces of logic. Justification in our
sense is therefore an intermediate stage between simple
motivation and logical justification. Thus, in the ex-
amples quoted, " Why do people say < strayed' ? " [instead
of saying lost] inclines to the * whys ' of motivation ;
" Why black coffee, all coffee is black" seems to appeal to
a logical reason (which would be a link between a reason
and conclusion) ; and the other two seem to be inter-
mediate ' whys' which appeal to a certain form of
spelling, etc.
To sum up. " Whys of justification " are an undifferen-
A CHILD'S QUESTIONS 171
tiated class before the age of 7 or 8. After the age of 7 or
8 this class is replaced such at least is our hypothesis
by two other classes. One of these, " logical justifi-
cation or reason," is to be contrasted with causal explana-
tion and motivation. The other, u justification of rules,
customs, etc." can be considered as intermediate between
logical justification and motivation. Before the age of
7 to 8, these two classes can therefore be united into one.
This gives us the following table :
Form of the question Matter of the question
Motivation Motive ..... Psychological actions
r ,i,c ,* f Justification proper . . . . Customs and rales
Justification^ . . - j - r--,
I Logical reason . Classification and connexion of ideas
In addition to this, it should be pointed out that there
are certain classes of questions beginning with such
words as 'how,' 'what is,' 'where , . . from,' etc.,
which correspond word for word with the classes of
whys J of which we have just spoken. This will supply
us with a very useful counter-proof.
2. "WHYS OF CAUSAL EXPLANATION." INTRO-
DUCTION AND CLASSIFICATION BY MATERIAL. We have
no intention of attacking the vast subject of causality
in childish thought. On the contrary, we have con-
centrated on the problem of the formal structure of
reasoning in the child. As it is very difficult to isolate
the study of causality from that of children's ideas in
general, the subject should by rights be excluded from
our investigation. Nevertheless, two reasons compel us
to broach the matter here and now. In the first place,
one of the objects of this chapter was to show that in
DePs case there are very few " whys of justification " ;
we must therefore analyse the sum of the answers
obtained, in order to become aware of this absence.
In the second place, it is quite possible to study causality
from the point of view of the structure of reasoning, and
172 STUDIES IN CHILD LOGIC
In particular of the influences of ego-centrism, without
encroaching too far on the actual domain of ideas.
We shall, in spite of this, however, say a few words
about these ideas, and then draw our conclusions in the
next paragraph on the structure of the questions which
have been asked.
" Whys of causal explanation " raise a number of
problems which are of paramount importance in the
study of the child intelligence. It is indeed a matter
for conjecture whether a child feels in the same degree
as we do the need for a causal explanation properly
so-called (efficient cause as opposed to finalism). We
ought therefore to examine the possible types of causality
which could take the place of causality properly so-called.
Stanley Hall has shown that out of some hundreds of
questions about the origin of life (birth) 75% are causal.
But he never stated his criteria. He simply pointed out
that among these causal questions a large number are
artificialistic, animistic, etc. 1 The difficulty therefore
still remains of classifying these types of explanation
and of finding their mutual relations to each other.
Now we had occasion in two earlier chapters (I and
III) to show that the child between 6 and 8 takes very
little interest in the * how ' of phenomena. His curiosity
reaches only the general cause, so to speak, in contrast
to the detail of contacts and of causal sequences. This
is a serious factor in favour of the sui generis character
of u whys of explanation" in the child. Let us try to
classify those of Del, starting from the point of view of
their contents, and not bothering about their form.
Classifying these ' whys ' by their contents consists in
grouping them according to the objects referred to in
the question.
In this connexion, 2 out of the 103 "whys of causal
1 Stanley Hall, "Curiosity and Interest," Pedag. Sem., Vol. X. (1903).
See numerous articles which have since appeared in the Pedag. Sem.
2 Out of 360 * whys' taken down in ten months, 103, viz. 28% were found
to be "whys of causal explanation."
A CHILD'S QUESTIONS 173
explanation " 88 refer to nature and 22 to machines or
manufactured articles. The 81 'whys' concerning
nature can be subdivided into 26 questions about
inanimate objects (inanimate for the adult), 10 about
plants, 29 about animals and 16 about the human body.
What is most remarkable about this result is the
feeble interest shown in inanimate physical objects.
This circumstance should put us on our guard from the
first against the hypothesis according to which DePs
* whys ' would refer to causality in the same sense as
ours. A certain number of the peculiarities exhibited
by the 'whys' concerning the physical world will enable
us to state the problem more precisely.
In the first place, some of Del's questions bear witness
to the well-known anthropomorphism of children. It
may be better described as artificialism^ but nothing is
known as yet about its origin or its duration. For
example : " Why [does the lightning make itself by itself
in the sky] ? Is it true ? But isn't there everything that is
needed to light afire with up in the sky?" These artificial-
istic questions, which are rarely quite clear, obviously
do not presuppose an efficient mechanistic causation
analogous to ours.
Other and more interesting ' whys ' raise the problem
of chance in the thought of the child. These for ex-
ample : Del had thought that Bern was on the lake.
" The lake does not reach as far as Bern " " Why? '' or
Why does it not make a spring in our garden ? " etc. Mile
V. finds a stick and picks it up. "Why is that stick
bigger than you" "Is there a Little Cervin and a Great
Ceivin? " No Why is there a Little Saleve and a Great
Saleve?" Questions of this kind abound with this
child ; we shall come across many more, and they will
always surprise us. We are in the habit of alotting a
large part to chance and contingency in our explana-
tions of phenomena. All "statistical causality," which
for us is simply a variety of mechanical causality, rests
on this idea of chance, t.e., of the intersection of two
174 STUDIES IN CHILD LOGIC
independent causal sequences. If there are no springs
in the garden, it is because the series of motives which
lead to the choice of the garden's locality is independent
of the series of causes which produced a spring a little
distance away. If the two lines of sequences had
crossed, it would only be by chance, as there is very
little likelihood of their crossing. But it is clear that
this idea of chance is derivative : it is a conclusion forced
upon us by our powerlessness to explain. The result
is, that the child is slow in reaching the sort of agnosti-
cism of ordinary adult life. For lack of a definite idea
of chance, he will always look for the why and where-
fore of all the fortuitous juxtapositions which, he meets
with in experience. Hence this group of questions.
Do these questions then point to a desire for caysal
explanation? In a sense they do, since they demand
an explanation where none is forthcoming. In another
sense they do not, since obviously a world in which
chance does not exist is a far less mechanical and far
more anthropomorphic a world than ours. Besides, we
shall meet with this problem of causality again, in con-
nexion with other varieties of ' why.'
The following questions, however, seem to really
belong to the order of physical causality :
(i ) " Why do they [bodies] always fall? " (2) ' ' It [water]
can rtm away, then why [is there still some water in the
rivers] ? " (3) " The water goes to the sea Why ? " (4)
" There are waves only at the edge [of the lake]. Why?"
(5) " Why does it always do that [stains of moisture] when
there is something there [fallen leaves]?" (6) "Will it
always stay there [water in a hole worn away in the
sandstone] ? No, the stone absorbs a lot Why? Will
it make a hole? No Does it melt?" (7) "WAy does it
get colder and colder as you go up [as you go north]?"
and (8) "Why can you see lightning better at night"
It is worth noting how difficult it is to determine
what part is played by finalism in these questions and
what by mechanical causality. Thus ' whys' 3, 7, and
A CHILD'S QUESTIONS 175
8 could easily be interpreted as finalistic questions:
you see the lightning better in order to ... etc. It
is only in questions i, 4, 5, and 6 that we can be at
all sure of any desire for a causal explanation, because
the objects are uncircumscribed, and clearly independent
of human or divine intervention. Lightning, on the
other hand, is, as we have just seen, spontaneously
conceived of as ( manufactured ' in the sky ; rivers, as
we shall show later on, are thought of as put into
action by man, etc.
In a word, these questions asked about the physical
world are far from being unambiguously causal.
Questions about plants do not throw much light on
the matter. Some point to a certain interest in the
circumstances of the flower's habitat: "Why are there
not any [bluebells] in our garden?" Others, more
interesting, refer to the life and death of plants : "Has
it rained in the night ? No. Then why have they [weeds]
grown?" Why do we not see those flowers about now [the
end of summer]?" " They [roses on a rose-tree] are all
withered^ why P They shouldrit die^ because they are still on
the tree " " Why does it [a rotten mushroom] drop off so
easily*}" We shall come across this preoccupation with
'death' in connexion with other questions, which will
show that this interest is of great importance from the
point of view of the idea of chance. The first group
of questions jraises the same problems as before: the
child is still very far from allowing its share to chance
in the nexus of events, and tries to find a reason for
everything. But is the reason sought for causal, or
does it point to a latent finalism?
Questions on animals are naturally very definite in
this connexion. About half of them refer to the inten-
tions which the child attributes to animals. "Does the
butterfly make honey ? No. But why does it go on the
flowers?" "Why do they [flies] not go into our ears?"
etc. These ' whys' ought s to be classified as "whys
of motivation," but we are confining this group to the
176 STUDIES IN CHILD LOGIC
actions of human beings. If it were extended to include
animals, there would be no reason for not extending
it to include the objects which at 6 or 7 the child still
openly regards as animated, such as (according to the
research work done at the Institut Rousseau on the ideas
of children) stars, fire, rivers, wind, etc. Among the
other questions, only four are causal, and curiously
enough, are again concerned with death :
" They [butterflies] die very soon, why ?" " Will there
still be any bees when I grow up ? " Yes, those that you
see will be dead, but there will be others. "Why?"
" Why do they [animals] not mind [drinking dirty
water] ? " " It [a fly] is dead, why ? "
The rest of the questions about animals are either
finalistic, or else those whys ' about special fortuitous
circumstances or about anomalies for which the child
wishes to find a reason.
" Why is it [a pigeon] like an eagle, why?" " If they
[snakes] are not dangerous, why have they got those things
[fangs]? " " Why has a cockchafer always got these things
[antennae]?" "It [an insect] sticks to you, why?"
[Looking at an ant]: "I can see green and red, why?"
"It [a cockchafer] can't go as far as the sun, why?"
[Del draws a whale with the bones sticking out of
its skin]; "You shouldn't see the bones, they don't
stick out. " Why, would it die? "
Some of these questions mean something, others
(those about the pigeon, the ant, etc.) do not. That
is because in the second case we bring in chance by
way of explanation. If our idea of chance is really due
to the impotence of our explanation, then this distinction
is naturally not one that can be made a priori. The
child, therefore, can have no knowledge of these shades
of difference ; hence his habit of asking questions in
season and out of season. Shall we adopt Groos' view
that curiosity is the play of attention, and interpret
all these questions as the outcome of invention?
A CHILD'S QUESTIONS 177
But this would not explain their contents. If childish
questions strike us as uncouth, it is because for the
child, everything can a priori be connected with every-
thing else. Once the notion of chance, which is a
derivative notion, is discarded, there is no reason for
choosing one question rather than another. On the
contrary, if everything is connected with everything
else it is probable that everything has an end, and an
anthropomorphic end at that. Consequently no question
is absurd in itself.
Questions referring to the human body will help us
to understand more clearly this relation between finalism
and those ' whys ' which are the negation of chance.
Here is an example of a definitely finalistic *why' where
we would have expected a purely causal one : Del asks
in connexion with negroes: "If I stayed out there for
only one day, would I get black all over?" (This question
without being a ' why * appears to be definitely causal.
The sequel shows that it is nothing of the kind) No
Why are they made to be [exist] like that? Although
too much stress must not be laid on the expression " are
they made to," it obviously points to a latent finalism.
There is therefore every likelihood that the following
questions will be of the same order :
" Why have you got little ones [ears] and I have big ones
although I am small?''' and " Why is my daddy bigger
than you although he is young?" " Why do ladies not
have beards?" " Why have I got a bump [on the wrist]?**
" Why was I not born like that [dumb]?" "Cater-
pillars turn into butterflies, then shall I turn into a little
girl?" No Why? " Why has it [a dead caterpillar]
grown quite small? When I die shall I also grow quite
small?" 1
Now here again, most of the questions are put as
though the child were incapable of giving himself the
answer: "By chance." At this stage, therefore, the
1 These last two questions correspond to two spontaneous ideas of child-
hood which are well known to psycho-analysts : that it is possible to change
one's sex, and that after death one becomes a child again.
M
178 STUDIES IN CHILD LOGIC
idea of the fortuitous does not exist; causality pre-
supposes a * maker/ God, the parents, etc., and the
questions refer to the intentions which he may have
had. Even those of the preceding questions which
come nearest to causality presuppose a more or less
definite finalism. Organic life is, for the child, a sort
of story, well regulated according to the wishes and
intentions of its inventor.
We can now see what is the part played by questions
about death and accidents. If the child is at this stage
puzzled by the problem of death, it is precisely because
in his conception of things death is inexplicable. Apart
from theological ideas, which the child of 6 or 7 has
not yet incorporated into his mentality, death is the
fortuitous and mysterious phenomenon par excellence.
And in the questions about plants, animals, and the
human body, it is those which refer to death which will
cause the child to leave behind him the stage of pure
finalism, and to acquire the notion of statistical causality
or chance.
This distinction between the causal order and the
order of ends is undoubtedly a subtle one if each case
be examined in detail, but we believe that the general
conclusions which can be drawn from it hold good.
Del has a tendency to ask questions about everything
indiscriminately, because he inclines to believe that
everything has an aim. The result is, that the idea
of the fortuitous eludes him. But the very fact that
it eludes him leads him to a preference for questions
about anything accidental or inexplicable, because
accident is more of a problem for him than for us.
Sometimes, therefore, he tries to do away with the
accidental element as such, and to account for it by
an end, sometimes he fails in this attempt, and then,
recognizing the fortuitous element for what it is, he
tries to explain it causally. When, therefore, we are
faced with a child's question that appears to be causal,
we must be on our guard against any hasty conclusion,
A CHILD'S QUESTIONS 179
and see by careful examination whether the finalistic
interpretation is excluded. It is not always possible to
come to a conclusion, and out of the 81 'whys ' referring
to nature, only one-tenth can be said to be definitely
causaL This is very little. Classification of the questions
by their contents cannot therefore correspond term for
term with a formal classification : the interest shown
in natural objects does not constitute a direct proof of
interest in mechanical or physical causality.
Before enquiring any further into the nature of childish
causality, let us examine the ' whys ' that relate to the
technical appliances of human beings, we mean machines
or manufactured articles. Out of 22 such questions,
two-thirds are simply about the intention of the maker :
" Why are the funnels [of a boat] slanting? " " Why have
two holes been made in this whistle?" These are the
questions which are continuous with the " whys of moti-
vation," but easily distinguishable from them, since the
question refers to the manufactured object. Only in a
few cases can there be any doubt as to the particular
shade of meaning. For example, before a picture of a
woman handing a cabbage to a little girl : " Why does
it always stay like that?" Does Del wish to know the
psychological intention of the artist or of the woman,
or is he asking why the drawing represents movement
by fixing a single position into immobility ?
The other fc whys' are more interesting: they refer
to the actual working of the machines, or to the
properties of the raw materials that are used :
" Why has it [a crane] got wheels? " " There are lamps
in our attic at home. When therms a thunderstorm, the
electricity cant be mended. Why?" After leaning too
heavily with his pencil on a sheet of paper: "Why can
you see through? " He traces a penny : " Why is this one
all right and not the other one?" His name had been
written in pencil on his wooden gun. The next day it
did not show any more : * ' Why do wood and iron rub
out pencil marks?" While he is painting: 41 When I
mix red and orange it makes brown, why ? "
i8o STUDIES IN CHILD LOGIC
Several of these questions do seem to call for a causal
explanation. But here, as in the questions about nature,
the definitely causal questions are concerned almost
exclusively with the element of accident, whereas those
which refer to a customary event [the question about
the crane or about the colours] seem as much concerned
with utility or motive as with cause. At any rate, we
have not found one indisputably causal question, even
among those concerning the working of machines. In
this respect, therefore, the questions of this group
confirm our previous findings.
3. STRUCTURE OF THE " WHYS OF EXPLANATION."
The reader can now see for himself how complex is
the problem of causality in the child, and how much
a classification based on the contents of the questions
differs from a formal classification, i.e., one relating to
the structure of ' whys ' and to the different types of
causality. We should like to be able to give such a
formal classification which would be homogeneous with
the rest of our work. Unfortunately, the present con-
ditions of knowledge render this impossible. To carry
out such a scheme, it would have been necessary to
examine Del in detail on all the natural phenomena
about which he had asked questions, and thus establish
a parallel between his questions and the types of ex-
planation which he gave. An enquiry which has since
been set on foot, and is now being carried out with the
collaboration of Mile Guex, will perhaps yield the desired
result. In the meantime, and pending the establish-
ment of formal types of "whys of explanation," let us
content ourselves with bringing some system into the
preceding considerations, and let us try to indicate what
is the general structure of DePs " whys of explanation."
There are five principal types of adult explanation.
First of all there is causal explanation properly so-called,
or mechanical explanation: "the chain of a bicycle re-
volves because the pedals set the gear-wheel in motion."
This is causation by spatial contact. Then there is
A CHILD'S QUESTIONS 181
statistical explanation, in a sense a special case of the
former, but relating to the sum of these phenomena
which are directly or indirectly subject to the laws of
chance. Finalistic explanation is used by common sense
in connexion with the phenomena of life: "Animals
have legs to walk with." Psychological explanation, or
explanation by motive, accounts for purposive actions :
" I read this book because I wished to know its author. "
Finally, logical explanation, or justification, accounts for
the reason of an assertion : "^ is larger than j/ 1? because
all its are larger than j/s." These various types nat-
urally encroach upon each other's territory, but in the
main they are distinct in adult thought and even in
ordinary common sense.
Now what we purpose to show is that in the child
before the age of 7 or 8, these types of explanation are,
if not completely undifferentiated, at any rate far more
similar to each other than they are with us. Causal
explanation and logical justification in particular are
still entirely identified with motivation ; because causa-
tion in the child's rnind takes on the character of finalism
and psychological motivation far rather than that of
spatial contact, and because, moreover, logical justifica-
tion hardly ever exists in an unadulterated form, but
always tends to reduce itself to psychological motivation.
We shall designate by the name of precausality this
primitive relation in which causation still bears the
marks of a quasi-psychological motivation. One of the
forms taken by this precausality is the anthropomorphic
explanation of nature. In this case, the causes of phenom-
ena are always confused with the intentions of the Creator
or with those of men, who are the makers of mountains
and rivers. But even if no * intention ' can be detected
in this anthropomorphic form, the * reason ' which the
child tries to give for phenomena is far more in the
nature of a utilitarian reason or of a motive than of
spatial contact
It will be easier to understand the nature of this
182 STUDIES IN CHILD LOGIC
precausality if we explain it at once by means of one
of the most important phenomena of the mental life of
the child between the years of 3 and 7 ; we mean that
which was discovered by specialists in the drawings of
children, and which has been most successfully char-
acterized by M. Luquet as " logical realism," or, as he
now calls it, " intellectual realism." The child, as we
all know, begins by drawing only what he sees around
him men, houses, etc. In this sense, he is a realist.
But instead of drawing them as he sees them, he reduces
them to a fixed schematic type ; in a word, he draws
them as he knows them to be. In this sense, his
realism is not visual, but intellectual. The logic of
this primitive draughtsmanship is childish but entirely
rational, since it consists, for instance, in adding a
second eye to a face seen in profile, or rooms to a house
seen from the outside. Now this intellectual realism
has a significance which, as we have shown elsewhere, 1
extends beyond the sphere of drawing. The child
thinks and observes as he draws. His mind attaches
itself to things, to the contents of a chain of thought
rather than to its form. In deductive reasoning he
examines only the practical bearing of the premises,
and is incapable of arguing as we do, vi formae, on any
given < data.* He does not share the point of view of
his interlocutor (see in this connexion Vol. II where we
shall meet again with the child's incapacity for formal
ratiocination). He contradicts himself rather than lose
his hold on reality. In this sense, he is a realist. But,
on the other hand, this reality to which he clings so
continuously is the outcome of his own mental con-
struction rather than the fruit of pure observation. The
child sees only what he knows and what he anticipates.
If his powers of observation seem good, it is because
his trains of thought, which are very different from
ours, cause him to see things which do not interest us,
and which it therefore astonishes us that he should have
1 Journal de Psychologic, 1922, pp. 223, 256-257, etc.
A CHILD'S QUESTIONS 183
noticed. But on closer observation one is struck by
the extent to which his vision is distorted by his ideas.
If a child believes that rivers flow backwards, he will
see the Seine or the Rhone flowing upwards towards
their sources ; if he believes the sun to be alive, he will
see it walking about in the sky ; if he believes it to be
inanimate, he will see it always motionless, etc. In a
word, the child observes and thinks as he draws : his
thought is realistic, but intellectually so.
The structure of childish precausality will now be
clear. Children's < whys' are realistic in the sense
that in Del's language, as we shall see in 4, there are
no genuine "whys of logical justification." Curiosity
is concentrated always on the causes of phenomena (or
actions), and not logical deductions. But this causation
is not visual or mechanical, since spatial contact plays
in it only a very restricted part. Everything happens
as though nature were the outcome, or rather the re-
flexion of a mental activity whose reasons or intentions
the child is always trying to find out.
This does not mean that the whole of nature is, for
the child, the work of a God or of men. These reasons
and intentions are no more referred to one single mental
activity than they are in the prelogical mentality of
primitive races. What is meant is that instead of
looking for an explanation in spatial contact (visual
realism) or in logical deduction of laws and concepts
(intellectualism), the child reasons, as he draws, accord-
ing to a sort of " internal model," similar to nature, but
reconstructed by his intelligence, and henceforth pictured
in such a fashion that everything in it can be explained
psychologically, and that everything in it can be justified
or accounted for (intellectual realism). Thus the child
invokes as the causes of phenomena, sometimes motives
or intentions (finalisrn), sometimes pseudo-logical reasons
which are of the nature of a sort of ethical necessity
hanging over everything (" it always must be so"). It
is in this sense that the explanations of children point
184 STUDIES IN CHILD LOGIC
to intellectual realism and are as yet neither causal
(spatial contact) nor logical (deduction), but precausal.
For the child, an event leading to an event, a motive
leading to an action, and an idea leading to an idea
are all one and the same thing ; or rather, the physical
world is still confused with the intellectual or psychical
world. This is a result which we shall frequently meet
with in our subsequent investigations.
Three independent groups of facts seem to confirm
our analysis of precausality in the child. The first is
the rareness of ' whys ' of pure causation and of ' whys '
of justification or logical reason properly so-called.
We showed in the last paragraph that out of 103 " whys
of causal explanation " only about 13, *., an eighth or
a seventh, could be interpreted as ' whys ' of causation
properly so-called, or of mechanical causation. We
shall show elsewhere, in 4, that "whys of logical
reason" are even rarer. Thus childish thought is
ignorant both of mechanical causality and of logical
justification. It must therefore hover between the two
in the realm of simple motivation, whence arises the
notion of precausality.
In addition to this, what was said about the notion
of chance and the element of fortuitousness also favours
the hypothesis of precausality. The child asks questions
as if the answer were always possible, and as if chance
never intervened in the course of events. The child
cannot grasp the idea of the * given,' and he refuses to
admit that experience contains fortuitous concurrences
which simply happen without being accounted for.
Thus there is in the child a tendency towards justifica-
tion at all costs, a spontaneous belief that everything
is connected with everything else, and that everything
can be explained by everything else. Such a mentality
necessarily involves a use of causality which is other
than mechanical, which tends to justification as much
as to explanation, and thus once more gives rise to
the notion of precausality.
A CHILD'S QUESTIONS 185
It should be remembered, however, that this tendency
to justify, though it is an essential factor in the pre-
causal explanation, is dependent in its turn upon a
wider phenomenon which we studied in the last chapter
under the name of * syncretism. 5 The incapacity for
conceiving of the fortuitous as such, or of the ' given *
in experience is reflected in the verbal intelligence of
the child. We have shown elsewhere 1 that up till the
age of about 11, the child cannot keep to a formal chain
of argument, i.e., to a deduction based on given pre-
mises, precisely because he does not admit the pre-
mises as given. He wants to justify them at all costs,
and if he does not succeed, he refuses to pursue the
argument or to take up the interlocutors point of view.
Then, whenever he does argue, instead of confining
himself to the data, he connects the most heterogeneous
statements, and always contrives to justify any sort of
connexion. In a word, he has a tendency, both in
verbal intelligence and in perceptive intelligence (and
the tendency lasts longer in the former than in the
latter type of mental activity), to look for a justification
at any cost of what is either simply a fortuitous con-
currence or a mere * datum.' Now in verbal intelligence
this tendency to justify at any cost is connected with the
fact that the child thinks in personal, vague and
unanalysed schemas (syncretism). He does not adapt
himself to the details of the sentence, but retains only
a general image of it which is more or less adequate.
These schemas connect with one another all the more
easily owing to their vague and therefore more plastic
character. In this way, the syncretism of verbal
thought implies a tendency to connect everything with
everything else, and to justify everything. Exactly the
same thing happens in perceptive intelligence. If the
precausal questions of the child have a tendency to
justify everything and connect everything with every-
thing else, it is because perceptive intelligence is
1 Loc. cit., Journ. Psych., 1922, p. 249, sq. et passim.
i86 STUDIES IN CHILD LOGIC
syncretistic, at any rate before the age of 7 or 8. In
view of this, intellectual realism can be thought of as
necessarily connected with syncretism by a relation of
mutual dependence. Syncretism, as we have already
shown, is the characteristic of confused perception
which takes in objects as a whole, and jumbles them
together without order. 1 The result of this is that
since objects are perceived in a lump and constitute
general schemas, instead of being diffused and dis-
continuous, childish realism can only be intellectual
and not visual. For lack of an adequate vision of
detail, and in particular of spatial and mechanical
contacts, syncretistic perception is bound to make the
child connect things together by thought alone. Or,
conversely, it can be maintained that it is because the
child's realism is intellectual and not visual, that his
perception is syncretistic. Be that as it may, there is
a relation of solidarity between syncretism and intellect-
ual realism, and enough has been said to show how
deeply rooted is the childish tendency to precausal
explanation, and to the negation of anything fortuitous
or * given.'
Finally, a third group of facts compels us to adopt the
hypothesis of precausality. A great number of i whys y
of causal explanation seem to demand nothing but an
interpretation of the statements made. When Del, for
example, asks: " Daddy says that it [lightning] makes
itself by itself in the sky. Why?" it looks as though we
were asking : " Why does Daddy say that? " Or, when
he asks why the lake does not reach as far as Bern, it
may seem that Del is simply looking for the reasons
which may exist for making this assertion. As a matter
of fact, this is far from being the case. Del cares very
little whether statements put forward are proved or not.
What he wants to know is something quite different.
When he asks: " Why is it [a pigeon] like an eagle?"
or " Why do I see [on an insect] red and green colours? "
1 Claparede, Psychologic de V Enfant, Geneva, 7th ed., p. 522 (1916).
A CHILD'S QUESTIONS 187
the question, though It has the same form, can obvi-
ously not receive the same interpretation. Sully and
his commentators will help us to understand cases of
this kind. This author has rightly pointed out that
if the questions of children frequently relate to new
and unexpected subjects, it is very often because the
child wants to know whether things are really as he
sees them, whether the new elements can be made to
fit into the old framework, whether there is a ^rule.' 1
But what should be especially noticed is, that this rule
is not merely factual : it is accompanied by a sort of
ethical necessity. The child feels about each assertion
that "it must be so," even though he is unable precisely
to find any definite justification for it. Thus at a certain
stage of development (from 5 to 6), a boy who is
beginning to understand the mechanism of a bicycle,
does not concern himself with the contact of the different
pieces of the machine, but declares them all to be
necessary, and all equally necessary. It is as though
he said to himself: "it is necessary since it is there."
The feeling of necessity precedes the explanation. 2 Its
meaning is just and as finalistic as it is causal, just as
ethical as it is logical. As a general rule, moreover, the
child confuses human necessity (moral and social, the
1 decus ') with physical necessity. (The idea of law has
long retained the traces of this complex origin). A
great many of the i whys * of children, therefore, do no
more than appeal to this feeling of necessity. It is
probable that the answer to the last few ' whys * we
have quoted is not only "because it always is so" but
also "because it should be, because it must be so."
The connexion will now be seen between this type of
explanation and precausality, which is precisely the
result of a confusion between the psychical or intellectual
1 See for example Buhler, Die geistige Entwicklung des Kindes^ 2nd ed.
Jena, 1921, p. 388 et seq.
2 J. Piaget, "Pour 1'etude des explications d'enfants" L* E&tcateur,
Lausanne and Geneva, 1922, p. 33-39.
i88 STUDIES IN CHILD LOGIC
world, or the world of ethical or logical necessity and
the world of mechanical necessity.
5 4.. " WHYS OF MOTIVATION." We have shown
6 T
that among the ' whys ' relating to nature and manu-
factured articles there are several which are not really
< whys ' of causal explanation properly so-called, but
questions connected more or less with motives, and
therefore leading back to the present category. This
category therefore predominates in Del's questions,
numbering altogether 183/360 of the total.
Many of these questions are concerned only with the
motive of a chance action or of an indifferent phrase,
and are not, therefore, particularly interesting. Here
are some examples :
"Are you having lunch here? No, I can't to-day
Why? " " Does this caterpillar bite? No Then why did
Anita tell me it did? Horrid of her I " " What is your
drawing supposed to be? You want to know everything.
Thafs greedy of you. Why do you want to know every-
thing, teacher? Do you think I am doing silly things? "
" Why is she frightened? " etc.
It is in this category that we find the earliest * whys '
of indirect interrogation: "Do you know why I would
rather you didn't come this afternoon ? "
Other "whys of motivation " relate less to purely
momentary intentions than to psychological explanation
properly so-called. It is in such cases as these that the
term * motive ' takes on its full meaning, both causal
and finalistic, for to explain an action psychologically
is really to consider its motive both as its cause and as
its aim. We can extend this meaning of "why of
motivation " to cover all questions concerning the cause
of an unintentional act or psychological event. For
example: " Why do you never make a mistake?" Be-
tween a motive and the cause of a psychological action
there are numerous transitional stages. We can talk
of the motive of a fear as well as of its cause, and though
we may not be able to speak of the motive of an in-
A CHILD'S QUESTIONS 189
voluntary error, we can do so In the case of one that
is semi-intentional. In a word, short of making them
definitely separate, we shall agree to place among the
" whys of motivation " all questions relating to psycho-
logical explanation, even when it is causal. Here are
some examples :
" Why do you teach me to count? " " Why does Daddy
not know [what day of the month it is], he is a grown-
up" <c And does my Mummy [love the Lord Jesus]?
Yes, I think so Why are you not sure?" " Why shall I
be able to defend you even if I don't take it [an iron bar]?
Because I am a boy? " " Why are angels always kind to
people? Is it because angels dorit have to learn to read
and do very nasty things? Are there people who are wicked
because they are hungry?" " Why can I do it quickly and
well now, and before I did it quickly and badly ? "
In all these cases we can see that the cause of the
actions referred to in the ' why* is inextricably bound
up with their aim and with the intention which has
directed them. The phenomenon is the same as in the
* whys' relating to nature, but in this case it is justified,
since these f whys ' relate to human actions. We may
therefore assert that among all the questions of the
child, "whys of motivation" are those which are the
most correctly expressed and the least removed from
our own manner of thinking.
Between these 'whys' and those relating to momentary
intentions there is naturally every intermediate shade of
meaning. For instance: "I like men who swim that
way. Why?" " Why are you not pleased tJiat I should
have killed him?" Thus it would appear possible to
establish two sub-categories among the "whys of motiv-
ation," one relating to momentary intentions, the other
to psychological states of a more lasting character. The
distinction, however, is unimportant. What would be
more interesting would be to bring greater precision
into the relation between "whys of motivation" and
"whys of justification." At times, it seems as though
igo STUDIES IN CHILD LOGIC
the explanation required by the child as an answer to
his ( why ' were something between logical explanation
(one idea bringing another idea in its wake) and psycho-
logical explanation (a motive bringing about an action).
For example : " Do you like mice best, or rats? Why?
Because they are not so fierce and because you are weak"
Cases like these help us to understand how "whys of
logical justification," which we shall study presently,
have gradually separated themselves from " whys of
motivation."
To the "whys of motivation" must also be added a
fairly abundant group of ' whys' (34 out of 183): we
mean those which the child expresses simply in order
to contradict a statement or a command which annoys
him. If these questions are taken literally and seriously,
they would seem to constitute "whys of motivation"
properly so-called, and at times even "whys of logical
justification " in the sense we have just instanced. But,
as a matter of fact, we are dealing here, not with genuine
questions as before, but with affirmations, or rather dis-
guised negations, which assume the form of questions
only as a matter of politeness. The proof of this is,
that the child does not wait for an answer. Here are
some examples: "Anita wouldn't, so I hit her You
should never hit a lady Why? She isn't a lady ..."
etc. "Up to here Why?" "Draw me a watch.
Why not cannons, etc." The child is apparently asking :
"Why do you say this?" or "Why do you want
this?" etc. Asa matter of fact the question simply
amounts to saying: "That's not true" or "I don't
want to." But it goes without saying that between
* whys ' of contradiction and those relating to intentions
there is a whole series of transitional types.
Finally, mention must be made of a class of * whys '
which hover between "whys of motivation" and
"whys of causal explanation," and which may be
called whys of invention. In these the child tells stories,
or personifies in play the objects which surround him,
A CHILD'S QUESTIONS 191
and It is in connexion with this romancing, that he
asks questions which, incidentally, do not admit of any
possible answer: " Why do you do that [rub with an
india-rubber] to the poor little table? Is it still old?"
" Do you know why I don't kill you? Ifs because I don't
want to hurt you "
5. "WHYS OF JUSTIFICATION." " Whys of justifi-
cation " are interesting in many connexions. They are a
sign of the child's curiosity about a whole set of customs
and rules which are imposed from outside, without
motive, and for which he would like to find a justifica-
tion. This justification is not a causal, nor even a
strictly finalistic explanation. It is more like the motiv-
ation of the last group which we described, but is to
be distinguished from it by the following character-
istic: what the child looks for under the rules is not
so much a psychological motive as a reason which
will satisfy his intelligence. If, therefore, we place
the * whys' of this category in a special group, it is
because they form the germ which after the age of 7
or 8 will develop into "whys of logical reason." In
the case of Del we can even see this gradual formation
taking place.
Del's "whys of justification" can be divided into
three sub-groups easily distinguishable from one another.
They are ' whys ' relati-ng i to social rules and customs,
2 to rules appertaining to lessons learnt in school
(language, spelling), and 3 to definitions. Of these
three, the third alone contains "whys of logical
reason." The first is still closely connected to pycho-
logical motivation, the second constitutes an inter-
mediate group.
Out of the " 74 whys of justification," 14 relate to social
customs. Among these, some point simply to psycho-
logical curiosity and might just as well be classed
under "whys of motivation." For example: " Why
in some churches are the gowns black, and [in] others they
are coloured?" Others come nearer to the idea of a
192 STUDIES IN CHILD LOGIC
rule: " Why is it forbidden [to open letters]? Would
he [the postman] be sent to prison? " etc.
This first group, as may be seen, is hardly In the
right place among " whys of justification." If we have
classified it in this way, it is simply because it is con-
nected through a chain of intermediate links with the
' whys ' relating to scholastic rules. Here is a transi-
tional case : "Why not ( in ' [in Alain]," or in connexion
with the spelling of 'quatre': "k? No Who said
not) was it the grown-ups in Paris ? " The ' grown-ups '
who settle the spelling of words are thus more or less
on the same level as those who make police regulations
and send postmen to prison.
The * whys' genuinely relating to scholastic rules
(55 out of 74) are much further removed from the ' whys '
of psychological motivation. Here are some examples :
" Why [are proper names spelt with capitals]? /
want to know" "You must always put a 'd j at the
end of ' grand* Why^ what would happen if you didrft
put any?" " Why is it ['bonsoir'] not spelt with a l cj
that makes coi" "You don't have to put a dot on a
capital I. Why?" " Why do you put full- stops here
[at the end of sentences], and not here [at the end of
words] ? Funny ! "
It is well known that in spelling and in grammar
children are more logical than we are. The large
number of " whys of justification " furnishes additional
proof of this. They are the exact parallel of the " whys
of causal explanation " with which we have already dealt.
Language, like nature, is full of freaks and accidents,
and the explanation of these must be sui generis and
must take into account the fortuitous character of all
historical development. The child, devoid alike of
the notion of chance and of the notion of historical
development, wants to justify everything immediately,
or is surprised at his inability to do so.
If we lay stress once more upon this rather trivial
fact, it is because these "whys of justification/' added
A CHILD'S QUESTIONS 193
to the already abundant "whys of causal explanation," and
showing the same tendency to justify at any cost, make it
all the more extraordinary that DeFs questions should be
so poor in " whys of logical reason." One would have
thought that since Del and the children of his age are in-
clined to justify everything, their language would be full
of deductive arguments, of the frequent use of ' because '
and 'why ' connecting one idea to another, and not a fact
to an idea or a fact to a fact. But this is not in the least
what happens. Out of the 74 "whys of justification/ 5
only 5 are " whys of logical justification or reason." It
is needless to repeat the reason for this paradox: the
child is not an intellectualist, he is an intellectual realist
Let us rather try to analyse the nature of this logical
justification, and find out how it differs from other
'whys.' ' Whys' relating to language furnish us with
several transitional cases along the path leading to the
true logical 'why.' These are the etymological 'whys':
" Why do you say ( strayed* when it means lost? " " Why
are there lots of words with several names, the lake of
Geneva, Lake Leman" "Why is it [a park in Geneva]
called Mon Repos" "Why c black coffee ,' all coffee is
black?" Just at first, it looks as though these were
genuine "whys of logical justification," connecting a
definition to an idea which serves as a reason for it.
This is true of the last of these ' whys,' which we shall
therefore class along with four subsequent examples
under logical justification. But the others aim chiefly
at the psychological intention. They are, moreover,
still tainted with intellectual realism. It is well known
that, for the child, the name is still closely bound up
with the thing ; to explain an etymology is to explain
the thing itself. Del's slip, "words with several names 3 *
is significant in this connexion. 1 Therefore we cannot
1 A child of whom we asked : "Have words any strength ? " answered that
they had if they denoted things that had strength, not otherwise. We asked
for an example. He mentioned the word 'boxing.' "Why has it got
strength" " Ok no! I was wrong" he answered, *'/ thought it was the
word that hit!"
N
194 STUDIES IN CHILD LOGIC
talk here about one idea being connected to another:
the ideas are connected to the objects themselves.
The only cases, then, in which one can say that there
is logical justification are cases of pure definition,
and cases of demonstration, in which the mind tries
to establish a proof in such a way as to render strict
deductions possible.
In definition, the question falls under the following
schema: "If you call all objects having such-and-
such characteristics #, why do you call this objects?"
Here the connexion is really between one idea and
another, or, to speak more accurately, between one
judgment recognized as such (an x is . . . etc.) and
another (I call such-and-such an object #), and not
between one thing and another. This distinction,
however subtle it may appear, is of the greatest im-
portance from the point of view of genetic psychology.
Up till now, the mind has dealt solely with things and
their relations, without being conscious of itself, and
above all without being conscious of deducing. In
logical justification, thought becomes conscious of its
own independence, of its possible mistakes, and of its
conventions, it no longer seeks to justify the things in
themselves, but its own personal judgments about them.
Such a process as this appears late in the psychological
evolution of the child. The earlier chapters have led
us not to expect it before the age of 7 or 8. The small
number of "whys of logical justification" asked by
Del confirms our previous treatment of the subject.
Similarly, in all demonstration the connexion holding
between < because' and i why' relates to judgments and
not to things. In the following example : " Why does
the water of the Rhone not flow upwards?" if an ex-
planation is expected the answer must be: "Because
the weight of the water drags it along in the direction
of the slope." But if a demonstration is expected, the
answer must be: "Because experience shows that it
does" or "Because all rivers flow downwards." In
A CHILD'S QUESTIONS 195
the first case the connexion connects the direction of
the water to the downward slope, it relates to the actual
things themselves and is causal. In the second case
the connexion relates to the judgments as such, and is
logical. Therefore all 'whys' of demonstrations are
" logical whys." But demonstration rarely operates
before the age of 7 or 8. The first two chapters showed
us that in their arguments, children abstained from any
attempt to check or demonstrate their statements.
In short, " logical whys " can by rights relate to any-
thing, since they include all 'whys' which refer to
definitions or demonstrations. Here are the only
questions of Del's which can be said to belong to this
group (in addition to the example about the black
coffee which we have just recalled) :
u Why [do you say 'torn cat']? A she-cat is a
mummy cat. A cat is a baby cat. . . . / want to write
* a daddy cat^ " " They are torrents Why not rivers ? "
"That isn't a bone, it's a bump Why? If I was
killed, would it burst?" " Is that snow? [question of
classification] No, it is rocks Then why is it white?"
The last of these ' whys ' is ambiguous ; it is probably
an elliptical form, meaning: "why do you say it is
rocks, since it is white?" At the same time, it may
very well be a simple "why of causal explanation."
There are therefore only four authentic " whys of logical
reason." They can be recognized by the fact that
under the interrogative word itself the phrase " why do
you assert that ..." can be understood; and this is
never so in the other categories. In a word, " whys of
logical justification " look for the reason of a judgment which
is recognised as such, and not of the thing to which the
judgment relates. < Whys ' of this kind are therefore
very rare before the age of 7 or 8. The child, while he
tries to justify everything, yet neglects to use the one
legitimate justification, that of opinions and judgments
as such. After the age of 7 or 8, however, these
questions will probably be more frequent We have
196 STUDIES IN CHILD LOGIC
fixed at 11-12 the age where formal thought first makes
its appearance, /.*., thought relating to hypotheses
which are held as such, and only seeking to ascertain
whether the conclusions drawn from these hypotheses
are justified or not, simply from the point of view
of deduction and without any reference to reality. 1 Be-
tween the period of pure intellectual realism (up till
7 or 8) and the beginnings of formal thought there
must therefore be an intermediate stage, in which
children try to justify judgments as such, yet without
for that matter being able to share the interlocutor's
point of view nor, consequently, to handle formal
deduction. The presence of " whys of logical justifica-
tion " must correspond to this intermediary stage.
In conclusion, the results of this section confirm
those reached in our study of the " why of causal
explanation." In the case of Del, there are no more
"whys of logical reason" than there are "whys" of
pure causality. Consequently, Del's mind must have
interests which are intermediate between mechanical
explanation and logical deduction. It is in this failure
to distinguish between the causal and the logical point
of view, both of which are also confused with the point
of view of intention or of psychological motive, that we
see the chief characteristic of childish precausality.
Finally, it may be of interest to point out a curious
phenomenon which supports the hypothesis that the
child often confuses notions which in our minds are
perfectly distinct. The peculiarity we are speaking of
is that Del occasionally takes the word ' why ' in the
sense of ' because/ and thus uses the same word to
express the relation of reason to consequence and that
of consequence to reason. 2 Here is an example which
happens precisely to be concerned with a logical
1 J. Piaget. " Essai sur la multiplication logique et les debuts de la pense
formelle chez I'enfant. Journ. Psych., 1922, pp, 222-261.
2 This confusion of why and because is easier to understand in the case of
French children. There is a certain degree of assonance between * pourquoi *
and f parce que,' [Translator's note.]
A CHILD'S QUESTIONS 197
'why' or * because': " Rain water is good Is it 'why
( = because) it is a spring? " Now, this is a phenomenon
which we have already noticed in connexion with ex-
planations between one child and another (Chapter III,
5), and which we shall meet with again in our study
of the conjunctions of causality (Vol. II). It occurs
frequently in ordinary life in children from 3 to 6.
We remember in particular a little Greek boy of 5 years
old who learnt French very well, but systematically
used the word 'why' instead of the word 'because'
which is absent from his vocabulary: " Why does the
boat stay on top of the water?" " Why ( = because)
it is light" etc. As a matter of fact this phenomenon
indicates only a confusion of words. But this confusion
shows how hard it is for a child to distinguish between
relations which language has differentiated.
6. CONCLUSIONS. The complexity of Del's * whys '
will now be apparent, as will also the necessity for classi-
fying them partly according to material, since it is
impossible to say straight away to which type of rela-
tion (strictly causal, finalistic or logical explanation,
etc.) they refer. The frequencies obtained out of our
360 * whys ' are summed up in the following table:
Whys of causal
explanation
(in the wide sense)
Numbers (roughly)
f Physical objects 26
Plants 10
Animals 29
Human body 16
Natural objects 81 22%
Manufactured objects 22 6%
Total 103 29%
(Properly so-called 143
Contradiction 34
Invention 6
Total 183 5o%
[Social Rules 14
Scholastic Rules 55
Whys of justification | Rules 69 19%
Logical reason or justification 5 i%
Total 74 21%
198 STUDIES IN CHILD LOGIC
Thus the "whys of motivation" outnumber all the
others. Does this preponderance indicate that the
other types of * why ' radiate from this group as from
a common centre? This would seem to be the case,
for the " whys of causal explanation " are connected with
motivation through a whole series of anthropomorphic
'whys, 1 finalistic 'whys,' and < whys ' which reveal
precausality itself. The "whys of justification,'' on
their side, are connected with those of motivation by
the series of ' whys ' relating to social usage and to
rules conceived of as obeying psychological motives.
The relations between the two groups of causal ex-
planation and justification are not so close. The idea of
precausality certainly presupposes a confusion between
causal explanation and logical justification, but this
confusion is only possible owing to the fact that both
are, as yet, insufficiently differentiated from psycho-
logical motivation. In a word, the source of Del's
'whys' does seem to be motivation, the search for an
intention underlying every action and every event.
From this source there would seem to arise two divergent
currents, one formed of ' whys ' which try to interpret
nature as a thing of intentions, the other formed of those
which relate to customs and to the rules associated
with them. Between those finalistic * whys' and the
* whys ' of justification interaction would naturally be
possible. Finally, causality proper would emerge from
the ' whys ' of precausality, and true logical justification
from the ' whys ' of justification. Such, approximately,
would be the genealogy of the whys asked by Del. We
shall try to sum it up in a table.
Is such a systematization as this the result of an
individual mentality of a particular type, or does it
mark the general character of child thought before 7
or 8 years of age? The answer to this question will
have to be supplied by other monographs. What
we know of other lines of research leads us to believe
that the schema is a very general one, but this sup-
A CHILD'S QUESTIONS 199
position must serve for the present only as a working
hypothesis.
GENETIC TABLE DEL'S WHYS
Cat/salt ft/
r properly So-called
r Customs, ru/es efc.
\.
property so-ca//e</ ' ~ ^>jusHficatfon
II. QUESTIONS NOT EXPRESSED UNDER
THE FORM OF 'WHY.'
Let us now approach the whole problem presented by
Del's questions. The ' whys ' served as an introduction,
forming as they did a clearly defined, partly homo-
geneous group, and capable of being classified in a
schema. The moment has now come to verify this
schema and to complete it with any information which
may be supplied by Del's other questions.
7. CLASSIFICATION OF DEL'S QUESTIONS NOT EX-
PRESSED UNDER THE FORM 'WHY.' It is even more
difficult than in the previous case to classify Del's
questions simply according to the material which is
the object of the child's curiosity. The same object,
say, a physical phenomenon, can give rise to questions
too widely different from one another: " When did it
happen." "How?" "Is it true that . . ." "What
is it that ..." We shall therefore have to use a mixed
classification, which will partly coincide with and partly
extend beyond that which we adopted in connexion
with the 'whys.' The important thing to remember
now is that for every < why ' there may be a correspond-
ing question of another form having the same meaning,
but that the converse does not hold.
200 STUDIES IN CHILD LOGIC
A first group is formed by questions of causal explana-
tion, these words being used in exactly the same sense
as above. Here are some examples: " [Talking of a
marble which is rolling down a slope] What is making
it go?" " What makes the lake run?" " Do you have
to have afire to make india-rubber?" To these will pro-
bably have to be added questions of the following form :
"What is he for?" [a greyhound], given the close
connexion between finalism and causal explanation in
the child's mind. Some of these questions are there-
fore exactly analogous to * whys/ others are asked from
a different point of view, but always referring to ex-
planation, either causal, precausal, or finalistic.
A second group, also very important and earlier in
appearance than the former (at least as regards ques-
tions of place), is that which we shall call questions
of reality and history. These questions do not relate
to the explanation of a fact or of an event, but to its
reality or to the circumstances of time and place in
which it appeared, independently of their explanation.
It is not: "What is the cause of :r?" but "Did x
happen, or will it happen?" or "When did or will
x happen?" or again "Where did x happen?" etc.
Such a class of questions obviously has no equivalent
among 'whys, 5 since the function of these is to relate
to the motive or the reason of facts and events, and
never simply to their history or their existence. Here
are a few examples : " Does he [the fish] find food? "
"Are there really any [men who cut up children]?"
" How soon is Christmas? " " Is Schaffhausen in Switzer-
land?" As will be seen, this type of question can have
a great many different shades of meaning history,
time, place, existence, but the central function remains
obvious. The criterion for determining whether a
question belongs to this group is therefore as follows.
Whenever a question relates to an object, a fact, or an
event other than a person or a human action, and when-
ever the child asks for neither the cause, the class, nor
A CHILD'S QUESTIONS 201
the name of this object, the question belongs to the
present category.
A third group resembles the last two in form, and
probably constitutes their common source ; it is the sum
of questions asked about personal activities, and about
persons themselves, excluding those about their names
or about social or scholastic rules. These are therefore
questions relating to human actions and intentions. At first
it might seem as though one could subdivide this group
into two smaller groups, one including questions about
the causes of actions and corresponding to < whys ' of
motivation, the other including questions about the
actions themselves, independently of their cause, and
corresponding to the last group of questions (reality
and history). But as a matter of fact these two points
of view merge imperceptibly into one another, and
cannot be made too rigidly separate. On the other
hand, it is useful to put in a class apart everthing
that concerns human actions, and thus to separate the
questions we are now dealing with from those discussed
in the last section. Here are some examples: "Did
you want to sit here this morning?" Yes "Because it
isn't fine to-day? " " Will you come? Perhaps " u May
I eat this pear? " " Would you rather have an ugly face
or a pretty face?" It is clear that the first of these
questions is exactly analogous to a "why of motiva-
tion " ; the others depart from psychological explanation,
and concern themselves more closely with matters of
fact, without any thought of cause. Nevertheless, the
group is quite homogeneous.
A fourth group is that which corresponds to certain
of the " whys of justification " ; these are all questions
relating to rules and usage; "How is it [a name]
written?" etc.
We can also distinguish a category of arithmetical
questions^ but its number is very small. The form
which these questions take is, for example: "How
much is 9 andg?"
202 STUDIES IN CHILD LOGIC
Finally, there is a whole group of questions of classifica-
tion and valuation, relating to the names of objects, to
their value, to the classes, to which they belong, and
to comparisons between them. We have placed in this
group questions implying judgments of value (evalua-
tion), for between the class referred to, in " Is it big?"
and the value in " Is it pretty?" we can find a whole
series of intermediate cases. Here are examples: " 1$
that a bee?" "Is that mountains?" " What are those
balls with 1 [quavers]?"" What is a cup?"" That's
pretty ', isn't it?" etc. This group of questions raises
certain borderland problems. It is occasionally hard
to decide whether a question should be classed in this
group or among the questions of reality or history, in
spite of the fact that in principle the criteria are quite
definite. There are, moreover, many transitional cases
between rules about names and classification, so that
the distinction between this group and the preceding
one is sometimes a very delicate matter.
In the main, however, these groups correspond to
certain fundamental functions of the mind, which are
distinct from one another, and for which when we
come to examine them in greater detail, we shall find
it quite possible to establish reliable criteria.
8. QUESTIONS OF CAUSAL EXPLANATION. Let us
try to verify by means of these questions the results
obtained from the corresponding ' whys J such as arti-
ficialism and finalism of questions relating to natural
objects, absence of any purely causal relations, etc.
We give from among the questions asked about
physical objects, those which seem to us the least
ambiguous :
[Del sees a marble rolling down a sloping piece of
ground]. " What is making it go?" It is because the
ground is not flat, it is sloping and goes down [a
moment later] "It [the marble rolling in the direction
of Mile V.] knows that you are down there?" [A few
seconds later]: "It is on a slope, isn't it?" (This
A CHILD'S QUESTIONS 203
question ought to be placed among questions of reality,
were it not for its obviously causal significance).
" What makes the Rhone go so fast?" "But what
makes the lake run?" [A few months later] " Ifs funny -,
the ground [here] is quite flat, how can it [the water] go
down ? "
"How do you make one [a river source]" "Do you
also have to have a spade to -make a source?" u But how
is the rain made up in the sky? Are there pipes, or running
torrents?" " [after the explanation] Then it comes un-
stuck? Then when it falls, it is rain?" " After that it
[the river] becomes a glacier?" "But it [the glacier]
melts , all of a sudden you dorft see it any more? " u Then
clouds often drop down [on to the mountains]?"
" [Talking about a magnet] I would like to know how
it happens?" "Look, it Attracts it [a key]. What makes
it move forward? "
* * But what does the snow do when we go tobagganing ?
Instead of melting, it stays nice and fiat? "
Thus the only questions of a truly causal nature are
those relating to phenomena for which a mechanistic
explanation has already been given to Del (function of
the slope, etc.). Now Del had put quite a different
interpretation upon these phenomena, as is shown by
his questions before the explanation. The last two
questions can indeed be regarded as instances of
mechanical causality, but with the following reserva-
tions. In the first place, the verbal form "to do"
(what does the snow do) should be noticed. Psycho-
logists have often been struck by it, M. Btihler, for
instance, rightly concluding from the frequent occurrence
of the verb machen that the child attributes anthropo-
morphic activity to ordinary objects. 1 But this may
be a mere aftermath of earlier stages, since verbal forms
always evolve more slowly than actual understanding.
What is more singular is that in both cases (questions
about the magnet and about the snow) the child seems
to be looking for the explanation in some internal force
residing in the object, and not in any mechanical
1 K. Btihlei ', Die geistige Entwicklung des Kindes,} ena, 1921, 2nd ed. p. 387.
204 STUDIES IN CHILD LOGIC
contact. Del may say that the magnet attracts the key,
this does not really satisfy him. Similarly, there is
surely a latent idea of force in the fact that the snow
does not melt. If these questions are causal, it is
obvious that the causality they invoke is more dynamic
than mechanical.
This dynamism is strikingly expressed in the question
about the marble : " Does it know that you are down
there?" The minimum hypothesis, so to speak, is
that we have here a question of romancing: Del
personifies the marble in play, just as, in his games,
he will lend life to a stone or a piece of wood. But
to say ' romancing ' takes us nowhere. We may well
ask, in a problem of this kind, whether the child could
do anything but invent. This leads us to the maximum
hypothesis: Does not Del attribute to the marble a
force analogous to that of a living being? A very
curious question about dead leaves will presently show
us that for Del, life and spontaneous movement are
still one and the same thing. 1 There is therefore
nothing surprising in the fact that the same question
should arise in connexion with a marble, the l why ' of
whose movements is not yet understood by Del. Even
if Del is romancing about the marble itself, the fact
that he should ask the question in this form and with
apparent seriousness is an index of the child's lack of
interest in mechanical causality and of his inability to
be satisfied by it. A case like this takes us to the very
root of precausal explanation. The child confuses
moving cause and motive because for him, phenomena
are animated with real life or with a dynamic character
derived from life.
Other questions endow men or gods with the power
of making river sources, rain, etc., by means of purely
human contrivances. Whether this ' artificialism ' as
1 A recent enquiry, of which we are not yet able to publish the results,
shows that Geneva boys up to the age of 7-8 look upon stars, fire, wind, and
eventually water, etc., as alive and conscious, because they move by themselves.
A CHILD'S QUESTIONS 205
M. Brunschwicg has called it, is earlier than the last-
mentioned type of causality or is derived from it is a
question which we do not wish to settle here, and one
which, incidentally, is outside the range of our subject.
We will content ourselves with pointing out that Del
does not generally try to find out exactly who is the
manufacturer of such-and-such a phenomenon (with the
exception of river sources and the Rhone). Most of
the corresponding whys should therefore be interpreted
as simply looking for an intention in phenomena, with-
out this intention being attributed to a given being.
Which brings us once again to precausal explanation
and the confusion of motive and mechanical cause.
From this point of view, it is possible to suppose that
animism preceded artificialism, both in the child and
in the race.
In short, these questions about physical objects, of
which only a very few admit of a genuinely causal
interpretation, confirm and define our hypothesis of
precausality by linking it up with the well-known
phenomenon of animism among very young children.
It may be thought that we are dealing too summarily
with these various types of childish explanations and
their affiliations, and that they require to be more
searchingly analysed, and compared with materials from
other sources. But our aim, be it said once more, is
not the analysis of causality, but the study of child
logic, and from that point of view it is sufficient for our
purposes to know that logical implication and physical
causation are as yet undifferentiated, and that this identi-
fication constitutes the notion of precausality.
The childish conception according to which moving
objects are endowed with an activity of their own gives
a special importance to Del's questions about life and
death. The reader will remember the result of our
study of * whys' relating to animals and plants. It
was that since chance does not exist for the child, and
all phenomena appear to be regulated by order and
206 STUDIES IN CHILD LOGIC
'decus,' life is a perfectly normal phenomenon, without
any elements of surprise in it, up till the moment when
the child takes cognizance of the difference between life
and death. From this moment, the idea of death sets
the child's curiosity in action, precisely because, if
every cause is coupled with a motive, then death calls
for a special explanation. The child will therefore look
for the distinguishing criteria of life and of death, and
this will lead him in a certain measure to replace pre-
causal explanation and even at moments the search for
motives by a conscious realization of the accidental
element in the world.
"Are they dead [those leaves]? Yes But they move
with the wind" "Is it [a leaf which Del has just cut
off] still alive now? ..." [He puts it back on the
branch] "Is it alive now?" "If it is put in water? It
will last longer Another day, and then? It will dry up
Will it die? Yes Poor little leaf! If it [a leaf] was
planted in blood, would it die [too] ? "
" Was that [a \.rz\ planted, or did it grow by itself? "
" What makes the flowers grow in summer? " "Daddy told
me that wystaria grows two seasons, spring and summer.
Then does it grow twice? "
The first of these questions shows us the confusion
between movement and life. This confusion is of very
great importance in understanding the precausal mind.
It enables us to see how, for the child, every activity is
comparable to that of life. Henceforth, to appeal to a
motive cause is at the same time to appeal to a living
cause, i.e., to one mentally based on a model endowed
with spontaneity if not with intentions. We can now
understand how a return shock will give rise to curiosity
about death, since the fact of death is an obstacle to
these habits of thought; and to curiosity about the
causes of life (later questions quoted above) since the
course of life can be disturbed by death.
Questions about animals point to the same preoccupa-
tions, likewise those about their powers and intentions.
A CHILD'S QUESTIONS 207
" If 'you kill him [a pigeon] at this little corner of his
wing, will he die?" " Does it [a caterpillar] know it has
got to die if it becomes a butterfly?" (cf. " Doesrit it [id.]
know it has to die very soon ? "), etc.
" [Reindeers draw sledges] Are they persons^ so that
they can hear what people say to them?" "[A moment
later] : How are horses driven?" " What is a greyhound
for?" etc.
The human body gives rise to analogous questions 1 :
" Do you die [of eating a chestnut] ? " " If 'you breathe
poison^ do you die?" etc." Who makes those little spots,
how [freckles on the arms]?"
We need press the point no further. All these
questions show that the order of causality imagined by
the child is hardly mechanical at all, but anthropo-
morphic or finalistic.
Finally, we should mention here a group of questions
about manufactured objects which is analogous to the
corresponding group of 'whys.'
" What are rails for? " "Is that machine there [which
is sifting sand] not of any use to that one [a crane] ? "
"If I have a boat and it is dipped in water and put in
the sun, it will stick again worit it?" " If you fire a
cannon on to a fire-work . , . the shell will go on to the fire
and it will burst, worit it? "
In conclusion, this section may be said to verify the
hypothesis advanced in connexion with "whys of causal
explanation," in particular as concerns the rareness of
strictly causal questions.
9. QUESTIONS OF REALITY AND HISTORY. Ques-
tions of this category are, by definition, those which
relate to facts and events, without relating to their cause
or to their causal structure. This criterion is not an
easy one to apply, and this group of questions merges
into the last by a whole series of intermediate stages.
1 Cf. "Is he dead" (a Geneva statue; Cartaret) Yes "^a// / be dead
too?" etc.
208 STUDIES IN CHILD LOGIC
There is nothing in this that need surprise us, since the
very notion of reality comes into being thanks only to
the relations of causality which the mind weaves be-
tween the facts of experience. Still, as we are in need
of reliable classifications about which every one can
agree, we shall have to adopt a criterion arbitrary, we
admit, but definite.
When the question which has been asked calls for
a causal answer, i.e., one beginning with the causal
( because/ or consisting of the phrase : " It was by God "
or " by man," then this question is undoubtedly causal.
But we have tacked on to this group such questions as :
"If you breathe poison do you die?" or "Does the
marble know you are down there?" which seem to be
questions of fact. In such cases, the criterion is more
subtle. These questions certainly do touch on causality,
since they amount to asking whether poison does or
does not cause one to die, whether the marble rolls in
a certain direction because it is conscious, or for any
other reason, etc. Whereas a question of pure fact like
this one : " Are there also little fishes round the edge? "
does not involve any search for the causal relation, nor
any use of it. We shall therefore adopt the following
convention for lack of a better one. When the relation
between the terms referred to in the question implies a
movement, an activity, or an intention, the question is
causal ; when the relation is purely static (existence,
description, or place) or simply temporal, the question
is not causal.
It is only by applying them, that we shall find out
whether these arbitrary distinctions are of any use. If,
for example, by applying them to several children of
different ages, one discovers a law of development, or
finds a means of distinguishing between different types
of inquisitive children, then the schema will be worth
preserving. Otherwise, it will share the fate of the
classifications of the grammarians and old logicians.
For the problems which occupy us here, problems of
A CHILD'S QUESTIONS 209
general and not of individual psychology, the schema
is of no importance whatsoever.
Questions of reality and history should, moreover, be
divided into various categories of which the first alone
is in any way likely to be confused with questions of
causal explanation. We mean the questions about facts
or events.
"Is it [this pool] very deep?" "/ can see myself in
your eyes, can you?" "Are there also little fishes round
the edge? " "Do they [rockets] go up as high as the sky? "
(This question can also be classed as about place)
" [Are the clouds] much, much higher than our roof?
yes I cant believe it!" (id.): " What is in there [in a
box]?' 5 "Are there whales [in the lake]?" [Looking at
a geographical map with the lake Zoug which he takes
for a hole]: "Are there holes?" "Are its [a snail's]
horns outside? " " Are there blue and green flies ? " etc.
This first group merges by a series of intermediate
stages into a second category which relates more
especially to place :
" Where do the big boats land?" " Where is German
Switzerland?" " Where is the Saint Bernard" " Then
Zermatt is not in Switzerland? " etc.
A third category consists of questions about time \
"How long is it till Christmas?" " Will my birthday
be on Monday ? I think it is Monday. Is it really ? I don't
know I thought grown-up people could think about these
things " [Note that the idea of " thinking about a thing "
is confused with that of " knowing," a mistake frequently
made by children].
A fourth group is made up of questions of modality,
z\e. y of those relating, not to facts and events, but to
their degree of reality. Between this group and the
others, there are naturally many intervening shades of
stages, but it is interesting, all the same, to consider it
by itself.
210 STUDIES IN CHILD LOGIC
"Are there really any? [men who cut up little
children]?" "Is it true [that it is poison]?" "Isn't
that a story?" etc. Along with these questions of
modality, mention may be made of the following state-
ment uttered by Del, which is none the less suggestive
for having undoubtedly been prompted by the tendency
to romance: "Jean [a friend] does not exist, because I
don't like him ! "
Finally comes the whole group of questions of im-
agination or invention relating to facts and events which
Del knows to be untrue.
" Is the little girl burnt up, now?"" That Mr J. [the
capital letter J] has eaten a lot, hasn't he?" "Are the
waves on the lake unkind?" [Is this romancing, or does
Del mean " are they dangerous? "]
These are the five categories into which it is possible
to divide Del's questions of reality and history. As
such, these questions tell us little that we do not already
know. Their chief value is that they partly enable
us to enquire into the nature of childish assumptions.
Meinong, it will be remembered, described as assump-
tions those propositions about which the subject reasons,
although he does not believe them. He showed that
assumption originated in the 'if of childish games,
in the affirmation which the child chooses to take as
the basis of imaginary deductions.
Now these assumptions give rise to a very serious
problem : what is the degree of reality which the child
attributes to them? The adult has, amongst others,
two kinds of assumptions at his disposal physical and
logical. Physical assumption is that which assumes a
fact as such, and deduces from it a relation between one
fact and another. " If the sun were to disappear, we
should not see any more," for instance, means that
between one given fact (disappearance of the sun) and
another fact (night) there is a relation of causality.
Logical assumption, on the other hand, simply assumes
a judgment as such, and deduces another judgment
A CHILD'S QUESTIONS 211
from it: " If all winged vertebrates are to be called
birds, then the bat is a bird." Here the relation is no
longer between two facts but between two judgments.
Now, of all the assumptions made by Del, not one
is logical. We have already given instances of assump-
tions in questions of causal explanation (" If you breathe
in poison, do you die ? ") Others refer to psychological
motivation (" Would your mother be sorry if they rang her
up and told her you were dead? Yes, she would come
and fetch you. . . . etc. And if I had gone away?
She would tell the police And if the police did not find
me p They would find you But if 'they didn't. . . . etc.")
Others refer to social usages ("Do the policemen forbid
it [tobogganing in the streets]? Yes. If I were a judge
could I do it?") For the most part they are questions
of reality and history, in which the child alters reality
at will, to see what would happen in such-and-such
conditions. They are " experiments just to see," the
work of imagination such as Baldwin describes it,
whose function it is to loosen the spirit from the bonds
of reality, leaving it free to build up its ideas into a
world of their own. For example :
(t If I was an angel, and had wings, and was flying in
the fir-tree, would I see the squirrels or would they run
away ? "
( ' If there was a tree in the middle of the lake, what
would it do?" But there isn't one. "I know there
isn't, but if there was . . . (id)."
"But then supposing it [the round of the seasons]
stopped one day ? "
" If I was to put a dragon and a bear together which
would win ? . . . And if I put a baby dragon ? "
All these assumptions, it will be seen, are physical ;
they are, in the words of Mach and Rignano, "mental
experiments." This therefore confirms our hypotheses
that before the age of 7-8 children do not care to deal
with logical relations. But it does more than this.
Childish assumptions point to a confusion between the
212 STUDIES IN CHILD LOGIC
logical and the real order of things, just as precausality
confuses logical implication and causal explanation.
In other words, the child, thanks to the notion of pre-
causality, conceives the world as more logical than it
really is. This makes him believe it possible to connect
everything and to foresee everything, and the assump-
tions which he makes are endowed in his eyes with a
richness in possible deductions which our adult logic
could never allow them to possess.
Indeed, the outstanding characteristic of children's
assumptions is that for us they contain no definite con-
clusion, such as they should contain for the children
themselves. We do not know what would happen " if
I were an angel," or " if I was to put a dragon and a
bear together." Del would like to know. He thinks
it possible to reason where we deem it impossible, for
lack of data. Since everything in nature seems to
him constructed, intentional, and coherent, what more
natural than that there should be an answer to every
1 if? The structure of childish assumptions therefore is
probably analogous to that of precausality confusion
of the causal or physical order (the real) with the logical
or human order (motivation).
The real, as we have seen, can in the last resort
be deformed at will by Del ("Jean does not exist because
I dorit like him"}. Thus childish assumptions deal with
a reality which is far more fluctuating than ours, one
which is perpetually shifting its level from the plane of
observation to that of play, and vice versa. In this
respect, reality is for the child both more arbitrary and
better regulated than for us. It is more arbitrary,
because nothing is impossible, and nothing obeys causal
laws. But whatever may happen, it can always be
accounted for, for behind the most fantastic events
which he believes in, the child will always discover
motives which are sufficient to justify them ; just as
the world of the primitive races is peopled with a wealth
of arbitrary intentions, but is devoid of chance.
A CHILD'S QUESTIONS 213
The result of this is that the idea of the possible is
far less precise in the child than in the adult. For the
adult, the possible is from one point of view simply a
degree of reality (physical possibility), and from another
point of view, the sum of all logical assumptions
(hypotheses forming the basis of logical deduction).
Thus the adult will be able to make an indefinite
number of deductions in the realm of the possible, or
of hypothesis, so long as he is able to conform to the
rules of logical deduction. But any illusion he may
have of building up reality in this way will disappear
the moment he remembers that the world of hypotheses
is subject to the world of observation. The child, on
the other hand, never makes logical assumptions, so
that for him the possible, or the world of hypotheses is
not one that is inferior to the real world, not a mere
degree of being, but a special world of its own,
analogous to the world of play. Just as the real is
crammed with motives and intentions, the possible is
the world where those intentions are laid bare, the
world where one can play with them unhindered and
unchecked. Hence those chains of suppositions which
we saw above: u If ... yes, but if ... yes, but
if ..." etc.
The possible is therefore not a lower degree of being,
it is a world apart, as real as the other, and an assump-
tion does not differ from a simple induction made about
the real world.
For the rest and this is the last point to be borne in
mind Del's deduction, like that of all children of his
age, is not pure (formal deduction), but is still deeply
tinged with intellectual realism. What constitutes the
validity of adult deductions based on assumptions (and
this applies to all demonstration) is that deduction
confines itself to connecting judgments to judgments.
In order to demonstrate the judgment already quoted.
" If the sun disappeared we should not see any more,"
one must perforce resort to logical assumptions and
214 STUDIES IN CHILD LOGIC
deductions of this kind: If (you allow that) daylight
is not due to the sun, then (you will have to admit that)
there must be daylight after sunset, because . . . etc."
Del, on the other hand, never attempts to demon-
strate. He never makes logical assumptions to see
where they would lead ; he reasons directly on the
imaginary model which he has made, and which he
regards as real.
In conclusion, these questions about reality corr-
oborate what we learnt from the questions of causal
explanation. The child shows signs of a perpetual
intellectual realism : he is too much of a realist to be a
logician, and too much of an intellectualist to be a pure
observer. The physical world and the world of ideas
still constitute for him an undifferentiated complex ;
causality and motivation are still thought of as one and
the same. Adults too, with the exception of meta-
physicians and naive realists, regard the connexion of
events and that of ideas as one, in the sense that logic
and reality constitute two series inextricably bound up
with each other. But the adult is sufficiently detached
from his ego and from his ideas to be an objective
observer, and sufficiently detached from external things
to be able to reason about assumptions or hypotheses
held us such. This brings about a twofold liberation
and a twofold adaptation of the mind. The child's
ideas, on the other hand, hinder his observations, and
his observations hinder his ideas, whence his equal and
correlative ignorance of both reality and logic.
10. QUESTIONS ABOUT HUMAN ACTIONS AND QUES-
TIONS ABOUT RULES. Like the corresponding 'whys/
questions about human actions relate sometimes to
psychological explanation properly so-called, some-
times to purely momentary actions. Here are some
examples.
" Who do you love best, me or mummy ? You You
shouldn't say that. Ifs a little bit naughty " " Does every
one love Jesus Yes Do you ? Yes If he was unkind,
A CHILD'S QUESTIONS 215
'would he punish us?" "Would you be a little bit funky
[climbing up a tree] ? " etc.
And the whole set of questions of the form : " What do
you think I am going to do, jump [or not jump] ? " " Then
do grown-up people make mistakes also ? / thought grown-
up people could think about these things ? " (already quoted
in connexion with a question about time), etc.
The only thing about these questions which is of
interest for our subject is the omniscience which the
child ascribes to adults. This circumstance is not with-
out influence in forming the child's anthropomorphic
view of nature. If the adult knows everything, foresees
everything, can answer everything if he only chooses
to, it must be because everything is harmoniously
ordained, and because everything can be justified. The
child's increasing scepticism towards adult thought is
of the greatest importance in this connexion, for this
it is which will give rise to the idea of the given as
such, of chance. Questions such as we have given
should therefore be very carefully studied if we wish
to be aware of the moment when scepticism is first
acquired by the child. At the period when the questions
were asked, faith in the adult was still considerable.
At the end of the year during which we studied Del
(7 ; 2) this faith had ceased to exist : " Then Daddy can't
know everything either, nor me neither ? "
As to questions about usages and rules, they are
continuous with those we have already discussed. It
is worth while, however, treating them separately ; in
the first place, because they correspond to a parallel
group of 'whys,' and secondly, because they in turn
merge into the questions of classification proper, which
constitute a very important group. In spite of this,
however, the necessity for making a separate class of
" questions about rules" may be questioned.
These questions therefore begin by being simply
psychological, but relating to social usages: "It is
always ladies who begin at games and parties, isn't it ? "
2ib STUDIES IN CHILD LOGIC
etc. But immediately there is a transition to school
rules: " Who said not [to spell 'quatre' with a k] Was
it the grown-ups in Paris ? "
Then comes a set of questions about rules as such :
"How do you spell [a name]?" "Should there be an
acute accent?" "Is that right?" etc. These questions
show no new features, and we have already pointed out
their affiliation with the corresponding * whys.'
ii. QUESTIONS OF CLASSIFICATION AND CALCULA-
TION. How does the transition occur from the interest
taken in rules, in "one does," or "one must" to the
logical interest, or search for a reason which will cause
a judgment to be adopted or rejected? We have seen
that Del's earliest "whys" were connected with defini-
tion, and therefore seemed to arise out of "whys about
rules " relating to language. The following questions
will confirm this affiliation.
There is a category half-way between questions about
rules and questions about classification, which is prob-
ably the root from which both have sprung. We
mean the collection of questions about names. Some-
times it is simply the search for the meaning of an
unknown word, sometimes it is an etymological analysis :
" What does year-end mean ? " " What does hob-nobbing
mean ? " " Who is Rodolphe ? " " What are rivers called
that run between mountains?" " Pric [He is reading],
yes does it prick?" " What does ' mar** [the first half of
the word ( Mardi '] mean ? "
There is a definite progression from these questions
about names to questions of classification. Questions of
classification are those which in face of a new object no
longer ask "what is it called?" but "what is it."
They also look for the definition of an already familiar
object.
"What is that, a pond?" "What is that, a
cockchafer?"
"What is a cuj"". . . a table-cloth?"". . . a
home" etc. (the name as such being already known).
A CHILD'S QUESTIONS 217
It is easy, therefore, to establish the genealogy of these
questions. As Sully and Compayre have said, children
believe that every object has received a primordial and
absolute name, which is somehow part of its being. 1
When very young children ask about an unknown
object " what is it?", it is the name of the object they
are enquiring for, and this name plays the part not
only of a symbol, but of a definition and even of an
explanation. Among the questions of rule and classi-
fication it is therefore the question of name which comes
first in point of time. But since this question of name is
both normative and classificatory, it is easy to under-
stand how it can give birth to questions so different
from one another as those of rules, of classification, and
finally of logical reason. One can, indeed, conceive of
all the successive stages, passing from nominal realism
to intellectual realism, and from intellectual realism to
logical justification.
Finally, we must add to classification questions of
evaulation (judgments of value) : " Is it pretty? " "Isn't
that fair? ""Isn't that right? " etc.
Questions of calculation, on the other hand, must be
put in a class apart. In Del's case their number is very
small owing to the age of the child and the individual
type to which he belongs: "My daddy told me that
1000 was 10 times 100? Yes And to make 10,000?
10 times 1000 And to make 100,000?"
III. CONCLUSIONS.
We must now bring forward some of the general results
obtained from Del's questions from the statistical point of
view, from the point of view of age, and from the point
of view of the psychology of child thought in general.
12. STATISTICAL RESULTS. For the purpose of
comparing Del's questions with each other from the
1 M. Rougier has suggested calling this phenomenon '* nominal realism"
in the theory of knowledge.
2i8 STUDIES IN CHILD LOGIC
point of view of constancy and age, we divided them
up into three lots of 250 successive questions, including,
of course, all the <whys. J These are questions 201 to
405 (September to November 3, 1921), 481 to 730 (March
3 to March 24, 1922), and 744 to 993 (June 3 to June 23,
1922). In this way we obtained the following table :
I. II. III.
I. Questions of causal explanation
j
16
8
Plants
6
3
3
Animals
14
5
17
Human body
8
3
3
Natural phenomena .
(40
(27)
Tso
Manufacture ....
14
5
17
Total .
"(55)
730
748)
II.
Questions of reality and history
Facts and events ....
24
19
5
Place
2
8
9
Time
I
10
7
Modality ......
3
i
I
Invented history ....
26
8
I
* Total .
(56)
746)
768)
Total of explanation and of reality .
(in)
(78)
(116)
III.
Questions on actions and intentions ,
(68)
(97)
(70
IV.
Questions on rules
Social rules
6
3
o
School rules
17
9
H
Total .
(23)
7l2)
(14)
Total of actions and rules .
(90
(109)
(85)
V.
Questions of classification
Name
18
3
19
Logical reason
2
i
3
Classification
25
52
23
Total .
"(45)
756)
745)
VI.
Questions of calculation ....
3
7
4
Total of calculation and classification
48
63
49
250 250 250
A CHILD'S QUESTIONS 219
Here are the results obtained with the same three
series from the verbal point of view :
I.
95
91
7
28
27
...
...
3
...
II.
122
53
13
18
21 5
4
7
3
6
III.
143
41
18
21
29
i
...
4
360 185 28 67 77
10
These tables extend over a period of ten months, each
series being separated from the succeeding one by an
interval of two months. They enable us to form a
certain number of conclusions.
To begin with, there is the relative constancy shown
by the three big groups first by the questions of
explanation and reality (in, 78, 116), then by the
questions relating to human actions and rules (91, 109,
85), and thirdly by the questions of classification and
calculation (48, 63, 49). This constancy is rather
interesting if it proves to be verifiable in subsequent
research. The development undergone by the questions
follows perhaps a law analogous to that which governs
the development of language, f It is a well-known fact
that while a subject's vocabulary will grow considerably
richer with increasing age, the proportions of its various
categories of words to each other, are always subject
to fairly rigid laws. I And the constancy of our big
groups of questions over the space of ten months remains
pretty much the same, in spite of very definite fluctua-
tions within each group.
The fluctuation which first attracts our attention is
the diminution of ( whys ' and the corresponding increase
of questions without any interrogative expression: 91,
53 and 41 ; and 95, 122 and 143. This must, of course,
be considered in relation to the diminution of questions
of causal explanation in comparison with the questions
of reality and history, which tend to increase in number.
Finally and this seems to contradict the last two facts
'whys,' although they are diminishing relatively to
220 STUDIES IN CHILD LOGIC
the number of questions (which does not mean that
they are diminishing absolutely), take on a more and
more causal character in the widest sense of the term.
To make quite sure of this, we cut out from questions 200
to 1125 three sections, each consisting of 60 successive
4 whys/ at a period when the questions were all being
taken down. Here is the result :
I II III
Whys of causal explanation . . 15 21 30
Whys of motivation ... 28 27 25
Whys of justification , . . 17 12 5
These figures may of course be the result of changes
in Del's particular interests, and they may be peculiar
to the year during which they were collected. We
therefore do not wish to establish general laws on such
a slender foundation. It may, nevertheless, be interest-
ing to see whether those three kinds of fluctuations are
independent of one another, or whether there is not a
certain solidarity between them. If the problem be
stated in this way, it will admit of a more generalized
treatment, even with the special data collected from Del.
Thus, on the one hand, the relative frequency of
' whys J diminishes; on the other hand, there is an
increase of questions of reality and history in comparison
to those of explanation ; finally, the sense of the ' whys '
becomes increasingly causal. These movements seem
to us to be closely connected with one another. It is
true that statistics can be made to prove anything, but
in this case statistical induction corresponds with the
results of qualitative analysis and clinical examination.
For one thing, if the frequency of ' whys * diminishes
in proportion to the bulk of the questions, this is
because between the years of 3 and 7 * why ' is really
a question which is used for every purpose, which
demands a reason for everything indiscriminately, even
when there is no reason present except through a con-
fusion of the psychological and the physical order of
A CHILD'S QUESTIONS 221
things. It is therefore quite natural that when these
two orders come to be differentiated, and when the idea
of chance or of ' the given ' first makes its appearance,
a large number of questions should break away from
the ' why ' form. They will then take on the form of
' how ' or of simple questions without any interrogative
words in them, and will concern themselves as much
with the consequences and inner mechanism of pheno-
mena as with their 'reason.' The decrease of 'whys'
would thus be an index of a weakening of precausality.
This weakening, it seems to us, can also be seen in the
increase of simple questions in so far as these show signs,
as compared to 'whys,' of a desire for supplementary
information.
Furthermore, the increase of 'whys' of causal ex-
planation in comparison with other ' whys ' is probably
due to the same reason. It is because precausality, or
rather the tendency to justify everything, is on the
decline that Del's curiosity is t less eager in seeking a
justification for rules in which no such motivation is
involved. It is because * whys J have become specialized
that the 'why' of explanation predominates. This,
incidentally, does not mean that the "whys of logical
justification " are condemned to grow less, for these only
make their appearance after the age of 7 or 8, and they
do so in connexion with any kind of demonstration.
In order to prove these assertions one would, of course,
have to separate the ' whys ' of precausality from those
of causality proper, and then take the percentages.
But since it would be impossible to carry this out with-
out making very arbitrary judgments, we must needs
content ourselves with suppositions. The diminution
of ' whys ' concerning the justification of rules, more-
over, is certainly an index in favour of the hypothesis
that Del is losing his desire to justify things at any
price, and that, consequently, precausality is giving way
to a wish for a more strictly causal explanation of
phenomena.
222 STUDIES IN CHILD LOGIC
From this point of view we are also enabled to under-
stand why questions of reality and history increase in
comparison to questions of explanation, always assum-
ing that this increase is not due to the arbitrary character
of the classification. If questions about facts and cir-
cumstances are multiplied, it is because the child gives
up the attempt to account for phenomena which are
simply given, and tries to gain a more detailed know-
ledge of the historical circumstances in which they
appeared, of their conditions, and of their consequences.
These results recall very clearly those obtained by
M. Groos in the fine work he has done on provoked
questions. By presenting children with any proposition
whatsoever, and then noting down the question which
it calls forth, M. Groos has shown that whatever the
age of the subject, from 12 to 17, questions of causality
taken in the widest sense of the term constitute a more
or less constant percentage (40%). But these causal
questions can be divided into regressive (cause), and
progressive (consequence). Now progressive questions
increase quite regularly with age. This result has been
roughly confirmed by the experiments carried out at
the Institut Rousseau on children under 12 (and above
g). 1 Therefore, the transformation of causal questions
into questions relating to consequences does not prove
that the general interest in causality has in any way
weakened ; it merely indicates that this interest is no
longer confined to the ' why ' pure and simple, but now
attaches itself to the details of the mechanism itself.
With regard to Del, our statistics enable us to con-
clude that he has gradually lost his interest in pre-
causality. The hypothesis can therefore be put forward
that the decline of precausality takes place between the
ages of 7 and 8. Now our earlier chapters have already
shown the importance of this age from the point of view
of the decline of ego-centrism, from the point of view
of the understanding between children, and above all
1 Intcrmgdiaire des Educatturs, 2nd year, 1913-14, p. 132 et stq.
A CHILD'S QUESTIONS 223
from the point of view of the mental habits involved in
genuine argument and collaboration in abstract thought.
This synchronizing is probably an index of important
correlations. Before attempting to establish these, let us
first see whether we cannot verify the statement that, as
Del approaches the age of 7 or 8, precausality tends to
give place to true causality.
, 13. THE DECLINE OF PRECAUSALITY. There is a
very simple method of measuring DePs evolution in
connexion with precausality ; that is, after a little time,
to ask him his own questions, or at least those whose
form clearly bring out their precausal character. We
chose 50 questions of causal explanation, etc., for this
purpose, and submitted them to Del himself when he
was 7 years and 8 months, telling him they were the
questions of a little boy of his age. Now the first point
to note is that Del had not the slightest idea that they
were his own questions. (It will be remembered that
he never noticed that all his questions were taken down).
Not only that, but he actually interspersed his answers
with remarks like these : " Ifs silly to ask that when ifs
so easy. Ifs silly. It doesn't go together. Ifs so [silly]
that I dorit understand a word of it" But this in itself
is not conclusive. A child's thought at 6 or 7 is still so
undirected, so unsystematized, in other words, it is still
so subconscious in the Freudian sense of the word, that
the fact of his forgetting questions asked by him a few
months ago and of his being incapable of answering
them does not prove much in the way of any change of
mentality. On the other hand, there is a very definite
hiatus between the type of answer and the actual form
of the question, a hiatus at times so strange that we felt
it incumbent upon us to make sure that the questions
which were asked really had a precausal significance.
We therefore tested them by submitting them to ten
children of 7 years old. Some of them answered us in
the definitely precausal manner which the Del of 6 to
7 expected when he asked the questions. Others
224 STUDIES IN CHILD LOGIC
answered us like the Del of 7 ; 2, thus showing that
they too had got beyond the precausal stage. 1
Here are some of Del's answers :
" Why is there a Little Sal eve and a Great Saleve?
Because there are two. There were two mountains stuck
close together, so people said that one was to be the great
Saleve, and the other the little Saleve"
" Why are they [negroes] made to be like that? It is
the sun, because in the negroes' country it is very hot, much
hotter than here."
"What makes it go [the marble]? Ifs because it is
going down-hill -Does it know you are down there ?
No, but it goes down to where you are"
"What makes the Rhone go so fast? Why! because
it goes down-hill a little"
"A little boy wrote his name on some wood. The
next day the name had gone. He asked : ' Why do
wood and iron rub out pencil marks ? ' Because you put
your hands on it and rub, and then it goes away. Is that
right? He made a mistake, because if you make pencil
marks on paper and take some wood and iron and rub, it
doesrft go away" Del has obviously not understood his
own question.
" The lake does not go as far as Bern, why? Because
Bern is far away and the lake is quite small. But the lake
of Geneva is big, but it doesrit go as far as Bern. If it was
the sea, it might, but it is not that country, that is not what
it is called"
" What makes the lake run? The Rhone"
" Was it [a tree] planted or did it grow by itself?
It was planted. A few flowers can grow by themselves"
"When I put red and orange together, it makes
brown, why? I don't know. My daddy couldn't know
everything, so neither can I" Here we see the signs of
scepticism in Del. He would not have answered in
this way at six-and-a-half.
1 It may be objected that Del has passed from the precausal stage to a
more advanced mental state because he has remembered the answers given
by the adult to the questions which he used to ask. This goes without
saying, but it is no explanation. The problem still remains to be solved,
why the child has accepted these answers and above all why he has assimi-
lated them without distorting them. It is of this capacity for assimilating
causal or natural explanations that we speak when we say that Del has entered
upon a new stage of development.
A CHILD'S QUESTIONS 225
"If they are not dangerous, why do they have those
things (poisonous fangs)? Because they live as we do.
We have nails and they aren't of any use."
Reading those answers, we begin to doubt whether
such a thing as precausality exists. It looks as though
Del had never had any but the most positive explana-
tions in mind, and as though any appearance to the
contrary were due simply to the clumsiness of his style.
But if this were the case he would never have asked
one of these questions. The explanation given by Del
of the fact that adders are not dangerous is significant
in this connexion. It amounts to saying that the
question should not be put, or is badly put. The same
applies to the answer about the size of the lake, and,
in a certain measure, to the answer about the two
Saleves, while the refusal to explain why red and
orange make brown is highly characteristic. In short,
in all these cases, if Del had expected the same answer
as he gives himself at 7 ; 2 he would not have asked the
question. With regard to the questions about the
negroes and the swiftness of the Rhone, it is obvious
that the very positive answers given by Del at 752
must not deceive us as to the anthropomorphic and
artificialistic character of the questions when he was 6,
otherwise the verbal form employed would be in-
comprehensible. We shall presently have occasion
to verify these statements when we come to examine
the way in which other children answered the same
questions.
There is, therefore, a complete discord between the
questions asked by Del and the manner in which he
answers them a few months later, and this seems to
indicate that the child has partly given up the use of
precausal explanations. The questions were originally
put as though a precausal explanation were possible, as
though everything in nature could be accounted for, as
though everything were animated by intentions, so that
the looked-for cause of phenomena could be identified
p
226 STUDIES IN CHILD LOGIC
with a psychological motive or a moral reason. The
answer given at 7;2 shows, on the contrary, that in his
mind the distinction is being formed between strictly
causal explanation, psychological motivation, and logical
justification. Not only does he give up the attempt to
account for everything, but the answers which he gives
are sometimes causal explanations properly so-called,
sometimes logical justifications. For instance he explains
the current of the Rhone by the slope, the colour of the
negroes by the sun (causal explanation). The answers
to the question about the two Saleves and to that about
the size of the lake began in both cases with a simple
logical justification (because there are two Saleves,
because Bern is far away from the lake). Precausality
would therefore seem to be on the wane in Del's mentality,
and the distinction to be growing between strictly causal
explanation and other types of relations.
Too much emphasis, however, must not be laid on the
contrast between Del's mentality at six-and-a-half and
at 7 ; 2. Even in the few answers which have been
reproduced here, neither causality nor logical justification
are present in an unadulterated form. The answer to
the question about the size of the lake is still very con-
fused in this respect. It amounts to ascribing to the
lake or withholding from it certain powers, just us
though it were not exactly a body endowed with
spontaneous activity (such as the wind and sun appear
to be to the child), but a big river, which goes where it
chooses. The lake could go to Bern if it were the sea,
but it is not called (i.e. it is not) the same as the sea,
therefore it cannot.
We decided to test these results by asking 10 children
of 7 to 8 those same questions of Del's. This procedure
showed in the first place that the questions were inter-
preted as requests for precausal explanation, at least
by most of the children, who can be regarded as only
slightly backward in comparison to Del ; and in the
second place, that between 7 and 8 many of the answers
A CHILD'S QUESTIONS 227
are already more or less causal or logical as the case
may be, just as those of Del at 7 ; 2.
Here are some examples of precausal answers :
"Why is there a great and a little Sal eve? For
children, and the great one for grown-ups (Au). Because
there is one for little children and the other for big ones (De).
Because of people who want to go into the little one or the
great one (Gia). The little one is to go on to> and so is the
great one (Ru)."
"My daddy told me that thunder made itself by
itself in the sky. Why? It was God who made it (Ri).
Because God made it (Au)."
"Why are they (negroes) made to be like that?
Because God punished them* They were naughty when they
were little (Au). Because they are dirty (Ga). God did it
(Go). Because they were born like that (Gia)," etc.
"What makes the Rhone go so fast? The water
(Ru, Ant, etc.), the boats (Ri)."
" What makes the lake run ? The machines [the locks]
(Ri), GW(Ru, Go, etc.), the rocks (Au)," etc.
"The lake does not go as far as Bern, why? // is
shut (Ru). Because it is stopped. There is a big wall
(Go). Because it isn't so big. There is another lake at
Bern. There's a lake in every country ( Au) . Because the
lake of Bern is another lake (De). Because they are not all
the same lakes (Ant). Because it is too far (Gia)."
These answers are probably all precausal. But they
also show the polymorphism of precausality, and we
cannot possibly at this point enter upon an analysis
of these childish explanations. That will be the work
of later enquiries. We must therefore limit ourselves
to the one conclusion which is of any value to the study
of logical reasoning in the child, viz., that precausality
points to a confusion between the psychical and the
physical order of things. The result is that in the child
logical justification will never appear in unadulterated
form, but will continually oscillate between justification
and psychological motivation.
14. CONCLUSION. CATEGORIES OF THOUGHT OR
LOGICAL FUNCTIONS IN THE CHILD OF 7. A question,
228 STUDIES IN CHILD LOGIC
says Claparede, " Is the conscious realization of a problem
or of the difficulty of solving it, z>., of the direction in
which to seek for its solution. To search effectually,
one must know what it is one is searching for, one must
have asked oneself a question. The nature of this
question will determine the whole orientation of subse-
quent research. Thus the function of the question is
quite clear : it is, an incitement to mental activity, in
a certain direction in view of readjustment. . . .
". . . Logicians have tried to catalogue . . . [different
kinds of questions], or rather the different kinds of
judgments which constitute their appropriate answers,
and they have given the name of category to the various
classes observed in this way. This enumeration of
different sorts of questions is of very little interest to
psychology. The number of questions that can be
asked is infinite, they are as many as the different ways
of being unadapted ; and the question of whether they
can be grouped under certain headings is only one of
secondary interest.
" It is more interesting to enquire into the biological
origin of these various types of questions. How did
the individual ever come to ask questions about cause,
aim, or place, etc. This problem of origins is the same
as that of knowing how the individual gradually came
to interest himself in the cause, the aim, and the place
of things, etc. And there is good reason for believing
that his interest was only directed to these ( categories '
when his action was unadapted to one of them. Need
creates consciousness, and the consciousness of cause (or
of aim, or of place, etc.), only arose in the mind when
the need was felt for adaptation in relation to the cause
(or the aim, etc.).
"When adaptation is purely instinctive, the mind is
not conscious of these categories, even though the
instinct in question acts as if it were ; the action here is
automatic, and its execution presents the mind with no
problems ; there is no failure to adapt, therefore there
A CHILD'S QUESTIONS 229
is no felt need, and consequently no consciousness of
any such need in the direction which would lead to its
satisfaction.
" We may note in passing how greatly our conception
of < categories ' differs from that of philosophers. Accord-
ing to the Faculty Psychology the categories are the
result of some sort of primitive mental intuition. But
observation shows that those categories only come into
being through some defect in adaptation. According
to Associationism, the categories are the result of
reiterated associations which have become inseparable.
But observation shows that precisely when association
reaches its highest degree of automatism (instinct,
habit), the individual is not conscious of the categories,
because, not having failed to adapt himself,-he has no
need to ask any questions." *
We have quoted this remarkable passage because,
having reached the end of this book, we can do no more
than express our complete agreement with it. In a
sense, we have gone further along the path of functional
psychology in asserting that the fact of becoming
conscious of a category will alter its actual nature. If,
therefore, we accept the formula: "The child is cause
long before having any idea of cause" it must be
remembered that we do so only for the sake of con-
venience. It is only as a concession to language (and
one which if we are not careful will involve us in a
thoroughly realistic theory of knowledge entirely out-
side the scope of psychology) that we can talk of
' causality ' as a relation entirely independent of the
consciousness which may be had of it. As a matter of
fact, there are as many types of causality as there are
types and degrees of becoming conscious of it. When
the child "is cause," or acts as though he knew one
thing was cause of another, this, even though he has
not consciously realized causality, is an early type of
1 Ed. Claparede, "La Psychologic de 1'intelligence," Scicntia 1917,
PP- 3*1-3-
230 STUDIES IN CHILD LOGIC
causality, and, if one wishes, the functional equivalent
of causality. Then, when the same child becomes
conscious of the relation in question, this realization,
just because it depends upon the needs and interests of
the moment, is capable of assuming a number of differ-
ent types animistic causality, artificialistic, finalistic,
mechanistic (by contact) , or dynamic (force) , etc. The list
of types can never be considered complete, and the
types of relation used nowadays by adults and scientists
are probably only as provisional as those which have
been used by the child and the savage.
The study of categories is, as M. Claparede rightly
maintains, a study of functional psychology, and vast
new horizons are opened to it by the law of conscious
realization. Here the psychologist meets on common
ground with the historian of science and the modern
logician. Traditional logic, whether we take the realism
of the Schools or Kant's apriorism, regarded the cate-
gories as fixed, and imposed on the mind and on things
once and for all, and in a definite form. This hypothesis
is psychologically false, and has been brilliantly attacked
by William James at a period when logicians themselves
had begun to abandon it. Renouvier and Cournot have
given to the theory of categories a turn which it is
no exaggeration to characterize as psychological, since
the task they have set themselves is to define the cate-
gories according to their genesis in the history of thought
and to their progressive use in the history of the sciences.
This is the point of view which Messrs Hoeffding, 1
Brunschvicg 2 and Lalande 3 have since very elaborately
developed. From this angle the problem of categories
must therefore be formulated in connexion with the
intellectual development of the child himself. The
genetician will therefore have to note the appearance
and use of these categories at every stage of intelligence
1 La pens & humaine.
2 Les ttapes de la philosophic mathtmatique.
3 Bulletins de la Soctite fran$ai$e de philosophic \ passim.
A CHILD'S QUESTIONS 231
traversed by the child, and to bring these facts under
the functional laws of thought.
It is in this spirit that we should like to build up out
of DePs questions a table which, though only approxim-
ately correct, might still serve to orientate subsequent
research. To do this we need only transcribe the
classifications we adopted, and consider them from the
genetic point of view.
In the first place, what relations can we allow to exist
between our classification of questions and our attempted
genealogy of whys ? Questions of the form * what is . , . ?'
and ( when?' are admittedly earlier than * whys 7 (Stern,
Mile. Descoeudres). But it can be definitely stated that
at the moment when < whys ' first make their appearance,
a reorganization of values takes place in the child's mind,
which enables us to see more clearly the relations uniting
the different categories of questions. We shall therefore
occupy ourselves only with the period extending from
the age of 3 to that of 7-8, i.e., with what Stern has called
the second questioning age.
In what circumstances do the first f whys' appear?
Approximately at the same age as the three follow-
ing fundamental phenomena : i The formation of two
distinct planes of reality. Up till the age of 3, the
real may be said to be simply what is desired. There
is, indeed, after i;g or 2 a yes and a no, a real and
an unreal, but without any further shade of difference.
At about 3, on the other hand, the imagined is some-
thing distinct from the real. According to Stern, this
is the age when we first meet with such words as
< perhaps,' l etc., which are precisely those which mark a
divergence between the imagined and the real. Again,
to quote Stern, there appear at the same date such verbs
as 'to think,' <to believe,' 2 etc. As we take it, the
1 In the lists given by Mile Descceudres (Le D&vdoppement de Tenfant
de deux ci sept am, Neuchatel, 1922) 'perhaps' occurs only in the language
of the child of 5, but we have ourselves noted it at 3.
2 According to the same authority "to think" is noted at 2 ; 9, "to
believe " at 5.
232 STUDIES IN CHILD LOGIC
advent of these words, whatever may be said to the
contrary, in no way indicates a distinction between
the psychical and the physical, or between thought
and thing, but a distinction between what is imagined
and what is perceived. 2 It is at about the same period
(2j9 and 3;io) that Scupin detected the earliest lies,
or, as P. Janet has so excellently described them,
" beliefs about the future" as opposed to beliefs about
the present. 3 Finally, it is also at about the age of
three that grammatical accident makes its first appear-
ance. Cases and tenses of a certain complexity, the
simpler forms of subordinate prepositions in a word,
the whole necessary apparatus for the beginnings of
formulated reasoning begins to be incorporated into
the language of the subject. Now the function of this
reasoning is to construct, over and above the immediate
world of sensation, a reality supposedly deeper than the
merely given world. And all these transformations
have this fundamental trait in common, that they indi-
cate an act of conscious realization. From now onwards
the child distinguishes between the real as it appears
immediately to his senses, and something which pre-
cedes events and underlies all phenomena. Let us
describe this something by the very comprehensive
term intention. The intentions of people and of things
sometimes conform to the wishes of the child, some-
times they do not ; hence the distinction between the
imagined or desired and the real. Hence, also, the
resistance put up by reality which necessitates lying.
Intentions can sometimes be detected at once, and fit
in spontaneously with the events ; at other times they
cannot, whence the necessity of reconstructing them,
of supposing their presence behind things, in a word,
of reasoning instead of simply looking on.
These changes, contemporaneous with the earliest
'whys,' are not altogether unrelated to this type of
question. Up till this age, reality coincided almost
entirely with desire, and existed on a single plane, so
A CHILD'S QUESTIONS 233
to speak, without the child having ever become clearly
conscious of intentions contrary to its own, or definitely
independent of them. The questions asked relate simply
to the names of objects and to the place which they
occupied after they have disappeared. Roughly speak-
ing, at about three years old the child takes cognizance
of the resistance set up by things and people ; there is
discord between desire and its realization. For a mentality
that has not yet learnt to distinguish between thought
and things, between animate and inanimate, between
ego and non-ego, this discord can only be conceived as
an intentional resistance on the part of people and
things. The real, henceforth, becomes crowded with
intentions ascribed first to other people, then to things,
whether these things are thought of as autonomous or
dependent upon persons. Thus the whole world be-
comes peopled in various degrees not, it is true, with
personified spirits, because at this age the child is still
unconscious of its own personal unity, and does not
think of ascribing intentions to definite 'IV but
of intentions that are impersonal, so to speak, or at
any rate improperly localized and multiform. Hence the
earliest 'whys,' 'why' being the specific question for
seeking the intention hidden behind an action or an
event.
The earliest * whys ' are generally asked in connexion
with human actions. The first ' why ' noted by Scupin
in the case of ' Bubi ' is of this order. The child's
mother was lying on the ground. The boy wants to
get her up: " JDu bis ya nicht tot warum stehste nicht
immersu auf?" The second one appears when the child
is forbidden to pull the petals off flowers. " Warum
denn?" But even where children begin with a "why
of explanation," it is difficult not to see in the expected
explanation not only a precausal explanation, but one
in which precausality is almost entirely confused with
psychological or intentional causality. " Why do trees
have leaves? "
234 STUDIES IN CHILD LOGIC
It is these intentions ascribed to people and to things
which will give rise to the types of question correspond-
ing to the principal categories of child thought. These
categories will therefore have an intentionalistic origin,
z>., they will arise from the conscious realization of
psychological operations relative to intentions, and not
from a mere observation of the world given in per-
ception. Moreover, the earlier categories of name and
place, etc., will join themselves to these categories of
intention, and together with them will form a single
whole.
This intentionalism gives rise to two fundamental
categories or primitive functions of thought : the ex-
plicatory function and the implicatory function. These
do not represent two separate departments of the mind,
but describe two moments which are present in all
mental activity. The explicatory function is the centri-
fugal moment, in which the mind turns to the external
world ; the implicatory function is the centripetal, in
which the mind turns inwards to the analysis of inten-
tions and of their relations.
The explicatory functions arise out of the need felt
by the child, as soon as he becomes conscious of inten-
tions, to project these into the world around him. On
the one hand, he finds himself surrounded by people
whose actions can be foreseen and whose motives can
be detected ; on the other hand, he is faced by a world
of phenomena and events which up till now have never
resisted his thought and therefore required no explana-
tion, but which have now become as great obstacles to
his fantasy as are people themselves. This duality has
to be abolished ; since there is a < why ' to human
actions, the same treatment must be applied to every-
thing which presents itself. Hence this universal desire
for precausal explanation which comes from confusing
psychological intentionalism with physical causality.
Thus the explicatory function has two poles psycho-
logical explanation and material explanation. These
A CHILD'S QUESTIONS 235
two poles are close together at first and not easily
distinguishable, but as time goes on they grow more
and more distinct, though always held together by
the fact that both are rooted in one and the same desire
for explanation.
Owing to the fact, moreover, that the idea of inten-
tion first appears through the resistance of reality, and
in particular through the resistance of persons, every-
thing seems to the child to obey some sort of necessity
which is both moral and physical. Everything seems
to him to be as it should be. So that the child's
tendency will be, not only to project intentions into
every object so as to explain events, but also to seek to
account for everything, to justify every event, and to look
for the connexions existing between intentions. Hence
the implicatory function. The explicatory function
was centrifugal in this sense, that from the intention
it sought to draw out the material consequence, the
resultant act or event. The direction of the Implicatory
function is, on the contrary, centripetal, in the sense
that from the intention it seeks to trace its way back to
the directing motive or idea. The explicatory function
tends towards things, the implicatory function tends
towards ideas or judgments. And child thought, being
at its origin equally removed from things and from
thought, occupies an intermediate position between
the two.
Thus the implicatory function also has two poles.
First a psychological pole in common with that of the
explicatory function and which causes the child to ask :
" Why do people do so ? etc." The "whys of justifica-
tion " which we collected from Del are naturally of a
much later date than these primitive questions, although
they constitute a special case of the * whys ' concerning
what ought to be. The other pole is made up of questions
about names, definitions, the reason for judgments, in a
word, about everything concerning logical justification.
Just as between psychological and physical explanation
236 STUDIES IN CHILD LOGIC
there are innumerable transitional cases, so also between
the implication of psychological actions (justification)
and the implication of names, classes and later on of
numbers, there is every type of intermediate example.
Thus the pole which is common to both functions, i.e.,
the psychological pole (psychological justification and
explanation) serves both as a starting point and as a
point of divergence for the two functions, explicatory and
implicatory, which are at first confused and then grow
more and more distinct. We shall call mixed, that
function of psychological justification and explanation
which partakes of the nature of explication and im-
plication.
This schema may be thought to apply only to 'whys,'
but it is obvious that other types of question, even of
earlier date, such as those of place ("where is . . . ?" etc.)
and of name ("who is ...?") are more or less in
corporated in it. As the explicatory function develops,
questions of place come more and more to resemble
the great group of questions of reality and history, to
which the desire for explanation gives its chief impetus.
Questions about names are originally independent, and
belong as such neither to the desire for explanation nor
to that for justification or implication ; but their function
is modified concomitantly with the development of the
implicatory function. The child finds that names which
originally were bound up in his mind with the object
can be subjected to an increasingly logical justification
(childish etymologies). This in itself tightens the bond
between questions of names and the implicatory function.
The same thing happens to questions of classification
and definition, definitions being at first, as is well
known, purely utilitarian, and then becoming increas-
ingly logical.
The main categories of child thought between the
years of 3 and 7-8 are therefore represented by the
following table :
A CHILD'S QUESTIONS 237
^ r . f Causality.
Explicatory function lor**- j 1
r y I Reality, time and place.
,,.,,, . f Motivation of actions.
Mixed function . . i T ^-^ * r 1
I Justification of rules.
T ,. r f Classification. Names.
Imphcatory function i XT , T . t , .
* J I Number. Logical relations.
To bring this chapter to an end, we must now try
very briefly to connect the results we have obtained
with the factors established in the earlier chapters, and
particularly with the ego-centrism of child thought.
In this chapter special stress has been laid on the
importance of precausality and consequently of intel-
lectual realism ; in other words, we have emphasized
the paradoxical fact that child thought is equally removed
from dealing with strictly causal explanation as it is
from dealing with logical justification properly so-called.
The whole mechanism of children's questions, as we
have studied it, can be accounted for by this funda-
mental fact.
What relation could there be between this fact and
the ego-centrism of child thought? A fairly close one
of mutual dependence, since (see 12) precausality tends
to disappear at the same age as ego-centrism, viz.,
between 7 and 8. In every strictly causal explanation
there is, after all, an effort to adapt oneself to the
external world, an effort to objectify, and, one might
almost say, to depersonalize one's thought. Without
this effort, the mind tends to project intentions into
everything, or connect everything together by means
of relations not based on observation, as is apparent
from the childish habit of justifying everything and of
conceiving nothing as fortuitous. Now ego-centrism
certainly hinders this effort towards the adaptation and
depersonalization of thought. It interferes with it
directly, in the first place, because the more the ego
is made the centre of interests, the less will the mind
be able to depersonalize its thought, and to get rid of
the idea that in all things are intentions either favour-
238 STUDIES IN CHILD LOGIC
able or hostile (animism, artificialism, etc.). But ego-
centrism is also an indirect hindrance, for in so far
as he is ego-centric, the child will not trouble to pit
his own ideas against those of others, and thus prove
what he has come to believe. He will therefore give
way to the primitive impulse of all thought, z\e. y he
will substitute for things as they are, a fragmentary
world of his own making in which everything has an
aim, and in which everything can be justified. But
there is also in the logical habit an effort towards internal
coherence and direction of thought, which is not spon-
taneously given to the primitive mind, but is a gradual
conquest of reason. Here again, ego-centrism is a real
obstacle to the acquisition of this desire for implication
or logical systematization. It is a direct obstacle,
because all ego-centrism is designed by its structure
to stand half-way between autistic thought which is
undirected,' z.., which as in day-dreaming hovers
about at the^ mercy of every whim, and i directed '
intelligence. ( Ego-centrism is therefore obedient to the
self s good pleasure and not to the dictates of impersonal
logic. It is also an indirect obstacle, because only the
habits of discussion and social life will lead to the
logical point of view, and ego-centrism is precisely
what renders these habits impossible.
We can now see that ego-centrism, while it does not
exactly explain the child's incapacity for true causal ex-
planation and logical justification, is nevertheless closely
connected with it. And we can understand how, as a
result of this, the child's mind is always hovering
between these two convergent paths, and is also equally
removed from both. This it is that gives rise to the
phenomena of precausality and intellectual realism,
both of which bear witness to this intermediate position.
And this it is also that gives rise to that tendency
in children to justify things at any price, or to connect
everything with everything else, which we have dealt
with at length in the course of this last chapter.
APPENDIX
FIFTY CONSECUTIVE REMARKS MADE BY
LEV. (See Chapter I)
I. Repetition.
II. Monologue.
III. Collective monologue,
IV. Adapted information,
V. Criticism and derision.
VI. Orders and threats.
VII. Question.
VIII. Answer.
NO.
TEXT
1 Lev puts his pencil down Geo's
neck. Geo cries out, " Lev ! "
It doesrit matter.
Lev begins to draw his hat again.
He shows his work :
2 Hook at things properly.
What are you looking at ?
3 The hat.
Lev repeats some words which one
of his companions is learning :
4 Luloidl celuloid!
Turning his drawing upside down
and addressing himself to no
one:
5 / want to see how it looks.
Ro brings some paper cigarettes.
He distributes them. Lev asks
for some :
6 How about me?
Lev goes back to his work. He
points to the little ribs of straw
on his hat and compares them
with the drawing :
CAT.
IV
IV
VIII
II
VII
REMARKS
This remark belongs
to IV because it is
part of a dialogue.
It calls forth an
answer and then re-
mark 3.
This is a request, but
expressed in inter-
rogative form ; it
therefore belongs to
category VI L
240
STUDIES IN CHILD LOGIC
No.
TEXT
7 There are just those things like
there ought to be, you see. (No
answer.)
To the other children :
8 You are moving the table.
He helps Go to play lotto. He
takes a card which Go does
not know where to put and
places it :
9 I think it goes here.
To Mile L. to whom Go has shown
a duplicate card :
10 Then if you lose one, there is one
left
Addressing himself to no one :
1 1 That is to make it hold together.
Pointing to one of the paper cigar-
ettes (without speaking to any-
one in particular) :
12 Look 'what Roger gave me. Roger
gave me that.
To Go, speaking of the lotto cards :
13 Yotfve got three of the same.
To someone else, showing his
cigarette :
14 It's like a cigarette, isn't it?
Geo is looking for the duplicate
card ; " Please, teacher, I can't
find the one that there's two
of."
15 No, I can see. -
Talking to himself as he draws :
1 6 First I shall do the fingers, so as
to make another round.
Has a hat got fingers ?
17 They are the things that go
below.
He shows his drawing :
1 8 There is my hat look at my hat,
There now, isrit that like I
Not very.
19 Still, it is just a little bit like.
CAT
IV
IV
IV
II
III
V
VII
IV
II
VIII
IV
IV
REMARKS
7 Addresses himself
to the community
in general without
expecting any re-
sponse.
8 Here Lev wants
to make himself
heard.
9 Clear case.
12 A doubtful case, but
as he has spoken
to the community
in general, no one
responds.
16 Clear case.
18 Might be III, but
has called forth a
response on the
part of a hearer.
APPENDIX
No.
TEXT
20 Look at my hat, it twists round
and round like a snail. The same
here too.
Geo asks the time. Lev, without
moving" from his work, mechanic-
ally repeats the question :
2 r Whafs the time ?
Talking of his work :
22 / kave done one half properly ,
teacher.
Mile L, does not hear. Another
child asks : " How is it a half? "
Lev does not listen but goes on
repeating :
23 One half properly.
A silence.
" Now you must write."
24 Write what?
" What there is on the hat" A
moment later Lev takes his hat
and looks,
25 IPs something- in English. I cart t.
Its " Mg
Drawing attention to what he is
doing :
26 Please teacher, one half is right,
yes, look.
Addresses himself to no one and
returns to his former topic :
27 1 ' dortt like that name in English,
Other children are building near
the table wheise Lev is drawing.
They are making a hut out of
leaves. Lev looks on :
28 I can make houses of leaves.
Ro is working with figures. Lev
gets up and announces :
29 I want fa see.
He approves of what Ro is doing.
Ro asks him nothing :
30 Yes, it is 5.
Ro makes a mistake. Lev :
31 You ought to count, down there.
Lev goes back to his place. He
talks without being listened to
by anyone :
CAT.
Ill
III
VII
III
IV
II
III
III
IV
IV
241
REMARKS
20 Is addressed to no
one in spite of the
form c Look. 3 A
clear case.
22 Might be III.
26 Might be 1 1 1.
30 Occupies himself
here with the work
of another.
STUDIES IN CHILD LOGIC
242
No. TEXT
32 I can make houses as big as trees,
can't /, teacher.
33 Mummy didn't give me a box, be-
cause she hadn't got a macaroni
box.
Lev looks on at Ro's work. Ro com-
plains that he has lost a ticket.
34 And the 6 isn't there. In another
box there is 6, 7, 8, 9.
Ro having followed this advice, and
looked for the 6 in the other box,
Lev returns to his drawing. He
makes no attempt to explain what
he is talking about. He is think-
ing aloud :
35 If I havertt took the things for
making the roof. I shall have
took them to-morrow.
Ro, whose eyes are blind-folded,
has to recognize by touch some
figures cut out in wood. He
makes a mistake. Lev who is
at the same table cries out :
36 Wrong!
Lev picks up a pencil which has
been dropped and hands it to
Bur;
37 Burny, take the pencil.
He takes some figures out of the box,
puts them before Bur, and asks :
38 What are those ?
Bur takes no notice, and does not
answer. Lev announces :
39 / can do much more than Bur.
He picks up a piece of wood, and
makes it roll on the table :
40 Now then, the van is going to start
the van has arrived the van
is going to start.
Another child looks at his drawing :
"What's that?"
41 That's the ribbon thafs come un-
done.
Lev is asked whether he won't put
the date on his drawing.
42 I'd rather put the name.
CAT.
Ill
III
IV
II
V
VI
VII
V
II
VIII
VIII
REMARK
32 This is not a ques-
tion.
APPENDIX
243
NO.
TEXT
43 He rises and leaves his place
shouting :
/ want to tell Mile L. something.
Someone is going- to write his
name on his drawing. He looks
on :
44 You mustn't write Leuane with
an e. With an a it is Leuana
in Georgian, without #, e ifs in
French.
Somebody tells Lev that he has
seen him disobeying his father.
45 Qh, no.
By way of changing the subject :
46 / shall write something else in
Georgian for you.
He looks to see what Bur is doing
and asks :
47 What is he doing?
48 To Bur :
Give meyeur fienciL
No.
Yes.
No.
49 c Yes* [in English]. Tfafs all I
know.
He leaves Bur and amuses himself
by folding bits of paper.
50 Look this is the way to do it.
(No answer.)
CAT,
III
IV
IV
IV
VII
VI
III
III
REMARKS
43 Addresses himself
to no one.
INDEX
Adolescent thought, 46
Alliteration, 121
Analogy, 47, 120, 121, 130, 135 ff.,
140 fL, 155, 157
Animism, 98, 172, 205
Anthropomorphic thought, 98, 165,
173 fL, 181, 207, 225
Art, 158 (Footnote)
Artificialism, 172, 173, 204, 225
Associationism, 132, 229
Assumptions, 210-214
Audemars and Lafendel, Miles, 5,
4*i 72
Autistic thought, 38, 43, 45, 114,
158 ff.
Baldwin, 2, n, 20, 41
Bally, 133
Belot, 98
Bergson, 132
Berguer, 42
Binet, 48, 109
Bleuler, 43
Borst, 86
Brunschvicg, 205, 230
Biihler, 131, 187, 203
Claparede, n, 86, 131, 132, 136, 155,
186, 228 fL, 230
Compayre", 217
Condensation, 158-9
Cook, O. F., 133
Cournot, 230
Death, idea of, 175-8, 206
Decroly, 132
Dementia prajcox, 101
Descoeudres, 25, 98, 112, 231
Deslex, 127
Drawing, children's, 116, 156, 182
Dreams, 43, 158
Drornard, 148 fE.
Etymology, children's, 149, 170^
216, 236
Faculty Psychology, 229
Fairies, belief in, 65
Ferenczi, 2
Ferriere, Ad., 120
Flournoy, H., 44
Freud, 2, 23, 158, 223
GestaltquaUttit, 131
Gesture, 42, 66, 69, 77, 118
God, 165, 168, 178, 181, 183
Groos, 176, 222
Guex, Mile Germaine, I, 50, 180
Hall, Stanley, 172
Hoffding, 230
Hysteria, 18
Imitation, n, 12, 41
Instinct of pugnacity, 26, 27
Interpretational maniacs, 148
Invention, 14, 125, 126, 130, 159,
1 60, 190, 204, 210
James, William, 230
Janet, 2, 3, n> 13, 16, 20, 41, 74, 75,
232
Jones, E., 2, 3
246 INDEX
Kant, 230
Kindergarten, 105
Lalande, 133, 230
Larrson, H., 159
Lies, 232
Luquet, 116, 182
Mach, 131, 211
Maison des Petits de I'Institut
Rousseau, 5, 14, 30, 38, 40,
48, 64, 112, 128, 176, 222
Meinong, 210
Meumann, 4
Meyenburg, Mile Hilda de, i
Mystics, 148
Mythological thought, 13, 43, 44
Parents, 48, 65, 98, 178
Piaget, J., 44, 134, 158, 161, 187,
196
Piaget, Mme V., 50, 76
Primitive races, 2, 3, 183, 212
Psycho-analysis, 3, 13, 43, 158, 177
Punning, 158, 159
Rassmussen, 164
Renan, 132
Renouvier, 230
Rignano, 75, 211
Roubakine, 120
Rougier, 217
Scepticism, 215, 224
Schuchardt, H., 133
Scupin, 164, 232
Simon, 48, 109
Spielrein, Mme, 2, 3
Stem, 2, 86, 98, 164, 231
Sully, 187, 217
Symbolism (of dreams), 158
Tachistocopic reading, 131
Terman, no
Theological ideas, 178
Transference, 158, 159
Veihl, Mile Liliane, 162, 163
Verbalism, 2, 78, 104, 105, 149
Word magic, 4, 14, 16
