Jean Piaget

If we examine the intellectual development of the individual or of the whole of humanity, we shall find that the human spirit goes through a certain number of stages, each different from the other, but such that during each, the mind believes itself to be apprehending an external reality that is independent of the thinking subject. The content of this reality varies according to the stages : for the young child it is alive and permeated with finality, intentions, etc., whereas for the scientist, reality is characterised by its physical determinism. But the ontological function, so to speak, remains identical : each in his own way thinks that he has found the outer world in himself.

This being so, two points of view are possible in the study of intellectual evolution.

The first of these is to choose a system of reference and agree to call " external reality ", reality such as it is conceived to be during one of the stages of mental evolution. Thus it would be agreed upon to regard as the external world reality as it is postulated by contemporary science, or contemporary common-sense. From this point of view, the relations of child thought to the external world would, in fact, be its relations to the universe of our existing scientific thought taken as the norm. In each explanation given by a child it would be possible to determine the part played by the activity of the subject and the part played by the pressure of objects, the latter being, by definition, objects as we now conceive them to be. And this would be Psychology, for the statements which this method led to would not claim to have any decisive bearing upon the Critical Problem in general.

Or else, the attempt to regard any system of reference as absolute can be abandoned. Contemporary common- sense or even contemporary science may be regarded as stages among other stages, and the question as to the true nature of external reality left open. And this would be Theory of Knowledge : this would be to place oneself above all the types of mentality that characterise the various stages of human development, and to seek to define the relations of the mind to reality without any preconceived notions as to what is mind and what is reality.

For our part, we shall confine ourselves to psychology, to the search, that is, for the relations between child thought and reality as the scientific thought of our time conceives it. And this point of view, narrow and question- begging though it appear, will enable us to formulate very clearly several outstanding problems. Does the external world (and by this we shall in future mean the world as it is viewed by science) impress itself directly on the child's mind, or are childish ideas the product of the subject's own mentality ? If the child's mind is active in the process of knowing, how is the collaboration effected between his thought and the data of the external world ? What are the laws which this collaboration will obey ? All these are the traditional problems of the Theory of Knowledge, which we shall be able to transpose into the particular sphere which we have just defined.

More exactly, the problems we are about to study are biological problems. Reality, such as our science imagines and postulates, is what the biologists call Environment. The child's intelligence and activity, on the other hand, are the fruit of organic life (interest, movement, imitation, assimilation) . The problem of the relation between thought and things, once it has been narrowed down in this way, becomes the problem of the relation of an organism to its environment. Is the organism entirely moulded by its environment in so far as intelligence is concerned ? If so, then we have, in terms of cognition, what may be called the empirical solution of the problem. Or does the organism assimilate the actions to its environment in accordance with a structure that is independent of these actions and that resists the pressure of all modifications coming from outside ? If so, then we have in terms of cognition what may be called the a priori solution. Or is it not rather the case that there is interaction between the two organism assimilating the environment to itself, but the environment reacting upon the structure of the organism ? Such is the solution which, in the domain of cognition, would imply a capacity for transformation in the categories of thought and an increasingly delicate adaptation of thought to things or of things to thought.

These, then, are the terms in which we set the problem. And if, in describing the results we have obtained in child psychology, we occasionally use words like empiricism, apriorism, etc., it must be remembered that we are not giving to these terms their strictly epistemological meaning, but are using them in a restricted and, as it were, in a purely psychological sense.

But, be it said in passing, it might perhaps be possible to make use in the Theory of Knowledge of the results acquired by our restricted method. Let us suppose, for the sake of brevity, that intellectual growth takes place along a straight line, in a linear series such that the stages A, B, C, . . . N follow one another without either interferences or changes from one level to another. We shall take the external world corresponding to stage G as absolute, and compare to it the external world corresponding to stages C, D, E, . . . etc. Such a comparison is without any epistemological bearing, since there is nothing to prove that G is decisive. But if, now, we take into account this very possibility of variation and regard the series C, D, E . . . G as capable of being extended, on the one hand, backwards, by the supposition of stages A and B, and, on the other hand, forwards, thanks to the future stages H, I, K . . . N, we shall discover the following : there will obviously exist a relation between the comparison of C, D, E to G and the comparison of G to H, I, etc. ; and the fragmentary conclusions obtained by the comparison of C, D, E to G will become a particular case of the general conclusions obtained by comparison of all possible stages.

To put things more concretely, it may very well be that the psychological laws arrived at by means of our restricted method can be extended into epistemological laws arrived at by the analysis of the history of the sciences : the elimination of realism, of substantialism, of dynamism, the growth of relativism, etc., all these are evolutionary laws which appear to be common both to the development of the child and to that of scientific thought.

We are in no way suggesting, it need hardly be said, that our psychological results will admit straight away of being generalised into epistemological laws. All we expect is that with the co-operation of methods more powerful than our own (historical, sociological methods, etc.), it will be possible to establish between our conclusions and those of epistemological analysis a relation of particular case to general law, or rather of infinitesimal variation to the whole of a curve.

The Child's Reality

How does the idea of reality constitute itself in the child's mind ? Any direct analysis of its origin is beyond our power ; the earliest stages precede language or are contemporaneous with the first spoken words, and any effort to reach the child's consciousness during these stages is fruitless, if one claims to go beyond mere hypothesis. But if we can content ourselves with conjecture, then it is best to try and extricate the laws according to which the idea of reality develops between the ages of 3 and n, and to extrapolate the guiding lines thus obtained so as to reconstruct the earliest stages. Moreover, as soon as we put this method into practice, we find that we can learn enough from the laws of evolution between 3 and n years, and that there is no need to attach any special importance to the original stage.

Three complementary processes seem to be at work in directing the evolution of reality as it is conceived by the child between the ages of 3 and n. Child thought moves simultaneously : i from realism to objectivity, 2 from realism to reciprocity, and 3 from realism to relativity. By objectivity we mean the mental attitude of persons who are able to distinguish what comes from themselves and what forms part of external reality as it can be observed by everybody. We say that there is reciprocity when the same value is attributed to the point of view of other people as to one's own, and when the correspondence can be found between these two points of view. We say that there is relativity when no object and no quality or character is posited in the subject's mind with the claim to being an independent substance or attribute.

Let us examine these processes more closely. In order to be objective, one must have become conscious of one's " I ". Objective knowledge can only be conceived in relation to subjective, and a mind that was ignorant of itself would inevitably tend to put into things its own pre-notions and prejudices, whether in the domain of reasoning, of immediate judgment, or even of perception. An objective intelligence in no way escapes from this law, but, being conscious of its own " I ", it will be on its guard, it will be able to hold back and criticise, in short it will be able to say what, roughly, is fact and what is interpretation.

So that in stating that the child proceeds from realism to objectivity, all we are saying is that originally the child puts the whole content of consciousness on the same plane and draws no distinction between the " I " and the external world. Above all we mean that the constitution of the idea of reality presupposes a progressive splitting-up of this protoplasmic consciousness into two complementary universes the objective universe and the subjective.

We have met with many examples of this realism of the first kind and of its progressive reduction. Children's ideas about thought may be taken as a first illustration of the phenomenon in question. The feeling of subjectivity and inwardness felt by the adult is, to a great extent, connected with the conviction of being the owner of a thought that is distinct from the things thought about, distinct from the physical world in general, and more internal and intimate than the body itself. This conviction only comes late in the child's development. During the earliest stages, the child believes that he thinks with his mouth, that thought consists in articulating words, and that these words themselves form part of the external things. The voice, being thus identified with thought itself, is regarded as a breath which participates with the surrounding air, and some children go so far as to say that it is identical with the wind in the trees, and that dreams are made of " wind ". They are quite incapable of distinguishing between thought and the things thought about. To use the expression chosen by M. EC- Delacroix, the sign " adheres " to the thing signified.

Later on, the child gives up this realism and localises thought inside his mouth, then in a little voice placed in the head ; he then gives up materialising thought and makes of it something sui generis which characterises the self as spirit (C.W., Chap. I).

The evolution of ideas about names is particularly suggestive from this same point of view. Word and name are about all that the child knows of thought, since he identifies thought with the voice. Now, names are, to begin with, situated in objects. They form part of things in the same way as do colour or form. Things have always had their names* It has always been sufficient to look at things in order to know their names. In some cases, this realism actually turns to magic: to deform the name is to deform the thing. Later on, names are situated in the adjoining air where the voice has uttered them, then in the voice, and finally in thought itself.

Dreams give rise to an equally definite realism. At first, they are thought to be pictures of air or light which come before our eyes from outside. At the earliest stage, the child thinks, naturally enough, that anyone could see the dream come into the room and go out again. Later on, the dream is believed to have an internal origin, but is conceived as coming out of the head or the stomach before appearing before the child. Finally, the child learns to distinguish between " being " and " seeming ", and localises the dream, first in the eyes, then in the head.

All these facts show that the localisation of the objects of thought is not inborn. It is through a progressive differentiation that the internal world comes into being and is contrasted with the external. Neither of these two terms is given at the start. The initial realism is not due simply to ignorance of the internal world, it is due to confusion and absence of objectivity.

Consequently, during the gradual and slow differentiation of the initial protoplasmic reality into objective and subjective reality, it is clear that each of the two terms in process of differentiation will evolve in accordance with its own structure. In the case of every object there will be a displacement of values which will modify the character of the object. Take, for example, the notion of "air", or of "wind". During the earliest stages, air is conceived as participating with thought : the voice is air, and, in return, the wind takes notice of us, obeys us, is " good at making us grow ", comes when we move our hands, and so on. When thought proper is localised in the self, and the participations between air and thought are broken, the nature of air changes by virtue of this fact alone. Air becomes independent of men, sufficient to itself, and living its own life. But owing to the fact that it is held to participate with the self, it retains at the very moment when it is severing these bonds, a certain number of purely human aspects : it still has consciousness, of a different kind perhaps than formerly, but its own nevertheless. Only very gradually will it be reduced to a mere thing.

This phenomenon is very general. During the early stages the world and the self are one; neither term is distinguished from the other. But when they become distinct, these two terms begin by remaining very close to each other : the world is still conscious and full of intentions, the self is still material, so to speak, and only slightly interiorised. At each step in the process of dissociation these two terms evolve in the sense of the greatest divergence, but they are never in the child (nor in the adult for that matter) entirely separate. From our present point of view, therefore, there is never complete objectivity : at every stage there remain in the conception of nature what we might call " adherences ", fragments of internal experience which still cling to the external world.

We have distinguished at least five varieties of adherences defined in this way. There are, to begin with, during a very early stage, feelings of participation accompanied sometimes by magical beliefs ; the sun and moon follow us, and if we walk, it is enough to make them move along ; things around us notice us and obey us, like the wind, the clouds, the night, etc. ; the moon, the street lamps, etc., send us dreams " to annoy us ", etc., etc. In short, the world is filled with tendencies and intentions which are in participation with our own. This is what we have called dynamic participation, in contrast to substantial participation, to which, however, it may lead.

A second form of adherence, closely allied to the preceding, is that constituted by animism, which makes the child endow things with consciousness and life.

A third form is artificialism (see C.W., Sect. III). The reader should be reminded at this point that artificialism in the child is not a theory which after reflection systematically takes man as the point of departure for everything. The terms must be reversed, and that is why artificialism has the same right to be classed among the adherences as animism. The child begins by thinking of things in terms of his own " I " : the things around him take notice of man and are made for man ; everything about them is willed and intentional, everything is organised for the good of men. If we ask the child, or if the child asks himself how things began, he has recourse to man to explain them. Thus artificialism is based on feelings of participation which constitute a very special and very important class of adherences in the sense that we have defined.

A fourth form is finalism : the starting-point and then the residuum both of animism and of artificialism, the deep and stubborn finalism of the child shows with what difficulty external reality frees itself from schemas due to internal and psychical experience.

A fifth form of adherence is constituted by the notion of force : things make efforts, and their powers imply an internal and substantial energy analogous to our own muscular force.

It is a striking fact that both the area of application and the strength of resistance of these adherences decrease progressively throughout the mental development of the child. And not only do these adherences lose ground little by little in correlation with each other, but their progressive disappearance seems to be proportional to the increasing clarity with which the child becomes conscious of his subjectivity. In other words, the better the child succeeds in dividing off the internal world from the external, the less stubborn are the adherences.

Three groups of facts may be mentioned in this connection. In the first place, as the child comes to notice the existence and the mechanism of his own thought, he separates signs from the things signified : thus, names cease to belong to the things named, thought is interiorised and ceases to participate with wind, dreams are no longer regarded as emanations of objects, and so on. Thus participations are loosened little by little, and even eliminated.

In the second place, in so far as the child discovers the existence and inwardness of his thought, animism, far from being strengthened is, through this alone, compromised and even completely destroyed. The decline of animism brings with it a progressive reduction of child dynamism. For so long as things seem to be alive and consequently active, the forces of nature are multiplied by the child ; and the elimination of life leads to a mechanisation of force which means ultimately an impoverishment of the actual notion of force. This very general process of evolution which leads the child from a dynamic to a mechanical view has been dealt with at sufficient length in connection with the details of children's explanations to render any further comment necessary.

Finally, as the child becomes conscious of his subjectivity, he rids himself of his egocentricity. For, after all, it is in so far as we fail to realise the personal nature of our own point of view that we regard this point of view as absolute and shared by all. Whereas, in so far as we discover this purely individual character, we learn to distinguish our own from the objective point of view.

Egocentricity, in a word, diminishes as we become conscious of our subjectivity. Now the decrease of egocentricity means the decrease of anthropomorphic finalism, and consequently the decrease of all the feelings of participation that are at the bottom of artificialism.

Progressive separation of the outer from the inner world, and progressive reduction of the adherences, such, in brief, are the two fundamental aspects of the first process which we defined as a passage from realism to objectivity. What we have just said about the relations between egocentricity and artificialism takes us on to the analysis of the second process, for it goes without saying that all these processes are closely related to each other, so much so, indeed, that they may be said to be completely indissociable.

The second characteristic process in the evolution of the idea of reality is the passage from realism to reciprocity. This formula means that the child, after having regarded his own point of view as absolute, comes to discover the possibility of other points of view and to conceive of reality as constituted, no longer by what is immediately given, but by what is common to all points of view taken together.

One of the first aspects of this process is the passage from realism of perception to interpretation properly so called. All the younger children take their immediate perceptions as true, and then proceed to interpret them according to their egocentric pre-relations, instead of making allowance for their own perspective. The most striking example we have found is that of the clouds and the heavenly bodies, of which children believe that they follow us. The sun and moon are small globes travelling a little way above the level of the roofs of houses and following us about on our walks. Even the child of 6-8 years does not hesitate to take this perception as the expression of truth, and, curiously enough, he never thinks of asking himself whether these heavenly bodies do not also follow other people. When we ask the captious question as to which of two people walking in opposite directions the sun would prefer to follow, the child is taken aback and shows how new the question is to him. Children of 9-10 years, on the other hand, have discovered that the sun follows everybody. From this they conclude that the truth lies in the reciprocity of the points of view : that the sun is very high up, that it follows no one, and that each sees it as just above him.

What we said just now about dreams is also to a certain extent germane to the present process : the child begins by regarding his own dreams as true, without asking himself whether every one dreams the same, as he does.

Side by side with this realism of perception and images, there is a logical realism which is far more important. We met with numerous examples of it in the course of our studies on child logic. Before the age of 10, on the average, the child does not know that he is a brother in relation to his own brothers. The ideas of right and left, of dark and fair, of the points of the compass, etc., are all subject to the law which is occupying us at the moment. These conceptions are at first regarded as absolute, so long as the personal point of view is accepted as the only possible one ; after that, the reciprocity of relations gradually begins to make itself felt (J.R., Chaps. II and III). In the present volume (as also in C.W.) we have pointed to several fresh examples of this process, examples which were of importance in forming the structure of reality.

Such are, above all, the ideas of weight and density. During the earliest stages, an object is heavy or light according to the immediate judgment implied by the child's own point of view : a pebble is light, a boat is heavy. Later on, other points of view are taken into account, and the child will say, for example, that such and such a pebble is light for him but heavy for the water, and that a boat may be light for the lake while it remains heavy for the child.

These last examples bring us to the third process which marks the evolution of the child's idea of reality : thought evolves from realism to relativity. This process is closely related to the last, and yet differentiates itself from it on certain points. During the early stage, the child tends to think of everything under the form of absolute substance and quality ; after that, bodies and their qualities seem to him more and more dependent upon each other and relative to us. Thus, substances become relations, on the one hand, because the mutual connection of phenomena has been seen, and on the other, because the relativity of our evaluations has been discovered. It would perhaps be as well to distinguish between these two aspects of " relativity ", but the second is, as a matter of fact, nothing but a combination of the first with the " reciprocity " of which we spoke just now. It will therefore be enough to point to this connection without complicating our classification.

The most striking example of this process is undoubtedly the evolution of the conceptions about life and movement. During the early stages, every movement is regarded as singular, as the manifestation, that is, of a substantial and living activity. In other words, there is in every moving object a motor substance : the clouds, the heavenly bodies, water, and machines, etc., move by themselves. Even when the child succeeds in conceiving an external motor, which already takes away from the substantiality of movement, the internal motor continues to be regarded as necessary. Thus a leaf is alive, even though it moves with the wind, i.e. it retains its spontaneity even though the wind is needed to set it in motion. Similarly, a cloud or one of the heavenly bodies remains master of its movements, even though the wind is necessary to start it on its path. But later on, the movement of every body becomes the function of external movements, which are regarded no longer as necessary collaborators but as sufficient conditions. Thus the movement of clouds comes to be entirely explained by that of the wind. Then these external motors are conceived as themselves dependent upon other external motors, and so on. In this way there comes into being a universe of relations which takes the place of a universe of independent and spontaneous substances.

Closely analogous to this is the evolution of the idea of force, since it is, as we saw, intimately connected with the idea of life.

The idea of weight supplies us with an excellent example of this advance towards relativity, and the evolution in this particular case is closely bound up with the advance towards reciprocity which we spoke of just now. During the earliest stages, weight is synonymous with strength and activity. A pebble sunk in water weighs on the water, even when the latter is motionless, and produces a current towards the surface. An object floats because, being heavy, it has the strength to keep itself up. Weight is an absolute thing : it is a quality possessed by certain bodies, a variant of that life, or substantial force which we have described. Later on, weight is regarded as relative to the surrounding medium : bodies float because they are lighter than water, the clouds, because they are lighter than air, etc. But the relation is still vague : the child simply means that for the water in the lake, such and such a boat is light, but no comparison has been made which introduces proportional volumes. The wood of the boat is regarded as heavier than an equal volume of water. Finally, between the years of 9 and 10, " lighter than the water " begins to mean that the body in question is, taken at equal volume, lighter than water. Thus do the ideas of density and specific weight make their appearance : absolute weight is succeeded, in part at any rate, by relative weight.

The explanation of shadows and of night also offers an example of the progression from substantialism to an explanation founded on relations. During the earliest stages, night and shade are substances that emanate from clouds and bodies in general, and which come and go more or less intentionally. In the later stages, night and shade are nothing but the effects conditioned by the spatial relations which regulate the diffusion of light.

In every domain the substantialist realism of perception is succeeded by explanation through geometrical and cinematic relations. Running parallel with this growing relativity of phenomena in relation to each other, can be seen a growing relativity of ideas and notions in relation to ourselves and our evaluations. Thus the establishment of relativity between phenomena leads to a relativity between the measurer and what is measured. The evolution of the notion of weight brings out very clearly this double development. On the one hand, as we have just seen, the weight of the body becomes relative to the medium constituted by the other bodies, and presupposes the establishment of a relation between weight and volume. On the other hand, the words " light " and " heavy " lose the absolute meaning they had during the earliest stages, and acquire a meaning that is relative to the units of measurement that have been chosen : the pebble is heavy for the water, light for us, etc. The absolute concept has become a relation. In such cases, the advance towards relativity ends by converging absolutely with the advance towards reciprocity of view-points ; in other words, the second and third processes as we distinguished them, finally merge into one.

Such, then, is the evolution of the notion of reality in the child. Three processes help to make it emerge from its initial realism and to orientate it towards objectivity. In what relation do these three processes stand to one another ? The first is of a purely social nature : the child replaces Ms own individual and egocentric point of view by the point of view of others and the reciprocity existing between them. The second of these three processes is of a purely intellectual order : substantialism of perception is replaced by the relativism of intelligence. The third process is both social and intellectual in character : in becoming conscious of his " I ", the child clears external reality of all its subjective elements, and thus attains to objectivity ; but it is, above all, social life that has forced the child to become conscious of his " ego ". Are we then to conclude that social factors determine the progress in the understanding of reality, or does this progress itself explain the development of social life ? Let us note, in the first place, that the three processes synchronise. All three begin very early, all three are very slow, they remain uncompleted at the close of childhood and survive throughout the intellectual development of the adult. There is therefore every reason to believe that they are interdependent.

As a matter of fact, we have here, as in the case of child logic, to suppose that social life is necessary to rational development, but that it is not sufficient to create the power of reasoning. Without collaboration between his own thought and that of others, the child would not become conscious of the divergences which separate his ego from that of others, and he would take each of his perceptions or conceptions as absolute. He would therefore never attain to objectivity, for lack of having ever discovered his own subjectivity. Without social life, he would never succeed in understanding the reciprocity of view-points, and, consequently, the existence of perspectives, whether geometrical or logical. He would never cease to believe that the sun follows him on his walks. He would be ignorant of the reciprocity of the notions of right and left, of dependence, in short, of relations in general. It is therefore highly probable that the relativity of ideas would elude him. This, at least, is what we endeavoured to show in an earlier volume (J.R., Chaps. II and III).

But at the same time, it would seem that reason, while it presupposes a social environment in which to develop, at one point transcends it. Once it has liberated the appearance of the logical norms in the child, the social environment enables him to become " permeable " to experience. And when this faculty has been acquired, the collaboration of logical reason and experience itself suffices to account for the intellectual development that takes place.

With this last remark we are led to analyse the evolution of the idea of reality from the point of view of the influence of environment on intelligence. Here, we are at once confronted with a paradoxical fact : compared with ourselves, the child is both closer to immediate observation and further removed from reality. For, on the one hand, he is often content to adopt in his mind the crude forms of actuality as they are presented in observation : one boat will float because it is light, another, because it is heavy, etc. Logical coherence is entirely sacrificed in such cases to fidelity to fact. The causality which results from phenomenism of this kind is not unlike that which is to be found in primitive races and has been wittily compared by M. Brunschvicg to causality as it was understood by Hume. Anything can produce anything : so long as two facts are given together in raw observation, the one may be considered the cause of the other. We shall give this the name of fihenomenistic causality. It is the starting-point of a large number of childish notions. The moon that follows us, the clouds that go with the rain, the heaviness or lightness of floating bodies all these are phenornenistic associations at the start, which later on lead the child to say : the moon moves along because I do, the clouds are the cause of rain, floating is determined either by heaviness or by lightness, etc. When the child is presented with a new and unknown fact, such as a toy engine, this phenomenistic turn of mind comes out very clearly: the child makes associations at random, connecting any one thing with another, and immediately takes these associations as causal.

But, in another sense, the child is far farther away from reality in his thought than we are. Reality, for him, is still overgrown with subjective adherences : it is alive and artificial ; words, dreams, and thought reside in external objects : the world is filled with forces.

Phenomenistic relations themselves take place against a background of dynamism, either magical or animistic. Thus the fact that the moon follows us is immediately interpreted by means of pre-relations, one of which makes the child think that he has power over the moon, the other that the moon is interested in 'him.

This paradoxical dualism of pure phenomenism on the one hand, and of magical dynamism, animistic or artificialist on the other, is a new manifestation of that dualism of juxtaposition and syncretism which we examined in our earlier volumes (L.T., Chap. IV, and J.R., Chap. I). Child thought proceeds by juxtaposition of its elements. There is a synthesis, but the terms juxtaposed in this way are embodied in subjective schemas, syncretism consisting in connecting everything with everything else in accordance with the hazards of a mental orientation that is subjective and egocentric.

The counterpart of this paradoxical dualism of phenomenism and egocentricity is the following : as it develops, the idea of reality tends to become both desubstantialised and desubjectified. Reality, as the child conceives it, is desubjectified with the years, in the sense that the adherences of animism, of artificialism, and of dynamism are progressively eliminated. But at the same time, reality becomes desubstantialised, in the sense that a universe of relations gradually takes the place of the universe of absolute substances which were assumed by primitive perception. Movement, weight, shadow, and night, force, etc., are all of them notions whose evolution is characteristic in this respect.

In short, from the point of view of the action of the physical environment upon the child, we are faced with a continual paradox : the child is both nearer to and farther from the world of objects than we are, and in evolving an adult mentality he both advances towards reality and recedes from it.

If by empiricism we mean the doctrine according to which intelligence is entirely moulded by its environment, then we must admit that empiricism fails to explain the paradox we have spoken of, and for two reasons. In the first place, the empirical hypothesis is in conflict with the circumstance that the more primitive childish intelligence is, the farther it is removed from what we call reality : the initial confusion between the ego and the external world, the existence of lasting subjective adherences, are sufficient to show that the physical environment does not imprint itself as such upon the mind of the child, but that it is assimilated by means of schemas that are drawn from internal experience. According to the opposite thesis, the most primitive thought would have, on the contrary, to be that which was nearest to external objects : pure phenomenism would have to exist to the exclusion of any adherence of internal origin. It may be objected that these adherences can be explained by the empirical hypothesis, in the sense that the mind first discovers its own internal sensations, and then having associated them together (in the phenomenist manner) proceeds to project the result of those associations into the external world. This projection would seem to account for the paradox mentioned above, but the question remains as to whether the child really discovers his own mind in the same way as he explores by pure phenomenism a new object presented from the external world. Now, obviously, the child does nothing of the kind. "For, while the external world is perceived by means of schemas of internal origin, internal phenomena (thought, speech, dreams, memory, etc.) are in their turn conceived only through schemas due to external experience. The child vivifies the external world and materialises the internal universe. Thus at no point is the child mind completely ruled by pure phenomenism. What one observes is a reciprocal digestion of objects by endogenous schemas, and of psychical experience by exogenous schemas. Such phenomenism as exists in the child is never pure, and those who want to revive empiricism in order to explain the point of contact between child thought and the external world will have to complete it by a theory of intellectual assimilation, which means that they will have to abandon empiricism as such.

In the second place, the development of the idea of reality in the child seems to us inexplicable on the empirical hypothesis for the reason, to which logicians have often drawn attention in another sphere, that a false notion does not in itself differ from a true one. Thus the observation that the heavenly bodies follow us about contains nothing incorrect in itself. Only the confrontation of one's own point of view with that of others, and the construction of a world of impersonal relations show the impossibility of the phenomenon. Further, the concepts of weight, of density, of force, of movement, etc., cannot be imposed by any experience : only an act of choice, due to the logical structure characteristic of the particular stage of intellectual development can account for the presence of one particular conception among the collection of possible conceptions. There is no truth without relations, and there can be no concepts without choice ; so there can be no facts without interpretation, and no interpretation without certain dominant mental tendencies. The construction of reality cannot, therefore, be the product of pressure exerted then and there by the physical environment : reality is built up by intelligence, which means that reality, as it appears to the child, is the fruit of a genuine collaboration between the mind and the world around it.

Does this mean that we are to regard knowledge as a free construction of the mind, which would mean admitting a more or less rigid apriorism ? It should be noted that from the point of view of modern biology such a theory is not absurd. According to a contemporary school, the environment does not act upon hereditary mechanisms : acquired characteristics are not transmitted. Thus the reaction of the organism to its environment is conditioned by a structure which is transmitted from germ-cell to germ-cell, without suffering any external change other than chemical influences capable of intoxicating the whole organism. Such a structure is therefore radically independent of its environment and of the influences which this environment may exercise upon the soma. From this point of view, intelligence itself might easily be thought to possess a fixed structure, and the sensations by means of which the environment imprints itself upon the mind might be thought to mingle and combine in obedience to laws that were completely foreign to this environment. An a priori theory of knowledge works in perfectly well with a psychological biology. Apart from any question of heredity, moreover, the " Gestalt " psychology of Wertheimer, Kohler, and Koffka, ^{[ 2 ]} would seem to constitute a revival of apriorism : for if every psychic synthesis implies the appearance of a general feature that is new and not reducible to the sum of its parts, it seems that the structure of knowledge must be irreducible to that of reality. Kohler, ^{[ 3 ]} it is true, claims to find " structures " in the physical world, which fact seems at variance with our interpretation. But it remains an open question whether the " Naturphilosophie " which at bottom M. Kohler is trying to establish alongside of his experimental psychology would not lose some of its value if the author were to express the problem in terms of Criticism instead of trying to retrace in the physical world the same structures as he finds in the human spirit.

Be that as it may, it is possible to conceive an apriorism defined within the limits of experimental psychology. But the a priori hypothesis is, to say the least of it, unnecessary for the interpretation of our results. In addition, it seems unable to account for two facts, with the second of which it is even in flat contradiction. The first is that, far removed though the child may be in certain respects from pure observation, his docility towards experience is nevertheless sufficient to give rise to a type of causality that is properly phenomenistic. Above all and this is the second fact the idea of reality undergoes with age a progressive transformation. This means that the categories of child thought are capable of evolution. Now, apriorism presupposes fixity, whereas every change in the actual structure of thought seems to show that this structure is plastic to the action of external things, whether this action be immediate or remote.

The truth, in short, lies half-way between empiricism and apriorism : intellectual evolution requires that both mind and environment should make their contribution. This combination has, during the primitive stages, the semblance of confusion, but as time goes on, the mind adapts itself to the world, and transforms it in such a way that the world can adapt itself to the mind. What is the mechanism of this adaptation ? This is the question we shall try to answer in 4. In the meantime, let us make a definite statement of the results of our analyses regarding the ideas of cause and of law.

Causality and the Child

The evolution of the idea of causality in the child follows very much the same lines as those we have just been observing in connection with the notion of reality. But it is important at this point to recall the facts in all their complexity. If we decide to do away with any arbitrary simplification, we shall find no less than 17 types of causal relation in child thought. Let us first make an analytical survey of these, and then try to establish the laws which control their evolution.

The first type is that of psychological causality, which is both causal and final ; let us call it the motivation type. For example, God or men send us dreams because we have done things that we ought not to have done. This type is, no doubt, the most primitive, but it is also the one that survives the longest. Its scope is reduced, however, as mental development proceeds, since things in general cease to be thought of as conscious or as specially made by men. But during the primitive stages the motivating relation is omnipresent. Elsewhere we have designated as pre-causality this tendency to take a psychological motive as the true cause of everything : there are two Saleve mountains, because there must be one for grown- ups and one for children, and so on.

The second type is that of pure finalism. This type overlaps with the preceding one to a certain extent, but it gradually separates itself from it. When the child says that the river flows so as to go into the lake, the river is not necessarily endowed with consciousness, nor the makers of things with a motive. There is simply finality, without either the origins or the consequences of this finalism being noticed by the child. It is much the same when we say, in accordance with ordinary common-sense, that ducks have webbed feet so as to swim better. Implicitly, of course, there is present some idea of a divine plan, or of conscious and voluntary effort on the part of the duck. But these links with psychological causality are not perceived or made explicit, which shows that finalism is to be distinguished from motivation.

A third type is constituted by phenomenistic causality : two facts given together in perception, and such that no relation subsists between them except that of contiguity in time and space, are regarded as being connected by a relation of causality. A fire lit under an engine or alongside of it is regarded as the cause of movement, long before the child has attempted to find a single intermediary between this fire and the wheels of the engine. A child will say that one pebble sinks to the bottom of the water because it is white, that another pebble is light because it is black, that the moon remains suspended in the sky because it is yellow and bright, and so on. Anything may produce anything.

This form of causality is undoubtedly independent of the preceding forms, since the connections which it implies are imposed by the external world itself. But we cannot, as Hume would have liked to do, regard phenomenistic causality as the only original form of causality in the child. For phenomenistic causality is essentially unstable ; as soon as it is established, a phenomenistic relation transforms itself into one that is animistic, dynamic, magical, etc. Thus the child who thinks that he is the cause of the movements of the moon always interprets this relation in a way that goes beyond the limits set by pure phenomenism. Up to the age of 4-5, he thinks that he is " forcing " or compelling the moon to move ; the relation takes on an aspect of dynamic participation or of magic. From 4 to 5 he is more inclined to think that the moon is trying to follow him : the relation is animistic. The child who attributes to the fire the movement of the engine, immediately lends to the fire a force, the capacity for making air, etc.

The phenomenistic relation is therefore essentially vicarious ; it clears the way for the dynamic and other forms of relations which follow immediately upon it. One even wonders whether the phenomenistic relation would exist if there were not other forms of relations to support it. Rather does it seem that the mind of the very young child, saturated as it is with dynamism, with finalism, with animism, with magic, with pre-causality, with artificialism, etc., when it is confronted with new phenomena establishes at random spatial and temporal contiguities, and sees relations between any one thing and another.

In short, though it cannot be reduced to causality by motivation, etc., phenomenistic causality, such as we find it in the child, would seem to be capable of existing in a mind already attuned to other forms of relation. Thus phenomenistic causality is to these other forms what, in our case, induction is to deduction properly so called : we make inductions independently of any sort of deduction, simply by empirical groping, but we do so because we are perpetually on the lookout for some possible deduction.

A fourth type of relation is participation. This type is more frequent than would at first appear to be the case, but it disappears after the age of 5-6. Its principle is the following: two things between which there subsist relations either of resemblance or of general affinity, are conceived as having something in common which enables them to act upon one another at a distance, or more precisely, to be regarded one as a source of emanations, the other as the emanation of the first. Thus air or shadows in a room emanate from the air and shadows out of doors. Thus also dreams, which are sent to us by birds " who like the wind " (C.W., Chap. Ill, 2).

Closely akin to participation is magical causality, a fifth type, magic being in many respects simply participation : the subject regards his gestures, his thoughts, or the objects he handles, as charged with efficacy, thanks to the very participations which he establishes between those gestures, etc., and the things around him. Thus a certain word acts upon a certain thing ; a certain gesture will protect one from a certain danger ; a certain white pebble will bring about the growth of water-lilies, and so on.

In its origins participation is connected with certain conditions of logical structure to which we shall return later. But participation and magic are connected even more closely with psychological causality. For not only does the child regard his desires as efficacious in themselves, but all realism presupposes a realism of thought and gesture, that is, a realism of signs in general. And this realism is the result of that initial confusion between the self and the external world which is the very thing which primitive psychological causality implies.

A sixth type, closely related to the preceding ones, is moral causality. The child explains the existence of a given movement or of a given feature by its necessity, but this necessity is purely moral : the clouds " must " advance in order to make night when men go to bed in order to sleep ; boats " have to " float, otherwise they would be of no use, etc.

Closely akin to psychological causality or finalism, but with an added element of necessity, moral causality is also related to that form of participation which we have called dynamic : external objects have intentions which participate with ours, and in this way our desires force them to obey us in accordance with purely moral or psychical laws.

The seventh type of relation is artificialist causality. Psychological causality or pre-causality is at the start neither purely moral nor purely physical. A given event is explained straight away by the intention or motive at the back of it, but the child does not ask himself how this intention has worked itself out in action. Since all nature both matter and consciousness is nothing but life, the problem does not arise. As soon as the two terms come to be differentiated, artificialist causality appears, at the same time as moral causality and in the nature of its complement : the event or object to be explained is then conceived as the object of human creative activity. This shows the family resemblance to the preceding types which are all capable of growing into artificialism or of finding themselves completed, thanks to this new type of relation.

An eighth type is animistic causality, or what might be called causality by realisation of form. The existence of a character or form is explained by an internal biological tendency that is both alive and conscious. The sun is what it is because, after having been made by men, it grows. Mountains have grown, etc. Clouds and the heavenly bodies move along because they are alive. This is the complement of artificial causality ; external motors act on things only if the latter possess an internal motor capable of carrying out the directions and commands received from without.

A ninth type, which is simply left over from the preceding, is constituted by dynamic causality. Once animism proper has been eliminated, there still remain in objects forces that are capable of explaining their activity and their movements. Thus, primitively, force is confused with life itself, but dynamism outlives animism, just as finalism outlives pre-causality. Throughout this book we have had occasion to point to the very general character of childish dynamism.

A tenth type of relation is explanation by reaction of the surrounding medium. It is, properly speaking, the child's first genuinely physical explanation. For all the preceding forms appeal either to motives or to intentions, either to occult emanations or to mystical manufactures. But reaction of the surrounding medium implies, and, for the first time, the need for defining the " how " of phenomenon, i.e. the need for continuity and contact. At first reaction of the surrounding medium still goes hand in hand with animistic dynamism. Only it completes this dynamism with a more exact mechanism. Thus the clouds are regarded as setting themselves in motion, but once this movement is started, the clouds are driven along by the air which they produce by their flight. Later on, reaction of the surrounding medium will serve to explain purely mechanical movements. Thus projectiles which are supposed to be devoid of any spontaneous movement are pushed along by the air which they make in moving. The prime motor is thus the hand that throws the projectile, and not an internal force, as in the case of the clouds. We have seen what universal use children make of explanation by reaction of the surrounding medium. The movement of clouds, of the heavenly bodies, of water, of air, of projectiles, of bicycles, of aeroplanes, of boats, of tops, the effects of centrifugal force all these are reduced to a schema which up till now was thought to be peculiar to Greek and mediaeval physics.

An eleventh type of causality is constituted by mechanical causality properly so called, i.e. explanation by contact and transference of movement : the wind pushes the clouds, the pedals make the bicycle go, etc. This form of causality appears between the years of 7 and 8. It is always the result of eliminating dynamism. The child who always begins by attributing these movements to the collaboration of two forces, one internal (the object's own force) and the other external, gradually comes to look upon the internal motor as unnecessary. At this point explanation becomes mechanical. Very often the schema of reaction of the surrounding medium serves as a transitional stage between the dynamic character of the early stages and the mechanical character of the later explanations which the child may offer of a given phenomenon.

A twelfth type of relation is what may be called causality by generation. The explanation of movement naturally admits much more easily of being reduced to the mechanical type than the explanation of how bodies are actually produced. And at the stage when children bring their ideas of movement in general under the heading of mechanism, they still look to artificialism and animism to explain the origin of things. How, then, will an attempt at a rational explanation of this origin first present itself ? We saw that in the matter of the heavenly bodies, of the clouds, etc., as soon as the child has given up the idea that they were made by men, he tries to think of them as being born out of each other. This is the type of relation which we shall call generation. The sun, for example, is regarded as a little living ball that has come out of a fiery cloud ; the clouds themselves have come out of smoke, of air, of fire, etc. This is simply an extension of the animistic idea, with the added notion of a transmutation of substances. The idea of such transmutations is often imposed by relations of dynamic participation : the child feels that there is a relation between rain and clouds long before knowing that the one comes out of the other. He begins by saying that the clouds " come with " the rain, then, from the moment that the rain ceases to be thought of as made by men, he imagines that it comes out of the clouds. Thus a mere participation between intentions and movements (clouds are at first believed to accompany rain, just as the sun and moon accompany us, not for any reason, but because they are made to) gives rise little by little to the idea of generation proper.

From this type of explanation to the thirteenth, namely, to explanation by substantial identification there is but a step. We shall say that there is identification when bodies that are born from each other cease to be endowed with the power of growth as it exists in living beings. It is not always easy to draw the line, but it is useful to note the difference. For instance, great progress has been made when the sun is no longer believed to have been born of a cloud, but is regarded as a collection of clouds that have " rolled themselves tip into a ball." In the first case, the sun is looked upon as a living being that is very small at first and gradually grows bigger. In the second case, it is regarded merely as matter resulting from the fusion and burning of other inert matters. It will be remembered how frequent were these explanations by identification between the years of 8 and 10.

Once this thirteenth type of explanation has detached itself from the preceding type, it quickly gives rise to the more subtle fourteenth and fifteenth types.

The fourteenth is characterised by the schemas of condensation and rarefaction. For it is not enough for the child to say that the sun has been made by clouds that have rolled themselves up into a ball, or that a stone is formed of earth and sand. The qualitative differences have to be explained, which separate bodies of similar origin. The child then makes the following perfectly natural hypothesis. That the qualities of the sun result from the fact that the clouds have been " well packed" (serres). That the hardness of the stone comes from the fact that the earth is " close " (serree). Thus the matter that makes up bodies is more or less condensed or rarefied. Naturally, the child does not seek, as did the early pre-Socratic thinkers, to reduce all qualitative differences to differences of condensation. Nevertheless, between the ages of 9 and 10, we can see a very general attempt at explanation by condensation. This shows with particular clearness in the evolution of the idea of weight. According to the very young children, bodies are heavy in proportion to their size, and the child has no notion of differences in specific density. The older ones, on the contrary, say that the water is light because it is " thin ", or " liquid ", whereas wood and stone are heavy because they are " big ", " thick ", " full ", and so on. In short, putting aside mistakes in the evolution of weight, water is a rarefied matter, whereas wood and stone are condensed matters.

The fifteenth type of explanation is, in a sense, simply an extension of the last : it is that of atomistic composition. From the moment that bodies are regarded as the result of the condensation or rarefaction of original substances, it follows inevitably that sooner or later they will be conceived of as made up of particles tightly or loosely packed together. This is the conclusion which the child comes to with regard to stones : the stone is made of little stones, which are made of grains of earth, etc.

The sixteenth type of explanation is spatial explanation. Thus the explanation of the cone-shaped shadow appeals, in the later stages, to principles of perspective. Similarly, the explanation of the rise in the water-level due to the immersion of solid bodies appeals, after the age of 9-10, to the volume of the immersed body. This is rather an advanced form of explanation and consequently only occasionally to be found in children.

Finally, the seventeenth type of explanation, the most subtle, but towards which most of the others tend, is explanation by logical deduction. A good example of this was supplied by the experiment of the communicating vessels : the level of the water is the same in both branches, so some of the children told us, because the water can go equally well in one or both directions, and this is what explains the final equilibrium. This is explanation by the principle of sufficient reason. All mechanical explanations, spatial, atomistic, etc., appeal sooner or later to the principle of deduction, and this type of explanation is therefore one of increasing frequency after the age of 10~11. For example, from the laws which he has observed in connection with the floating of boats, the rise in the water level, the child gradually draws explanations which imply concepts, such as density, specific weight, and so on. These concepts are pure relations ; they are chosen in view of deductions to be made, and are not imposed by facts.

Having distinguished these seventeen types, we can now lay down three main periods in the development of child causality. During the first, all the explanations given are psychological, phenomenistic, finalistic, and magical (types I-VI). During the second stage, the explanations are artificialist, animistic and dynamic (types VII-IX), and the magical forms (III and IV) tend to diminish. Finally, during a third period, the preceding forms of explanation disappear progressively and give place to the more rational forms (X to XVII). Thus the first two periods are characterised by what we have called pre-causality (in the widest sense of the word), i.e. by the confusion of relations of a psychological or biological type in general with relations of a mechanical type, and true causality does not appear till about the age of 7-8 (third period).

Three processes seem to us to characterise this evolution : the desubjectification of causality, the formation of series in time, and the progressive reversibility of the systems of cause and effect.

The first process is very definite. Causality, like the whole of reality, is at first teeming with subjective elements. No distinction is drawn between motivation and physical causality (types I, II, VI) or between muscular and manual activity and mechanical action (types VII and IX), or again between the influence of mind on body or of the body on itself, and the influence of external objects on each other (types III, IV and VIII). As to phenomenistic causality, it is, as we saw, essentially vicarious and unstable. In the course of our studies on child psychology we had expected to fix upon 7-8 as the age before which no genuinely physical explanation could be given of natural phenomena. Our present enquiry entirely confirms this expectation. After 7-8 the more positive forms of causality gradually supplant the others, and we can say that at the age of about 11-12 the evolution is completed. There is, therefore, in the domain peculiar to causality a process of evolution exactly similar to that to which we drew attention in speaking of reality : confusion of the self and the universe, then progressive separation with objectification of the causal sequences.

The second process is peculiar to causality : it is the constitution of temporal series. What strikes one most about the child's more primitive forms of causality is the immediate and almost extra-temporal character of the relation. It was the same with participation : the moment we have made a certain movement in a room, the air rushes into our hands through closed windows. As soon as we bring a copy-book up to the table, the shadows of the sky or of the trees come and interpose themselves between our hand and the wood of the table. As soon as we walk along the street, the sun or the moon begin to move. Not a thought is given to the question of distance or of how long the action would have to take in travelling from the cause to the effect. Joined to this relative immediacy, is a remarkable absence of interest as to " how " phenomena occur. Thus, according to the very youngest children, the pedals make the wheels go round without being in any way attached to them, simply by influence. The fire makes the wheel of the engine turn, even 50 centimetres away. There is no contact, during the primitive stages, between cause and effect. Immediacy of relations and absence of intermediaries, such are the two outstanding characteristics of causality round about the age of 4-5. But such features are completely absent from children of 11-12 in subjects of which they know nothing. Thus, it is more or less impossible for a child of 10 to understand how a motor-car works. Nevertheless, the child presupposes pipes, cog-wheels, chains, and belts to act as intermediaries between the petrol and the wheels. From 7-8 onwards, excellent explanations from memory are to be found concerning bicycles, whereas before this age, the various parts were believed to act on one another, but never in the same order. In the sphere of nature, the establishment of contacts and of series in time is also very definite. The step from pre-causality to mechanical causality in the explanation of movement is a good example ; and the explanation by reaction of the surrounding medium an explanation which serves as a transition between the two extreme types of causality marks precisely the presence of a need for a chain of intermediate links between cause and effect. There is, therefore, in every sphere a constant progression towards the establishment of series of intermediaries, and series ordered in time.

We shall not go so far as to say, however, that even in the earliest forms of the causal relation there is not a feeling of before and after. Indeed, it may be that although cause and effect are infinitely close, the child still refuses to reverse the terms. This is a special problem, which we hope to take up again with I. Meyerson. All we are saying for the moment is that between cause and effect there is no series properly so called : there is nothing at all, and progress consists precisely in establishing chains of intermediaries such that each should be the effect of the one that precedes it and the cause of the one that comes after it.

A third process in the evolution of causality is the progressive establishment of reversible series. At first sight this aspect of the question seems to be in contradiction with that which we have just been discussing, but it will soon be seen that this is not the case. If we examine a mechanism of any complexity that has been correctly understood by a child of 8-10, we shall always find that it is a reversible mechanism. When the stone has been understood to be composed of little particles of earth, the child admits that the stone can be decomposed into earth. When a child has understood how the pedal of a bicycle makes the wheel move round, he sees that by turning the wheel the pedals can be made to turn. The child believes that the cloud is made of smoke and that this is how it produces fire : from this he concludes, sooner or later, that the cloud can turn itself back into fire and in this way give rise to thunder and to the heavenly bodies. This reversibility does not exclude the existence of a series in time. Only, the series in question is one that can happen in two different directions.

Now, if we really look into the matter, we shall find that all the more advanced forms of explanation in the child are reversible. Mechanical causality is obviously so. So also is causality by substantial identification, i.e. by transmutation of the elements. Children, at any rate, have no doubts on this point : air produces fire, and fire, air ; water produces smoke, and smoke, water, etc. Causality by atomic composition follows the same rule, so much so, indeed, that the child always becomes involved in vicious circles when he applies it for the first time : the stone is made of earth, and vice versa. Later on, these vicious circles cover more ground, i.e. a greater number of intermediaries are introduced between the extreme terms of the series. But the explanation is still circular. As to spatial and deductive explanations, it goes without saying that they are reversible, since they omit the time element.

The primitive forms of causality, on the contrary, are all irreversible. Take, for example, psychological, magical, finalist causality, etc. : an action or a motive explains a given phenomenon, but the reverse is inconceivable. Artificialism is in the same case : men made the universe, but the universe did not make men. Participation raises a special problem : its immediacy seems to imply reversibility. But in every participation there are emanations and there is the source from which they issue. This is therefore anything but reversibility. Of all the types of causality that come before that of mechanical causality, explanation by reaction of the surrounding medium is the one that points most clearly to a beginning of reversibility. But this is precisely the type that clears the way for the higher forms of explanation.

The progress from irreversibility to reversibility is thus continuous. This process, moreover, seems extremely natural if we bear in mind the manner in which the idea of reality grows up in the child. For the primitive universe is both strewn with subjective adherences and very near to immediate perception. Now, in so far as it is tinged with the child's subjectivity, this universe is irreversible : the flow of consciousness, psychological time, the whims of desires and actions which follow one another without order or repetition all these things are projected in their entirety into the external world. Similarly, in as much as it is near to immediate perception, the child's universe is irreversible, for perception never shows us the same sun nor the same trajectory, nor the same movements twice. Events cannot happen over again in the same way. It is the mind that builds up reversible sequences underneath perception. To the extent that the child's universe is removed from these constructions and close to the immediately given, it is irreversible. Thus the advance towards reversibility shown by the development of child causality follows exactly the same lines as those underlying the processes defined in connection with the idea of reality.

It is not our business to compare these results with the conception which M. E. Meyerson has so ably defended in his works on explanation in the sciences. Anyone can see the affinity between the three processes we have distinguished and that " identification " to which M. Meyerson has reduced the progress of causality. At the same time, we prefer the word reversibility to the word identity for characterising the causality of the later stages, because, even if it is true that to explain always means in the last resort to deduce, it is not by any means so sure that to deduce means the same as to identify. One can also say that deduction is simply, as Mach and Goblot have shown, a construction, and this view would tend to bring together causality and legality, after having kept these two notions strictly apart.

What conclusion are we to draw from our study of child causality as regards the influence of physical environment on the growth of intelligence ? Here, again, we can only note that the more primitive the ideas of the child, the further removed are they from the physical environment as we know it. All the early forms of causality magic, finalism, animism, artificialism, and, above all, dynamism, are inexplicable if we do not allow that between environment and consciousness there come to be interposed schemas of internal origin, i.e. psycho-physiological schemas. The starting-point of causality is a non- differentiation between inner and outer experience : the world is explained in terms of the self.

Are these facts, then, a confirmation of Maine de Biran's psychological realism? We have already mentioned, in connection with the idea of force, what seemed to us to be the difficulties of such a doctrine. With regard to causality in general it will therefore be enough to say that if the world is interpreted by the very young child in terms of his own " I ", the " I " in its turn is explained in terms of external experience. We have no more direct cognisance of the self than we have of external objects. Participation and magical causality show, on the contrary, that it is for lack of having discovered his own subjectivity that the young child feels his gestures, his words, and his thoughts, to be bound up with the objects themselves. We do not therefore, as Maine de Biran maintained, begin by discovering internal causality and then proceed to transfer it into objects. Causality is the result of a sort of bodily contact between the organism and the world, which is prior to consciousness of self, and this bodily contact takes us back to the notion of an assimilation of things by thought, a notion to which we shall return later in 4.

On the other hand, to make causality into an a priori form, fixed once and for all in the structure of the mind, is to raise insuperable difficulties. For, after all, why should not causality appear from the first in its completed form ? Why does it evolve to the extent of giving rise SUMMARY ANJD CONCLUSION 273 to 17 forms, distinguishable among children alone ? Why is it dependent upon the influence of environment ? If these are, in causality, signs of a structure which eludes empirical explanation, it will have to be admitted that this structure is plastic, and this leads us back once more to the hypothesis of an assimilation of external objects by the organism, such that the objects modify the organism, and such that the organism in its turn adapts things outside to its own peculiar structure.

The Child's Idea of Law

The notion of law presents in the child, as indeed in the whole history of thought up to modern times, two complementary features universality and necessity. Law is a constant and necessary relation. M. E. Meyerson has done very useful work in drawing attention to the difficulties of the position, and in distinguishing clearly between legality and causality, legality being simply generality, while causality alone could serve as a foundation for necessity. But from the genetic point of view, the concepts of natural and of social law have always reacted upon each other. For the child, who alone concerns us here, law is inconceivable without necessity. Let us therefore try, without the help of any preconceived ideas, to trace the development of these conceptions.

We can distinguish three periods in the evolution of law in the child. Each of these is characterised by the peculiar relationship in which generality and necessity stand to one another. During the first, generality is non- existent ; as to necessity, it is purely moral, physical determinism not having been separated from the idea of social obligation. During the second period, these two types of necessity are differentiated, and generality comes into being. During the third period, generality is established, and physical determinism is accompanied by logical necessity, which is the last term in the evolution from moral necessity.

The first period lasts till about the age of 7-8. During this time, there are no natural laws. Physical and moral determinism are completely confused with each other. More exactly, any law observed to hold among external objects is regarded as a social law, and things are believed to behave in accordance with rules that are imposed upon them from outside. This will be recognised as the combination, to which we have so often drawn attention, of animism and artificialism : nature is a society of living beings of whom man is the master and at the same time the creator. All recurring movements are explained primitively in this way. The movements of the sun and the moon, that of the clouds, the return of night, the course of rivers and of waves all these are subject to the same principle : things have obligations towards us. Before the age of 7-8 we found no example of movement regulated by purely physical laws. There are always two motors which ensure the movement, thanks to their collaboration : an internal motor, which is the obedient will of the moving object, and an external motor, which is at first man himself, and then certain other bodies which play the part of masters or of more vigorous enemies (such as the sun driving away the clouds and the night summoning them).

We saw the counterpart of these facts when we came to analyse childish animism (C.W., Sect. III). When children between 5 and 8 are asked whether the sun could go away if it wanted to, they always answer that it could : if it does not go away, it is because it " has to shine a little longer", or because "it has to lighten us during the day ". Clouds cannot go because they show us the way, etc. In short, if there are natural laws at work, it is not because the bodies in question are physically determined ; they could perfectly well evade the law if they wished to. It is simply that they are obedient.

This is why for the child, as for Aristotle, the two notions which for us are to characterise the world of laws violence and chance come under no law whatsoever. With regard to violence, we cannot claim to have found in children any explicit belief that could be compared to the distinction made by Aristotle. But the whole of the moral conception of law of which we were speaking just now, shows that natural movements are, in the child's eyes, not violent, but harmonious and free. This implicit belief is constantly coming to the surface. Thus a boy of 6 told us that the sun is clever " because he wants to make it warm " ; we are clever " when we don't do what we ought not to do " ; as to clouds, they are " not clever ", because " they try to fight the sun ".

We have dealt at great length elsewhere (L.T., Chap. V) with the question of chance, and shall therefore content ourselves now with recalling our results. Before the age of 7-8 the child seeks, as far as possible, to eliminate chance from nature. The very way in which he formulates his " whys " shows that for him everything has a reason, even when to us it seems fortuitous and contingent. Now, whatever contradicts this conception provokes, by the mere fact of doing so, the maximum of curiosity on the part of the child. And this is why we find the child trying to find the reason or justification for a whole number of facts which for us are inexplicable because they are due to chance : why there is a big and a little Sal&ve mountain, why pigeons are like eagles, why one person has smaller ears than another, etc. In short, law may be arbitrary in the sense that the will of gods or of men may be capricious ; but chance is banished from nature, for everything admits of justification or of motivation, since everything in nature has been willed.

To conclude, during the first period the necessity of law is entirely moral, and physical necessity is simply the lining as it were of this moral necessity, i.e. it is simply dependent upon the force and authority of the masters of nature. What is the type of generality that could go hand in hand with such a conception of necessity as this ? History has shown over and over again that to a moral conception of natural law there corresponds a belief in the possibility of numerous exceptions* These exceptions are of two kinds : miracles and the resistance of external things (monsters, etc., conceived by Aristotle as the resistance of matter in relation to form) . The child thinks in exactly the same way. Corresponding to the confusion existing between natural and moral law, there is, during this first stage, a complete absence of generality in the laws of nature.

We have repeatedly had occasion to note this absence of generality. First of all, in child physics : according to one and the same child a given boat will float because it is heavy, another, because it is light ; a given body immersed in water will raise the level because it is big, another will fail to do so although it is heavy, and so on. There are certainly laws, but the exceptions are as frequent as the rule. As a rule, the water in rivers flows down, but it might just as well flow upwards. Sometimes the wind drives the clouds before it, sometimes they move by themselves.

The possibility of miracles is, of course, admitted, or rather, miracles form part of the child's conception of the world, since law is a moral thing. Children have been quoted who asked their parents to stop the rain, to turn spinach into potatoes, etc.

During the second period, on the contrary, we see two processes at work which are complementary to each other and take place between the years of 7-8 and 11-12 ; on the one hand, moral necessity and physical determinism become differentiated, and on the other hand, law becomes general.

We have seen numerous examples after 7-8 years of the manifestation of physical determinism. Thus the movements of water and clouds come fairly soon to be attributed to purely mechanical causes : the water cannot do otherwise than go down the slope, the cloud is bound to move forward when there is wind, and so on. At 8 years old on the average, the mechanism of a bicycle is completely understood, and this fact alone points to a mentality that is beginning to bend to the idea of uniform and physically determined causal sequences.

But the clearest index of all is the appearance of the idea of chance. At about 7-8 the child begins to admit that there are things which serve no particular purpose and events due solely to chance encounters. Thus the arbitrary and capricious element, which during the preceding period went hand in hand with the conception of law, has turned into chance, which means that it has lost its moral aspect and has taken on a purely physical character.

It goes without saying that moral necessity is not changed into physical determinism at a stroke. Up to the age of 11-12, many natural laws are still thought of as moral. The movements of the sun and the moon, for example, are those which are interpreted the longest as obeying purely moral laws. Moreover, determinism conquers only the details of events, whereas the body of natural laws taken in their most general aspect remains moral in character. For instance, the child may know that the formation of rain and the movement of clouds are due to purely physical processes, but he continues to believe that if there are clouds it is "for the gardens", Finalism dies hard.

As to the generality of law, it naturally grows in proportion as moral necessity decreases. As soon as the movement of rivers is interpreted physically, water is considered as always flowing in the same sense, etc. This is the age when the child seeks to avoid contradictions, and when he begins to understand that a law either is general or is not.

Finally comes the third period, which sets in at about l0-11. During this period, the generality of law naturally takes deeper root. But what becomes of necessity ? For a general law is not, as such, necessary. The child may very well discover the absolute regularity of a given physical law (such as that light bodies float, and that heavy ones sink, etc.) but there is no physical necessity that can account in his eyes for this regularity. Apart from cases of apparent constraint (as when the wind pushes the clouds) there is no physical necessity. What makes a law necessary in our eyes is its deducibility : a law is necessary if it can be deduced with a sufficient degree of logical necessity from another law, or from sufficient geometrical reasons. Thus a paradox attends the evolution of law in the child : as the generality of laws increases their necessity grows less (in so far as this necessity, as during the first two periods, is moral). For as the child abandons the idea of moral necessity for justifying laws, he is faced with a mere generality of fact that is, however, devoid of any foundation whatsoever. When we ask the child why water goes down whereas smoke goes up, he can answer that heavy bodies fall whereas light bodies rise, which is certainly a general law ; but when we question him further as to why this is so, he can answer nothing. The younger ones invent a moral reason to get out of the difficulty, but the older ones are nonplussed. Are we then to admit that moral necessity disappears without leaving a trace, or shall we, during this third period, find this necessity reappearing in a new form ? We believe this last solution to be the right one, and that after the age of 10-11, moral necessity becomes logical necessity.

For at about this age various attempts at deduction and logical justification of laws manifest themselves. We recalled just now the attempt to reduce the principle of the communicating vessels to an explanation by sufficient reason. During the same age, we find explanations of floating by the notion of density or by the relation of weight to volume or form, all of which are attempts in the right direction. Questions about shadows, about immersed bodies, etc., also give rise to so many deductions of laws.

But what possible affinity can we allow to exist between moral and logical necessity, two types which at first sight seem so radically distinct ? Let us imagine a universe controlled by moral necessity. Now take from it all direct influence, all consciousness and will, and also all mystical activity exercised by man upon things. Much will still remain. There will still be the ideas of order, of organisation, of regularity, coherence, and intelligibility. Thus there remains the possibility of explaining one group of phenomena by taking another group as our starting point. All this is characteristic of logical necessity ; for at its root lies the conviction that every law that has been observed empirically must admit of justification by an appeal addressed no longer to the will and the emotions, but to reason.

What conclusions can we draw from our analyses from the point of view of the influence of environment ? We cannot honestly account for the evolution of the idea of law by any direct action of the physical environment. The idea of law, like those of cause and reality, is much more remote from the child than it is from us. From the very first, we see that the feeling of the moral necessity of law comes before any exact knowledge of the law itself. From the first, we can observe a fusion of physical experience and internal feeling. No doubt only " conditioned reflexes " and motor anticipations will supply the contents of the most primitive laws discovered by the child, but this purely empirical attitude in no wise explains the " obligatory " character of law.

This feeling of obligation, thanks to which the child co-ordinates things as they are presented to him, can therefore only be of internal origin. Nothing in nature herself can give the child the idea of necessity. To immediate perception nature is full of whims, and there are no laws that do not admit of numerous exceptions. The child sees this himself when he begins by refusing to allow any generality to the laws of nature. But why does he not stop there ? On the contrary, we are faced with the paradoxical fact that the earliest relations perceived are conceived as morally necessary even before having become intelligible. This strikes one particularly in the explanations given of floating or of movements in general : wood stays on the water because it must, the moon follows us because it is forced to do so, the river flows because it must flow, etc. Sully long ago noted this feeling of a rule, but these primitive rules have a meaning that is far more moral than logical : they imply, not generality, but simply obligation. The same thing strikes one also in the explanations children give of machines : all the parts present must be there for the machine to go, says the child, before having understood, or even guessed at the true r61e of these parts ; thus the lamp, or the brake, or the tyres of a bicycle, will seem as necessary to movement of the whole as do the wheels and the pedals.

If, then, the feeling of necessity is not due to the pressure of the physical environment, is it perhaps the outcome of social surrounding ? For one may well wonder how a child would think who was removed from the authority of his parents. Is not the concept of nature simply the outcome of comparisons with family life rather than the result of pre-relations properly so called ? Only one of the various theories put forward to account for the moral obligation attaching to the idea of law tends to make us subscribe to this hypothesis. We are thinking of the fine work done by M. Bo vet on the conditions of obligation in the conscience. ^{[ 4 ]} There is obligation in so far as the orders are given by persons for whom we feel respect, respect being a s^i^ generis mixture of fear and love. There is obligation in the measure that a relation exists, not between person and person, but between the small and the great.

It would therefore be in the relation of the child to its parents that we should have to look for the origins of law. Only, in M. Bovet's view, this relation would go a long way back. And indeed, observation shows the extreme importance and the great precocity of this attitude of mingled fear and love, which according to M. Bovet is precisely what constitutes respect. The child's conceptions about natural law would therefore seem to have their roots in a very universal and a very primitive reaction.

We must not therefore say that the child's concept of nature is based simply on an analogy with family life. What is at work is a genuine pre-relation. Moral obligation forms part of the very structure of the child's mind, if we admit that the feeling of obligation is derived from the earliest contacts of the child's will with that of its parents, for these early contacts condition the whole of the child's mental life.

Assimilation and Imitation

In studying the evolution of the idea of reality in the child, we found in the primitive stages a dualism that was little short of paradoxical. On the one hand, the universe of the child is closer to immediate perceptions, closer to external things than the universe of the adult. On the other hand, it is more subjective, more permeated with characters that are, in fact, taken from internal experience. The same phenomenon appears in the evolution of the ideas of cause and law. The most primitive laws which the child lays down take account of every feature and every detail of immediate perception : they are not general, they have exceptions, and the explanations are over-determined. They point to a type of necessity borrowed from inner experience, and the pre-causality which goes with them is marked by a confusion of motivation with physical connections.

The same dualism is to be found in the logical structure of child thought. Child thought proceeds by juxtaposition. Judgments are not interrelated, because each espouses the object in all its detail and takes no account of the judgments that came before it. But there is syncretism, which means that the lack of objective relations is made good by an excess of subjective relations.

In brief, there is dualism everywhere realism on the one hand, subjective adherences on the other. It should be noted in passing that the situation is closely analogous to what M. Brunschvicg has so clearly shown to exist in the mentality of primitives. " With regard to relations of causality, the metaphysics of dynamism mingle with the phenomenism of contingency ", ^{[ 5 ]} We would therefore seem to be in the presence of a very general feature of the evolution of thought. But we shall confine ourselves, in what follows, to the psychology of the child, which we have chosen as our special province.

In order to explain this dualism, we shall be obliged to carry into the sphere of psychology the epistemological distinction between the form and the matter of knowledge. All we need do is to agree upon the meaning which we wish to give these terms in the domain of practical experience. The line which we proposed to follow at the opening of these general conclusions will supply us, in this connection, with a perfectly natural solution.

If we agree to take as our system of reference, nature as science describes her, then we shall call matter, or content of the child's knowledge all that experience and observation impose upon the child. And we shall name form of the child's knowledge everything that the child adds to this matter, that is the pre-relations and pre- notions which we, as adults, have shed. The choice of system of reference is, we repeat, a convention, but we are making use of this convention quite consciously, and shall not allow it to lead us into epistemological realism of any kind.

In the sphere of biology, moreover, the distinction between form and matter, such as we have just described, has a very definite meaning. The matter of knowledge belongs, as a special case, to the sum of influences which the environment exercises on the organism. These influences are in the main transmitted through the medium of substances that are absorbed as food, as the energy of heat, light, movement, sound, etc. Among these energies, some are not accompanied by consciousness, others liberate psychical states and reactions, and thus constitute a collection of pressures exercised by the physical environment on the motor activity of the subject and on the intelligence that is bound up with this activity. The form of consciousness, from the biological point of view, is, on the contrary, a special case of those structures which the organism imposes upon matter and assimilated energies. The organism has a structure which is retained by assimilation and which, moreover, conditions by selection the choice of substances and energies to be absorbed. Thus the influence of the environment can never be pure: every external stimulus presupposes an internal reaction, and what is assimilated by the organism is always and necessarily the result of an external influence and an internal digestion, which process of digestion may equally well be mechanical or motor as chemical.

Actually, indeed, it is impossible even so much as to think of " an environment " and " an organism " without abstraction. There exists between these two terms a complexus of relations, of changes and reactions which implies complete physico-chemical continuity. For it is impossible to mark the boundaries within an organism between what is living, what is functioning, and what is the collection of substances already assimilated or already rejected and hardened into deposits. The environment is being perpetually modified by the organism, and a given reaction which has just taken place will never take place again, as such, because of the fact that this reaction is a perpetual " becoming ".

Psychic life being conditioned by organic life, there seems no reason why it should escape these laws. The origin of knowledge is no doubt fraught with mystery, but however far back we go, we always find sensations and movements. At one time, psychologists laid all the stress on sensations, which led them to believe that knowledge was simply a replica of external reality. But they soon realised that there are no sensations without movements and that knowledge consists to a great extent of accomplished or anticipated movements. Now, once this has been admitted, it will have to be conceded, at least as a hypothesis, that motor schemas have already been formed at the moment when knowledge takes its rise, and that, in relation to knowledge, these schemas play the part of form which is independent of matter. There seems no reason why psychological assimilation should consist in reproducing the environment as it is, when we remember that all physiological assimilation is performed as the function of a structure that persists and that conditions every influence coming from the environment. On the contrary, there is a strong presumption in favour of the view that these two kinds of assimilation are partly analogous, both consisting in giving to matter coming from outside a form conditioned by the structure of the organism.

These biological hypotheses, it will be seen, account for the facts we have observed, and are necessary for that purpose. The whole structure of the child's idea of reality rests on a primitive lack of differentiation between the self and the external world, that is on a fusion of organic experience and external experience. External perceptions are moulded into muscular sensations ; the contents of organic consciousness is amalgamated to external things. Sometimes we have complete fusion of the element of muscular sensation with the external element, as in the idea of force. Sometimes we have progressive dissociation between the self and external objects, as in the evolution of the notion of cause. Thus the analogy is clear that exists between this confusion of the self with the world on the one hand, and the continuity of the organism with its setting on the other. In both cases, we have two terms in relation, but each of these terms exists only as a function of the other, and the interchange between them makes it impossible to dissociate them without mental work.

Let us try to define this initial complexus more closely.

Every fresh external influence exercised upon the organism or the mind presupposes two complementary processes. On the one hand, the organism adapts itself to the object which exercises this influence : in this way there is formed a sort of motor schema related to the new object. This is what we shall call, in a very wide sense, imitation. On the other hand, this adaptation implies that between the new movements and the old habits there is a certain continuity, i.e. that the new movements are partly incorporated into already existing schemas. This incorporation we shall call assimilation.

There is no need at this point to dwell upon the importance of imitation in mental development. J. M. Baldwin has done enough to emphasise this importance and to show that imitation transcends the limits usually set by the word. We shall go even further in this direction. Imitation can be by gesture and by movement, as when the child who plays at being its model who is learning to talk, to walk, etc. Drawing is imitation. But imitation can also be of thought, thought being a compressed form of action. In all these imitations there is a motor element, and this is why it is worth while reducing all these processes to imitation by gesture. And this, indeed, is what M. Delacroix has done, showing with his usual penetration that every perception " imitates " the perceived object. ^{[ 6 ]} As to assimilation, insufficient attention has been paid to the important part it plays. Assimilation must not be confused with analogy, with that perpetual tendency to reason by analogy which has been taken as the characteristic of elementary intellectual reactions. ^{[ 7 ]} For even if analogy is derived from assimilation, the latter is quite a different thing at first. There is analogy when two percepts or two concepts of the same order are reduced to one another. Thus when we see what we believe to be a tree, what we actually are perceiving is a green patch, an oblong shape, etc., and by immediate fusion we identify this perception with other earlier analogous perceptions and in this way are enabled to recognise a tree. In such a case there is analogy because the terms compared are on the same plane of reality, that is, they both are borrowed from external experience. But assimilation takes place when percepts, formless in themselves, and incapable of being completed by elements drawn from the same plane of reality, are worked into schemas taken from another plane of reality, into schemas, that is to say, which were there before the experience of the kind in question, and which are conditioned by the structure of the organism.

We have met with typical examples of assimilation in connection with the genesis of the idea of force or of animism in general. Inanimate objects are assimilated to living beings who act with effort and will. Now this assimilation is not the result of a mere judgment of analogy. The feeling of effort is bestowed upon objects in quite a different manner. It is introjected into the object before having been conceived as characterising the self, or at any rate at the same moment. For the internal experience is not more directly accessible than the external. On the contrary, it is only after having assimilated the activity of external bodies to his own muscular activity that the child turns this new-made instrument upon himself and, thanks to it, becomes conscious of his internal experience. Thus the child learns to know the force of external objects through his own, and his own through that of external objects. By which we mean that lie does not directly see force anywhere, but conceives it, thanks to the relation existing between a schema which is prior to knowledge and to the contents of this knowledge.

In short, the subjective adherences which we observe during the primitive stages of intellectual evolution cannot be due simply to judgments of analogy, because there are not in this case two separate terms, known separately, which are compared and then identified with one another.

There is fusion which takes place prior to any knowledge of the terms compared. This fusion is what we have called assimilation, and is something which cannot be understood unless one imagines the existence of schemas already formed by action and into which are merged the elements of knowledge in the process of formation.

We are fully aware of all that remains obscure in this notion of assimilation. But our results up to date do not permit of any further analysis of the subject. In order to grasp the mechanism of assimilation one would have to investigate the zone that lies between organic and intellectual life. It is therefore from an analysis of the first two years of the child's life that we may hope for light on the subject. All we can do for the moment is to postulate the existence of this process of assimilation. For the confusion made by the child between the self and the external world seems to us to constitute a $m generis relation. It is not a relation between two terms situated on the same plane, because one of the two terms is not an object of cognition but a factor in cognition, in that it imprints its structure upon the other term. One would therefore have to turn to the a priori synthetic relations which Kant has foisted upon the Theory of Knowledge in order to find an adequate comparison. But in Kant's theory, the a priori form is fixed and cannot be modified by experience, whereas the schema of assimilation is plastic. Experience changes it and fresh schemas are constantly emerging under the pressure of facts. This is what M. Brunschvicg has brought out so admirably in dealing with the history of the sciences. Moreover, as indeed M. Brunschvicg himself admits, form and matter are inseparable in episternology. Form alone is nothing : it can neither be defined nor give rise to a cognition. Only, M. Brunschvicg has intentionally severed the connection between his psychological reflections and biology, whereas we, on the contrary, feel constrained, as psychologists, to look for the continuity between cognition and life. Assimilation in this respect appears to us as the biological equivalent of judgment. It has been said, with truth, that the act of judgment cannot be reduced to any other. But it is none the less legitimate to seek for what may be its biological roots, provided, of course, it be remembered that judgment can only be reduced to assimilation in so far as it is already potentially contained in it.

Having said this, we can now return to the study of the paradox of child thought which necessitated the more detailed account we have given of these various ideas. Child thought is at once more realistic and more subjective than ours. But if we appeal to the concepts of assimilation and imitation, this dualism becomes comprehensible. For assimilation and imitation are at the root of two strictly antagonistic tendencies which arise, when the organism is confronted with something new (a circumstance which is all the more likely to occur when the child is very young). Assimilation consists in adapting the object to oneself by divesting it of all its irreducible characters. Imitation consists in adapting oneself to the object by abandoning attitudes which could obstruct this adaptation and by taking up an attitude that is entirely new.

Le Dantec's deplorable neglect of experimentation in biology is well known, but to him belongs the merit of having discovered certain very general and synthetic formulae such as are made in thermo-dynamics for lack of any possible analysis of detail. Now, Le Dantec has shown in a very striking manner this contrast between assimilation and imitation with regard to organic life. ^{[ 8 ]} The organism left to itself tends to assimilate its environment, it tends, that is, to persist exactly as it was before and to deform the environment so as to subject it to this assimilation. But the environment resists and influences the organism. According to the strength of this resistance the organism is forced to change, and each of these variations consists in a sense, in an imitation of the object which is exercising its constraining power. Biology, concludes Le Dantec, can be summed up as the struggle between assimilation and imitation.

In the psychic life of the child, which is closer to organic life than ours, this antagonism still shows clearly ; that is, if we agree to regard assimilation and imitation in the sense in which we have defined these terms, namely as special cases of the process described by Le Dantec. Assimilation and imitation work in opposite directions, so that each pulls the mind its own way. Any mental attitude during the primitive stages will therefore consist in a compromise between these two tendencies and not in their synthesis.

This situation will serve to explain the paradox which we referred to above. Sometimes the child makes an effort to imitate reality, and then he is servile in his acceptance of the outlines and curves of direct perception. In such cases, one has the impression of a purely empirical, a purely " phenomenistic " mind, which does no more than establish relations between any one thing and another, provided experience has allowed of their being brought together. This is thinking by juxtaposition. For in so far as imitation triumphs, assimilation is repressed. When, on the contrary, assimilation has the upper hand in the thought process, the child seems not to trouble in any way about objective observation, and rushes headlong into dynamism, animism, and participation.

Such a situation as this is not an accident in the history of thought, but a biological necessity. Lacking collaboration, the two tendencies, the imitative and the assimilative, lead to no coherent result. Because he fails to imitate correctly when he is assimilating, the child deforms reality in assimilating it to himself, and because he fails to assimilate when he is imitating, he becomes the victim of direct perception instead of constructing a world of intelligible relations.

But such an equilibrium as this is unstable, and assimilation and imitation soon begin to collaborate. It may even be questioned whether the definition of the whole of thought does not lie precisely in this collaboration.

As imitation and assimilation become complementary to one another, their characters change. Assimilation ceases to deform, i.e. to alter reality in terms of the self. The assimilating schemas become more and more flexible in yielding to the demands of external things and of experience. Now a non-deforming assimilation is synonymous with understanding or deduction, which means that it does no more than bend the data of experience to the exigencies of logical structures corresponding to the various stages of development. Under the influence of assimilation, imitation in its turn loses its servility and grows into intelligent adaptation to the external world. Deduction and experience then become the two opposite poles of one and the same effort of thought which synthesises the formerly antagonistic tendencies, assimilation and imitation.

From the point of view of the evolution of the ideas of reality, etc., the effect of the process we have been describing is quite clear, and it would not be difficult to reduce it to these various factors. It would be sufficient to remember that assimilation and imitation are not only reactions to the physical environment, but also to the social environment. From this point of view, the deforming assimilation of the primitive stages is synonymous with egocentricity, and imitation synonymous with social imitation. At first only a compromise is effected between these two tendencies ; but by collaborating progressively throughout the mental development of the child, imitation becomes adaptation to others, and assimilation turns into understanding and a sense of reciprocity. So that we can now see how the processes which transform primitive realism into objectivity, reciprocity, and relativity, are all based upon the progressive collaboration of assimilation and imitation. As to the progressive reversibility of thought, which brings with it the progressive reversibility of causal sequences, we have shown elsewhere in what manner it results from the same tendencies (J-R-, Chap. IV, 4).

Child Logic

It may be of interest now to see whether our present study confirms the results previously arrived at concerning child logic. In this way we shall be able to establish what are the relations between this logic and the structure of reality, of causality and of legality as these are conceived by the child.

Let us remind the reader once again that in questioning the children about the phenomena of nature we did not reach their spontaneous thought, but a thought that was necessarily systematised and consequently deformed by the very fact of the interrogatory. Further, and this is the fundamental point, the most original and the most important part of the answers which the children gave us had never been communicated to anyone before it was given to us. Children do not talk amongst themselves about their conceptions of nature, and in so far as they put questions to adults upon the subject, these tend to annul the purely childish character of their conceptions. And yet these conceptions are constant in the towns where we were able to question children, and they are to be found amongst nearly all children of the same mental age. Nothing is more striking in this respect than the very simple experiments which are completely removed from anything that the children can have been taught. Such is, for example, the experiment of the pebble placed in a glass of water, so as to make the level of the water rise : all the younger children say that the water rises because the pebble is heavy, and all the older ones say that it rises because the pebble is big. The convergence here is extremely interesting.

This secret and yet constant character of childish views about the world shows very clearly that before the interrogatory the spontaneous thought of the subject must have been made up more of images and motor schemas than of conceptual thought, such as could be formulated in words. We have here a general confirmation of the hypothesis we put forward earlier, and according to which child thought is not social but egocentric, and consequently intermediate between autistic and logical thought. In autistic thought, intellectual work is carried on by means of images and motor schemas. In logical thought, word and concept replace these primitive instruments. These two processes mingle in the child's mind, the first retaining its power in so far as thought is secret and unformulated, the second undergoing development in so far as thought becomes socialised.

This explains why the thought of the children we questioned was so lacking in logic. We were able, within each sphere, to establish special stages, but it would be extremely difficult to establish inclusive stages for the reason that during these early years the child is still very incoherent. At the age when the child is still animistic, artificialist, or dynamic in his way of thinking on some points, he has already ceased to be so on others. He does not reap the benefits of a progress in all the domains where this progress is bound eventually to make itself felt. Corresponding stages are at varying levels, because the influence of one belief upon another takes place unconsciously and not thanks to a conscious and deliberate generalisation. Thus child thought is in no way organised. There are, of course, certain remarkable correlations between one given achievement and another. (We may recall the correlation mentioned in Chap. VII, 3 of this volume.) But this is not the sign of discursive and reflective logic, it merely indicates the existence of a certain coherence between the warring parts of an organism which is unable as yet to release instantaneously such synergy as may exist. There is therefore not deduction, but juxtaposition, devoid of systematic logical multiplication and addition. The concepts of life, of weight, of force, of movement, etc., are not concepts properly so called, they are not defined by means of exact logical additions or multiplications, but they are those conglomerate concepts of which we have spoken elsewhere (J.R., Chap. IV, 2-3).

But leaving these more general considerations, let us see whether our present results tend to confirm the analysis which we formerly attempted to make of transduction, i.e. of the childish method of reasoning (J.R., Chaps. IV and V).

Childish transduction is opposed to adult deduction by the possession of three fundamental characteristics.

1. Transduction is, in the first place, purely a mental experiment, by which we mean that it begins by simply reproducing in imagination events such as they are or could be presented by immediate reality. For instance, having noticed that the presence of stones in a river produces tiny waves, the child explains the movement of the river by appealing to other stones which are supposed to have set it in motion.

2. Transduction is carried on by predicative judgments, or by certain simple judgments of relation. It might be better to say judgments of pre- relation, relations being conceived simply as properties : " the stone has force ", etc. For to do no more than combine the data supplied by immediate perception is to forget the part played in perception by the self or by the personal point of view : it is, therefore, to take a false absolute instead of objective relations as a foundation for reasoning. Thus when a child says that a boat floats because it is heavy, he does so because, in his mind, the weight of the boat has not been compared to its volume nor to the weight of the water, but has been evaluated as a function of the subject's own point of view, taken as absolute. In the same way all those instances of reasoning which bear upon the concepts of force, life, and movement, will be found to contain false absolutes, mere pre-relations, simply because the laws of physics have not been desubjectified.

3. Owing to the fact that it does not reason by relations but is a simple combination of judgments, transduction does not attain to the strict generality of deduction but remains an irrational passage from particular to particular. When the child seems to be deducing, that is, to be applying the universal to the particular, or to be drawing the universal from the particular, he does so in appearance only, owing to the indeterminate character of the concepts employed. Here is an example : a boy tells us that large-sized or " big " bodies are heavier than small ones ; yet a moment later he declares that a small pebble is heavier than a large cork. But he does not, for that matter, give up his first affirmation, he only declares that the stone is heavier than the cork " because some stones are bigger than corks ". Thus the character " big " has not at all the same meaning as for us. It does not define a class, it is transmitted by syncretistic communication to analogous objects : since there are big stones, little stones participate in their bigness, and thus acquire weight. At other times, the child reasons only for special cases and does not generalise at all : one boat floats because it is heavy, another because it is light, and so on. In short, either we have a juxtaposition of special case reasonings without generalisation, or we have apparent generalisation, but generalisation by syncretism and not by correct logical addition and multiplication.

Our interpretation of transduction is therefore that it moves from particular to particular, regardless of contradictions, because it is ignorant of the logic of relations, and that there is mutual dependence between this ignorance of the logic of relations and the fact that reasoning occurs simply by mental experimentation. This interpretation entails three debatable points which in our former studies we perhaps failed to analyse in sufficient detail, unacquainted as we were with much of the material which our present study has brought to light. Let us then submit these three points to somewhat closer examination.

The first runs as follows. It may be objected that transduction differs in no way from adult deduction, unless it be by insufficient elaboration of the material dealt with. In other words, between the child and ourselves there is no differ