The first crude attempts about the age of three or four to draw animal forms exhibit great incompleteness of conception and want of a sense of position and proportion. In one case the head seems to be drawn, but no body—if, indeed, head and body are not confused; and in others where a differentiation of head and trunk is attempted there is no clear local separation, or if this is attempted there is no clear indication of the mode of connexion (see, for example, Fig. 43 (a)). In the case of animals the side view is for obvious reasons hit on from the first. But, needless to say, there is no clear representation of the profile head. As a rule we have the front view, or at least the insertion of the two eyes. Both eyes appear in Mr. Cooke’s illustrations of drawings of the cat by children between three and four (Fig. 43 (b)), as also commonly in drawings of horses. The position of the eyes is often odd enough, these organs being in one drawing by a boy of five pushed up into the ears (Fig. 43 (c)).^[273] The front view of the animal head along with profile body appears occasionally in savage drawings also.^[274] In some of children’s drawings we see traces of a mixed scheme. Thus I have a drawing by a boy of five in which a front view is reached by a kind of doubling of the profile (Fig. 43 (d)).

More remarkable than all, perhaps, we have in one case a clear instance of the scheme of the human face, the features, eyes, nose, and mouth being arranged horizontally to suit the new circumstances (Fig. 44 (a)). With this may be compared the accompanying transference of the animal ear to the human figure, though this suggests—especially in view of the pipe—a bit of jocosity on the part of the young draughtsman (Fig. 44 (b)).

The forms of both head and trunk vary greatly. In a few drawings I have found the extreme of abstract treatment in the drawing of the trunk, viz., by means of a single line, a device which, so far as I have observed, is only resorted to in the case of the human figure for the neck and the limbs. An example of this was given above in Fig. 1 (p. 334). The following drawing of a dog by a little girl between five and six years old illustrates the same thing (Fig. 44 (c)).^[275] On the other hand we see sometimes a tendency to give the trunk abnormal thickness, as if the model used had been the wooden toy-horse, as in the accompanying drawing by a boy of five (Fig. 44 (d)). Rectilinear instead of rounded forms occur, and the head is often triangular, these rectilinear contours being probably suggested by the teacher in his model schemes (see Fig. 44 (e)).

The legs are of course all visible. The strangest inattention to number betrays itself here. As we saw, a child in beginning his scribble-drawing piles on lines for the legs (see above, p. 334, Fig. 1). A girl between three and four years of age endowed a cat with two legs and a bird with three (see Fig. 45 (a) and (b)).^[276] A boy in his sixth year drew a quadruped with ten legs (Fig. 45 (c)). They are often drawn absurdly out of position. In more than one case I find them crowded behind, as in the accompanying drawing of some quadruped by the same little girl that drew the cat and the bird, and in a drawing of a mouse by another child about the same age, viz., three and a half years (Fig. 45 (d) and (e)). They commonly remain apart from one another throughout their course, following roughly a parallel direction. But this simple scheme is soon modified, first of all by enlarging the space between the fore and the hind legs, and then by introducing some change of direction answering to the look of the animal in motion. This is most easily effected by making the fore and the hind pair diverge downwards, as in Fig. 43 (b) and (c) (p. 373). In rarer cases the divergence appears between the two legs of the fore and of the hind pair (Fig. 45 (f)). The knee-bend is early introduced as a means of suggesting motion. Either the legs are all bent backwards, as in Fig. 45 (g) (cf. above, Fig. 44 (e)); or, with what looks like a perverted feeling for symmetry, each pair is bent inwardly, as in Fig. 45 (h). The forms are often extraordinary enough, a preternatural thickness of leg being not infrequently given, and the knee-joint occasionally taking on grotesque shapes as if the little draughtsman had just been attending a class on the anatomy of the skeleton. The hoof is drawn in a still freer manner, various designs, as the bird-foot, the circle, and the looped pattern, appearing here as in the case of the human foot (Fig. 45 (i) and (j); cf. Figs. 43 (c) and 44 (a) (p. 373)).

In this unlearned attempt to draw animal forms the child falls far below the level of the untutored savage. The drawings of animals by the North American Indians, by Africans, and others, have been justly praised for their artistic excellence. A fine perception of form is, in many cases, at least, clearly recognisable, the due covering of one part by another is represented, and movement is vigorously suggested. Lover though he is of animals, the child, when compared with the uncivilised adult, shows himself to be woefully ignorant of his pets.

Men on Horseback, etc.

Childish drawing moves as the dialectic progress of the Hegelian thought from distinction and antithesis to a synthesis or unity which embraces the distinction. After illustrating the human biped in his contradistinction to the quadruped he proceeds to combine them in a higher artistic unity, the man on horseback. The special interest of this department of childish drawing lies in the fresh and genial manner of the combining. To draw a man and a horse apart is one thing, to fit the two figures one to the other, quite another.

At first the degree of connexion is slight. There is no suggestion of a composite or mixed animal, such as may have suggested to the lively Greek imagination the myth of the centaur. The human figure is pitched on to the quadruped in the most unceremonious fashion. Thus in many cases there is no attempt even to combine the profile aspects, the man appearing impudently in frontal aspect, or what would be so but for the lateral nasal excrescence, as in the accompanying drawing by a boy of five (Fig. 46).

With this indifference to a consistent profile there goes amazing slovenliness in attaching the man to the animal, and this whether the front or side view of the human figure is introduced. No attempt is made in many cases to show attachment: the man is drawn just above the quadruped, that is all. It seems to be a chance whether the two figures meet, whether the feet of the man rest circus-fashion on the animal’s back, or, lastly, whether the human form is drawn in part over the animal, and, if so, at what height it is to emerge from the animal’s back. Various arrangements occur in the same sheet of drawings (see Fig. 47 (a), (b) and (c)).

When this overlapping takes place the presence of the animal’s trunk makes no difference in the treatment of the man. He is drawn with his two legs just as if he were in relief against the horse; and this arrangement is apt to persist even when a child can draw a rude semblance of a horse and knows at what level to place the rider. So difficult to the little artist is this idea of one thing covering another that even when he comes to know that both the legs of the rider are not seen, he may get confused and erase both (see above (p. 376), Fig. 45 (f)).^[277]

The savage is in general as much above the child in the representation of the rider as he is in that of the animal apart. Yet traces of similar confusion do undoubtedly appear. Von den Steinen says that his Brazilians drew the rider with both legs showing. Andree gives an illustration, among the stone-carvings (petroglyphs) of savages, of the employment of a front view of the human figure rising above the horse with no legs showing below (Fig. 48 (a)).^[278] Even among the drawings of the North American Indians, in which the horse is in general so well outlined, we occasionally find what appear to be the germs of confusions similar to those of the child. Thus Schoolcraft gives among drawings from an inscription on a buffalo skin one in which we have above the profile view of a horse the front view of a man, with arms stretched out laterally while the legs are wanting.^[279] A clearer case of confusion is supplied by the following drawing, also by a North American Indian, in which the lines of the horse’s body cut those of the rider’s legs (Fig. 48 (b)).^[280]

The same tendency to show the whole man where the circumstances hide a part appears in children’s drawings of a man in a boat, a railway carriage and so forth. Ricci has shown that the different ways in which the child-artist puts a human figure in a boat are as numerous as those in which he sets it on a horse. The figure may stand out above the boat or overlap, in which last case it may be cut across by the deck-line and its lower part shown, or be clapped wholly below the deck, or again be half immersed in the water below the boat, or, lastly, where an attempt to respect fact is made, be truncated, the trunk appearing through the side of the boat, though the legs are wanting.^[281] A man set in a house, train, or tram car, is seen in his totality (Fig. 49 (a) and (b)). It is much the same thing when a child flattens out a house or other object so as to show us its three sides, that is to say one which in reality is hidden (Fig. 49 (c) and (d)). With these habits of the child may be compared those of the savage. The impulse to show everything, even what is covered, is illustrated in a drawing of a singer in his wigwam by an Indian (Fig. 49 (e)).^[282] Even where colour comes in and one thing has to be hidden by a part of another thing the savage artist, like the child insists on drawing the whole. This is illustrated in a curious custom, the drawing of two serpents (in dry, coloured powder) by North American fire-dancers. They are drawn across one another, and the artist has first to draw completely the one partly covered, and then the second over the first.^[283]

The child’s drawing of the house, though less remarkable than that of the man and the quadruped, has a certain interest. It illustrates, as we have just seen, not merely his determination to render visible what is hidden, but also his curious feeling for position and proportion. In one case I found that in the desire to display the contents of a house a girl of six had actually set a table between the chimneys. The accompanying drawing done by the boy C. at the age of five years five months illustrates the fine childish contempt for proportion (Fig. 50). A curious feature in these drawings of the house is the care bestowed on certain details, pre-eminently the window. This is even a more important characteristic feature than the chimney with its loops of smoke. Some children give a quite loving care to the window, drawing the lace curtains, the flowers, and so forth.

Résumé of Facts.

We may now sum up the main results of our study. We find in the drawings of untrained children from about the age of three to that of eight or ten a curious mode of dealing with the most familiar forms. At no stage of this child-art can we find what we should regard as elements of artistic value: yet it has its quaint and its suggestive side.

The first thing that strikes us here is that this child-delineation, crude and bizarre as it is, illustrates a process of development. Thus we have (a) the stage of vague formless scribble, (b) that of primitive design, typified by what I have called the lunar scheme of the human face, and (c) that of a more sophisticated treatment of the human figure, as well as of animal forms.

This process of art-evolution has striking analogies with that of organic evolution. It is clearly a movement from the vague or indefinite to the definite, a process of gradual specialisation. Not only so, we may note that it begins with the representation of those rounded or ovoid contours which seem to constitute the basal forms of animal organisms, and proceeds like organic evolution by a gradual differentiation of the ‘homogeneous’ structure through the addition of detailed parts or organs. These organs in their turn gradually assume their characteristic forms. It is, perhaps, worth observing here that some of the early drawings of animals are strongly suggestive of embryo forms (compare, e.g., Fig. 45 (b) and (d), p. 375).

If now we examine this early drawing on its representative side we find that it is crude and defective enough. It proceeds by giving a bare outline of the object, with at most one or two details thrown in. The form neither of the whole nor of the parts is correctly rendered. Thus in drawing the foot it is enough for the child to indicate the angle: the direction of the foot-line is comparatively immaterial. In this respect a child’s drawing differs from a truly artistic sketch or suggestive indication by a few characteristic lines, which is absolutely correct so far as it goes. The child is content with a schematic treatment, which involves an appreciable and even considerable departure from truthful representation. Thus the primitive lunar drawing of the human face is manifestly rather a diagrammatic scheme than an imitative representation of a concrete form.

In this non-imitative and merely indicative treatment there is room for all sorts of technical inaccuracies. Form is woefully misapprehended, as in the circular trunk, the oblong mouth, the claw foot, and so forth. Proportion—even in its simple aspect of equality—is treated with contempt in many instances (cf. the legs of the quadruped and the bird in Fig. 45 (a), (b), and (c) (p. 375)). What is no less important, division of space and relative position of parts, which seem vital even to a diagrammatic treatment, are apt to be overlooked, as in drawing the facial features high up, in attaching the arms to the head, and so forth. Even the element of number is made light of, and this, too, in such simple circumstances as when drawing the legs of an animal.

One of the most curious of these misrepresentations comes into view in the third or sophisticated stage, viz., the introduction of more than is visible. This error, again, assumes a milder and a graver form, viz., (a) the giving of the features more distinctly and completely than they appear in the object represented, and (b) the introducing of features which have no place in the object represented. Examples of the first are the introduction of the nasal angle into the front view of the human face; the separation throughout their length of the four legs of the horse; and such odd tricks as detaching the reins of the horse from the animal, as in Fig. 51 (a). Illustrations of the second are numerous and varied. They include first of all the naïve introduction of features of an object which are not on the spectator’s side and so in view, as the second eye and the second arm in what are predominantly profile representations. With these may be classed the attempt to exhibit three sides of a house. Closely related to these errors of perspective is the exposure of objects or parts of objects which are covered by others. It is possible that the spread-eagle arrangement of the two joined arms is an attempt to represent a feature of childish anatomy, viz., the idea that the arms run through and join in the middle of the trunk. A clearer example of this attempt to expose to view what is covered is the exhibition of the whole human figure in a boat, house or carriage. With this may be compared the disclosure of the whole head of a horse when drinking, as in the accompanying drawing by a boy of five (Fig. 51 (b)), of the whole head of the man through his hat (see above, p. 350, Fig. 20 (b)), and of the human limbs through the clothes (Fig. 41, p. 371).

A class of confusions, having a certain similarity to some of these, consists in the transference of the features of one object to a second, as when a man or quadruped is given a bird-like foot (Figs. 7 (d) and 43 (c), pp. 342, 373), and still more manifestly when the facial scheme of the man is transferred to the quadruped or vice versâ (Fig. 44 (a) and (b), pp. 373, 374).

These last errors clearly illustrate the tendency to a conventional treatment, a tendency which, as I have observed already, runs through children’s spontaneous drawings. This free conventional handling of natural forms has been illustrated in the habitual drawing of the mouth and eyes, and still more strikingly in that of the hands and feet.

Paradoxical though it may seem, these drawings, while in general bare and negligent of details, show in certain directions a quite amusing attention to them. Thus, we find at a very early stage certain details, as the pipe of the man, insisted on with extravagant emphasis; and may observe at a somewhat later stage in the elaborate drawing of hair, buttons, parasol, and so forth, a tendency to give some feature to which the child attaches value a special prominence and degree of completeness.

The art of children is a thing by itself, and must not straight away be classed with the rude art of the untrained adult. As adult, the latter has knowledge and technical resources above those of the little child; and these points of superiority show themselves, for example, in the fine delineation of animal forms by Africans and others.^[284] At the same time, after allowing for these differences, it is, I think, incontestable that a number of characteristic traits in children’s drawings are reflected in those of untutored savages.

Explanation of Facts.

Let us now see how we are to explain these characteristics. In order to do so we must try to understand what process a child’s mind goes through when he draws something, and to compare this with what passes in the mind of an adult artist. The problem has, it is evident, to do with drawing from memory or out of one’s head, for though the child may begin to draw by help of models, he develops his characteristic art in complete independence of these.

In order to draw an object from memory two things are obviously necessary. We must have at the outset an idea of the form we wish to represent, and this visual image of the form must somehow translate itself into a series of manual movements corresponding to its several parts. In other words, it presupposes both an initial conception and a correlated process of execution.

In psychological language this correlation or co-ordination between the idea of a form and the carrying out of the necessary movements of the hand is expressed by saying that the visual image, say, of the curve of the full face, calls up the associated image of the manual movement. This last, again, may mean either the visual image of the hand executing the required movement, or the image of the muscular sensations experienced when the arm is moved in the required way, or possibly both of these.

The process of drawing a whole form is of course more complex than this, each step in the operation being adjusted to preceding steps. How far the movements of the draughtsman’s hands are guided here by a visual image of the form, which remains present throughout, how far by attention to what has already been set down, may not be quite certain. Judging from my own case, I should describe the process somewhat after this fashion. In drawing a human face we set out with a visual image of the whole, which is incomplete in respect of details, but represents roughly size and general form or outline. This image is projected indistinctly and unsteadily, of course, on the sheet of paper before us, and this projected image controls the whole operation. But as we advance we pay more and more attention to the visual presentation supplied by the portion of the drawing already produced, and only realise with any distinctness that part of the projected visual image which is just in advance of the pencil.

It is evident that the carrying out of such a prolonged operation involves a perfected mechanism of eye, brain and hand connexions; for much of the manual adjustment is instantaneous and sub-conscious. At the same time the process illustrates a very high measure of volitional control or concentration. Unless we keep the original design fixed before us, and attend at each stage to the relations of the executed to the unexecuted part, we are certain to go wrong.

Practice tends, of course, to reduce the conscious element in the process. In the case of a person accustomed to draw the outline of a human head, a cat or what not, the operation is very much one of hand-memory into which visual representations enter only faintly. The movements follow one another of themselves without the intervention of distinct visual images (whether that of the linear form or of the moving hand). There is an approach here to what happens when we put last year’s date to a letter, the hand following out an old habit.

Now the child has to acquire the co-ordinations here briefly described. He may have the visual image of the human face or the horse which he wishes to depict. This power of visualising shows itself in other ways and can be independently tested, as by asking a child to describe the object verbally. But he has as yet no inkling of how to reproduce his image. That his inability at the outset is due to a want of co-ordination is seen in the fact that at this stage he cannot draw even when a model is before his eyes.

The process of learning here is very like what takes place when a child learns to speak. The required movements have somehow to be performed and attached to the effects they are then found to produce. Just as a child first produces sounds, partly instinctively or spontaneously, partly by imitating the seen movements of another’s lips, etc., so he produces lines by play-like scribble and by imitating the visible movements of another person’s hand. The tendency to imitate is observable in the first loop-formations, and possibly also in the abrupt angular changes which give a zig-zag look to some of these early tracings.

In this early stage we see a marked want of control. The effort is spasmodic and short-lived: the little draughtsman presently runs off into nonsense scribble. The want of control is seen, too, in the tendency to prolong lines unduly, and to repeat or multiply them, the primitive play-movements being very much under the empire of inertia or habit, i.e., the tendency to repeat or go on with an action. The effect of limitating natural conditions in the motor apparatus is illustrated, not only in the slightly curved form of these first scribble lines, but in the general obliquity or inclination of the line; it being manifestly easier for the hand when brought in front of the body to describe a line running slightly upwards from left to right (or in the reverse direction) than one running horizontally. The want of control by means of a steady visual image is further seen in the absence of any attempt at a plan, at a mapping out of the available space, and at an observation of proportion.

It might be thought that, though a child at this inexperienced stage were unable to produce the correct form of a familiar object, he would at once detect the incorrectness of the one he sets down. No doubt, if he were in the attitude of cold critical observation, he would do so: in fact, as Mr. Cooke and others have shown, he sees the absurdities of his workmanship as soon as they are pointed out to him. But when drawing he is in another sort of mood, akin to that imaginative mood in which he traces forms in the plaster of the ceiling, or in the letters of his spelling-book. He means to draw a man or a horse, and consequently the formless jumble of lines becomes, to his fancy, a man or a horse. His first drawings are thus, in a sense, playthings, which, like the battered stump of a doll, his imaginative intention corrects, supplements, and perfects.

With repetition, and that amount of supervision and guidance which most children who take a pencil in hand manage to get from somebody, he begins to note the actual character of his line-effects, and to associate these with the movements which produce them. A straight horizontal line, a curved line returning upon itself, and so forth, come to be differentiated, and to be co-ordinated with their respective manual movements.

We may now pass to the second stage, the beginning of true linear representation, as illustrated in the first abstract schematic treatment of the human face and figure.

A question arises at the very outset here as to whether, and if so to what extent, children re-discover this method of representation for themselves. Here, as in the case of child-language, such as ‘bow-wow,’ ‘gee-gee,’ tradition and example undoubtedly play their part. A parent, or an older brother and sister, in setting the first models, is pretty certain to adopt a simple scheme, as that of the lunar face; and even where there is no instruction a child is quick at imitating other children’s manner of drawing. Yet this does not affect the contention that such manner of drawing is eminently childish, that is, the one a child finds his way to most readily, any more than the fact of the nurse’s calling the horse ‘gee-gee’ in talking to baby affects the contention that ‘gee-gee’ is eminently a baby-name.

The scanty abstract treatment, the circle enclosing two dots and the vertical and horizontal lines, points to the absence of any serious attempt to imitate a form closely and fully. It seems absurd to suppose that a child of three or four does not image a human face better than he delineates it; and even if this were doubtful it is certain that when he sets down a man without hair, ears, trunk, or arm, his execution is falling far short of his knowledge. How is this to be accounted for? My explanation is that the little artist is still much more of a symbolist than a naturalist, that he does not in the least care about a full and close likeness, but wants only a barely sufficient indication. This scantiness of treatment issuing from want of the more serious artistic intention is of course supported by technical limitations. The lunar face with the two propping lines answers to what the child can do easily and comfortably. Much more than his elder brethren our small limner is bound by the law of artistic economy, the need of producing his effects with the smallest expenditure of labour, and of making every touch tell.

Defects of executive resource and of manual skill appear plainly in other characteristics. The common inclination of the lines of the legs points to the unconscious selection of easiest directions of manual movement.^[285] The unduly lengthened arm and leg, the multiplication of legs—as seen most strikingly in the case of the quadruped—illustrate the influence of motor or muscular inertia. There is, too, a noticeable want of measurement and management of the space to be covered, as when one eye is put in so large as to leave no room for a second, or when filling in details from above downwards the eyes are put in too near the occipital curve, and so all the features set too high up. The same want of measurement of space may contribute to the child’s habit of drawing the trunk so absurdly small in proportion to the head; for he begins with the head, and by making this large finds he has not left, within the limits of what he considers the right size of figure, space enough for the trunk.

Very noticeable is the influence of habit in this abstract treatment. By habit I here mean hand-memory, or the tendency to combine movements in the old ways, though this is commonly aided, as we shall see, by “association of ideas”. Thus a child falls into a stereotyped way of drawing the human face and figure; line follows line in the accustomed sequence; the only variation showing itself is in the insertion or omission of nose, ears, or arms; these uncertainties being due to fluctuations of energy and concentration. A child’s art is, in respect of its unyielding sameness, a striking example of a conservative conventionality. He gets used to his pencil-forms, and pronounces them right, to the greater and greater neglect of their relation to natural forms. Habit shows itself in other ways too. Notice, for example, how a child, after adding the trunk, will go on inserting the arms into the head as he used to do. Such a habit is an affair not only of the hand but of the eye. The arms have by repeated delineation come in the art-sphere to belong to the head.

Coming now to the more elaborate and sophisticated stage of five or thereabouts, in which the shape of eyes, mouth, and nose is shadowed forth, the difficult appendages as hands and feet attempted, and the profile aspect introduced, we notice first of all a step in the direction of naturalism. The child like the race gets tired of his bald primitive symbolism, and essays to bring more of concrete fulness and life into his forms. Only this first attempt does not lead to a continued progress, but stops short at what is rude and arbitrary enough, substituting merely a second rigid conventionalism for the first.

This transition indicates an advance in technical skill; hence we find a measure of free and bold invention, as in the management of the facial features, e.g., the scissors-shaped nose, and still more in the treatment of hands and feet, which is at once exaggerative, as in the big burr forms, and freely conventional, as in the leaf-pattern for the hand, and the wondrous loop-designs for the foot.

Yet though this freer treatment shows a certain technical advance it illustrates the effect of the limitations of the child’s executive power. Thus the new partially profile figures are very apt to lean, looking as if they were falling backwards. It is probable that the wide-spread tendency to make the profile face look towards the spectator’s left rather than his right is due to the circumstance that the eye can much better follow and control the pencil in this case than in the opposite one. In the latter the hand is apt to interfere with seeing the line of the face, especially if the pencil is held near its point.

Habit, too, continues to assert its dominion. The tendency noticeable now and again, even among English children, to treat the feet after the manner of the hands illustrates this. Habit is further illustrated in the tendency to a transference of forms appropriate to the man to the animal; or, when (owing to the interposition of the instructor) the drawing of animals is in advance of the other, in the reverse process; as when a cat is drawn with two legs, or a horse is given a man’s face, or the human form develops a horse’s ears, or a bird’s feet. With these may be compared the transference of a bird-like body and tail to a quadruped in Fig. 45 (i), p. 377. The accompanying two drawings by a child of six show how similar forms are apt to be used for the man and for the animal (Fig. 52).

But the really noticeable thing in this later sophisticated treatment is the bringing into view of what in the original is invisible, as the front view of the eye as well as both eyes into what otherwise looks a side view of the face, the two legs of the rider and so forth. Here, no doubt, we may still trace the influence of technical limitations and of habit. The influence of the former is seen in the completing of the contour of the head before or after drawing the hat: for the child would not know how to start with the lines which form the commencement of the visible part of the head. The influence of habit is also recognisable here. A child having learned first of all to draw the front view of the eye, the two eyes and the two legs side by side, tends partly as the result of organised hand-trick, partly in consequence of ‘association of ideas,’ to go on drawing in the same fashion in the new circumstances. A specially clear illustration of this effect of habit already alluded to is the introduction of the front view of the nose in the mixed scheme. These cases are exactly paralleled by the Egyptian drawing in which while one shoulder is pulled round the other is left in square front view (see above, p. 369, Fig. 39 (b)). Still, habit does not account for everything here. It does not, for example, explain why the child brings into view three sides of a house. The technical deficiencies of the small draughtsman, his want of serious artistic purpose, seem an insufficient explanation of these later sophistries. They appear to point plainly to certain peculiarities of the process of childish conception. We are compelled then to inquire a little more closely into the characteristics of children’s observation and of their mental representation of objects.

We are apt to think that children when they look at things at all scrutinise them closely, and afterwards imagine clearly what they have observed. But this assumption is hardly justified. No doubt they often surprise us by their attention to small unimportant details of objects, especially when these are new and odd-looking. But it is a long way from this to a careful methodic investigation of objects. Children’s observation is for the most part capriciously selective and one-sided. They apprehend one or two striking or especially interesting features and are blind to the rest. This is fully established in the case of ordinary children by the wondrous ignorance they display when questioned about common objects. It is hardly necessary to add that their spontaneous untrained observation is quite unequal to that careful analytical attention to form-elements in their relations which underlies all clear grasp of the direction of linear elements, the relative position of the several parts of a figure, and proportion.

This being so it maybe said that defects of observation are reflected in children’s drawing through all its phases. Thus the primitive bare schematism of the human face answers to an incomplete observation and consequently incomplete mode of imagination, just as it answers to a want of artistic purpose and to technical incapacity. How far defective observation assists at this first stage I do not feel sure. Further experimental inquiries are needed on this point. I lean to the view already expressed, that at this stage manual reproduction is far behind visual imagination.

When, however, we come on to the delineation of an object under its different aspects the defects of mental representation assume a much graver character. We must bear in mind that a child soon gets beyond the stage of recalling and imagining the particular look of an object, say the front view of his mother’s face, or of his house. He begins as soon as he understands and imitates others’ language to synthesise such pictorial images of particular visual presentations or appearances into the wholes which we call ideas of things. A child of four or five thinking of his father or his house probably recalls in a confused way disparate and incompatible visual aspects, the front view as on the whole the most impressive being predominant, though striking elements of the side view may rise into consciousness also. With this process of synthesising aspects into the concrete whole we call a thing there goes the further process of binding together representations of this and that thing into generic or typical ideas answering to man, horse, house, in general. A child of five or six, so far from being immersed in individual presentations and concrete objects, as is often supposed, has carried out a respectable measure of generalisation, and this largely by the help of language. Thus a ‘man’ reduced to visual terms has come to mean for him (according to his well-known verbal formula) something with a head, two eyes, etc., etc., which he does not need to represent in a mental picture because the verbal formula serves to connect the features in his memory.

Hence when he comes to draw he has not the artist’s clear mental vision of the actual look of things to guide him. He is led not by a lively and clear sensuous imagination, but by a mass of generalised knowledge embodied in words, viz., the logical form of a definition or description. This, I take it, is the main reason why with such supreme insouciance he throws into one design features of the full face and of the profile; for in setting down his linear scheme he is aiming not at drawing a picture, an imitative representation of something we could see, but rather at enumerating, in the new expressive medium which his pencil supplies, what he knows about the particular thing. Since he is thus bent on a linear description of what he knows he is not in the least troubled about the laws of visual appearance, but setting perspective at naught compels the spectator to see the other side, to look through one object at another, and so forth.

Since the process at this sophisticated stage is controlled by knowledge of things as wholes and not by representations of concrete appearances or views, we can understand why the visible result does not shock the draughtsman. The little descriptor does not need to compare the look of his drawing with that of the real object: it is right as a description anyhow. How strongly this idea of description controls his views of pictures has already been pointed out. Just as he objects to a correct profile drawing as an inadequate description, so he objects to a drawing of the hind part of a horse entering the stable, and asks, ‘Where is his head?’ We may say then that what a lively fancy did in the earlier play-stages childish logic does now, it blinds the artist to the actual look of what his pencil has created.

Use soon adds its magic force, and the impossible combination, the two eyes stuck on at the side of the profile nose, the two legs of the rider untroubled by the capacious trunk of the animal which he strides, the man wholly exposed to view inside the boat or carriage, gets stereotyped into the right mode of linear description.

All this shows that the child’s eye at a surprisingly early period loses its primal ‘innocence,’ grows ‘sophisticated’ in the sense that instead of seeing what is really presented it sees, or pretends to see, what knowledge and logic tell it is there. In other words his sense-perceptions have for artistic purposes become corrupted by a too large admixture of intelligence. This corruption is closely analogous to what we all experience when we lose the primal simplicity of the eye for colour, and impart into our ‘visual impressions,’ as we call them, elements of memory and inference, saying, for example, that a distant mountain side is ‘green’ just because we can make out that it is grass-covered and know that grass when looked at nearer is of a green colour.

I have dwelt on what from our grown-up standpoint we must call the defects of children’s drawing. Yet in bringing this study to a close it is only just to remark that there are other and better qualities well deserving of recognition. Crude, defective, self-contradictory even, as these early designs undoubtedly are, they are not wholly destitute of artistic qualities. The abstract treatment itself, in spite of its inadequacy, is after all in the direction of a true art, which in its essential nature is selective and suggestive rather than literally reproductive. We may discern, too, even in these rude schemes a nascent sense of values, of a selection of what is characteristic. Even the primitive trunkless form seems to illustrate this, for though, as we have seen in a previous essay, the trunk plays an important part in the development of the idea of self, it is for pictorial purposes less interesting and valuable than the head. However this be, it is clear that we see this impulse of selection at work later on in the addition of the buttons, the pipe, the stick, the parasol and so forth.

It is to be noted, too, that even in these untutored performances, where convention and tradition exercise so great a sway, there are faint indications of a freer individual initiative. Witness, for example, the varying modes of representing hair, hands, and feet. We may say then that even rough children in elementary schools who are never likely to develop artistic talent display a rudiment of art-feeling. It is only fair to them to testify that in spite of the limitations of their stiff wooden treatment they express a certain individuality of feeling and aim, that like true artists they convey a personal impression. These traits appear most plainly in the later representations of action, but they are not altogether absent from the earlier statuesque figures. Compare, for example, the look of alert vigour in Fig. 5 (a) (p. 339), of grinning impudence in Fig. 6 (a) (p. 341), of provoking ‘cheekiness’ in Fig. 20 (b) (p. 350), of a seedy ‘swagger’ in Fig. 32 (p. 362), of inebriate gaiety in Fig. 17 (p. 348), of absurd skittishness in Fig. 24 (b) (p. 354), of insane flurry in Fig. 26 (a) (p. 355), of Irish easy-goingness even when somebody has to be killed in Fig. 34 (p. 363), of wiry resoluteness in Fig. 29 (a) (p. 359), of sly villainy in Fig. 38 (p. 365), and of demure simplicity in Fig. 26 (c) (p. 356); and note the delicious variety of equine character in Fig. 45 (f) (p. 376) and following.

If a finer æsthetic feeling is developed the first rude descriptive drawing loses its attractions. A friend, a well-known psychologist, has observed in the case of his children that when they try to draw something pretty, e.g., a beautiful lady, they abandon their customary mode of description and become aware of the look of their designs and criticise them as bad. This seems to me a most significant observation. It is the feeling for what is beautiful which makes a child attend closely to the bare look of things, and the beginning of a finer observation of forms commonly takes its rise in this nascent sense of beauty. Indeed, may one not say that only when a germ of the æsthetic feeling for beauty arises, and a child falls in love with the mere look of certain things, can there appear the beginnings of genuinely artistic work, of a conscientious endeavour to render on paper the aspect which pleases the eye?