Art - Linear And Aerial Perspective - Part 1
( Originally Published Early 1900's )
THE word perspective is familiar to us all. With its meaning we have had more or less experience which may, or may not, be cause for congratulation upon our arrival at a subject concerning which we have some knowledge. The subject is certainly not new. Anaxagoras and Democritus wrote geometrical treatises upon it centuries ago, and many not unworthy successors have done so since their time. But with the geometrical side of perspective I do not purpose to deal, for the reason that in actual painting it has not been usually considered by the painters of the past, and among those of the present it is not even generally understood. Those of you who may care to follow up the study of this geometrical side would better read Mr. Ruskin's treatise on the subject. What I may have to say about the subject today will be almost entirely from the artist's point of view.
Perspective is, perhaps, not so much an end of painting in itself as it is a means of obtaining certain effects. As a means it is its object to show upon flat surface the dimensions and intensities of objects at varying distances by just gradations of form, contour, color, and light. There are two kinds of perspective at least, and I am not sure but that there should be three classes of it, described respectively by the adjectives linear, aerial, and chromatic. But for our purposes the second adjective is expansive enough to include the third and ruder the latter unnecessary.
The proper use in painting of linear perspective produces a lessening in the size of objects by recession, and an apparent convergence of lines toward a given focus, called technically " the point of sight." It gives us upon a small scale the representation of an effect continually seen in nature. A glance down a long street reveals this effect to us every day of our lives. The rows of trees and the buildings, as the eye follows the top line of them, appear to run down from the upper sky to the rim of the horizon or the point of sight. The bed of the street, the curb-stones, the sidewalks appear to run up from our feet to this same point of sight. Again, it will be noticed that the walls of the buildings and the sides of the trees appear not only to run down, but to run in, until, if the street be very long, the rows of buildings, trees, curb-stones, and sidewalks come to meet in the distance—the lines of convergence appearing to run from every direction toward a horizon centre. Again, these lines of convergence compress and contract all the objects as they recede from us, and allow them to expand as they approach us. We often see this illustrated at the railway station when looking up or down the tracks fora belated train. As the train approaches the station we see the locomotive grow larger and larger; as the train moves away the rear car becomes smaller and smaller.
The principle by following which this effect is produced in art is not difficult of comprehension. Imagine the sun upon the horizon line shooting out shafts of light from it in all directions, the distance between the shafts widening of course with the radiation ; place a fete-simile of this sun—the point of sight—in the centre of a picture-frame so that the top, bottom, and sides of the frame shall cut off the ends of the flying shafts, and we shall then have a skeleton of perspective. In the street scene, of which I spoke a moment ago, if the sun were placed at the end of the street the lines of the curbstones and sidewalks would follow certain of these flying shafts of light directly toward the centre ; the lines of the tops of the houses, the trees, and the telegraph-poles would follow other shafts higher up, and, were the picture specially composed as a perspective effect, the breaks in the clouds would be arranged in their lines to conduct upon the upper shafts of light directly to the centre again. Upon whatever line or shaft the eye might fall it would inevitably be led to the point of sight, or, as we have supposed, the sun itself.
To be sure such bits of nature as the railway and the street scene do not arise continually as exemplars of nature's perspective principles, but there is, nevertheless, a law of diminution and convergence underlying every scene, whether it be a positive example or not. We hardly need to be told that there is no such thing in reality as two parallel lines running together, but to our eyes they appear to run together. Perspective is in itself one proof in many that painting represents not reality, as our " realists" would have it, but only the appearance of reality. It is merely a semblance of things resulting from the eye's inability to grasp distant objects as they actually exist. The compensation, however, for the inability of the eye to see things in the distance in their real relations is, that by seeing them in perspective we gain a breadth and depth of view not otherwise obtainable. We are enabled to .see not one object alone but many objects, all held together by a common bond of unity. This is true of the perspective in a painting. For the eye could not grasp, in either depth or breadth, the whole of the scene upon canvas were it not that by the lines or shafts of light, of which I have spoken, the vision is conducted down a converging path toward the point of sight, which is usually the point of interest as well. One object of linear perspective in painting, then, aside from its giving the appearance of distance, is to obtain unity—to enable us to grasp the whole scene at a glance—and without it a picture would be disjointed, and comprehensible only by examining one part of it at a time as we do the passing scenes of a panorama. This unity by perspective may be seen well exemplified if standing in a room we look out through a pane of glass, imagining the window-sash a picture-frame. Miles in depth and miles in width appear within the square of a few feet ; and if we imagine further that the landscape is painted on the glass instead of seen through it, we shall have the correct perspective of a picture.
This you will understand is the general principle of perspective that I am trying to explain, and the illustrations given show linear perspective in its simplest form only (Fig. 7). It has its many complexities, involving problems which are scarcely worth our time puzzling over ; but I think it necessary to say they exist, that you may not think every picture which lacks a diamond-point composition, converging lines of trees, buildings, or telegraph-poles, and a sun, or at least a tunnel of concentrated light precisely in the middle of it, is therefore bad or lacking in perspective. The point of sight, or the place where the sight should be drawn, is a matter of choice with the painter. It is generally near the centre of the piece, as in Leonardo's "Last Supper," for instance but there is no particular reason that it should be so. The painter, like the photographer, may focus his picture where he pleases ; but, also like the photographer, whatever point he focuses becomes a centre of interest from which the shafts or lines radiate. Some of the Dutchmen, like Van Goyen, Cuyp, and Van de Velde, were very fond of placing the point of interest off at the extreme side, and leading up to it by long rows of buildings, the descending masts of ships, or the retiring ranks of trees or hills. But the goal of interest in any one of their pictures is generally well defined and easy to discover, no mattter where placed, because of the convergence of line and light toward that spot. The fly's parlor of a spider's web is not always placed in the centre of a given space, nor in the centre of the web, but our eye naturally seeks it because all the lines of the web upon which the spider travels lead directly to it.
The point of sight may be shifted right or left provided we shift our station-point to correspond ; and there is no reason why it should not be raised up by a high horizon line, as the needs of the foreground and middle distance may require, or, on the other hand, placed law down, as the needs of the upper sky may require, the visual ray which comes to the eye being raised or lowered again to correspond. There is nothing very arbitrary about perspective except the point of sight, which should be the loadstone of the picture to attract the eye of the spectator. As I have observed, in speaking of light-gradation, this point of sight will not always be so apparent as in a Claude or Turner sunset, and in many pictures you will have some difficulty in finding it at all ; but if the perspective be good the lines, whether apparent or not, will converge, and the eye will be led to some one point in the picture —the point of interest, the point of sight, and generally the point of light. Two notable instances of the violation of this rule are Raphael's " Transfiguration ", and the large " Marriage in Cana," by Paolo Veronese, in each of which there are two points of sight, two horizon lines, and two base lines. But this, instead of being a virtue of the pictures is perhaps a fault, because of the confusion brought about by conflicting points of interest. In the presence of such a dual composition the eye is embarrassed by riches, and like the historic donkey of Buridan, it starves between two measures of oats.
The effect of linear perspective upon objects is that as they recede from us they appear to decrease in size in a geometrical ratio. To explain this, let us change the flaming sun, which I have been using for illustration, into so commonplace a thing as a bicycle wheel, and we shall have quite as good a perspective skeleton upon which to construct a picture. Let the hub be the point of sight ; let a line drawn directly across the middle, just below the hub, be the horizon ; let the lower spokes be the middle distance and foreground ; the lower rim the bottom of the frame or the station-point where the spectator stands ; the upper spokes the sky ; the upper rim the zenith, or the top of the frame ; and the side spokes and rims the wings of the picture and the sides of the frame. Now, if we suppose in the right-hand corner foreground some palaces in a row, as in Turner's Carthage pictures, or Claude's "Embarkation of the Queen of Sheba," the first one of which just fits in between two spokes of the wheel at their widest end, and that the row recedes in the distance, perfect linear perspective would require that the height of the palaces decrease in proportion to the narrowing of the converging lines or spokes. On the canvas the buildings might appear two feet in height in the foreground, one foot in the middle distance, and farther back only half a foot; for, as the lines or spokes narrow, the objects shrink correspondingly, not only in height but in width, until at last lost in distance or centred in the hub.
Of course the buildings would not be the only objects so affected. The clouds in the upper zenith would also be caught, and compressed as it were, between the spokes of the upper half, and receding down toward the hub would decrease in all their dimensions. The objects on the ground or on the water, whether bushes, trees, men, animals, boats, or ships, would undergo a similar process of size-degradation. The natural result of such a proportionate contraction of objects would be that only the larger objects would hold out in the distance, and that the smaller ones would be completely absorbed or blotted away. The grass, the small bushes, the stones, the human beings would vanish long before the trees, the ships, and the palaces ; and, as Leonardo has wisely remarked, by the very abandonment of the small things and the recognition of the large bodies only, would the distance be increased and the illusion of perspective made more complete.
It may be well to bear in mind, however, that in practice few painters ever cover their canvases with lines like the spokes of a bicycle wheel ; or, as I have said, know, or care to know, anything about the geometrical side of perspective. If the painter's perspective be true, it may be planned and scaled by lines, but he does not consider the geometrical theory of form-shrinkage to gain the practical truth of perspective. He simply draws nature as he sees it, trying as far as possible to get rid of the abstract literal knowledge which he possesses about objects, and striving to record only the impression received by his eyes. This is not an easy thing to do, for the memory of objects is continually influencing the eye. A man a mile away from us has as many lines and shades about him as at any other distance, but we do not see them. He counts to our eyes as nothing but a spot of color on the landscape, though we may think he possesses more distinctness. When he steps up into the middle distance he becomes more like a man, though he is still only a horse-post-looking affair with a hat at the top. When he comes into the foreground, however, not only the lines of the body but those of the face and its features, the hands, the clothing, all come out distinctly. The relative height or breadth of the man at the varying distances, instead of being geometrically ascertained by a skeleton of converging lines, is caught in a very primitive manner by holding out the handle of a paint-brush at arm's-length, getting the man in a line of sight, shutting one eye, and indicating the height or breadth in inches by a thumb-mark on the brush handle. And for practical purposes, perhaps, this is as good a way as any. It is not quite accurate, but accuracy is the bane of the fine arts, for no other reason than because it is accuracy. Preciseness and primness, exactness and conventionality are synonyms in the art vocabulary, and any one of them is likely to make a painting mechanical, impersonal, and unsympathetic.