Perspective in Perspective

Lawrence Wright

• Publisher: Routledge
• Year: 2017
• ISBN:
  9781315412993
• Edition: 1
• Topic: Psicologia
• Subtopic: Storia e teoria della psicologia

Originally published in 1983, this book is about the way we see things – or think we do, which is by no means the same – and about the ways in which we have tried to reproduce that visual concept in diagrams, pictures, photographs, films and television. Whatever the medium, if any degree of realism is intended, some use of perspective is inevitable, and some understanding of it can aid the appreciation of the result. But here the technicalities of perspective geometry are treated as far as possible non-technically, by a common-sense approach. Students, would-be artists or architects, are warned in the Preface that they will travel second-class in the author's train of thought (the 'general reader' coming first), but they may well find the journey worthwhile in that it provides a background to a subsequent, more detailed studies.

Lawrence Wright shows that every form of perspective representation has some innate falsity, but that most such forms offer an adequate makeshift; that rules of geometry often need to be bent; that labour-saving dodges and shortcuts exist. As he says, perspective drawing, like politics, is an art of the possible. In reading this book, beginners may find it all simpler than they had supposed, though the established expert may in some interesting respects find just the opposite. The general reader may thereafter find himself seeing things – and representations of them – in a new light.

The object, the eye and the picture 1

We are often exhorted, especially by politicians on the defensive, to ‘get things in perspective’ but, when you come to think of it, this is an odd metaphor to choose for seeing things in their true relative proportions. If that is the aim, would it not be better to see things in plan, elevation and section, drawn to scale? In that form, shapes and sizes are readily comparable by eye and precisely measurable, whereas in perspective no solid object or two-dimensional surface can be represented in its right shape, unless it is seen head-on, in which case it can hardly be said to be in perspective. As for right size, this is the thing that perspective especially falsifies. Another version of the cliché would have us ‘stand back and get things in perspective’, but in practice, the more we stand back, the less do parallels seem to converge, and the less do similar objects at varying distances seem to vary in size; by standing back we tend to get the perspective out of things.

What, then, is so especially ‘true’ about a picture in perspective? A possible answer is that, unlike any other form of two-dimensional representation, it shows things ‘as we see them’. But even if that were so, is the way we see things true, and reliable in interpreting them?

Some of us may recall moments of startled wonder, in infancy, at our first experience of natural phenomena that we soon came to take for granted, but who can recall feeling the slightest surprise at the curious way that solid objects expanded as our pram approached them, then shrank as they receded behind? Or that, when we toddled across the kitchen, the floor tiles under our feet were square, but all the others were changing, twisted diamond shapes?

The apparent diminution of objects with increasing distance was noted at least as long ago as the seventh century BC when—according to a story on tablets from Assyria—one Etana, being carried up by an eagle to the heavenly throne of the goddess Ishtar, was so clear-headed in emergency as to observe that the earth became smaller and smaller until it reached a vanishing point.

This diminution and the apparent convergence of parallel lines are of course only different aspects of a single optical effect. Artists must long since have been at least vaguely aware of it, but for thousands of years none seems to have tried to reproduce it. A real man did not change his size with distance; a table was the same height at both ends whatever the angle of view; to depict things otherwise would be false. Although a man might indeed be drawn less than life-size, if only for want of a large drawing surface, the resulting figure still looked like a man, whatever one’s distance from the picture. One man might be drawn smaller than another in one and the same picture, but this had nothing to do with recession or spatial depth. There is a suggestion of relief in some prehistoric animal paintings, but it arises only from light and shade. In a particular age and society there may be a local convention requiring pictured objects to vary in size according to their supposed distance, but there is no universal law to that effect, nor that a picture must be read instantaneously as a whole: may not the eye, after a brief general assessment, rove at will—or under the subtle guidance of the lines of composition—from this item to that? Is this not, indeed, just what the eye normally does? Nor on the other hand does anything but local custom forbid an artist from simulating the natural effects of distance if he so wishes, though in so far as this is an innovation he may not expect to be understood. When and if some early genius was first to achieve a perspective effect, he was doubtless howled down—at least as loudly as artists were to be howled down by our grandfathers for rejecting perspective as an unnecessary nuisance.

On many occasions there is much to be said for this rejection. Perspective was to prove a good servant but a bad master. By its means, any competent draughtsman could produce a seemingly accurate reproduction of the visual effects of a given moment and a particular viewpoint, demanding no more insight or creative imagination than does a technical photograph. From time to time, and especially in the late nineteenth and early twentieth centuries, countless mere illustrators came to be rated as artists in direct proportion to the ‘lifelikeness’ of their work. ‘We do not always realize’, says Herbert Read, ‘that the theory of perspective developed in the fifteenth century is a scientific convention; it is merely one way of describing space and has no absolute validity.’ Drawing is essentially a two-dimensional art, and the imitation of three-dimensional effects savours of trickery: the early Italian practitioners called their perspective exercises dimostrationi, showpieces, as if only proofs of skill in geometry. There is something incongruous in making an illusory hole in the picture surface, or having an object seem to protrude from it. Fine drawing can be fine without attempting relief. A picture in perspective has, as such, no aesthetic superiority over one that is not; though if an artist does elect to use realistic convergences and foreshortenings, we are entitled to expect him to use them with some consistency, and to object at least to any obvious blunders. Van Gogh’s drawings and paintings of his bedroom at Aries (1.1) were not based on geometrical construction, and would be none the better if they had been; there are subtleties of observation in them that a professional perspector would have missed. But Holman Hunt’s etching, A Day in the Country (1.2), is clearly intended as realistic perspective, and it does not need a professional eye to spot its glaring faults. Guardi’s painting of Venice (1.3) is a superb rendering of colour and atmosphere, and its forms are perfectly legible; of those who notice its wildly contradictory convergences, many would think these of little consequence, but some might find them disturbing, and a reproduction of the picture hard to live with. Had it been drawn a little more truly—without demanding the painstaking accuracy of a Canaletto—would any aesthetic value have been lost? Turner would have painted this subject much more freely, but his underlying perspective, though less obtrusive, would have been more plausible. One may distinguish between purposeful adjustment and unintentional error.

1.1 Van Gogh: The artist’s room in Aries. 1888.

1.2 W. Holman Hunt: A day in the country. Etching, 1865.
For the figures and the nearby house, the artist’s eye level is just above the ground-floor window cill, but for the house at top left it is at least 30 feet higher. Even if the nearer woman, taken as a standard of comparison, is 6 feet tall, the omnibus from the Bank is being pulled by Shetland ponies.

The original meaning of the Latin perspectiva had nothing to do with what we call perspective, either as a visual effect or as a means of representing it. The root is perspicere, to see clearly, or to look through. Ars perspectiva was the ancient Greek optikē, the science of sight, or optics; quite distinct from the arts of geometrical drawing and painting. It is important to distinguish between ‘natural’ and ‘linear’ perspective. Natural perspective refers to the way we see things, and the geometry of sight is invariable. Linear perspective refers to the ways we use to reconstruct this view in pictures, and depends on variable and more or less arbitrary methods of geometry. A perspective can, of course, be drawn by direct observation, ‘by eye’, without use or even knowledge of geometry. Even then, the artist is not drawing what he sees; he is drawing from memory selected parts of what he saw, or supposed he saw, before he turned his attention from the object to the drawing. Neither an artist nor a geometrician—not even a photographer—has ever produced a perspective wholly ‘true’ in that it corresponded with the visual image received at the actual scene. This is not a mere philosophical quibble. The discrepancies are substantial and they have practical implications. When God created the world there were no treatises of linear perspective to guide him, and he has arranged its optical effects and our reception of them according to a different system. As William Blake put it,

This life’s five windows of the soul

Distorts the Heavens from pole to pole,

And leads you to believe a lie

When you see with, not thro’, the eye.

1.3 Francesco Guardi: Venice, S. Pietro di Castello, 1770–80.

The first known treatise on optics was by Euclid (c. 1300 BC). He thought that visual rays were emitted by the eye, an idea that persisted for centuries, whence perhaps the expressions ‘cast an eye’ or jeter un regard; but this reversal of the facts did lead him to correct conclusions. He made diagrams much like those of a modern textbook to show why receding parallel lines seem to converge; that the further parts of a ground plane look higher; he explained how foreshortening makes a chariot wheel seem oval. All perspective theories are based on Euclid. But he did not note that converging parallels, if produced far enough, appear to meet at a vanishing point. He looked for a simple law governing the proportion between the real and apparent sizes of an object at varying distances, but failed to find it, because he was thinking only of angles: to double the distance of an object (in linear units) is not to halve the angle it subtends to the eye (in degrees). In due course, Leonardo da Vinci was to find the reason why ‘a second object, as far away from the first as the first is from the eye, will appear half the size of the first, though they be the same size really.’ The basic difference in their observations was that Euclid was considering only the ‘visual cone’ involved in natural perspective, whereas Leonardo had in mind linear perspective and the projection of rays through a picture plane.

Lucretius (c. 98–55 BC) noted how a colonnade appears to vanish to the apex of a cone. He subscribed to the Epicurean theory of optics, that objects continually throw off eidola, minutely thin surface films which traverse the air and on meeting the eye produce vision; an odd concept leading to the even odder conclusion that the heavenly bodies are no bigger than they appear to be—though not to the equally valid conclusion that a distant man is smaller than a near one. Ptolemy, in a treatise on optics probably written between AD 127 and 141, showed how to trace on a mirror the outlines of a reflected object, and gave an explanatory diagram which might have led him to the projection method of perspective drawing. The science was thereafter handed on—at a most leisurely rate of development—through Galen, Alhazen and Vitellio, to the Italian architect-painters of the early Renaissance.

Leonardo compared the working of the human eye to that of the camera obscura, but this resemblance can be misleading. The retina of the eye, on which the image falls, is concave, forming more than half of the interior of a sphere (1.4). No geometry exists whereby this image could be reproduced correctly on paper. The problem is as insoluble as that of making a flat undistorted map of the spherical surface of the earth, or of flattening a deflated toy balloon. The ‘retina’ of the camera obscura is a plane surface (a white wall, a fabric screen or a sheet of paper) as that of a modern camera is a flat plate or film, and from light rays similar to those entering the eye it produces very different results. Although the image of the object in the eye changes very slightly in size and shape as eye movements bring it to various parts of the retina, it changes far less than the camera image does when the camera is rotated a few degrees, shifting the image from the centre to the edge of the field (1.5). If photographic plates were made concave, like the retina, the design of lenses would be simplified, but no other advantage would result. A flat enlargement from the concave negative, however complex the optical means used, would re-establish the distortion towards the edges. A concave positive print, large enough to be viewed from its centre, might bear some slight resemblance to a retinal image. But the purpose of a photograph (or of a realistic picture) is not to reproduce the retinal image...