Well, I have wandered far and fancifully from the undertaker king and his enigmatic nightmare of whales and Jonahs. But imaginative wandering is the privilege of the ignorant. When one doesn’t know one is free to invent. I have seized the opportunity while it presented itself. One of these days I may discover what the picture is about, and when that has happened I shall no longer be at liberty to impose my own interpretations. Imaginative criticism is essentially an art of ignorance. It is only because we don’t know what a writer or artist meant to say that we are free to concoct meanings of our own. If El Greco had somewhere specifically told us what he meant to convey by painting in terms of Black Holes and mucus, I should not now be in a position to speculate. But luckily he never told us; I am justified in letting my fancy loose to wander.

(From Music at Night)

A painter or a sculptor can be simultaneously representational and nonrepresentational. In their architectural backgrounds and, above all, in their draperies, many works even of the Renaissance and the Baroque incorporate passages of almost unadulterated abstraction. These are often expressive in the highest degree. Indeed, the whole tone of a representational work may be established, and its inner meaning expressed, by those parts of it which are most nearly abstract. Thus, the pictures of Piero della Francesca leave upon us an impression of calm, of power, of intellectual objectivity and stoical detachment. From those of Cosimo Tura there emanates a sense of disquiet, even of anguish. When we analyze the purely pictorial reasons for our perception of a profound difference in the temperaments of the two artists, we find that a very important part is played by the least representational elements in their pictures—the draperies. In Piero’s draperies there are large unbroken surfaces, and the folds are designed to emphasize the elementary solid-geometrical structure of the figures. In Tura’s draperies the surfaces are broken up, and there is a profusion of sharp angles, of jagged and flame-like forms. Something analogous may be found in the work of two great painters of a later period, Poussin and Watteau. Watteau’s draperies are broken into innumerable tiny folds and wrinkles, so that the color of a mantle or a doublet is never the same for half an inch together. The impression left upon the spectator is one of extreme sensibility and the most delicate refinement. Poussin’s much broader treatment of these almost non-representational accessories seems to express a more masculine temperament and a philosophy of like akin to Piero’s noble stoicism.

In some works the non-representational passages are actually more important than the representational. Thus, in many of Bernini’s statues, only the hands, feet and face are fully representational; all the rest is drapery—that is to say, a writhing and undulant abstraction. It is the same with El Greco’s paintings. In some of them a third, a half, even as much as two thirds of the entire surface is occupied by low-level organic abstractions, to which, because of their representational context, we give the name of draperies, or clouds, or rocks. These abstractions are powerfully expressive, and it is through them that, to a considerable extent, El Greco tells the private story that underlies the official subject matter of his paintings.

At this point the pure abstractionist will come forward with a question. Seeing that the non-representational passages in representational works are so expressive, why should anyone bother with representation? Why trouble to tell a high-level story about recognizable objects when the more important low-level story about the artist’s temperament and reactions to life can be told in terms of pure abstractions? I myself have no objection to pure abstractions which, in the hands of a gifted artist, can achieve their own kind of aesthetic perfection. But this perfection, it seems to me, is a perfection without rather narrow limits. The Greeks called the circle “a perfect figure.” And so it is—one cannot improve on it. And yet a composition consisting of a red circle inscribed within a black square would strike us, for all its perfection, as being a little dull. Even aesthetically the perfect figure of a circle is less interesting than the perfect figure of a young woman. This does not mean, of course, that the representation of the young woman by a bad artist will be more valuable, as a picture, than a composition of circles, squares and triangles devised by a good one. But it does mean, I think, that Nature is a richer source of forms than any textbook of plane or solid geometry. Nature has evolved innumerable forms and, as we ourselves move from point to point, we see large numbers of these forms, grouped in an endless variety of ways and thus creating an endless variety of new forms, all of which may be used as the raw materials of works of art. What is given is incomparably richer than what we can invent. But the richness of Nature is, from our point of view, a chaos upon which we, as philosophers, men of science, technicians and artists, must impose various kinds of unity. Now, I would say that, other things being equal, a work of art which imposes aesthetic unity upon a large number of formal and psychological elements is a greater and more interesting work than one in which unity is imposed upon only a few elements. In other words, there is a hierarchy of perfections. Bach’s Two-Part Inventions are perfect in their way. But his Chromatic Fantasia is also perfect; and since its perfection involves the imposition of aesthetic unity upon a larger number of elements it is (as we all in fact recognize) a greater work. The old distinction between the Fine Arts and the crafts is based to some extent upon snobbery and other non-aesthetic considerations. But not entirely. In the hierarchy of perfections a perfect vase or a perfect carpet occupies a lower rank than that, say, of Giotto’s frescoes at Padua, or Rembrandt’s Polish Rider, or the Grande Jatte of Georges Seurat. In these and a hundred other masterpieces of painting the pictorial whole embraces and unifies a repertory of forms much more numerous, varied, strange and interesting than those which come together in the wholes organized by even the most gifted craftsmen. And, over and above this richer and subtler formal perfection, we are presented with the non-pictorial bonus of a story and, explicit or implicit, a criticism of life. At their best, non-representational compositions achieve perfection; but it is a perfection nearer to that of the jug or rug than to that of the enormously complex and yet completed unified masterpieces of representational art—most of which, as we have seen, contain expressive passages of almost pure abstraction. At the present time it would seem that the most sensible and rewarding thing for a painter to do is (like Braque, for example) to make the best and the most of both worlds, representational as well as non-representational.

Within his own Byzantine-Venetian tradition El Greco did precisely this, combining representation with abstraction in a manner which we are accustomed to regard as characteristically modern. His intention was to use this powerful artistic instrument to express, in visual terms, man’s capacity for union with the divine. But the artistic means he employed were such that it was not possible for him to carry out that intention. The existence of a spiritual reality transcendent and yet immanent, absolutely other and yet the sustaining spiritual essence of every being, has frequently been rendered in visual symbols—but not symbols of the kind employed by El Greco. The agitation of quasi-visceral forms in an overcrowded and almost spaceless world, from which non-human Nature has been banished, cannot, in the very nature of things, express man’s union with the Spirit who must be worshiped in spirit.