Go toSemiotic Review of Books Home Page
Go toSRB Highlights
Go toSRB Archives


SRB Archives

This review appeared in Volume 2 (2) of The Semiotic Review of Books.

Perspectives on Perspective

by Charles Lock

Cassirer, Panofsky, and Warburg: Symbol, Art, and History. By Sylvia Ferretti, translated by Richard Pierce. New Haven and London: Yale University Press, 1989. 304 pp. ISBN 0-300-04516-6

Semiotics of Visual Language. By Fernande Saint Martin. Bloomington and Indianapolis: Indiana University Press, 1990. 271 pp. ISBN 0-253-35057-3

Changing Images of Pictorial Space: A History of Spatial Illusion in Painting. By William V. Dunning. Syracuse, NY: Syracuse University Press, 1991 .272 pp. ISBN 0-8156-2505-7

To understand a painting it is generally considered a good idea to look at it. But to understand that aspect of visual art known as perspective, it is just as important to observe the observers. For perspective sets up that symmetrical and reflexive relationship between the viewer and the viewed that enables another person, a "meta-viewer", to infer much about the one by looking at the other.

In "traditional perspective", as we may label all those versions -- orthogonal, aerial, tonal, etc. that have prevailed from the fifteenth century up to the recent past, the viewer is offered an optimal point of view. That point of view corresponds to the spot from which the painter viewed the scene before representing it, and in the intervals of the process of painting it. As the spot tends to be some distance from the depicted foreground, on the canvas itself, we recognize in film the act of painting less by brush and palette than by the mating ritual, moving close and backing off, over and over, performed between the artist and his easel. By contrast, the viewer moves quite unthinkingly towards the optimal view-point and stays there; only with conscious purpose does the viewer step out of that determined or enchanted circle, to move close to the painting, within arm's reach, in order to observe the marks of the brush, the textures, the signs of making.

It is a useful exercise to go into an art gallery and avert one's eyes from the paintings. By restricting one's visual information to the positions and movements of the other visitors one is forced to make guesses about the paintings: the distance between the viewer and the painting tells us something about the visual depth of the painting, and the size of the view-point -- that is, the number of people who can share the optimal space -- tells us about the visual field. With a little practice one can guess, from the viewers alone, the type of painting: still-life, landscape, portrait, group-portrait, interior. And if one can make no confident guess on the basis of the viewers' positions and movements, it is a safe guess that the painting dates from either before 1400 or after 1900.

Traditional perspective is comforting and reassuring not because it enables us to know where we stand, but rather because it dictates to us where we ought to stand. It is of course integral to the "awe" that great art inspires, that in its presence we are so physically controlled, almost (hence the awe) in spite of ourselves.

And painting that fails to dictate its spatial terms to us is a betrayal and an imposture. It is this, rather than the messiness, technical incompetence, or offensive distortions, that caused outrage in the early viewers of Impressionist and Post-lmpressionist works; and that continues to draw popular mockery and opprobrium on Cubism, Suprematism, Abstraction and other recent movements, even while Post-Impressionism rules today as a standard of middle-brow taste. In short, the absence of perspective is a sign of anarchy, a failure of crowd control, an abnegation of art's responsibility to order not only the pictorial space but the space of living.

This may help to explain the almost inevitable and rather distinctive snobbery that goes with an appreciation of modern art. For there is a conceptual sleight of mind by which the terms of the debate have switched from the aesthetic to the philosophical. Modern art is bearable, even enjoyable, to those viewers for whom what is viewed is less important than the freedom with which we can carry out our viewing. In the polemics of art criticism since about 1870 the advocates of modern art have been asked to justify the grotesque shapes, the failure of mimesis, the refusal to please. But to the emancipated viewer these are simply not problems: rather, they are the occasions and means of emancipation.

It is then no coincidence (as we say) that perspective could be treated theoretically, as a way and not the way of seeing and representing, only when the viewer-subject had been philosophically emancipated from transcendental status. Apart from artistic practice, and the often opaque manifestos of early twentieth century painters, there appears to be no earlier instance of perspective understood in a theoretical manner than Panofsky's "Die Perspektive als 'Symbolische Form' " of 1924-25, translated into French in 1929, and -- astonishing -- not yet published in English, even though there is an English translation available in typescript from the Warburg Institute (Dunning, 241). Silvia Ferretti, whose Italian book of 1984 has now been translated, sets Panofsky's essay in the context of his debate with the "neo-post-Kantianism" of Cassirer.

Once Alberti had derived the system of perspective from the uniformity and consistency of space, it was space that became an a priori category, and perspective became the neutral medium of the representation of space. Only when space ceased to be understood as uniform could one think again about perspective. Combining the histories of art, science and philosophy, Ferretti shows how, with the help of Cassirer's philosophical historicism, Panofsky understood that the uniformity of space implies its endlessness. We may note here a further debt to Koyré, whose "closed cosmos and open universe" find their representational parallels in the bounded space of pre-Renaissance art and in the infinity of space as implied and indicated by the "vanishing-point". Between Alberti and Cassirer is Kant, whose philosophy of the subject can be compared to Renaissance perspective, even if Panofsky hesitates to make aetiological attributions:

Panofsky compares the Renaissance function of perspective to that of Kantian criticism, in that it rationalizes the subjective visual impression to the point of being able to construct an empirical world upon a solid and yet infinite foundation. (Ferretti, 208).

This, Ferretti assesses in ambivalent terms: "Here we are witnessing a true apotheosis of Kantian criticism"(207). Because Panofsky was aware of the rejection of Renaissance perspective in the art of his contemporaries, he "kept a prudent distance from Cassirer's optimistic belief in the fulfilment of the evolution of the symbol" (203). Ferretti fails to pursue the implications of this, and to inquire into his subsequent abandonment of a theoretical criticism, concerned with the semiotics of space, in favour of an altogether more conservative mode of scholarship in iconology.

Running through Ferretti's study is a fairly critical attitude towards Gombrich, explicitly for matters under dispute in his "intellectual biography" of Warburg, and implicitly for Gombrich's broader championing of Warburg as patron of a regressive art historical discipline. Warburg and Panofsky were, on the evidence adduced by Ferretti, working towards a radicalism, of concept and discipline, when something happened (not, one assumes, unrelated to certain events of the 1930's) and the discipline over which Panofsky and then Gombrich have been so dominant over the past half century relapsed into an intellectually powerful but still traditional mix of the history of ideas, antiquarianism and connoisseurship.

Yet it must be acknowledged that Gombrich has always been very faithful to certain axioms of semiotics, especially those that would insist on relating any theory of signs to theories of perception. One's frustration and bewilderment at Gombrich's massive achievement is focused on the avoidance of connection: no other art historian has been so insistent on the need to understand the psychology and the physiology of representation, yet his iconological scholarship proceeds with an objectivity apparently oblivious of its limitations. In the light of Ferretti's study it is intriguing to speculate what might have happened had Panofsky and others developed the critique of Cassirer's theory of the symbol, and accepted with greater theoretical conviction the worth of twentieth-century painting. Art history would have been more prominent in theoretical debate.

If art history is to make contributions beyond its well-defined borders, it will be indebted to scholars such as Fernande Saint-Martin, whose Semiotics of Visual Language brings to a wider audience her Sémiologie du langage visuel of 1987. As theoretical treatise and as a basic hand book, Saint-Martin's work acknowledges the tardiness of the discipline and the need to do double duty. The appendices attempt to ground a "repertory of forms" through a system of notation which would permit the systematic description of any aspect of visual language. With usage and familiarity such a scheme could become a valuable instrument in treating of "visible marks" in ways which would have nothing to do with "images" and "representations" and all the other residue of mimesis. It would be the equivalent in artistic scholarship of the basic metrical schemes by which poetry can be described without reference to its reverentially.

The enterprise is saved from the look of folly by Saint-Martin's consciousness of its extraordinary belatedness, and, interestingly enough, she acknowledges her debt to the early work of Panofsky.

Unfortunately, pictorial semiotics has been slow to recognize the fact that cultural norms do not assure an absolute status to the artificial perspective of the Renaissance, whose restrictive and ideologically regressive character had for a long time been pointed out by artists as well as a number of art theorists.

But, following Panofsky s research, it can be said that any system of perspective is restrictive, because it is based on a specific point of view....

Saint-Martin then goes on to classify no fewer than twenty-six systems of perspective. One questions not whether there might not actually be twenty five or twenty-seven, but whether all systems are equally repressive, and indeed whether all systems of perspective are in fact based on a single point of view. The instances which Saint-Martin herself provides would seem to contradict the claim. What one apprehends is the usefulness of the tactic for Panosky: he could at the same time uncover the philosophical and ideological bases of Renaissance perspective, and continue to champion it and uphold it as normative, because all other systems are equally suspect. This is a familiar manoeuvre of conservatives inconvenienced and embarrassed by their own intellectual powers. Why Saint-Martin should give her assent is more problematic. For she writes very well about twentieth-century painting, and teases out the often recalcitrant links between manifestos and artistic practice, notably show of Kandinsky, Albers, Mondrian, and Hofmann. Without the vulgarity of grading these artists on scales of repression and emancipation, it should still be possible to acknowledge difference, to recognize that some systems of perspective may be more or less restrictive than others, or at least restrictive in different ways and for different purposes.

In his rather sketchy and conventional "history of spatial illusion in painting" William Dunning makes the interesting point that no sooner had Renaissance perspective been abandoned than the flatness of the picture plane became the new dogma: by the middle of the twentieth century critics and artists alike took it as a given -- even if, for traditionalists, this was a given within a hiatus. To that extent Panofsky, and Saint-Martin, may be right: the prevailing system of perspective, whatever it might be, has a way of looking inevitable. But whereas Renaissance perspective has the inevitability of the way the world looks, the unbroken picture plane quickly acquired the inevitability of the way modern art looks.

These recent studies suggest that two related and long overdue projects may be ripe for realization: a theoretical critique of the history of spatial representation, and a historical critique of the theories of spatial representation. None of these books, it must be said, takes us beyond John White's extraordinary work of 1957 (looking more impressive and proleptic with each passing decade), The Birth and Rebirth of Pictorial Space. But these and other recent works have made such an advance seem possible, soon.

Charles Lock is Associate Professor of English, Erindale College, University of Toronto, author of Thomas Hardy: The State of the Art and a forthcoming biography of John Cowper Powys.

Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives