"Poussin’s fame generally danced to the tune of the times."

               H.-W. van Helsdingen[1]

            How complex can interpretations of pictures legitimately be? The development of art history is marked by movement from simpler to more complex interpretations.[2] Is Vasari’s account of Leonardo’s Last Supper simpler than Leo Steinberg’s book-length account? Most people think so, but Paul Barolsky has convincingly argued that Vasari is a highly subtle writer who tends to be read in overly simple ways. Certainly, however, Vasari’s account is much briefer. It is easy to give a sociological explanation of this change. Where Vasari’s commentaries are terse, ours tend to be long-winded. Vasari, a busy painter, wrote as a sideline. But once the academic institutions were established, inevitably interpretation became more elaborate. If nowadays you are a graduate student, publishing an account of some much discussed painting requires demonstrating that your precursors missed something.

                  This move from simplicity towards complicity occurs in diverse branches of art history. We find it in debates about Piero della Francesca and Caravaggio. Piero’s personality is elusive, and so analysis is guided just by what we see. Caravaggio, by contrast, has a much-discussed identity, though discussion of his life, and its relationship to his painting is highly controversial. The problem, it may sometimes seem, is that interpretation is speculative because these artists are historically distant. But that cannot be the whole story, for the same development appears, also, in discussion of Robert Rauschenberg’s collages and Robert Ryman’s minimalist paintings.[3] During Rauschenberg’s lifetime, opposed interpretations appeared. And Ryman describes his art in ways that are not consistent with the claims of his best-known champions.

                   Recently, in reaction to ever more elaborate commentaries, there have been attempts to present simple interpretations. Perhaps Piero’s and Caravaggio’s paintings are over-interpreted by bookish academics. Once we recognize that no contemporary of the artists describes them in our modern terms, then it is natural to wonder if our reconstructions are plausible. Just as some highly privileged people embrace simplicity, so when we have many elaborate interpretations, some commentators are attracted by simplicity. Nowadays, however, the practice of simple interpretation requires sophisticated supportive argumentation.

                  The goal of interpretation is to understand pictures as the artist intended. We want to reconstruct the meaning of Piero’s, Caravaggio’s, Rauschenberg’s and Ryman’s art in their terms. Why not, then, settle for commentary contemporary with the artist? This strategy will work with Rauschenberg and Ryman. But since we have almost nothing from Piero’s contemporaries, and only hostile incomprehension from Caravaggio’s, with them a different strategy is required. Sometimes, of course, it has been argued that interpretation itself is a creative act, like making the painting it describes.  True enough, but we want that our art historical accounts be true to the pictures. Caravaggio has become a culture hero, and so novels about him appear. The best one, Christopher Peachment’s Caravaggio, uses frankly speculative reconstruction of the artist’s life to project interpretation of the pictures.[4] Suggestive as his account is, there is a difference in kind between such fiction and art historical analysis.

           Most academic interpretations build upon the existing structure of debate. Once, for example, Erwin Panofsky presented his iconographic analysis of Jan van Eyck’s Arnolfini Marriage, then inevitably other scholars elaborated that account, and applied it also to other paintings. And, also predictable, was the more recent reaction against such accounts. Inspired by feminism, art historians now are aware that Panofsky says nothing about the patriarchal view of marriage presented by van Eyck. As our epigraph indicates, interpreters always are of their time, which is to say that as the general culture changes, so too do interpretative styles.

           The question, still, which concerns the philosopher of art history is how such changes are consistent with the basic goal, truth in interpretation. We hope that we have progressed, and so have more truthful interpretations than did our ancestors. But once we acknowledge that interpretation is influenced by contemporary ways of thinking, is this belief plausible? Perhaps interpretations improve. If the contemporaries of the old masters lacked the proper vocabulary required to describe these works of art, then only now can Piero and Caravaggio be adequately described. Alternatively, we might relativize the truth in interpretation. Vasari had one legitimate style of interpretation, Panofsky another and we still another. The old pictures we describe have not changed, but we may legitimately, i.e. truthfully, describe them differently. On this view, an ahistorical analysis of truth in interpretation cannot be acceptable.

           I came into art history from philosophy some twenty years ago because I found these questions exciting. Here, then, I take them up from a different perspective, with reference to Nicolas Poussin. In a sequence of published and forthcoming essays, I revise my Poussin’s Paintings (1993), taking into account more recent scholarship.[5] Commentary on Poussin poses questions about truth in interpretation in especially pointed form. His correspondence says much about his art.[6] And in his century, elaborate accounts appeared. Because he is identified as the philosopher-painter, there is a tendency to assume that he was extremely erudite. Once Anthony Blunt offered influential highly complex accounts, other scholars, in response, offered still more elaborate interpretations.

           When Poussin depicts a landscape, some commentators think that every person and architectural element must have a source in some written account. If he himself did not read a book, at least it could have been known to some learned colleague. (He was friendly with some Roman intellectuals.) Whatever Poussin depicts can, with ingenuity, be matched with some written commentary. But surely he could have depicted figures and a city in a landscape without reference to a text? With modernist art, we are accustomed to think that the painter might have invented his composition. Why, then, should we not give Poussin the same freedom? Why must his pictures always be controlled by some source? Once we recognize that any image can, with ingenuity, be linked with some appropriate text, then we should recognize that establishing word-image correlations is not illuminating. Within the older literature, this contrast was anticipated when Anthony Blunt and Denis Mahon compared concern with Poussin’s social and political context with interest in Poussin as a painter.

            The great 1994 Poussin retrospective, in Paris and London, marking his four-hundredth birthday, did not come to New York. But in 2008, in delayed recompense, there was a large exhibition "Poussin and Nature" at the Metropolitan Museum of Art. The goal of Keith Christiansen, the New York curator was to focus on the landscape paintings. As he notes in his catalogue essay, Poussin is not normally thought a landscape painter.[7] In the seventeenth-century, Poussin’s landscapes were given lessor value than his history paintings. Even in Joshua Reynolds’s time, it was taken for granted that elevated subjects made history paintings essentially more important than landscapes. Reynolds’s Discourses, with their surprised judgment of the importance of his friend, Gainsborough, display this bias. Ernst Gombrich traced the rise of pure landscape painting.[8] Without mentioning Poussin, he identified the way in which the creation of this novel genre required a new public, "the consumer of collector, who creates the demand."  Poussin, Christiansen is suggesting, responds to that demand.

                  What is a pure landscape painting? If Constable is our model, as he is Gombrich’s model in Art and Illusion, then a landscape is a painting in which any human figures are mere adjuncts to nature herself. Farmers, mill workers or shepherds can appear, but not identifiable people. So, for example, paintings showing Hannibal crossing the Alps or Goethe meditating in the Campagna cannot be pure landscapes. To put this distinction another way, a history painting and even a portrait of Goethe can only be fully understood by reference to some text identifying the people present.  But a pure landscape merely shows an attractive or stormy countryside because its farmers, mill workers and shepherds are not identifiable individuals. Landscape painting is a purely aesthetic genre.

           Christiansen calls Poussin the landscape painter a "proto-romantic."[9] I would put this point in a stronger way. Once Constable’s pure landscapes appeared, Clement Greenberg’s modernist separation of literature from visual art was in the wings. Ambitious painting no longer needed to ally itself with texts. Of course, visual modernism did not only involve landscape painting. More often, in fact, it was associated with Baudelairian urban subjects. But once this genre was developed, then the way was open to thinking in formalist terms that a significant painting’s subject was essentially irrelevant to judging its aesthetic value. As I reconstruct Christiansen’s concerns, one goal is to make Poussin acceptable to contemporary taste by focusing on the landscapes.

           In the nineteenth century, two very different Poussins appear, the history painter and "a painter of nature whose works could be appreciated with little or no literary knowledge."[10] To understand the history paintings, we must know the story depicted. But with a landscape, there is no deeper, hidden meaning requiring exegesis. What then has gone wrong, Christiensen suggests, is the recent tendency of scholarship to treat the landscapes as if they were history paintings. He seeks simpler interpretations.

           How, then, should we understand the title of the exhibition, "Poussin and Nature"? Poussin painted only a few pure landscapes. Some early pictures, Landscape with a Nymph and Sleeping Satyr might count.

             But Landscape with Orpheus and Eurydice, Landscape with Diogenes, and Blind Orion are not pure landscapes. And classifying Landscape with a Man Killed by a Snake and the pair of paintings, Landscape with a Storm and Landscape with a Calm is difficult. Perhaps the former merely depicts a dead man and the latter just present two very different weather reports. But if, as many commentators have claimed, these pictures illustrate some literary conception, then they, as much as Poussin’s classic history paintings, can only be understood in relation to a text. 

           You cannot tell if a picture is a pure landscape just by looking at it. Envisage an example of the form Arthur Danto has made familiar.[11] Consider two indiscernible pictures. The first, Landscape with a Storm, just shows what the title identifies, a story. The second, visually identical, is a commentary on the need for political stability in times of strife. No mere act of connoisseurship could allow us to distinguish between these paintings. This is an imaginary example, but the problem for Poussin scholarship is that we don’t know how to classify his pictures. Maybe there is more to his landscapes than meets the eye, but perhaps they are just attractive scenes. In the history paintings, Christ Healing the Blind for example, human figures are large and set in the foreground. But although the human figures in Landscape with Phramus and Thisbe, which the Metropolitan includes in "Poussin and Nature," are relatively small, you cannot understand that picture without knowing the story of Phramus and Thisbe.

           Christiansen claims that William Hazlitt’s account marks a fundamental shift in the way that Poussin was understood. Focusing on the landscapes, Hazlitt argues that Poussin was among painters (more than any one else) what Milton was among poets. There is in both something of the same pedantry, the same stiffness, the same elevation, the same grandeur, the same mixture of art and nature, the same richness of borrowed materials, the same unity of character.[12]

           This Romantic way of thinking, Christiansen argues, marks a decisive change in taste, both in the way that Poussin was understood, and in the larger culture. In his essay "On the pleasure of painting" Hazlitt notes how painting exercises the body. It is a mechanical as well as a liberal art. To do any thing, to dig a hole in the ground, to plant a cabbage, to hit a mark... to attempt to produce any effect, and to succeed, has something in it that... carries off the restless activity of the mind of man. [13]

           Poussin’s contemporaries did not say anything like this. On the contrary, they tended to focus on the artist’s intellectual concerns.

           Hazlitt’s brief account of Orion certainly is very suggestive. But you need only set it alongside Ernst Gombrich’s iconographic analysis, and the literature it spawned to see the difference between belles-lettres writing and serious scholarship.[14] Hazlitt offers only a very slight account of the features that fascinate modern day commentators. Christiansen, I hasten to add, does not want that we return to the terms of Hazlitt’s account, but offers a suggestive novel argument.

           Christiansen’s highly original account draws on an analysis of Poussin’s patrons. In 1649 and 1650, Nicolas Poussin made his two self-portraits. The second, more famous one, painted for Paul Fréart de Chantelou, has been the object of exhaustive analysis.

             A great deal has been said about the three, or four canvases behind the artist; the words on the first; the woman depicted on the second; and, even, the book which Poussin holds and his ring. This self-portrait demands iconographic analysis, for the arrangement of pictures and the woman who has a third eye, set in closely calculated relation to the painter, are obviously enigmatic. The second, painted for another of Poussin’s French collectors, Jean Pointel, is more straightforward, and so has been less discussed. From 1945 until the end of state socialism, it was relatively inaccessible in East Berlin.

           That Chantelou and Pointel were patrons with different sorts of interests explains why Poussin offered them different self-portraits. Chantelou was most interested in the beautiful history paintings, which are intricately composed; and Pointel, in the sublime landscapes, which show the disasters associated with nature going seriously wrong—floods ( Winter), unjust political executions ( Landscape with the Ashes of Phocion), and of course the deaths by snakebite and suicide that have already been mentioned.  

            Without mentioning the sublime, Richard Wollheim wrote that Poussin shows correspondence between nature—nature considered broadly as the backdrop to human action—and what might be called mental fecundity... that unbounded capacity of the mind... to generate an indefinite profusion of thoughts,     memories, images, wishes, hopes, fears. [15]

           Nature, both understood literally as the landscape scenes and in Wollheim’s broader terms, he alludes to a psychoanalytic conception of human nature, is challenging to a seicento painter because it seems to be formless. But he did not anticipate Christiansen’s analysis, which, so far as I can see, is entirely original.

           I myself would be happier if Christiansen did not ask that we "return to the magic of the pictures themselves," as if their visual qualities were transparently accessible.[16] But this is a minor complaint. The real difficulty I have with his analysis lies in the terms of debate. Either Poussin’s pictures are highly subtle illustrations of literary references, or they are beautiful landscapes. I agree with Christiansen: the development of ever more elaborate iconographic interpretations no longer seems productive. But what as yet remains inadequately described, I believe, are the purely visual qualities of Poussin’s landscapes. These paintings, though not physically very large, open up enormously deep spaces. In Landscape with Phramus and Thisbe, for example, Thisbe is in the foreground; a fleeing man is behind her; in the third layer of the space, a lion attacks a horseman; behind them other figures respond to the story; and, still, the space goes farther back. And in Landscape with a Man Killed by a Snake, we have the dead man in the foreground; the man who responds and the woman who sees this running figure, but not the cause of his fear; while behind them figures lounge and, behind them, other men work on a boat, all far in front of a distant cityscape. Poussin inspires contemplative thought.  Here, I follow Wollheim, because these deep walk-in landscapes demand prolonged attention. But he almost always blocks access by setting rocks and trees in the foreground.[17]

           Nicolas Poussin certainly was not Paul Cézanne. But insofar as his landscapes demand a visually sensitive response, he led the way towards modernism.

             In an interesting way, the Metropolitan’s presentation of Poussin as Hazlitt’s proto-modernist was deconstructed by its installation alongside large Courbet and Jasper Johns exhibitions. Walking between these three displays, what to me seemed most striking was how different were Poussin’s deep spaces from those of his modernist successors.  After looking at his paintings, I found it hard to properly attend to Courbet and Johns. Insofar as Poussin’s landscapes are not governed by some prior text, they are proto-modernist pictures. But, again following Greenberg, if modernist painting is defined by flatness, then Poussin’s characteristic deep spaces make him an anti-modernist. But here I only open up what will, I hope, be ongoing discussion.

            **This essay is for Edward Olszewski, in thanks for his generous support of my writing and honest leadership of my department. Thanks are also due to Svetlana Alpers and Bill Berkson for accompanying me through this exhibition; to Keith Christiansen and Ivan Gaskell for comments; and to Margaretha Lagerlšf.

           

---