I. W. J. T. MITCHELL 4 W. J. T. Mitchell representation; that very exhaustion, however, may signal an enhanced power at other levels (in mass culture and kitsch, for instance) and a potential for renewal in other forms, other places. Notes 1. This approach to landscape aesthetics is most fully developed in the influential work of Ernst Gombrich, particularly his essay "The Renaissance Theory of Art and the Rise of Landscape," in Norm and Form: Studies in the Art of the Renaissance (Chicago, 1966). See also Kenneth Clark, Landscape into Art (Boston, 1963), which popularizes and universalizes Gombrich's claim. 2. See, for instance, Reading Landscape: Country—City—Capital, ed. Simon Pugh (Manchester, 1990): "This collection of essays proposes that landscape and its representations are a 'text' and are, as such, 'readable' like any other culniral form" (2-3). ONE Imperial Landscape Theses on Landscape 1. Landscape is not a genre of art but a medium. 2. Landscape is a medium of exchange between the human and the natural, the self and the other. As such, it is like money: good for nothing in itself, but expressive of a potentially limitless reserve of value. 3. Like money, landscape is a social hieroglyph that conceals the actual basis of its value. It does so by naturalizing its conventions and conventionalizing its nature. 4. Landscape is a natural scene mediated by culture. It is both a represented and presented space, both a signifier and a signified, both a frame and what a frame contains, both a real place and its simulacrum, both a package and the commodity inside the package. 5. Landscape is a medium found in all cultures. 6. Landscape is a particular historical formation associated with European imperialism. 7. Theses 5 and 6 do not contradict one another. 8. landscape is an exhausted medium, no longer viable as a mode of artistic expression. Like life, landscape is boring; we must not say so. 9. The landscape referred to in Thesis 8 is the same as that of Thesis 6. 6 W. J. T. Mitchell We are surrounded with things which we have not made and which have a life and structure different from our own: trees, flowers, grasses, rivers, hills, clouds. For centuries they have inspired us with curiosity and awe. They have been objects of delight. We have recreated them in our imaginations to reflect our moods. And we have come to think of them as contributing to an idea which we have called nature. Landscape painting marks the stages in our conception of nature. Its rise and development since the middle ages is part of a cycle in which the human spirit attempted once more to create a harmony with its environment. —Kenneth Clark, Landscape into Art (1949) We have come a long way from the innocence of Kenneth Clark's opening sentences to Landscape into Art. Most notably, perhaps, the "we" for whom Clark speaks with such assurance can no longer express itself outside of quotation marks. Who is this "we" that defines itself by its difference from "trees, flowers, grasses, rivers, hills, clouds" and then erases this difference by re-creating it as a reflection of its own moods and ideas? Whose history and whose nature is "marked" into "stages" by landscape painting? What disruption required an art that would restore the "human spirit" to "harmony with its environment"? Recent criticism of landscape aesthetics—a field that goes well beyond the history of painting to include poetry, fiction, travel literature, and landscape gardening—can largely be understood as an articulation of a loss of innocence that transforms all of Clark's assertions into haunting questions and even more disquieting answers. "We" now know that there is no simple, unprobiematic "we," corresponding to a universal human spirit seeking harmony, or even a European "rising" and "developing" since the Middle Ages. What we know now is what critics like John Barrell have shown us, that there is a "dark side of the landscape" and that this dark side is not merely mythic, not merely a feature of the regressive, instinctual drives associated with nonhuman "nature" but a moral, ideological, and political darkness that covers itself with precisely the sort of innocent idealism Clark expresses.1 Contemporary discussions of landscape are likely to be contentious and polemical, as the recent controversy over the Tate Gallery's exhibition and the monograph on the works of Richard Wilson suggest.2 They are likely to place the aesthetic idealization of landscape alongside "vulgar" economic and material considerations, as John Barrel! and Ann Bcrmingham do when they put die English landscape movement in the context of the enclosure of common fields and the dispossession of the English peasantry.3 I might as well say at the outset that I am mainly in sympathy with this darker, skeptical reading of landscape aesthetics and that this essay is Imperial Landscape 7 an attempt to contribute further to this reading. Our understanding of "high" art can, in general, benefit considerably from a critical perspective diat works through what Philip Fisher has called the "hard facts" embedded in idealized settings.4 My aim in this essay, however, is not primarily to add to the stock of hard facts about landscape but to take a harder look at the framework in which facts about landscape are constituted—the way, in particular, that the nature, history, and scmiotic or aesthetic character of landscape is constructed in both its idealist and skeptical interpretations. As it happens, there is a good deal of common ground in these constructions, an underlying agreement on at least three major "facts" about landscape: (1) that it is, in its "pure" form, a western European and modern phenomenon; (2) that it emerges in the seventeenth century and reaches its peak in the nineteenth century; (3) that it is originally and centrally constituted as a genre of painting associated with a new way of seeing. These assumptions are generally accepted by all the parties in contemporary discussions of English landscape, and to the extent that they provide a common grammar and narrative shape for criticism, they foster a kind of mirror symmetry between the skeptical critique and the idealist aesthetic it opposes. Clark's opening paragraph, for instance, may be read as still true if only its key terms are understood in an ironic sense: the "different structure" of nature is read as a symptom of alienation from the land; the "reflective" and imaginary projection of moods into landscape is read as the dreamwork of ideology; the "rise and development" of landscape is read as a symptom of the rise and development of capitalism; the "harmony" sought in landscape is read as a compensation for and screening off of the actual violence perpetrated there. The agreement on these three basic "facts"—let us call them the "Western-ness" of landscape, its modernity, and its visual/pictorial essence— may well be a sign of just how well founded they are. If critics of radically different persuasions take these things for granted, differing mainly in their explanations of them, then there is a strong presumption that they are true. The modernity of European landscape painting, for instance, is one of the first lessons landscape historians pass on to their students. Ernst Gombrich's classic essay "The Renaissance Theory of Art and the Rise of Landscape" (1953), with its story of the "revolutionary" emergence of a new genre called landscape in sixteenth-century European painting, is still die basic reference point for art-historical treatments of this topic.5 Kenneth Clark expresses the lesson in its most general form: "People who have given the matter no thought are apt to assume that the appreciation of natural beauty and the painting of landscape is a normal and enduring part of our spiritual activity. But the truth is that in times when 6 W. J. T. Mitchell We are surrounded with things which we have not made and which have a life and structure different from our own: trees, flowers, grasses, rivers, hills, clouds. For centuries they have inspired us with curiosity and awe. They have been objects of delight. We have recreated them in our imaginations to reflect our moods. And we have come to think of them as contributing to an idea which we have called nature. Landscape painting marks the stages in our conception of nature. Its rise and development since the middle ages is part of a cycle in which the human spirit attempted once more to create a harmony with its environment. —Kenneth Clark, Landscape into Art (1949) We have come a long way from the innocence of Kenneth Clark's opening sentences to Landscape into An. Most notably, perhaps, the "we" for whom Clark speaks with such assurance can no longer express itself outside of quotation marks. Who is this "we" that defines itself by its difference from "trees, flowers, grasses, rivers, hills, clouds" and then erases this difference by re-creating it as a reflection of its own moods and ideas? Whose history and whose nature is "marked" into "stages" by landscape painting? What disruption required an art that would restore the "human spirit" to "harmony with its environment"? Recent criticism of landscape aesthetics—a field that goes well beyond the history of painting to include poetry, fiction, travel literature, and landscape gardening—can largely be understood as an articulation of a loss of innocence that transforms all of Clark's assertions into haunting questions and even more disquieting answers. "We" now know diat there is no simple, unproblcmatic "we," corresponding to a universal human spirit seeking harmony, or even a European "rising" and "developing" since the Middle Ages. What we know now is what critics like John Barrel! have shown us, that there is a "dark side of the landscape" and that this dark side is not merely mythic, not merely a feature of the regressive, instinctual drives associated with nonhuman "nature" but a moral, ideological, and political darkness that covers itself with precisely the sort of innocent idealism Clark expresses.1 Contemporary discussions of landscape are likely to be contentious and polemical, as the recent controversy over the Tate Gallery's exhibition and the monograph on the works of Richard Wilson suggest.2 They are likely to place the aesthetic idealization of landscape alongside "vulgar" economic and material considerations, as John Barrell and Ann Bermingham do when they put the English landscape movement in the context of the enclosure of common fields and the dispossession of the English peasantry.3 I might as well say at the outset that I am mainly in sympathy with this darker, skeptical reading of landscape aesthetics and that this essay is Imperial Landscape 7 an attempt to contribute further to this reading. Our understanding of "high" art can, in general, benefit considerably from a critical perspective that works through what Philip Fisher has called the "hard facts" embedded in idealized settings.4 My aim in this essay, however, is not primarily to add to the stock of hard facts about landscape but to take a harder look at the framework in which facts about landscape are constituted—the way, in particular, that the nature, history, and semiotic or aesthetic character of landscape is constructed in both its idealist and skeptical interpretations. As it happens, there is a good deal of common ground in these constructions, an underlying agreement on at least three major "facts" about landscape: (1) that it is, in its "pure" form, a western European and modern phenomenon; (2) that it emerges in the seventeenth century and reaches its peak in the nineteenth century; (3) that it is originally and centrally constituted as a genre of painting associated with a new way of seeing. These assumptions are generally accepted by all the parties in contemporary discussions of English landscape, and to the extent that they provide a common grammar and narrative shape for criticism, they foster a kind of mirror symmetry between the skeptical critique and the idealist aesthetic it opposes. Clark's opening paragraph, for instance, may be read as still true if only its key terms arc understood in an ironic sense: the "different structure" of nature is read as a symptom of alienation from the land; the "reflective" and imaginary projection of moods into landscape is read as the dreamwork of ideology; the "rise and development" of landscape is read as a symptom of the rise and development of capitalism; the "harmony" sought in landscape is read as a compensation for and screening off of the actual violence perpetrated there. The agreement on these three basic "facts"—let us call them the "Wcst-ern-ness" of landscape, its modernity, and its visual/pictorial essence— may well be a sign of just how well founded they are. If critics of radically different persuasions take these things for granted, differing mainly in their explanations of them, then there is a strong presumption that they are true. The modernity of European landscape painting, for instance, is one of the first lessons landscape historians pass on to their students. Ernst Gombrich's classic essay "The Renaissance Theory of Art and the Rise of Landscape" (1953), with its story of the "revolutionary" emergence of a new genre called landscape in sixteenth-century European painting, is still the basic reference point for art-historical treatments of this topic.5 Kenneth Clark expresses the lesson in its most general form: "People who have given the matter no thought are apt to assume that the appreciation of natural beauty and the painting of landscape is a normal and enduring part of our spiritual activity. But die truth is that in times when 8 W. J. T. Mitchell the human spirit seems to have burned most brightly the painting of landscape for its own sake did not exist and was unthinkable."6 Marxist art historians replicate this "truth" in the narrower field of English landscape aesthetics, substituting the notion of ideology for Clark's "spiritual activity." Thus Ann Bermingham proposes "that there is an ideology of landscape and that in die eighteenth and nineteenth centuries a class view of landscape embodied a set of socially and, finally, economically determined values to which the painted image gave cultural expression."7 Neither Bermingham nor Barrel! makes the explicit claim for world-historical uniqueness that Clark does; they confine their attention quite narrowly to the English landscape tradition, and to even more specific movements within it. But in the absence of any larger perspective, or any challenge to Clark's larger claims, the basic assumption of historical uniqueness remains in place, subject only to differences of interpretation. A similar point might be made about the visual/pictorial constitution of landscape as an aesthetic object. Bermingham regards landscape as an ideological "class view" to which "the painted image" gives "cultural expression." Clark says that "the appreciation of natural beauty and the painting of landscape is" (emphasis mine) a historically unique phenomenon. Both writers elide die distinction between viewing and painting, perception and representation—Bermingham by treating painting as die "expression" of a "view," Clark by means of the singular verb "is" that collapses the appreciation of nature into its representation by painting. Clark goes on to reinforce the equation of painting with seeing by citing with approval Ruskin's claim in Modem Painters that "mankind acquired a new sense'" along with the invention of landscape painting. Not only landscape painting, but landscape perception is "invented" at some moment of history; the only question is whether this invention has a spiritual or a material basis.8 There arc two problems with these fundamental assumptions about the aesthetics of landscape: first, they are highly questionable; second, they are almost never brought into question, and the very ambiguity of the word "landscape" as denoting a place or a painting encourages this failure to ask questions. But the blurring of the distinction between the viewing and die representation of landscape seems, on the face of it, deeply problematic. Are we really to believe, as Clark puts it, that "the appreciation of natural beauty" begins only widi the invention of landscape painting? Certainly die testimony of poets from Hesiod to Homer to Dante suggests that human beings did not, as Ruskin thought, acquire a "new sense" sometime after the Middle Ages that made them "utterly different from all the great races that have existed before."9 Even the more restricted claim that landscape painting (as distinct from perception) has a uniquely Imperial Landscape 9 Western and modern identity seems fraught with problems. The historical claim that landscape is a "postmedieval" development runs counter to the evidence (presented, but explained away as merely "decorative" and "digressive" in Clark's text) that Hellenistic and Roman painters "evolved a school of landscape painting."10 And the geographic claim that landscape is a uniquely western European art falls to pieces in the face of the overwhelming richness, complexity, and antiquity of Chinese landscape painting." The Chinese tradition has a double importance in this context. Not only does it subvert any claims for the uniquely modern or Western lineage of landscape, the fact is that Chinese landscape played a crucial role in the elaboration of English landscape aesthetics in the eighteenth century, so much so that le jardin anglo-chinois became a common European label for the English garden.12 The intrusion of Chinese traditions into the landscape discourse I have been describing is worth pondering further, for it raises fundamental questions about the Eurocentric bias of that discourse and its myths of origin. Two facts about Chinese landscape bear special emphasis: one is that it flourished most notably at the height of Chinese imperial power and began to decline in the eighteenth century as China became itself the object of English fascination and appropriation at the moment when England was beginning to experience itself as an imperial power.13 Is it possible that landscape, understood as the historical "invention" of a new visual/pictorial medium, is integrally connected with imperialism? Certainly the roll call of major "originating" movements in landscape painting—China. Japan, Rome, seventeenth-century Holland and France, eighteenth- and ninctcenth-century Britain—makes the question hard to avoid. At a minimum we need to explore the possibility that the representation of landscape is not only a matter of internal politics and national or class ideology but also an international, global phenomenon, intimately bound up with the discourses of imperialism. This hypothesis needs to be accompanied by a whole set of stipulations and qualifications. Imperialism is clearly not a simple, single, or homogeneous phenomenon but die name of a complex system of cultural, political, and economic expansion and domination that varies with the specificity of places, peoples, and historical moments.14 It is not a "one-way" phenomenon but a complicated process of exchange, mutual transformation, and ambivalence.15 It is a process conducted simultaneously at concrete levels of violence, expropriation, collaboration, and coercion, and at a variety of symbolic or representational levels whose relation to the concrete is rarely mimetic or transparent. Landscape, understood as concept or representational practice, docs not usually declare its relation to imperialism in any direct way; it is not to be understood, in my view, as a mere 10 W. J. T. Mitchell tool of nefarious imperial designs, nor as uniquely caused by imperialism. Dutch landscape, for instance, which is often credited with being the European origin of both the discourse and the pictorial practice of landscape, must be seen at least in part as an anti-imperial and nationalistic cultural gesture; the transformation of the Netherlands from a rebellious colony into a maritime empire in the second half of the seventeenth century indicates at the very least how quickly and drastically the political environment of a cultural practice can change, and it suggests the possibility of hybrid landscape formations that might be characterized simultaneously as imperial and anticolonial.16 Landscape might be seen more profitably as something like the "dreamwork" of imperialism, unfolding its own movement in time and space from a central point of origin and folding back on itself to disclose both Utopian fantasies of the perfected imperial prospect and fractured images of unresolved ambivalence and unsuppressed resistance. In short, the posing of a relation between imperialism and landscape is not offered here as a deductive model that can settle the meaning of either term, but as a provocation to an inquiry. If Kenneth Clark is right to say that "landscape painting was the chief artistic creation of the nineteenth century,"17 we need at least to explore the relation of diis cultural fact to the other "chief creation" of the nineteenth century—the system of global domination known as European imperialism. The "Rise" of Landscape Man had not only reconquered his rights, but he had reentered upon his possession of nature. Several of these writings testify to the emotion which those poor people felt on beholding their country for the first time. Strange to relate! those rivers, mountains, and noble landscapes, where they were constantly passing, were discovered by them on that day: they had never seen them before. —Michelet, History of the French Revolution (1846) Thence up he flew, and on the Tree of Life, The middle tree and highest there that grew, Sat like a cormorant; ... ... nor on the virtue thought Of that life-giving plant, but only used For prospect, what, well used, had been the pledge Of immortality... . -Milton, Paradise Lost IV: 194-96;198-201 Imperial Landscape 11 When does landscape first begin to be perceived? Everything depends, of course, on how one defines the "proper" or "pure" experience of landscape. Thus, Kenneth Clark dismisses the landscape paintings that adorned Roman villas as "backgrounds" and "digressions," not representations of natural scenery in and for itself. Landscape perception "proper" is possible oniy to "modern consciousness," a phenomenon that can be dated with some precision. "Petrarch," Clark tells us, "appears in all the history books as the first modern man," and so it is no surprise that he is "probably the first man to express the emotion on which the existence of landscape painting so largely depends; the desire to escape from the turmoil of cities into the peace of the countryside." Clark might admit that some version of this emotion appears rather frequently in the ancient genre of the pastoral, but he would probably insist that the enjoyment of the view "for its own sake" is not quite achieved prior to "modern consciousness." Petrarch doesn't just flee die city in good pastoral fashion for the comforts of the country; he seeks out the discomforts of nature. "He was, as everyone knows, the first man to climb a mountain for its own sake, and to enjoy the view from the top."18 A fact that "everyone knows" hardly requires an argument, but Clark goes on to give one anyway. The unique historical placement of Petrarch's perception of landscape at the originary, transitional moment from ancient to modern is "proved" by showing that Petrarch himself lives in both worlds, is both a modern humanist and a medieval Christian. Thus, Clark notes that at the very moment Petrarch is enjoying the view, "it occurred to him to open at random his copy of St. Augustine's Confessions to a passage that denounces the contemplation of nature: "And men go about to wonder at the heights of the mountains, and the mighty waves of the sea, and the wide sweep of rivers, and the circuit of the ocean, and the revolution of the stars, but themselves they consider not."19 Properly abashed by rhis pious reminder, Petrarch concludes that he has "seen enough of the mountain" and turns his "inward eye" upon himself. What Clark's "historical" narrative of the development of landscape ignores is that St. Augustine's admonition is itself testimony to the antiquity of the contemplation of nature. Long before Petrarch and long before St. Augustine, people had succumbed to the temptation of looking at natural wonders "for their own sake." Numerous other "originary moments" in the viewing of landscape might be adduced, from Jehovah's looking upon his creation and finding it good, to Michelet's French peasants running out of doors to perceive the beauties of their natural environment for the first time. The account of landscape contemplation that probably had the strongest influence on English painting, gardening, and poetry in the eighteenth century was 12 W. J. T. Mitchell Milton's description of Paradise, a viewing, we should recall, that is framed by the consciousness of Satan, who "only used for prospect" his vantage point on the Tree of Life. The "dark side" of landscape that Marxist historians have uncovered is anticipated in the myths of landscape by a recurrent sense of ambivalence. Petrarch fears the landscape as a secular, sensuous temptation; Michelet treats it as a momentary revelation of beauty and freedom bracketed by blindness and slavery; Milton presents it as the voyeuristic object for a gaze that wavers between aesthetic delight and malicious intent, melting "pity" and "Honor and empire with revenge enlarged" (iv. 374; 390). This ambivalence, moreover, is temporalized and narrativized. It is almost as if there is something built into the grammar and logic of the landscape concept that requires the elaboration of a pseudohistory, complete with a prehistory, an originating moment that issues in progressive historical development, and (often) a final decline and fall. The analogy with typical narratives of the "rise and fall" of empires bcomes even more striking when we notice that the rise and fall of landscape painting is typically represented as a threefold process of emancipation, naturalization, and unification. The article "Landscape Painting" in The Oxford Companion to Art provides a handy compendium of these narratives, complete with "origins" in Rome and the Holy Roman Empire of the sixteenth century and "endings" in twentieth-century Sunday painting. Landscape painting is routinely described as emancipating itself from subordinate roles like literary illustration, religious edification, and decoration to achieve an independent status in which nature is seen "for its own sake." Chinese landscape is prehistoric, prior to the emergence of nature "enjoyed for its own sake." "In China, on the other hand, the development of landscape painting is bound up with . . . mystical reverence for the powers of nature."20 The "other hand" of landscape, whether it is the Orient, the Middle Ages, Egypt, or Byzantium, is preemancipatory, prior to the perception of nature as such. Thus, the emancipation of landscape as a genre of painting is also a naturalization, a freeing of nature from the bonds of convention. Formerly, nature was represented in "highly conventionalized" or "symbolic" forms; latterly, it appears in "naturalistic transcripts of nature," the product of a "long evolution in which the vocabulary of rendering natural scenery gained shape side by side with the power to see nature as scenery." This "evolution" from subordination to emancipation, convention to nature has as its ultimate goal the unification of nature in the perception and representation of landscape: "It seems that until fairly recent times men looked at nature as an assemblage of isolated objects, Imperial Landscape 13 without connecting trees, rivers, mountains, roads, rocks, and forest into a unified scene."21 Each of these transitions or developments in the articulation of landscape presents itself as a historical shift, whether abrupt or gradual, from ancient to modern, from classical to Romantic, from Christian to secular. Thus, the history of landscape painting is often described as a quest, not just for pure, transparent representation of nature, but as a quest for pure painting, freed of literary concerns and representation. As Clark puts it, "The painting of landscape cannot be considered independently of the trend away from imitation as the reason d'etre of art."22 One end to the story of landscape is thus abstract painting. At the other extreme, the history of landscape painting may be described as a movement from "conventional formulas" to "naturalistic transcripts of nature."23 Both stories are grail-quests for purity. On the one hand, the goal is nonrepresenta-tional painting, freed of reference, language, and subject matter; on the other hand, pure hyperrepresentational painting, a supcrlikcness that produces "natural representations of nature." As a pseudohistorical myth, then, the discourse of landscape is a crucial means for enlisting "Nature" in the legitimation of modernity, the claim that "we moderns" are somehow different from and essentially superior to everything that preceded us, free of superstition and convention, masters of a unified, natural language epitomized by landscape painting. Benedict Anderson notes that empires have traditionally relied on "sacred silent languages" like the "ideograms of Chinese, Latin, or Arabic" to imagine the unity of a "global community."24 He suggests that the effectiveness of these languages is based in the supposed nonarbitrarincss of their signs, their status as "emanations of reality," not "fabricated representations of it." Anderson thinks of the nonarbitrary sign as "an idea largely foreign to the contemporary Western mind," but it is, as we have seen, certainly not foreign to Western ideas of landscape painting. Is landscape painting the "sacred silent language" of Western imperialism, the medium in which it "emancipates," "naturalizes," and "unifies" the world for its own purposes?25 Before we can even pose this question, much less answer it, we need to take a closer look at what it means to think of landscape as a medium, a vast network of cultural codes, rather than as a specialized genre of painting. The Sacred Silent Language The charming landscape which I saw this morning, is indubitably made up of some twenty or thirty farms. Miller owns this field, Locke that, and Manning 14 W. J. T. Mitchell the woodland beyond. But none of them owns the landscape. There is a property in the horizon which no man has but he whose eye can integrate all the parts, that is, the poet. This is the best part of all these men's farms, yet to this their land-deeds give them no title. —Emerson, Nature (1836) I have been assuming throughout these pages that landscape is best understood as a medium of cultural expression, not a genre of painting or fine art. It is now time to explain exactly what this means. There certainly is a genre of painting known as landscape, defined very loosely by a certain emphasis on natural objects as subject matter. What we tend to forget, however, is that this "subject matter" is not simply raw material to be represented in paint but is always already a symbolic form in its own right. The familiar categories that divide the genre of landscape painting into subgenres—notions such as the Ideal, the Heroic, the Pastoral, the Beautiful, the Sublime, and the Picturesque—are all distinctions based, not in ways of putting paint on canvas, but in the kinds of objects and visual spaces that may be represented by paint.26 Landscape painting is best understood, then, not as the uniquely central medium that gives us access to ways of seeing landscape, but as a representation of something that is already a representation in its own right.27 Landscape may be represented by painting, drawing, or engraving; by photography, film, and theatrical scenery; by writing, speech, and presumably even music and other "sound images." Before all these secondary representations, however, landscape is itself a physical and multisensory medium (earth, stone, vegetation, water, sky, sound and silence, light and darkness, etc.) in which cultural meanings and values are encoded, whether they are put there by the physical transformation of a place in landscape gardening and architecture, or found in a place formed, as we say, "by nature." The simplest way to summarize this point is to note that it makes Kenneth Clark's title, Landscape into Art, quite redundant: landscape is already artifice in the moment of its beholding, long before it becomes the subject of pictorial representation. Landscape is a medium in the fullest sense of the word. It is a material "means" (to borrow Aristotle's terminology) like language or paint, embedded in a tradition of cultural signification and communication, a body of symbolic forms capable of being invoked and reshaped to express meanings and values. As a medium for expressing value, it has a scmiotic structure rather like that of money, functioning as a special sort of commodity that plays a unique symbolic role in the system of exchange-value. Like money, landscape is good for nothing as a use-value, while serving as a theoretically limitless symbol of value at some other level. At the most Imperial Landscape 15 basic, vulgar level, the value of landscape expresses itself in a specific price: the added cost of a beautiful view in real estate value; the price of a plane ticket to the Rockies, Hawaii, the Alps, or New Zealand. Landscape is a marketable commodity to be presented and re-presented in "packaged tours," an object to be purchased, consumed, and even brought home in the form of souvenirs such as postcards and photo albums. In its double role as commodity and potent cultural symbol, landscape is the object of fctishistic practices involving the limitless repetition of identical photographs taken on identical spots by tourists with interchangeable emotions. As a fetishized commodity, landscape is what Marx called a "social hieroglyph," an emblem of the social relations it conceals. At the same rime that it commands a specific price, landscape represents itself as "beyond price," a source of pure, inexhaustib e spiritual value. "Landscape," says Emerson, "has no owner," and the pure viewing of landscape for itself is spoiled by economic considerations: "you cannot freely admire a noble landscape, if laborers are digging in the field hard by."28 Raymond Williams notes that "a working country is hardly ever a landscape," and John Barrcll has shown the way laborers are kept in the "dark side" of English landscape to keep their work from spoiling the philosophical contemplation of natural beauty.29 "Landscape" must represent itself, then, as the antithesis of "land," as an "ideal estate" quite independent of "real estate," as a "poetic" property, in Emerson's phrase, rather than a material one. The land, real property, contains a limited quantity of wealth in minerals, vegetation, water, and dwelling space. Dig out all the gold in a mountainside, and its wealth is exhausted. But how many photographs, postcards, paintings, and awestruck "sightings" of the Grand Canyon will it take to exhaust its value as landscape? Could we fill up Grand Canyon with its representations? How do we exhaust the value of a medium like landscape? Landscape is a medium not only for expressing value but also for expressing meaning, for communication between persons—most radically, for communication between the Human and the non-Human. Landscape mediates the cultural and the natural, or "Man" and "Nature," as eigh-teenth-century theorists would say. It is not only a natural scene, and not just a representation of a natural scene, but a natural representation of a natural scene, a trace or icon of nature in nature itself, as if nature were imprinting and encoding its essential structures on our perceptual apparatus. Perhaps this is why we place a special value on landscapes with lakes or reflecting pools. The reflection exhibits Nature representing itself to itself, displaying an identity of the Real and the Imaginary that certifies the reality of our own images.30 The desire for this certificate of the Real is clearest in the rhetoric of 16 W. J. T. Mitchell scientific, topographical illustration, with its craving for pure objectivity and transparency and the suppression of aesthetic signs of "style" or "genre." But even the most highly formulaic, conventional, and stylized landscapes tend to represent themselves as "true" to some sort of nature, to universal structures of "Ideal" nature, or to codes that are "wired in" to the visual cortex and to deeply instinctual roots of visual pleasure associated with scopophilia, voyeurism, and the desire to see without being seen. In The Experience of Landscape, Jay Applcton connects landscape formulas to animal behavior and "habitat theory," specifically to the eye of a predator who scans the landscape as a strategic field, a network of prospects, refuges, and hazards.31 The standard picturesque landscape is especially pleasing to this eye because it typically places the observer in a protected, shaded spot (a "refuge"), with screens on either side to dart behind or to entice curiosity, and an opening to provide deep access at the center. Appleton's observer is Hobbes's Natural Man, hiding in the thicket to pounce on his prey or to avoid a predator. The picturesque structure of this observer's visual field is simply a foregrounding of the scene of "natural representation" itself, "framing" or putting it on a stage. It hardly matters whether the scene is picturesque in the narrow sense; even if the features are sublime, dangerous, and so forth, the frame is always there as the guarantee that it is only a picture, only picturesque, and the observer is safe in anodier place—outside the frame, behind the binoculars, die camera, or the eyeball, in the dark refuge of the skull. Appleton's ideal spectator of landscape, grounded in the visual field of violence (hunting, war, surveillance), certainly is a crucial figure in the aesthetics of the picturesque. The only problem is that Appleton believes this spectator is universal and "natural." But there are clearly other possibilities: die observer as woman, gatherer, scientist, poet, interpreter, or tourist. One could argue that they are never completely free from the subjectivity of (or subjection to) Appleton's observer, in the sense that the threat of violence (like the aesthetics of the sublime) tends to preempt all other forms of presentation and representation. Appleton's landscape acsdictic applies not just to the predator but to the unwilling prey as well. We might think of Appleton's "predatory" view of landscape, then, as one of the strategies by which certain conventions of landscape are forcibly naturalized. Nature and convention, as we have seen, are both differentiated and identified in the medium of landscape. We say "landscape is nature, not convention" in the same way we say "landscape is ideal, not real estate," and for the same reason—to erase the signs of our own constructive activity in die formation of landscape as meaning or value, to produce an art that conceals its own artifice, to imagine a representation Imperial Landscape 17 that "breaks through" representation into the realm of the nonhuman. That is how we manage to call landscape th-.i "natural medium" in the same breath that we admit that it is nothing but a bag of tricks, a bunch of conventions and stereotypes. Histories of hndscape, as we have seen, continually present it as breaking with convention, with language and textuality, for a natural view of nature, just at; diey present landscape as transcending property and labor. One influential account of the European origins of landscape locates it in the "free spaces" of medieval manuscript illumination, an "informal space left vacant by the script" in "the margins and bas-ie-pages of manuscripts" where the painter could improvise and escape from the demands of doctrinal, graphic, and illustrative subordination to "the severe lines of the Latin text" for a romp with nature and pure painting.12 This double semiotic structure of landscape—its simultaneous articulation and disarticulation of the difference between nature and convention—is thus the key element in the elaboration of its "history" as a Whiggish progress from ancient to modern, from Christian to secular, from the mixed, subordinate, and "impure" landscape to the "pure" landscape "seen for itself," from "convention" and "artifice" to the "real" and die "natural." These semiotic features of landscape, and the historical narratives they generate, are tailor-made for the discourse of imperialism, which conceives itself precisely (and simultaneously) as an expansion of landscape understood as an inevitable, progressive development in history, an expansion of "culture" and "civilization" into a "natural" space in a progress that is itself narrated as "natural." Empires move outward in space as a way of moving forward in time; the "prospect" that opens up is not just a spatial scene but a projected future of "development" and exploitation.33 And this movement is not confined to the external, foreign fields toward which empire directs itself; it is typically accompanied by a renewed interest in the re-presentation of the home landscape, the "nature" of the imperial center.34 The development of English landscape conventions in the eighteenth century illustrates diis double movement perfectly. At the same time as English art and taste arc moving outward to import new landscape conventions from Europe and China, it moves inward toward a reshaping and re-presentation of the native land. The Enclosure movement and the accompanying dispossession of the English peasantry are an internal colonization of the home country, its transformation from what Blake called "a green & pleasant land" into a landscape, an emblem of national and imperial identity. Pope's "Windsor Forest" is one such emblem, epitomizing British political and cultural sovereignty ("At once the Monarch's and the Muse's seats"), and its imperial destiny, figured in the "Oaks" that provide the material basis for British commercial and naval power: "While 18 W. J. T. Mitchell by our Oaks the precious loads are born, / And realms commanded which those Trees adorn" (lines 31-32). Decline and Fall If, indeed, the reader has never suspected that landscape-painting was anything but good, right, and healthy work, I should be sorry to put any doubt of its being so into his mind.... I should rather be glad, than otherwise, that he had formed some suspicion on this matter.... We have no right to assume, without a very accurate examination of it, that this change has been an ennobling one. The simple fact that we are, in some strange way, different from all f the great races that existed before us, cannot at once be received as the proof j of our own greatness. f —Ruskin, "On the Novelty of Landscape" |: The "realms" that proved most dramatically vulnerable to the "Oaks" of f British sea power at the height of the eighteenth- and early nineteenth-century landscape movement were the islands of the South Pacific and 'f the larger continental prize of Australia. Between the first voyage of Cap- / tain Cook in 1768 and the voyage of Darwin's Beagle in 1831, die British established unrivaled naval supremacy in die South Pacific and planted colonies that would develop into independent English-speaking nations. The ease of this conquest makes it of special interest for the understanding ; of landscape. Unlike the colonial landscapes of India, China, or the Middle East, the South Pacific had no ancient, urbanized, imperial civilizations or military establishments to resist colonization. Unlike Africa, it pre- L sented few land masses inaccessible to British "Oaks."35 Unlike North I America, it did not quickly develop its own independent pretensions to ;| be an imperial, metropolitan center.36 The scattered cultures of Polynesia i were seen, in Marshall Sahlin's phrase, as "islands of history," the last » refuges of prehistoric, preeivilizcd people in a "state of nature."37 The J South Pacific provided, therefore, a kind of tabula rasa for the fantasies j of European imperialism, a place where European landscape conventions could work themselves out virtually unimpeded by "native" resistance, >■ where die "naturalness" of those conventions could find itself confirmed by a real place understood to be in a state of nature. Bernard Smith's European Vision in the South Pacific documents this process in encyclopedic detail, noting the way that specific places were | quickly assimilated to the conventions of European landscape, with Tahiti ,-represented as an arcadian paradise in the style of Claude I.orrain and t New Zealand as a romantic wilderness modeled on Salvator Rosa, com- ,: plctc with Maori "banditti."38 Australia was a bit more difficult to codify, ':■ Imperial Landscape 19 not because of any native resistance (the aborigines were probably subjugated and erased from the landscape more quickly than any other people in die South Pacific), but because of the ambivalence in England's own sense of what it wanted to sec there—a fearsome, desolate prison for transported convicts, or an attractive pastoral prospect for colonial settlers.39 But Smith's account of the development of the South Pacific landscape suggests that ambivalence about the proper forms of representation, and about the "independence" or "otherness" of the colonized landscape, is constitutive of its perceived nature. Here is Smith's overview of the story his book will tell: The opening of the Pacific is ... to be numbered among those factors contributing to the triumph of romanticism and science in the nineteenth century world of values. Whilst it will be shown how the discovery of the Pacific contributed to the challenge to neoclassicism in several fields, more particular attention will be given to the impact of Pacific exploration upon the theory and practice of landscape painting and upon biological thought. For these two fields provide convenient and yet distinct grounds in which to observe how the world of the Pacific stimulated European thought concerning the world of nature as a whole; in the case of the former as the object of imitation and expression, in the case of the latter as an object of philosophical speculation.40 The ambivalence of European vision ("Romantic" versus "scientific," "neoclassicism" versus "biological thought," "imitation and expression" versus "philosophical speculation") is mediated by its absorption into a progressive Whig narrative that overcomes all contradictions in the conquering of the Pacific by science, reason, and naturalistic representation. The caicial moments in Smith's accounts of landscape painting are typically found in "fearless attempts to break with neo-classical formulas and to paint with a natural vision."41 Smith treats the Pacific as a spatial region that was there to be "opened," "discovered," and constructed as an object of scientific and artistic representation, one that reserves all "challenges" and historical, temporal movements for the internal unfolding of European thought, its overcoming of its own attachment to artifice and convention. The real subject is not the South Pacific but European imperial "vision," understood as a dialectical movement toward landscape understood as the naturalistic representation of nature. Empires have a way of coming to an end, leaving behind their landscapes as relics and ruins. Ruskin seems to have sensed this even as he celebrated the "novelty" of landscape, questioning whether "we have a legitimate subject of complacency" in producing a kind of painting (and its associated feelings) that reveals us as "different from all die great races that have existed before us."42 Kenneth Clark says that "landscape paint- 20 W. J. T. Mitchell ing, like all forms of art, was an act of faith" in a nineteenth-century religion of nature that seems impossible today;43 for Clark, abstract painting is the successor to landscape, a logical outgrowth of its antimimctic tendencies. Perhaps abstraction, the international and imperial style of the twentieth century, is best understood as earning out the task of landscape by other means.44 More likely, the "end" of landscape is just as mythical a notion as the "origins" and developmental logic we have been tracing. But there is no doubt that the classical and romantic genres of landscape painting evolved during the great age of European imperialism now seem exhausted, at least for the purposes of serious painting.45 Traditional eighteenth- and nineteenth-century landscape conventions are now part of the repertory of kitsch, endlessly reproduced in amateur painting, postcards, packaged tours, and prefabricated emotions. That doesn't mean that beautiful scenery has lost its capacity to move great numbers of people; on the contrary, more people now probably have an appreciation of scenic beauty, precisely because they are so estranged from it. Landscape is now more precious than ever—an endangered species that has to be protected from and by civilization, kept safe in museums, parks, and shrinking "wilderness areas." Like imperialism itself, landscape is an object of nostalgia in a postcolonial and postmodern era, reflecting a time when metropolitan cultures could imagine their destiny in an unbounded "prospect" of endless appropriation and conquest. As a conclusion to this essay, I would like to examine two imperial landscapes that exhibit in quite contrary ways diis "precious" and "endangered" condition in the rcarview mirror of a postcolonial understanding. The first is New Zealand, a land that is virtually synonymous with pristine natural beauty, a nation whose principal commodity is the presentation and representation of landscape; the second is the "Holy Land," the contested territories of Israel and Palestine. It is hard to imagine two landscapes more remote from one another, both in geographic location and in cultural/political significance. New Zealand is at the periphery of European imperialism, die last and remotest outpost of the British Empire, an unspoiled paradise where the nineteenth-century fantasies of ideal, picturesque, and romantic landscape would seem to be perfectly preserved. The Holy Land has been at the center of imperial struggle throughout its long history; its landscape is a palimpsest of scar tissue, a paradise that has been "despoiled" by conquering empires more often than any other region on earth. The juxtaposition of these two landscapes may help to suggest something about the range of possibilities in colonial landscape—the poles or antipodes between which the global features of imperial landscape might be mapped in (say) Africa, India, China, the Americas, and die South Pacific. More important than any global mapping, however, is the Imperial Landscape 21 possibility that a close reading of specific colonial landscapes may help us to see, not just the successful domination of a place by imperial representations, but the signs of resistance to empire from both within and without. Like all scenes framed in a rcarview mirror, these landscapes may be closer to us than they appear. Circumference and Center Columbus's voyage on the round rim of the world would lead, he thought, back to the rocks at its sacred center. —Stephen Greenblatt, Marvelous Possessions New Zealand would seem at first glance to be the site of least possible resistance to the conventions of European landscape representation. Its sublime "Southern Alps," its picturesque seacoasts, lakes, and river valleys, and its sheep-herding economy make it seem tailor-made for imposition of European versions of the pastoral. The fact that New Zealand was originally colonized by missionaries who rapidly converted the Maori inhabitants to Christianity redoubles its identity as a "pastoral" paradise.46 If Australia was imagined as a prison-scape for the incarceration of die British criminal class, New Zealand was thought of as a garden and a pasture in which the best elements of British society might grow into an ideal nation, bringing tiie savage inhabitants into a state of blessed harmony with this ideal nature. It's hardly surprising, then, that landscape painting has always been the dominant mode, in New Zealand art, and that this painting has consistently been bound up with questions of national identity. New Zealand represents itself as a nation of backpackers, mountain climbers, shepherds, and Sunday painters (a glance at any travel brochure will confirm this), a refuge from the problems of modern civilization, a nuclear-free English socialist utopia in the South Pacific. The hegemony of New Zealand landscape had, however, a contradiction built into it from the vety first. How could New Zealand present itself as a unique place widi its own national identity, while at the same time representing itself with conventions borrowed from European landscape representations? How could it reconcile its desire for difference with its equally powerful desire to be the same? An answer was suggested in the early eighties by Francis Pound, a New Zealand art critic who caused a storm of controversy by questioning the uniqueness and originality of New Zealand landscape painting. Pound shows that the history of this painting, like that of its European predecessors, has largely been told as the familiar story of the movement from convention to nature, from the Ideal to the Real, and that this story underwrites a progression from 22 W. J. T. Mitchell cultural colonialism and dependency to national independence. Pound exposes what he calls the "fallacy . . . that there is a 'real' New Zealand landscape with its 'real' qualities of light and atmosphere" and suggests that this naturalism is nothing but a "critical myth," a "fantasy of the truth" that was devised for the purpose of "inventing a country."4 Pound insults the desire for difference by continually discovering the same in New Zealand painting, showing that no new conventions of landscape were invented in New Zealand; on die contrary, New Zealand painters simply imported European conventions and absorbed the alien land into them. Landscape painting in New Zealand is thus a derivative, attenuated simulacrum of styles, techniques, and conventions invented elsewhere. A good example of this pictorial colonization is John Alexander Gilfil-lan's Native Council of War (1855), which inserts the "native" Maoris into the familiar landscape conventions of the Claudean picturesque (see fig. 1.1). However much we may admire the beauty and the technical skill of this painting, there can be no doubt that it is a throwback to an earlier style of landscape painting, not the discovery of a new style or a new reality. The clearest evidence of this fact is the placement of the cultural "others," the Maori war council, into die serpentine "tine of beauty" that had been understood, at least since Hogarth, as the iconic form of visual curiosity, of access to the varieties of visual experience. Pound attacks the nationalist/naturalist claims of traditional New Zealand art history with an internationalist discourse of cultural relativism and conventionalism. The result is a considerable refinement in the understanding of diat history and the visual conventions diat constitute it. But his replacement of naturalist fantasies with historical conventions raises a new problem that he is quick to acknowledge: All the aforesaid may arouse the objection that it is merely a sevenfold classification of eighteenth and nineteenth century landscape into pigeonholes—the holes labelled the God in Nature, the Ideal, the Sublime, the topographical, the picturesque, the sketch, and the Impressionist. It can be answered with the assertion that previously in the critical literature all New Zealand landscapes were stuffed into two pigeonholes: the true and the untrue to New Zealand. The present text's classifications have the advantage of offering concepts used at the time of the paintings, rather than those that merely answer the nationalistic concerns of critics a century later.48 Pound replaces the binary oppositions of a retrospective nationalist mytii with the historical categories appropriate to the self-understanding and self-representation of New Zealand landscape. The question is whether this move doesn't simply reinstate the categories of imperial landscape conventions without questioning their specific historical function. Imperial Landscape 23 1.1 J. A. Gilfillan, A Native Council of War (1853). Courtesy of the Hocken Library, Dunedin, New Zealand. Pound's analysis of the Gilfillan landscape, for instance, like most of his commentaries, tends to reduce the painting to an itemized list of its conventional elements with their appropriate emotional epithets—the picturesque, the sublime, the beautiful—and the sentiments appropriate to these conventions are simply recirculated. The discourse of imperial landscape is reinstated in die name of history, but at the expense of its historical function in the formation of a colonial and national identity. A historical, as distinct from a historicist, understanding of this sort of painting would, in my view, not simply retrieve their conventionality but explore die ideological use of their conventions in a specific place and time. Gilfillan, Pound tells us, "made precise pencil sketches from life" of all the figures in this design except two—the "Titianesque" woman and the seated man beside her, who serve as "lead in figures." These figures are inserted on die threshold of the painting, the transition space between 24 W. J. T. Mitchell observer and observed. They "sit in" for the European observer, reassuring us that the Maori see things as we do, while maintaining their difference. The key figure in mediating this difference is the bare-breasted woman, the Renaissance Venus who plays the role of eye-catcher, a titillating bit of soft-core colonial pornography, an emblem of native "nature" opening herself for easy access to the imperial gaze while her husband's back is turned (the husband, by contrast, covers his nakedness, holding himself in).49 These figures of access are the only "invented" elements in the painting, the only features that were not "drawn from life"; they are also its most conspicuously conventional and derivative feature, the element that declares most explicitly the fantastic sameness of colonial representations of difference. This idyllic absorption of the strange into the familiar comes to seem all the more fantastic when we come to it with the information that the painter's wife and three children had been killed by a Maori raiding party just eight years before this painting was finished in the relative safety of Australia. Gilfillan's painting allows (understandably perhaps) for no resistance to conventions of European landscape, except perhaps for the slight indication of compositional dissonance in the way the oval, canoe-shaped circle of the Maori war council cuts off the serpentine access route. By-contrast, Augustus Earle's Distant View of the Bay of Islands, in spite of a title that seems to announce nothing more than another picturesque scene, offers quite pointed, if subtle resistance to European conventions (see fig. 1.2). Francis Pound's commentary enumerates the specific conventions evoked by the painting: "Earle uses the traditional system of planes of shadow alternating with planes of light; and though the landscape itself has not allowed him the assistance of any convenient tree, he has managed to place an appropriate repoussoir [a picturesque side-screen] at the right," in the form of the carved figure.50 The problem with this reading is not just that it immediately reduces the painting to a familiar code and a conventional response ("the effect is of solemn splendor"). The real problem is that it doesn't push conventionalism far enough and draws back to an appeal to what nature—"the landscape itself"—allowed the painter to do. But one of the key principles of the picturesque tradition was that it allowed the painter to introduce a convenient tree (or to cut it down), in accordance with the demands of the convention. As William Gilpin puts it: "Though the painter has no right to add a magnificent castle, he may shovel the earth about him as he pleases . . . he may pull up a paling or throw down a cottage."51 Gilfillan's introduction of the Titianesque Venus illustrates precisely this inventive license. If Earle were simply following picturesque conventions, the "landscape itself" would have had nothing to say about the matter. And if the "un- Imperial Landscape 1.2 Augustus Earle, Distant View of the Bay of Islands, New Zealand (ca. 1827). Watercolor, 26 x 44.1 cm. Courtesy the Rex Nan Kivell Collection, National Library of Australia. couthly carved figure" at the right (to use Earie's words) is really to be seen as a stand-in for the picturesque side-screen or "lead-in" figure, it is a singularly awkward, ineffectual, and ironic one. It does not, like die traditional repoussoir, provide a dark refuge for the viewer to hide behind, nor does it provide a convenient stand-in for the beholder's gaze within the composition. On the contrary, it is a hazard, an emblem of an alien vision that stares back into die space of the beholder. The function of the figure in Maori culture is to stand guard over tabooed territory, to separate the sacred, forbidden landscape from the territory surveyed and traversed by the European traveler and his Maori companions.52 The caned figure may, like the lopped tree trunks that so often appear in the foregrounds of early New Zealand landscapes, allude to the traces and vestiges of the picturesque side-screen; the figure may "stand in the place of" the repoussoir, but it does so only to show that convention displaced by something else. What is this "something else"? It is certainly not "die landscape itself" or nature but another convention for organizing and perceiving the landscape, one that contends with and reshapes the convention that Earle carries as a picturesque traveler. That convention is the Maori experience 26 W. J. T. Mitchell Imperial Landscape 27 and representation of landscape symbolized by the carved figure, who | stands as rigidly erect and still as the halted European traveler, gazing on | a holy land. So far as I know, Pound is correct to say that "the Maori | did not paint landscape," but he is seriously mistaken in claiming that f "landscape, the pictorial attitude to the land, stopping still just to look at § it, is purely an imported convention."53 Earle's picture suggests a more | complicated situation. The Maori statue indicates at a minimum that f "stopping still just to look" at the land is so important to the Maoris that | they erect a statue to keep surveillance over a place. Nor is this surveillance I confined only to the carved figure. The Maori bearer on the left seems to | be hesitating as he walks, turning to the side to scan the tabooed territory, J while raising his war club slightly to ward off a potential threat. The f Maori warrior just ahead of the European traveler, moreover, seems to | be joining in the western gaze, looking out toward the opening clouds | and horizon. His musket, upright posture, and European garments sug- | gest that he is the Maori chief in this party and that he is able to make | the transition from Maori sense of taboo landscape to a sharing in the § European appreciation of "prospect" about as easily (and perhaps at as 1 great a cost) as he is able to replace a war club with a musket.54 | This intermingling of landscape conventions runs deepest, however, | not in the explicit iconographical signs but in the odd, somber composi- If tion and coloring. Most notable is the way the picturesque convention of gj the serpentine line from foreground to background is cut off in this pic- I ture and turned back on itself. A tiny reminder of the serpentine appears | in the middle distance, just to the left of the Maori chief, but it is only a ;J vestige or trace of the convention, not a fulfillment of it—in much the j same way the carved figure reminds us of the picturesque side-screen while | eliminating its function as refuge. In the place of the serpentine access J route, the composition deploys a crescent, canoe-shaped hollow, envel- I oping a procession of equally scaled figures across the shallow surface of * the painting. This effect, which is reminiscent of the treatment of figures ^ in a bas-relief, is heightened by the flattening of the perspective by alter- *J nating bands of light and dark monochrome wash across the painting. The effect is of an oval or circular procession, advancing up toward us on