James Cahill dies at 87; scholar of Chinese art

Visual, Verbal, and Global (?): Some Observations on Chinese Painting Studies James Cahill

(Note: The following is a somewhat shortened combination of two papers written for a double event organized by Professor Chi-sheng Kuo at the University of Maryland, November 13-14, 2005. The first, titled "Visual and Verbal Approaches in Chinese Painting Studies" was for a one-day symposium on the theme “Chinese Painting Studies in Postwar America”; the second was a “position paper” written in preparation for a public conversation held the next day with Professor James Elkins of the School of the Art Institute of Chicago, on the theme of “Chinese Painting Studies and the Project Is Art History Global?.” My argument is thus two-pronged, addressed both to the visual-verbal controversy and to Elkins’s proposals for “globalizing” art history. I want especially to thank Professor Hong Zaixin, who also presented a paper at the symposium, for advice and help throughout the writing of this paper.)

This first portion of my paper will be an argument for the importance of a visual approach, in conjunction with the verbal or documentary approach—certainly not instead of it--in Chinese painting studies. My choice of this topic was occasioned by a situation I know about principally through correspondence and conversations with Chinese colleagues, but also from my own observations: the verbal-visual controversy in art history circles in China’s academia today. And in that context, I must begin by acknowledging the technical and physical difficulties that our colleagues in China have had in setting up programs for studying their own tradition of art. I became acutely aware of those problems when I went around art academies and universities there in the 1970s-80s; and I understand from Chinese colleagues that in spite of some improvement, the problems are still serious: slide collections not widely available to academics; expensive reproduction books in limited availability; access to major painting collections still not as widespread as we are accustomed to enjoying in the U.S., and so forth. About all this I can only feel the deepest sympathy, and I join others in hoping that new technologies, such as relatively inexpensive visual databases of digitized images, will work to alter the balance toward more incorporation of the visual into the teaching and scholarship of Chinese painting in China.

Everything I write, then, in commenting on this controversy on the basis of my limited understanding of it and coming down heavily on the side of the visual, or at least of better balance, acknowledges this problem in China and is addressed only at deliberate, theoretically-argued choices to downplay the visual. My understanding of this controversy, and especially of the powerful anti-visual faction in Chinese academia, is based also on a paper by Professor Ding Ning of Beijing University which he has contributed to James Elkins’s long-term project The Art Seminar, and which I read through the kindness of Elkins. Professor Ding’s paper, titled “Verbal Above Visual: A Chinese Perspective,” begins by outlining the historical background to what he calls the “emphasis on the verbal” in Chinese scholarship, as opposed to the visual, which he says has been “effectively marginalized.” He goes on to describe the present situation in graduate programs in art history in China, without himself ever coming out in favor of their heavily verbal emphasis; in fact, he writes in his final paragraph as his own opinion that “The interpretation of images should be the primary and uppermost task of art history.” But he does not hold out much hope that this will come about soon or easily.

I myself was made aware of this widespread disinterest in the visual from the time I began lecturing and publishing in China: journal editors who came up after one of my lectures (illustrated with slides) and wanted to publish my text, but were unconcerned about the pictures, even when I pointed out that the lecture was virtually meaningless without them; translations of my writings published in China with a crucial painting missing from the illustrations, and nobody seeming to notice its absence, even though it is discussed at length in the text. I must add, however, so as not to exaggerate the real situation, that I have also read excellent recent writings on Chinese painting by Chinese specialists that make skillful and effective use of the visual materials, and worked with publishers there who are sensitive to the need for adequate illustrations. What I am arguing against is only a tendency, or proclivity, one well recognized by others as well as myself, and acknowledged by those engaged in it.

Studies of Chinese painting in the U.S. and Europe since around 1950 have undergone a remarkable development through the coming together of a number of traditions of studying it. From China, artist-connoisseurs such as Wang Chi-chien (C. C. Wang) and Chang Dai-chien, collectors such as Wan’go Weng, and scholars such as Wen Fong, Nelson Wu, Wai-kam Ho, and Chu-tsing Li, later Ju-hsi Chou, Mayching Kao, and Fu Shen, have brought with them their backgrounds in Chinese connoisseurship and scholarship. Osvald Siren, a Finn who lived and worked in Sweden, applied to Chinese painting what he had learned from Bernard Berenson and others. German art historians, notably Max Loehr but also Ludwig Bachhofer as his work was carried on by his students, especially Harrie Vanderstappen, applied to it their training in the methods of that great branch of art history. From Japan, scholars such as Shûjirô Shimada and Kiyohiko Munakata brought to the field some of the special strengths of the Japanese tradition of appreciating and writing about Chinese paintings. The British school of writing about art and aesthetics was drawn on productively by Michael Sullivan, John Hay, Roderick Whitfield, later Craig Clunas and others. American scholars, notably Laurence Sickman, Alexander Soper, Sherman Lee, and Richard Edwards came from the art history and sinology programs in our universities to make major contributions to Chinese art studies. (I leave out American specialists younger than myself, which means virtually all those extant—a clever tactical move, I think.) And all of us American scholars learned from these people, as well as from specialists who continued to work in China, Japan, and Europe; and all of us, whatever our backgrounds, learned from each other, trying to incorporate some of what we found most useful into our own work. The interaction between us and the approaches we represented, although often contentious, has always, I think, been beneficial; and the outcome, although still far short of real synthesis (an ideal that in any event can never be realized, nor should it be), can be recognized as a rich, multi-cultural and pluralistic product, less a method than a cluster of different methods, that transcends any of the lineages that went into it. It cannot by any means be dismissed simply as “Western art history,” since it is far more than that.

Now, while all these different strands that have fed into the present state of Chinese painting studies have followed different patterns of research in documentary materials, they have also brought with them different ways of looking at the paintings, reading the paintings, dealing with them visually. The important thing for my present argument is that they all have had, along with their diverse practices of research in written materials, a seriously-pursued visual component, and that this has been central to most of the achievements that have led up to our present, I believe fundamentally healthy, state.

The first grand get-together of Chinese painting specialists, a two-day “Palace Museum Exhibition Post-mortem Symposium” organized by myself and held at Asia House in New York in October 1962, was almost entirely visual: a number of us, including most of those I listed earlier, argued for two days about paintings, especially early landscapes, that had been in the great “Chinese Art Treasures” exhibition from the National Palace Museum in Taiwan, debating issues of dating, authorship, and authenticity. We did this almost entirely on the basis of visual observation, with slides of the paintings to which we could refer on the screen before us and our audience. This was, I feel, a good start for the field, and it was followed by a long period during which arguments of that kind continued, for better or worse, to be central to our interactions. When asked why this was so, we were inclined to reply that any other mode of dealing with the paintings, if pursued before we had established a relatively solid foundation in generally accepted datings and attributions for the key works, would be like building on sand. That is the answer I would give, for instance, to Svetlana Alpers in the later 1960s, after I had moved to Berkeley, when she wondered why we were still pursuing this hopelessly backward kind of art history. Historians of Western art, for whom this was no longer a problem to anything like the same degree, heirs as they were to several generations of hard-working Europeans (and a few Americans, such as Berenson) who had built for them the kind of comfortable foundation we were now trying to build for ourselves, found it hard to understand our preoccupation with establishing a more or less secure “corpus” within our materials of study. We are still far short of having accomplished that desirable goal; but we are a lot closer to it than we would have been if we had “ridden off madly in all directions,” so to speak, from the beginning.

Chinese specialists who lean toward the verbal approach also, of course, have their own system of dealing with authenticity problems, one that belongs within a distinctly Chinese tradition: it relies heavily on seals and inscriptions, colophons and catalog references. Without wanting to downplay the importance of that system, I would see it as a supplement, never a substitute, for judgments based on visual studies of the paintings. It has been used effectively by some Western scholars as well, notably Richard Barnhart, who combines it with high-level visual studies.

The next point I want to make, and make strongly, is that even those who have thought of themselves, and have been thought of by others, as primarily book-readers have managed to combine their strengths in that pursuit with valuable kinds of engagement with the paintings. Let me exemplify that statement with three names (and I hope that the only survivor among them will not object to being in the list): Wai-kam Ho, Shûjirô Shimada, and John Hay. John, to take his case first, is perhaps the most widely read among us in general aesthetic theory, and the one who has most brilliantly brought his broad and deep understanding of it into his writings. But when he has had occasion to write about paintings, he has done that also in unexpected and enlightening ways, as when he makes a reading of Huang Gongwang’s “Fuchun Mountains” scroll in terms of Chinese geomancy.

Wai-kam Ho was trained in the two best Chinese history programs of his time in China, at Lingnan and Yenjing Universities, before he came to this country around 1950 and took another degree in Chinese history and Asian art at Harvard. It would be superfluous and presumptuous to speak of the extraordinary strengths of his writings, which must be known to this audience. When I chaired a Chinese painting delegation for a month-long visit to China in 1977, Wai-kam was the only one among us who took no photographs of the hundreds of paintings we saw; instead, he studied them intensely while we were popping our flashguns, and wrote constantly in his notebook, noticing details that the rest of us missed. Wai-kam, while basically an historian, adapted with difficulty (he was notoriously unhurryable) but in the end successfully to his role as museum curator, exercising his visual faculties in helping Sherman Lee to choose and write about paintings for the Cleveland Museum of Art, and in organizing exhibitions, with the great “Century of Tung Ch’I-ch’ang” (1992) as the climax.

I will admit that Wai-kam was one of those I had in mind when I expressed in a 1976 article[1] some impatience with heavily text-based kinds of scholarship, which I referred to as “artless studies of art.” But that was a brief methodological complaint, balanced by things that Wai-kam found to complain about (with good reason) in my own writings, and did not hurt our relationship, which was over the years one of mutual respect.

Shûjirô Shimada was at his strongest when dealing with the kinds of paintings especially preserved and appreciated in Japan, the Song-Yuan period works that had mostly been brought there centuries earlier, with which he was closely familiar from long study; those were the main materials for much of his teaching at Princeton. But he also knew intimately Japanese collections of Ming-Qing paintings such as the former Kuwana Tetsujô collection, most of it later owned by Hashimoto Sueyoshi, and the great collection of Sumitomo Kan’ichi with its masterworks by Bada Shanren and Shitao. He took me to see these and other collections during my Fulbright year in Japan in 1954-5, at the same time that he was teaching a course at Kyoto University that involved a close reading of Guo Ruoxu’s Tuhua Jianwen Zhi, using Soper’s new translation as a supplementary text. Here were visual and verbal approaches separately exemplified at their highest; and his long article on the yipin or “untrammeled” style, which I later translated, offered an equally exemplary model of how they could be combined and made to interact. Like other Japanese scholars of his generation, Shimada wrote only about paintings he had studied in the originals--his book on Song-Yuan painting, like Yonezawa Yoshiho’s on Ming painting, used only works in Japanese collections. This was both a strength and a weakness, the weakness not to be overcome until, from the later 1950s on, Suzuki Kei, Kohara Hironobu, and others of the younger generation began to travel abroad and see collections there. But above all, Shimada’s practice represented a dependence on first-hand visual experience of the works, and expressed a deep trust in that experience.

I was surprised at first to see Shimada making careful sketch copies of paintings he was shown; later I realized that this was a common Japanese practice. I encountered it again in Nishimura Nangaku, a kanteika or “authenticator”--professional connoisseur--whom I was taken to meet while I was working on the early Nanga-school master Sakaki Hyakusen, since he had written about that artist. When I showed him photos of Hyakusen works in the U.S., he made sketch copies of them from the photos before he delivered his judgments on their authenticity. The same practice was common, of course, to the Kano-school painters who had performed the same function in earlier times, and who have left us albums and scrolls of shukuzu, reduced-size sketch-copies, that they had made from paintings they were shown; these are still valuable as records of works now mostly lost.

Behind this practice, which is peculiarly Japanese—I cannot recall seeing a Chinese connoisseur, or a German art historian, doing the same, although there may well be those who did it—behind this lies the Japanese concentration on the visual image over the execution of the work, a concentration that sets their mode of appreciation apart from the Chinese. In its extreme form it prefers the simple, isolated image that is central to both Zen and tea-ceremony taste, producing as it does a sharp, immediate visual experience that works like a metaphor for Zen enlightenment. (It was some adherent of that taste and doctrine who cut the fisherman in his boat from a Ma Yuan painting to suit it for hanging in the toko-no-ma of a tea-ceremony room, another who cut off the left end of Yujian’s painting of the “Waterfall on Mt. Lu” to simplify and focus the image, another who cut up what was presumably a Muqi handscroll to produce the famous “Six Persimmons,” “Hibiscus,” and “Chestnuts” pictures.) The Japanese have not until very recently, by contrast, paid much attention to those areas of Chinese painting that were, for Chinese literati connoisseurs, the highest peaks of the art: Huang Gongwang, Ni Zan, and other Yuan masters, Shen Zhou and Wen Zhengming in the Ming, Dong Qichang and the Four Wangs and their following in the later periods. Such works were not totally absent in Japan, but they were rare, and generally under-appreciated, except by certain Nanga artists who had become dimly aware, largely through reading, of this disparity between Japanese and Chinese connoisseurship. On the whole, the Japanese never quite understood or adopted the Chinese brushwork aesthetic until quite recent times; their appreciation of Chinese paintings came mostly through their responses to them as pictorial images, apprehensible by the eye and not to be intellectually theorized as they were in China.

The difference between this mode of appreciation and that of Chinese connoisseurs such as C. C. Wang and Chang Dai-chien, with both of whom I also spent a lot of time looking at paintings, was striking. It is true that Chang Dai-chien made small sketches of passages from well-known paintings from memory while we talked about those paintings during my first meeting with him in Kyoto in 1954; and it is true also that Chinese professional artists commonly made small sketch-copies, or fenben, from paintings they saw for later use. The difference lies in a certain divorcement of these Chinese artists and connoisseurs from those who practiced academic scholarship, a divorcement that may have prefigured, I now realize, the visual-verbal controversy in present-day China. The Chinese have, that is, their own practitioners of visual approaches, but they tend not to be the same who teach art history in universities.

The importance of the visual within Chinese painting studies in the U.S. is indicated also by the time, effort, and resources that have been devoted to building slide collections and photographic archives, and disseminating these materials to art history programs and museums. Unhappily, as I noted at the beginning, nothing comparable has been possible in China, and that failure is in some part responsible for the relative weakness of the visual direction in art history there. The ease and inexpensiveness, for us, of making 35 mm. color slides from Chinese paintings—their paper and silk surfaces, for one thing, did not reflect light from a flashgun as varnished oil paintings do, so the problem of glare was avoided—and the need for multiple shots of a single work (whole and details, leaves of an album or sections of a handscroll)--have made these a prime resource for our research and teaching, more so for some of us than the black-and-white photos on which earlier art historians depended.

As my students will recall, I began every seminar with a long “slide-show,” which might go on over several two-hour sessions: showing slides of paintings of the kinds we were to deal with and talking about them, trying to instill a common visual acquaintance with our materials in the seminar participants, along with a tentative organization of these materials and equally tentative observations about them. Often these opening formulations had changed a good deal by the end of the seminar. But they started us off on the right basis, I think: ideas and information were linked from the beginning with the paintings. I am not offering my own practice of this as any kind of model, but as a working method, it still seems to me a good one. In my Anhui school seminar, the one that led to the 1981 “Shadows of Mt. Huang” exhibition, four pairs of students explored issues raised by the paintings and the historical circumstances around them: the relationships of the paintings to Anhui pictorial printing; to the topography of the region, especially Huangshan; to developments in painting theory in that period; and to economic factors and patronage, the Huizhou merchants and their culture.[2]

Slide-making is, of course, rapidly becoming obsolete, replaced by digital imagery. This means that Chinese scholars and teachers can, if they choose, “leap-frog” the slide-making era entirely and adopt the new technologies, as their funding and facilities permit.

Another argument for the visual approach is this: if visual mastery of special areas of art is not taught, who will become the museum curators, the dealers, the auction specialists? Arnold Chang, who was in charge of Chinese painting for Sotheby’s auction house in New York during the great period of the 1970s and 80s, was enrolled in my heavily visual three-semester Chinese painting course and seminars at Berkeley, then studied traditional Chinese connoisseurship with C. C. Wang, wanting, as he saw it, to “get the best of both traditions.” Will aspirants to that kind of career in the future be able to enjoy similar training? What are the possibilities open to anyone who wants to pursue careers of these kinds in China, where auctions of paintings and calligraphy are booming?

Museum curators, of course, can come from other backgrounds than art history programs. One type of specialist found among our ranks has been the sinologue-in-the-museum: Wai-kam Ho was a good example, and Aschwin Lippe at the Metropolitan, whose doctoral dissertation was a translation of the Li Kan treatise on painting bamboo, but who went on to organize the first exhibition of a regional school of painting, his Nanking School exhibition held at China House in New York in 1955.

I have commented on Max Loehr’s strengths in the “position paper” for tomorrow’s conversation (see below), so will not do so here; in any case, the contributions of the German art historians to the visual approach scarcely need be pointed out, since they virtually invented it. Osvald Siren, at one time the best-known European specialist in Chinese art, should have been a central figure in visual studies of Chinese painting, since he was a pupil of Bernard Berenson and set out to do for Chinese painting what Berenson had done for Italian. But Siren never had a really good eye for painting, as Alexander Soper pointed out in a review of Siren’s seven-volume Chinese Painting, Leading Masters and Principles which had been published in 1956-58. I had worked with Siren on that in Stockholm for three months in early 1956, and knew too well his weaknesses, not only in judging authenticity (he had purchased two terrible “Xu Wei” fakes, along with others as bad, for the collection of the National Museum in Stockholm) but also in distinguishing individual and regional styles within Chinese painting, and in writing perceptively about them. Nor were those weaknesses compensated by sinological strengths: he scarcely read Chinese, and depended on others for translations and information from texts. He was essentially a compiler, a gatherer; I remember talking with Jan Fontein in Amsterdam on my way back from working for Siren, and saying that for our generation he was like the person who goes through the blackberry patch picking all the berries that are within easy reach, leaving it for us later people to scratch our hands getting the harder ones.

Vastly superior models among Western scholars were two of my colleagues at the University of California in Berkeley, Svetlana Alpers and Michael Baxandall: I attended many of their lectures, read their writings, talked with them at every opportunity, argued with them (especially with Svetlana), but also received invaluable guidance from them. Svetlana’s Art of Seeing, Michael’s Painting and Experience in 15^th Century Italy, along with their other writings: one could hardly do better in choosing models for joining visual studies of painting to historical research and theoretical concerns. Michael and I became friends, sharing a favorite pop-thriller writer (Ross Thomas), and attending each other’s lectures; I probably quote him more often than anyone else outside the Chinese art world, mostly because what he writes is always intelligible, and addresses productively the major problems, offering original and workable ways of negotiating some of them, along with admonitory rules for how not to do it. I mention this personal example simply to point out the obvious: that our visual treatments of Chinese paintings must be heavily dependent, for some part of their methodology, on the much-longer-established and more highly developed practice of our Western-art colleagues. If we oblige our students to take courses and seminars with those colleagues, as we should, that is the reason. And if they end up improperly applying patterns they learn from Western art history to their Chinese materials, the gain, I feel, still heavily outweighs the danger.

Another virtue of the visual approach is that it permits us to explore areas of Chinese painting that are ignored, or nearly so, in Chinese texts. All serious students of Chinese painting and its literature know that Chinese writings on painting are unmatched, at least until very recent times, in their volume and sophistication, But we know also that these writings are partial in both senses: they are partial to the literati or scholar-amateur artists and their works, and they write about only a part of Chinese painting, paying little or no attention to large areas of it that we may want to study today. A visual approach allows us to work in these areas where documentation is lacking, using the paintings themselves, the relationships between them, and their analyzable imagery as our data. Some of these areas will belong to the kinds that traditional Chinese connoisseurs consider low-class, but that should not deter us. I myself still believe that distinctions can and should be made between significant works of art and pictures that belong rather to studies of visual culture, but it is probably best not to assign any work or group of works to one or the other category too quickly or too firmly. A great many paintings exist, especially in foreign collections, that have been turned into “fakes” by dealers who added spurious signatures and early attributions to what were originally honest Ming-Qing works; often we can restore these to their proper art-historical positions, even determine their authorship, on the basis of their style and imagery. C. C. Wang was expert at this; Richard Barnhart has done it in some of his work, as have others, including myself. Others who have dealt with undocumented or poorly documented materials include Marsha Haufler in her studies of paintings by women or late-period Buddhist works, and Lothar Ledderose on the “Kings of Hell” series mostly preserved in Japan. Still others have done unexpected things within the canon on the basis of visual evidence, in ways unauthorized by the literature: Wen Fong’s “structural analysis” studies and his collaboration with Sherman Lee on the “Streams and Mountains Without End” monograph, Ellen Laing’s studies of the paintings of Qiu Ying, Jerome Silbergeld on Gong Xian and the Li-Guo School in the Yuan, Richard Vinograd’s writings on Yuan landscape and on portraiture, Jonathan Hay in his recent book on Shitao and his articles on Jin Nong and Luo Ping—all these, along with many others, have gone far beyond where written sources could carry them in illuminating and expanding important areas of Chinese painting history.

My own work in recent years has been mostly of this kind, a radical departure from my early period in which, for instance, I scanned reams of Song and later writings searching for quotes and clues that could be used in putting together a tentative account of literati painting theory (which made up the long opening section of my doctoral dissertation.) Now I scan auction catalogs and other sources of “low-class” and vernacular paintings of the kinds that make up the materials for my book Pictures for Use and Pleasure: Urban Studio Artists in High Qing China, and for my project of trying to define a body of Ming-Qing paintings done mainly for an audience of women, as well as for other projects similarly risky and largely unsupported by written evidence. Studies of intercultural exchanges in the arts—China and the West, Japan and China—of the kind that some of us have engaged in, and that have opened up in a healthy way in recent years, are likely to be similarly dependent on readings of the visual materials, and to go more or less undocumented, because of the touchiness of these issues and the reluctance of many writers, especially those verbally oriented, to recognize that the interchanges took place at all. The lack of textual evidence cannot be taken as a deterrent to our pursuing these studies—if the visual evidence is compelling, open-minded and open-eyed viewers will accept it, and the skepticism of others will in the end be inconsequential.

I learned long ago to shrug off—and I would strongly advise others to do the same--the frequently-heard criticisms of visually-oriented studies on the grounds that they lack any solid foundation. I remember well Noel Barnard, with whom I was working in the early 1960s on the Freer Gallery’s catalog of its Chinese ritual bronzes, deriding the stylistic approach of Max Loehr, which I was trying to apply in my catalog entries on style and chronology, as merely a matter of personal feeling, with no objective basis or value. Robert Bagley, a Loehr student, was met with similar reactions when he tried talking about bronze styles with Chinese specialists while working on the Metropolitan’s “Great Bronze Age of China” exhibition of 1980. My own writings on painting have frequently encountered similar responses. The typical charge is, “He doesn’t have any real evidence,” by which the speaker or writer means, of course, written or documentary evidence. One feels frustrated, typically, because a visual argument that is entirely convincing, even blindingly obvious, to anyone open to reading visual data will make no impression on those who are not; one can only point, saying “Look! Look!” and collapse into despair when they don’t look, or won’t. And as often as not, the one who complains about “lack of evidence” knows perfectly well that because the aspect or area of painting under consideration lies outside the limited range of what the traditional Chinese critics and theorists felt to be worthy of their attention, there isn’t going to be any written evidence for it. To insist, then, on documentation as the only legitimate basis for studying that area of Chinese painting is in effect to rule out serious consideration of it—which is exactly the underlying purpose and message of such criticism: Stay within the boundaries that we and our predecessors have drawn for you! That implicit admonition should be taken as a strong motivation for moving even more determinedly into the would-be forbidden territories, opening them up for further investigation, breaking the taboos wherever we find them.

Other than that, there is nothing I know of that we can do in the face of determined mistrust of the visual except to go on teaching our students what we ourselves know to be true about it, and continuing to practice visual studies of Chinese painting (along with textual and theoretical) as responsibly and convincingly as we can, both to provide useful models for others who may wish to do it, and simply to keep the practice going.

I will end this section by stating several things that I hope will happen before too long, as they will need to if I am to have the pleasure of watching them happen. First, I hope that the visual-verbal controversy can be divorced, in the thinking and arguing of both Chinese and foreign scholars, from the issue of foreign vs. indigenous modes of art history; they need not be linked, and should not be. Our discussions could then proceed with less danger of touching uncomfortably on cultural sensitivities. Second, I hope that more of our colleagues in China will recognize that Chinese painting studies as they have come to be practiced in the U.S. and Europe, and increasingly in other parts of the world as well, are not by any means simply “Western,” but embody methods and insights brought by Chinese as well as Japanese and Western scholars who have been engaged in them. For those in China now to adopt from these studies what seems useful to them, of their own volition, should thus carry no onus of betraying the indigenous tradition. Thirdly, I hope that as digital imagery and other new technologies encourage Chinese scholars to move toward a better balance of visual and verbal in their work and their teaching, they will develop distinctively Chinese ways of looking, of engaging visually with the paintings, drawing on the great Chinese tradition of visual study and connoisseurship of individual works. (I certainly do not mean to imply that this kind of balance and synthesis is not being accomplished at all today, but only that it is not yet widespread enough to constitute a general collective practice.) By combining this Chinese mode of visuality, in which the reading of seals and inscriptions becomes a part of the visual experience as one appreciates their design and calligraphic quality as one reads their texts, with attention as well to brushwork and other aspects of style and to the painting as a picture, adding to these the Chinese specialists’ unmatched mastery of the documentary sources, they will bring into being a practice that can be considered a truly Chinese history of Chinese painting. The interaction between that and the present multicultural one practiced outside China will, I think, be healthy and highly productive.

Some time in 1991 I began a correspondence with someone I had not met named James Elkins, who had sent me an essay he had written titled “Chinese Painting as Object Lesson.” I read it with great interest, having had thoughts in a similar direction myself, and sent him a very positive response, along with some suggestions for minor improvements. His essay began:

“Chinese landscape painting can be an ‘object lesson,’ that is, an analogy for understanding the course of Western painting from antiquity to postmodernism and beyond. It is possible, I will suggest, to make a reading of the Chinese tradition, and specifically of its developing sense of its own history, that runs parallel to essential developments in Western concepts of the history of painting.”

I myself still believe this contention to be true, and well stated. Its author, however, has in a complicated way pulled away from it without really renouncing it, After his essay in its original form had been rejected by several journals (for what I believe to be bad reasons), he himself rewrote it into a new essay titled “Chinese Landscape Painting as Western Art History.” This also remains unpublished in English, but has been published in Chinese translation.[3] In this he takes a very different direction, arguing instead that any history of Chinese painting written today must by necessity follow the patterns of our familiar Western art history, so that (if I understand correctly the implications of his argument) what he saw before as “object lesson” and “analogy” are no more than products of the (Western) way in which the art-historical account is written. This argument has led, in turn, to his present concern with the possibility of a “global art history” or “world art history” within which writings about the art of non-Western cultures might escape this subjection to European modes of thinking and writing about art, and not have imposed on them the narratives of Western art, such as those centered on space representation and illusionism (the latter associated especially with the writings of Ernst Gombrich.) I do not mean now to engage myself in this problem as such, but only to respond to two of his proposals for ways this “global art history” might be achieved, as Elkins sets them forth in his review of David Summers’s book Real Spaces.[4]

I must insert here a proper disclosure: I am not myself widely read in aesthetics or philosophy more generally, or nearly so widely in the general literature of art history as Elkins, who is an art-historical and art-theoretical polymath and omnivore. I have an uncomfortable feeling that in my conversation with him, I will sound like someone responding with trivialities to matters of serious import. I can only say that my arguments as they will be made below, however naïve they may sound and however weak in theoretical grounding, are not trivial for me, but represent deeply-held convictions formed over quite a few years of teaching and writing. Anyone who feels that an untheorized art history is not worth practicing will be frustrated by what follows. Also, my pairing of “teaching and writing” represents accurately my view on that matter: the two practices have scarcely been separate in my work. So references to art history throughout what I write below are to both pedagogy and published scholarship: these are only, for me, different ways of reaching an audience, a readership, a body of students.

In the final pages of his review of Summers’s Real Spaces, Elkins offers five possibilities for responding to the problem of moving toward a global art history, from the most “intellectually conservative” to the most radical. He sees Real Spaces as midway along the scale. I myself would, I am afraid, have to be located still further back, either still committed to his first, “Art history can remain essentially unchanged as it moves into world art” (he finds this the “most potentially destructive of the coherence and interest of art history”) or groping into his second, “Art history can redefine and adjust its working concepts to better fit non-Western art.” I would suggest that good art historians working in non-Western fields do this second without necessarily formulating it as an objective, and without separating themselves clearly from the “potentially destructive” first option.

Elkins’s third option, “Art history can go in search of indigenous critical concepts”—and, as he elaborates with examples, adopt the native terms for these into its vocabulary--is attractive in theory but, I think, impractical if carried much beyond present practice, to the point where it would constitute a new mode of art-historical writing. Is each foreign writer on a non-Western artistic tradition to choose and employ a set of native terms for what he or she sees as the key concepts, and expect the reader to learn them all—along with the different set of terms used by each other foreign writer? I am certainly not opposed to introducing the native words for a limited number of ideas and qualities that are unfamiliar to foreign readers, and explaining them—I could make a list of the ones I myself introduced in early writings on literati painting theory, and others have done the same. But we mostly explain the indigenous concepts using our own vocabulary, pointing out how empty or void are not exact equivalents for the Chinese kong (Elkins’s example) but continuing to employ what we take to be the closest equivalents. We explain the non-Western concepts as best we can—in my case, concepts derived mostly from reading Chinese texts, or from colleagues’ readings of them--and try to take account of them in our own writings. But we can never assume simply that these provide a “right” way of interpreting the foreign works of art, as opposed to our own “wrong” way. (Later on I will give reasons for saying that.)

Elkins’s fourth option, “Art history can attempt to avoid western interpretive strategies,” also sounds attractive but also breaks down, I think, when we consider how it would be applied. He writes that Professor Cao Yiqiang of the China Academy of Fine Art in Hangzhou is “interested in adopting elements of a ninth-century Chinese art historical text by Zhang Yanyuan, called Record of the Famous Painters of All the Dynasties,” in constructing an art history based in the Chinese tradition. “Adopting elements” is a fine but limited objective; going much further would, I think, be like contemporary specialists in Italian painting basing their inquiries only on those issues that concerned Vasari. It might not, on the other hand, be so radical a move as Elkins believes. He returns to Zhang Yanyuan’s book a few pages later to remark that what makes it “so different from contemporary art history is its author’s [Confucian] insistence that painting promotes filial piety and the health of the community.” But those sentiments are largely limited to the introductory section of Zhang’s book, which, as is common in introductions to Chinese books, proclaims pieties that are left behind once the author turns to his main matter. (My colleague Cyril Birch, a Chinese literature specialist, once advised one of my students about these prefaces that she should “take them very seriously but don’t believe them.”) Once Zhang begins the substantive discussions that make up the main body of his book (along with treatments of individual artists), he is as absorbed as we cultural outsiders might be in such matters as the styles of great and lesser masters, the distinguishing of schools and artistic lineages, standards of quality, and formulating a kind of “narrative” for the early periods. Elkins contrasts Zhang’s purported “Confucian purpose” with “current scholarship on Chinese painting, with its emphasis on politics, identity, and patronage.” But in fact we can learn quite a lot from Zhang’s book about politics and patronage, if not identity, as factors in the context of painting up to his time, and not much about Confucianism. Cao Yiqiang himself writes that Zhang Yanyuan’s book is “strikingly similar” to Vasari’s, and notes that Zhang “surveyed this development [of early painting] from the point of view of a gradual improvement in the representation of natural life—he claimed that ‘a better likeness was obtained in later portraits,’ for instance. . .”[5] (Zhang Yanyuan could, as I cannot, adopt that approach without being accused of imposing the approach of Gombrich onto Chinese painting.)

The Chinese literature on painting is extensive, highly sophisticated, and in considerable part devoted to concerns that match well with our own. It would seem all the more self-evident, then, that there would be little loss and much to be gained if, in our studies of Chinese painting, we were to follow Elkins’s fourth option, “attempt to avoid Western interpretive strategies,” and embrace the assumptions and attitudes that underlie Chinese writings about it. And yet I have come to believe, after some decades of working in the field and being actively engaged in the directions it has taken, both here and in China, that there are several compelling reasons why that is just what we should not do.

The first and basic reason is that the extant Chinese writings about painting, which must necessarily be our guides and sources for whatever understanding we can reach of traditional indigenous approaches, can never be accepted simply as telling us what “the Chinese” thought and believed about the paintings. For one thing, even when the artist and the writer are the same person (as in the case of Dong Qichang), the exigencies of writing, the pressures on the writer to take positions in accord with factors in his time, place, social class, etc., were strong, even constituting determinants, which might be quite different from the factors operating on the artist as he painted. (Baxandall remarks that “Even [the artist’s or maker’s] own description of his own state of mind . . .[can] have very limited authority for an account of intention of the object; they are matched with the relation between the object and its circumstances, and retouched or obliquely deployed or even discounted if they are inconsistent with it.“)[6] Painting and writing about painting in China, that is, often go in quite different directions. I have even come to believe, although I would be hard put to argue it on a theoretical ground without falling into the pitfalls that beset anyone making this kind of argument, that many of the most interesting and significant developments in Chinese painting, the complex artistic stratagems used by painters that can make their works so fresh and absorbing, go quite unrecognized and undiscussed in Chinese writings of their time and later. This is true despite the amazing richness of the Chinese literature of art and the frequently sharp observations made by its writers.

A major reason for the gap between painting and writing about it is that the prestigious writers were, virtually by definition, members of the literati class, and so were committed, especially from the Yuan period onward, to its particular viewpoint and special system of values. Their writings are imbued with the doctrines of literati or scholar-amateur painting, doctrines that were formulated in some part to support the practice of the literati artists, to deflect attention from their weaknesses and proclaim their strengths as defining the loftiest criteria of value.[7] I have observed elsewhere that at a time when so little else of the old, self-serving rhetoric of elites has been allowed to stand, this one has enjoyed a surprising tenacity. And I have been working, along with others, to recognize and try to reconstruct, necessarily in a very limited way, the large areas of Chinese painting that were deliberately excluded from the literati writers’ account of it, and so have stood a poor chance of being preserved. Trying to understand, and give some voice to, the preferences and attitudes of the silent non-literati majority of consumers of Chinese paintings, the ones who did not write the books but acquired and enjoyed most of the pictures, is in my belief a legitimate pursuit. It can, of course, be realized only very tentatively and imperfectly, but the same is true of many other legitimate pursuits.

Those who advocate a narrow dedication to “the Chinese tradition” in studying Chinese painting—and who, as we learn from the paper by Professor Ding Ning cited earlier, make up the strongest faction within the present-day practice of art history in China’s academia—seem to assume that the ways in which “the Chinese” understood and appreciated the paintings can be ascertained by reading the surviving literature about it. But that is only partly true, because, as noted earlier, that literature is, with a few exceptions, heavily partial in both senses: it applies only to a part of its assumed subject, Chinese painting, and it is thoroughly partial, i.e. biased, toward that part, and dismissive of the rest. Everyone knows the example of Chan or Zen painting, which, because it was not valued by Chinese critics and collectors, would be virtually lost to us if it were not for the preservation of a large body of it, including some masterworks, in Japan. Other big and important areas of Chinese painting have not fared even that well, and have to be put together from surviving, mostly misidentified scraps.

If, then, we were to move to a mode of studying Chinese painting dependent on Chinese writings on the subject, most of the advances we have made over the past half-century in working for a better balance and looking into neglected areas of Chinese painting would be sacrificed, and some of the most promising directions that Chinese painting studies have taken would be blocked. I tried to identify some of these neglected areas in a 1997 lecture delivered at Yale, the Hume lecture, titled “Toward a Remapping of Chinese Painting,” and much of my own work in recent years has been on the undocumented areas. I will return to this issue in a moment; I want first to point out another basic flaw in the idea that surviving Chinese texts on painting can be accepted as conveying to us fully the thinking of artists and critics of the time the paintings were done.

The writings of Dong Qichang and other major Ming-Qing and earlier critics, theorists, and colophon writers are based on their wide acquaintance with major paintings--many hours or days of studying them, traveling widely to see collections, developing a level of connoisseurship which even when it may be flawed was deep and extensive. Their writings were done in relation to this, and based on this—a kind of visual archive in memory, which allowed them to make judgments of works they saw by comparing them with the large data-banks stored in their heads, and to write about different schools and artists and styles by drawing on this. They wrote often about particular paintings, and even when they didn’t, their writings were charged with their close visual knowledge of a great many individual paintings. And they assumed a comparably broad visual experience of major artists and paintings in their readers—who, if they were reading the colophons at all, must be members of that small elite who had access to old and original works--and they made references to them on that assumption.

The problem is that in that pre-photography age there was no way for them to convey their deep visual engagement with paintings in their writings. Woodblock reproductions were of virtually no use, since they could not transmit the aspects of the paintings that mattered most to them. Copying by hand was an option, as in the well-known Xiaozhong Xianda or “Great Revealed in the Small” album, reduced-size copies of early masterworks owned by Wang Shimin with facing inscriptions by Dong Qichang; but the limitations of copies were also severe and well recognized. So the writings of these people come down to us more or less bare: imbued with, but physically lacking, the visual element—only as texts. And later people, through the Qing dynasty (when the great early paintings were mostly absorbed into the imperial collection, and accessible only to a very few in the court) and down to recent times, have had little or no opportunity to see and study major early works. In the 1930s-40s and later a few collectors and connoisseurs and museum people, such as the late C. C. Wang and Xu Bangda, have been able to see more; but academics mostly haven’t, and opportunities for students are even fewer. So they get a false and vastly reduced sense of what Dong Qichang and the others really knew and felt and believed. And they sometimes write as though staying within the realm of words-- theories and arguments and colophons, divorced from real visual engagement with the paintings--were adequate and even desirable, since (they argue) it continues a native Chinese tradition. But it continues it only in a greatly diminished form, robbed of what it was really based on in the visual experience of the writers they study.

For us outside China, then, to embrace the “indigenous Chinese” tradition of art history, as we would necessarily have to derive it from these texts, would oblige us to come down heavily on what I firmly believe would be, for us, the wrong side in the verbal-visual controversy discussed earlier.

Moreover, as I pointed out in the first section, what is dismissed by some academics in China as a “Western” or “foreign” way of studying Chinese painting is really the product of some sixty years of richly multicultural interaction between specialists who came from a number of traditions, Chinese and Japanese as well as German, English, and others. That the interaction took place largely in the U.S. has more to do with the economic and technological advantages we have enjoyed, along with the immigration of so many major art historians and collectors to the U.S. during and after the Second World War and the growth of major collections here, than with any factors particular to our own culture. To partake of the achievements of this very fruitful interaction can only augment and enrich, I believe, without in any way betraying, the more exclusively Chinese tradition of scholarship. Setting up the two as somehow irreconcilable alternative choices seems to me in itself a falsification of where we stand. (To write that is, of course, to contradict directly James Elkins’s contention that Chinese art history can only be Western art history, an argument that after a lot of thinking I still cannot entirely comprehend and certainly cannot follow.)

Now, on to my reasons for believing that any mode of art-historical teaching and writing that does not involve close readings and analyses of the works of art themselves is short-changing its readers and students. In arguing this, I have sometimes used analogies: one could, in principle, study and teach poetry without reading any poems, much less attempting close readings of poems, but that would be a very impoverished and ineffective way of teaching poetry. No one can reach any adequate understanding of, for instance, uses of metaphor or allusion in poetry without being provided with good examples of their usage in particular poems, just as (I once argued) any full and balanced understanding of Dong Qichang’s advocacy of fang or creative imitation must be based on a visual study of how it works in particular paintings.[8] The same is true of music: we can read articles or hear lectures on composers and compositions and historical developments, and those are by no means without value; but teaching music without playing musical examples for the students to hear would deprive one’s teaching of what should be its heart. (I have also pointed out, however, that we historians of the visual arts have a great advantage over teachers of poetry or music: where they must discuss the poem and then read it, or talk about the composition and then play it (in whichever order), we can enjoy a perfect three-cornered simultaneity: the painting, perhaps with details, on the screen, ourselves talking, and the students or lecture audience listening to us and gazing at the painting, all at once. This is an advantage we should exploit as fully as we can.) In any case, what matters is that the learner be afforded immediate experiences of the works themselves, whether the experiences be visual, aural, or literary; reading or hearing about the works can never suffice.

The objection will be raised: why need any teacher’s voice be there at all? Isn’t it better for the student/viewer to have a fresh, unmediated experience of the work? And that objection would lead to another proposal for avoiding a “visual art history,” a proposal that I will introduce only to try immediately to shoot it down, since it is another that sounds plausible in theory but has fatal weaknesses when one thinks of how it might be put into practice. The proposal is this: why not adopt the Chinese verbal mode of teaching, and let the students and readers, equipped with the knowledge they have gained, go on to seek out the paintings and look at them on their own? Everyone who has eyes can look at paintings; why does anyone need to be “taught” to do it?

The answer to that lies in the paradox that Michael Baxandall somewhere writes about as the “man in the bus” situation: the man is excitedly pointing out to his fellow passengers what they can see outside with their own eyes. What, then, is his function, if any? When we teachers stand before a painting (or a slide image of one) and talk about it, are we doing no more than that? But Baxandall, and any good art historian, knows that ideally we are doing much more: we are conveying a reading of the painting, perhaps some kind of analysis of it, that opens new areas of perception in the viewer-listener’s mind.

This brings me to a function of art-historical lecturing and writing that is scarcely taken into account in the Elkins discussions, as I read them, or more generally in theoretical accounts of art history, but that has been centrally important to my writing and teaching and the reception to it. I recall a student in one of Max Loehr’s classes on Chinese painting asking me, partway through the semester, “What’s so good about this? He just stands up there and describes the painting, things we can see as well as he can.” In fact, what Max Loehr did over the semester was implant in us a visual understanding of the course of Song and earlier painting, especially landscape, its issues and stylistic developments and structures of relationships as he saw them, that has been the basis for my own accounts of it ever since, different as those have been from his. The way he taught the ritual bronze vessels of Shang-Zhou China, never neglecting archaeological or epigraphical evidence but laying out also a convincing morphology of décor styles and shapes, established in our minds a framework within which the individual object, seen as a “move” within a Kublerian series,[9] might be intelligible and deeply satisfying, in a way that it could never be without that context. It was not that he fitted the styles into some pre-existing Wolfflinian or other pattern or “narrative,” but that the study of morphology of forms in related series in European art had alerted him to possibilities that non-art-historians did not recognize: for example, that an abstract pattern can metamorphose over time into an image—a pair of circles, for instance, into the eyes of an animal mask. And after outlining the earliest development of Chinese ritual bronze décor styles and vessel shapes on this pattern, Loehr was proven to be basically right, in a way independent of cultural traditions, when later excavations of pre-Anyang sites largely corroborated his proposed sequence of styles. Three other scholars, the pre-eminent bronze specialists in China, Japan, and Europe (as was recalled in a tribute to Loehr’s achievement by Alexander Soper),[10] had all got it backwards by assuming that the series must have begun with the representational image, which could then “dissolve” into an abstract pattern. This was not, then, a matter of the Western art historian imposing the patterns he knew onto the foreign materials, but of his having been sensitized, through his training, to possibilities to which those without that background were blind.

Loehr himself made this point strikingly at the beginning of the first seminar he gave in Ann Arbor, which was on the Shang-Zhou bronzes. He put a pile of unlabeled photographs of these on the table and invited us to arrange them into some kind of sensible order. As the only one in the seminar who had already been exposed to the bronzes, I stayed out. All the others were graduate students with backgrounds in other fields. By the end of about an hour, they had sorted the photos into an order that was a not-bad approximation of what we knew from other evidence to be the historical sequence of styles and shapes. Loehr had, of course, chosen examples that were susceptible to this kind of ordering, but his point was nonetheless made.

The writings of Bachhofer, Loehr, Soper, and others on the early periods of Chinese landscape painting drew in this way on visual and conceptual skills that had been developed in part through their study of other traditions of pictorial art, a consideration that in no way invalidates their work. Later studies making use of new archaeological finds and other materials have refined, in part corrected, but not replaced the narratives they laid out. None of them, on the other hand, was able to deal effectively with post-Song painting, to which I have recently applied the term “post-historical”;[11] and it is there that the Chinese narratives, with their emphases on local schools or movements, “imitating” old masters, the aims and doctrines of literati painting, the brushwork or “hand” of the individual master, and other factors independent of any Gombrichian “pursuit of likeness,” seem to be the suitable referents for an art-historical account. They are in fact the referents on which I have mainly tried to base my own accounts of the later periods of Chinese painting. But these, in turn, are mostly unsuited for dealing with painting of the early periods, where traditional Chinese connoisseurs are at their weakest.

Elkins sets up a bad model for cross-cultural work in art history, one in which foreign patterns of constructing “narratives” are imposed onto the indigenous tradition, and assumes that because all attempts must follow that model, the whole enterprise is contaminated. I would rather believe that there can be good and less good, appropriate and less appropriate, ways of applying an outsider’s understanding to a foreign body of art, and that we must make our choices as perceptively and honestly as we can, trying to avoid a priori assumptions about how this or that choice will distort our subject.

Lecturing and writing of the kind I advocate can open new circuits, so to speak, in the minds of its listeners/readers, like installing software that allows them to read and understand a new set of texts that were inaccessible or meaningless before (I am not a computer person and the analogy may be faulty), equipping them to respond in a sensitized and cultivated way to other, related works of art. I would argue--and the idea is a common one, virtually a truism--that much of the value of a work, as a quality of the experience of it by a viewer, derives from its relationship to others made before and after. This is especially true of works by artists who belong to a self-conscious artistic tradition (as Elkins recognized the Chinese and European traditions of painting to be) and who are taking part, whether or not consciously, in vast games—not divorced from their lives in the real world, like the players of Hesse’s Glasperlenspiel (the “Glass Bead Game”), but deeply engaged not only with the past of their culture but also with their own time and place and society, within which their work is created and experienced. And the place of the work within that context, from which a major part of its value derives, will not be perceived fully unless the mind of the reader/viewer is prepared by lots of previous experience of related works, and lots of thinking—his own or a teacher’s, ideally both--about the relationships between these: What are the really original creations (Kubler’s “prime objects”)? How do others derive from them? What are truly successful and satisfying solutions to the formal and aesthetic “problems” that obtain within this series? Of course, other contexts than the visual can also be fed into this mind-conditioning process as well, and should be: contextual matters, for instance, of the kind Baxandall writes about in his “inferential art history.”

A good teacher, who has arrived at a clear and convincing set of formulations of all these and conveys them effectively to her or his students, can take the students a long way in this process. The student does not, of course, accept the teacher’s account as the final word, as I did not accept Loehr’s; but he or she has a model and a tentative structure to accept, modify, or reject—any of these being better than beginning “cold.”

And that is what is wrong with the idea that the student can simply come to the work of art visually unprepared and experience fully its expressive impact and aesthetic value. I have met people who claim they can recognize quality in any work of art of any time and tradition; this, again, is not entirely untrue, especially if they have had enough previous experience with other kinds of art, but is only a limited truth. Examples I have used of sensitive and intelligent people “coming cold” at an unfamiliar body of art and being unable to make qualitative and other distinctions within it include the Chinese connoisseur Wang Chi-chien going around the European painting galleries of the National Gallery of Art with me in the early 1960s and remarking that the paintings all looked more or less alike and “have no brushwork,” and John Canaday responding in a New York Times review to my “Fantastics and Eccentrics in Chinese Painting” exhibition of 1967 by writing that the paintings, beautiful as they were, did not look so different in his eyes from all the Ming-Qing paintings he had seen before—what was so “eccentric” about them? I have cited also a reviewer of a concert of Indian music who noted that when the musicians played an erotic raga the Indians in the audience were unable to sit still, while the non-Indians sat “like bumps on logs.” Music is an especially telling case: someone who has not listened seriously to classical European music, for instance, cannot tell Mozart from Beethoven, much less Mozart from Haydn. Discernment is not, of course, the whole content of the appreciation and enjoyment of art, but it is a major component of it, virtually a requisite for it.

Art history, then, is in some part a matter of implanting the right set of visual referents in the viewer/listener’s mind so that the work will be perceived in relation to those, and pointing out the relationships as you have come to understand them. For reasons I have never understood, theorists are inclined to leave out this function when discussing art history; perhaps they feel it belongs properly to pedagogy, or to criticism, and so is beneath their notice. But the truly successful art historian (or literary historian, or music historian), the one whose lectures and writings are exciting to listeners and readers, the one who gets letters saying “You have changed my life,” is the one who understands the importance of this process, has built up in his or her mind the formulations and structures of relationships for the bodies of art he or she teaches, and can convey these to others by instilling some form of them, with enough time and slides, in their minds. Individual teachers, of course, do this better or less well; all of them, I think, should at least attempt it, even when their principal strengths are elsewhere. Many will simply adopt existing “narratives,” whether the traditional Chinese one or some adaptation of one they learned from their teacher, or one put together from reading attempts by others to formulate new “histories.” Many, perhaps most, will modify existing accounts to put more emphasis on their own strengths and interests. What matters is that the students will have a visual grounding into which they can fit their experiences with paintings, and which they can in turn adapt and use if they themselves become teachers. Teachers of the history of Chinese painting have fewer well-established and widely accepted “narratives” to choose among than their colleagues in Western art studies, a circumstance that makes their choices all the more interesting and important.

To reinforce my argument, let me quote a passage by a literary historian named Geoffrey Galt Harpham that I read recently and found myself resonating with.[12] After writing about the consequences of “the anti-humanistic spirit that, in varying degrees, animated all [the theorists who have dominated literary studies in recent decades],” with their “determination to undermine concepts of human creativity, human freedom, and the human capacity for self-awareness,” he ends by re-asserting some old beliefs about the value of literature and the teaching of it: “The distinctive form of aesthetic pleasure we take from the literary experience gives us the sense that we are being deepened, empowered, and enriched even as we are being entertained or charmed. . Such a complex experience is difficult to theorize, professionalize, or politicize, but it is a vital and essential dimension of literary study and should be maximized wherever possible.” What good teachers of literature teach, he believes, “is not just a set of facts about an archive of texts, important as the record of literary history is; they also inculcate an informed and disciplined responsiveness not directly connected to advantage, utility, or immediate needs. Transmitting the literary heritage in all its astonishing variety, scholars are engaged in the constant rekindling of the capacity to experience aesthetic pleasure and the sense of imaginative freedom, even wonder, that accompanies that experience. . . . Such a project may not satisfy many short-term interests, but it is not without honor, and those who are engaged in it have no need to question the value of their work.”[13]

Within such a project, the structures or patterns (“narratives” if you will) into which one organizes the materials are not, I think, nearly so determined as Elkins argues by one’s cultural tradition and background. He would have it that all Western (European and American) art historians are inextricably bound to the Gombrichian narrative built around gains in illusionism. I myself, as I argued earlier with references to Bachhofer, Loehr, and Soper, believe that that is not an inappropriate approach, if applied very loosely, for Chinese painting through Song, since it corresponds in a general way with what, in my understanding, the Chinese artists are “up to” in this early period (see Zhang Yanyuan cited above, who agrees.) But it makes no sense after that; my “histories” of Yuan and later painting have taken their form, I think, from the materials they treat and the issues as they are discussed in Chinese writings and as I see them, not from any pre-existing Western model. That I have constructed a “narrative” for these later periods is true, but it has not, so far as I can see, derived from the Gombrichian or Wolfflinian or any other pattern that I have been forced to impose on them simply by being a Western art historian.

If someone objects, and someone will, that the procedure I advocate is necessarily going to be shot through with the special assumptions and biases of my time and situation, I will agree, but add that I do my best to let it or make it derive from the body of art itself, or from my limited understanding of it, and of pertinent factors in its cultural setting. They will say: but you can never make it all the latter, so as to eliminate the former; I will reply, yes, but it is worth doing anyway. Baxandall (to cite him once more) somewhere points out that to avoid attempting something because you cannot do it perfectly is like telling a runner not to run the hundred-yard dash because he cannot run it in no time at all.

There is no reason I can think of why art history in this sense, as I have tried to practice it and believe it should be practiced, cannot be global, if one means by that, applicable to any body of interrelated works of art, of whatever culture and period. Of course the practitioner, in preparation for teaching and writing, will attempt to reach an understanding of relevant indigenous concepts that will inform his or her formulations, and will strive to avoid imposing “narratives” absorbed through the study of some other culture onto the art of this one—as distinct from the practice, praised earlier in the writing and teaching of Max Loehr and others, of drawing on one’s familiarity with such patterns in other cultures in recognizing them in the material being studied. I will be the first to acknowledge the difficulty of making the distinction (yes, I know about unconscious biases) and of practicing what I advocate, and to admit my own failings sometimes in attempting it. But I am not persuaded by arguments that it is basically impossible, and so not worth even attempting.