CLP 47: 2001 "Uses of Sketches by Chinese Painters." Paper for conference on "Sketches and Creativity,"

Uses of Sketches by Chinese Painters

(Paper for Lavin/Millon conference on Sketches and Creativity, May 23-25, 2001) James Cahill

I’m pleased to be one of the speakers on this occasion, honoring Hank Millon and Irving Lavin, especially because I have a comfortable feeling of being here in a familiar role. There was a period of some years in the 1970s-80s when I found myself serving on a number of important art-history committees and boards, frequently with one or both of these two eminences as fellow members or chairs, committees on which I represented the non-West. This was a time when the field was realizing the need to open up a bit to that large outer region, shall we call it, which had remained somewhat outside traditional Western art history, and I and a few others were recruited, in effect, to represent it on the great committees and boards. Also in CAA: when Irving co-organized a session on child, primitive, and mad artists, I was brought in to present a Chinese mad artist. I’m teasing them, of course; actually, both have been extremely supportive over the years, both of me and of non-Western art studies, and I’m grateful for that, and happy to be back serving my simple function again, at an event honoring their retirements. Resuming my familiar role of showing how the Chinese weren’t like us, I’ll speak on why Irving’s European model of the sketch as facilitating or even embodying creativity doesn’t appear to apply to the practice of Chinese artists, and suggest a few possible reasons why not.

S,S. A single case that might fit that model--and even this is not entirely clear--is a group of sketches from the Buddhist cave site of Dunhuang, 8th-9th century in date (and thus predating by centuries any comparable extant materials from the West, so far as I know), drawn with brushes in ink on paper (a Chinese invention), sealed in a cave in the 11th century, acquired by the great French sinologue Paul Pelliot, and now in the Bibliotheque Nationale in Paris. The sketches were published first by Jao Tsung-i, and later, with an extensive study, by my former student Sarah Fraser. Most of them appear to be preparatory sketches for wall paintings, or parts of wall paintings, or for portable paintings on silk, and indeed many can be matched up closely with surviving finished works in both forms.

S. S. Some, however, are what Fraser terms “practice sketches,” identifiable as that by their being loosely drawn, partial images in random arrangements filling otherwise unused areas of the paper. Some of these reveal the artist trying out different ways to draw parts of his picture (as at left, the head and foot); others, like those seen above the guardian figure at right, are simply sketches of unrelated motifs, and would appear to represent the anonymous workshop artists training their hands toward the goal of being able to draw freehand the full images, in preparatory drawings for finished works.

S--. A comparable finished painting from Dunhuang, one of those now in the British Museum, impresses us as less powerful, for reasons that have often been observed. That same preference for spontaneity over finish lay behind the reputation of the divine Wu Daozi, greatest of Chinese figure masters, who painted grand figural compositions on the walls of temples in 8th century Chang’an, the capital. He reportedly instructed his assistants to color his works only lightly, or even leave them uncolored, so as not to obscure the brush drawing.

S,S. From descriptions of his paintings we can imagine his figures in dynamic postures, turning in space, with some parts convincingly foreshortened. (No work of his survives, even in reliable copies; these are two more of the Dunhuang sketches) And all this he accomplished without underdrawing of any kind, making his images come into being swiftly, as if magically--an early critic writes that Heaven seemed to have lent him its creative powers.

S,S. (Two more.) Crowds would assemble to watch in awe as he drew freehand a perfect circle as a halo around the head of some divinity. The unmatched ability of Tang-period artists to render volume in line drawing must have culminated in his work. Of course we can assume that Wu Daozi, and later artists as well, in their long pursuit of this transcendental skill, covered all the waste paper they could find with practice sketches like those from Dunhuang. But no other examples survive, nor is there any written record of the practice, so far as I know.

S.S. What do survive, and are mentioned in texts, are preparatory drawings, huagao in Chinese. Xia Wenyan, in his 1365 Tuhui Baojian, writes: “The preparatory drawings of old artists are called sketches (fenben); many connoisseurs have collected and treasured them. This is because, in their rough and unplanned look, they have a wonderful spontaneity.”[1] Surviving fenben are not especially rough or unplanned, but rather highly finished preliminary drawings or cartoons for wall or scroll or screen paintings. These are two examples from a large group with an old attribution to Wu Daozi; they appear to be 13th or 14th century in date.

S,S. We are sometimes unsure whether a linear design that could be such a cartoon for a temple wall painting is that or, alternatively, a careful copy made after the wall painting to preserve its composition in detail, for use in subsequent restoration or replacement. This work, a handscroll on silk loosely attributed to the 11th century figure master Wu Congyuan, generates far more energy in its dynamic procession of Daoist figures than the narrow confines of a handscroll (only about 40 cm. in height) can comfortably contain. It must have been either a cartoon for a wall painting or a copy after one.

S.S (Cui Bo, Guo Xi.) Artists who painted large, complex compositions normally began by making sketches directly on the silk or paper; in modern times, and presumably earlier, this was done with a charcoal stick, which left markings that either were covered by the painting or could be brushed off when the work was finished. That this was the common practice is attested by the singling out for special praise of artists who did not make them, but attacked the silk directly, with their entire compositions (so the critics believed) clearly in their minds. One such was Cui Bo, whose Hare and Jays at left is dated 1061. The great landscapist Guo Xi, whose Early Spring dated 1072 is on the right,

S -- (here a detail of the middle right) would sometimes avoid the undesirable planned look by directing the plasterers preparing a palace wall to throw or daub on the plaster roughly, so that he could use the resulting chance configurations as the basis for his landscape composition. (Early Chinese painting can boast a number of such predecessors to the Western artists of Janson’s “Image Made by Chance.”)

The medium of ink and colors on silk, unlike oils on canvas, didn’t allow any overpainting, and only very minor correction, so that we can’t expect to find pentimenti or detect changes in plan through X-ray, as we can with European easel paintings. When we can detect them at all, it’s through close observation and assumptions about the artist’s intent. (Show place in middle right)

S --. In a remarkable and unnoticed passage in his Early Spring, Guo Xi must have intended originally to turn a sketched-in area (here blacked out) into one of the sloping banks of earth, like the one beyond it, with bare trees growing out from it.

S --. But, in a radical last-minute decision, realizing that to fill it in that way would work against the whole effect of erosion and instability that is the theme of his picture, a vision of the world in process, he opens it up instead to a view into what appears to be a subterranean hollow, making the earth mass above overhang implausibly, adding more trees that turn awkwardly to grow back into the newly-opened space. Such a passage reveals a great deal about the working methods and priorities of Chinese painters.

S,S. I have been speaking up to now about professional artists, academy masters, technically trained studio painters. But, as I’m sure many of you know, Chinese painting after the 13th century or so is dominated, at least in critical writings and collectors’ preferences (which have largely determined what has been preserved), by scholar-amateur artists, the so-called literati painters. (Two examples, from the 12th and early 14th centuries.) Working in principle from inner motivations, they scorned preliminary sketches along with technical finish generally, since they aimed at capturing both momentary feeling and something of their personal cultivation and character in brushstrokes. The idea of working toward a “better” image by way of sketches would have been antithetical to their purpose.

S,S. (Detail from a painting of grapes by Riguan, 13th cent.; pomegranates by Bada Shanren, 17th cent., both Ch’an or Zen Buddhist monk-artists, another category of amateur painters.) By relying on inherited type-images without being bound to them, and by repeating certain motifs and subjects in which their hands had became practiced, while giving room also to momentary impulse, they could produce accomplished and original pictures as if spontaneously, on demand.

S,S. (Sections of Zhao Mengfu’s “Village by the Water,” 1302, at right; and at left, from Huang Gongwang’s “Dwelling in the Fuchun Mts,” painted over the three-year period 1347-50.) Landscape paintings by literati masters, even when elaborate in composition, had to maintain a certain transparence, displaying the hand of the artist so as to suggest sustained spontaneity.

-- S. (Another sec’n of Huang Gongwang’s masterwork, the most admired, coveted, and written about of literati landscapes.) According to Huang’s inscription, he began by sketching out his entire composition (on the roll of paper, that is) all in one day, and then added to it from time to time over a period of three years, until it was--somewhat arbitrarily--judged to be finished and could be presented to the intended recipient, who had grown impatient. The quasi-improvisational look of the work, in which the hand of the artist remains clearly visible, reflects this manner of creation. Here we might find some agreement with Irving’s model in that a kind of sketchiness is in effect equated with creativity. But it is sketchiness that is not a stage on the way, but survives into the finished work, as visual testimony to the spontaneity of its execution.

S,S. Landscapes by the scholar-amateur masters, then, like their simpler pictures, were in principle made without preparatory sketches. Even when radically experimental (like these two by the early 17th century landscapist Dong Qichang) they were required in theory to retain some roots or basis in older painting, expressed in stylistic references that are sometimes abstruse, making this a connoisseur’s art. Xing si or lifelikeness had been placed scornfully by a 14th century theorist at the bottom of a list of criteria for judging quality in paintings, and this attitude, fundamental to literati painting, continued to prevail among critics and to affect artists’ practice. The artist, then, had in principle no impetus for Gombrichian changes of the inherited schemata in the direction of truth to nature. Moreover, since the viewer was not located in relation to the scene by any kind of systematic perspective, adjustments to correct that aspect of the picture were not needed. And so forth: the hypothetical ideal form toward which he might work through successive sketches would have had no meaning for the Chinese painter.

I should add once more, to avoid misunderstanding, that neither reliance on pre-existing type-forms nor resonances with old styles diminished the originality and creativity of these paintings, at least the successful ones, any more than they do in works by Picasso or Stravinsky or Ezra Pound. By no means do they move these Chinese pictures into Irving’s “medieval” model, in which the past is simply replicated endlessly. Western authorities as distinguished as Gombrich and Arthur Danto have misunderstood this aspect of Chinese painting. I’ve devoted some recent writings and lectures (including one given here at Princeton) to correcting these wrong readings of later Chinese painting; but I won’t even try to summarize that argument now, except to say that as early as the 14th century, I believe, Chinese painting entered into what Hans Belting and others call a post-historical phase.

S,S. Literati painters did sometimes make sketches, but for other purposes. Huang Gongwang, the artist of the landscape handscroll we just saw, advises the painter to carry a brush with him on strolls in nature to sketch interesting trees (and, by implication, other things) as he encounters them. Literati landscapists such as Huang’s younger contemporary Ni Zan (right) and Dong Qichang (left) made sketchbooks and scrolls of trees and rocks that may be instances of that practice, although the relationship of their imagery with that of nature is tenuous. (Dong, even though he was quite incapable of depicting a house convincingly, continued to try.)

S,S. In doing such compilations of motifs these artists are in some part following a much older type, the artist’s repertory scroll such as this one, which supplies models for any conceivable way one might want to portray a horse. (Such a taxonomic freedom may call to mind, for a western viewer, Foucault’s famous citation of Borges and “a certain Chinese encyclopedia.”)

S, S. Professional artists of the kind with which we began did go on using preparatory drawings and sketches; and I will conclude with a brief account of the uses to which they were put. Some have been mentioned already. Studio artists would make sketch copies of compositions and motifs from old paintings to which they had access, to be used as needed in their own works. The 17th century master Gu Jianlong does this in a preserved album, one of a great many he reportedly made. (In this leaf he copies the pine trees and rock from a Ming handscroll now in the Freer Gallery of Art.)

S,S. Portraitists would make preliminary drawings of their subjects, either to record their likenesses for later use or to show to their sitters to get their approval before doing the finished portrait. What I take to be an example of the former is an album of portraits of eminent men of Zhejiang Province, dating from the mid-17th century and exhibiting the new light-and-dark shading and other techniques adopted from European art that revitalized Chinese portraiture at this time, and made Chinese viewers exclaim in wonderment that the painted images were exactly like reflections in a mirror. The pictures in this album are usually seen as finished portraits, and I myself published them as that, realizing only later that some features of them--the coarse paper on which they are painted, inscriptions that have the character of informal artists’ jottings--show them to be preliminary drawings. As such, they present their subjects “warts and all,” as we would say, with uncompromising realism; the same people would doubtless have looked more bland and benign in the finished portraits.

S,S. In 1781 the Yangzhou master Luo Ping made this preliminary sketch for a portrait of the famous litterateur Yuan Mei and gave it to him for approval, which in this case was withheld. Before returning it to the artist Yuan Mei wrote a long, facetiously philosophical inscription at the top, playing with the question of whether he himself or the artist’s image was the real Yuan Mei, and adding that the people of his household complained that it made him look like the man who delivered vegetables to the back door. Again, the final image would probably have been less harsh, but it was never made. I first published this preliminary sketch in 1959, and Richard Vinograd discussed it much more seriously in his 1992 book on Chinese portraiture, but neither of us put it in what I now believe to be the right context.

S,S. Painters doing major pictures for important clients, and most of all court artists working for the emperor, were required to prepare and submit detailed cartoons before they could carry out the finished work. When the Italian Jesuit artist Giuseppe Castiglione (Chinese name Lang Shining), who served under three Chinese emperors, was preparing to paint his famous “Hundred Horses” scroll in 1728, he made a cartoon of the whole large-scale composition, which is now in the Metropolitan Museum of Art. These are two details.

S --. And this a corresponding detail from the finished painting. The cursive drawing of some of the background in the cartoon, contrasting with the precise linear drawing of the figures, may reflect Castiglione’s practice of sometimes leaving the background to be filled in by Chinese assistants in his studio.

S,S. Similarly, before a team of five painters including Chen Mei executed the well-known Qing-period version of the “Qingming Festival” scroll in 1735, a highly detailed and finished preliminary drawing was made--and signed, somewhat surprisingly, by an entirely different artist, whose name does not appear on the finished scroll. The circumstances behind this odd division of labor remain to be clarified.

S.S. Even under a system with such built-in precautions, a finished painting might displease the client and have to be redone. We know from literary references about paintings that were returned to the artist by dissatisfied customers; I will conclude with the case of a court painting done for the 18th century Qianlong Emperor, who kept an especially tight control over his academy artists. It is a New Year’s picture portraying the emperor and some of his consorts and sons on the porch of a palace building, and it survives in two versions. One, at right, is dated 1738 and signed by five artists, including the same Castiglione and Chen Mei. The other is unsigned, but has been preserved in the palace in Beijing. Those who have written about the two versions have not speculated on their relationship; I would guess that the unsigned picture (at left) was a first version that was rejected by the emperor, and look for reasons why. One is surely that the unsigned work situates the emperor off-center, and within a strong perspectival pull into depth that works to draw the viewer’s attention away from him. The other reason is that he is portrayed in a manner too relaxed, too anecdotal, bending his head to look down at a child in his lap.

-- S. The same is true of other participants in the scene, such as these boys who watch one of their number setting off a firecracker: they are more relaxed and individualized than the highly formal requirements of court painting could accomodate. For us, these may be just the features that make the picture more appealing, the other stiffer and colder. We can surmise that the artists themselves may also have preferred the earlier version, but in China, and especially in the imperial court, their preferences didn’t count for much.

I’m quite aware that some or even most of the Chinese practices I’ve outlined have approximate equivalents in European painting, but limits of time and my competence have ruled out serious analytical comparisons. I’m inclined nonetheless to recognize important differences, and especially, for our present concern, to believe that the Chinese use of sketches stands apart from the somewhat dichotomized developmental pattern of the symposium statement, in which the medieval practice, tending to inhibit creativity, gives way to the Renaissance and later practice that encourages it. This formulation is based, I think, in deep Western beliefs about ideal forms or images toward which the artist might strive, images aesthetically or naturalistically or expressively better than those from which they depart, and also about how the best art arises from conditions in which the artist is least subject to outside demands. Neither quite corresponds with Chinese reasons for preferring one image or work over another. Painting in China for which the artists claimed an inner motivation of self-expression, chiefly literati painting, emphasized spontaneity in such a way as to make sketches counter-productive, while painting done by the professional masters, responding (in principle) to outer motivations, located decisions about altering or “improving” the image more in the client than in the artist. How far these non-correspondences betray real differences between the European and Chinese artistic traditions. and how far they only reflect differences in the ways the concepts and practices within those traditions are formulated, is a larger and more difficult question that I will, in the end, leave unsatisfyingly open. Thank you.

3017 Waipuna Rise
Honolulu, HI 96822
August 9, 2001

Dr. Therese O’Malley
Associate Dean, CASVA
National Gallery of Art
Washington, D.C. 20565

Dear Therese,

I got back from some weeks in China to find your letter of 18 July. I do remember that you and Irving said from the beginning that the papers would be published, but I’m afraid I put this at the back of my mind and didn’t think about how practicable this was for me. And I put together a paper for the conference that seemed right for that occasion, a short account of a large subject, the uses of sketches in Chinese painting, using lots of slides. But there is no way I can turn this into a publishable paper. The things I showed are far-flung, in many collections, and some were introduced briefly only as examples of some type or phenomenon, and getting photos and permissions for all of them, or even for enough to illustrate a paper with footnote references for others, would be a large task that I can’t now undertake, even with a much later deadline than September 4. For reasons of funding, the book I’ve been working on for several years must be finished by the end of this year, to be published (along with several others in the series, The Culture and Civilization of China) around the end of next year; and my travels, and talks with a Chinese collaborator in Beijing who hasn’t even finished his part, leave me in a very tight situation, unable to take on anything time-demanding.

This is entirely my fault; I should have done a different kind of paper, less dependent on visuals, and undertaken to get the photos and permissions much sooner. And I apologize for this to you and Hank and Irving and others. Yours will be a very distinguished and variegated volume, but China, alas, won’t be there, through my bad planning.

James Cahill

[1] Xia Wenyan, tuhui Baojian, preface 1362 (HSCS ed. chap. 1, p. 3)

Latest Work

  • Conclusion Conclusion
    VI Conclusion It is time to draw back and look, if not at the whole Hyakusen, at as much of him as we have managed to illuminate in this study. Dark areas remain, and doubtless many distortions, but...

Latest Blog Posts

  • Bedridden Blog
    Bedridden Blog   I am now pretty much confined to bed, and have to recognize this as my future.  It is difficult even to get me out of bed, as happened this morning when they needed to...