In the introduction to Performing the Visual: The Practice of Buddhist Wall Painting in China and Central Asia, 618–960, Sarah Fraser describes her project as an inquiry into the medieval artist’s practice through close analysis of several of the sixty-five ink sketches from the ninth and tenth centuries that were preserved in the sealed Cave 17 of the Mogao Grottoes at Dunhuang and are now mostly in the British Museum in London and the Bibliothèque nationale de France in Paris. The author examines the relationships between the sketches and the finished murals and silk banner paintings from the ninth- and tenth-century Buddhist cave-shrines at Dunhuang, considers what such relationships reveal about the ways workshop artists produced murals and painted banners, and analyzes the attitudes of ninth-century art historians toward art as process and product.

Fraser asks important new questions: What was the economic structure that supported the artisan painter? What was his relationship with the Buddhist establishment and with lay donors? What is the relationship between artists’ drafts and the finished murals and banners? What can we know about workshop practices from extant sketches? Is there a connection between the practices of anonymous workshop artisans and the work of celebrated mural-sketchers such as Wu Daozi? What were contemporaneous critics’ attitudes toward sketches and drawing? What biases and agendas influenced the critics of the ninth century? What is the relationship between the ubiquity of temple mural production in the late Tang Dynasty and the later change in taste that favored figure painting done in monochrome ink drawing?

In chapter 1, Fraser describes the economic structures for artisan painters at Dunhuang. These artists worked in the painting academy established in the 930s by the governing Cao clan, or in government painting workshops, which were temporary bodies put together to execute a particular project. There was also something called the “painting guild” (huahang), whose relationship to these other entities is unclear. Recovered government documents suggest that artisans were also requisitioned to practice their particular skills to satisfy their government requirement for corvée labor, while votive inscriptions reveal that many guilds and individual artists were wealthy enough to be themselves donors of Buddhist images. Strangely, in the example the author presumably selected to illustrate the workings of these institutions, the restoration of Dunhuang Cave 96 by Cao Yuanzhong and his wife, Fraser interprets the record of repair to say that the workmen were “hired from another location” (38) and were paid with meals of cooked food and wine. This appears to contradict the notion of an established government bureau for art production under the control of the Caos, which would likely have paid its workers a regular salary in grain, just as it paid other government employees in the Tang. If the Caos did not use their painting academy to train and support artists for their own projects, why did they use it?

Chapter 2 discusses the practices of the wall painter as they may be reconstructed from the extant sketches and murals, in particular, those of The Magic Competition. Fraser explains the theme and analyzes the compositional structure of these murals as a “panorama of good and evil” (85), showing which scenes from the scripture were chosen and giving some thoughtful reasons why. The Magic Competition, which debuted in Cave 9, ca. 892, was given its finest form in Cave 196 in 893–94 and had become rote by 920–40, when it was painted in Cave 146. These grottoes were all sponsored by elite local families, and Fraser makes the interesting proposal that the murals of The Magic Competition be regarded as “a type of frontier court art” (89). She shows that this and a handful of other themes were done again and again in the grottoes and calls this the repertoire of the artisans. Fraser observes that these oft-repeated subjects are found in murals in matching pairs on opposite walls of the grottoes, revealing “strict standards of measurement for the compositional spaces. Within each cave, the widths of each of the individual compositions arrayed along a wall are nearly identical” (87). I had the impression that this intentional creation of identical mural spaces was part of the artisan’s battery of techniques for efficient and effective mural production, but after the author discusses how the interiors of grottoes were supposed to resemble constructed wooden architectural temple complexes—not only in including all the elements of a temple complex, such as gates, corridors, and halls, but also in being decorated with painted imitation of wooden architectural elements, and that “the regular intervals in murals imitate the bays created by pillars in buildings” (93)—I was left wondering if the mural size was dictated by the artisan’s practice or by the imperative of imitating the regular intervals between bays.

The problem I had with this chapter is unfortunately the underlying premise of the whole book, that is, that the sketches recovered from the sealed Cave 17 have a demonstrable relationship to the murals. The author lays out her argument:

It is possible that the drafts were made after completed paintings by a few, unrelated artists and that these works on paper are just the notes of an unorganized group of painters. The drawings, however, can be firmly linked to the hundreds of finished paintings at the site, suggesting that they are an index to organized workshop procedures. (54)

A few of the sketches probably relate to the murals, such as a scroll of sketches of the subsidiary scenes (fig. 2.6) for a depiction of the advent of Maitreya on earth, which are similar to the subsidiary scenes in the mural in Cave 196. The likeliest case for a sketch being a preparatory drawing is the one showing Sarasvatī and groups of worshipers, from the Sutra of Golden Light (fig. 2.9a), where the groups of figures are labeled with their identities and with numbers that may indicate their position within a finished mural, such as the one on the south wall of Cave 196.

Yet even if we accept the notion of a relationship between the sketches and the murals, what proves that the sketches are preliminary drawings? Moreover, if these are not preliminary drawings, then Fraser’s insistence that these sketches reveal something of the way painters worked needs to be rethought. Her notion that some sketches are conceptual and served to work out a mural’s overall composition, while others of individual figures were models or references, holds only if the sketches are indeed preliminary drawings.

The discussion of the use of pounces, also found in the sealed cave, is more credible (102–8). A pounce is a drawing on paper whose outlines have been evenly perforated with pinpricks, so that when the pounce is placed against a wall, powder can be applied to the wall through the holes, leaving a dotted outline of the figure on the wall for the artist to paint and embellish. Pounces were used to produce the “wallpaper” effect of the identical figures of the Thousand Buddhas in the ceilings of the grottoes, and variations were then created in the painting of figures by varying the mudras and the coloring. Some pounces are about the same size as the figures in the Thousand-Buddha-pattern ceilings in Caves 146 and 196, two of the grottoes containing The Magic Competition murals, while a bigger pounce was probably used to make the larger Buddha figures on the walls of Cave 61.

Chapter 3 explores three topics: 1) the relationship of sketches to copybooks (which turns out not to be very significant); 2) miniature copy sketches of post-Tang wall paintings; and 3) the relationship of proto-baihua (I believe the author means baimiao) practices in the Dunhuang ateliers and literati painting styles associated with sketching. Fraser notes that Chinese scholars consider the (probably fourteenth-century copy of the) monochrome ink painting of 1050 by Wu Zongyuan, called The Star Gods Procession, to be a miniature copy of a wall painting that was perhaps made to facilitate any later necessary repairs. Fraser says that the work may instead be a painting that was produced to obtain a patron’s approval for a planned mural, and that afterward it circulated as an independent art object. I like that theory indeed, but her conclusion that paintings such as this influenced the development of baimiao painting by the literati, who adopted the “sketchiness” of workshop sketches to signify spontaneity, must remain in the realm of conjecture.

In chapter 4, Fraser examines Dunhuang sketches for silk banners and ritual-practice diagrams. Although the author asserts that “banner sketches were either traced onto clean painting surfaces or used to make direct copies” (132), the evidence presented does not support this conclusion. The sketches are often, though not always, the same size as the banners and are as complete in detail. Fraser considers the close similarities between the sketches and the finished banners to indicate that the sketches were traced onto the banners, but it seems that the nature of the relationship between the two, which I would characterize as close but not exact, actually argues against tracing, which should produce an exact replica of the sketch on the banner. The sketch and the banner may instead be two versions of the same figure by the same artist, who was trained since childhood to draw certain figures in a certain way and whose many efforts in different mediums would look alike yet never be exactly the same. Further, why would a workshop artisan, who possessed an ingrained, organic ability to draw certain types of figures, take the time to elaborately trace a figure onto a silk banner, when he could probably draw the figure in his sleep?

Chapter 5 surveys the history of scholarship on The Magic Competition, from its beginnings in the Sutra of the Wise and Foolish to its development into vernacular literature and mural painting. The tale of The Magic Competition, which ends with the Buddhist believers led by Shariputra defeating Raudraksha and the heretics in six contests of strength, was written as a dialogue–debate, similar to those discussions that actually occurred at the Tang court; the legend is described this way in painting as well. Dunhuang audiences may have enjoyed this story and its pictorial versions as encapsulating the tensions between Tibetans and Chinese in the eighth and ninth centuries, which ended in a Chinese victory. Shariputra and his cohort are depicted as racially Chinese, while Raudraksha and company have the features of foreigners. In Fraser’s view, the composition of the mural scenes in The Magic Competition is “linked to” but not necessarily preceded by the verbal forms used in the performance of this narrative by storytellers. Moreover, the “[p]erformative movement conveyed through dramatic poses of the body signaled the arrival of vernacular expression” (165), which is to say that the murals, concomitant with the development toward dramatic action and emotion in the telling of the story, now depict the narrative with emphasis on the six contests and characters with dramatic poses and extreme facial expressions. The author disagrees with Wu Hung’s theory that the murals are independent of the bianwen, which are the texts for the oral recitation of the story (“What is Bianxiang? On the Relationship between Dunhuang Art and Dunhuang Literature,” Harvard Journal of Asiatic Studies 52, no. 1 [spring 1992]: 111–92), though she does agree that the murals did not serve as visual props for the performance of the story. Rather, Fraser considers the murals homologous to the oral performance, the main similarity being the episodic presentation.

In chapter 6, the author states, “In the imagination of viewers and critics during this period [the ninth century], a new link forms between monochrome sketching, spontaneity, artistic performance, and motion depicted in two-dimensional space” (197). The two personifications of this trend were the mid-ninth-century art chroniclers Zhang Yanyuan and Zhu Jingxuan, and the artist upon whom they set their hopes was the eighth-century court and professional painter Wu Daozi. Zhu Jingxuan praised Wu Daozi for capturing hundreds of miles of river landscape in his mind and reproducing it on the walls of the palace without benefit of preparatory sketches. Fraser believes this anecdote reveals that “the performance of sketching itself—promoting the act of sketching to an aesthetic object to be appreciated on its own terms—becomes the center of discussion” (199). My central objection to this assessment of the “act of sketching” is the definition of “sketching.” Was Wu Daozi’s painting considered sketching, underdrawing, or finished painting? In Zhu Jingxuan’s anecdote, it says “Wu finished more than three hundred li of the scenery in a single day” (198). If we compare this to what we have of Tang landscape painting—such as that seen in the polo murals in Prince Zhanghuai’s tomb of 706, which was executed in monochrome brushwork to which no colors have been added—we might conclude that Wu Daozi’s landscape paintings were not sketches but finished paintings. By contrast, in terms of Buddhist and Daoist figures, where such a mural drawn by Wu Daozi might have been considered finished by the ninth-century critics, Zhang Yanyuan’s frequent lament about Wu’s paintings that “artists have spoiled them in finishing them in colours” (William Acker, Some T’ang and Pre-T’ang Texts on Chinese Painting, 2 vols. [Leiden: Brill, 1954–74], 1:266, 268, 272) strongly suggests that the owners of the mural did not consider it finished as an ink drawing alone.

Regarding the anecdote about Wu Daozi painting the river landscape without sketches, Fraser says “With this tale of ‘The sketch in the mind, not in the hand’ one may identify a central tenant [sic] of Tang aesthetic theory—the ideal artist relies on intuition rather than learned technique” (199). If we repair to Zhang Yanyuan’s entry on Wu Daozi in Lidai minghua ji, however, we find that the artist relied on three distinct sources, two of which were learned techniques: 1) “he learnt his brush methods from Chang Hsü [active first half of eighth century] which shows yet again that the use of the brush in calligraphy and painting is identical”; 2) “his spirit borrowed from the creative powers of Heaven”; and 3) “there must have been some orally transmitted secret for this which no one [now] can know” (Acker, Some T’ang and Pre-T’ang Texts, 1:179–80). Zhang Yanyuan’s agenda is transparent: it has nothing to do with elevating sketching to the height of painting and everything to do with elevating painting, seen as an artisan craft, to the level of calligraphy, the gentleman’s art. Zhang’s second source agrees with Fraser’s argument that Tang critics privileged “intuition,” “spontaneity,” and the Daoist “notion of effortless skill” (213) in artists, but the third source shows that Zhang equally credited the training that Wu Daozi received from his teacher. No paintings remain from the hand of Wu Daozi, so I sympathize with the author’s desire to reach the artist somehow through the sketches that survived at Dunhuang, but I am far more inclined to agree with her assessment that Wu Daozi was admired by Zhang Yanyuan as part of a trend toward ink monochrome painting rather than for expressing an aesthetics of sketching, since Zhang was demonstrably anticolor. Fraser is right on the money with her statement that “Zhang’s disdain of color and celebration of linear expression may be connected to a general rejection of foreign ideas and cultures in the late Tang” (225). From the old Confucian associations of color with superficial appearance and the charms of women, with the advent of Buddhism, color and gilding were used as a proselytizing device; and to the Confucian-minded color came to mean the seductive lure of Buddhist icons, even if to Buddhist believers it signified the living efficacy of icons. Zhang Yanyuan wanted to privilege the force of nature or the spirit of Zhang Xu working through Wu Daozi, not the power of the dharma operating through some craftsman or of its own volition.

Although I admire the questions posed in Performing the Visual, with their focus on anonymous workshop artisans, perspective on actual practice, and awareness of class attitudes, many more issues are raised than can be answered definitively. There is nothing wrong with that; in fact, it is tremendously helpful for writers of seminar syllabi. In the end, even though I remained unconvinced of some of the assertions that underlie these intriguing questions, I salute Fraser’s boldness in posing them and stand in awe of the amount of worldwide travel and difficult study required to master the materials necessary to engage in such an ambitious project.