What is 'Chinese narrative illustration.'

Julia K. Murray

In modern writings on Chinese pictorial art, the term narrative illustration is loosely applied to a diverse array of works. Although there seems to be an implicit consensus that the designation is appropriate to pictures that illustrate a story, the limits of this territory are unclear: Is it necessary for the pictures to expound the story visually, as in depictions of the life of the Buddha [ILLUSTRATION FOR FIGURE 1 OMITTED], or may they simply evoke or refer to it, however cryptically, as in assorted pictures of a boat near a bluff that are taken to illustrate Su Shi's 1082 account of his visit to the Red Cliff [ILLUSTRATION FOR FIGURE 2 OMITTED]? Should the term narrative illustration apply to any image that has a textual basis of any kind, even if it is a lyric poem or a catalogue record? Is it appropriate to extend the designation, as many scholars do, to generic scenes, such as depictions of livelihood and amusement [ILLUSTRATION FOR FIGURE 3 OMITTED], even passages of landscape? On the premise that an overly inclusive understanding of narrative illustration is not very useful for the purpose of analysis or theory, I will propose here a more limited and rigorous definition for the category.

A significant portion of Chinese pictorial art consists of representations of stories and texts.(1) Such pictures not only embody and express cultural ideals and values, as traditional and modern writers have repeatedly observed, pictures also played a part in forming and disseminating social norms and political authority, functions explored in more recent scholarship. However, narrative illustration was not conceived as a separate genre of painting (hua)(2) in premodern histories or criticism. Instead, it has become an entity in the study of Chinese art largely because of conventions adopted from Western art historical writing and has now become so naturalized that art historians have tended to forget that there is no traditional Chinese term for or concept of such a genre or critical category.

In modern Chinese, the combinations xu shi hua and gu shi hua have been formulated as terms meaning "narrative painting" by analogy with the terms xu shi shi and gu shi shi for "narrative poetry" (see App. below).(3) The compound xu shi (literally, "to tell the matter") is found in traditional Chinese literary criticism, but with the limited meaning of recounting an action in prose, as opposed to verse or song;(4) it is not used to characterize an entire piece or to name a category of pieces in which events are narrated. Xu shi hua might thus be translated as "to tell the matter in pictures" and might potentially be used to imply full-blown pictorial storytelling, such as a detailed pictorial exposition of the life of the Buddha [ILLUSTRATION FOR FIGURE 1 OMITTED]. The term gu shi, which is used by Sima Qian (145-86 B.C.E.) to mean "old matters" or "events of antiquity," has been adapted in modern literary criticism to denote the fictional "story."(5) Accordingly, gu shi hua would seem to mean "picture of an ancient event" or "story - picture," and it might suggest a more concise representation than xu shi hua - a single scene that epitomizes the story, as in Li Song's allusion to Su Shi's Rhapsody on the Red Cliff (Chi bi fu, [ILLUSTRATION FOR FIGURE 2 OMITTED]), rather than a detailed depiction. Nonetheless, gu shi hua and xu shi hua are used interchangeably, despite their apparent potential for specifying single-scene and multiple-scene illustrations, respectively.

In traditional histories and criticism of Chinese painting, subject matter served as the major classifier.(6) Accordingly, the pictures now grouped together in the modern category of narrative illustration were not necessarily considered similar in earlier periods. Narrative illustrations are found under headings such as Figures (ren wu), Buddhist and Daoist Subjects (variously fo dao, dao shi, and xian fo), and Ghosts and Spirits (gui shen). The Figures genre was sometimes subdivided into more specific types, such as Beautiful Women (shi nil), Peasants (tian jia), and Barbarian Tribes (fan zu). An important consequence of this focus on subject matter is that the traditional categories that contain narrative illustrations also include paintings that do not refer to a story, such as depictions of social roles [ILLUSTRATION FOR FIGURE 3 OMITTED].

An analogous situation exists in the field of Chinese literature. According to the si bu (four-part) system of bibliographic classification, which has been in use since the seventh century, the major genres of writing are classics (jing), history (shi), philosophical discourse (zi), and belles lettres (ji). All four categories contain many works that recent Western literary criticism would identify as narratives, and narrative technique is found across a broad spectrum of writings. Nonetheless, there is no traditional term or separate genre for narrative in Chinese literature, as literature specialists have long recognized.(7)

Unlike scholars of Chinese literature, however, art historians have not given much consideration to the problem of defining narrative in an artistic tradition whose significant categories were formulated differently from those of the European tradition. The issue was obliquely raised by John Hay in an article largely devoted to Zhao Gan's tenth-century handscroll Along the River during Winter's First Snow (Jiang xing chu xue tu).(8) Discussing the diverse associations of the theme of the fisherman in literature, Hay proposed three categories - moral narrative, literary narrative, and genre narrative - which he considered to have "approximate equivalents in painting."(9) However, Hay did not explain this tantalizing formulation further, leaving readers to infer definitions from his examples. My own reading leads me to think that Hay's categories are defined by varying criteria, rather than by a consistent standard that would make the groups mutually exclusive. "Moral narrative" seems to refer to pictorial representations with a didactic purpose, "literary narrative" to pictures that embody poetic symbolism,(10) and "genre narrative" to subject matter drawn from everyday life. In other words, the first category is defined by purpose, the second by expressive technique, and the third by theme. Accordingly, the groupings can easily overlap, and a single painting may well fit in more than one. A case in point is the illustration to the poem "In the South There Are Fine Fish" (Nan you jia yu) from the Book of Odes (Shi jing), which Hay offers as an example of moral narrative [ILLUSTRATION FOR FIGURE 4 OMITTED].(11) The painting is based on a canonical Confucian text that was believed to express moral intent; the rhetorical images of the poem are depicted (flying birds, trees with gourds) or implied (fish); and the picture shows two humbly dressed men who are fishing with a basket and a net. If the picture is a narrative illustration at all, surely it fits all three of Hay's categories: moral, literary, and genre narrative.

Moreover, Hay's categories are indefinite enough that virtually any pictorial representation of human action could arguably be admitted somewhere. Subsequent writings by other scholars have invoked his terms in analyzing an immense range of pictorial works. For example, in Beyond Representation, under the heading "Narrative Representation," Wen Fong uses Hay's categories under the slightly modified names of moral-symbolic, literary, and genre-descriptive narrative. Among the many works that Fong discusses are a bronze vessel bearing repetitions of a generic hunting scene; mass-produced tomb files with mix-and-match stamped motifs of trees, horses, and human figures; daily-life scenes of men farming or women tending children; and portraits of human or animal subjects.(19) The variety encompassed by his examples implies a formulation of narrative representation that includes any picture whose significance is independent of the artist and whose mode of depiction is objective and descriptive. The diversity of the works that meet these criteria is not problematic for Fong's major purpose, which is to argue (in a later chapter) that literati artists from the late twelfth century onward created an altogether different kind of visual art, one that was self-referential and self-expressive. Indeed, Fong's concept of narrative fits well with the dichotomy between the objective ("narrative") and subjective ("lyric") modes posited by literary theorists.(13) However, it is difficult to study the methods and development of narrative technique when using a definition that includes such disparate works.

In similar fashion, Wu Hung uses a broad conception of narrative in order to emphasize the newness of an "iconic" mode of representation, which he finds emerging in Chinese pictorial art during the second century C.E., under the influence of Buddhist imagery.(14) As Wu describes it, the composition of an icon exhibits bilateral symmetry and the image is presented from a frontal perspective, so that it engages the viewer directly.(15) Nothing is happening within the scene, and the image appears to be timeless. Indeed, it is not bound to a specific time or meaning, for its interaction with the viewer (often a worshiper), causes its significance to change with each encounter. By contrast with this foreign-inspired mode of representation, Wu argues, the indigenous Chinese tradition is based on the "episodic" mode, which is characterized by self-contained, asymmetrical compositions, whose elements are rendered in profile or three-quarter view. Anchored to a specific time and place, the figures relate to one another or to objects in the scene, not to the viewer, who is "a witness, not a participant." Wu characterizes this mode as narrative because the interactions take place within the picture; elsewhere, he notes also that episodic compositions are based on narrative literature.(16)

Another common practice in discussions of Chinese painting is to use the term narrative to describe the compositions of long handscrolls that take the viewer on a journey through time and space as the scroll is unrolled. In the extreme case, even the individual motifs that constitute the landscape, such as trees and rocks, have been called narrative.(17) The basis for such designations is the element of narration that involves "telling" or, more neutrally, "presenting,"(18) a process that implies the passage of time. In accordance with traditional viewing practices (as opposed to modern museum display), only a portion of a handscroll is viewed at one time. As the scroll is unrolled, each section is revealed in turn, only to be rolled out of sight when the viewer moves on to the next portion.(19) Thus, the composition is experienced sequentially; moreover, lengthy works also would have been painted section by section. A case in point is Zhang Zeduan's Qingming shanghe tu (variously translated as Going Upriver for the Qingming Festival, Qingming Festival on the River, or even Peace Reigns on the River(20)), a twelfth-century tour de force that is often characterized as narrative. This handscroll depicts a parade of human activity through a continuous terrain that extends from the countryside through a bustling suburb and on into a walled city [ILLUSTRATION FOR FIGURE 5 OMITTED]. Over seventeen feet in length, the painting encompasses literally hundreds of figures and dozens of man-made structures. The profusion and clarity of the unusually realistic details have led many writers to interpret the painting as a narrative of life in the Northern Song capital, Bianliang.(21)

Concepts of narrative illustration that accommodate so much diversity have the disadvantage of blurring the distinctions among early types of pictorial representation and obscuring their evolution. In order to say anything meaningful about a category called Chinese narrative illustration, it seems desirable to formulate a more restrictive definition that identifies a consistent body of material. Even though this category will be an artifact of modern Western scholarship rather than an indigenous ordering of knowledge, I would argue that it is reasonable to discuss Chinese cultural phenomena in nonindigenous terms if the criteria and motivations are made explicit. Indeed, in a certain sense, non-Chinese scholars of the late twentieth century can write about Chinese culture of earlier periods only from an outsider's viewpoint, unless they work entirely within the discourse of traditional connoisseurship.(22) Otherwise, even if modern researchers deliberately frame questions that respect the parameters of traditional categories, it is hard to imagine that their conclusions will not be somehow influenced by the premises of modern scholarship. Recognizing this, I believe that it may be both valid and illuminating to group and analyze certain pictures under the nonnative rubric of Chinese narrative illustration. However, I would agree with the literary theorist Tzvetan Todorov that a category defined by outsiders should not be characterized as a genre if traditional critical discourse indicates that the culture did not perceive it as such.(23) Moreover, a definition for narrative illustration cannot simply be adopted from Western art history because, even there, the existence of such a genre is typically assumed rather than clearly defined? Accordingly, it seems necessary to consider some basic problems of definition. To set the stage for such a discussion, I will briefly refer to some ideas from Western literary theory that I have found useful in thinking about Chinese narrative illustration.

Literary Definitions of Narrative

Narration exists in a variety of symbolic systems, but most of the theoretical discussion and analysis has focused on its manifestations in literature and film. Scholars generally agree that a fundamental marker of narrative is action, which produces change, the element that most clearly distinguishes narration from description. Another fundamental element of narrative is time. Although time may also figure in description, time in a description is continuous (Todorov's "duration-time"), while in a narrative, time is divided into discontinuous units ("event-time").(25) Moreover, a narrative's discontinuous moments occur in a sequence, sometimes with causal links between units of time. Although a nonnarrative "exposition" may also present its elements in a sequence, as in a list,(26) Nelson Goodman observes that exposition lacks the underlying coherence of purpose that is characteristic of a narrative sequence.(27) Indeed, according to Hayden White, to "narrate" means to impose a structure and meaning on a sequence of events in a way that gives them moral coherence.(28) They may be implicitly endorsed, censured, or presented neutrally, depending on the moral system that underlies the account. This coherence distinguishes "history" from "annals" and "chronicles," neither of which situates the events within a comprehensive interpretive framework.

Structuralist critics, including Goodman and Seymour Chatman, stress a distinction between the time-sequence of events in a story and the time-sequence in the recounting of the story, a distinction that is captured by the terms "story time" and "discourse time," respectively.(29) Thus, the events may be recounted in a different order from that in which they "actually" happened, as in a flashback; or they may occupy different amounts of time, as when a narrator elaborates a fleeting occurrence into an extended account or collapses hundreds of years into a single sentence. Analogies with Chinese painting readily spring to mind, such as two contrasting approaches seen in handscrolls called Raising the Alms-bowl (Jie bo tu), which illustrate the Buddhist parable of the conversion of Hariti, Mother of Demons (Guizimu).(30) One type of composition focuses on the pivotal moment in which Hariti has exhausted her demonic forces in a vain attempt to rescue her youngest baby from imprisonment under the Buddha's begging bowl [ILLUSTRATION FOR FIGURE 6 OMITTED]; however, her subsequent repentance and conversion to Buddhism are foreshadowed by the appearance of the Buddha and his beatific assembly at the beginning (right end) of the handscroll composition [ILLUSTRATION FOR FIGURE 7 OMITTED]. Another type of composition follows what appears to be the "natural" order of the story. Various special warriors within the demon army are individually introduced [ILLUSTRATION FOR FIGURE 8 OMITTED], the onslaught on the bowl is shown in detail, a despairing Hariti is presented in her final moment of resistance, and the Buddha appears at the end of the scroll, waiting to receive her submission.

Arguing against the structuralists' differentiation of story time from discourse time, Barbara Herrnstein Smith points out that the distinction is premised on the unstated belief that it is possible for the two kinds of time to correspond exactly, which implies that all other relationships represent deviations by discourse time.(31) She contends instead that the ideal of a one-to-one correspondence is a false premise. Moreover, she proposes that the "story" itself is an ever-changing entity; there is no Platonic core underlying all the manifestations of a narrative. Rather, she maintains that there are only versions, or "retellings," which are constructed in relation to other versions for various specific purposes and within particular social contexts. The features of each version are determined by the interests, functions, and circumstances for which it was created, and these depend in turn on the backgrounds, expectations, and motivations of narrator and audience. Here, too, it is easy to think of analogies in Chinese pictorial art. For example, innumerable sets of pictures depicting the life of Confucius were made in China from the mid-fifteenth century onward [ILLUSTRATION FOR FIGURES 9, 10 OMITTED].(32) The various illustrated biographies of Confucius differ, often radically, in the number and type of episodes that are included in the set, media employed, quantities made, degree of artistry attained, and intended viewers. The purposes and ideological content of each pictorial "retelling" of Confucius's life story are also affected by the sociopolitical situation of the sponsor(s) or patron(s). Among the many differences between the two examples reproduced here, perhaps the most striking is that the 1989 government-approved rendition [ILLUSTRATION FOR FIGURE 10 OMITTED] shuns the considerable body of supernatural lore about Confucius, a prominent element in many premodern versions. The scene of his birth from a sixteenth-century example [ILLUSTRATION FOR FIGURE 9 OMITTED] shows the arrival of the sage not only being celebrated by a troupe of musicians hovering in the clouds but also marked literally by words on the newborn's chest, which proclaim him "created to stabilize the world [zhi zuo ding shi]." Auspicious portents of his birth and the superhuman efficacy attributed to him served to characterize Confucius as a being who was created, supported, and endorsed by the impersonal forces of heaven. Such "superstitious" elements could not be part of a retelling in which Confucius was to represent the model citizen in an atheistic communist utopia, particularly one whose Cultural Revolution had smashed the "four olds" (si jiu): old ideas, old culture, old customs, and old habits. Though shorn of his preternatural powers in the 1989 pictorial biography [ILLUSTRATION FOR FIGURE 10 OMITTED], Confucius nonetheless dwarfs other figures in many scenes, perhaps evoking the Great Leader of post-1949 China.

Narrative in Painting

In considering the implications of these discussions for the problem of defining narrative illustration in painting, I think it is necessary to specify a particular kind of content as a fundamental element of any proposed definition: the picture must depict or refer to a story, where "story" means the presentation of one or more events that occur in a sequence of time and bring about a change in the condition of a specific character. Even though the literary definition of a story requires that the account be overtly constructed (that is, it is not simply a statement of "fact"), it makes no difference for present purposes whether the story purports to be true or not, nor whether its narrator can be identified.(33) Nor does it matter whether the story is written down or orally transmitted. In these respects, the purely literary definitions diverge from those of the visual media. What does matter is that something happens.(34)

Another important element in the definition of narrative illustration is its function within the larger system of society: What does it do or aim to do? Having already limited my range of material to pictures that depict or refer to a story, I would further propose that narrative illustrations record, affirm, inform, instruct, indoctrinate, proselytize, propagandize, or even entertain the viewer; sometimes they may perform more than one of these functions simultaneously. Comparing narrative illustrations with paintings that perform other functions may help to bring them into clearer focus. For example, cultic icons are primarily intended to be worshiped or to provide a focus for the performance of rituals; portraits commemorate and solidify social relationships; decorative paintings adorn surfaces in order to please the patron's eye, impress his or her associates, and/or create an auspicious environment; and so forth.

A third aspect to consider in defining narrative illustration is the mode of presentation: How is the story communicated visually? The many possibilities here result from the interaction of three variables: format, compositional structure, and conceptual approach. Major formats for the illustration of narrative subjects in China include walls, tablets, screens, scrolls, fans, albums, and books. On occasion, even vessels and furnishings may be decorated with narrative themes. Each of these surfaces presents a particular set of formal possibilities and limitations related to the size, shape, material properties, and general configuration of the pictorial space. The format in which the picture is executed thus may influence the treatment of the subject in certain ways. In addition, the choice of format may also affect the viewer's interpretation of the illustration and its significance.(35)

The second variable, compositional structure, refers to the spatial arrangement of the pictorial elements. Here, a modified version of Vidya Dehejia's terminology for depictions of narrative in early Indian art may be useful.(36) The simplest construction is a monoscenic composition, which depicts a single moment within one structural frame. For example, Figure 11 shows the beautiful consort Yang Guifei cutting off a lock of hair to send Tang emperor Minghuang, by whom she has been temporarily banished from the palace, and this gesture induces him to restore her to favor. A single picture may also depict multiple moments in the story, resulting in a conflated composition if no characters are repeated within a unified pictorial space, or a synoptic one if repetition occurs without an indication of temporal sequence. In practice, it is very difficult to find any clear example of a conflated composition in Chinese pictorial arty but synoptic views are more common. The 1989 pictorial biography of Confucius [ILLUSTRATION FOR FIGURE 10 OMITTED] portrays him in each of several different episodes that are juxtaposed within a single frame.(38) Another type of synoptic scene often appears when the content of a dream or thought is illustrated in the same composition with the dreamer or thinker, as in Li Gonglin's well-known depiction for chapter seventeen of the Classic of Filial Piety (Xiao jing), which shows a minister sitting at home and thinking about serving his ruler.(39) In this type, however, the mental image is enveloped by a cloud that emanates from the head of the protagonist, somewhat separating the subscene from the main part of the picture. The two moments thus might seem to coexist in different realms rather than to occupy a shared space on equal terms.(40)

Sequential compositions present multiple episodes in a linear series, whether arranged in a continuous flow or segmented into more or less distinct units, which may be uniform or irregular in size.(41) Successive events may be depicted one after another on a common background, with the figures repeated as needed, as in some scrolls illustrating the Rhapsody on the Goddess of the Luo River (Luoshen fu, [ILLUSTRATION FOR FIGURE 12 OMITTED]).(42) In the section reproduced from the scroll in the Beijing Palace Museum, a landscape resembling stage scenery provides a continuous setting for three apparitions of the lovely goddess, who performs various alluring actions described in the rhapsody. More typically, sequential scenes are separated to some extent. In some illustrations, the individual episodes are marked off unambiguously by intervening passages of text, as in most renditions of Eighteen Songs of a Nomad Flute (Hujia shiba pai, [ILLUSTRATION FOR FIGURE 13 OMITTED]), which first recounts and then portrays each successive stage of the abduction, captivity, ransom, and return of Cai Yan (Lady Wenji), from 195 to 207 C.E.(43) In depictions of other subjects, the divisions between scenes are more subtly conveyed by the strategic placement of landscape elements, architecture, furniture, or empty space. This way of structuring a long series of events is used in a mid-sixth-century illustration of the Life of the Buddha [ILLUSTRATION FOR FIGURE 1 OMITTED], which is painted in three narrow registers on both sides of the pitched ceiling inside Dunhuang cave 290. The repetition of the segmentation device, whether text or pictorial motif, may create a visual consistency that reinforces the fundamental coherence of the "whole story."

One last category of compositional structure comprises large planar compositions, such as pictures on walls, whose size makes it possible to arrange multiple episodes in a variety of ways that are not limited to a simple linear sequence based on relative chronology.(44) For example, episodes that occur in the same location but at different points in the story may be grouped together, as in some later illustrations of events in the life of the Buddha. A detail from a mid-twelfth-century mural in the Yanshansi monastery [ILLUSTRATION FOR FIGURE 14 OMITTED] juxtaposes several unrelated scenes within a large and grand palace, but they depict events that occurred at various times. Other strategies for organizing a large rectangular composition include clustering events that involve the same people, dividing the pictorial space to reflect opposing forces of good and evil, and arranging elements to create a harmonious visual effect.(45) Sometimes the logic underlying the organization of the composition is difficult to explain at all. It may be possible to gain further insights into spatial logic by broadening the scope of examination to include examples from related cultures, such as the Ajanta cave murals in India and folding screens in Japan.(46)

The third element of a mode of presentation is its conceptual approach, a term I use to refer to the most general kind of relationship between a story and its pictorial treatment. Visual images may allude to or symbolize a story by representing its main character or characters, its setting, significant props, and so forth, without suggesting any particular event within it. For example, in The Red Cliff [ILLUSTRATION FOR FIGURE 2 OMITTED], the portrayal of men in a boat near a cliff refers to Su Shi's entire narrative about visiting the Red Cliff with his friends. A depiction may summarize or epitomize the story by presenting an episode or scene that serves to recall the whole, as "Yang Guifei cuts off a lock of hair" [ILLUSTRATION FOR FIGURE 11 OMITTED] brings to mind Yang Guifei's life in the palace before and after her dramatic offering of hair, the action that caused the emperor to reverse her banishment. These kinds of conceptual approaches evoke a story and its conventional interpretation without actually recounting it.

If more than one event is represented in pictures, the scenes are likely to be highlights excerpted from the story. For example, the several different versions of Goddess of the Luo River [ILLUSTRATION FOR FIGURE 12 OMITTED] present varying numbers of scenes showing the evolution of the poet's ill-fated romance with the goddess.(47) The more detailed and complete the pictorial account - most often seen in pictorial biographies such as that of the Buddha [ILLUSTRATION FOR FIGURE 14 OMITTED] - the closer the approach to what is often called pictorial narration. In the extreme case, which is rarely encountered in Chinese art, the story develops literally scene by scene in pictures [ILLUSTRATION FOR FIGURE 1 OMITTED].(48) The conceptual approach selected for illustrating any particular subject probably is related in some way to the function intended for the picture, or the nature of the space in which it is to be depicted. However, these suggestions by no means exhaust all the possibilities, and further research should yield useful insights into the significance of the conceptual approach chosen for various situations.

Having discussed format, compositional structure, and conceptual approach separately, and recognizing that each encompasses a range of possibilities, I return to the interaction of these three major variables in constituting the modes in which Chinese narrative illustration is presented. In theory, many combinations are possible, but in practice, a more manageable number account for the great majority of narrative representations. The most frequently encountered modes of presentation are: (1) monoscenic compositions that symbolize or epitomize the story in the handscroll, hanging scroll, and fan formats (for example, [ILLUSTRATION FOR FIGURE 2 OMITTED], The Red Cliff, and [ILLUSTRATION FOR FIGURE 11 OMITTED], Yang Guifei cuts off a lock of hair); (2) sequential compositions that excerpt highlights from the story or narrate it pictorially in the handscroll, album, and book formats ([ILLUSTRATION FOR FIGURE 12 OMITTED], Goddess of the Luo River, and [ILLUSTRATION FOR FIGURE 13 OMITTED], Eighteen Songs of a Nomad Flute); and (3) planar compositions that excerpt the highlights of the story in the mural format ([ILLUSTRATION FOR FIGURE 14 OMITTED], the later Scenes from the Lip of Buddha).

Based on my discussion in the entire preceding section, I would contend that the classification "narrative illustration" should be limited to pictures whose content can be related to an oral or written story in which something happens, and whose representation evokes that story to produce some kind of transformative effect on viewers. They may variously be taught, persuaded, reassured, distressed, entertained, and so on. To "illustrate a narrative" might mean as little as adding a thematically appropriate picture to embellish a self-contained story-in-words, or as much as presenting the story through pictures. If the picture itself does not portray specific action, its identification as a narrative illustration will depend on the viewer's ability to recognize the connection to a verbal story. For example, a viewer must have some knowledge of Su Shi's account of his outing to the Red Cliff in order to identify Li Song's work [ILLUSTRATION FOR FIGURE 2 OMITTED] as its symbolic representation, rather than simply a generic picture of men taking a boat ride.(49) By contrast, a composition that conveys its story pictorially (such as the detailed presentation of the Buddha's life in Dunhuang cave 290 [ILLUSTRATION FOR FIGURE 1 OMITTED]) is easily recognized as a narrative illustration even if the content of the story is unfamiliar to the viewer.

Paintings That Are Not "Narrative"

To make these proposed definitions clearer, I will briefly discuss some categories of picture that are intentionally excluded, although they have been called narrative illustrations by other scholars. First of all, I would rule out any pictorial exposition in which there is no action or reference to action. On these grounds, landscape panoramas and topographical pictures clearly do not qualify, even though the sites depicted may be provided with written identifications or longer textual accounts [ILLUSTRATION FOR FIGURE 15 OMITTED]. Nor would I include the many handscrolls and albums that purport to depict various non-Chinese peoples and their customs [ILLUSTRATION FOR FIGURE 16 OMITTED]. Often given titles like Pictures of Foreign Tributaries (Zhi gong tu), such illustrations are embellishments of texts that use descriptive generalizations to catalogue the exotic appearances and livelihoods of these populations.(50) Although specific characteristics may differ from one regional group to another, these differences are simply variations within the larger category of "otherness" that underlies both the texts and pictures. At this level, all of the non-Chinese peoples are the same and completely interchangeable with one another. The generic otherness of the non-Chinese peoples is invoked to underscore the implicit superiority of the Chinese and their customs. What distinguishes such pictures from narrative illustrations on the same theme is the absence of a specific story in which something happens.

Another group of pictures that I would not characterize as full-fledged narrative illustration is the representation of activities that have no protagonist, as in the paintings Going Upriver at Qingming [ILLUSTRATION FOR FIGURE 5 OMITTED] and Zhao Gan's Along the River during Winter's First Snow, the scroll that prompted John Hay to set out his three categories of narrative illustration.(51) Although these paintings show many individuals engaged in various activities, such as traveling, working, and relaxing, all these figures are generic, not characters in a story. Similarly, I would exclude depictions of generalized or recurrent activity, such as illustrations on the themes of Agriculture and Sericulture (Geng zhi tu, [ILLUSTRATION FOR FIGURE 17 OMITTED]), which have been produced in considerable numbers from the twelfth century onward.(52) These portray an idealized sequence of the separate procedures involved in rice farming (in twenty-one scenes) and silk production (in twenty-four scenes), traditionally considered the archetypal occupations for men and women, respectively. Each stage of activity is pictorialized and, in most sets, accompanied by a eulogistic poem. Rather than portraying the experiences of specific individuals in a particular time and place, however, the illustrations present a timeless and universalized ideal of a harmoniously ordered livelihood.

Likewise depicting generic activities, but much earlier in date, are the representations of funerary rites and processions on Han tomb walls. These symbolize the deceased's journey from earth to a realm of eternal existence, encompassing both a movement in space and a change in status [ILLUSTRATION FOR FIGURE 18 OMITTED]. Wu Hung has described such scenes as "transitional narratives" because they come between static tableaux relating to death and to the afterlife and, in his view, turn the entire pictorial program into a grand narrative.(53) This observation is valid in a metaphorical sense, but the pictures themselves are highly conventionalized scenes of cavalcades, sacrifices, and feasts, and they seem not to represent particular events or specific individuals. On these grounds, they do not fit a more restrictive definition of narrative illustration.

Finally, I would also set apart the pictures made for prescriptive texts in which action does not take place, such as the Classic of Filial Piety, Ladies' Classic of Filial Piety (Nu xiao ting), and Admonitions of the Court Instructress (Nu shi zhen).(54) The first two offer a systematic exposition of norms and ideals for male and female conduct, respectively, with eighteen short chapters each; the last is a more impressionistic and randomly organized collection of maxims and advice for ladies of the court. Such works teach the principles that underlie familial and social relationships. If an illustrator depicted a social encounter in order to suggest its associated principle, the result might be visually indistinguishable from an illustration of a story. In the Admonitions scroll, such a strategy, is used to embody the pronouncement that a woman must gracefully accept the inevitability that her appeal will decline and her husband will come to favor other women [ILLUSTRATION FOR FIGURE 19 OMITTED]; the picture shows a man holding up his hand as if to repel the woman facing him, whose presumed distress is signaled by the fluttering of her scarves. Although the image depicts an action involving individuals, it is not linked to a story, and thus is not a narrative illustration. On the other hand, some prescriptive texts incorporate narrative elements in the form of parables or biographical anecdotes, which serve to elucidate points of discussion. For example, in the Ladies' Classic of Filial Piety, the story of Wise Lady Fan is introduced to demonstrate an appropriate way for women to use their wisdom,(55) and the story of Lady Qiang is recounted to emphasize that a reputation for unimpeachable chastity is even more important than life itself. As implied by the illustration of Lady Qiang on the flooded terrace [ILLUSTRATION FOR FIGURE 20 OMITTED], Lady Qiang chooses to drown rather than go to safety with a man other than her husband.(56) Pictorializations of these embedded narratives do accord with the criteria by which I define narrative illustration.

It is important to note that pictures in the excluded groups often exhibit exactly the same modes of presentation as narrative illustrations. Many are handscrolls containing excerpted highlights in a sequential composition, and many others are monoscenes that epitomize the story. On the one hand, this visual similarity reinforces my contention that the mode of presentation cannot be used as the primary criterion for deciding which pictures are narrative illustrations. Structure alone does not determine whether a picture is narrative, because the content and purpose of the work must also be those defined as narrative. On the other hand, the frequency with which nonnarrative material is pictorialized using the same modes of presentation as narrative subjects suggests that the two kinds of pictures are intended to have similar effects on the viewer. Idealized pictorial accounts that are presented as normative, such as Agriculture and Sericulture [ILLUSTRATION FOR FIGURE 17 OMITTED] and Pictures of Foreign Tributaries [ILLUSTRATION FOR FIGURE 16 OMITTED], serve to inculcate a worldview that is based on a fundamental division between the civilized center, China (zhong), and a barbarian exterior (wai). To be Chinese, one must perform proper rituals (li) and observe the norms of filial piety (xiao). Such propositions also underlie the illustrations of many explicit stories, such as Eighteen Songs of a Nomad Flute [ILLUSTRATION FOR FIGURE 13 OMITTED], which follows Lady Wenji of the Han from the time of her capture by northern nomads, through twelve years of her life as wife and mother in the steppe grasslands, to her ransom and return to China, an outcome that required her to abandon her "barbarian" family. In a sense, any purportedly objective pictorial assertion of "things that are" (such as Fig. 5, Going Upriver at Qingming) or "things that should be" (such as Fig. 17, Agriculture) has the same import as an illustration of a story whose resolution reaffirms the cosmic and social orders. Whether they depict generic themes or illustrate particular stories, expository illustrations provide visible embodiments of core values in Chinese culture.

I would also venture to suggest that traditional categorizations of painting demonstrate both the tacit acceptance of social and political norms and a lack of interest in how they are maintained. As discussed above, traditional texts on painting divide their material into categories based on motif-level subject matter, such as Beautiful Women, Peasants, and Landscapes. There is no apparent concern for discriminating among the various purposes of pictorial art or the forms of its presentation, which may help to explain why "narrative illustration" was not constructed as a critical category in traditional texts on Chinese pictorial art. However, with a more restrictive definition of Chinese narrative illustration with which to identify a consistent body of examples, it should be possible to analyze the evolution of subject matter and illustrative strategies over the centuries, and to assess the impact of diverse factors such as religious movements and socioeconomic changes. A better grasp of the scope and purposes of narrative illustration would also provide firmer grounding for studies of individual works, different types of illustration, and iconological program. Finally, a clearer understanding of the position of narrative illustration within Chinese visual culture as a whole would contribute toward a revised and more balanced view of painting practice in late imperial China. Although critical and scholarly attention has been focused for nearly four centuries on the purportedly self-referential, self-expressive paintings of the literati, recent scholarship now amply demonstrates that other kinds of painting were also important.(57) Along with portraits, religious icons, decorative paintings, and misty vistas, narrative illustrations were a significant component of Chinese pictorial art. They were made in great numbers and at every level of quality, and their social functions are at least as interesting as those of literati painting.

Appendix

Characters for Chinese Terms

dao shi fan zu fo dao gu shi (or) gu shi hua (or) gu shi shi gui shen Guizi mu hua iji dozu ji jing li ren wu shi (history) shi (matter) shi (event) shi nu si bu si jiu tian jia wai xian fo xiao xu shi hua xu shi xu shi shi zhi zuo ding shi zhong zi

Notes

A preliminary version of this paper was prepared for the Open Session on Asian Art, organized by Susan Huntington, at the College Art Association conference in Boston, 1996. I wish to thank Gene Phillips and Robert Bagley for their insightful comments and helpful suggestions on the revised versions. I am also grateful to the Smithsonian Institution and American Council of Learned Societies for grants in support of my research on narrative illustration in China.

1. Nonetheless, China is omitted from the ostensibly global discussion of "Narrative Art" in The Dictionary of Art, ed. Jane Shoaf Turner, vol. 22 (London: Macmillan, 1996), 510-23. This omission is all the more surprising given that The Dictionary of Art otherwise devotes a great deal of space to the arts of China.

2. The term hua refers not just to paintings done with a brush but also to picture making in a range of media such as images carved or incised in stone, molded in clay, or printed with woodblocks.

3. On the terms for narrative poetry, see Dore J. Levy, Chinese Narrative Poetry (Durham, N.C.: Duke University Press, 1988), 2, 152 n. 3.

4. Andrew Plaks, "Towards a Critical Theory of Chinese Narrative," in Chinese Narrative: Critical and Theoretical Essays, ed. Andrew Plaks (Princeton: Princeton University Press, 1977), 310.

5. As indicated in the App., "Characters for Chinese Terms" (see above), two different characters for shi may be used in the compound gu shi. The second form of the compound, more often employed in modern writing to mean "fiction," was used in the broader sense of "story" from the late Ming onward (for example, by Xie Zhaozhe in Wu za zu; see Sewall Jerome Oertling, Painting and Calligraphy in the Wu-tsa-tsu [Ann Arbor, Mich.: Center for Chinese Studies, University of Michigan, 1997], 123, 194). Some modern art historians, such as Xu Bangda, use this form of shi in the further compound gu shi hua (as in "Song ren hua Ren wu gu shi ying ji Ying luan tu kao" [The Anonymous Song Figural Narrative should really be Welcoming the Imperial Carriage: A Study], Wen wu, 1972, no. 8, 61). Others, such as Shih Shou-chien, prefer the first shi in that compound (as in his book Fengge yu shibian: Zhongguo huihua shi lunji [Taipei: Yunchen wenhua, 1996], 115).

6. In charting the categories named in Chinese texts on painting from the Period of Disunion through the Song period, Lothar Ledderose argued that differences in groupings imply changes in the conceptualization of painting over time; see Ledderose, "Subject Matter in Early Chinese Painting Criticism," Oriental Art 19 (1973): 69-83.

7. For particularly useful discussions, see Levy (as in n. 3); various articles in Plaks (as in n. 4); Sheldon Hsiao-peng Lu, From Historicity to Fictionality: The Chinese Poetics of Narrative (Stanford: Stanford University Press, 1994); Patrick Hanan, The Chinese Vernacular Story (Cambridge, Mass.: Harvard University Press, 1981); and Robert Hegel, "Traditional Chinese Fiction: The State of the Field," Journal of Asian Studies 53, no. 2 (May 1994): 394-426.

8. A. John Hay, "Along the River during Winter's First Snow: A Tenth-Century Handscroll and Early Chinese Narrative," Burlington Magazine 114, no. 830 (May 1972): 292-303.

9. Ibid., 298.

10. Other scholars have understood this type simply to include illustrations based on a literary source, but Hay's remarks suggest something less straightforward, more like pictures that display a poetic sensibility, such as the Cleveland Museum's album leaf Watching Deer by a Pine-shaded Stream, attributed to Ma Yuan (active ca. 1190-ca. 1225); reproduced in James F. Cahill, Chinese Painting (Geneva: Skira, 1960), 83 (under the title Scholar and Servant on a Terrace).

11. Hay (as in n. 8), 303, fig. 37. For more extended discussion of this painting, see also Julia K. Murray, Ma Hezhi and the Illustration of the Book of Odes (Cambridge: Cambridge University Press, 1993), 63-65.

12. Wen C. Fong, Beyond Representation: Chinese Painting and Calligraphy, 8th-14th Century (New York: Metropolitan Museum of Art, 1992), chap. 1, esp. 21.

13. See Yu-kung Kao, "Chinese Lyric Aesthetics," in Murck and Fong, 47-90.

14. Wu Hung, The Wu Liang Shrine (Stanford, Calif.: Stanford University Press, 1989), esp. 132-40.

15. The term icon is used more broadly in art historical scholarship than Wu's characterization implies, and many compositions that are not frontal or bilaterally symmetrical are so categorized. For relevant discussions, see Sixten Ringbom, Icon to Narrative: The Rise of the Dramatic Close-Up in Fifteenth-Century Devotional Painting, rev. ed. (Doornspijk, the Neth.: Davaco, 1984); and Hans Belting, Likeness and Presence: A History of the Image before the Era of Art, trans. Edmund Jephcott (Chicago: University of Chicago Press, 1994). Wu's argument is valid for a particular kind of icon, the cultic image, a sculptural or painted image that serves as the focus of liturgical or devotional activities.

16. Wu (as in n. 14), 133, 360 n. 60.

17. Maxwell K. Hearn, "The 'Kangxi Southern Inspection Tour': A Narrative Program by Wang Hui," Ph.D. diss., Princeton University, 1990, 260. The twelve-scroll set projects a highly constructed account of the Kangxi emperor's 1689 Southern Inspection Tour, and each segment is provided with an explanatory text. The pictures may indeed be called narrative illustration according to criteria I discuss below; here I simply disagree that landscape elements in themselves may "narrate."

18. For a visual medium such as painting or film, "showing" would be more appropriate and specific.

19. This procedure is nicely demonstrated and explained by Wu Hung (partially following Jerome Silbergeld) in The Double Screen: Medium and Representation in Chinese Painting (Chicago: University of Chicago Press, 1996), 57-61. In a continuous composition, there are no predefined sections as such; rather, the viewer determines which portion of the painting to frame between the two rolls of the scroll.

20. This translation is proposed in Valerie Hansen, "The Mystery of the Qingming Scroll and Its Subject: The Case against Kaifeng," Journal of Sung-Yuan Studies 26 (1996): 196-97; she takes the substance of her argument from Hsiao Ch'iung-jui, "A New Look at the Title of Ch'ing-ming shang-ho t'u," in Proceedings of the International Colloquium on Chinese Art History, 1991, vol. 1 (Taipei: National Palace Museum, 1992), 111-37.

21. For an introduction to the scroll, see Hansen (as in n. 20), 183-200; Linda Cooke Johnson, "The Place of Qingming shanghe tu in the Historical Geography of Song Dynast)' Dongjing," Journal of Sung-Yuan Studies 26 (1996): 145-82; and Julia K. Murray, "Water under a Bridge: Further Thoughts on the Qingming Scroll," Journal of Sung-Yuan Studies 27 (1997): 99-107. The classic detailed study remains Roderick Whitfield, "Chang Tse-tuan's Ch'ing Ming Shang Ho T'u," Ph.D. diss., Princeton University, 1965. For a complete reproduction and extensive documentation of the scroll, see Zhongguo lidai huihua: Gugong bowuyuan cang hua ji, vol. 2 (Beijing: Wenwu Press, 1981), 60-83, app., 9-12.

22. For this insight I thank one of the anonymous readers of the initial version of this article for the Art Bulletin. This reader also reminded me of Ernst Gombrich's useful and relevant observations on the problems of defining genres and interpreting earlier works even from one's own culture in his essay "Aims and Limits of Iconology," repr. in Symbolic Images: Studies in the Art of the Renaissance, 2d ed., vol. 2 (London: Phaidon, 1978), 1-25.

23. Tzvetan Todorov, Genres in Discourse, trans. Catherine Porter (Cambridge: Cambridge University Press, 1990), 17. In considering the validity of this proposition for painting in general, Quitman E. Phillips has argued that the categories meaningful to literary men (who produced the critical discourse) probably differed considerably from those of practicing painters, particularly when the two came from separate social groups (private communication, September 1996).

24. A brief but explicit discussion of the lack of consensus on the definition of "narrative" is given in Marilyn A. Lavin, The Place of Narrative: Mural Decoration in Italian Churches, 431-1600 (Chicago: University of Chicago Press, 1990), 1-3, 9. The meaning of "narrative" in writings on Western art also varies according to what is being set in opposition to it. In addition to icons, the nonobjective representations of modern art have been characterized as nonnarrative, their emergence even marking the "death" of the 19th-century narrative.

25. Todorov (as in n. 23), 28.

26. An analogy in painting might be the presentation of figures in a procession.

27. Nelson Goodman, "Twisted Tales; or, Story, Study, and Symphony," in Mitchell, 99-115.

28. Hayden White, "The Value of Narrativity in the Representation of Reality," in Mitchell, 1-23, esp. 22-23. White developed his definitions in relation to European historiography; thus, their appropriateness for other areas may be open to question.

29. Goodman (as in n. 27); Seymour Chatman, "What Novels Can Do That Films Can't (and Vice Versa)," in Mitchell, 117-36. See also Lavin (as in n. 24), 9-10.

30. For illustrations and further discussion, see Julia K. Murray, "Representations of Hariti, the Mother of Demons, and the Theme of 'Raising the Alms-bowl' in Chinese Painting," Artibus Asiae 43, no. 4 (1981-82): 253-84.

31. Barbara Herrnstein Smith, "Afterthoughts on Narrative," in Mitchell, 209-32.

32. See Julia K. Murray, "The Temple of Confucius and Pictorial Biographies of the Sage," Journal of Asian Studies 55 (1996): 269-300; and idem, "Illustrations of the Life of Confucius: Their Evolution, Functions, and Significance in Late Ming China," Artibus Asiae 57, nos. 1-2 (1997): 73-134.

33. On the other hand, Richard Brilliant has suggested that a narrative picture may have as many as three narrators: the artist who presents the story in pictures, the viewer who "reads" it, and (sometimes?) the protagonist within it; see Brilliant, Visual Narratives: Storytelling in Etruscan and Roman Art (Ithaca, N.Y.: Cornell University Press, 1984), 16-17. This point is further developed by Peter Holliday in his introduction to Narrative and Event in Ancient Art, Cambridge Studies in New Art History and Criticism (New York: Cambridge University Press, 1993), 3-13.

34. Drawing on scholarship in communication theory, Wolfgang Kemp proposes that the transformation that occurs (that is, exactly what happens) must be significant to the narrator and/or audience; see Kemp, "Narrative," in Critical Terms for Art History, ed. Robert S. Nelson and Richard Schiff (Chicago: University of Chicago Press, 1996), esp. 65-66.

35. Wu (as in n. 19) provides an illuminating discussion of some of the issues raised by paintings on screens and in handscrolls.

36. Dehejia. Another useful discussion, based on European art, appears in Lew Andrews, Story and Space in Renaissance Art (Cambridge: Cambridge University Press, 1995), 120-26. Dehejia's categories are criticized and additional terms are defined for Japanese pictorial art in Quitman E. Phillips, "The Price Shuten Doji Screens: A Study of Visual Narrative," Ars Orientalis 26 (1996): 1-21.

37. Pao-chen Chen has argued that an incised stone slab at Wu Liang Shrine depicts five moments in the story of Jing Ke's attempt to assassinate the king of Qin in a single composition, without any repeated characters; see Pao-chen Chen, "The Goddess of the Lo River: A Study of Early Chinese Narrative Handscrolls," Ph.D. diss., Princeton University, 1987, 113-17. However, the composition is more easily explained as depicting the moment that Jing Ke's attempt has failed.

38. Kohara Hironobu ("Narrative Illustration in the Handscroll Format," in Murck and Fong, 255) would call both conflated and synoptic compositions iji dozu ("different moments, same picture"). As Dehejia notes (382 n. 18), the terminology used by writers on Western art does not necessarily distinguish whether or not the protagonist is repeated. However, it seems useful to make such a distinction. Concerning the synoptic mode specifically, Dehejia emphasizes that the temporal sequence of the episodes is not indicated by formal means.

39. The scene is reproduced and discussed in Richard M. Barnhart et al., Li Kung-lin's Classic of Filial Piety (New York: Metropolitan Museum of Art, 1993), 147-49.

40. An interesting discussion of Western conventions for representing dreams appears in Sixten Ringbom, "Some Pictorial Conventions for the Recounting of Thoughts and Experiences in Late Medieval Art," in Medieval Iconography and Narrative: A Symposium, ed. Flemming Anderson (Odense: Odense University Press, 1980), 38-69.

41. Segmented compositions largely correspond to Dehejia's linear category, as well as to Kurt Weitzmann's cyclical narrative; see Weitzmann, Illustrations in Roll and Codex, rev. ed. (Princeton: Princeton University Press, 1970), 17-33.

42. Kohara (as in n. 38, 256-62) has argued that continuous narrative in China appeared as the result of omitting texts between scenes, as seen in a comparison between the Beijing Luoshen scroll and the Liaoning version, which includes passages of text interspersed among the pictorial motifs. Moreover, he points out that continuous narrative did not arouse much interest in China, whereas in Japan it developed to a cinematic sophistication in the celebrated Shigisan engi emaki. The Beijing Luoshen scroll is fully reproduced in Zhongguo lidai huihua: Gugong bowuyuan cang hua ji, vol. 1 (Beijing: Wenwu Press, 1978), 2-19. The Liaoning version appears in Liaoning sheng bowuguan cang hua ji, vol. 1 (Beijing: Wenwu Press, 1962), 1-12.

43. For a complete reproduction of this particular version, see Wen C. Fong and Robert A. Rorex, Eighteen Songs of a Nomad Flute (New York: Metropolitan Museum of Art, 1974); others are discussed and reproduced in Robert A. Rorex, "Eighteen Songs of a Nomad Flute: The Story of Ts'ai Wen-chi," Ph.D. diss., Princeton University, 1974.

44. This group corresponds to Dehejia's "narrative networks" (388).

45. These different kinds of logic for the organization of large planar pictorial surfaces are discussed in Wu Hung, "What Is Bianxiang?" Harvard Journal of Asiatic Studies 52, no. 1 (June 1992): 111-92, esp. 159-69.

46. For a useful discussion of large planar compositions in the Ajanta murals, see Dehejia, 388-92; for Japanese folding screens, see Phillips (as in n. 36).

47. The contents of several extant versions of the theme are analyzed in great detail in Pao-chen Chen (as in n. 37).

48. Robert Bagley has suggested (personal communication, January 1997) that the extreme case is realized in Sergei Eisenstein's film notebooks, where the composition of figures, costumes, and scenery is drawn for every scene and from each camera angle. In terms of scenes, one of Eisenstein's notebooks is as complete as the corresponding portion of the finished film.

49. Nonetheless, viewers with relatively superficial knowledge could identify the subjects of symbolic pictures that were ubiquitous in their visual culture. Depictions of the Red Cliff outing even appeared on decorative objects. For a stimulating discussion of this point, see Craig Clunas, Pictures and Visuality in Early Modern China (London: Reaktion Press, 1997), esp. chap. 2. Gombrich (as in n. 22, 3) also comments on the importance of "the beholder's share," that is, the knowledge and experience that the viewer brings to a work.

50. One group of Qing works within this genre has recently been studied by a social historian; see Laura Hostetler, "Chinese Ethnography in the Eighteenth Century: Miao Albums of Guizhou Province," Ph.D. diss., University of Pennsylvania, 1995.

51. See above at n. 8.

52. A lucid introduction to this subject is given by Thomas Lawton, Chinese Figure Painting (Washington, D.C.: Smithsonian Institution, 1973), cat. nos. 7, 8; also useful is the bibliography and discussion in Joseph Needham, Science and Civilisation in China, vol. 4, Physics and Physical Technology', p. 2, Mechanical Engineering (Cambridge: Cambridge University Press, 1965), 166-69.

53. Wu Hung, "Beyond the 'Great Boundary': Funerary Narrative in the Cangshan Tomb," in Boundaries in China, ed. A. John Hay (London: Reaktion Books, 1994), 81-104, esp. 91.

54. For an introduction to the Classic of Filial Piety and complete reproductions of the earliest surviving illustrations, see Barnhart et al. (as in n. 39). For the Ladies' Classic, see Julia K. Murray, "Didactic Art for Women: The ladies' Classic of Filial Piety," in Flowering in the Shadows: Women in the History of Chinese and Japanese Painting, ed. Marsha Weidner (Honolulu: University of Hawaii Press, 1990), 27-53. For the Admonitions, see Hsio-yen Shih, "Poetry Illustration and the Works of Ku K'ai-chih," Renditions 6 (Spring 1976): 6-29; and Osvald Siren, Chinese Painting: Leading Masters and Principles, vol. 3 (New York: Ronald Press, 1956), pls. 11-15.

55. Lady Fan, a consort of King Zhuang of Chu, used her wisdom to enable the state to prosper by opening the king's eyes to the shortsighted and selfish actions of his chief minister, Yu Qiuzi. As a result, the king began taking counsel with the sage Sun Shuao and soon was able to achieve hegemony over the other feudal princes; see Murray (as in n. 54), 41, fig. 11.

56. Ibid., 32. Lady Qiang became stranded on a terrace amid rising floodwaters, and her husband, King Zhao of Chu, sent an envoy to rescue her. In his haste to reach her before the terrace was engulfed, the envoy forgot to bring the official tally that verified his mission. Rather than commit an impropriety, Lady Qiang refused to accompany him - and drowned.

57. I am thinking particularly of studies such as Richard M. Barnhart, Painters of the Great Ming: The Imperial Court and the Zhe School (Dallas: Kimbell Museum of Art, 1993); James E Cahill, The Painter's Practice: How Artists Lived and Worked in Traditional China (New York: Columbia University Press, 1994); idem, The Lyric Journey, The Edwin O. Reischauer Lectures, 1993 (Cambridge, Mass.: Harvard University Press, 1996); Clunas (as in n. 49); Richard Vinograd, Boundaries of the Self: Chinese Portraits, 1600-1900 (New York: Cambridge University Press, 1992); Marsha Weidner, ed., Latter Days of the Law (Honolulu: University of Hawai'i Press, 1994); and Wu Hung (as in n. 19).

Frequently Cited Sources

Dehejia, Vidya, "On the Modes of Visual Narration in Early Buddhist Art," Art Bulletin 122, no. 3 (Sept. 1990): 373-92.

Murck, Alfreda J., and Wen C. Fong, eds. Words and Images: Chinese Poetry, Calligraphy, and Painting (New York: Metropolitan Museum of Art, 1991), 247-66.

Mitchell, W.J.T., ed., On Narrative (Chicago: University of Chicago Press, 1981).

Julia K. Murray is professor of art history at the University of Wisconsin, where she also chairs an interdisciplinary East Asian Studies Program. She received her Ph.D. in Chinese art and archaeology from Princeton University and is the author of books and numerous articles on the arts in traditional China [Department of Art History, University of Wisconsin, Madison, Wis. 53706].

COPYRIGHT 1998 College Art Association
COPYRIGHT 2004 Gale Group