Toward a Structural Analysis of Chinese Landscape Painting Author(S): Wen C

Home , Chinese painting

Toward a Structural Analysis of Chinese Landscape Painting Author(s): Wen C. Fong Source: Art Journal, Vol. 28, No. 4 (Summer, 1969), pp. 388-397 Published by: College Art Association Stable URL: http://www.jstor.org/stable/775311 Accessed: 14/09/2010 16:08

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=caa.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

College Art Association is collaborating with JSTOR to digitize, preserve and extend access to Art Journal.

http://www.jstor.org Wen C. Fong

Toward a Structural Analysis of Chinese Landscape Painting

A critical and historical study of Chinese painting has long suffered from the lack of an acceptable method for dating paintings by style. Mr. Li Lin-ts'an of the Na- tional Central Museum in Taiwan has published a series of three articles dealing with this problem.' In "Rules for Dating Chinese Painting," Mr. Li lists the study of material, technique, period style, personal style, signature and colophon, and catalogue description as "six methods for In "The of Ink-bamboo he dating."2 Dating Painting," 4,:t:::I samples eighty-eight bamboo paintings attributed to famous masters, from the tenth through the eighteenth centuries, and attempts diagrams illustrating the technical development of the bamboo stalks and leaves. These diagrams show, for instance, "ring joint technique," "plain joint technique," "dotted joint technique," "natural leaves," "star-shaped leaves," "feathered-shaped leaves," etc.3

Mr. Li's classification of motifs and techniques con- Fig. i. Detail of Mirror, early 8th century, Sh6s6in Treasury, Nara, Japan. tinues in the tradition of the "Mustard Seed Garden at the base of the nail. Mr. Painter's Manual" of the seventeenth century. By arrang- finger Chang's demonstration is in at least two ing his motifs chronologically, he hopes to establish cer- methodologically meaningful respects: his tain criteria for dating. His demonstration suffers, how- firstly, since wall-painting examples are archaeologi- ever, from two serious difficulties: firstly, he is faced with cally discovered and dated, he does not have to concern the problem of circularity: that of having to date a style himself with the problem of later copying and imita- by means of examples which themselves need to be tion;5 secondly, by describing not only the shape of the nail but also how it is on the he dated; secondly, motifs are easily imitated and perpet- finger grown finger-tip, uated in the copies. Even if we assume that all of Mr. is observing a morphological detail, which, if verified by Li's samples are correctly dated and authentic, his dia- all archaeologically dated examples, may constitute a period that grams of motifs merely illustrate, as in the "Mustard characteristic governs all figure paintings of Seed Garden Painter's Manual," the technical traditions that period. There is, from the point of view of descrip- tive a difference between Li's of the various masters' manners. They provide no clue method, significant "star- and that recede into the for the actual dating of a painting, or a copy, in the man- shaped leaves" Chang's "nails fin- ner of a given master. ger-tip"; the former merely identifies a two-dimensional

I SLi Lin-ts'an, "Chung-kuo-hua tuan-tai-yen-chiu-li [Rules Chinese "in" An interesting appendix to Li's "Rules for Dating for Dating Painting], Studies Presented to on His The Chinese Painting" shows seven illustrations by the fa- Tung Tso-pin Sixty-first Birthday," Bulletin of Institute mous contemporary Chinese painter Chang Ta-chien (or the of History and Philology, Academia Sinica, extra volume no. Chang Dai-chien), demonstrating the development of the 4, 1961, pp. 551-582; "Chung-kuo-hua- shih ti Chinese Paint- drawing of the hand as seen in Buddhist wall-paintings ch'ung-chien [the Reconstruction of at Tun-huang.4 Mr. Chang notes, for instance: in the ing History]." Ta-lu tsa-chih, XXXI, no. 5, 1965, pp. 1-5; on Northern Wei period, the drawing of fingers shows nei- "Chung-kuo mo-chu-hua-fa ti tuan-tai-yen-chiu [Study the Ink-bamboo The ther joints nor nails; during the reign of K'ai-yiian, (713- Dating of Chinese Painting]," Na- 742), the hand is plump and soft and has "nails that re- tional Palace Museum Quarterly, Vol. I, no. 4, 1967, pp. cede into the finger-tips": during the middle T'ang, the 25-79. nails "grow over the finger-tip, tapering to a rounded 'Op. cit., pp. 554ff. point"; in early Sung hands, there is a short straight line ' Op. cit., betweenpp. 78-79. SOp.cit., figures14-17. ' Mr. Fong teaches Chinese art at Princeton. This paper is There is, of course, the problem of repair and repainting a statement he wrote some time ago in preparation for a in wall-painting, which often complicates the task of sty- book on Chinese landscape painting. M listic analysis.

ARTJOURNAL XXVIII 4 388 . i:o i* :i o i::i:i oio o* :j j:j: ii:i? :70 ::: :7 ...... :: '.",,X ?::? -:-:-:K ?.?? ..,. a.. ? - . . I,:." iiiiii~:iiiiiiiiiiiii:iiiiiiiii lil~i~ i i iilii~i : lii~ii~iieiiiiiiiiiPO',"iii R I I ::1, .? ;:,-.??:??:.;-:.. ;?.... : :..- -.- ...,...i:,:...- :..:.,::.::::?j:...?<-..,? , , I:i?r::::~ ~~::::-?l:III::::::::II::i::-:::::~~~7 7 7 : 7 77l- ,.. . "?:.; , X 1. .::::???, :?i:::MI'. Q .: -11 ? ..... ::.:: ::: I I I :i? ,,,?i??::M'? :..:"...... ?- - . .k?.. ? .. ?.- . 1;? I "'ii'ijir K?*::::I .1I - I * : I 1. .1.? , . I..

sp .1.1?%? - .I ..--m. l. ii~i' -. ???.? ? ---.. ?..,., " ? -1 ?s~-lI? ~? . a Ii~n~ii~~2~i~~#~k~l, ?1 ~?PP'~~ ?,1," 1 , -? . :1 .iiE~ifiii'~r3ii,; , ? ,?. ,i *?- ? ? - &? '' ,? I ? - ?.,., , -?. , - .1,.1 - ., ,.?*-, - , - . ",I -? ,..- --11, iiiii*ii0k?. - ?-.?.?, K '* , ?, " ,-,, ? I , - i ? ?, - ? - ?,, -T , `-?, -,I?A.? ?:? -o ? - I .-..?.,. ---:?i.:iii-.'?*:::i--.l?l~i -?,.,*",:.,; " 'iil~ii:~9~i~?is~Pe~...... i~iii

?X. , .? ?.. w 1.1. -, :.*?:::;:?. ..; _- .? - ,x

I~iilli~fii''~~'i~~l~ij~i~SWIM~ :?Ni:il?? : ?. . . , ? .~ii~~i~'.~...... ::??:?:, ...... ---- :: :.? .. .,: .?:m.::::;::::. 7.,?: '*

......

Fig. 2. Landscape with Riders on an Elephant, 8th century, Sh6s6in Treasury, Nara, Japan. motif without indicating its structural relationship with Fig. 3. Scenes Illustrating the Life of Buddha, 9th century, silk banner from Tun- British Museum. other parts of the painting, while the latter, by showing huang. concern for the relationship between two motifs, the nail painter's observations far enough to describe the chang- and the finger-tip, begins to describe a structural configu- ing structural configurations of the hands of the various ration. Mr. Chang, however, did not carry his sensitive periods. For the early-T'ang period, he merely noted that

389 Fong: Toward a Structural Analysis of Chinese Landscape Painting "the brushwork is gentle and supple; it is capable of describing some of very difficult hand gestures." When we try to identify and describe an individual manner, we usually note its special form elements, motifs and techniques on the one hand, and its unique expressive qualities on the other. When we try to classify a style, however, we interpret the stylistic peculiarities of an individual work as specific solutions to generic structural problems.6 While neither motif nor quality gives adequate evidence for fixed positions in time, morphological analysis dealing with successive visual structures in history provides a key for dating a painting. From the structural point of view, before a painting of whatever form elements, motifs or techniques can express a certain philosophy or mental outlook, it resents first a solution to the problem of delineation, modelling and composition. Form relationships seem to change without direct relationship to meaning.' An obvious example is that despite the Chinese painter's avowed lack of interest in "form-likeness," they nevertheless successfully mastered illusion in painting. Every Chinese painting is at once representation, decoration and abstraction; it is the arranging of form elements to create a semblance of nature that exists in its own right. From the representational point of view, Chang's illustrations show the development of the drawing of the hand from a two-dimensional silhouetted shape to a three-dimensional and fully articulated, grasp- ing organ: each stage is characterized by certain struc- Fig. 4. Autumn, Eastern Mausoleum at Ch'ing-ling, Eastern Mongolia, ca. 1030. tural problems and solutions. The Northern Wei hand was neither joints nor nails, because it is conceived as a ponents. The early T'ang emphasis on complex hand-ges- silhouetted form without organically differentiated com- tures reflects an interest in conquering the technical difficulties in representing a hand. Both the short nails "that recede into" the finger-tip and long ones "that grow SIn his well-known article on "Style" (in Anthropology over" the tip show the middle-T'ang concern for organic Today, edited by A. L. Kroeber, Chicago 1953, pp. 287- details. Finally, the "short straight line at the base of the 312), Meyer Schapiro notes that the word "style" is gen- finger-nail" seen in the early-Sung paintings represents an erally used to describe three different aspects of a work of increasing interest not only in modelling but in decora- art: 1) form elements or motifs, 2) form relationships, 3) tive stylization as well; both tendencies are typical of the qualities (including an over-all quality which we may representational art of the tenth century. call "expression"). Following this definition, we might II say that the Chinese critics have traditionally emphasized form elements and qualities, but neglected form relation- The modern notion of a "period style" is based on ships, in their stylistic descriptions. See my article, "Chi- Wolfflin's famous assumption that "every artist finds cer- nese Painting: A Statement of Method," Oriental Art, tain visual possibilities to which he is bound. Not every- new series, vol. IX, no. 2, summer 1963, pp. 73-78; also thing is possible at all times."8 As an abstract concept my article, "The Problem of Ch'ien Hsiian," The Art which deals with the structural principles rather than Bulletin, XLII, September 1960, p. 188. specifically identifiable motifs and qualities of a work of 'In The Shape of Time (Yale University Press, 1962), art, however, a "period style" exists only as an idea. Professor George Kubler writes: "the structural forms ' can be sensed independently of meaning. We know from Heinrich Wilfflin, Principles of Art History, 1915, trans- linguistics in particular that the structural elements un- lated by M. D. Hottinger, Dover paperback edition, New dergo more or less regular evolutions in time without York, p. 11. Italics added. George Kubler points out that relation to meaning. . . . Similar regularities probably the limits of the existing state of knowledge "confine orig- govern the formal infrastructure of every art" (pp. vii- inality at any moment so that no invention overreaches viii). the potential of its epoch." (op. cit., p. 65).

ART JOURNALXXVIII 4 390 While it is a natural process of the mind to comprehend which in all ages and in all climes inexorably induces art- facts through generalization, the historian is caught for- ists to produce works of art in a certain preordained ever in a seemingly absurd circle of having to understand fashion." individual facts in terms of a general theory although the It is of course distressing to those who value the in- latter can be formulated only on the basis of individual dividuality of a work of art to have that individuality ig- facts.9 Faced with a paucity of established stylistic facts in nored by generalizations and classifications. To most Chinese art history, earlier Western art-historians tended scholars, documents, literary evidence, and above all, the to lean too heavily on metaphors and Western analogies in individual qualities of the artist and his work remain the their characterizations of Chinese stylistic developments. central important concerns of art history. Professor Max Professor Ludwig Bachhofer, for instance, saw Chinese art Loehr has suggested that perhaps the dating of copies is as going through the familiar Wdlfflinian cycles of graphic, not important; "As long as we have no means of ascer- plastic and ornate-or, archaic, classic and baroque taining the authenticity of individual works and attribu- stages.1' However useful it was as a pedagogical device, tions [by documentary and historical means], the histo- Bachhofer's dating of individual objects, on the basis of a rian is constrained to concern himself with the question Stilgeschichte on the W6fflinian model, appeared dog- of the authenticity, not of discrete works but of their matic; in one reviewer's words: "There is firstof all an a styles."12 He makes a careful distinction between "au- priori framework into which works of art in evolutionary thenticity" and "importance"; copies and imitations of progression are made to fit .... [Bachhofer is] an art-his- famous masters' works can be very important, while torian who regards style as the be-all and end-all of art archaeological evidence may be authentic but unim- history: style is a kind of sinister autonomous force portant.13 "The importance of a work," he writes, "de- pends largely on [the historian's] insight into its one- ' Erwin describes this circulus methodicus as an Panofsky time stylistic newness." A new style is a new idea .... The situation." See in the Visual Arts, "organic Meaning historian is interested in the inceptions of styles, not in Anchor Books, New York 1955, 8-10, and Doubleday pp. their perpetuation.14 In his quest to understand an "im- n. 3. E. H. Gombrich discusses the as 35, problem follows: portant" stylistic "idea," he prefers the evidence of later "The the historian's seems to me paradox of position pre- copies to that of the archaeologically recovered works of that the cherished can be cisely particular only approached the period, trusting himself to the "importance" of the on a the spiralling path through labyrinth of general "idea" in the copies. He lines up all the copies and attri- theories, and that these theories can be out only mapped butions in a distinct manner, meticulously studies and those who have reached the Think of the by particular. tabulates their motifs, then makes an intuitive leap to an adventure an ancient which exciting of deciphering script "insight into its one-time stylistic newness."'5 is not far from everybody's mind today. The individual " inscription is studied for what we can learn of the secrets Review of Bachhofer's book by Benjamin Rowland, Jr., of the script, and the script in its turn for what it will tell in The Art Bulletin, XXIX, 1949, pp. 139-141. us of individual inscriptions. To divorce the one from the 2 Max Loehr, "Some Fundamental Issues in the History of other would not only be foolish, it would be impossible." Chinese Painting," The Journal of Asian Studies, XXIII, '"Ludwig Bachhofer, A Short History of Chinese Art, no. 2, p. 187. " Pantheon Books Inc. New York, 1946. According to Pro- Loehr writes: "It is conceivable that we might arrive fessor Bachhofer, the development of Shang bronzes "took at a fairly accurate idea of the history of Chinese painting its natural course, from the simple to the complicated," on the basis of copies and imitations, if these are under- moving from the "graphic" to the "plastic" and finally stood in their stylistic sequence, and it is equally conceiv- to the "ornate"; the Chou begins with "a new cycle started able that a body of undubitable original works (if there on a new basis with very simple tectonic forms and ended is a way of establishing their genuineness) may not yield with complex atectonic forms." Similarly, "it was impos- an historically intelligible sequence." (Ibid.) sible . . . to keep sculpture from completing its cycle [of 14Ibid., p. 188. archaic, classic and baroque]." In painting, the major " In his recent book, Chinese Landscape Woodcuts from cycle, according to Bachhofer, ends with the "baroque" an Imperial Commentary to the Tenth-century Printed phase of Southern Sung. With Chao Meng Fu (1254-1322) Edition of the Buddhist Canon (Cambridge, The Belknap there began a "neo-Classicism in which many artists saw Press of Harvard University Press, 1968), Professor Loehr salvation from the utter destruction of form wrought by makes a "catalogue of [sixteen] motifs typical of the four a baroque style ... [by turning] deliberately to the linear woodcuts and of various paintings [attributed to the T'ang art of the great T'ang masters." "Mannerism" and "ec- and Sung periods] in which the same motifs occur" (p. lecticism" dominated the remaining centuries, with ap- 42). He then summarizes the qualities which these motifs parently only brief interruptions such as when neo-classi- share in common: "They are bold, even drastic motifs.... cism was fully re-instated by one of the great painters of There is an element of exaggeration in them, not con- the sixteenth century, Ch'iu Ying. trolled by rational restraint nor leavened by the experi-

391 Fong: Toward a Structural Analysis of Chinese Landscape Painting Fig. 5. Fragment from Khara-khoto, Inner Mongolia, ca. 1160-12, ink on silk.

Fig. 6. Wall-painting at Tomb of Feng Tao-chen in Ta-t'ang, Shansi, ca. 1265.

But a true historical synopsis must embrace both Nara, Japan) are of only limited value, because they rep- conceptual discipline and individual facts. How can we resent the work of anonymous craftsmen rather than of formulate an historical development of Chinese painting ranking artists and, as such, tell us little of the great cre- which, in short, combines the idea of periodic change in ative moments of the time. The significance of such data, pictorial structure (or form-relationships) in painting, however, lies in their indubitable authenticity. Archaeo- with the knowledge of continuous individual manners logical materials showing early Chinese landscape paint- characterized by individual motifs (including form ele- ing through the late-thirteenth century are found from ments and techniques) and expressive qualities? Japan to innermost Asia, and these offer a clearly defin- We must study the archaeologically recovered early able stylistic development. That widely scattered works works for the only remaining evidence of fixed visual po- should appear in a linked sequence of change is impor- sitions during the early periods. It has frequently been tant. Even though these works' may not mark the stylistic pointed out that archaeological data (including firmly frontiers of their times, they indicate a set of visual posi- datable works such as those in the Sh6s6in Treasury in tions that must be taken into account whenever the dating of an attributed work is in question.

ence of consciously explored visual reality" (p. 52). See III review by Richard M. Barnhart to appear in Artibus The development of landscape painting shown by Asiae. archaeological evidence from the pre-T'ang (before 7th

ART JOURNALXXVIII 4 392 century) to the early Yiian period (late 13th century) is one ranging from ideographic motifs to the creation of illusionistic space. The principal elements in Chinese landscape painting are mountains (or rocks) and trees. Ar- chaic representations of mountains and tress closely re- sembled their ideographic forms: t (shan) showing three mountain peaks, a "host" flanked by two "guest" peaks, * (mu) describing both forking branches above and anchoring roots below. The first important compositional discovery was that overlapping triangular mountain motifs suggest recession (fig. 1). By the seventh and eighth centuries, fixed compositional schemas developed. All the three principal schemas, later described by Kuo Hsi (active ca. 1060-1075) as the "high-distance" (kao- yiian), "flat-distance" (p'ing-yiian) and "deep-distance" (sheng-yiian) views, can be seen among eighth-century paintings in the Sh6s6in, Nara, Japan (fig. 2), and ninth- century ones from Tun-huang (fig. 3).16 These three compositional schemas have been basic to Chinese landscape paintins ever since; the picture-plane dominated by vertical elements, the picture-plane filled by a series of horizontal elements, and the picture-plane divided vertically between these two alternatives. In the T'ang and early- Sung examples, space is compartmentalized, a picture is entered in stages, each with a suggested receding plane tilted at a different angle towards the viewer. As seen in the wall-paintings at Ch'ing-ling in East Mongolia, dated around 1030 (fig. 4), which represents a panoramic "flat- distance" view, individual motifs are organized on an ad- ditive principle; they are seen part by part, and motif by motif. In the silk fragments of landscape sketches from the Central Asian site of Khara-khoto, dated archaeologically before the early-thirteenth century as (fig. 5), and the re- cently discovered wall-painting at Ta-t'ung in northern China, dated 1265 (fig. 6), spatial continuity developed. This was done first through the fragmentation of mountain masses (fig. 5). Disconnected silhouettes of dissolved forms, ranging continuously through space, are united by the mist or void around them. Finally, physical integra- tion of landscape elements is achieved through the estab- lishment of a consistent, receding ground-plane. In the Fig. 7. Fan K'uan (ca. 990-1030), Travellers among Streams and Mountains, Na. wall-painting of 1265 (fig. 6), mountains and trees are or- tional Palace Museum, Taiwan. ganic masses. Brushstrokes are fused and blurred; they suggest forms seen through atmosphere. Each mode of representation corresponds to a way of seeing. Archaic graphic conventions (cf. fig. 1) reduced, transposed and re-created nature; the words of the late 16 Four paintings on biwas in the Shas6in, all dated before fourth-century landscapist Tsung Ping explain this ap- 756, show the three principal schemas: "Sitting under a proach: "A vertical stroke of three inches may equal a Mountain" represents the "high-distance," "Hawks and height of several thousand feet; a horizontal passage of Ducks" represents the "flat-distance." "Tiger Hunt" and ink of a few feet may represent a distance of a hundred "Musicians on an Elephant" (fig. 2) represents the "deep- miles.""1 During the late-T'ang and early -Sung period distances." In the 9th century Buddhist silk banner from Tun-huang (fig. 3), the top scene is a "deep-distance," the 'TP'ei-wen-chai shu-hua-p'u, chilan v/2a. Translated by middle is a "high-distance," and the bottom one is a "flat- Alexander C. Soper, "Early Chinese Landscape Painting," distance." The Art Bulletin, xxiii/2, June 1941, p. 164.

393 Fong: Toward a Structural Analysis of Chinese Landscape Painting represented no mere retinal impressions of nature, but images of the macrocosm. In Chang Huati's (twelfth century) words: "Painting distinguished the 'black' of loom, heaven from the 'yellow' of the earth; it disclosed the se- ... crets of the yin and yang of creation. ... Whatever can be comprehended through the figures of the diagrams [of the Book of Changes] may be represented with physical form.1s The Southern treatment of land- --oXX::. . z Sung simplified ...... scape forms in mist, archaeologically exemplified by the .. Karaikhetsfragment (fig. 5), is described in a text by Han ...... Cho (12th century).'9 In Sailboat in Rain attributed to, M"M '10 $Iwo, and acceptable as by, Hsia Kuei (ca. 1190-1230) at Boston camwon ATwo (fig. 8), there is no ground-plane that actually links or w holds the objects, but the space depicted is unified and ...... Wo continuous. Frontal silhouettes of mountains and trees are made to float and fade into a void, representing a SM 4JN mist which ties the elements in a sequential fashion, MOM motif by motif. The illusionistic technique shown by the wall-paint- RM "'-Mow- of 1265 is in a text ...... ing (fig. 6) explained by Huang ------Kung-wang (1269-1354), which in turn perfectly describes Fig. 8. Hsia Kuei (ca. 1190-1230), Sailboat in Rain, Boston Museum of Art. the drawing and brush technique seen in the famous ------

...... ii

W.: iii =M**?:?!!iiiiiii

Fig. 9. Chao Mang-fu (1254-1322), Autumn Colors in the Ch'iao and Hua Mountains, dated 1296, National Palace Museum, Taiwan.

(cf. fig. 2, 3 and 4), philosopher-landscapists systemati- handscroll by Huang, Dwelling in the Fu-ch'un Moun- cally translated natural phenomena into different sets of tain of 1350 (fig. 10). The most important work for the interdependent yin-yang relationships. As motifs, there study of early Yiian painting is the short handscroll Au- were for instance "earthen mountains" (t'u-shan) and tumn Colors in the Ch'iao and Hua Mountains, dated "rocky mountains" (shih-shan), densely foliaged trees and 1296, by Chao Meng-fu (1254-1322) (fig. 9)..20 As a work bare branches. In technique, there were brushed lines that exemplifies the Yiian scholar-painting aesthetics, the and inked dots or washes. The "principles" (li), so often painting is well known for its use of archaic motifs and discussed by early-Sung theorists, referred to both the calligraphic brushstrokes. Yet in spite of their differences principles of nature and the principles of pictorial struc- in form elements and brush idioms, there are great structure. What was observed in nature must be articulated in tural similarities between this handscroll and the ar- theoretical principles as well as pictorial forms. The part-by-part compositions of the Northern Sung, here " P'ei-wen-chai, chiian b/7a. seen in the magnificent hanging scroll in the Palace "See Osvald Sirdn, The Chinese on the Art of Painting. Museum in Taiwan, attributed to, and commonly ac- Peiping 1936, pp. 81-87. cepted as by, Fan K'uan (ca. 990-1030) (fig. 7) by reveal- " See Chu-sing Li, Autumn Colors on the Ch'iao and Hua ing different views of landscape in a controlled sequence, Mountains, Ascona, 1965.

ART JOURNAL XXVIII 4 394 Fig. 10. Huang Kung-wang(1269-1354), Dwelling in the Fu-ch'anMountains, dated 1350, National Palace Museum,Taiwan. chaeologically discovered wall-painting of 1265 (fig. 6). painters were concerned with the problems of creating Both paintings show an illusionistic technique of creat- depth and recession and the treatment of forms in space, ing forms with fused brushstrokes of mixed ink tones, the Ming painters turned, more and more, to problems and an integrated spatial organization with a physically of surface organization and decorative values in painting. described ground-plane. The similarities indicate that, in The very complex details resulting from the "conquest of the second half of the thirteenth century, an anonymous illusion" in painting demanded new organization professional wall-painter in the North and a great schol- through pattern and stylization (fig. 11). Calligraphic ar-painter from the South, despite their differences in mannerisms and archaizing motifs were explored, in the expressive intent and artistic reputation, were at approxi- Ming period, for decorative purposes, and at the very end mately the same level in solving the structural problems of Ming and in early Ch'ing, as abstract forms in space of landscape representation. (fig. 12). It was precisely at the moment when illusion was IV mastered that leading Yiuan painters sought increasingly for the extra-representational qualities in painting. Ni Visual structure alone, of course, does not fully ex- Tsan (1301-1374) expressed the Yiian interest in "idea- plain style; an artist's style changes not because of evolu- writing" (hsieh-i), when he wrote: "I do not seek for tionary law, but because of conscious stylistic choice.22 In form-likeness, merely using painting to amuse myself."21 the second half of the thirteenth century, for instance, Likeness, a matter of eliciting recognition, or the lack of Chao Meng-fu (fig. 9) chose to paint in a calligraphic it, can be understood of course only within the context of idiom, while the Ta-t'ung wall-painter (fig. 6) used a the visual structure of the time. Although the Yiian paint- more conservative ink-wash idiom. Art-historically, this ers applied calligraphic techniques to painting, they did not paint calligraphic abstractions. In the accepted works " Max Loehr has formulated this problem well: "Tenta- by Chao Meng-fu, Huang Kung-wang and Ni Tsan, indi- tively I would conclude ... that changes of style are not vidual brush-strokes are subordinate to representation, al- caused by immanent forces; that 'immanence' is a con- ways describing and modelling form. After the Yiian, struct derived from an apparent logicality in sequences brushwork increasingly assumed an independent expres- of style; that this logicality stems from the rational and sive quality, and eventually dominated the representa- conscious act of innovation achieved by an individual tional form (figs. 11 and 12). Structurally, while the Yiian artist; and that without the creative individual's mind there would be no change, no sequence, no logicality, and no inevitability to speculate upon." ("Some Fundamental " P'ei-wen-chai, chiian villb. Issues .. ," p. 189).

395 Fong: Toward a Structural Analysis of Chinese Landscape Painting Fig. 11. Wen Cheng-ming (1470-1559), Landscape in the Manner of Wang Meng," dated 1535, National Palace Museum, Taiwan.

difference is more important than the fact that they both worked within an illusionistic structure. Similarly, it means little to say that Ming painting is "decorative" and landscape is "abstract," unless we can relate li i'~:i iXi Ch'ing isiilfX~i the Ming surface decoration and the Ch'ing abstract space to the scholar-painters's expressed interest in callig- raphy and the aesthetics of hsieh-i, or "idea-writing." ii i iiii :I;:?::xiiiiiiiiiiii:;-iii:::::::I.::i - _::::::iIl iiiiI iii The aim of structural is to reconstruct the i analysis formal problem to which stylistic changes must corre- spond as linked and purposeful solutions. In this reconstruction, historical and literary records, artistic treatises, as well as attributed works which constitute the bulk and essence of the available visual material must necessarily i:i''.ii'i:i?;-'?:!iii::i:I iiiiiiiiiiiii;)iii'l''i'ii??.; ?i'iii!!iii'i!i?~~i play an even more important role than the archaeological evidence. Literary records usually ignore the common :''''':':::::'''l:iiiii~i''iii~iii iiiiiliiiiiiiiiiiiiiiriiiiiiiii~ii~iiiiiiiiiilii iiii~i ...... and stress the unique; annals of art record the great moments of creative invention in much the same way as dy- nastic histories emphasize the heroic exploits of the great leaders. The stylistic development deduced from the ar- iiiiiiiiiil'i:il'rijiiii i~iii~iiij iiiiiiiiiliiiiiiii::::::::::::::::::::::::::::iii chaeological evidence, on the other hand, represents a history of style without knowledge of individual contri- : ii~ii~iiiiiii!•iii~i :'::::':'::':':':'::l:'::r::::::: r:i iiiili~i~l~il~i~iiiiii:ILL::iiiiiiiiiiiiibutions. With the help of literary records, individual ~ii~iji'~~iiiii~iij:iiii~~ contexts and critical purposes can be reconstructed. 'iiiiiiiiiii~~iiil:~ ~\ii:Iiiiiiiiiiiiiiiiiii:::iiiririiii iiii•i~iiiilllliiiiliiiiiiiiili~iiiiiiiiii~ii Attributed works must be studied in the light of not iiiiiiiiiiiiiiiiiiii!•i~ii:-:iiii'iiii!!!i..:'?iiiiiiiiiiiiiiii•: only archaeological and literary evidence, but also all iiliil:::::::::i::lil:::::, liririii'iiiiiii-iiilil~it:.:. •'ii??~~jiiiii:iiiiiiiii~ii_iiiiii ::::::::::::::::?::iiiiiiil other attributed works. I that the con- iiiiiiiii~iiiiiiiiiiiii'iiii iiiiii :::::::::::.:::::::: suggest following sii~iC~liiiiiiaiii- ::i:::::::::::::?? siderations may eventually bring order to the complexi- !iiiiii~i~ii':ii~iiiii~iiii!iiii•iiiiii?:-???:iiriiiiiii-e:l iiiiisijjiiiiiiliiiii~iiiiii ties of Chinese painting history: First of all, we assume i!i::iiji- that the manners of the ancient masters underwent visi- i•!i'?ii~ii::.l' iis-~':;iiiiiii~:liili:i;~'8i iiii~iiiiii~iiiiiiiiiiiii':iible structural alterations in each century at the hands of their admiring imitators. When a painter paints in the :ii~iiil~iliririiii::: ir~:il I manner of an ancient master, he borrows first the ob- *ii:::::i:::::::::jjiiij i:::i:i~ii~i-i~i~i~i-:iiiiiiiiiiiii:iiiiiiiiij il:::ii vious identifying brush idioms, form elements and com- iiiiiiiiiiiiii iiiii::::-:positional motifs. If he hopes to produce a close likeness .;•:?::;.,is.?ii?iii:?l• of his model he also tries to capture its expressive qualities. In expanding the original solution and giving it i fresh understanding, however, the copyist deviates from the original and makes subtle structural changes, thus bringing his work to a new visual position. The copyist, •iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiin short, shows in his work not the real ancient master, but a transmitted and transformed image of him. While qualitative differences are difficult to argue about, structural changes can be more easily detected and described.23 Secondly, since the visual material in the Chinese painting field abounds in copies and imitations, it lends itself to Professor George Kubler's idea of "formal se- " An exact tracing copy may preserve much of the original structure, but it suffers from a lack of spontaneity in exe- cution. For various methods of forgery, see my article, "The Problem of Forgeries in Chinese Painting," Artibus Asiae, Vol. XXV, 2/3, 1962, pp. 95-119.

ART JOURNALXXVIII 4 396 ~I I: :ri?iiiiiiiiii:i~i:~ I?:::::?::i~::~:~~i~It:~-

I:ili ~ a ~i~:~:':~':'i~::':'~~''~'~9Z~;~li~iii~i ::::iia.:?:?::?:::::::::iii:::~~?:??:::

:liiiijiiiiii.i:l'~'ili'iiiiiiiii':~:""i?,?:L~jiiii;i:iii:ili:-i:ijii:iiiiiii ii~9~i~i~(?i'ilii:iiII'"':':::::::;:::'" :i'i:::-:~lx:;iiiii::-i:i:i,:i:::;r;-~iiiiii:~:.::.:iiiiiiiiiiii:_:?g???i?i;~ i.~i~~.8'-':;:~'?x?-~~iioi'~':'i~i-i'li-x:i:i:;i-l:l::::-l:i:i:i:lli:;l-:::ji--:::::: -:::i:::i:l:l:::iii:i.-.i:i:i:l:::lli:l. ~:~:iii~jjj:??.:I~?i'::B1:-:ili-iiiiia'i:i,:i::::::i:i::::l:iiiiiiiiiiiil:i:?j::-:::::::-:: :::::::::::::j:I:-::-:-:-i:j~:: ..: :::-:-:::::i:i:l:l"r:i:r:::il::i::- ::::l:liliijiiijriI::ii ~:-::':::'~~i~~:~ii:I--::-:-:::::ii::1::?:::::::::::::-i-:::::_:_:-::-.:::::::::::::'I:i--:i-i:-:-i-i:i--i::;;?i-iiiri::':ii::-::-?::::::::: .i~ii:::_-:i:'l-li-z~~~j~giii~,:-li ~:iZiiilliliii:::::: ::::::::::::-;: ----?: ::- ::: :::: ::2::-:_il-:jl::l~lll'l:I:l~~l-:j:~?:jd) :':':':':':':'-::::':-:''::'-'----'-'---:-:: ..:.. ..-.- --:i--:l:':??i:i:???~??_i:.iiiiiiiuii:i

i''i?i:'?':i:i;?Ti?:'::?'??:?ri?i`'i;:i-i-i:i-i:i:i:i::'::I??'::-::-:::':-::'-~:::::ri:i:i-i

il::1:-~.:iiiiii::iii:i:???;ii:iiii4.iii I!iiiilgii)iliiiliiiiiiiiii:iiiii:iii il:l:i:iiiiiii;-:--;::-:llii:i:iiiii-;i :::::::::j:::;::i:::-i:?:::l::::i:::j::: ::l:I:l:iil:l:i:'-'-:::-::::::?:I::::::::::::::::::::::::::I:I:I:i:i:i:i:i:i:li:i:i: ~'"iiiiiiiiiiii-i:iii-iiiiii.::::_::i-: :---:-:_:lr1::1:::-:::?:I:::::::;r:r:::::?:~ i~~i-ii:iiliiiili:i:iiiiii.il;.ii:iiji::'::::::::::::-::::::::::::::: ::--:-?-:- ::--:--:I::-:::--:-::?i:I:ir:::_:::'?:fi iiliiiliiii r~i~ii~:i:~~j''i,--:ii:r;":iii:i::i;i;?i::i:::l;::::::;:::--::-?::-:?- i:-: i~:ii-:ii~i~~i~i i:i:~?:~iiiiiiiiiiiiiiiiiiiiiiiiiii-iiii:-i::' :-I::: :--: -::l:i::'--r---: -:'--:-'-': :iiiiiiiiiiiiiiiiiiiiiiiiiiii;~?-'?'..:l ~i:izl:'iii:?i~iii?:iii.~;~:':'''''l'''':::::::::::::::-:::::::-':::::::-:-:-:::. :: : :::::: :i~iii.ii;iiiliii:iiiiiiiiii:iirliiiiis'::-:i:slii.:~:Liiiiilf~iiriiiiil'~'~ ::iii?iiii-iii:iiiiiiiiili:::-i?iiii?i:i -:::I_;:;::::::~:--::-:-::j,:-ji::-;=:r-::i-:j:::ii::::::i-::;:;-:?:--?::-:i:::::i-::;:::_~i~:: ~-j~~i:::;~-~ ~b~i~~'iii:i~Xi~~'~:i"liiiiiii':':i-~:i..iiii:i:',Qs:::::1:::::-'''''i::'I"'i:'~,-,~i'''ii'i:i::~:-:::::1:::1::::'::;:':':':':':":' :::::''_'h':'i':::?:?ii:-i~ "i:::::::::::-:i:-1:::::l:::i;::::::::::~'liiiiiiiiiiiiiiii:il-lii:ii?i.i;i.i:lii~~iX :::l::::i::::_:::::::i:i:i--:i::::-:l;:: ~i~~~~~-~g~i:: :?;,;.~:::::::::.:.::.:::::-~ ziid:i~iiiii::iiii-'?::::::':"::::::::: :::::::::::::::::::::::::::::::::::::::::::::'::::: ::I:I~-..;.-..::,~?l,~i~i:lii.ii :~:j'i:~-''Fi:in~.:-:i~:::''ili~:~~..~i i:r:ia-:l:i::::::r::I~~?ili:l:l:;-l:i:::::::: ,i :iiii:iiiiiiiiiiiiiiiiiiIlIiI::::::::i-liil'~ii i ::::-~:::::::i:?:i:?:::::::"iI:::::::::i:::: :.:..:-i:::::i:i9:lii;ii':?'~:~E-::i-l::::: iQil~li

Fig. 12. Tung Ch'i-ch'ang (1555-1636), Landscape in the Manner of Huang Kung-wang, Cleveland Museum of Art.

quences" and "linked solutions."24 Copies and deriva- painting is structurally, a work of the period to which it tions of a single composition, done in various periods, is attributed; secondly, that together with literary and form one kind of sequence. Imitations and forgeries of other attributed material, the painting not only contrib- works of a given master, done in various periods, form a utes to the understanding of the personal style of the second kind of sequence. Signed works of well-known master, but also explains the transmitted image of the painters of different periods that are deliberately couched master's manner in later periods; and finally, that the in the distinct manners of some earlier masters, say, "wrong" painting can be explained and placed in a later "Huang Kung-wang," "Wu Chen," "Ni Tsan" or "Wang period, within the attributed master's stylistic sequence, Meng," form yet a third group of sequences. Properly or tradition. When the best of the attributed works are studied, all the works in different sequences should ap- established as original masterpieces, or their close copies, pear in series of linked solutions, beginning with the they will reveal the great moments of creative progress. original work, or its closest copy, and passing through successive stages of replication and transformation. Since these sequences, or continuous traditions, form parallel El Greco stylistic movements, they will corroborate, enrich and modify each other, eventually filling out a general stylis- These saints do not believe tic development through the different periods. Thirdly, to prove the authenticity of an individual That God can forgive. Not work, we must go beyond structure. To prove that one of two attributed works in a stylistic sequence is an original and the other a later imitation; or forgery, we must give All the Prophets's reassurances in order the following evidence: firstly, that the "correct" Can shake their prideful intelligence. 4 Op. cit., pp. 33ff. By characterizing stylistic development in terms of "sequences" and "linked solutions," Professor No, not even love, given fully Kubler avoids the difficulties of both the biographical ap- proach, which limits stylistic description to an individual Or received, makes any life span, and a Stilgeschichte on the Wdl;flinian model, difference. which implies a necessary sequence of styles. (See ibid., p. 36). -Thomas B. Brumbaugh

397 Fong: Toward a Structural Analysis of Chinese Landscape Painting