CHANG TA-CH'IEN'S FORGERIES OF OLD-MASTER PAINTINGS
Paper by James Cahill for symposium "Chang Dai-chien and His Art, Sackler Gallery, Washington D.C., Nov. 22, 1991. (Note: the Ss in boldface are references to slide changes.)
(Introductory: compliments on exhibition, catalog. Fu Shen's surprising success in getting owners of fakes to lend them, and have them published in catalog, as that.)
This is a paper, or perhaps better a slide show with commentary, of which the public presentation has been long delayed--for nearly thirty years. I planned to deliver something like this at the first great gathering of Chinese painting specialists, the "post-mortem conference" for the Chinese Art Treasures Exhibition, held at Asia House Gallery in New York in October of 1962. My presentation has grown to be much larger, and now has a quite different purpose: to commemorate and in a curious way honor its subject, Chang Ta-ch'ien, as a great forger of Chinese old-master paintings, instead of warning the world about them, as I meant to do then. Since much of what I will present comes from personal experience, I hope you will forgive a somewhat autobiographical tone.
(S,S) I met Chang first in Kyoto in 1955, when I was a Fulbright student and he had come to work with the publisher Benrido on the four volumes of reproductions of his collection, Ta-feng-t'ang ming-chi. He was staying at Kyoto's most elegant ryokan, the Tawaraya, and I visited him there. Since he had studied textile making in Kyoto for two years from 1917 and had been back to Japan often since then, he knew the city well, and spoke some Japanese, so that we could communicate; also, he was traveling with the art critic Chu Hsing-chai, who served as his English interpreter. My memory is of sitting with Chang drinking tea and talking about particular paintings; he had a brush and paper in front of him, and was sketching passages from them as we talked--I would ask "What do you think of the so-called Ch'ien Hsüan in Detroit?" and he would do a detail from it, perhaps a frog and dragonfly, as he replied. This was my introduction, and an extremely impressive one, to his extraordinary visual command of the whole past of Chinese painting.
Other memories include being at painting viewings with him, arranged by my then-teacher Shûjiro Shimada--Osvald Siren was also in Kyoto at the time, and I recall a meeting at the Fujii Yurinkan when Chang wrote a colophon on a handscroll of theirs attributed to Hui-tsung, listing in it all our names; it will no doubt puzzle scholars of the future. I remember the dealer Kawai Shôgadô, whom Chang owed an obligation for a T'ang silver box Kawai had given him, asking him to paint an album of landscapes in the styles of early Ch'ing masters such as Shih-t'ao, Pa-ta, K'un-ts'an etc.; Kawai's intent, as he told me, was to use the album in detecting Chang's forgeries of those masters. But Chang, as always, was one step ahead: his "imitations" of them in this album were recognizably in their styles, but were quite different from the styles he would use when forging them. Again, a dazzling performance of which I was already able to appreciate the subtleties.
Other reminiscences are irrelevant to this paper, but are hard to suppress. Being introduced to Szechwan cuisine by Chang at the restaurant at Roppongi that he patronized, when I visited him in Tokyo; audaciously persuading this famous personage to come with me to the Yûshima Seidô, the Confucian temple at Ochanomizu that sold Chinese antiques, to give me courage to spend $150 of my Fulbright stipend on a handscroll painting (it was the "Fishermen" handscroll by Wu Wei). Chang pronounced it genuine, and I bought it.
I was not aware, then, that Chang was producing forgeries of T'ang-Sung paintings as well as Ming-Ch'ing; that realization came after I returned to the Freer Gallery in 1956. I brought with me color transparencies I had made from a remarkable painting that had appeared in Japan, owned by the dealer Ogiwara, which I hoped the Freer would buy. (S,S) It was a Bodhisattva holding a glass with flowers, purportedly from Tun-huang, with an inscription containing a date corresponding to 757. Most of you will be familiar with it from Wen Fong's excellent 1962 article on forgeries in Chinese painting, in which this work is analyzed skillfully and shown to be a fabrication, copied from a figure still on the wall of one of the caves at Tun-huang. (I deliberately begin with a case in which I myself was taken in, to avoid any implication in what follows of criticizing the connoisseurship of others--we were all taken in, at one time or another.) The Freer did not pursue the painting, but not because of any doubts I had about it.
(S,S) Some time after this, the dealer Joseph Seo brought us for consideration another would-be Tun-huang painting, also dated 757, with the same donor (or his wife) mentioned in the inscription; we borrowed it to study. (It is reproduced in color in one of the Ta-feng-t'ang volumes.) Now our suspicions were aroused a bit, by the circumstance of two related works appearing within a short time. The Freer's Japanese mounter Takashi Sugiura looked at it carefully and pronounced the damning verdict that this purported T'ang painting was in fact done on Japanese silk (he was quite firm and specific about this). (S--)With the owner's permission we took pigment samples that were analyzed by John Gettens of the Freer's technical laboratory; he told us that one of the pigments, the yellow as I recall, was not used until some time in the 19th century. (Note also the flattening, completely non-volumetric drawing.)
(S,S) The detection of this one focused my suspicions on a recent Freer purchase (for which I wasn't responsible--it had been accepted for purchase by the director Archibald Wenley before I joined the staff in 1957), the handscroll representing "Three Worthies of Wu-chung," which presented itself as a work of around the time of Li Kung-lin, possibly by him. We put this through the same kind of technical examination, and it also flunked. (S,S) Besides the yellow, a white pigment--titanium dioxide--again was impossible for the early date claimed for the painting. (The technical details are no doubt still preserved in the Freer's files.) (S--) And Sugiura pointed out that although the silk was rent everywhere, the individual threads had not decomposed, and still had their flexibility and tensile strength. This was enough in itself to indicate artificial aging.
It was clear that these three works, along with others that appeared over the next few years, had identifiable physical and stylistic characteristics that could not be easily matched in genuinely old paintings. (S--) The areas of heavy pigment were rubbed and abraded in a more or less uniform way, and sometimes partly blackened, as if by pigment discoloration. (This is a detail from another Buddhist painting, included in Chang's posthumous gift to the Palace Museum in Taiwan). (S--) Light spatters of dilute ink onto the silk were meant to represent mildew spots partly removed through washing in remounting. The silk itself had a distinctive color and look, the product of artificial aging. Shûjirô Shimada remarked to me that the inscriptions on these paintings, which purportedly spanned centuries in date, seemed all to be from a single hand. (Fu Shen has studied them and concluded that it is probably the hand of Chang himself, although he also reports another opinion that it was done by his third wife.) (S--) The brush-drawing is strangely lifeless, representing presumably Chang's attempted re-creation of the purity and impersonality of brushwork as he had seen it in anonymous early paintings such as those at Tun-huang. (S--) Structural faults can be detected, drawing that does not "read" representationally, such as the non-organic drawing of a Bodhisattva's hand pointed out by Wen Fong in his article. (S--) Most cleverly, clues are planted in hard-to-read seals and inscriptions, which, when deciphered and identified, prove to match up with old, somewhat obscure records. Scholar-curators followed these clues with increasing excitement, like Hansel and Gretel following the trail of bread-crumbs left by the witch to lead them to her house where they would be eaten; they arrived at the conclusion that Chang had intended them to, and felt pride in having uncovered a long-neglected masterwork. The construction of these false trails itself required an unusual knowledge and ingenuity, a kind of scholarship-in-reverse that creates the data instead of uncovering and analyzing it.
By this time, I was of course catching on to what was happening, and who was behind it. The project of following Chang Ta-ch'ien's tracks and trying to detect his fabrications became a fascination for me; it was like playing a complicated game with a very capable adversary. (S,S) For anyone working in the field of Chinese painting at that time, Chang was a more or less inescapable figure. For instance, after my Skira book on Chinese painting was published in 1960, more than one Chinese friend (one of them, I recall, was Cheng Te-k'un) told me that while my selection was generally good, I had made one major mistake: reproducing the Sumitomo Shih-t'ao "Waterfall on Mt. Lu" (a detail from which was on my title-page.) When I would argue for the authenticity of this great work, which I knew well from visits to Sumitomo Kan'ichi in Oiso, they would play their trump card: they had been personally assured by Chang Ta-ch'ien himself that he had painted it. These reports disturbed me, and I took the trouble of checking into the matter. What I discovered was that the painting had come to Japan, and entered the collection of Kuwana Tetsujô, as early as 1908, when Chang was nine years old. After that I had my counter-argument ready, for those unwilling to accept the simple truth that the painting was far beyond Chang's capacity: he was indeed precocious, but not that precocious. It would appear that part of his effort of obfuscation was to lay claim to works that in fact were not his.
| 
