Problems of Painting and Viewing in the Eighteenth Century

Timon Screech

Stories of painters' lives and pictures of artists at work are rich sources of understanding how the creation of images was interpreted. But other information can be found in another sort of myth: the myth of the superlative painter. China and Japan generated a number of these, two of which were particularly prominent. One told of Wu Daozi of the Tang, and the other of the somewhat earlier Zhang Sengyao. Wu was celebrated throughout the history of Chinese and Japanese art criticism, and although not a single work by him had survived from antiquity, his praise dotted the writings of every art commentator. One classic source may be quoted here - a treatise by the major Kano artist, Yasunobu, published in 1680, the year of his death. Wu drew, he said, with marvellous 'naturalness and truth' (shizen to makoto), and Yasunobu continued,

When Wu Daozi painted a dragon, it stirred its fins making clouds gather and rain fall. Yet even the painter did not know how it was that the dragon came to move. This was painting of the highest order.[1]

Again a few pages later,

Wu Daozi painted five dragons in the Inner Chambers of the Emperor Ming's palace. Thereafter, whenever it rained, the fins of the dragons would stir, ushering in smoky-ink clouds. This happened because Wu Daozi believed fervently in the Way of Painting and often chanted the Diamond Sutra, to which he adhered firmly. That was how he came to understand the very heart (honshin) of visual qualities.[2]

The tale became proverbial: "paint a dragon and heaven clears" (garyu tensei), ran an old saw.

Similar achievements were attributed to Zhang Sengyao. Zhang was enumerated with Gu Kaizhi, and Lu Tanwei as the "Three Greatest Painters [of all time]". "Zhang captured the flesh, Lu the bones and Wu the spirit," wrote an ancient source.[3] By this claim Zhang actually surpassed Wu, for while Wu's dragons moved on the surface of the paper, Zhang's came fully to life, shuffled off their inky bonds and flew away.[4] By the Edo period Zhang and Wu were firmly set in the Japanese artistic pantheon. When Yanagisawa Kien famously left stormed from the Kano School in high dudgeon in the early years of the eighteenth century, (as he recounted in his work Hitorine of 1724), he announced he would be putting hiself under the tutelage of Wu Daozi and Zhang Sengyao.[5]

Towards the end of that century, or in the beginning fo the next, a painter living in the Osaka region, Nagayama Koin, produced a single screen showing Zhang at work. The artist is sen having just put the final touches to the painting of a dragon [Fig. 1]. Yet Zhang's page is quite blank: his dragon, fully rendered, has assumed life and is ascending heavenwards to the wonderment of an on-looker.

Zhang was the archetype of perfection in art, and the model for all East-Asian painters, and his myth formed the prevailing standard of the highest achievement. Zhang's authentic œuvre was scarce (only one work by him existed in Edo Japan) and like Wu's tales of his genius took on qualities essentially irrelevant to him as an individual painter, referring rather to an ideal type. Those who wished to follow excellence were urged to subscribe to the notion that painting was the creation of a life urge. Painting was said to wrest generation out of the mean grasp of nature, and by making the artist one with the natural world, allow him to participate in creation itself. This was not a matter of following nature by capturing the appearance of forms, but of equalling nature by imbuing pictures with such seminal power that neither paper not silk could contain them, and they would live. The process of generation by painting was mystical, and the ability acquired in a way that even the artist did not fully understand. "Painting of the highest order" is magical.

The force of the Zhang and Wu myths as archetypes was such that other great painters came to be accredited with the same powers. In Japan, Japanese artists were said to have achieved this. Shohaku was one. It was recorded that he painted a large dragon for a temple in Ihosaki one summer, as he completed it, an unseasonable storm unaccountably brewed out of blue sky (incommoding the parishioners who had chosen that day for a picnic).[6] Even when the storm had cleared people were terrified to look at the picture, fearing that the dragon might spring out at them.

The myth of Zhang is the quintessential East Asian myth. It may be objected that the Western tradition has something similar in Daedalus and Pygmaleon who produced works that lived. These, however, are different on close inspection. Firstly, Daedalus and Pygmaleon were sculptors, not painters, and so their works can be said to have existed in the three dimensions of the empirical world to start with, meaning that animation did not affect their form, only their internal qualities; secondly, at least in the case of Pygmaleon, the sculpture was said to come to life through direct divine intervention, as a specific act of grace, and for all Pygmaleon's expertise the mystery was nothing to do with his own skill. Thirdly and importantly, the ability of those sculptors to make their work live was predicated upon total verisimilitude. The Chinese legends differ on all these points: their works start as two-dimensional; they came to life habitually, and although the creation of living painting is a mystery, and at least in Yasunobu's retelling of the tale tied to acts of religious devotion, it is not the result of supernatural intervention; most crucially, nowhere is it stated that the images moved because they were realistic in any formalistic sense.

Later commentators on the Chinese legends were skeptical of the claims made for Wu and Zhang on precisely the grounds that the images they made were not realistic, and so could not have lived in the human world. This was true in the case of dragons which are, after all, mythical beasts. In China, Shixiang Jushi raised the issue in 1632,

I have privately wondered how a painting that was not real [i.e. not on a real subject] could do this… What about the lightening and thunder cracking the wall after Sengyao had painted in the eye-pupils of the dragons? What about he five dragons painted by Wu Daozi in the Inner Hall which could give off mist when it rained? In these cases it is not possible to treat the paintings as real.[7]

Shixiang was perhaps too literal-minded, but his objection had some cogency and he was not the only one to raise such doubts. Ranga artists also criticised the Chinese myths, and Kokan flatly refused to accept the possibility of creating life forces. He took as his point of departure another Chinese text,

In the New Thesis (Xinlun; J: Shinron) of Liuzi [J: Ryushi] it says,

"If an artist paint a picture of [the bodhisattva] Seishi, then however beautiful he is, you cannot make love him; if a sculptor carve a peach, no matter how realistic it may be, you cannot eat it."

No matter how desirable Seishi may be, he is only a picture, and the peach it cannot be savoured because it is only wood. [A picture] has the name of what it represents but not the essence.[8]

Ranga artists accordingly did not consider themselves much served by existing myths.

Kokan visited Nagasaki in 1788, to find out about Western painting, but improving not only his technique, he also had the chance to clarify his theoretical positions for he heard the most famous of all Western myths of painting: the legend of Zeuxis - a narrative that far outweighs Pygmalion or Daedalus in its centrality in the Western artistic tradition.[9] Upon returning to Edo the following year, Kokan painted a Zeuxis, showing the artist standing before the famous depiction of grapes which he made so realistic that birds flew in to peck at the painted fruit [Fig. 2]. Kokan offers this myth as the governing legend of Ranga, for although he painted the figures in European clothes, he signed the canvas of grapes with his own name in Roman letters, 'Kookan'.

Kokan may have heard the legend of Zeuxis from a Dutch informant in Nagasaki. Certainly the story became rather well known in Japan around this time. In 1801, Yamamura Saisuke discussed it in his Seiyo zakki, and included further details, such as the fact that Zeuxis had been painter to Alexander the Great.[10] Saisuke concluded that Zeuxis was, "unmatched in any artist in the ten thousand nations." The Lord of Kaga also owned a Dutch book giving illustrations of the tale, which Kokan apparently saw.[11] It has been plausibly suggested, moreover, that the source of Kokan's picture is another Western book, J.L. Gottfried's world history, the Historisches Chronick, originally published in German, but brought to Japan in the Dutch translation of 1660.[12] The original story, which comes from Pliny, is longer than the truncated form Kokan showed, for Zeuxis had in fact painted his grapes as part of a trial of skill with another artist, Parrhasius. Pliny told how Parrhasius' return offering was covered by a curtain, and when Zeuxis went to draw it back, he found the curtain was the painting.[13] Gottfried showed the encounter between the two artists at the point where the crowd has turned from admiring Zeuxis' work to look at Parrhasius's (in Pliny the trial actually takes place over two days). Kokan omits Parrhasius. This makes a drastic change to the meaning of the tale, for while Zeuxis had fooled the birds, Parrhasius had fooled a man.[14] Zeuxis provided a perfect substitute for the depicted object, but Parrhasius did something more subversive. As Pliny put it, when Zeuxis realised Parrhasius' skill, with modesty that did him honour, he yielded up the palm, saying that whereas he had managed to deceive only birds, Parrhasius had deceived an artist.[15]

As proposed by Kokan, replication represents the pinnacle of achievement, and trompe l'oeil is not discussed. The work of art isproposed as an exact duplication of the phenomenal world, so close that the innocent eye of nature (here birds) is fooled.

The differences between the European and East Asian myths are, I think quite clear. Both extol the power of the creator of images, but the kind of power diverges. In one case, the picture breaks nature's monopoly on the generation of life, in the other, the painting replicates nature with what has been called an Essential Copy.[16]

Acceptance of a division between form and content of a depicted thing was understood in Japan to govern all Dutch painting, and yet so strong was the tradition of equating great art with the creation of life that even painters whose own beliefs would not have supported the notion were attributed generational powers. Okyo was ascribed a feat like Zhang's. In that version the dragon is replaced by a ghost, but the effect is the same and Okyo is said to have recoiled as the strange form he had depicted peeled itself off the picture to haunt him. Extant sketchbooks show that Okyo worked his compositions over and over in a way incompatible with the idea of 'finishing' a work in a specific instant credscendo, as the legendary Chinese masters had. It may have been unease with the jarringness of applying the myth of Zhang to a Westernising artist that led to adaptation of the tale. It began to be said that the painted ghost had haunted Okyo less because he had imbued it with such emotional charge than because he had copied the appearance of an real corpse so exactly that the deceased person (not the picture) had been conjured up.[17] But Okyo was not the only Westernist so stranglely praised. Yuhi, a Nagasaki artist was held to have inadvertently brought a tiger to life while copying it from a Western-style painting.[18] The painted tiger, an impatient sitter by all accounts had, it was said, lost concentration, dropped its head and began to look away during the process; Yuhi jovially chid it by tickled it on the nose with a bamboo pole, and then returned to work. So absorbed was he that he did not notice the uncommonness of the incident, but several onlookers were amazed, and when the tiger began to look provoked, they fled.

The Chinese legends were a part of the Japanese heritage, and they could be reattributed to native painters with impunity (even if with a degree of tongue-in-cheek). It was a little harder to know what to do with the novelty of Zeuxis, with or without Parrhasius. Kokan, perhaps wishing to claim that existing manners of Japanese painting could never be Zeuxian, did not attempted to integrate them. Others tried, however, and experimented with indigenised Zeuxises. Akinari was one such, and he attempted a synthesis in another story in the Ugetsu monogatari.[19] Akinari contrasted his Japanised Zeuxian painter with a maker of Zhang-like images. Both painters, who were called Narimitsu and Kogi in the story, are praised for their skill, and it is noticeable how there is no attempt to set up an open debate between the two as there was in the Hakuraku-ten encounters, indeed debate is eschewed, and Kogi is even said to be Narimitsu's teacher to suggest continuity. The achievements of the artists, however, stand in sharp contrast.

The story, Muo no rigyo tells how Kogi, a monk from ancient times living in the Temple of Mii-dera on Lake Biwa, liked to paint carp. He would take a boat out on the lake to see the fish sporting, and paint them there all day long. Kogi never sold his pictures, though he was perpetually solicited to do so, for he loved the carp as pets. Eventually, realising he was near to death, Kogi loaded his pictures into a boat, took them onto the lake, and scattered them over the water. The carp at once started to life, and floating away from the picture surface, swam off into the depths. After Kogi's death Narimitsu is introduced,

His pupil, Narimitsu, received his wonderful skill (shinmyo) and was famous in his day. An old chronicle says that when he painted a cock on a shoji in the Kan-in Palace, and a living cock came up and kicked it.[20]

The difference between Kogi and Narimitsu is exactly that between Zhang and Zeuxis. Kogi, we read, painted 'very precisely' (kuwashiki), but his genius did not lie in his skill at copying and transcribing, but in his emotional involvement with his painting, which eventually brought them to life. Narimitsu's achievement, however, is to fool the innocent eye of a cock.

In 1825, Hokusai's pupil Hokkei produced a print that might almost have been an illustration of the Narimitsu story. He shows a living cock kicking a painted bird done on a tsuitate screen [Fig. 3].

Kogi and Narimitsu were historical figures and the Kan-in Palace was a real place (although long since destroyed) and Akinari took care to point out that the event was well attested, noting, "this account comes from an old anthology of tales", but Akinari did not trouble to identify the source. That was not really the point of his remarks. He was less interested in attesting the story through citation of documents, than in demonstrating that the Narimitsu myth was antique and therefore indigenous.[21] To him, the Essential Copy was not derived, as Kokan would have it, from the Dutch. Akinari's Zeuxian precedent was untainted by the West. There is no doubt that if well dredged, the East-Asian tradition could yield tales of such achievement, just as Pygmaleon and Daedalus exist in the West, but these are very much counter-trends. I believe the timing of Akinari's writing makes it impossible that he was aware of the claims being advanced in Ranga circles, and that he was consciously reacting against them. Akinari's Tales are difficult to date for he says in the Preface that they were completed in 1768, but they did not come out in print for a further eight years, and by 1776 Lord Satake's treatises were already written, and many of the Ranga positions fully articulated. It is quite improbable that Akinari could not have had these in mind as he wrote.

Akinari's attempt to slide seamlessly from Kogi to Narimitsu (from Zhang to Zeuxis) without acknowledgement of the enormous split that this entailed, cannot, I think, in the end, be accepted. The differences are insurmountable and the co-existence of the types associated with Zhang and Zeuxis cannot be so natural as Akinari protests.

Life and the Genius

Ranga artists were happy enough to remove the demand to be the creator of life from the shoulder of artists. Kokan, for example, wrote in his Seiyo gadan of 1799,

There are many Western pictures to be found in Japan today, and they are called Horurando [Holland] pictures. In the countries of the West, people do not paint as artists here, but capture reality (shashin). Many Japanese and Chinese-style artists find the Dutch manner odd and are not inclined to study it or put it into practice, preferring to dismiss the manner as precision work (saiga) and not art. This response is foolish.[22]

The differences between art and non-art, and thus between artist and non-artist were irrelevant to Kokan. It may reasonably be asked where his interpretation of pictures leads. Many critics found the position, wedded to capture of external appearance, devoid of spirit and (literally) lifeless. Tani Buncho, for example, another formidable force in later Edo painting, wavered over the issue. In his early career, Buncho had taken the name Shabutsu-do, "the artist who copies down matter" with its decidedly Westernist ring, but subsequently he altered this to Shazan-do, "the artist who draws mountains", affiliating himself with more namga-based (literati) concerns. Buncho expressed his anxiety,

I used to have a large number of Western pictures in my collection, but I tend to find they are like the images people engrave on seals - without real meaning (imi) to them. When you try to appreciate Western pictures on a profound level, you always feel there is something lacking.[23]

In contrast to these misgivings, Buncho fully supported the old sort of painting of Wu and Zhang, and indeed he copied Zhang's only known work existing in Japan.[24]

The distinction between Westernist and nanga painting is expressed here via its relationship to the viewer - whether or not it conveys meaning to him or her. In the legend of Zeuxis, rewards were allotted in proportion to the extent that the viewers were fooled, and thus painterly worth is properly to be measured only after the artist has put down his brush, the paint is dry, and a viewer has come upon the scene. By contrast, it was irrelevant to Zhang what his viewers thought, or even, indeed, if he had any viewers at all. His pictures were prone to spring to life and be gone, and pushed to its logical conclusion, this myth annihilated the role of the viewer, allowing place only for viewers of the process of painting. Akinari noted as much in the case of Kogi, who was so great painter that his pictures came to life, and, "on this account, his work has not come down to us today."[25]

Even when a picture did not literally animate and depart, the finished image would only be an afterglow of the exhilaration of production. Nanga accepted this and so celebrated process, setting particular store by the ideal of a community of like-minded individuals who produced work in front of each other, publicising the enthusiasm of the moment. Nanga ideals of painting fused viewer and painter together by emphasising co-production. Roles could be mixed, with one man painting, another adding something, another inscribing it, and so on through the group, delaying the time when the picture would have to be called finished and put aside. It was more important, therefore, to produce at the right time, than to make something that would last. This was the case in gesaku too, which was also a community movment. Sanba, for example, was reckoned to be a thoroughly third-rate poetaster and, "he left behind not a single celebrated verse", but he was nevertheless immensely sought-after because what he did, he did impromptu, performatively. Moreover, Sanba's verses were often made as complements to existing work, prolonging its act of creation and raising viewing to the level of creating, melding them. "Only Sanba," wrote Yukimaro, "and the late Enba managed to compose poems on the spot, based on the paintings, to give to others in the party."[26] The final work was a by-product of the activity and almost an excrescence from the conviviality of the occasion.

So it was that many nanga painters left no name behind them. After their deaths, with the inevitable splitting up of their groups, their work ceased to have any great significance. Even the great late eighteenth century artist Ike Taiga (regarded nowadays as the representative nanga) was forgotted totally only thirteen years after his death.[27] Kokan commented on this phenomenon with reference to a painter, Chiku Gen'iku, who had died a generation before,

He loved a good drink, and was a sociable and refined person (furyu gajin). He painted ink orchids [viz. Harunobu's nanga artist and Nandara's subjects]. Everyone would come to him with paper and he would dash off several sheets and hand them over - there was a vast quantity of his work around. Many people had his pictures, and everyone thought him a particularly interesting person. But his family has died out now, and those who were once in his circle have all passed away. As a result his orchid pictures have become just so much waste paper, and no one now has heard of him.[28]

Gen'iku's painting was a kind of play, and, the game over, his pictures were expendable. Lord Satake, comparing Dutch paintings with Japanese, came to the conclusion that the latter were indeed little more than accompaniments to jollifications, performances, and so, in his estimation, inevitably trivial; Kokan too, on more than one occasion, dismissed the entirety of Chinese painting as "playthings." (ganro).[29]

The practise of sekiga, or painting live in exchange money was simply a commercialised extension of this community ideal. Sekiga were made by needy nanga painters in preference to selling finished works. Happy to admit to staying with rich patrons as 'guests in ink' (bokkyaku), nanga artists were loath to admit to selling finished works. Painting before a patron in exchange for lodging was at least performing, albeit debasedly for payment. These highly visible presentations allowed those outside the community of producers to experience the vagaries of nanga practice, and sekiga sessions provided food for myths. There are no tales of Ranga painters at work because no one saw them; Kokan's Artist in his Studio paints alone. Economic circumstances did force Kokan (and presumably others) to perform sekiga sometimes, particularly when travelling, but the titles of the works made on those occasions suggest that the pictures were done in one of the non-Western styles at which his early training had also made in adept.[30]

Wu Daozhi was famous for working fast, and his companions need not have remained by his side for very long before seeing the miracle appear. This was not the case with Ranga. Producing Western-style pictures was slow, punctilious and hardly a spectator sport. Yuhi, when a student, had presented his master, Nanpin (whose style was strongly influenced by European manners) with a work on which he had spent two full days; this would have been a long time for a nanga work, but Nanpin was annoyed with what seemed to him a hurried, slap-dash attitude, and forbade Yuki in future to bring anything on which he had spent less than thirty days.[31] Kokan's journals reveal that while his non-Westernist sekiga sessions lasted part of an evening, on days when he painted in the Western style, no other activity was undertaken.[32]

In the spring of 1799, the incompatibility of Western-style painting with nanga notions of convivial production was conclusively demonstrated at a party held by a wealthy Kyoto rice merchant, Nakai Hidetake.[33] Fifty painters and calligraphers were invited to a gala that lasted two days. Okyo was now dead, but his pupil Goshun and son and heir Ozui were present, as were Hara Zaichu, Watanabe Nankaku, Okamoto Toyohiko and an assortment of other celebrities. Kokan had painted Hidetake's portrait two years before while on the way to Nagasaki, and so he was invited down from Edo too. The works were presumably produced extempore during the festivities with each artist performing for the others. Kokan, however, was unable to finish his piece during the time provided. His picture did not appear, and had to be sent in a month later.[34] Less that a year earlier, again passing through Kyoto, this time staying with the painter Gessen, Kokan had made the point, "I had hoped to make a Dutch-style picture that night, but as they are painted with oils, it is not feasible to do one in the space of a single day."[35]

The Western picture, being a transcription of nature, should betray no evidence of the flourish of the hand that made it, but had to be worked and reworked in silent concentration. The concomitant removal of the personality of the painter was hailed by protagonists as the major advantage of the style. Kokan's diary Shunparo hikki, completed 1811, indicated how it was that the personality of the artist could be expunged,

Dutch painting, contrary to our Japanese painting or to that of the Chinese, has no brush-dharma, no brush-thought, and no brush-energy. It confines itself to depicting things as they really are.[36]

These three aspects of the 'brush mind' (hippo, hitsui and hissei) are among the dearest doctrines of traditional painting, forming the contours of the mental state that it was believed allowed creative genius to flourish. Kokan rules them out; Lord Satake disliked brush mind too, "the opinion prevails" he wrote, "that pictorial spirit is preferable to formal likeness… but we must not permit ourselves to draw only according to feelings and sensations."[37] Such was the emphasis placed on the objectivity of Ranga pictures that it might be supposed a machine should do the job of painting. There was such a machine, and many Westernists did think it was better.

Kokan sings the praises of a Dutch contrivance that had been imported and could be used to create automatic pictures, cutting out artistic personality, and cutting out painting as performance. The device was the camera obscura, introduced from Holland at the end of the eighteenth century. Kokan described the box as, "a contrivance to let one depict scenery so that it will appear just as if one were actually walking through the place."[38] The camera obscura was well enough known for one of the major compendia of Dutch knowledge to include reference to it: the Ransetsu benwaku, an encyclopedia published as a question-and-answer session between Arima Gencho, domainial doctor of Fukui, and his Dutch-studies teacher Otsuki Gentaku in 1799, stated the following,

My question to you is this: there is a kind of device made from a box fitted with a glass lens that projects landscapes or human figures so that they can be copied down. In Japanese this is called a 'truth-copying lens' (shashin-kyo). Are the original foreign ones called the same thing?

My reply to you is this: there are several of these boxes to be found in the houses of enthusiasts in this country. The device is extremely cunningly made and is indeed called a 'truth-copying lens'.[39]

The book illustrated the contraption, and labelled it in a transcription of the Dutch term, Donkel kammer.

No great skill was needed to operate a camera obscura, and it allowed the painter to be almost literally dispensed with. Gentaku claims that boxes were to be found in Japan, apparently in some numbers, although the rate of diffusion is unclear. Imported cameras would have been expensive, but domestic versions could have been improvised without undue trouble once the gist of the structure was understood. The only part difficult to obtain in Japan would have been the lens, but home-made cameras could be jerry-built without a lens so long as a fuzzy and inverted image was tolerated.[40] The point of the camera obscura has often been misunderstood, and it has been confused with viewing devices but it was in fact a machine for making pictures. Sometimes the reflection might not be taken down, and the image observed in all its transience, as a picture in its own right, but this was to use the device by halfs. Paper should be inserted into the box and the images traced and preserved. Exposing a camera to friends or fairground customers was sure to cause a sensation. Bakin went up to Kansai from Edo in 1802, and saw a pin-hole camera in Osaka. The device was not a box but a literal camera - a room - darkened to pitch blackness and with a hole in the wall. There was no lens and so the image fell up-side down and could only be focused by moving a board that served as the screen backwards and forwards. The demonstration was carried out in the house of Fukami Shobei, a wealthy wholesaler, and his son, Sataro, who were obviously in the habit of showing the effect off. Bakin recorded the experience in his diary, Kiryo manroku. The projection he saw must have been a remarkably clear since, even allowing for hyperbole, he claimed to have been able to make out the whole the garden outside, complete with pond, bamboo grove and willows, and even a piece of child's practice calligraphy hung up to dry on which he could go so far as to read the date.[41]

Bakin's diary was only published posthumously, but news of Shobei's room circulated widely in Edo thanks to Bakin's committing his observations to the more immediate context of a kibyoshi. The work was entitled Kage to hinata chinmon zue and published in 1803 with illustrations by Shigemasa (this is in fact the book mentioned above that bewan with Bakin being visited by the publisher's boy) [Fig. 4]. The texts of the diary and the kibyoshi are nearly identical, but in the former, Bakin added an interesting observation: the projection looked just like a Dutch painting.[42] How did he come to gain this impression from the image of a garden that sounds so Japanese? The precision of the colours and the shading might have seemed European, but what was most foreign, it may be surmised, was the automatic creation of the work, formed by itself without the imposition of any artistic personality. Bakin looked on virgin representation.

---

[1] Gadô yôketsu, reproduced in Sakazaki Tan, Nihon garon taisei vol 1 (Arusu, 1929), p.11.

[2] Gadô yôketsu, p. 23.

[3] Zhang Huaiguan (a. mid-8th century), quoted in Susan Bush and Hsio-yten Shih, Early Chinese Texts on Painting, (Harvard University Press, 1985), pp. 79-80.

[4] The two legends were often conflated or confused, for while Zhang's feat was the greater, Wu was the more famous artist, see for example, Tachibana Morikuni's Oshukubai, of 1740 (Museum of Fine Arts, Boston).

[5] He also named Gu Kaozhi and Lu Tanwei; see Yoshihiro Yonezawa & Chu Yoshizawa, Japanese Painting in the Literati Style (Heibonsha, 1974), p. 16.

[6] Hashimoto Kansetsu (1883-1945); this annecdote is cited in the notes to the Shohaku Exhibition held at the Museum of fine Arts, Boston, April, 1991. The Museum takes the story to refer to their own vast dragon screens.

[7] In the Preface to a work by Ling Mengchu, quoted in Judith Zeitlin, Pu Songling's "Giaozhai zhiyi" (unpublished Ph. D dissertation, Harvard, 1989), p. 263. Shixiang's name is a pseudonymous and his real identity is unknown.

[8] Mugen dôjin zakki (Toyo Bunko 309, 1977), p. 272; the work referred to is the Confucian classic compiled in the first century B.C.

[9] The myths of Pygmalion and Daedalus have been all but ignored in the West, and as David Freedberg has recently mused in the case of Pygmalion, "It seems so peculiarly apt for illustration, that one can only wonder that its vogue - before the nineteenth century - was not greater", The Power of Images: essays in the history and theory of response (Chicago University Press, 1989), p. 340.

[10] Cited in Sugano Yo, Nihon dôban-ga no kenkyû, (Bijutsu Shuppan-sha, 1974), p. 224.

[11] The book was Jan van Vondelen's Vorstelijcke warande der Dieren (Amsterdam, 1699); copies of certain of the drawings, thought to be by Kôkan, are preserved in Kanazawa, see Sugano Yo, "Edo ki denrai no bijutsu kankei Ransho" inKobijutsu 53, 1977.

[12] This claim is now accepted as standard, see, inter alia, Sugano, Nihon dobanga no kenkyu, p. 223.

[13] Natural History, Book 35, Loeb Classical Library (Harvard University Press, 1952), pp. 309-11.

[14] It is posssible that an accompanying screen once existed, and this would have had to stand on the right for compositional reasons; this could have shown Parrhasius, which would fit with Gottfried's original plate. However, I consider this possibility unlikely, not least because a similar screen by Kokan, the Western Scholars in Disputation, now in the Yabumoto Collection, Tokyo, has no pair.

[15] Natural History, p. 311.

[16] This phrase is used by Norman Bryson in,Vision and Painting (Yale University Press, 1983), p. 13 and passim.

[17] Kindai meika shoga dan, of Anzai Un'en (1830), cited by Onishi Hiroshi (et al.), Geijutsu-ka densetsu (Perikan-sha, 1989), p. 236.

[18] The work is said to have been 'Nanban' i.e. either a seventeenth-century Western work or a Japanese work in the Western style, see Onishi, Geijutsu-ka densetsu, p. 239.

[19] Ueda Akinari shu, pp. 70-6.

[20] Ueda Akinari shu, p. 76.

[21] The collection has been identified as the Kokon chômon-shû, of 1258 edited by Tachibana no Narisue, which contains many artists legends.

[22] Reproduced in Nihon shiso takei 64, Yogaku (Iwanami Shoten, 1964), p. 59.

[23] Bunchô gadan, quoted in Sasaki Sei'ichi, Nihon kinsei bijutsu-shi vol. 1 (Ruri Shobo, 1986), p. 231.

[24] See Max Loehr, The Great Painters of China (Phaidon, 1980), p. 27.

[25] Ueda Akinari shu, p. 76.

[26] Gesaku rokkasen, quoted in Leutner, Shikitei Sanba, p. 36. Karasu-tei Enba was a famous Edo wit, and the founder of rakugo.

[27] Kokan visited the Taiga-do studio in 1789 (Taiga having died in 1776), but had to go twice before gaining any information about it, and even when he succeeded in contacting the present resident, he found that the man had never heard of Taiga or his equally famous painter wife Gyokuran, see Saiyu nikki, pp 195 & 201.

[28] Mugen dojin hikki (Toyo Bunko 309, 1977), p. 243.

[29] For example, Shunparô hikki, reproduced in Nihon Shiso taikei (2nd ed.), vol 2 (1975), p. 21, and Seiyô gadan, p. 492.

[30] Several sekiga sessions are recorded in the Saiyû nikki (Toyo Bunko, 461, 1986), for example, pp. 15, 51 & 128; a Western-style painting by Kôkan in the Powers Collection may be the result of a sekiga, but, significantly, the work is unfinished and little more than a roughly produced assemblage of monochromatic sketches.

[31] Onishi, Geijutsu-ka densetsu, p. 227.

[32] Saiyû nikki, for example, p. 202.

[33] Naruse Fujio, "Shiba Kokan no Kansei 11-nen" in Yamato Bunka 1985, p. 32-40.

[34] Naruse does not suggest any reason why Kôkan's picture should have been late; he does however suggest that the offering may have been the extant Fisherman on a Beach.

[35] Saiyû nikki, p. 53.

[36] Shunparô hikki, p. 22.

[37] Gahô kôryô, p. 71.

[38] Shunparô hikki, p. 22.

[39] Reproduced in Edo kagaku koten sosho (Inawa Shuppan, 1979) vol. 17, p. 45.

[40] Lensless camera obscuras had been known before, and indeed, the simple fact that outside views can be projected into darkened interiors is likely to be discovered independently in many cultures; Gentaku referred to Chinese examples, which he called ringa-kyô. Hokusai's Fugaku hyakkei of 1834 had a scene entitled Saiketsu no Fuji showing a man demonstrating a naturally produced camera obscura, and Bakin (see below) referred to a similar effect produced in the house of one Niwa Yuzaemon on Omiya-dori which projected the pagoda of the Toji outside, and in another room in Shinshu. It seems, however, that the pictures were reckoned to be marvels of the peculiar rooms concerned, rather than effects that could be created anywhere.

[41] Kage to hinata chinmon zue, from the first printed edition in the National Diet Library, p. 1 verso.

[42] Kiryo manroku, section 23.