The Japan Society of the UK

Home Education Events Library Lectures Reviews

Japan Society Lectures Upcoming Lectures
and other events
Recent Lectures
Past Lectures
Joint Seminar Series

JAPAN SOCIETY LECTURE

Kano Sansetsu (1590-1651) and the Song of Everlasting Sorrow (Chogonka) - Sources and Resources, from Tang China to Edo Japan
A lecture given to the Japan Society and Japan Research Centre at SOAS, University of London, on Tuesday 7th November, 2006,
by Shane McCausland (Chester Beatty Library, Dublin)

ABSTRACT The focus of this lecture is the pair of picture-scrolls entitled Song of Everlasting Sorrow (Chogonka gakan), once owned by Louis Gonse (1846-1921) and now in the Chester Beatty Library collection. Created by the Kyoto Kano School master Sansetsu in the early Edo period, the painting depicts the tragic love story of the Tang emperor Xuanzong and his beloved concubine Yang Guifei, as told by the great Tang poet Bai Juyi (772-846; Haku Rakuten in Japanese) in his epic ballad, 'Song of Everlasting Sorrow' (806). The subject has had an intriguing history in Japan, ever since it was first mentioned in the Tale of Genji but it is the style of the painting that compels interest here. It opens a window on how Sansetsu (1590-1651) organized his painting workshop and raises questions about the artist's pictorial sources in Ming China. This lecture also provides an opportunity to introduce a recently published facsimile of the Chogonka scrolls, Kano Sansetsu ga 'Chogonka gakan' (The 'Song of Everlasting Sorrow' picture-scrolls painted by Kano Sansetsu; Tokyo: Bensey Publishing, 2006), the second volume in an ongoing project to publish the major picture-scrolls and Nara ehon (Nara picture-books) in the Chester Beatty Library's Japanese collection.

TEXT The topic of the talk is a painting that tells the tale of a tragic love-affair: the one between the ageing Tang (618-907) emperor Xuanzong (685-762; r. 712-756), also known as Minghuang, and his beautiful, famously 'plump' concubine Yang Guifei (719-756; Japanese, Yogiri). This relationship is variously cited as the cause of, or the pretext for, the great An Lushan rebellion in 756, which almost toppled the Tang dynasty. The story can be told using a dozen or so slides, as follows:

Fig. 1: In his later years on the throne the emperor Xuanzong tires of government and increasingly devotes his attention to beautiful women. Yearning for 'a lady to topple a dynasty', he sends scouts throughout the empire to bring beauties into the imperial harem. One of these new arrivals, a lady of the Yang clan, is seen one day by the emperor stepping frailly out of the hot spring at Huaqing. He is smitten.

Fig. 2: In a scene depicting their nuptial bedchamber, the emperor is shown to have eyes for none among his several thousand concubines but the Yang lady, now promoted to the rank of Precious Concubine (Guifei). Over the ensuing decade and a half to 756, she uses her power and influence to advance members of her own clan, while the emperor, consumed by a grand passion for his lady, neglects the important business of government.

Fig. 3: In 756 rebellion breaks out, led by Yang Guifei's adoptive son and a favourite of the emperor, General An Lushan, who believes he can unite the empire under a better government than Xuanzong's. Xuanzong, Yang Guifei and their inner circle flee in the night, as the rebel armies approach the Tang capital Chang'an (modern Xi'an) from the east. Their aim is to escape through the mountains of southern Shaanxi to the Sichuan region in the south west, the Yang clan's power base. The scene in the picture shows the rebel army of An Lushan torching and looting the imperial palace at Chang'an.

Fig. 4: A 'satellite view' from 650 km above earth (courtesy of Google Earth; not shown) and a topographical map give an idea of the mountainous terrain through which the imperial party intends to escape - along the 'plank roads' of southern Shaanxi, across the Jian'ge Pass and down into Shu (the ancient name for Sichuan).

Fig. 5: The imperial party begins its journey west from the capital along the Wei River, but is held up after just 50 km at Mawei, where the imperial bodyguard mutinies, demanding the deaths of the most powerful members of the Yang clique, including Yang Guifei. With the rebels approaching the capital not far behind, the emperor, for the sake of the dynasty, has no choice but to accede to this demand. The scene depicts Yang Guifei, seated on a mat, about to be strangled to death with a silken cord.

Fig. 6: The scene shows the emperor inconsolable with grief, having just been told the deed has been done. The imperial party must keep moving, but the emperor is unable to look upon the spot at Mawei where Yang Guifei was killed, and her body thrown in a ditch.

Fig. 7: A slide shows the imperial party passing along perilous plank roads affixed to sheer cliffs and in the near darkness under the towering cliffs of the Emei Mountains. The emperor arrives in Shu, and continues to govern the state until news arrives that the Heir Apparent has usurped the throne. Again, Xuanzong has no choice but to accede to this.

Fig. 8: After a year's exile in Shu, and once order has been restored throughout the empire, Minghuang is permitted to return to the capital as 'retired emperor'. The slide shows him back in the imperial palace, where he is reminded at every turn of his past life with Yang Guifei: alone with his thoughts, he gazes on a lotus pond in bloom (Yang Guifei was often compared to lotus flowers), while beyond is a row of willow trees (trees to which she was also compared). The fallen autumnal leaves on a bridge beyond reflect his sombre mood.

Fig. 9: The retired emperor, tormented by grief for his lost love, hears of a Daoist wizard who has the ability to find and speak with the spirits of the deceased. Minghuang summons him and sends him to find the soul of Yang Guifei.

Fig. 10: The Daoist is seen searching the heavens and the depths of the ocean.

Fig. 11: The Daoist hears tell of a lady living in the isles of the immortals across the sea, and flies there. He asks for her at the door of the palaces of the immortals and is granted an interview with the soul of Yang Guifei. He is seen preparing to return to the world, bringing with him love tokens for Minghuang.

Fig. 12: In this pictorial retelling, the tale ends on an image of a 'bridge to heaven' which symbolises the eternal divide between the worlds of the living and the dead.

One might say that such a tale had all the features to give it a keen edge in society, politics, literature and painting through history. It has beauty (slide: Yang Guifei) hidden, discovered and resented; kingship, including examples of good and bad government (slide: Minghuang); romance and passion (slide: nuptial bedchamber); rebellion and violence (slide: torching of a palace - and how painters love to portray a good fire!); mutiny and execution (slide: Yang Guifei's execution); a perilous journey (slide: escape through the mountains); mourning and longing (slide: Minghuang grieving); the supernatural and the afterlife (slide: the Daoist).

The topic has indeed been nestled deep in the artistic imagination in China since the great Tang poets Li Bai and Bai Juyi (772-846; Japanese: Haku Rakuten), among others, wrote about it. Li Bai's famous poem is entitled 'Hardships of the Road to Shu' ('Shu dao nan'), and Bai Juyi's 60-couplet ballad composed in AD 806, is 'Song of Everlasting Sorrow' ('Chang hen ge'; Japanese: 'Chogonka'). For both these poets, this subject was one of both romantic escapism and grave admonition.

Similarly, it has lain deep in the artistic imagination in Japan since the Heian period (794-1185). An allusion to it appears in the opening paragraphs of the early eleventh-century novel, Tale of Genji (see Chapter 1, The Pawlonia Court), where the unnamed Japanese emperor's favourite is likened to Yang Guifei by courtiers as negative example - to imply that his dallying with this beauty will lead to neglect of state affairs and ultimately to rebellion and chaos. In a wonderfully ironic insight on the part of the Genji's writer, however, there is the sense that the emperor is not at all disapproving of this comparison, which sees his favourite likened to one of the most beautiful Chinese women of all time, Yang Guifei. As the narrative unfolds, it will emerge that prince Genji is, of course, the product of this tragic union.

By the middle of Chapter 1, when this favoured lady is hounded to death by jealous courtiers, the historical parallel becomes even more obvious, as Minghuang becomes the self-image for bereaved emperor in Genji. The latter is said to become 'addicted' to illustrations of Chogonka by the ninth-century Japanese emperor Uda and recites verses on the subject by Japanese and Chinese poets. In a brilliant paradox, the emperor keeps on revealing his blind-spot vis-�-vis the moral-political lesson of the story.

Chogonka illustrations
The Chogonka tale of the emperor and his concubine was evidently a painting subject in China by the mid-Tang dynasty; it was also painted in the succeeding Song (960-1279) and Yuan (1271-1368) periods, as evinced by surviving early images, especially of mountain scenes relating to Li Bai's 'Shu dao nan'. Not surprisingly, it was something of a touchy subject in China, and was widely known by euphemisms like the 'picking melons' picture (Minghuang's concubines are said to have picked melons by the roadside during the flight to Shu) and 'Minghuang escaping the summer heat'. The Tale of Genji is itself evidence that the subject was also painted in Heian Japan, by one emperor painter himself. However, neither these nor those early Chinese paintings are the main focus here. The purpose is rather to look at one of the most polished extant versions of Chogonka, painted during in the early Edo period (about 1600-1868), by Kano Sansetsu, master of the Kyoto Kano school from 1635 to his death in 1651.

The inscription and signature at the end of this scroll reads (see fig. 12): 'Kano, of many generations, Sansetsu, painted this for the first time'. Quite what is behind the curious and seemingly contradictory words 'of many generations' and 'for the first time' has long exercised the minds of viewers. The 'of many generations' implies that the artist was deeply aware of his Kano lineage; and there are a number of images of this general topic by Kano School masters from the late Momoyama (1573-1603) and early Edo Japan, among them Mitsunobu (1561/5-1608), upon which he could have drawn. They are in the main highly decorative, making lavish use of gold, but are less 'narrative paintings' than genre-type scenes or 'scenes from the life of' Minghuang and Yang Guifei, and as such, are almost closer to the Chinese tradition.

The subject was also painted by Sansetsu's Kano School master Sanraku (1559-1635). Popular Nara ehon (Nara picture-book) versions dating to the sixteenth and early-seventeenth centuries are also known. This awareness of a Kano lineage contrasts with the claim to uniqueness in the words 'painted this for first time'. The problem arises as to exactly what that means: did Sansetsu illustrate the topic in emaki (narrative handscroll) format for the first time? Did he invent/reinvent the scenes, type-forms and iconography? Or did he simply mean this was his first attempt at the subject? These are some of the questions we will keep in mind as we look more closely at the painting.

The painting is a two scroll set in Chester Beatty Library. Each scroll measures about 33 cm high by 11 m in length. Each is painted on six sheets of silk, measuring about 180 cm in length. The context of its creation in mid-seventeenth-century Kyoto remains a puzzle. On this point, Miyajima Shin'ichi has suggested a possible connection with the Kujo, one of the leading families of the city. He has noted that Kujo Kanezane (also known as Fujiwara no Kanezane), the founder of this clan, which was one of the five so-called regent houses of Japan (sekke), presented a version of the Chogonka scrolls to the cloistered emperor Go-shirakawa (r. 1155-58) 'presumably with the intention of admonishing him'. Further, it is known that Sansetsu worked under the patronage of the Kujo family in Kyoto¹, but beyond this, we must draw our own conclusions.

The scroll set was once owned by Louis Gonse (1846-1921), who seems to have acquired it in the later nineteenth century and it was subsequently bought by Chester Beatty, probably between the wars, entering his Library in Dublin when he retired to that city after World War II. On Chester Beatty's death in 1968, the Chester Beatty Library became a charity maintained by the Irish state and the painting now belongs in this National Cultural Institution. In the 1990s, the scroll set was restored and remounted through the generosity of friends in Japan, and pictures of the scenes of Yang Guifei's execution before and after restoration show the transformation². This year (2006), the scroll set is published in a handsome facsimile volume by Bensey Publishing of Tokyo, Kano Sansetsu ga 'Chogonka gakan' (The 'Song of Everlasting Sorrow' picture-scrolls painted by Kano Sansetsu; see www.bensey.co.jp). The scroll set was also published in a CD-ROM made in 1998 by two former curators at the Chester Beatty Library, Yoshiko Ushioda and Clare Pollard. Despite the breakneck speed of development in the fledgling 'information age', this interactive compact disk still holds its own.

There are a number of approaches one could take in exploring this painting, such as the dating problem (which will be touched on), the provenance, or the narrative technique, but this study will address the scrolls primarily from angle of their production in a studio practice. There are three aspects to this. The first we come to in a moment. The second is to explore some of the pictorial and stylistic sources in China - to try to go beyond the usual comparisons, such as of the mountain scene with the anonymous Song painting, Minghuang Escaping to Shu (National Palace Museum, Taipei), which was not a direct source, although its traditions would have been - to ask what Chinese artists, paintings and traditions were regarded as sources and to consider how these might have added to the authenticity of the painting. The third is, by way of conclusion, to try to relate Kano Sansetsu's painting practice to contemporary practice in China, and we will touch on that at the end.

First, however, we attend to Sansetsu's workshop practice itself, and try to imagine the role of master and the uses to which he put his assistants, who included his son and disciple Eino (1631-97). This can be done by examining two virtually identical versions of the Chogonka composition that appear to have come from the Sansetsu studio in the years 1646-7: the one in Dublin, and another in a private collection in Japan (the Kuki family). One scene of the latter version was published in Kokka no. 44 in 1893 (see fig. 13: the escape through the mountains), and it appeared again in a rare 1916 publication before disappearing for much of the last century. The scroll resurfaced again in late 1980s, seemingly in the same family collection, when a set of black-and-white study photographs were made. At present, this version is still part of a private collection in Japan, where it may not be available for first-hand study. Discussion of it in this paper is based on the set of study photographs in the Chester Beatty Library placed there by Mrs Ushioda.

Fig. 13 The escape through the mountains to Shu. From Kano Sansetsu (1590-1651), Song of Everlasting Sorrow picture-scrolls (Chogonka gakan). Detail from a set of three handscrolls; ink and light colours on paper. Private collection, Japan. After Kokka (1893)�.

What is the relationship between the two versions? First of all, the Kuki version is on three scrolls, and is painted on small sheets of paper about 25 cm wide. Said to be in ink and light colours, the decorative details in the Kuki version are much simpler than in the Dublin one. At first glance, it resembles one of Sansetsu's preparatory drawings, such as those made for his screen compositions, but a closer look shows it to be far more detailed.

The last Kuki scroll bears an inscription including a date in the first lunar month of the year corresponding to 1646; a note to say the painting was done in 36 scenes; and a signature reading 'Kano Hokkyo' (Hokkyo being an honorific title Sansetsu received in 1647 for executing an imperial commission)³. In addition, the three scrolls of the painting feature inscriptions of selected passages from Bai Juyi's ballad, with occasional identifications of places, above or beside each of the three dozen scenes. The Dublin painting has none of these texts from the ballad.

With regard to the overall plotting of the narrative, there are reasons to think that the two-scroll format of the Dublin version was the intended final form. The death of Yang Guifei, for instance, occurs at the exact midway point of the Dublin version (fig. 14). She is about to be executed at the end of the first scroll, while at the beginning of the second, the emperor is pictured inconsolable with grief at the news of her death. As such, the cruel deed itself happens off-stage, as it were, between the scrolls, and the event acts as a searing visual caesura within the overall plotting. In Bai Juyi's ballad, this event occurs just one third of the way through the 60-couplets of the text.

Fig. 14 The end of the first scroll and beginning of the second. � Trustees of the Chester Beatty Library, Dublin.

Looking now at the phrasing of the narrative in the Chogonka scrolls, we find this develops, roughly, in 12 sets of three scenes, rather like 'acts' of a play. Each 'act' of three scenes takes place over one 180 cm length of silk in the Dublin version. There are, for instance, three scenes describing the love affair; the rise of the Yang clan through favouritism and nepotism; the escape through mountains; the retired emperor in mourning back in Chang'an; and the emperor engaging the Daoist. The implication is that this length of about 180 cm, which corresponds to one length of silk in the Dublin version, determined to a large extent the organisation of whole narrative set.

What can be said about qualitative differences in the execution of the two paintings? At even a cursory glance, it is clear that the Dublin version is more lavish. It is painted on silk, which takes decorative detail and infilling better and it makes prominent use of bright colours and gold ink - the kind of task that could be entrusted to assistants in the studio or workshop. The Kuki version is described as being in light colours on paper, and is accordingly more linear in effect. Looking closely one may even observe one or two 'misreadings' of the Kuki version in Dublin version, for instance, in the area depicting the sleeping groom in one of the scenes describing the rise of the Yang clan from obscurity due to Yang Guifei's influence.

Comparing the tack of the two horses (the one seen between the two trees to the right of the sleeping groom) in the Kuki version, the leathers used for the bridle, reins and martingale are all most naturalistically painted in different widths, according to their purposes: the reins are thin; the bridle thicker and the martingale is thicker still. In the Dublin version, these differences are somewhat lost (fig. 15). A particularly noticeable difference is how the martingale relates to the horse's musculature. In the Kuki scroll, it is tight across his front quarters and digs into his muscles, creating a visible groove; in the Dublin scroll, this groove of muscle still appears in the animal's front quarters, but the thin martingale rides up above it - removing any naturalistic reason for the groove.

Fig. 15 A horse and resting groom from the scenes depicting the rise of the Yang clan. � Trustees of the Chester Beatty Library, Dublin.

One might say that Dublin version is richer and more 'finished', but it may also be described as somewhat less lyrical, and lacking the spontaneity and naturalness inherent in a more linear style. Nonetheless, it is evident that a firm executive hand underlay the conception and structure of the Chogonka composition, which tallies with Sansetsu's insistence that he 'did it for first time', but there are also interesting suggestions of delegation in the execution of the painting, as one would expect in a polished version on silk from a Kano School studio.

We now go back to a moment earlier in the production process to look at Kano Sansetsu's background and his research on stylistic sources in China. Sansetsu joined the Kyoto Kano school under Sanraku and later married his daughter. On Sanraku's death in 1635 he became master, a position he held until his own death in 1651. The Dublin painting dates to toward the end of that period: 1646-7. Despite his leadership role in the Kano School, Sansetsu could be described as something of an eccentric, or a genuine original.

Although he never visited China, Sansetsu was a sinophile. His signatures proudly display this, being sometimes written, as on the Dublin Chogonka, in clerical script, which even in seventeenth-century China would have been construed as an antiquarian, scholarly mode of writing. Like his Kyoto contemporary and rival Kano Tan'yu (1602-74), who had an encyclopaedic knowledge of Chinese paintings in Kyoto collections, he admired Chinese painting and would have known, if not that painting itself, then the tradition of Minghuang's Flight to Shu in Taipei, which had been transmitted to Japan via Yuan masters like Wang Zhenpeng and followers, like Li Rongjin. Sansetsu was a scholar of painting: he compiled Honcho gashi (The Painting of this Realm), completed by his son and pupil Eino. In addition, he was in tune with the Kyoto patrons of the China School, as evidenced by his many paintings depicting Chinese subjects, such as his West Lake screens depicting the Southern Song capital, Hangzhou.

The resources available to Sansetsu about China and about this subject comprised not just the oeuvre of earlier Kano School masters, including Mitsunobu and others, but also Kyoto collections that included Chinese paintings. Although knowledge of these resources survives in the form of the many sketches of Chinese paintings made by Kano Tan'yu, Sansetsu's access to them is less clearly understood. Perhaps more in vogue as contemporary sources about China were imported woodblock-print-illustrated books of literature and encyclopaedia, including such books as Hainei qiguan (Extraordinary Views of the World [i.e. China]) of 1609, which catered to a growing market for tourist guides and travel literature as well as Sancai tuhui, the illustrated encyclopaedia of the same date.

It is worth noting that the social-economic context for these publications across East Asia comprised a dramatically expanding literacy rate and new opportunities for social mobility; broadening global horizons, as new knowledge and knowledge systems were being introduced, by Jesuits for example; and great leaps in print technology, which also contributed to the boom in the publishing industry.

Broadly speaking, the stylistic features of the Dublin painting belong to the East Asian narrative tradition. In the treatment of figures, for instance, the social elite are by convention portrayed in a dignified manner and given mask-like faces. The low-life and riff-raff, including porters, grooms and rebels are, by contrast often rendered in burlesque caricature. The pictorial depiction of architecture, scenery, time and season is also conceived so as to suggest mood and to create atmosphere. The 'plank roads', for example, are an appropriate image of the perils faced by Minghuang and his entourage as they escape the capital and of the precariousness of human affairs. Elsewhere, intimate interiors conjure late-night passion, while towering, far-off palaces create an 'other-worldly' effect.

What stylistic sources from China can be identified, and what role might Sansetsu have assigned to them? The point in asking these questions is not to suggest that China can resolve the puzzle of this painting for us. But it should be borne in mind that despite its long history in Japan, the narrative is Chinese and that Sansetsu was from the China School; we would assume that he could have wished to invest some degree of cultural authenticity into his painting by means of a judicious convergence with Chinese sources that were either known or becoming known in his circle of patronage. In a recent article on the Dublin Chogonka painting in University Press (June 2006), Itakura Masaaki identified various Chinese painters and their works as sources, including Wu Bin (active 1568-1626), Tang Yin (1470-1524) and Qiu Ying (1494-1552) - painters active in the century and a half to the mid 1640s, who were all representative of mid-to-late Ming literati painting culture.

The figure flying through the sky in Tang Yin's handscroll Dreaming of Immortality in a Thatched Hut (Freer Gallery, Washington DC), is certainly an attractive-looking precedent for Sansetsu's figure of the Daoist magician flying about in the Chogonka. Tang Yin's oeuvre is also a likely source for some of Sansetsu's garden scenery. Qiu Ying is another possible source, particularly for the inner palace scenery, including courtyards, colonnades, and furnishings, and indeed for the figures of elegant palace ladies. These subjects are all clearly visible in Qiu Ying's masterful handscroll Spring Morning in the Han Palace (National Palace Museum, Taipei). Qiu Ying is also a most plausible source for the ethereal palace exteriors and polychrome effects of the immortal realms⁴. All these Qiu Ying effects would have been known in Japan through the many Qiu Ying fans transmitted there⁵. In addition, a number of other painters and paintings could be cited, but space is limited.

Let us take a closer look at just one section of the Chogonka painting with significant links to Chinese painting: the section describing the escape through the mountains to Shu. Mooted as a source here by Professor Itakura was the late Ming painter Wu Bin, a master of 'weird' mountains. Any number of his paintings in attenuated formats of attenuated forms could be identified as sources for the needle-pointed peaks and strangely-shaped mountains we find here: Steep Ravines and Flying Cascades (National Palace Museum, Taipei) of before 1610; Thousand Cliffs and Myriad Ravines in the Palace Museum, Beijing; Landscape of the Road to Shanyin in the Shanghai Museum (fig. 16); and so on. Most of these exemplify the late Ming vogue for 'the strange' (qi), which often takes form in exaggerated and sometimes romanticised regional stereotypes. The 'plank roads' of southern Shaanxi is an example of a local infrastructure that had by late Ming become a diversion for the armchair (or daybed) tourist.

Fig. 16 Wu Bin, Scenery along the Road to Shangyin. Detail of a handscroll. Shanghai Museum.

Other important sources known to Sansetsu include the illustrated books, Hainei qiguan and Sancai tuhui. Intriguingly, both have pictures of the 'Plank Roads to Sichuan'⁶ and of 'Mount Emei' (see figs 17 & 18, for the Sancai tuhui prints). Comparing them, it is evident from the lesser pictorial legibility that the Hainei qiguan images were either cribbed from those in the Sancai tuhui, or were less well copied from a mutual source. Curiously, the Hainei qiguan pictures make a greater claim to cartographic accuracy by virtue of the place name captions that have been somewhat arbitrarily added. Let us look more closely at one of these, the 'Plank Roads to Sichuan'. Geographically, these road lie southwest of Baoji (a city due west of Xi'an along the Wei River) and lead south through the precipitous terrain toward the mountain passes, of which the Jian'ge Pass is perhaps the best known, into the Sichuan basin.

Fig. 17 �Plank Roads to Shu�. Scene from the Chinese illustrated encyclopaedia Sancai tuhui (1609) (with author�s annotations).

Fig. 18 �Mount Emei�. Scene from the Chinese illustrated encyclopaedia Sancai tuhui (1609).

Said to have been built in the Qin dynasty (221-206 BC), these roads were once the only route from Xi'an down to Sichuan. The Hainei qiguan print is somewhat fanciful as a topographical picture, but effectively shows how the mountains loom above the main river arteries bounding the region: the Wei to the north; the Jialing River to the west, and the Han River cutting from east to west through the middle. The illustration may roughly be taken as a view looking west (see annotations to fig. 17). Baoji is positioned at the bottom right; above it is the Dasan Pass (about 30 km to the south-west), and above that is Fengling, probably the peak above the county town of Feng County, about 50 km south-west of the Dasan Pass. Baocheng (called Liangzhou in the Tang), which lies some 150 km to the south of Baoji on the Han River, is named to the left of Baoji at the bottom. The Jiangling River is named in the bottom left corner. The pass to the south at Qipan, roughly 150 km south-south-west of Baocheng, is named in the top left corner.

This print of the 'Plank Roads to Sichuan' must have been of some interest to Sansetsu, if only as corroboration of the wooden cliff-face structures seen in paintings by artists like Wu Bin, or for a general iconography of 'snow-capped peaks' in the distance, and a river-below-mountain effect. Nonetheless, although it is perhaps out of character for an art historian to admit it, the role played by these images in shaping this mountain passage must be said to pale in significance next to the effect of certain lines in Bai Juyi's ballad. We find here a quite literal interpretation of the lines, 'The clouded plank road twisted up to the Jian'ge Pass/ Under Mount Emei they went in small numbers/ The day-bright colours of the imperial banners faded in the darkness.' In the painting, the party proceeds from Jian'ge Pass along the plank-roads and under the gloomy, enclosing cliffs of Mount Emei immediately after, as if these places were close rather than hundreds of miles apart: Sansetsu quite literally borrows Bai Juyi's poetic licence. It would seem from such a direct interpretation from the literature that the plotting of the scroll vacillated between being more text or more image driven. This textual conceit would no doubt have been appreciated by the more literary minded of Sansetsu's patrons, and yet pictorial references to Chinese paintings and pictures also lay within.

We may recall that this mountainous section of the scroll does not stand alone, but lies between two depictions of Mawei, the place of Yang Guifei's execution. Before it, is a passage of painting describing how Minghuang was unable to gaze upon the spot (as Bai Juyi put it in the ballad). After it, is an interesting scene of him passing Mawei which is painted in the manner of a Tang 'space-cell' (recalling scenes from paintings like the Tang painter-poet Wang Wei's iconic Wangchuan Villa scroll), and this merges into a scene of his entourage arriving back at the capital from exile. Although the mountain passage represents a discrete section of the story, it also takes its place within the whole. Citing Wu Bin or Hainei qiguan as sources, therefore, does not give full satisfaction: these 'type forms', or references to them, are deployed in the service of this much grander plot, and appear little more than playthings in hands of the maestro.

We come now to the third and perhaps the most challenging part of this essay, where we undertake to compare Sansetsu's workshop practice with that of a contemporary in China, namely Chen Hongshou (1598-1652). Chen Hongshou is much loved eccentric artist, who lived near Hangzhou. As a painter, he blurs the categories of 'professional' and 'scholar'. He passed the lower-level provincial examinations but never qualified to take the more prestigious higher level exams that would have opened his path to an appointment in the civil bureaucracy and resorted, albeit reluctantly, to making his living from painting. His work, which often included illustrations of ancient narratives, is known for its ironic humour, and for reviving ancient figural styles, such as the 'grotesque' style of tenth-century Chinese Buddhist portraiture and the lyrical outline of literati figural masters including Li Gonglin (c. 1041-1106). The latter style is visible in the Portrait of He Tianzhang (fig. 19) in Suzhou Museum, which is roughly contemporary with the Chogongka painting, dating to the mid 1640s.

Fig. 19 Chen Hongshou (1598-1652) and disciple Yan Zhan, Portrait of He Tianzhang. Handscroll; ink and colours on silk. Suzhou Museum. Photo courtesy of the Suzhou Museum.

In this painting, Chen Hongshou's patron and subject, He Tianzhang, is seen seated in the luxuriant surroundings of his garden, surrounded by the trappings of elegance and taste: a long stone table, a mature pine, a beautiful lady, and a musician who also brews fine tea. What is perhaps unexpected about this painting is what the artist declares in the inscription, where he writes that he painted the 'caps and clothes', that is, the figures, and that 'disciple' (menren) Yan Shuizi (Yan Zhan) filled in the scenery. The figures are indeed beautifully painted in lyrically flowing outlines, lines that pulsate with rhythmical energy, as we have come to expect in the finest works by this consummate figural artist. True to the artist's word, the execution of the painting evinces a division of labours, with the scenery patently executed by another hand, as the 'muddier' and less self-assured handling of ink indicates.

The Portrait of He Tianzhang offers a comparison with the scene of the nuptial bedchamber in the Dublin Chogonka painting. Particularly we may compare the two beauties (figs 20 & 21). Although different ideals of beauty are portrayed here, for Yang Guifei is the model of the Tang 'plump' beauty whereas the lady of He Tianzhang is the willowy Ming type with bound feet, there is a remarkable convergence of iconography. The endlessly trailing tassels of both ladies' drapery, for instance, is rendered in that Gu Kaizhi (c. 344-c. 406) idiom well-known from early paintings like the Admonitions of the Court Instructress (Ch. N�shi zhen tu; J. Joshi shin zukan) in the British Museum. Even so, there is a distinction in the lineament style. Chen Hongshou's outline harps back to that 'gossamer thread'-like delicacy celebrated in Gu Kaizhi's figures, later taken up by distinguished Song scholar-painters like Li Gonglin. By contrast, Sansetsu's line appears more modular, as if pieced together in phrases; the ink has a blacker, more organic and brittle quality that recalls the style of consummate technicians like the Yuan painter Wang Zhenpeng, whose work could be seen as a later, Yuan interpretation of the Li Gonglin mode and who would have been known to Sansetsu through transmissions to Japan in the fourteenth century.

Fig. 20 Emperor Xuanzong and Yang Guifei in their nuptial bedchamber. � Trustees of the Chester Beatty Library, Dublin.

Fig. 21 Detail of fig. 18.

Another point of comparison is the role of assistants. For Chen Hongshou, the apprentices or 'disciples' in the studio were useful fillers-in of paintings he apparently did not have the time or inclination to finish. We do know that on occasion they painted for him, even signing his name, at the request of patrons; and are beginning to understand that they also created multiple versions of paintings such as 'birthday presentation pictures' in his name⁷. The creation of the Chogonka painting appears, for its part, to have represented a moment in the apprenticeship of Sansetsu's son and disciple Eino, who was about 16 in 1646. This painting was the inspiration behind Eino's later five-volume woodblock-printed book Chogonka zusho published in the early 1680s, making it roughly contemporary with the Chinese artist Wang Gai's highly influential woodblock compendium of past artists' styles, Mustard Seed Garden Manual of Painting (Jieziyuan huazhuan; first published in 1679). Eino's book features an annotated transcription of Bai Juyi's ballad interspersed with illustrations based mainly on the three dozen scenes in his father Sansetsu's painting (compare fig. 8 with figs 22 & 23). The British Library has a hand-coloured version of this illustrated book, which is yet to be studied in depth.

Fig. 22 Gazing at a lotus pond, Minghuang is reminded of his lost love Yang Guifei. From Kano Eino (1631-97), Chogonka zusho (5 vols, 1677). After Sakakibara Satoru, �Chogonka e no koto� [The matter of the Song of Everlasting Sorrow painting], in Chiesuta Biti Raiburari (series Hizo Nihon Bijutsu Taikan, vol. 5 [Tokyo: Kodansha, 1993]), pp 276-280.

Fig. 23 Fallen leaves on a flight of steps. Source as in fig. 22.

By way of conclusion, we now step back and see what broader perspectives can be taken on this juxtaposition of Chinese and Japanese paintings. In the first instance, the comparison shows that while the Edo artist has a rather eclectic regard for Chinese painting from pre-Tang to Ming China, but with an emphasis on the mid to late Ming, the Ming artist is more focussed on a specific figural tradition going back to Gu Kaizhi via Li Gonglin. Still, there is a remarkable convergence, quite a shared legacy. The artists' sources may differ somewhat, but the art of both carries a contemporary sense of 'belatedness' (the term is David Wang's), an effect of looking to an idealised if flawed Chinese past, one that could no more be recovered in Ming China than it could in Edo Japan.

As for their studio styles, these are both characterised by seemingly contradictory elements: they are collaborative productions, yet they also exhibit strong claims to individuality, suggestive of those aspirations to autonomy which in part come to define modernity. There are signs in these paintings that forms of cultural exchange were going which lie mostly under our radar, but now demand we pay closer attention to the wider 'professionalisation' of painting over the seventeenth century in China and Japan. I put these ideas forward for your judgement; and I recommend this new facsimile volume from Bensey to help you reach it.

Endtnotes
1. Miyajima Shin'ichi, 'Chogonka zukan', in Zaigai Nihon geijutsu no shufuku (The restoration of Japanese artworks in overseas collections), nos 33-34.
2. Ibid.
3. Problems arising from this signature could not be addressed in the context of the lecture, and must be deferred until a first-hand examination of the Kuki scroll itself has been undertaken. These include the somewhat strange placement of the name to the right of the inscription itself. Possibly the space left for the later insertion of a signature was insufficient. In addition, there is an apparent contradiction in the dating, with the scroll dated to the first lunar month of the year corresponding to 1646 by inscription but with Sansetsu only having received the honorific title Hokkyo in the autumn of the following year.
4. E.g., Qiu Ying, Pavilions in the Mountains of the Immortals, dated 1550. Hanging scroll; ink and colours on paper, 110.5 x 42.1 cm. (Illustrated in Fong & Watt, Possessing the Past, pl. 204.)
5. I am grateful to Matthew McKelway for this observation.
6. See Hainei qiguan, juan 8 (1994 reprint, pp. 552-553), for the 'Plank Roads'.
7. See, e.g., Shane McCausland, 'Chen Hongshou and his Studio - a Methodological Brief', in Bridge to Heaven: Essays in Honor of Professor Wen C. Fong (Princeton: Princeton University Press, forthcoming in 2007).

Select Bibliography
(Japanese Titles available on following PDF here (PDF Document)

Bai Juyi (772-846), 'Song of Everlasting Sorrow' (Ch. Chang hen ge; J. Chogonka), AD 806.
Chiesuta Biti Raiburari emaki ehon kaidai mokuroku [Descriptive catalogue of picture-scrolls and picture-books in the Chester Beatty Library]. Tokyo: Bensey Publishing, 2002.
Chiesuta Biti Raiburari. Series Hizo Nihon Bijutsu Taikan, vol. 5. Tokyo: Kodansha, 1993.
Hainei qiguan . 1609; reprint, Zhongguo gudai banhua congkan erbian (Shanghai: Shanghai guji chubanshe, 1994), vol. 8. Itakura Masaaki, 'Kano Sansetsu ga kaita Chogonka zu - itan kara koten e' [The Song of Everlasting Sorrow painted by Kano Sansetsu - from heterodox to classic], University Press, 404 (June 2006), pp 6-12.
Kano Eino (1631-97), Chogonka zusho, 5 vols, 1677. BL: Or.74.cc.7 volumes 1-5.
Kano Eino (1631-97), Honcho gashi.
Kawaguchi Hisao, Chogonka emaki [The Song of Everlasting Sorrow picture-scrolls]. Tokyo: Taishukan shoten, 1982.
Kitano Yoshie, 'Kano Sansetsu hitsu Saiko zu byobu' (The West Lake screens painted by Kano Sansetsu), Kokka no. 1227 (1998).
McKelway, Matthew, Cityscapes: Folding Screens and the Political Imagination in Late Medieval Kyoto. Honolulu: University of Hawai'i Press, 2006.
Miyajima Shin'ichi, 'Chogonka zukan', in Zaigai Nihon geijutsu no shufuku (Restoration of Japanese art in European and American Collections; Tokyo: Chuokoron-sha Inc., 1995), nos 33-34.
Owen, Elizabeth Marie, 'Love Lost: Qian Xuan (c. 1235-1307) and Images of Emperor Minghuang and Yang Guifei'. PhD dissertation, Yale University; Ann Arbor, 2005.
Sakakibara Satoru, 'Chogonka e no koto' [The matter of the Song of Everlasting Sorrow painting], in Chiesuta Biti Raiburari (series Hizo Nihon Bijutsu Taikan , vol. 5 [Tokyo: Kodansha, 1993]), pp 276-280.
Sorimachi, Shigeo, Japanese Illustrated Books and Manuscripts in the Chester Beatty Library, Dublin, Ireland. Tokyo: The Kobunso, 1979, no. 87.
Wang Qi and Wang Siyi (comps), Sancai tuhui (1609). Reprint of Wanli edition in Shanghai Library; Shanghai: Shanghai guji chubanshe, 1988. [Also Heaven and Earth; also Morrison collection has almost complete Qianlong reprint of 1609 edition] Ushioda, Yoshiko, Tales of Japan: Three Centuries of Japanese Painting from the Chester Beatty Library, Dublin (Alexandria, VA: Art Services International, 1992), no. 29.

Home Events Library Lectures

About us Membership and Support

©The Japan Society, Swire House, 59 Buckingham Gate, London SW1E 6AJ
e-mail: info@japansociety.org.uk, Tel: (020) 7828 6330, Fax: (020) 7828 6331
Registered charity 1063952