‘Gaze Upon Its Depth’: On the Uses of Perspectival Painting in the Early-Modern Chinese Village
Hannibal Taubes
In the late nineteenth and early twentieth centuries, a small group of villages in northern Hebei province began to adorn their opera stages and temples with Western-style perspective drawings.[1] These drawings, painted in vigorous blue shades on the walls flanking the temple and stage rooms, depict fantastic amalgamations of Chinese and Western-style buildings, receding into pictorial space: immense multilevel mansions (Fig. 1), pagodas and minaret-like spires (Fig. 2], long rows of buildings lining the streets (Fig. 3), tall domes, and opulent porticoes. Around the bases and atop the roofs, belvederes and porticoes of these structures, complex scenes take place: godly personages sit in state, complex processions set out and return (Fig. 4), more groups of characters gesture or strike operatic poses, while others just lean on the balustrades, taking in the view. This paper attempts to ask the basic questions about these images. What are these pictures? Who are the people and stories in them? Why were they drawn on these walls? What did they mean to the people who drew them?
Prosaically, this paper represents a case study of the early-modern mural-painting traditions of a single Chinese county in northern Hebei province, describing the state of rural artistic traditions on the eve of the introduction of Western-style perspective, and then examining how these traditions changed with Western contact. I contend throughout this paper that these murals represent structural elements of complex compositions of village space, and are not comprehensible without this context. To this end, I begin by giving a brief history of these villages from 1500, focussing on religion, opera, and spatial discourses. In the second section, I narrate the arrival of perspectival techniques and particularly the use of these techniques in opera stage murals, where they were probably first applied. I identify two crucial themes within the stage murals for the ‘Pagoda of Gazing in the Four Directions’ (Si wang ting 四望亭) and the ‘Mansions of the Western Seas’ (Xi yang lou 西洋樓). I argue that these ‘exotic’ architectural drawings establish the village stage as a crucial point of contact with the modernising world. In the final section, I re-examine the logic of both stage and temple murals, focussing on the idea of the temple and stage as ‘porticoes’ or entranceways that derived their power from access to holy, fictional, or otherwise ‘foreign’ spaces that extended away behind the altar or scaenae frons. I then show how in the nineteenth and early twentieth centuries this logic was reworked with Western-inspired painting techniques including perspectivalism and the use of cast shadows, catalysing a renaissance in rural mural painting that produced some of the most beautiful and complex works in the history of this art.
Beyond their simple artistic worth, these images are significant for a number of reasons. Scholarly study of the reception of Western artistic techniques in China, for reasons of preservation, access, and documentation, has tended to focus on images produced either in the imperial court,[2] by members of the broader upper classes,[3] or in well-known artistic centres in the southern cities of Jiangnan.[4] To date, there has been little writing in any language on Chinese temple mural painting after about 1400, or on small-scale temple paintings not connected to elite sponsorship.[5] What painting traditions existed in the early-modern countryside, and what effect Western perspectival techniques might have had on the construction or experience of painted temple or theatre space, are unexplored questions. In a country that was overwhelmingly rural until the turn of the twenty-first century, these images represent a uniquely village creativity, allowing us to see villagers as themselves ‘actors’ in early-Modern Chinese cultural history. From another perspective, it has long been an established fact in theatre studies that Chinese opera did not generally use stage sets. While the rural stage murals examined here were not sets or scene backgrounds in the strict sense, they certainly were elaborately painted tableaus intended to establish the stage as a theatrical space, and thus represent an important expansion to our preconceived ideas of the Chinese theatre. For these reasons, an examination of the apparently recherché painting traditions of a small north-Chinese county takes on some importance, as a case study of how the rural majority of Chinese people in their villages produced art, religion, theatre, and, ultimately, modernity.
1) Fortress, Temple, and Stage
These images are located in Yu county, Zhangjiakou city, Hebei province (Hebei sheng, Zhangjiakou shi, Yu xian 河北省張家口市蔚縣), about 150 kilometres west across the mountains from Beijing. The area is one part of a larger historical and geographical region known as Xuan-Da 宣大, which encompasses the contiguous area between the cities of Xuanhua and Datong north of the Taihang 太行 and Yan 燕 mountains. Over the course of this paper, I will speak specifically about Yu county while making repeated references to the culture and history of broader Xuan-Da, which encompasses it. I came to know about the images over the course of a survey of rural antiquities undertaken in this and adjacent counties over the winter and spring of 2013–14, and shored up by multiple subsequent visits over the last few years, particularly 2018 spent in the field.[6] The stages and temples on which these murals are painted are scattered across around fifty small villages and perhaps a thousand square kilometres.[7] Few of these villages today have even nominal populations of over a few thousand people, and pre-Communist numbers would have been smaller. Almost none of these images have been reproduced, and, to my knowledge, only one brief analytical treatment of them exists — a chapter in a PhD thesis by Liu Wenjiong 劉文炯.[8] In the following pages, I will attempt to describe the rural cultural world in which these images appeared, with particular emphasis on the historical development of religious, theatrical, social, and artistic space.
These stage and temple drawings must be understood in the context of the history and culture of the Yu county village. Both the temples and the stages were religious structures and they formed one component in physically and symbolically complex arrangements of village space. The villages of Yu county were established in their present form in the first half of the sixteenth century, when they were reconstructed and fortified en masse in response to continual Mongol raiding from the steppe.[9] The religious system of these villages — that is, the temples that existed within, atop, and around the village walls — seems to have appeared at the same time and as part of the same process.[10] After the Longqing Peace from 1567, Yu county, in its isolated mountain basin, did not again see significant warfare essentially until the Japanese invasion in the 1930s. As evidenced by the stele record, from around 1700 the region entered a long period of relative prosperity, which saw the maintenance of old structures and forms of worship, with the much slower addition of new temples within the existing system. This prosperity peaked in the long Qianlong reign (1735–96) but continued through the nineteenth century (troubled elsewhere in China, but peaceful in Yu county) and even into the first decades of the twentieth.[11] This period of fading but still-evident prosperity and peace in the late nineteenth and early twentieth centuries is the one that produced the perspectival drawings that are the topic of this paper.
Temples in Yu county are usually one to three bays (jian 間) in width, often built upon a low stone and earth plinth (Fig. 5). Other temples, particularly those built into the north wall of the fort, are tall, stepped structures of multiple levels. (Fig. 6) Some of the larger and more important temples have a walled yard (yuanzi 院子) extending in front of them, although the majority of small village shrines do not. The main gods are universally depicted on the rear wall opposite the door. These depictions were either statues set on an altar, paintings on the wall, or in some cases both; almost no pre-Revolution statues now survive in Yu county. For some iconographies, the altar extends around the room on three sides, so that subsidiary statues of the main god’s retinue flank the worshipper as he enters. Offerings and incense are set on the altar or a table in front of it; gongs, bells, incense-burners, votive dhvaja pillars, and steles are placed around the shrine or in the courtyard as appropriate.[12]
Rural Xuan-Da religion was, broadly speaking, a system in which compositional or spatial relationships carried defining symbolic meaning. This was true both of the arrangement of villages and the compositions of village art. The village itself marked the power of particular ritual buildings via their command over lines of sight. From the sixteenth century on, the new fortress villages were usually dominated by a tall tower at the centre of the northern wall that gazed over an axial road stretching down to the main gate (Fig. 7). A male, martial god — frequently the Perfected Warrior (Zhen wu 真武) or the Jade Emperor (Yu huang 玉皇), but sometimes other gods — would be enthroned in a shrine on this panoptical tower, while a temple to a female deity — usually Avalokiteśvara (Guan yin 觀音) or the Daoist ‘Goddesses’ (Niangniang 娘娘) — would sit at the southern end of the axis. Although the specific buildings and their placement could vary according to the situation of individual villages, on the whole this axial relationship between temples was the main structuring principle of the physical space of Yu county villages from the sixteenth century onward.[13]
The Xuan-Da religious system was also one that paid attention to the placement and motion of gods through both three-dimensional and painted space. Particular gods are depicted in temple murals as being positioned in particular ways. The Perfected Warrior and the Jade Emperor are generally seen seated in state on the north wall of a temple, gazing down the axial line of the village, with rows of warriors lining either side of the room so as to emphasise the spatial metaphor of an imperial court. (Figs 8a & b) Other gods are depicted as moving through space to fulfill some important function of their cult. The Dragon Kings (Long wang 龍王), all-important as weather gods in semi-arid north China, are invariably painted processing in a circular rain-dispensing expedition around the space of the shrine room, wrathful and thunderous on the left-hand wall, pacific and shedding rainbows as they return after the storm on the right (Figs 9a & b). The God of the Five Ways (Wu dao shen 五道神) and the Horse King (Ma wang 馬王) are depicted in similar processions, in this case pursuing demons and, returning on the right-hand wall, dragging them back in chains.[14] As a broad set of principles relevant to these later perspectival paintings, Yu county temples were positioned with spatial intent, the artistic depictions in and around those temples expressed aspects of that intent, and the gods who inhabited those temples and those depictions were thought to be located in particular places and to move through space in particular and intentional ways.
Opera stages in Yu county are small, freestanding structures, constructed of timber and brick, usually three narrow bays in width. (Fig. 10) They are generally raised up on a stone plinth, typically higher than those of the temples, between one and two metres in height. With a few exceptions, the stages have three walls, with the fourth side open to the audience. Usually, there is no back door onto the stage and the only entrance is from the front, although some larger stages do have a small opening at the rear through which performers could slip in and out. The stages were usually divided into a proscenium and a backstage area by an elaborate scaenae frons ‘screen wall’ (geshan qiang 隔扇牆) of painted wooden panels. (Figs 11a & b) This scaenae frons, however, was not a solid boundary; it invariably had two doors leading in and out on either side, and elaborate lattice windows towards the centre. These openings could be used for dramatic purposes or closed off with curtains to create a true backstage area, as the occasion demanded (Figs 12a & b).[15]
Some of these villages had opera stages within or around the walls during the period of their creation in the Ming dynasty, but it doesn’t seem that all of them did.[16] From the eighteenth century on, there are a number of steles that record the refurbishment of old stages and the establishment of new ones in villages previously without them.[17] As we will see below, the apparent difficulty with which many of the stages are fitted into the packed space of these villages suggests that many of them were late additions into the system. In any case, by the start of the twentieth century, nearly every village in the county had at least one stage. Of these, perhaps two hundred now remain, in various states of repair and collapse.[18] The perspectival drawings that are the subject of this paper are almost never accompanied by clear dates. A few examples can be fixed to the late eighteenth or early nineteenth centuries. However, from the style and the topics depicted therein it is clear that bulk of them cannot have been painted before the year 1800, and that the majority were produced in the late nineteenth century or the first half of the twentieth.[19] The Guangxu reign (1875–1908) particularly may have represented something of a golden age for Yu county opera generally: dated graffiti on Yu county stages seems to have peaked in the last decades of the nineteenth century (Fig. 13).[20]
These stages were votive structures and all of them were located across from temples.[21] The ritual use of opera in rural China is well documented, and Yu county is not an exception.[22] Although the content of the plays was usually secular, in the sense that the operas were, for the most part, not ‘mystery plays’ specifically depicting acts of the gods,[23] the operas were performed ritually as offerings during rainmaking rituals,[24] temple fairs,[25] and autumn festivals to repay and entertain the gods after the harvest.[26] The ritual uses of opera are actually frequently depicted in temple murals themselves. The lower right-hand wall (according to the Chinese reckoning)[27] of the all-important Dragon King temples almost always contained depictions of rituals performed in autumn to thank the gods at the end of the successful harvest. Here, we find images depicting processions of Daoist priests accompanied by shawm bands (Fig. 14),[28] masked saishe 賽社 or shehuo 社火 processions (Fig. 15), and, not infrequently, images of opera being offered on stages facing the temple (Figs 16 & 17a & b). We will return to this type of mural later, but for now it is enough to note that for such ritual performance to be effective, a direct line of sight between the temple and the stage was necessary. The god had to physically see the opera.[29]
Of course, these walled villages were tightly packed spaces, and free space for an opera stage across from a suitable temple was not always easy to come by. Villagers went to great lengths and exercised considerable architectural creativity to ensure the lines of sight between temple and stage. One stele from 1750 records a village argument over the creation of such a stage, since the only proper spot for it — across from a temple to Lord Guan (Guan gong miao 關公廟) — was already occupied by a temple to Avalokiteśvara (Guan yin miao 觀音廟).[30] The assembled villagers (zhong 眾) were stymied until ‘an elderly man, racking his brains and with emotion’ (baisou dong nian kairan 白叟動念慨然) came up with the solution of shifting the Avalokiteśvara temple to the top of the gatehouse facing north and putting the stage in its previous place, thus satisfying the geomantic requirements of both structures. In another village a geomantic temple and a stage were folded into each other to form one awkwardly conjoined ‘franken’-building, in which the back wall of the stage is the front wall of the temple.[31] In at least three other villages the stage has been cleaved in half on either side of the axial road so that it doesn’t impede the central geomantic axis. When an opera was to be performed, the gap over the road could be boarded over and foot traffic diverted around or under it.[32] The point here is that creating and maintaining these lines of sight between votive structures was not necessarily easy, and villagers put a great deal of effort, resources, and ingenuity into doing so.
These processes, proceeding differently in each individual village, resulted often in the creation of complex architectural assemblages of many buildings and open spaces between them, linked by criss-crossing lines of sight. With a temple on one side, a stage on the other, enough open ground between them to fit the assembled villagers in, perhaps the fortress gate on the third side and a few other shrines and shops set adjacent, these lines between temple and stage often defined what are effectively village squares. The areas created in this way were and are the main centres of public life in the villages, comparable to the fora and agorae in the classical West. Today they are often busy and multipurpose spaces: farmers thresh grain, shopkeepers chat with their customers, itinerant peddlers lay out their wares on carts and blankets, children chase and play, old men sit sunning themselves on the temple steps or opera plinths, smoking, and exchanging gossip.[33] Between the stage and the temple, Chinese villagers sang their opera, worshipped their gods, conducted their buying and selling, and held their public meetings. We should not overlook these village squares as among the principle spaces in which Yu county villagers eventually met and took part in the modern world.
It is a truism of Chinese village opera that it instructs,[34] and that it forms one of the main ways in which illiterate villagers take part in larger historical and national narratives.[35] Although we know very little about it, we do know that these stages were sometimes used for news-sheet re-enactments of current events. Arthur Henderson Smith, writing in 1889 of China generally, remarked that anti-foreign plays — for instance, depicting the Tianjin Massacre of 1870 — were frequently performed in villages, and that local events such as disputes between counties would be rapidly written into dramas and performed to acclaim in those areas.[36] These stages and the squares in front of them could function as fora to relate and discuss local and national events. And of course opera troupes circulated, often very widely, moving frequently between the cities and the villages.[37] Temple fairs were commercial fairs, and commercial fairs also had opera, held on stages in front of temples.[38] It’s quite probable that for a villager in nineteenth or twentieth century China, even into the 1980s and ’90s, the gatherings held in these spaces and performances represented one of the main venues for contact with the wider world.
Of course, to restate the obvious, these opera stages were also used for performing operas. In a largely illiterate society, these stages allowed Chinese villagers to interact both with local artistic forms and with the broad stream of Chinese national literature. Wang Zhijun and Tian Yongxiang in Difang juzhong gaishuo enumerate a broad typology of performances on Yu county stages as follows: Zither Tunes (Xianzi qiang 弦子腔), Lolo Tunes (Luoluo qiang 羅羅腔), Sai Operas (Sai xi 賽戲), Great Operas (Da xi 大戲), Seedling Songs (Yang ge 秧歌),[39] Operas of the Way and Emotion (Dao qing 道情), Playing with Children (Shua hai’er 耍孩兒), Lantern-Shadow Plays (Deng ying xi 燈影戲), Plays on Stilts (Gaoqiao xi 高蹺戲), Shanxi Operas (Jin xi 晉戲), and Assorted Tales of the Pear Garden (Liyuan zashuo 梨園雜說).[40] Although the differences between each of these genres go far beyond the limits of this paper, the reader will appreciate the immense variety and sophistication of the rural performing culture in these regions. From the opera titles found throughout Wang and Tian’s book, it is clear that the rural Yu county repertoire included material in both national and local circulation: plots were drawn from pan-Chinese story cycles like that of the Three Kingdoms (San guo 三國), plus a great variety of plays popular in Peking or Shanxi opera, all freely merging with locally produced performance and song. These rural plays cover the full range of human experience, including love songs, political dramas, histories, comedies both secular and mythological, performances with votive or exorcistic content, and from the twentieth century onward, ‘new’ plays both Republican and Communist.
If this was the religious, spatial, and performing culture that produced these structures and the images on them, then what of the culture of fiction and theatricality within these villages, surely relevant to what seem to be essentially fantasy landscapes on opera stages? We must be wary of projecting onto rural Chinese life a Protestant dividing line between fiction and religion. Similar, perhaps, to classical Greece and Rome, Chinese society generally was one in which religion could be freely used as the raw material for fictional invention, as in Journey to the West (Xi you ji 西遊記) or even Dream of the Red Mansions (Hong lou meng 紅樓夢),[41] and in which fictional invention could, without any apparent cognitive friction, again form the basis for religious worship, as in Romance of the Three Kingdoms (San guo yanyi 三國演義) and Romance of the Investiture of the Gods (Feng shen yanyi 封神演義).[42] In Yu county villages, such fiction was consumed on the opera stage as part of elaborate, and often elaborately theatrical, religious rituals.[43] The gods themselves might populate the stage at times,[44] while the gods’ temples themselves often held elaborate panelled hagiographies apparently drawn from popular literature.[45] Religion, opera, and fiction were interpenetrating realms in these villages, and we divide them at our own intellectual peril.
Nor can we assume that villagers were naïve about these relationships, or that the theatricality of the village stage was in any way less sophisticated than that of the urban or literati-produced opera. One fascinating window into rural ideas of religion, gaze, fiction, and theatricality are the ‘opposing couplets’ (duilian 對聯) that were written or inscribed on the two central pillars of the stage.[46] Theatricality is front and centre here. We read: ‘We make real an empty meal — and discuss all the examples of history / we make True (zhen) a False (jia) image — and act out all the strange affairs of Now and Then’ 虛飲作實談論歷代典故 / 假像變真扮演古今奇事. A repeated theme in many of these couplets, and one which will become very important later in this paper, is the way in which the physically small and bounded space of the stage can form an entrance into much larger historical and social realms. ‘People call it a house (fang), but I call it a mansion (lou) / lords, dukes, kings, and ancestors — they’re all inside of it’人家叫房我叫樓 / 公侯王爺在裏頭. ‘The surface-area isn’t big, but it has families, it has nations, it has all beneath heaven / there’s not many people, but they are sons, they are fathers, they are gentlemen ministers’ 地面不大有家有國有天下 / 人數無多為子為父為君臣. Another longer poem refers to mirage-visions, ‘the cities of the sea, towers upon giant oysters’, 海市蜃樓 and ‘all the realms of illusion, from first to last’ 乾坤幻境. Thus the epigraphic culture of these stages describes them as a place of spatial recession into fictional space — a place at the boundary of true and false, between our reality and fantastic other worlds.[47]
To summarise here for the purposes of the remainder of the paper, the temples and spatial system of Yu county villages appeared in recognisable form in the early sixteenth century. The opera culture that we see there today achieved maturity later, in the eighteenth and nineteenth centuries, including a high-point at least of stage graffiti around the Guangxu reign (1875–1908). This culture was sophisticated and intentional about the creation of physical and symbolic space. Villages were planned settlements, and temples were constructed within and around this space with clear symbolic intent. Stages were located across from these temples so that the gods could watch the opera. The physical opposition between these two structures formed village squares that were the centre of most types of village events, including religious, operatic, commercial, and political. Important for understanding the perspectival murals, these spaces and the events held in them were the main gateways via which Chinese villagers interacted with the broader world. The epigraphic culture found around these structures emphasised the spatial liminality of both the stage and the temple as trespasses between religion, fiction, and worldly society. The operas, songs, and rituals performed on these stages belonged to a rich and sophisticated circulation of local and trans-local traditions, all of which freely mixed religion and fiction.
2) The Mansion and the Pagoda: Western Exotica in Yu County Village
The history of Yu county village art prior to the turn of the eighteenth century is difficult to trace. The Yu County Museum possesses many medieval-period objects of common types that have been the subject of studies in English. These include Northern Wei-period carved Buddhas set within arched niches, nested śarīra (sheli 舍利) reliquaries from the base of Liao-period stūpas, and Liao-period tomb murals depicting gateways flanked by musicians. From this, we should conclude that Yu county was well within the mainstream of medieval-period north-Chinese visual culture, and that later mural traditions should represent an outgrowth of these earlier genres. Importantly, recent studies on these types of medieval objects have emphasised the connection between their holiness as religious objects and various types of spatial recession, as well as the connection between theatrical space and entrances to other worlds, notably that of the afterlife.[48] Given that the fortified villages and temples of rural Yu county were created in the sixteenth century, it seems reasonable to hypothesise that the main elements of village mural traditions took their present form in that period as well. This, however, must remain speculative until more dated examples come to light.[49] The outstanding exceptions to this lack of clearly Ming-period murals are the immense Water-and-Land (Shui lu 水陸) images of Previous-Fort Monastery (Gu cheng si 故城寺), dated from ceiling plaques to 1507.[50]
From around 1700, however, clearly dated and well-preserved murals become comparatively numerous. The situation at this earliest extant level in Yu county seems to have been one of relative conformity in composition and subject combined with great stylistic heterogeneity. All of these village temples have, on the central wall, images of the enthroned deities seated in state (Fig. 18). While three Dragon King side-wall procession murals from 1698 (Fig. 19), 1709 (See Figs 9a & b), and 1730 (Fig. 20) seem almost unrelated in style, the roster of different figures and their relative positions in the procession are all the same. Importantly, for later developments, the inner (northern) side of two out of three compositions depicts a portico-fronted building extending back behind the frame. This is the Crystal Palace (Shui jing gong 水晶宫) from which the expedition of the Dragon Kings departs on the left-hand wall and to which it returns on the right. A similar procession of the Goddesses from 1724 departs farther afield stylistically from anything else in Yu county, to the extent that we may suspect some Tibeto-Mongol influence (Figs 21a & b). Although the portico-fronted building at the interior of the image and the small figures beneath are here missing, the main composition of the procession of the gods is similar to that of the Dragon King images and to other undated Goddess processions from around Xuan-Da.[51] Clearly dated eighteenth-century examples of panelled story narratives and martial images of divine generals indicate that these genres and topics were well established in Yu county by this point as well. Thus we may describe Yu county village mural art before the clear appearance of Western perspective as capable of great beauty and stylistic variety, but possessing a limited range of subjects and compositions.
By the year 1700, however, European perspectival techniques and pictorial themes had already begun to have an influence on the broader sphere of Chinese art. Merchants and missionaries first brought European images to the coastal port cities towards the end of the Ming dynasty. By the mid-seventeenth century, the southern cities of Yangzhou and Suzhou were centres of large-scale production of woodblock ‘foreign images’ (yang hua 洋畫) depicting exotic Western scenes, as well as painted images of all types that adopted various aspects of Western perspective, shading, or composition.[52] The eighteenth century saw the integration of European artists into Qing imperial court production in northern China, especially after the arrival of the celebrated Jesuit painter Giuseppe Castiglione (Lang Shining 郎世寧, 1688–1766) in Beijing in 1715. By the twentieth century, images of Western cities and scenes both real and imagined could be viewed all across rural and urban China via woodblock ‘yearly pictures’ (nian hua 年畫) and travelling zograscope displays. Woodblock perspectival drawings also made frequent appearances in reformist and modernising literature and illustrated newsprint.[53] Although there is very little literature describing this process, Western pictorial techniques certainly had an influence on religious mural painting as well. Perspectival murals depicting imperial processions were commissioned at Mount Tai in Shandong as early as 1677,[54] and by the end of the nineteenth century, perspectival techniques were being used in murals even on the Tibetan border in modern Qinghai province.[55]
It is difficult to date precisely when perspectival painting first appeared in Yu county villages. The vast majority of the extant perspectival murals in Yu county clearly date from the nineteenth or early twentieth centuries, judging by both style and the adjacent grafitti. The majority of these are located on the walls flanking opera stages. We have seen that these stages were a centre of the public life of the village, and that opera troupes were one of the main agents of cultural circulation within these villages. Opera stages were a site naturally associated with spectacle, gaze, modernity, and encounters with the outside world; it is not surprising then that Western themes and pictorial techniques of receding space were painted there. We have also seen that the period of the appearance of opera stages in rural Yu county (beginning in the sixteenth century, but gradually achieving ubiquity only over the course of the eighteenth and nineteenth centuries) roughly matches the period of Western art’s transfusion into China. Thus it may not be necessary to search for developed opera stage mural traditions prior to Western influence.[56] It seems likely to me, although unproven, that perspectival murals first appeared on stages, which were, at the time, a relatively new site for mural painting and one not necessarily limited by religious prescription. Only afterwards did these paintings begin to spread into the temples.[57]
Liu Wenjiong has calculated that perspectival architectural paintings represent somewhat over 40 per cent of extant opera stage murals in Yu county, with the rest being images of folding screens, with or without figures peeking around the edges (Fig. 22).[58] These paintings appear on the two flanking walls of the stage, exterior to the scaenae frons and perpendicular to the gaze of the spectators and the deity in the opposing temple. It is important to emphasise that these images were not backdrops in a strict sense; although they were visible to spectators sitting at angles to the stage, the only people who would have seen them head-on during the performance were the actors themselves. The use of perspectival technique is loose. It is apparent that this technique was new to the artists, and they used it with varying success. Some of the drawings succeed at simple point perspective, with long lines of buildings receding down an axial street. In other compositions, attempts at Western perspective are applied more unevenly to the slant of the railings of the structures, which become a crazy zig-zag of impossibly layered stories, with recession in height indicated by the reduced size of the upper figures. In still more compositions, the artists do not even attempt formal point perspective, instead achieving the effect of receding space by foregrounding some buildings and placing others on a distant horizon line behind it (Figs 23a & b). Although many permutations are possible, in the main these images contain two main structures or compositional elements, which I refer to here as the pagoda (ting 亭) and the mansion (lou 樓).[59] Each of these seems to derive from a different source, and I will treat them separately below.
The ‘pagoda’ that usually occupies the centre or outer side of these compositions appears to derive, at least in some cases, from an anonymous mid-Qing novel called The Full Tale of the Green Peony (Lü mudan quan zhuan 綠牡丹全傳).[60] The story is also known as The Full Tale of the Pagoda of Gazing in the Four Directions (Si wang ting quan zhuan 四望亭全傳). The novel was published in 1800. From that year until the end of the dynasty (1911–12), it was reprinted 26 times, and, over the course of the nineteenth and twentieth centuries, a series of operas and later films were based on it.[61] The story concerns two heroes, — young circus-performer named Jade-Lotus (Hua Bilian 華碧蓮) and her love interest, a somewhat bumbling scholar-aspirant named Luo Hongxun 駱宏勛. Luo Hongxun, Jade-Lotus, her father, her spry old grandmother, and a band of swashbuckling friends all take up arms against the evil Empress Wu Zetian. At the end of the story, aided by the famous Judge Dee (Di Renjie 狄仁杰), the band of heroes defeats Wu Zetian and restores the rightful heir to the throne of China.
The relevant scene for our purposes occurs roughly a third of the way through the novel, over the nineteenth and twentieth chapters. By this point, the two heroes have professed their love to each other but have since become separated. Jade-Lotus and her father, Hua Zhenfang 花振芳, are driving a herd of horses in through the gate of Yangzhou city (Yangzhou fu 扬州府), where they arrive at a tower called the Pagoda of Gazing in the Four Directions (Si wang ting 四望亭). Here they find an old friend of Hua Zhenfang’s named Yu Qian 余謙, who has vowed in front of a crowd to catch an escaped monkey, which is clambering about on the roof. The monkey, by this point, has ascended the Pagoda of Gazing in the Four Directions, and Yu Qian is labouring around after it, steadily losing face. Hua Zhenfang calls Yu Qian down, and instead sends Jade-Lotus up to nab the monkey. In a series of acrobatic leaps she makes it up onto the roof of the pagoda. Granny Hua (Hua Nainai 花奶奶) leaps up after her. Jade-Lotus pursues the monkey up to the highest gable. Thereupon comes the moment of high drama:
Jade-Lotus slides herself stealthily towards the monkey. She reaches out an arm to grab him. The monkey sees that Jade-Lotus is blocking the escape route to the right: he’s got no empty space to escape through. The animal panics, and uses all his might to leap, hoping that he can jump right over Jade-Lotus’s head. For many years, though, nobody has maintained the Pagoda of Gazing in the Four Directions. The wood is rotten, the mortar and bricks split asunder — Jade-Lotus together with the monkey go plummeting down. On the ground the people exclaim: ‘How terrible! Someone’s fallen down!’
As Jade-Lotus falls, Hua Zhenfang, Yu Qian, and the Ba brothers are all dismayed and without any recourse. Jade-Lotus has no way of saving her own life. The only hope comes from one young man beyond the fourth or fifth ring of spectators, who shouts: ‘You still haven’t moved to save her — what are you waiting for!’
With a single leap he’s there, and catches Jade-Lotus with both hands, and holds her in his embrace, sitting down into the dust. Everyone shouts: ‘What an extraordinary hero! Without him, she’d have been crushed to meat-mush!’ Hua Zhenfang and a whole crowd all run over and with one glance they see — the person who’d saved Jade-Lotus was none other than Luo Hongxun!62
Jade-Lotus lays unconscious, pressed against Luo Hongxun’s chest. For a brief moment, the narrative shifts into the first person as she regains consciousness: finding herself laying sweat-covered in the arms of her long-separated beloved, she decides to keep her eyes closed for a while longer. Only when her father has tactfully removed her onto a charpai does she admit to being awake.[63] The various characters then turn to acrimony over the prize money.
This, then, is identifiably the scene represented on the opera walls in Yu county. At least five of these murals have buildings labelled ‘Pagoda of Gazing in the Four Directions’ and have visible pictures of Jade-Lotus and the monkey perched dramatically on the gables (Figs 24a & b). In two more compositions, the building is not labelled but nonetheless the telltale figures of a girl and a monkey are visible. In two of them you can also see an older woman balanced on a lower roof, and this must be Granny Hua. In some cases, the whole cast of characters is visible standing around the base of the tower, looking up and gesticulating at Jade-Lotus and the monkey, although it’s difficult to tell who is who (Fig. 25). Given how heavily damaged most of the murals are, it seems certain that there were once many more figures visible in all of the scenes. Many more villages just have depictions of tall towers or pagodas without any visible or extant labels or figures.
What did this image represent to Yu county villagers? The scene itself is engaging, kinetic, and affecting. It features acrobatic stunts, a rooftop chase, a fall from a high place, and a last-minute rescue. It also contains the unexpected reunion of lovers, a moment of deeply felt romantic emotion (qing 情), and the first-person expression of female erotic desire.[64] It’s fair to suppose that the image of a sensuous woman knight-errant (nü xia 女俠) leaping about on an exotic southern pagoda would represent an engaging interruption into the social and architectural world of the average nineteenth- or early twentieth-century Yu county peasant. One can also imagine a whole variety of positive explanations for this scene’s prevalence, although all of them are very speculative. It could have represented a rebus,[65] or it had a relationship with village ritual.[66] Liu Wenjiong has suggested a link to the village stage as a site of sexual and gender exploration, and this may be so.[67] In early twentieth century Shanghai, the scene was performed with elaborate rope-work special effects (jiguan bujing 機關佈景); perhaps Yu county people somehow witnessed this impressive modern spectacle and attempted to depict it.[68] The name Pagoda of Gazing in the Four Directions references the theme of architecturally established lines of sight, which we have seen is an important part of these stages’ function, and to which we will return again.
But none of these arguments quite satisfies. Among other things, the operas based on Green Peony do not seem to have been particularly popular in Yu county, at least within living memory. Wang Zhijun and Tian Yongxiang’s volume on Yu county opera mentions hundreds of titles performed within the county in the pre-Communist era.[69] As far as I can find, none of the various opera titles derived from Green Peony appear there.[70] So these plays may have once been performed on Yu county stages, but there’s no written record of it, and nobody around now seems to remember it. Perhaps no explanation is needed. One stage-painter enjoyed the novel and painted the scene; from there it became popular and emulated within the community of painters and audiences around Yu county. It seems safe, however, to say that Yu county villagers painted the world they wanted to see. And in village after village, that world was a realm of swashbuckling women warriors, free-booting and foul-mouthed proletarian heroes, chases, humour, excitement, architectural exotica, and an easy-going eroticism, love, and adventure.
The second building in the opera stage murals, the ‘Mansions’, always appears on the inner side of the composition, where it abuts the edge of the scaenae frons and the doors that pass through it. The images depict oblong, multistorey buildings with rows of windows facing the street, galleries along the upper stories, Italianate flourishes over the doors, long colonnades, tall minaret-like towers, and even domes — something truly exotic in eastern China. These structures are identified in several of the cartouches. One image-caption has them as The Mansions of the Western Seas (Xi yang lou 西洋樓), (see Fig 25). Another cartouche reads Yi da gong 意大宮 — ostensibly, The Intentions-Great Palace. This name is as awkward in Chinese as it is in English. Given the foreign form of the buildings, it seems more likely that the full name should be The Italian Palace (Yidali gong 意大利宮). The Chinese painter, finding Yidali ‘Italy’ a mouthful, simply lopped off the last syllable and gave the palace a more acceptably Chinese two-character name (Fig. 26). Another such mansion, in this case labelled The Mansion [sic] of Gazing in the Four Directions (Si wang lou 四望樓) has a row of nonsensical ‘Western’ alphabetic characters written along the top, although it is unclear to me even whether they are drawn from the Roman or Cyrillic scripts (Figs 27a & b).[71]
The mansions of the Western Seas (Xi yang lou 西洋樓) was not only a generic name for Western-style buildings; it originally referred to a specific set of structures located in the imperial Garden of Perfect Radiance (Yuan ming yuan 圓明園) outside of Beijing. The buildings and the gardens, maze, and fountains around them were constructed by Jesuit artisans, headed by Guiseppe Castiglione, for the Qianlong Emperor over the course of the late eighteenth century. The mansions were then destroyed by foreign troops during the second Opium War in 1860. At the time these buildings were constructed, they were strongly associated with theatricality and perspectival modes of viewing.[72] Yu county people would certainly have been aware of these structures, since they sat along the well-travelled road to the capital. The multistorey Western buildings were adjacent to the northern wall of the Garden and would have been visible over that wall from the outside. After the violent destruction of the Garden, travellers had free access to the ruins, which still existed in substantially complete form well into the twentieth century. Yu county people could hardly have failed to notice, and taken imaginative interest in, the immense ivory-white ruins of the foreign palaces that rose out the fields along the carriage road to the capital.[73] One such depiction, in the Halls of the God of Wealth (Cai shen dian 財神殿) and of the God of Literature (Wenchang 文昌) in the Six Gods’ Temple of Li [Family] Fort (Li buzi liu shen miao cai shen dian 李堡子六神廟) may actually be copied directly from the mansions or from one of the various sets of lithographs that were produced depicting them (Figs 28a & b).[74] The moniker ‘The Italian Palace’ on another of these depictions strengthens the general connection to Castiglione’s buildings in Beijing.
However, the majority of these images are not literal depictions of the Jesuit-built mansions. A more proximate source for all of these Yu county depictions may be found in perspectival ‘yearly pictures’ and peep-box show or zograscope images that circulated widely through eighteenth- to twentieth-century Chinese society. The 23-volume Collected Chinese Woodblock Yearly Pictures (Zhongguo muban nianhua jicheng 中國木板年畫集成)[75] contains hundreds of perspectival prints from the late Qing and Republican periods. Dozens if not hundreds of these showcase exotic Western-style buildings of exactly the type found in the Yu county images, including long multi-storey façades, domes, and high, almost futuristic towers. We may point to specific depictions that may have been the source of particular depictions of the Pagoda of Gazing in the Four Directions at Yangzhou, which display the same scene from Green Peony.[76] (It is relevant to point out here that the Pagoda of Gazing in the Four Directions is a real structure that still stands in Yangzhou city, and that Yangzhou was a major source of such prints.) Any of these images could have served as the model or inspiration for the Yu county opera-stage drawings. Many fantastical architectural scenes of this sort circulated as peep-box views, known in Chinese as ‘scenes of the Western seas’ (xi yang jing 西洋景) or the closely homophonous ‘lenses of the Western seas’ (xi yang jing 西洋鏡).[77] The name and format of these devices underscores the link between perspectival drawings, Western architecture, technological innovations in the creation of spectacle, and the emphasis on the linear, modern gaze.
Thus it seems closer to the truth to say that these stage drawings referenced a broader representation, popular across China from the eighteenth century until the Communist takeover, of a sort of architectural and visual ‘occidenterie’:[78] an exotic new type of building, and an exotic new way of seeing those buildings. Although drawn from the broader stream of Chinese visual culture, this representation had specific connotations within Yu county. Perspectival architectural drawings were in no way unnatural to a society that used architecturally established lines of sight as one of its main physical structuring principles. We have also seen how the open communal squares thus established between stage and temple were among the main spaces in which villagers interacted with the broader world, and how the epigraphic culture associated with these structures stressed the ability of the physically bounded stage to include ‘all beneath heaven’ (tianxia 天下).[79] A geographic and architectural imagination that extended to Beijing, Yangzhou, and even Europe was appropriate to these spaces. It was via these squares, stages, murals, opera performances, and commercial fairs that Yu county people first met the oncoming modern world, represented it to each other, and ultimately integrated it into their own cosmology and sense of self. These images of the Pagoda of Gazing in the Four Directions and the Mansions of the Western Seas are one artifact of that process.
4) The Portico of the Gods’ Realms
James Cahill has commented on the introduction of Western art to China that, ‘In China as in France, [artistic] appropriations were liberating rather than confining, giving artists the courage to break out of old habits that had become stultifying’.[80] This was certainly true in Yu county. The last 150 years of Yu county mural art, from ~1800 to 1950, were a period of extraordinary innovation. As these Westernising themes and perspectival techniques moved from the stage into the temples, traditional compositions were promiscuously remixed and reinvented. Rather than representing the breakdown of traditional mural-painting culture, I argue that this period represents its lost renaissance. Yu county artists combined Western techniques of depicting receding depth with the traditional spatial logic of the stage and temple to create what are undoubtedly the most complex and innovative compositions in the history of this art. To this end, I will first return to the earliest stratum of extant murals in the early eighteenth century. I argue that their efficacy as religious images derived from a combination of the recursion of the temple room with the realm of the gods depicted, and the evocation of a hidden space behind the altar from which the gods’ power emanated. I will then show how this logic was reworked with the new perspectival techniques, allowing for unique compositions that stretch across multiple buildings and surfaces and establish complex and efficacious connections between our world and the holy, fictive, or otherwise ‘exotic’ realms beyond.
Perhaps the most consistent single element of all of these depictions on temples and stages, both early and late, is a large portico (menlang 門廊) opening on the inner side of the composition (Fig. 29). In fact, we could reasonably refer to the entire genre of images examined in this paper as ‘portico-pictures’, since the great roofed portico on the inner side of each composition is their most consistent and distinctive feature. This structure is a flat, raised plinth, sometimes with railings but always bounded by pillars supporting the roof. Especially in the later images, the ‘sky-flower panels’ (tianhua ban 天華板) on the portico ceiling became an object of obsession for painters, who distorted space to bizarre degrees to display the perspectival recession of this grid (see Figs 1, 2, 26, 28b, et cetera). In the later opera stage images, this portico fronts the structure that I have earlier referred to as the Mansion and provides an entrance to its interior. In the temple images both early and late, this structure represents the façade and gateway into the gods’ palace.[81]
It is worth pointing out here that the physical porticoes of real Xuan-Da temples remain symbolically important places today. During a visit in the summer of 2017, the Daoist cleric resident in the Goddesses Temple of the Cock’s-Crow Postal Station (Ji ming yi 雞鳴驛) remarked to me, unprompted, that the gods would sit on the small portico space jutting out of the front of the temple in order to watch the opera being performed on the stage opposite. This raised and roofed entranceway, apparently empty, was thus one of the most important symbolic spaces of the temple complex, strongly associated with divine, spectactorial gaze from the temple to the stage, and with the actual historical performance of opera at temple fairs (Fig. 30). It is probable that in many temples the statue of the god would be physically carried out and set on this space during rituals and performances; I have seen this done in other parts of north China, although not in Xuan-Da proper. These porticoes are also, of course, the gateway or anteroom through which the devotee must pass in order to reach the inner sanctum.
The above examples point to two important interpretive aspects of these depictions, apparently unrelated but in fact intertwined. The first is that these porticoes do, after all, represent gateways, or epi-structures around gateways, that lead into sacred or otherwise alteric space. The second is that there is, from the earliest examples until the latest, a persistent conflation and even recursion between the space of the depiction on the wall and the three-dimensional space in which it was set, either temple, stage, or fortress. We will examine these propositions in tandem so that their relationship becomes clear.
To begin with, the earthly temple to the god was frequently conflated with the god’s celestial (or submarine) palace. The name of the palace in which the god is thought to reside is often used as a metonym for the physical temple building, both in the stele texts and on signs over the gateways of the actual temple buildings. Thus a Dragon King temple might have a gate inscribed with the words The Crystal Palace,[82] a Temple to the Perfected Warrior might be called The Palace of the Northern Dipper (Bei dou gong 北斗宮),[83] or a temple to the Goddesses referred to as The Travelling Palace of Mount Tai (Tai shan xing gong 泰山行宮),[84] etc. We have already noted that the interior of the temple is often creatively furnished to indicate the grotto or palace in which the god is thought to hold court. In this way, the temple space mimics the divine space of the gods. In the case of the portico images, however, a slightly more complex logic is at work. The flanking-wall porticoes are invariably painted on the interior side of the wall, with the building extending back outside the frame of the image. Thus in these earliest extant images, the position of the portico suggests the exterior façade and gateway of a mysterious interior or ‘backstage’ through these painted gates. This holy interior, the realm of the gods, is located at or extends away behind the altar where the statues sit.
As Dragon King temples were the most common type of large temple in Yu county, the spatial logic of these depictions is most developed in those compositions, and many intact examples exist. In these temples, the inner side of both lateral walls almost always holds an image of both the Crystal Palace and, beneath it, a smaller recursive image of the temple building itself in which the image and the shrine is held (Fig. 31). The rain-giving procession of the dragons around the space of the shrine room (that is out and back from their palace, and around the human world) is almost always echoed by miniature, sympathetic images of mortals on the ground beneath, fleeing from dragon-sparked lightning, then plowing, planting, harvesting, and, finally, forming themselves into a votive procession with a shawm band or shehuo mummers’ parade that arrives to give thanks at the gates of the recursive temple. As we have seen above, several extant images even show opera performances performed on stages facing the temple gate (see Figs 14, 15, 16 and 17). The visual recursion between the small temple below and the great palace above, and between the painted image and the physical temple space, is clear — by facing north towards the altar in the temple, the devotee is able to access the Crystal Palace of the Dragon Kings. The space of the workaday human world is coterminous with the circuit of the Dragon Kings’ dominion, and the physical temple is the anteroom to their divine realm.
That the rear walls of Chinese temples were understood to recess into holy space is spectacularly confirmed by two temples in areas immediately adjacent to Yu county that have, on the central/rear wall, chiaroscuro trompe-l’œil images of hanging prayer beads that cast painted shadows on the wall — something genuinely unprecedented in traditional Chinese art. One of these images was painted in 1783, while the other is undated but appears to be in early nineteenth-century style, indicating the early spread of Western painting techniques into rural Xuan-Da, and their use from the very start to compose sacred space. In both cases, the realistic illusion of these shadow-casting beads serves to emphasise the spatial recession of the wall into a specifically feminine interior space located through the gates of a portico structure. In the 1783 temple, this is apparently the interior quarters of the palace of a draconic water god and his wife (Figs 32a, b & c), and in the other case, it is a wall fronting the inner sanctum of the Goddesses (Figs 33a & b).[85] The former temple also makes it clear that this harem interior is the fount or source of the divine fecundity — from under the table in the interior quarters and over the rooftops of the portico on the side wall pour great waves of life-giving water. The optical illusion of the beads in particular breaks the boundaries between painted surface and physical space, allowing the gods’ realm, and their blessings, to inhabit our mundane world.
Similar to the case of the porticoes in front of the temple, this idea of a secret, often specifically feminine interior behind the altar corresponds to the actual construction and ritual use of physical temple space. I have visited several temples across Xuan-Da where a male god was worshipped in two separate buildings — a front hall where he ‘did work’ (ban gong 辦公), and a rear hall, which was his living quarters (zhufang 住房, qin gong 寢宮). In one of these rear halls, the statue of the God of Walls and Moats (Cheng huang ye 城隍爺) could be found enthroned together with that of his wife, referred to simply as ‘the Goddess’ (Niangniang). The room had been painted in the 1990s with murals of the god’s household furnishings including a television, radio, fan, tea thermos, bookshelf, et cetera. Today, villagers consider this rear hall a location possessing a particular power of miraculous response (lingying 靈應) within the temple complex, especially efficacious to prayers for human fertility.[86]
This unseen rear space and the portico that is its entrance are important from the earliest extant images to the last, Western-influenced ones. However, the size and centrality of the side-wall portico structure grew over time, as well as the range of contexts where it might appear. In a God of Wealth mural dated to the Daoguang reign (1820–50) at the Cock’s-Crow Postal Station (Ji ming yi cai shen miao 雞鳴驛財神廟), the portico is still rendered in traditional isometric Chinese style, but it has expanded to cover nearly half of the wall. Here, the structure seems to represent the gateway of a wealthy devotee’s house. Towards this house, approach the God of Wealth from a celestial cloud and bearded Central Asians on the ground level, all bearing gifts (Fig. 34). By the late-nineteenth or early-twentieth century, the portico was depicted in ostentatiously Western style, with the receding grid of tiles on the ceiling now expanded to fill the entire composition; gods, who, in earlier murals would have processed across the mortal world, now sit upon this stage in state. While the earliest images already employ trompe-l’œil, European techniques allowed the effect to be perfected. The Yu county heavens had been Westernised, both in their architecture and in their mode of depiction, and the evocation of this hidden interior receding behind the altar wall had become the central theme of the murals.
This spatial recession is even more the case on opera stages, where the painted portico fronts the Mansion of the Western Seas and appears to represent both the entranceway to the Mansion and a stage in itself. In this case, the mysterious interior to which the gates lead is the backstage, actual and imagined, hidden beyond the two perforating gates of the scaenae frons (Fig. 35). Marvin Carlson has called this space ‘the hidden “other” world of the actor, the place of appearance and disappearance, the realm of events not seen but whose effects conditioned the visible world of the stage’.[87] Stage performers would traditionally enter the stage through the right-hand door and depart from the left.[88] Thus the circular, clockwise motion of the temple gods through the mortal world is mirrored by the circular motion of stage performers across the theatrical ‘world’ of the stage, while the painted portico-stages in the murals mirror the actual stage space. A portico is particularly useful as a sign in this context because it can represent a stage just as well as it can represent the pavilion at the gate of a palace; both structures are, after all, just a flat plinth with pillars supporting a roof. And this equivocation points to the shared function of these two types of physical structures. Both temples and stages, as we have seen, are, in some sense, anterooms built around the access points to other worlds, be they exotic-to-fictional (the Mansions of the Western Seas) or holy (the various divine palaces) or both. Moreover, the symbolic ability of these physical structures and spaces to access phantasmagorical and efficacious realms derives precisely from their equivocation or mirroring with the painted images of gateways on their walls.[89]
This understanding gives us the key to understanding many of the more unusual perspectival paintings around Yu county, and allows us to make sense of the innovation that these perspectival techniques unleashed. The undated Goddesses Temple of Stone-Waste Fort (Shi huang bu niangniang miao 石荒堡娘娘廟), while ostensibly traditional in its layout, is a fascinating example of this creativity.[90] The two side walls of this room depict the procession of the three goddesses out and back from their mansion, here rendered in skillful perspectival style. Although the lower part of the composition is destroyed so we cannot see if there are small mortal figures at the base as there would be in a Dragon King temple, it is clear that the circuit of the room represents the goddesses’ dominion, which is to say, our world. And from interior of the mansion, the goddesses’ blessings come welling into our space: children, healthy sons, thronging the balustrades, peaking out from the windows, and gambolling among the pillars and on the portico steps. The rear wall of the temple room, behind where the votive statues would once have stood, depicts painted screens from behind which lovely musician-attendants peer. These are the same screens that also appear on the walls of the opera stage, south across a plaza, suggesting the homology of the goddesses’ powerful, feminine interior to the fictive space extending laterally out and back from the theatre walls (Figs 36a & b, see also Fig 4).[91]
But beyond this now-standard layout, perspective technique and the gendered space have encouraged the artist to play with surface and depth, illusion and meta-image. The painted Mansions are perforated by windows and gates, revealing glimpses of more hidden spaces from which children bound or beautiful women peer. The screens that block our view back from the altar are themselves painted with interior and exterior scenes, views through moon gates, misted landscapes stretching back. As if to intentionally toy with the trompe-l’œil effect, the artist’s style playfully shifts across each panel of the room and the painted screens. Our gaze moves from the regal musician-attendants to the still, voyeuristic scenes on the screen, reminiscent of ‘beautiful-woman pictures’ erotica (meiren hua 美人畫),[92] then out through the portico gates to the riotous, demonic calvacade of the Goddesses’ procession across the exterior world, and finally up to triangular panels in the rafters, where cartoonish images depict two women seated in a mountain grotto (Fig. 37), and the moon-goddess Chang’e with her rabbit, pounding the elixir of immortality. The artist here has used Western perspective as one among many stylistic tools to break the boundaries between flat and three-dimensional space, producing a work that fascinates with its glimpses into illusory feminine realms as much as it functions as an efficacious religious and theatrical tableau.
Other works are still more innovative in their composition. Song Family Village (Song jia zhuang 宋家莊) has a ‘split-centre opera stage’ (chuan xin xilou 穿心戲樓), built as a square archway over the main village street (Fig. 38). Carved stone panels on either side of the stage depict actors peaking out from half-opened doors, emphasising the function of the stage as a gateway. The two flanking walls of this stage have murals depicting a receding perspectival view through the centre of a portico structure, which is signalled by the pillar-couplets and the sky-flower panelled roof. At the end of the perspectival corridor, the climactic scene of the opera Officials All Around the Bed (Man chuang hu 滿床笏) is taking place. The scene describes the joyous pageant of the sixtieth birthday of the main character, Guo Ziyi 郭子儀, who has, over the course of the play, defeated the famous rebel An Lushan 安禄山, and brought peace to China. His seven sons have all attained high office and his grandson has just achieved the first place in the examinations (zhuangyuan 狀元). Together with all the daughters and servants, they gather around to congratulate him.[93] The receding view through the portico in the stage-painting mirrors the physical view of the spectators and the god, through the split-centre opera stage and down the axial road of the fort (Fig. 39). To drive home the parallelism, a wooden plaque (bian 匾) originally hung over the stage with the epigrammatic words ‘Gaze Upon Its Depth’ (Guan qi shen 觀其深).[94] This pointed double-entendre points both to the play itself, to the visual recession of the perspectival drawings, and to the physical and fictive space of a stage that was also an actual gateway. In this way, the equivocation of the painted space, the actual space, and the fictive space of the play allows the desideratum of the painting — worldly success — to share the space of the fort.
Probably the most exceptional creation of this period, however, was the Temple to the Perfected Warrior of Wang and Liang [families] Fort (Wang Liang zhuang zhen wu miao 王良莊真武廟). This spectacular shrine, now half plundered and half collapsed, was unique within Yu county in its determination to show on each wall the interior of this mysterious divine space, of which in all other cases we see only the gateway and façade. Located at ground level in a curious northern barbican space of the fort, the temple was accessed only via a tiny gate through the inner fortress wall labelled The Palace of the Dark Emperor (Xuan di gong 玄帝宮) (Fig. 40). A line of sight stretched through this diminutive gate and down the axial street of the fort to an opera stage in the southern barbican. Within the temple, on the two side walls flanking the shrine room, were two immense full-wall images of the palace of the Perfected Warrior, composed in traditional isometric style. This painted palace was part temple, and part fortress, accessed by a fortified gate facing south, with corner towers in the style of the Forbidden City. Inside the palace, the Perfected Warrior was seated within a raised and roofed portico, flanked by his retinue of richly adorned generals, beautiful palace ladies, and fairy girls with slim swords. On the floor of the fortress-palace, beneath the portico, a mortal supplicant knelt (Figs 41a, b & c).[95] The mirroring argument was clear: the temple is an entrance to the palace of the god, which is the fortress, which is the god’s axial domain. Therefore, the worshipper may identify himself with the painted supplicant on the wall, who has entered now into the god’s realm and is communicating directly with the deity.
The rear (northern) wall of the room, which survived only partially when I first saw it in 2013, has now totally collapsed. One of three original panels survived: on the right side of the altar and facing south, so that the right (eastern) side of the image led into what was presumably once the centre of the depiction. Originally, statues would have stood on the altar-top, implying that the images on the wall represented the imagined space behind the physical and spiritual position of the god. Thus the image represents a glimpse beyond the scaenae frons of the Perfected Warrior’s heaven and into the mysterious interior beyond. In the mural, the palace had become a Western-style mansion, painted in clumsy but effective perspectival style in shades of deep blue and black, with a gallery of fantastic multistorey buildings receding towards the right, sharp spires and upper stories, walkways and pavilions. This was, evidently, the ‘rear palace’ (hou gong 後宮), the antaḥpuram or seraglio, for it was populated by strolling women in beautiful robes. At the far right or inner side of the receding line of structures, was a giant portico that led off-frame towards the vanished central panel of the main wall. Through the pillars of the portico we saw, distantly, even deeper apartments, an inner room, two servant girls in conversation, and a languid lady in red robes, drowsing upon her boudoir bed (Fig. 42).
All the elements we have identified above were present in this assemblage: the opposition between temple and stage; the physical construction of village space in a way symbolic of the spiritual positions of the gods; the spatial, textual, and visual recursion between the palace of the god and both the temple and the fortress; the use of this almost metafictive mirroring as a gateway or means of access into fantastic alternate realms, most commonly a holy interior located behind the altar wall; the function of the portico as one particularly powerful symbolic structure via which one could do this; the still-incomplete architectonic and pictorial ‘Westernisation’ of the village heavens; and, at a basic level, the extraordinary willingness of nineteenth and early-twentieth century village artists to experiment with the composition of space. This, then, was the effect of contact with the West and with perspectival painting in the Yu county village: a renaissance of artistic creativity and a reinvention of religious and theatrical space, which bloomed unexpectedly in these remote villages and was then cruelly and utterly cut off when the axe of Communism fell.
4) Conclusion
Jeehee Hong, speaking of Liao, Jin, and Yuan dynasty depictions of stages in north-Chinese tomb art, argues that the side passages of stages or ‘ghost-gates’ (gui men 鬼門) were used symbolically to represent the boundary of ‘fictive “elsewhere” ’ — a gate between worlds, in this case, between the world of the living and the world of the dead.[96] Only actors could move between these ‘ontologically heterogeneous spheres’.[97] In the context of the Yu county opera paintings, these observations can be usefully combined with those of scholars working on Western-influenced perspectival drawing. Columbia scholar Shang Wei 商偉 has recently attempted to show how, in the eighteenth and nineteenth centuries, newly introduced Western perspectival techniques fused in the popular literature with pre-existing Chinese associations of mirrors, lenses, and magic, and techniques of gazing. He tells us:
What arrived with perspectival techniques and Western viewing lenses was not concepts of reproduction or representation, nor was it the fixed forms of realism. Instead it was an exploration of the subjectivity and uncertainty of vision, the discovery of receding space, and an interest in true and false, emptiness and reality, existence and nonexistence. In this way the indigenous language of phantasmagoria and illusion (meiying mohuan 魅影魔幻) was reinvigorated and endowed with new meaning.[98]
These perspectival drawings in some ways represent the imaginative synthesis of all these discourses, and discourses local to Yu county as well. The Yu county stage and temple drawings do seem very much to belong to the world of which Shang Wei speaks: the world of Supplement to the Journey to the West (Xi you bu 西遊補), A Tower for the Summer Heat (Xia yi lou 夏宜樓), and Dream of the Red Mansions — all stories in which architectural spaces, devices of viewing, and the magical realms and beings of Chinese folk religion provide portals into new and bewitching fictional universes. These drawings also suggest, as Jeehee Hong does, gateways into other fictional or ontological spaces, receding off laterally from the divine or spectatorial line of gaze. And the stages and temples do belong, as well, to the very specific world of the Yu county fortress — a world in which the symbolic power of particular buildings was defined according to their imperial command over architecturally defined lines of sight. These images, axial views into fantastic and otherworldly architectural landscapes, serve to underscore the spatial symbolism from which Yu county temples derived their efficacy, and to establish the village opera stage as a powerful and separate space in its own right. They became structures through which we gaze into other possible universes, branching off from ours, at once here and elsewhere, and new, and old, and other.
Finally, and at the risk of stating the obvious, these scenes are the products of a formidable creativity and intellectual engagement. If we go looking for a Chinese intellectual and artistic response to the Western world, we should not neglect to seek it on the temples, stages and public squares of the Chinese village. When these illiterate farmers discovered the West, their response was one of admirable curiosity and imagination. They took these foreign techniques and adapted them to their own sophisticated indigenous language of space, gaze, architecture, fantasy, and religiosity. They painted perspectival pictures on their temple and opera stage walls and then they filled those newly created spaces with the things they wanted to see: light, colour, action, and excitement; exotic cities, skyscraping spires, and fantastic palaces; new technologies of spectacle; holy sanctums of the ancient gods; tough, smart, working-class heroines who took up arms against the powerful and unjust; and dizzying, axial views down the main streets of the new world.
East Asian Languages and Culture
University of California, Berkeley
htaubes@berkeley.edu
twosmallblocks@gmail.com
EAST ASIAN HISTORY 43 (2019)
1 All photographs were taken by the author. Many of the mural sites described in this paper are not in any way protected or even maintained, and several murals reproduced here have been plundered by thieves or otherwise destroyed since I took the photos. For this reason, I have cited the locations of these murals (that is, provided the village and temple name) in those cases where I am satisfied that they are not in danger, but elected to withhold the locations in cases where the site is vulnerable. Interested academic researchers may contact me at twosmallblocks@gmail.com or htaubes@berkeley.edu for precise locations and image galleries of all of the sites described here.
I am thankful to the following individuals, as well as innumerable others, for their help at various points in the creation of this paper: Zhao Wei 趙偉, Li Xinwei 李新威, Sophie Volpp, Lam Ling-hon 林凌瀚, Hannah Theaker, Kārlis Rokpelnis, Cao Xinyu 曹新宇, Hung Lichien 洪儷倩, Stephen Jones, David Johnson, Peter Bol, my advisor Jacob Dalton for his preternaturally accepting attitude towards my tangents, all the donors to my mural documentation fundraising campaign, and every Chinese villager who has assisted me over the last five years.
2 Works on the Chinese reception of Western artistic techniques have become legion. A particularly useful summary of recent scholarship is Wang Cheng-hua, ‘Wither Art History? A Global Perspective on Eighteenth Century Chinese Art and Visual Culture,’ The Art Bulletin 96.4 (2014): 379–94. Studies particularly relevant to this paper and not cited elsewhere are: Marco Musillo, The Shining Inheritance: Italian Painters at the Qing Court, 1699-1812 (Los Angeles: The Getty Research Institute, 2016); Kristina Kleutghen, Imperial Illusions: Crossing Pictorial Boundaries in the Qing Palaces (Seattle: University of Washington Press, 2015); Ibid., ‘Staging Europe: Theatricality and Painting at the Chinese Imperial Court,’ Studies in Eighteenth-Century Culture 42.1 (2013): 81–102. There are many others.
3 Petra Ten-Doesschate Chu and Ning Ding, ed. Qing Encounters: Artistic Exchanges Between China and the West (Los Angeles: The Getty Research Institute, 2015).
4 For instance: John Lust, Chinese Popular Prints (Leiden: Brill, 1996); Yachen Ma, ‘Picturing Suzhou: Visual Politics in the Making of Cityscapes in Eighteenth Century China’ (PhD diss., Stanford University, 2006).
5 For introductions to the extant corpus of late-imperial temple murals in Hebei and Shanxi provinces, see Wang Hui 王輝, ‘Hebei xiancun siguan bihua de fenbu yu ticai’ 河北現存寺觀壁畫的分布於題材, Wenwu chunqiu 56 (2000): 33–38; Chai Zejun 柴澤俊, Shanxi fosi bihua 山西佛寺壁畫 (Beijing: Wenwu chubanshe, 2006). Many other collections of reproductions exist.
6 Money for this original survey was kindly provided by the Gardener Traveling Fellowship of Harvard University.
7 I’ve been able to identify 50 structures (stages, temples, a few gatehouses) with drawings influenced by perspectival techniques, in 43 villages. These range from perfectly preserved full-wall murals to a few lines of paint on the one standing wall of a collapsed opera stage. Estimating extremely broadly here, this may represent something like three-fourths of all such images now extant in Yu county and its immediate surroundings. It is a very small fraction of the number of such images that existed before the Cultural Revolution. I have been to most of the villages in Yu county with old structures extant, but not all buildings were necessarily open to me at the time that I went. .
8 Liu Wenjiong 劉文炯, ‘Shui zhong bu: Ming-Qing zhi ji Yu-zhou cunbu kongjian de jiegou zhuanxing’ 水中堡:明清之際蔚州村堡空間的結構轉型 (PhD diss., China Central Academy of Arts, 2014), pp.202–46. So far as I know, the only discussion of these images in a Western language is in Pedro Luengo, ‘Yuánmíng Yuán en el siglo XVIII: Arte entre la diplomacia y la filosofía; entre Europa y Pekín,’ Araucaria: Revista Iberoamericana de Filosofía, Política y Humanidades 18.35 (2016): 193–216, at pp.210–12. Several sets of these murals have also been reproduced in Chinese publications: Hebei sheng gudai jianzhu baohu yanjiusuo 河北省古代建築保護研究所 and Yu xian bowuguan 蔚縣博物館, ed. Yu xian gu xilou 蔚縣古戲樓 (Beijing: Kexue chubanshe, 2014), p.346 and p.201; Yu xian bowuguan 蔚縣博物館 ed., Yu zhou simiao bihua 蔚州寺廟壁畫 (Beijing: Kexue chubanshe, 2013), pp.242–43 and pp.250–51.
9 The study of the village fortresses of Yu county is still in its infancy. This narrative derives from my own survey of fortress gatehouse inscriptions in Yu county and the surrounding areas. I have been able to photograph a total of 248 gatehouse inscriptions in 409 fortresses. Of these, 74 record a date apparently of the original construction of the gate, in some cases including dense information about the organisational structure of the fort at the time, which was typically based on the lijia 里甲 system. The number of Yu county fortress villages can be dated as: <1475: 0; 1475–99: 1; 1500–24: 8; 1525–49: 44; 1550–74: 7; 1575–99: 0; 1600–24: 2; 1625–49: 0; 1650–74: 0; 1675–99: 0; 1700–24: 0; 1725–49: 0; 1750–74: 1; 1775–99: 1; 1800–24: 2; 1825–49: 3; 1850–74: 2; 1875–99: 1; 1900–24: 2; 1925–49: 0. I strongly question the validity of dates after 1600 as foundation dates for the fort; many of these dates can be refuted by mentions of these forts in gazetteers prior to the stated date on the gate. My guess is that all or most of these post-Ming dates actually represent cases in which the fortress gate was refurbished and the new inscription simply did not specify this.
Chinese studies of Yu county fortresses so far have been hampered in their analysis by the lack of an accurate table of gatehouse dates. Important studies of Yu county villages consulted here include: Liu Wenjiong, ‘Shui zhong bu’; Deng Qingping 鄧慶平, ‘Huabei xiangcun de baozhai yu Ming-Qing bianzhen de shehui bianqian – yi hebei Yu-xian wei zhongxin de kaocha’ 華北鄉村的堡寨與明清邊鎮的社會變遷 – 以河北蔚縣為中心的考察, Qingshi yanjiu 清史研究 3 (2009): 19–27; Luo Deyin 羅德胤, Yu xian gu bu 蔚縣古堡 (Beijing: Qinghua daxue chubanshe, 2007), and others cited elsewhere. I intend to produce my own study of this in the future.
10 For information on this religious system, and corroborating information about the time of its appearance, see the works of Willem Grootaers: Willem Grootaers, ‘Les Temples Villageois de la Régon au Sudest de Tat’ong (Chansi Nord), Leurs Inscriptions et Leur Histoire,’ Folklore Studies 4 (1945): 161–212; Ibid., ‘Rural Temples around Hsüan-Hua (South Chahar): Their Iconography and Their History,’ Folklore Studies 10 (1951): 1–116; Ibid., ‘Temples and History of Wanch’uan 萬全 (Chahar): The Geographical Method Applied to History,’ Monumenta Serica 13 (1948): 209–316. Precisely tabulating Grootaers’ data is a complex task, but even a cursory examination will reveal that the immense bulk of the dates for which he was able to establish the existence of village cults fall into the period between 1447 and 1570 — that is, the period of the mass fortification of this region. A partial bibliography of English-language works dealing with north-Chinese popular religion in general, not cited elsewhere, include: Prasenjit Duara, Culture, Power, and the State: Rural North China, 1900–1942 (Stanford CA: Stanford University Press, 1988); James Flath, ‘Temple Fairs and the Republican State in North China,’ Twentieth-Century China 30.1 (2004): Adam Yuet Chau, Miraculous Response: Doing Popular Religion in Contemporary China (Stanford: Stanford University Press, 2005); 39–63; Thomas Dubois, The Sacred Village: Social Change and Religious Life in Rural North China (Honolulu: University of Hawai'i Press, 2005); Vincent Goossaert, The Taoists of Peking, 1800-1949: A Social History of Urban Clerics (Cambridge MA: Harvard University Asia Center, 2007); Daniel Overmyer, Local Religion in North China in the Twentieth Century: The Structure and Organization of Community Rituals and Beliefs (Leiden: Brill, 2009); Vincent Goosaert, ‘Is There a North China Religion? A Review Essay,’ Journal of Chinese Religions 39.1 (2011): 83-93; Wang Jinping, In the Wake of the Mongols: The Making of a New Social Order in North China, 1200-1600 (Cambridge MA: Harvard University Asia Center, 2018).
11 This narrative is derived from Deng Qingping’s extremely useful collection of stele texts from around Yu county: Deng Qingping 鄧慶平, Yu xian beiming jilu 蔚縣碑銘輯錄 (Guilin: Guangxi shifan daxue chubanshe, 2009). Taking only those 71 steles in the collection that refer to villages, as opposed to the walled county seat or independent monasteries, I was able to identify 121 instances of creation or reconstruction of various structures around the village (mainly temples). This is an extremely imperfect way to gather dates; it’s often unclear what should be considered a repair (new paint around the eves as part of a general refurbishment of the village temples?) or new founding of structures (a new storage shed in a pre-existing temple?). The reader who wishes for a full case-by-case methodology and a spreadsheet may email me. It should also be pointed out that the sample size is too small to draw real conclusions, since the data can be easily throw off by, for instance, one stele that records the repair of six village structures. With the above strong caveats, the physical history of Yu county villages as seen by their stele record would go as below. I’ve divided the data into 50-year periods, and separately tabulated New Founding of Structures (N) and Repair and Refurbishment of Pre-Existing Structures (R): 1501–50: (N) 5 (R) 1; 1551–1600: (N) 1 (R) 2; 1601–50: (N) 1 (R) 1; 1651–1700 (N) 2 (R) 0; 1701–50 (N) 5 (R) 6; 1751–1800 (N) 8 (R) 29; 1801–50 (N) 6 (R) 17; 1851–1900 (N) 3 (R) 20; 1901–50 (N) 4 (R) 7. From a qualitative standpoint, there is a marked increase after 1800 in the number of newly founded things which are not temples — that is, stages, wells, special seating area for women to watch the opera, flagpoles, etc. This suggests either that the roster of village temples was essentially complete by this point or that there was no longer money for large-scale construction projects. It should also be pointed out that the surge of repairs during the second half of the eighteenth century suggests that most of the structures involved had originally been built some hundreds of years before (that is, in the sixteenth century), so that they began to grow dilapidated en masse after about 1750. This corroborates the evidence from Grootaers’ data and the narrative of the Yu county gate-plaques.
12 Grootaers, Hsüan-hua, p.12; Wanch’üan, p.219; Tat’ong, p.167 all give detailed de-scriptions of these temples and their accoutrement.
13 Grootaers does not dwell on this type of relationship in his surveys, but a close reading indicates that they also existed in his regions, in some cases from the original period of fortification. See for instance: Grootaers, ‘Tat’ong’, p.195, p.197 and p.199; Ibid., ‘Hsüan-hua,’ p.48 and p.57; Ibid., ‘Wanch’üan,’ p.240 and p.248. Important recent Chinese analytical studies of these axial temples specifically in Yu county include: Liu Wenjiong, ‘Shui zhong bu,’ pp.106–42; Wang Xinlei 王新磊, ‘Yu Xian gu bu zhong de Zhen wu xinyang’ 蔚縣古堡中的真武信仰, Hebei beifang xueyuan xuebao (Shehui kexue ban) 33.4 (2017): 51–54; et al.
14 Grootaers refers to this type of image as the ‘Pursuit of the Evil Ones’, and mentions it frequently. See, for instance, Grootaers, ‘Hsüan-Hua,’ pp.13 and 33; Ibid., ‘Wanch’üan’, p.220.
15 The ur-source on Yu county opera stages is an internally published study produced as part of a series on materials for cultural history (wenshi ziliao 文史資料): Hebei sheng Yu xian zhengxie wenshi ziliao weiyuanhui 河北省蔚縣政協文史資料委員會, Yu xian gu xilou 蔚縣古戲樓 (Yu county: Yu xian hongsheng zhuanyan yinzhi chang, 2008). This was then reworked with diagrams and photographs to produce a second book of the same name: Yu xian bowuguan 蔚縣博物館 and Hebei sheng gudai jianzhu baohu yanjiusuo 河北省古代建築保護研究所 ed., Yu xian gu xilou 蔚縣古戲樓 (Beijing: Kexue chubanshe, 2014). The two books share some material but differ in other places; I have cited the latter book throughout due to its availability outside Yu county. Useful analytical studies of opera stages in Yu county are: Wang Penglong 王鵬龍 and Liu Jinping 劉晉萍, ‘Hebei Yu xian gubu yu miaoyu: minjian yanju kongjian de chanshi’ 河北蔚縣古堡於廟宇:民間演劇空間的闡釋, Zhongguo xiqu xueyuan xuebao 37.3 (2016): 45–52; Deng Dijiao 鄧弟蛟 ‘Yu xian gu xitai diaocha yanjiu’ 蔚縣古戲台調查研究 (Masters diss., Shanxi Normal University, 2017).
16 Yu xian bowuguan et al., Yu xian gu xilou, pp.384–404, gives a list of 233 opera stages in Yu county and their attributions. Of these, eight are attributed to the Ming dynasty (1368–1644); another eight are attributed to the Republican or Communist periods, and the rest are mostly mid- or late Qing. The situation regarding the eight putative Ming stages is unclear. The tables in the rear and the main text of the book disagree in many cases on the attribution. Even in the cases where the two agree, I have my doubts: the ‘Ming’ stages look to me no different from the Qing stages. The only stage which seems to me indisputably Ming is one located in Great Previous-Fort Village (Da gu cheng 大故城), which has beautiful mid-Ming brackets (dou gong 斗拱) under the eaves. Interestingly, despite the early date of this structure, its location (across from a Temple to Lord Guan (Guan gong miao 關公廟), defining a small square in the village centre) follows the same pattern as the later stages. Although there may have been few stages in rural Yu county during the Ming dynasty, the opposition between temple and stage was already established at this time, as was the use of that opposition to define open space within the village.
17 An excellent collection of transcribed Yu county stele texts mentioning opera stages or performances can be found at Deng Dijiao, Yu xian gu xitai diaocha yanjiu, pp.168–96.
18 According to Yu Xian Bowuguan et al., Yu xian gu xilou, p.3, exactly 233 stages now exist within the county, of originally over 800 that existed in the early twentieth century. I’m not sure if either of these numbers are necessarily scientific: many stages exist only as bare stone plinths or piles of rubble, and many more are located deep in the roadless mountains. I have visited 187 such stages in 409 villages of Yu county and counties immediately surrounding it, plus probably another dozen-odd stages in monastery courtyards and a further hundred or so more across a broad area of northern China. I also noted in my surveys of Yu county the locations of another 38 stages now vanished, although this was far from systematic and the absence of a stage in my survey should in no way indicate the absence of one historically. The point is, there are about 200 extant stages in Yu county, and there were many more before the Cultural Revolution; nearly every village had one.
19 Early dated examples include the murals at Stone-Waste Fort (Shi huang bu 石荒堡), discussed in the final section of the paper, which may have been painted in the eighteenth century. Another important piece of evidence are the murals at the Monastery of the Peaceful Sage in East Liu [Family] Village of neighbouring Guangling county (Guangling xian dong liu tuan cun an xian si 廣靈縣東留疃村安賢寺). The panels around the rafters contain perspectival paintings, dated from steles to a repair in 1818.
20 The graffiti (ti bi 题壁) on the stage walls support the dates 1800 to 1950 as the general period from which we may expect to find intact writings or drawings on stage walls. Traditionally, when one performed on a stage, it was common to write a small graffito on the backstage walls. These lively and often extremely colloquial writings contain information about the performers, the performance, messages to other performing troupes, jokes, poems, sexual innuendo, apotropaic formulae, and other random comments. They are often accompanied by ink drawings, topics of which include stages and performances, actors and tumblers, impressions of opera masks, soldiers, animals, calligraphic or possibly exorcistic marks, phalluses, bound (and therefore erotic) feet, sexual acts between both humans and animals, and sundry other things. The rear walls of these stages do also have a tradition of large-scale colour muralling, located behind the scaenae frons where only actors would see the images. Common topics include opera actors and immense qilin 麒麟 beasts howling at the moon. I have photographed some hundreds of stage inscriptions and graffiti. Yu Xian Bowuguan et al., Yu xian gu xilou, pp.370–81, gives a transcribed selection of written graffiti from 62 different stages, collected during a survey in 1984. By my count in this list, 186 dates are mentioned, not including a few more, which, for whatever reason, give only the regnal period and not the year. The graffiti count is: 1801–25: 1; 1826–50: 14; 1851–75: 29; 1876–1900: 69; 1901–25: 38; 1926–50: 13; 1951–75: 2; 1976–2000: 2.
This gives us a basic timeline for things written on opera stage walls. The earliest date from which we should expect to see extant images on stage walls is around 1800; weathering on these three-sided structures would efface anything earlier than this. The high point of Yu county opera graffiti, and quite possibly of Yu county opera itself, was the Guangxu reign (1875–1908). This one reign accounts for half (85 out of 169) of all dated stage graffiti recorded in Yu xian gu xilou. After this period the number of attested performances drop off quickly, no doubt due to the impoverishment and conflict that finally in the twentieth century reached up even into remote and mountain-ringed Yu. This list could be expanded and refined by dates from other stages than the ones listed in the book, but I think the basic contours would remain the same. (Another such list of graffiti can be found at Deng Dijiao, ‘Yu xian gu xitai diaocha yanjiu,’ pp.153–67.)
21 See Wang Zhijun 王志君 and Tian Yongxiang 田永翔, Zhongguo yu zhou minsu wenhua jicheng: difang juzhong gaishuo 中國蔚州民俗文化集成:地方劇種概說 (Beijing: Zhongguo xiju chubanshe, 2012), p.1. I’m not aware of a single pre-Revolution stage in Yu county that was not located across from some type of shrine. The stages that appear to lack temples generally turn out to be exceptions that prove the rule. I visited one stage located in Warm-Springs Town (Nuan quan zhen 暖泉鎮) that is a private ‘household stage’ (jia tai 家台) located within the mansion of a wealthy merchant family. The stage faces down the main axial line of the courtyards so as to be visible from the rear hall (zheng dian 正殿) where the ancestral shrine would have stood. Even private plays in private homes were offerings to the ancestors. In another village nearby (‘Su and Shao [Family] Fort’, Su shao bu 蘇邵堡) there is an opera stage apparently without a temple. Locals related to me when I visited in January 2014 that a ‘grass canopy’ (cao peng 草棚) originally stood across from the stage, and that the statues of all of the gods of the villages would be processed out from their temples and into this temporary grandstand whenever an opera was performed.
22 Representative studies in English include David Johnson, ‘Actions Speak Louder than Words: The Cultural Significance of Chinese Ritual Opera,’ in ed. David Johnson, Ritual Opera, Operatic Ritual: “Mu-lien Rescues his Mother” in Chinese Popular Culture (Berkeley: Chinese Popular Culture Project, 1989) pp.1–45; Barbara Ward, ‘Not Merely Players: Drama, Art, and Ritual in Traditional China,’ Man (New Series) 14.1 (1979): 18–39; an introductory article about such rituals in northern China can be found at: Ekaterina Zavidovskaya, ‘Celestial and Human Audience of the Traditional Opera xiqu in Modern Shanxi and Shaanxi provinces,’ Problemy Literatur Dal’nego Vostoka 2 (2012): 485–97.
23 With some exceptions; such ‘mystery plays’ did exist and were performed on these stages too. The main occasion for these seems to have been the winter sai operas 賽戲 in the first lunar month, which involved elaborately re-enacted divine combats and exorcisms intended to purify the village for the new year. For descriptions of these in Yu county or areas immediately adjacent, see Wang Zhijun et al., Difang juzhong gaishuo, p.28 for a narrative of the rituals, and p.165 for examples of the votive songs sung during these performances. David Johnson, Spectacle and Sacrifice: The Ritual Foundations of Village Life in North China (Cambridge MA: Harvard University Asia Center, 2009), pp.69–84 contains a detailed description and analysis of such a sai opera performed in a village just over the mountains from Yu county to the north.
24 See Grootaers, ‘Hsüanhua,’ p.14, and ‘Wanch’üan,’ p.14, for corroborating statements.
25 Wang Zhijun et al., Difang juzhong gaishuo, p.128 and p.134, gives two potential calendars of these events.
26 Deng Qingping, Yu xian beiming jilu, p.466. The text tells us that the stage was built to ‘Repay and report the autumn harvest, in the ninth month we sing operas to make offerings to worship the Dragon Heavens. The people will be happy in their work and, each peaceful in their appointments, give thanks and repayment.’ 答報秋成九月間唱戲獻供敬 龍天人民樂業個安分酧謝. Almost all of the other steles in Deng Dijiao’s list (Yu xian gu xitai diaocha yanjiu, pp.168–96) document similar votive use of these stages. An interesting early description of rural rainmaking rituals in counties immediately adjacent to Yu is found in: E. De Vleeschouwer, ‘K’i Yu 祈雨’ Folklore Studies 2 (1943), pp.39–50.
27 Here and throughout, I follow the Chinese practice of assigning left and right within architectural space according to the perspective of the enthroned god or master of the house — that is, seated in the northern-most hall (zhengdian 正殿) and facing south towards the gate or stage.
28 For Daoists and shawm bands in North Chinese villages, see particularly the studies of Stephen Jones, including Ritual and Music of North China: Shawm Bands in Shanxi (Aldershot: Ashgate, 2007); In Search of the Folk Daoists of North China (Aldershot: Ashgate, 2010); Daoist Priests of the Li Family: Ritual Life in Village China (St Petersburg, FL: Three Pines Press, 2017).
29 Yu county villagers today, when asked why opera stages always face temples, will simply shrug and say ‘shen hao kan xi’ 神好看戲 ‘The gods like watching opera’. The Qing-period stele texts that deal with stages put this in slightly more circuitous language, but it seems to me that the basic reasoning of this statement was true then as it is today.
30 Deng Qingping, Yu xian beiming jilu, p.412. The village in question is Water [Gully] West Big Fort (Shui [jianzi] xidabu 水[澗子]西大堡).
31 Ren Family Gully (Ren jia jian 任家澗); locals tell me that the temple and stage at West Shop (Xi dian 西店) once had a similar arrangement.
32 These three villages are the Song Family village (Song jia zhuang 宋家莊), Yang and Juan [Families] town (Yang juan zhen 陽眷鎮), and Little Expeditionary Valley-Mouth (Xiao tan kou 小探口).
33 One of the most important and most universally overlooked aspects of Chinese monumental architecture is that such buildings are frequently constructed atop low, flat-topped stone plinths which extend a few feet beyond the walls of the building. Temple, stage, and gatehouse plinths in present Yu county are visibly smoothed by generations of village bottoms. Glancing down while ascending temple stairs, one frequently notices chess-boards cut into the stone. Susan Naquin has argued that temple courtyards in Beijing were among the few ‘public spaces’ in the city, and thus played an important part in constituting urban civil society. See her Peking: Temples and City Life, 1400-1900 (Berkeley: University of California Press, 2000), xxx–xxxi and 88–89.
34 See, for instance, Yin Jianhong 殷建宏, Xitai yu shehui: Ming-Qing Shanxi xitai yanjiu 戲台與社會:明清山西戲台研 (Beijing: Zhongguo shehui kexue chubanshe, 2009), p.68. Although I haven’t directly cited it elsewhere in the text, I should acknowledge here the debt of my thinking in this paper to this insightful book.
35 For one of many such statements available in the literature, see David Johnson, ‘Opera Imagery,’ in eds Judith Zeitlin and Li Yuhang, Performing Images: Opera in Chinese Visual Culture (Chicago: Smart Museum of Art, University of Chicago, 2014), 44–57, at p.54.
36 Arthur Smith, Village Life in China: A Study in Sociology (New York: Fleming H. Revell Company, 1899), p.65.
37 A primary source for this movement is again the graffiti on the stage walls. Yu Xian Bowuguan et al. Yu xian gu xilou, pp.370–82 gives transcriptions of a whole series of these writings, many or most of which seem to have been written by performers not local to the village in question. Performers are attested from far-off metropoles such as Beijing (p.370), Zhangjiakou (p.371) and Datong (p.372), as well as many other smaller but still distant locations. Opera troupes originating in Yu county are themselves attested in nineteenth-century stage graffiti from neighbouring Yangyuan 陽原, Huailai 懷來, and Chicheng 赤城 counties (p.382). The latter county is a good hundred miles away, across rugged passes. The point is that there was a constant circulation of performers not only from village to village, but from county to county, and from the cities to the villages and back.
38 For an anthropological description of some of these fairs held in Yanggao county 陽高縣 next door to Yu, see Jones, Ritual and Music, p.72 and onward.
39 Elsewhere in northern China, ‘Yang ge’ is a genre of folk song, but in Yu county the term refers to the indigenous county opera style.
40 Wang Zhijun et al., Difang juzhong gaishuo, p.1. The book on opera stages by the Yu County Museum (Yu Xian Bowuguan, Yu xian gu xilou, pp.13–14) gives a fascinating further typology of the ‘occasional’ types of operas, at which plays from the above genres might be performed. The list goes as follows: plays to consecrate the statues in monasteries or temples (kai guang xi 開光戲; plays to repay vows to the gods, for instance for cure from sickness or the successful conclusion of a lawsuit (huan yuan xi 還愿戲); plays on market days (kai shi xi 開市戲); plays performed at marriages (qing hun xi 慶婚戲); plays to celebrate the decennial birthdays of elderly members of the community (qing shou xi 慶壽戲, or tang hui xi 堂會戲); plays to celebrate the building of an opera stage (kai tai xi 開台戲); plays at temple fairs (miao hui xi 廟會戲); plays to beg the gods for rain (qi yu xi 祈雨戲); plays at funerals (fa sang xi 發喪戲); plays to celebrate the posthumous ritual marriage of those who died young (si hun xi 死婚戲); plays to thank the gods after a harvest, which the editors note was actually the most common type of play (xie cha xi 謝茬戲); ‘punishment plays’ (also called fa xi 罰戲) demanded if the opera troupe was late to the occasion or otherwise remiss; ‘closing-the-box plays’ (feng xiang xi 封箱戲), which were performed in the tenth month at the end of the summer opera season in order to ‘shut the gods’ gate’ (fengbi shen men 封閉神門) and ritually signal the winter absence of the gods; and plays expressing Buddhist or Daoist moral or religious truths (Fo Dao xi 佛道戲). A similar list is found in Wang Zhijun et al., Difang juzhong gaishuo, pp.128–34.
41 For a cogent introduction to the relationship between religion and fiction in China, see Meir Shahar, ‘Vernacular Fiction and the Transmission of Gods’ Cults in Late Imperial China,’ in ed. Meir Shahar and Robert Weller, Unruly Gods: Divinity and Society in China, (Honolulu: University of Hawai‘i Press, 1996), pp.184-211. Another book-length study of the religious elements in these three classic pieces of Ming–Qing fiction, and one that has strongly informed the thinking of this paper, is Li Qiancheng, Fictions of Enlightenment: Journey to the West, Tower of Myriad Mirrors, and Dream of the Red Chamber, (Honolulu: University of Hawai‘i Press, 2004).
42 Mark Meulenbeld, Demonic Warfare: Daoism, Territorial Networks, and the History of a Ming Novel (Honolulu: University of Hawai‘i Press, 2015), represents one recent study of the complex inter-relationship of a novel and religious worship.
43 David Johnson, Spectacle and Sacrifice, p.116 gives a fascinating description of a exorcistic ritual performed in a village in southern Shanxi in which an entire gruesome mock execution of a scapegoat was framed in specifically operatic terms. Spectators were threatened that if they misbehaved, they too would be taken not to hell, but to another stage representing hell, where presumably an operatic representation of their torture and execution would also take place. He points to other sets of rituals performed next-door to Yu county in Hunyuan 渾源, where the whole village would turn out during the New Year’s festivities to enact events from two popular novels, Shuihu zhuan 水滸傳 and Feng shen yanyi 封神演義 (ibid., p.84). Wang Zhijun et al., Difang juzhong gaishuo, p.31 records a ritual expulsion of the Draught Demon (Han ba 旱魃) in Yu county. Although to the Western reader the ritual seems purely exorcistic, the main events take place on the opera stages, and the terms used at least in Wang and Tian’s book are operatic: the events are referred to as an ‘opera’ (ju 劇 and xi 戲), the ritual as a ‘performance’ (yan 演, yanju 演劇); the character of the Draught Demon plays the part of a ‘clown’ (chou 丑, p.32). The whole performance concluded with the ‘troupe head’ (banzhu 班主) sitting on the stage and telling stories from books (shuo shu 說書) late into the night. As Johnson puts it, ‘… the boundaries between religion and drama collapse — […] creating deep ambiguities’ (Johnson, Spectacle and Sacrifice, p.116).
44 Again, see Wang Zhijun et al., Difang juzhong gaishuo, pp.28–32.
45 Willem Grootaers, ‘Hsüan-hua,’ p.65 and ‘Wanch’üan,’ p.247 for descriptions of temples to Guanyu that contained long panelled narrations telling the famous events of this god’s life, presumbly drawn from the popular novel San guo yanyi. At least one painted narration of Feng shen yanyi exists in Yu county, and many more were painted in the counties of Hunyuan 渾源 and Ying 應縣 adjacent to the west.
46 The technical term for these is yinglian 楹聯 ‘front-pillar couplets’, which refers to paired stanzas written vertically on long wooden plaques affixed to the front pillars of stages or porticoes (menlang 門廊).
47 These couplets appear in an appendix to the Yu County Museum’s book on opera stages (Yu Xian Bowuguan, Yu xian gu xilou, p.383). The book does not say from which village each poem was collected, and I have not seen any of them myself in my survey. One of the main editors of the book, and the director of the Yu County Museum, Li Xinwei 李 新维 (personal communication, 21 July 2017) explained that the couplets in this collection were compiled from the few pre-revolution wooden pillar plaques which do survive, but mainly from the memory of elderly villagers about what had once been written on their stages. Although such verses seem slightly arcane to the modern Westerner, it seems very plausible to me that Yu county villagers would remember this type of information accurately. The opera stages are the centre of a swirl of fond memories for elderly villagers, and questions about this structure always elicit interested and interesting discussion. Even after the physical destruction of the written poems, these stage couplets would have remained treasured bits of village lore.
48 See, for instance, Eugene Wang, Shaping the Lotus Sutra: Buddhist Visual Culture in Medeival China, (Seattle: University of Washington Press, 2005): pp.238–321; Ibid., ‘Of the True Body: The Famen Monastery Relics and Corporeal Transformation in Tang Imperial Culture,’ in Body and Face in Chinese Visual Culture, eds. Wu Hung and Katherine Tsiang (Cambridge MA: Harvard University Asia Center, 2004), 79–118, at pp.97–118; Jeehee Hong, Theater of the Dead: A Social Turn in Chinese Funerary Art, 1000–1400 (Honolulu: University of Hawai‘i Press, 2016).
49 Several publications propose Ming-period dates for various murals around Yu county, but given the lack of unambiguous textual evidence or a comprehensive stylistic study of Yu county art, I remain sceptical.
50 These murals are reproduced in full in Hebei sheng gudai jianzhu baohu yanjiusuo 河北省古代建築保護研究所 and Yu xian bowuguan 蔚縣博物館 ed., Gucheng si bihua 故城寺壁画 (Beijing: Kexue chubanshe, 2010).
51 It should be noted that, in the broader north-Chinese context, procession murals to and from a portico-fronted palace are attested from at least the sixteenth century onward. The most famous example is the sublime Ming-period murals at the Temple of the Holy Mother in Tian village outside of Fenyang city in central Shanxi (Fenyang shi Tian cun sheng mu miao 汾陽市田村聖母廟). Closer to Yu county, the Temple to the God of Fire at Ever-Peaceful Township in Yanqing district (Yanqing qu yong ning zhen huo shen miao 延慶區永寧鎮火神廟) probably represents an early Qing (seventeenth century) example. The Goddesses murals at Cock’s-Crow Postal Station in Zhuolu county (Zhuolu xian ji ming yi 涿鹿縣雞鳴驛) and South Liu [Family] village (Nan liu zhuang 南留莊) in Yu county are undated but probably from the eighteenth century. Other examples will be discussed later in this paper.
52 James Cahill, Pictures for Use and Pleasure (Berkeley: University of California Press, 2010), pp.68–70. See also Lust, Chinese Popular Prints, for a book-length study of such images.
53 Ellen Johnston Laing, ‘Reform, Revolutionary, Political, and Resistence Themes in Chinese Popular Prints, 1900–1950,’ Modern Chinese Literature and Culture, 12.2 (2000): 123–75, at pp.142–48 and throughout.
54 Cheng-hua Wang, ‘Whither Art History?,’ p.387.
55 Qinghai sheng wenhua ting 青海省文化庭, Qutan si 瞿曇寺 (Ürümqi: Xinjiang keji weisheng chubanshe, 2000), p.158.
56 Liu Wenjiong, ‘Shui zhong bu,’ pp.217–18 suggests that Ming-period stages in Yu county have images of screens, while Qing-period stages have images of architectural drawings. I am sceptical of this in the extreme; among other things, I see little proof for the existence of more than one or two un-renovated Ming-dynasty stages in Yu county to begin with, let alone the existence of Ming-dynasty stage murals. It is also clear that paintings of screens continued to be produced on stage walls into the twentieth century, right alongside the perspectival images.
57 I have not seen significant research on opera-stage murals elsewhere in north China, and thus I add here a few extremely scattershot notes based on my own travels around Shanxi and Hebei. Perspectival or Western-influenced paintings on opera stage walls certainly did exist outside of Yu county. I have seen perspectival images at the Temple of the Eastern Marchmount at Pu county (Pu xian dongyue miao 山西蒲縣東嶽廟), Fu village of Dai county (Dai xian fu cun 代縣富村), and the Monastery of the Mountain of Enlightenment in Lingqiu county (Lingqiu xian jueshan si 靈丘縣覺山寺), all in Shanxi province. Another, published set of perspectival murals from from central Shanxi can be found in Zhao Peiqing 趙培青 and Li Jingming 李晶明, Yu xian gudai bihua lu 盂縣古代壁畫錄 (Taiyuan: Sanjin chubanshe, 2014), pp.3–18. Perhaps most intriguingly, Scottish missionary Alexander Williamson reports seeing murals depicting ‘continental cities’, that ‘succeeded wonderfully well with the perspective’ in the City God Temple of Taiyuan in the 1860s (Taiyuan fu chenghuang miao 太原府城隍廟). These images, however, had recently been defaced on orders of the city government, perhaps because the painted buildings had crosses on top. Alexander Williamson, Journeys in North China, Manchuria, and Eastern Mongolia, with some account of Corea, Vol. I (London: Smith, Elder & Co., 1870). pp. 305–306.
However, despite these interesting examples, outside Yu county such ‘exotic’ images seem to have been exceptions rather than the rule. The majority of extant rural stages north and east of Yu in Zhangjiakou cityship have images either of folding screens or of dynamic life-sized images of actors and tumblers. In northern Shanxi province to the west of Yu, most stages (and many temples) have austerely painted murals depicting scrolls hung on nails, usually a shan-shui landscape painting flanked by vertical couplets (duilian 對聯). Although less interesting in themselves as artistic productions, these images do sustain the connection (discussed below) between temple and stage walls, and the use of both surfaces as sites for trompe-l’œil painting. I have not visited villages or stages on the south-east slope of the Taihang mountains. Very provisionally, I suggest that the interest in perspectival and otherwise Western-influenced images represented a broad turn in rural north-Chinese opera and religious visual culture from at least the turn of the nineteenth century on, but that Yu county may have been unusual in the pervasiveness of this interest and in the particular form that these images took.
58 Liu, ‘Shui zhong bu,’ p.217. A few stages have unique decorations. One stage has massive battle scenes flowing across both flanking walls, in which the various heroes and villains of the Feng shen yanyi clash with each other amidst swirling clouds. (Bu Family North Fort [Bu bei bu 卜北堡]). Another one has the slightly Boschian trope of various scenes taking place in the interior of gigantic flora, particularly fruit. A general holds up the head of his decapitated enemy and strikes a martial pose inside of a gargantuan pumpkin, two finches perch on branches within a leafy stemmed apple, a domestic conversation is conducted between two women standing in a fine mansion located inside of a huge tulip, etc. This seems to be the work of an eccentric genius.
59 I intentionally use the vague and exogenous word ‘pagoda’ here to encompass several overlapping Chinese-language terms, referring to a broad range of structures. The tiered multistorey towers depicted in these images are referred to in the image-captions and in the source novel as both ting 亭 and lou 樓; in other contexts, the same types of buildings can be called ta 塔 or ge 閣. In fact, the English word ‘pagoda’ conjures up roughly the correct range of different structures, and the faintly chinoiserie connotation of the word fits very well with exotic Western phantasmagoria (‘occidenterie’) of these images as a whole.
60 To my knowledge, the first source to identify this theme in print is Hebei sheng Yu xian zhengxie wenshi ziliao weiyuanhui, Yu xian gu xilou, p.200. Liu, ‘Shui zhong bu,’ pp.216–32 provides a more detailed analysis of its appearances.
61 My three main sources on this are: Anon., Lü mudan quan zhuan 綠牡丹全傳 (Beijing: Baowentang shudian 寶文堂書店, 1985); Margaret Wan, Green Peony and the Rise of the Chinese Martial Arts Novel (New York: State University of New York Press, 2009); Zhang Yahui 張雅惠, ‘Xiayi xiaoshuo “Lü mudan quan zhuan” de liuchuan yu gaibian’ 俠義小說《綠牡丹全傳》的流傳與改編, Dongfang renwen xuezhi 7.2 (2008): 207–26.
62 Anon, Lü mudan quan zhuan, pp.121–22: 華碧蓮悄悄挨近猴邊,才待伸手去捉,猴子見華碧蓮擋住右邊,無空逃走,那畜生發急,用力一跳,欲從華碧蓮頭上跳過。不料這四望亭多年未曾修理,不料朽爛,灰磚裂開,華碧蓮同猴子俱墜下來了。眾人齊道:“不好了,掉下人來了!”華碧蓮從亭上掉下,華振芳、余謙並同巴氏弟兄俱偕驚慌無措。華碧蓮自料性命難保。只見四,五叢人之外有一少年,大叫一聲:“還不救人,等待何時?”將身一縱,跳過來將華碧蓮雙手接住,抱在懷中,坐在塵埃。眾人齊道:“難得這個英雄,不然要跌為肉泥!”華振芳同眾人跑過來一看,接住華碧蓮的不是別人,正是駱弘勛大爺。
63 Margaret Wan, Green Peony, p.13 notes in respect to this and other scenes that ‘Green Peony is exceptional among the martial romances in giving a psychological description of Hua Bilian’s feelings for Luo Hongxun, from initial infatuation to secret bliss when he catches her in his arms after she falls off a roof’.
64 It also contains some extremely colourful language. ‘Your mother’s dog-stinking donkey-farts! You could be the Son of Heaven standing in the middle of the court, and you’d still have to pay me what you owe me.’ (‘放你娘的狗臭驢子屁!就是朝中的太子,許我的也要給我。’) Anon., Lü mudan quan zhuan, p.124.
65 ‘The Pagoda of Gazing in the Four Directions’ is not so far off from si wang ting (死亡亭) ‘death is stopped’. This type of painted or carved rebus is extremely common in architectural decoration around Yu county and China generally.
66 New year ‘Sai opera’ rituals in Xuan-Da habitually included the exorcistic pursuit, capture, and ritual banishment from the village of a demonic figure dressed as a monkey. See Johnson, Spectacle and Sacrifice, p.69 and Wang Zhijun et al., Difang juzhong gaishuo, p.31. Temple murals depicting the expedition of the God of the Five Ways (Wu dao shen 五道神, a common crossroads god) show the god capturing evil spirits. The evil to be pacified is, in some cases, depicted as a lecherous monkey with long quail feathers in its cap, similar to those worn in opera performances.
67 Liu, ‘Shui zhong bu,’ pp.229–43. Jing Anning in his study of the Song dynasty murals at Guangsheng Monastery in central Shanxi has suggested that opera performances contained sexually titillating themes in order stimulate the ‘mixing of Yin and Yang’ and allow the Dragons’ life-giving rain magic to work. This represents another possible explanation for these murals, see Jing Anning, The Water God’s Temple of the Guangsheng Monastery: Cosmic Function of Art, Ritual, and Theater (Brill: Leiden, 2002) pp.144–99.
68 This association is extremely tempting, but, unfortunately, the geographic distance and late attestation of the mechanised performances make it difficult to confirm a connection. The Shanghai Jingju Theatre Company (Shanghai jingjuyuan 上海京劇院) website notes that the earliest performance of Green Peony known to them was in 1875. By 1915, the performances ‘used mechanical sets, real swords and real lances, suspension by ropes, and other special effects in order to attract customers’ 以機關佈景,真刀真搶,穿插走繩索等各類特技為號召。[Anonymous], ‘Hong bi yuan’ 宏碧缘, Shanghai jingjuyuan上海京劇院, [Aug 2008], (). Quoted in Zhang Yahui, ‘Xiayi xiaoshuo “Lü mudan quan zhuan”,’ p.212. Another interview with an elderly Shanghai actor, published in the 1962 edition of Shanghai xiju上海戲劇, describes how Jade-Lotus would leap up onto each level of the stepped pagoda, cry out ‘Ai-ya!’ 哎呀! as the topmost brick turned beneath her feet, and then plummet dramatically down into Luo Hongxun’s arms, all accomplished by means of ropes and pulleys. Even in the 1960s, the old performer seems to have been impressed by his memories of it; he emphasises how ‘new and fresh’ (xinxian 新鮮) it seemed at the time. (Lü Jian 呂健, ‘ “Hong bi yuan” de jiguan bujing’ 宏碧緣的機關佈景, Shanghai xiju 上海戲劇 7 [1962]: p.24.) It’s possible that Yu county people travelling in the south could have witnessed these performances, or that such mechanised performances could have taken place in Beijing, but I have no record of it.
69 Wang Zhijun et al., Difang juzhong gaishuo.
70 That is, Lü mudan 綠牡丹, Hong bi yuan宏碧緣, Taohua wu 桃花塢, Si jie cun 四杰村, Jiaxing cun (‘Jiaxing village’ 嘉興村, Hua bilian duo zhuangyuan 花碧蓮奪狀元, etc. See Zhang Yahui, ‘Xiayi xiaoshuo “Lü mudan quanzhuan” de liuchuan yu gaibian’ for a fuller discussion of these various descendant titles.
71 These have been discussed at Liu, ‘Shui zhong bu,’ p. 219. The appellation ‘Mansion of Gazing in the Four Directions’ should indicate that it would be unwise to draw too clear a line between the ‘Mansion’ and the ‘Pagoda’. They were, rather, two of many potential compositional elements at the artist’s disposal, and could be combined or rearranged at will.
72 Lithographs of the Garden produced by the Manchu artist Yi Lantai 伊蘭泰 are reproduced in several sources, but fully in Régine Thiriez and Ellen Lawrence, The Delights of Harmony: The European Palaces of the Yuanmingyuan & the Jesuits in the 18th Century Court of Beijing (Worcester, MA: Iris and B. Gerald Cantor Art Gallery, College of the Holy Cross, 1994). We read in these lithographs of ‘observatories’ (guan 觀) in the garden: The Observatory of the Distant Seas (Yuan ying guan 遠瀛觀, one of the mansion buildings), The Observatory Facing Out (Fang wai guan 方外觀, a belvedere), The Observatory of the Fountains (Shui fa guan 水法觀). These ‘observatories’ must have taken a form quite similar to the porticoes in the opera stage depictions, about which more later. Western linear perspective was also important. Sights within the garden included a ‘Mountain of Perspective’ (Xian fa shan 線法山), as well as a ‘Perspective Painting’ (Xian fa hua 線法畫). The ‘Perspective Painting’ stood on the far side of a lake at the end of the garden, the final view; it depicts a European village according to the European gaze, receding into the perspectival distance.
73 See Régine Thiriez, ‘Old Photography and the Yuanmingyuan,’ Visual Resources 6.3 (1990): 203–18.
74 The white colour of the buildings, the angles of the roofs, and the ornamentation over the windows all suggest that these images were copied either directly from the real Mansions or from the lithographs. For an examination of this, see Pedro Luengo, ‘Yuánmíng Yuán en el siglo XVIII,’, pp.210–12.
75 Feng Jicai 冯驥才, ed. Zhongguo muban nianhua jicheng, ershisan juan 中國木板年華集成,二十三卷 (Beijing: Zhonghua shuju, 2011).
76 Reproduced in Feng Jicai 馮驥才, ed. Zhongguo muban nianhua jicheng: Taohua wu juan (xia) 中國木板年畫集成:桃花塢卷 (下) (Beijing: Zhonghua shuju, 2011), p.331.
77 Two representative studies of this are: Kristina Kleutghen, ‘Peepboxes, Society and Visuality in Early Modern China,’ Art History 38.4 (2015): 763–77; and Shang Wei 商維, ‘Bi zhen de huanxiang: xiyangjing, toushifa yu daguanyuan de menghuan meiying (shang, zhong, xia)’ 逼真的幻象:西洋鏡、透視法與大觀園的夢幻魅影 (上、中、下), Cao xueqin yanjiu 1 (2016): 95–117; Cao xueqin yanjiu 2 (2016): 103–23; Cao xueqin yanjiu 3 (2016): 38–63.
78 Kristina Kleughten, ‘Chinese Occidenterie: The Diversity of “Western” Objects in Eighteenth Century China,’ Eighteenth-Century Studies 47.2 (2014): 117–35, at p.117.
79 Yu xian bowuguan, Yu xian gu xilou, p.383.
80 Cahill, Pictures for Use and Pleasure, p.68.
81 This type of image also existed in the imperial court. One image titled ‘The Qianlong Emperor Watches a Play’ (Qianlong guan ju tu 乾隆觀劇圖) depicts the emperor seated on one side of a courtyard, watching opera on the other side. In this case, both the area where the emperor sits and the stage upon which the actors stand resemble the Xuan-Da portico images. Zhongguo xiqu zhi weiyuanhui 中國戲曲志委員會, Zhongguo xiqu zhi: Hebei juan 中國戲曲志:河北卷 (Beijing: Zhongguo zhongxin, Xinhua shudian beijing faxingsuo jingxiao, 1993) has the image reproduced in the photographic front matter without page numbering. The theme was the subject of Westernising interpretations from an early point, including multiple images of the Qianlong emperor seated on pavilions by Guiseppe Castiglione (1688–1766)[ in conjunction with Chinese artists, and Western-influenced mid-eighteenth century woodblock ‘yearly pictures’ examples. See, for instance, Kleughten, Imperial Illusions, pp.108, 109, 142, 145, 161, etc. Of particular note is the image in the Musée Guimet titled ‘Kazaks Offering Horses in Tribute to the Emperor Qianlong’, reproduced at Kleughten p.161. Both the composition and the subject are clearly echoed particularly by the God of Wealth (Cai shen 財神) depictions in the Xuan-Da temples.
82 For instance, the Dragon King Temple of South Lü Family Village (Nan lü jia zhuang 南呂家莊) in Yu county; there are many others.
83 Kaiyang Fort (Kaiyang bu 開陽堡) in neighbouring Yangyuan county, for instance, has a sixteenth-century stele that refers to the axial Temple of the Perfected Warrior as a ‘Palace of the Dark Emperor’ (Xuan di gong 玄帝宮). For other examples of these metonymic naming conventions, see Grootaers, ‘Wanch’üan,’ p.249.
84 For instance, the 1785 stele at Stone-Waste Fort (Shi huang bu 石荒堡), reproduced in Deng Qingping, Yu xian beiming jilu, pp.432–33.
85 The 1783 temple appears to be devoted to a local water deity whose cult differs only slightly from that of the Dragon Kings. The shrine is known as the Holy Shrine of [Musical] Harmonies (Lülü shenci 律呂神祠), located at the edge of a marsh in a village called Holy Creek in Hunyuan county (Hunyuan xian shen xi cun 渾源縣深溪村). The 1783 repair stele in the courtyard explains that ‘ “Harmony” refers to the modulation of yin and yang’ 蓋律呂著調理陰陽之謂, hence the regulation of rainfall. Two statues are enshrined in the main room, one male and one female. The murals match the standard Dragon King processions figure-by-figure, except instead of the usual five dragons of the five oceans, only one male god rides out and back on the two walls, while the female god awaits him at the palace.
86 This location was the Temple to the God of Walls and Moats at Holding-the-Gate Fort of Yanggao county (Yanggao xian zhenmen bu chenghuang Miao 陽高縣鎮門堡城隍廟). Small nooks in this rear hall are full of little infants made of clay, deposited there by women to ‘return the child to the god’ after prayers for reproductive fertility had been answered (personal communication, shrine caretaker, 13 and 14 June 2018). According to villagers at North Tower Gate Village of nearby Ying county (Ying xian bei lou kou cun 應縣北樓口村, personal communication, 26 August 2018), the Temple to the Northern Marchmount (Bei yue miao 北岳廟) on the mountain-top by their village once had similar front and back halls for the god to work and rest, although the rebuilt rear hall is now a shrine to the Buddha (Fo dian 佛殿). This type of arrangement may be very old. Jing Anning makes the fascinating suggestion in his study of the Water God’s Temple at Guangsheng Monastery that the rear-wall trompe-l’œil murals of female attendants offering food were originally painted in the early fourteenth century as a substitute for a physical ‘rear palace’ (hou gong 後宮) building that had been destroyed in an earthquake (Jing, The Water God’s Temple, pp.128–29).
87 Marvin Carlson, Places of Performance: The Semiotics of Theatre Architecture (Ithaca: Cornell University Press, 1989), p.131. Quoted in Judith Zeitlin, The Phantom Heroine: Ghosts and Gender in Seventeenth Century Chinese Literature (Honolulu: University of Hawai‘i Press, 2007), p.146.
88 Zeitlin, The Phantom Heroine, pp.144–45. On modern Yu county stages, these gates are often marked ‘The General Exits’ (chu jiang 出將) and ‘The Minister Enters’ (ru xiang 入相); none of the pre-revolution stages I’ve seen have such plaques, but they may once have existed.
89 Jeehee Hong has made a very similar argument about Liao-period tombs, specifically about the efficacy of three-dimensional carving to represent an intermediate or ‘third realm’ between flat and receding space: ‘These sculptural elements [in the tomb] possess an anomalous sense of depth and volume that complicates the binary spatial division in the tomb, breaking the conceptual distance between pictorial and real spaces. They are projected into the real space to an extreme degree, as if they could transcend the border between the two spaces and eventually belong to both. [...] As the counterpart of the hidden realm of the dead existing beyond the architectural surface, this tomb space is redefined as a realm still conceptually accessible to the living.’ (Jeehee Hong, Theater of the Dead, p.100).
90 The steles outside the temple describe the re-creation of the temple and opera stage from scratch with images and paintings in a new location in 1710, and then the repair of the temple walls and gates in 1785. (These are reproduced in Deng Qingping, Yu xian beiming jilu, pp.428–29, 432–23.) If we accept either of these dates as the date of the mural paintings, then this beautiful room represents by far the earliest perspectival drawings now extant in Yu county. However, not all of the steles are now legible, and there may have been later reconstructions. The dated grafitti in the opera stage adjacent (which has murals in the same hand as those of the temple) are all from the nineteenth century.
91 I have not touched on the issue of painted images of screens on the rear walls of temples and the side walls of opera stages elsewhere, but they are nevertheless an interesting subject. It should be pointed out that a screen, just like a perspectival drawing, represents a sort of trompe-l’œil —implying a hidden space behind. A screen is also a potent way to play with surface and depth. Wu Hung sums up this logic when he defines three uses of the screen: ‘The screen as a three-dimensional object that differentiates an architectural space; the screen as a two-dimensional surface for painting; and the screen as a painted image that helps construct a pictorial space and supply visual metaphors’. Wu Hung, The Double Screen: Medium and Representation in Chinese Painting (London: Reaktion Books, 1996), p.24.
92 See Cahill, Pictures for Use and Pleasure, pp.149–98 for a description of this genre and its connection to Western-inspired ideas of inner space.
93 Zhongguo shehui kexueyuan wenxue yanjiusuo guben xiqu congkan bianji weiyuanhui中國社會科學院文學研究所古本戲曲叢刊編輯委員會, ed. ‘Man chuang wu’ 滿床笏, in Guben xiqu congkan wu ji 古本戲曲叢刊五集 (Shanghai: Shanghai guji chubanshe, 1986,) Vols 6 and 7. The relevant final scene, ‘A Circle of Officials’ (Hu yuan 笏圓), begins on p.77.
94 According to elderly villagers at Song Family Fort, this plaque was destroyed around the time of the Cultural Revolution; its existence is mentioned in Yu Xian Bowuguan et al., Yu xian gu xilou, p.374. On the rear of the stage, there remains a second plaque by the same calligrapher (one Feng Guohua 馮國華), with the words ‘Repeatedly Celebrate the Harvest Plenty’ (Lü qing nian feng 屢慶年豐), referencing the votive use of the stage across from a temple to the Perfected Warrior (Zhen wu miao 真武廟). These murals themselves have an interesting history. When I first visited the village in 2014, they were not visible. Since then, the stage has been repaired, in the course of which the mud plaster that had covered the walls since the Cultural Revolution was scraped away, revealing the murals beneath. One of my volunteer documentarians (Hannah Theaker) photographed the murals on the renovated stage in January 2017. Since then, they have been vandalised with a spray-paint bottle.
95 One aspect of this image that I don’t fully understand is the fact that there were two Perfected Warriors per palace — one on each visible inner face of the courtyard, to make a total of four. Three of these survived at the time I first saw the image, and one survives now. It could be that two of these four figures represent Tianpeng 天篷 — an obscure counterpart of the Perfected Warrior. Alternatively, and this seems more likely to me, it may be that the artist simply wanted to include the maximum number of scenes and attitudes and so drew a figure on each of the visible inner porticoes.
96 Jeehee Hong, Theater of the Dead, p.130. As noted above, the type of tomb art Hong is talking about also existed in Yu county.
97 Ibid., p.117.
98 Shang Wei, ‘Bi zhen de huanxiang (xia),’ p.25. 我想强调指出的是,伴随着透视法和西方视镜而来的,并非模仿、再现的观念,也不是写实主义的固定模式,而是对视觉主观性和不确定性的探索,是对纵深空间的发现和对真假、虚实和有无的执着兴趣,本土的魅影魔幻话语也因此而被再度激活并赋予了新的意义。