Conceptual framework

Studies of late imperial Chinese history and religion over the past several decades have moved from James L. Watson’s influential conception of “standardization,” perceived as a largely top-down model that attempts to account for the high degree of social and political integration in late imperial China, to case studies of temples, cults, rituals, and the communities that support them that demonstrate the wide cultural diversity found below the surface of any real or imagined standardized unity.Footnote ¹ Taken together, such studies also reveal the great ingenuity and flexibility of various social actors or groups who navigate and negotiate among disparate levels of social, political, and religious power to assert their own agency, sometimes through, sometimes against, and, in some cases, complexly hidden within, the broader discourse of unity and orthodoxy. One of the main thrusts of these later studies has been to challenge or complicate the oft-perceived top-down dynamic of Watson’s standardization process. In doing so, some of these studies gesture towards the conceptual antithesis of orthopraxy, “heteropraxy,” typically defined as rituals, customs, and social formations that are directly at odds with the late imperial state’s accepted range of social practice.Footnote ²

This discursive move can be illuminating, but like all dichotomies, the imperial-versus-local and orthopraxy-versus-heteropraxy rubrics oversimplify and thus sometimes obscure many of the dynamic social interactions that they hope to make visible. What, for example, is the imperial state, if not itself a complex amalgam of social agents negotiating political hierarchies and local or regional cultural expectations in a careful dance between top and bottom, the local and the imperial, the canonical and the customary? Following from this, state orthopraxy can hardly be neatly juxtaposed to local heteropraxy if all actors who represent the state are themselves embedded in a web of diverse social praxis that stretches across multiple social groups, scales, and hierarchies. This approach, then, draws our attention away from black-and-white distinctions, directing it instead to the dynamic space between poles that are, in many cases at least, simply heuristic devices. The approach I pursue below attempts to explore and better understand the behavior of social agents enmeshed in complex social and cultural contexts (contexts that are themselves, in turn, nested within multilayered hierarchies) by demonstrating some of the dynamic ways representatives of the late imperial state worked within the rubric Watson calls “standardization” that are neither top-down nor bottom-up, but rather responsively fluctuate between top and bottom.Footnote ³ I will begin this examination on the decidedly small scale, on the premise that it is precisely the day-to-day decisions, actions, words, motivations, and strategies of such social actors (and others, at all levels) that recursively shape, and are simultaneously shaped by, Chinese culture, through what P. Steven Sangren calls the “mutually constitutive relationships between individuals and cultures.”Footnote ⁴

Specifically, this case study is aimed at highlighting how highly literate social actors often negotiated complex social interactions that might tend towards heteropraxy by invoking and/or interpreting orthodoxy, typically in the form of canonical texts. Those texts, however, did not float in an ideal bubble apart from the social worlds that produced and transmitted them, but rather were available to social agents only via many layers of interpretation. Thus, citing a canonical tradition in an attempt to navigate some social, political, or economic situation was always a contingent endeavor, one that provided literate social agents at least two creative opportunities: the opportunity to select and activate some existing or potential thread of canonical interpretation (thus making a contribution, big or small, to doctrine, hermeneutics, and orthodoxy); and the opportunity, through such recourse to the canon, to offer a novel, productive solution to the problem at hand (thus making a contribution to either broader social cohesion and political unity—that is, to orthopraxy—or to alternative structures of wielding power—that is, potentially at least, to heterodoxy and heteropraxy). My interest, as must be obvious by now, is in not exclusively one or the other, but rather in the way that skillful interpretation occupies, negotiates, defines, and mines the middle ground between the two.

Moreover, I am particularly interested in examining such acts of interpretation and negotiation as a complex form of self-projection or self-realization by social agents hoping to create, sustain, and/or expand their various identities (personal, professional, familial, etc.) through successful navigation of perceived social, religious, and hermeneutical pathways of influence. Success at navigating these pathways, especially those that move vertically across social boundaries, often turns on understanding how the deep ambiguities of language, cultural symbols, and social institutions contribute to the rise of divergent registers. Social actors who are adept at the arts of communicating and interpreting among various registers, for example religious leaders or successful civil service candidates, also become adept at producing and projecting visions of integration (speaking across social registers is itself an act of integration), helping contribute to the fantasy-cum-reality of the orthodox and unified empire, under which rubric so much diversity still thrives.

I will focus on a single case study—primarily a single temple inscription—to illuminate the importance of this sort of interpretation and, perhaps potentially, even a sort of virtuosic improvisation on the part of the author. The former I take to be one of the very real products of the late imperial civil service examination system and the broader range of activity centered on classical learning—the ability to interpret contemporary social and political events through a canonical lens and offer interventions. The latter I see as one of the core promises of Confucian self-cultivation (broadly speaking, at basically any time in the history of Confucianism)—that Confucian ritual, moral, and textual mastery produces an individual (ideally, the junzi 君子, or “gentleman”) capable of immediately grasping the moral, emotional, and ritual circumstances of any social or political situation and, in the flow of events, producing the precisely suitable words and actions to meet the occasion. This sort of virtuosic social and moral improvisation, while not impossible, is easier to portray in storytelling than to achieve in actual life, so we must be aware of its possible discursive production and manipulation. As will become clear, the narrative that comprises this particular inscription appears to be highly staged, and following the translation, consideration will be given to the possibility that the text existed, and circulated, in other forms.

Along the way, I hope to demonstrate the unique insights that this particular type of historical source affords us. Temple inscriptions are typically focused on or composed in response to highly specific events or needs; they are usually tied to very specific sites or localities; their authors, sponsors, and even the names of craftsmen who worked on various aspects of the physical forms they took are sometimes known or knowable to us. Moreover, their siting (at a sacred place) and authorship (educated elites) means temple inscriptions are often created at the intersection of various highly charged vectors of ambition, desire, sentiment, and self-expression. Composing and erecting an inscribed stele at a temple is a ritualized or sacred performance, a historiographical or authorial act, and an assertion of local (sometimes hermeneutical) power and authority. For these reasons and more, the authorial voice of many temple inscriptions is particularly reflective, and the content can move from mundane (but to the scholarly eye, precious) details of village life to quite lofty ruminations about the spirit world. Yet temple inscriptions have most often been employed in modern scholarship as unproblematic documentary evidence concerning the histories of temples or of local political, social, or religious developments; their discursive complexity remains largely unrecognized and thus understudied.Footnote ⁵

Finally, between text and meaning, social tension and improvised resolution, there must be ample space for the sort of interpretation noted above. Even relatively narrow definitions of “the canon” in China are still characterized by diversity and ambiguity.Footnote ⁶ In spite of regular attempts by literati, officials, religious specialists, and social agents of all sorts to clarify, specify, or prescribe both meaning and behavior, social practice, too, by virtue of its complexity, tends toward ambiguity. This ambiguity can be productive, particularly insofar as it may simultaneously allow for or elicit a broad range of understandings, behaviors, and sentiments.Footnote ⁷ In the stele inscription examined below, the author presents a scene in which the gap of ambiguity between empire and village, orthodoxy and heteropraxy, becomes a space for both a sophisticated and highly prescriptive classical erudition and, conversely, also for a deeply sympathetic and locally informed tolerance of the diversity of social and religious practice at the village level. Examining the ways that the author navigates this ambiguity invites us to consider the political, intellectual, and perhaps even moral significance of the piece (and other interventions like it), as well as something about its possible audiences.

Sacrifices to the matchmaker on high and the spatial-temporal logic of canonically sanctioned ritual

The stele inscription in question is located at a temple in Dayangquan village 大陽泉村, outside of modern-day Yangquan city 陽泉市, in Shanxi province. The inscription is dated 1782, and records a purported exchange between the inscription’s author, Zhang Peifang 張佩芳 (1732–1793, awarded jinshi degree in 1757) and an unnamed (and perhaps imaginary) member of the local, rural community. Dayangquan was Zhang’s hometown, and though he spent his entire official career serving in various posts in Anhui and what is now neighboring Jiangsu, he was home in Shanxi in the fall of 1782, and in preparation to return to his post he visited a prominent temple in the village, called the Guangyu Shrine 廣育祠, which had just been restored the previous year. These details of Zhang’s life we learn from the inscription itself. Zhang was never a widely known figure; he participated in the compilation of local gazetteers in some of the places he served (almost exclusively in modern-day Anhui province), as well as in his home county of Pingding, in Shanxi, and it has been possible to piece together details of his life and examples of his other writings from these sources, but there is nothing resembling this inscription in his other preserved works. During his career, in addition to overseeing or participating in the writing and compiling of local gazetteers, he authored a handful of other temple inscriptions and at least one essay on water control. From his surviving works as well as biographical mentions of him in gazetteers, we might describe him as primarily a historical geographer who worked at the small scale of village, town, and county.Footnote ⁸ Among scholars and officials of his era, few outside of Shanxi or Anhui would have known much, if anything, about Zhang Peifang, and the existence of the Guangyu Shrine in Dayangquan was surely a matter of no significance outside of Pingding county. And yet, in choosing to present a detailed translation and analysis of the inscription he wrote at this small temple, I am asserting its importance in presenting a vantage point from which to observe the way rural temple inscriptions can capture vivid details of the sort of social and interpretive negotiations I have discussed in more general terms above. From the late imperial era, there are, in fact, many thoughtful inscriptions from temples like this one, typically written by relatively unknown authors, scattered across Shanxi and many other parts of China. A broader survey of these would certainly be useful, but nothing is quite as powerful as a close examination of a single example.

During his visit to Guangyu Shrine, Zhang’s inscription tells us, someone asked him the identity of the spirit of the temple. Their entire exchange, as well as Zhang’s description of the temple, are illustrative of the sort of “work” that is done at temples across China, both today and in historical times, a work I have described elsewhere as “negotiation” among (to simplify) representatives of the state, educated elites, ritual or religious specialists, and locals.Footnote ⁹ The tone of the inscription, as with those studied in that previous research and many more in Shanxi, is casual and even a bit intimate. It begins directly with Zhang’s description of the temple:

The interaction thus far is telling. It exemplifies a type of highly uneven exchange that can still be seen transpiring to this very day across much of rural China. In such exchanges, a member of a rural community avails himself of an educated urban visitor’s presence to learn something authoritative about his own local culture (Zhang would surely have been easily identifiable, by his clothing, demeanor, and comportment, as a powerful and educated man). Zhang Peifang was once a resident of this village as well, but it is not clear from the inscription if his interlocutor knows this, or, in fact, if it is a salient issue. More to the point is the fact that Zhang, an educated degree-holder (returning from the outside, if not an actual outsider), can serve as a mediator for this local man, a mediator between local social and religious practice and the imperial state, or, we might say, between local memory and history writ large.Footnote ¹²

Taking the narrative for the moment as a record of an actual conversation, it is notable, and perhaps somewhat doubtful, that Zhang Peifang’s local interlocutor seems not to have known the identity of the deity to whom his village temple was dedicated. It is equally notable that he expected Zhang to know, and was apparently eager to hear, perhaps even embrace, Zhang’s response. Other than this somewhat unusual confusion, Zhang’s inscription thus far, with its attention to the small details of what was likely a fairly generic village temple, reads as a largely believable (if clearly highly narratively staged) and straightforward account of a brief and, so far, uneventful visit to his community’s temple. The temple still stands today.Footnote ¹³

Temples to Gaomei, who appears in early canonical texts as the recipient of royal prayers concerning childbirth, can be found throughout Shanxi and occasionally elsewhere in China, though not with the frequency of the most common goddesses of childbirth, e.g., Songzi Guanyin 送子觀音.Footnote ¹⁴ The spirit’s name carries the meaning of “matchmaker on high” (or, Divine Matchmaker), and insofar as the choice of a suitable marriage partner and the consummation of marriage are steps toward the most important creative act of all royal subjects, the simultaneous reproduction of the patriarchy and the populous, the real domain of Gaomei’s spiritual efficacy can be boiled down to one thing: the birth of boys.Footnote ¹⁵ The name of the temple, Guangyu Shrine, expresses the main function of the temple as a place to pray for offspring—guang meaning “widespread” and yu referring to conception, childbirth, and the rearing of children. Yet the temple’s name makes no direct reference to the identity of its main spirit.Footnote ¹⁶ To whom were the villagers of Dayangquan praying? According to the narrative Zhang presents, the villager’s question about this prompts Zhang to offer a quick history lesson, and then further to pen a short reflective piece that ends up as the inscription at the temple. This snippet of village life, as we shall see below, provides Zhang an opportunity to reach into his vast repertoire of textual expertise and produce, apparently rather effortlessly, a plausible historical narrative for his hometown temple, one with an aura of imperial sanction and venerated tradition. The inscription continues:

What a splendid story Zhang tells. If we have been inclined to accept the general outlines of his narrative to this point, his local interlocuter’s willingness and ability to ask just the right questions to allow Zhang to produce his canonically rooted critique of local custom now strains credulity. Zhang roots his excursion in a passage from the Liji 禮記, the Record of Rites, a part of the Five Classics. These works, the Book of Documents 尚書, Book of Odes 詩經, Book of Changes 易經, Spring and Autumn Annals 春秋, and a collection of ritual texts including the Rites of Zhou 周禮, and the Record of Rites, are the oldest and most revered of all classical works in the high literary tradition. The ritual classics are, among other things, the authoritative sources consulted for information on how the rulers of antiquity designed and carried out sacrifices and other activities at altars, shrines, and temples. They are frequently cited when temples or shrines become the site of contention, as canonical support for a certain position on orthodoxy versus local custom, in the very manner seen here.

Zhang’s citation of the Liji marks the beginning of a series of hermeneutical moves that will intervene in local religious and ritual practice. His very first maneuver may slip by the contemporary reader, or, once it is brought to attention, surprise by its acceptance of folk custom: it is appropriate, he says, that the temple should be dedicated to a female deity. The high literary tradition, and particularly the Five Classics, leave relatively little room for accommodation to female deities, but Zhang is well aware that folk custom across Shanxi province, and indeed probably all of China that he was familiar with, embraced shrines, temples, and worship dedicated to female figures, especially concerning matters of childbirth.Footnote ¹⁸ The prominent role of female deities in Chinese folk religion is worthy of much more attention, and I will return to this issue below, but here suffice to say that Zhang’s first intervention with his hometown temple is accommodating. He manages to find canonical authority for prayers to a female deity concerning childbirth. Let us not miss the irony that the canonically celebrated prayers, here, are explicitly made for the birth of a male heir; his interpretive move is only an accommodation, not a retreat from patriarchal and patrilineal principles.

Zhang’s next move is no surprise; after rooting his intervention in the canonical tradition, he turns to history, specifically a historical overview of imperially sanctioned spatial configurations of ritual activity at temples or shrines honoring spirits that include “ancient emperors.” Understanding this somewhat arcane section of the inscription is crucial to comprehending the power of the interpretive move Zhang will make at the conclusion of the piece. He begins his analysis with the example of the Northern Qi (550–577); as we will see below, meaningful details about sacrifices to Gaomei are scarce prior to this time. Zhang’s description of the placement of spirit images (or, more likely, spirit tablets, which were typically preferred at the elite level) evokes a set of complex canonical spatial arrangements coded in the “Yueling” chapter of the Record of Rites, which are designed to reproduce, both within the ritual calendar and within imperial architectural design, the spatial and temporal configuration of the orderly cosmos. The attention given to proper spatial arrangements, in even a small rural temple, hinges on this sense that cosmic order is mediated and reinforced even at the microcosmic scale of local temples, shrines, and altars. The relationships between the “Yueling” chapter’s textual description of a calendar and the authoritative spatial arrangement of spirit images in the dynastic histories Zhang cites are most easily grasped in visual form.

Figure 1 shows the “Yueling” calendrical and spatial configuration in radically simplified form, portraying only the season, direction (time and space) and the ancient emperor associated with that season/direction. Because the calendar, divided into four seasons of three months each, needs to be mapped onto two important systems that are based on the number five (the four directions plus a center; and the canonically important Five Phases, wuxing 五行, wood, water, metal, fire, and soil), a placeholder season of a single day is posited between summer and autumn to occupy the center. This is a nearly universal solution to numerological problems of fours and fives in early texts. Each direction is associated with one of the Five Phases, the east with wood (or more broadly “vegetation”), the south with fire, the west with metal, and the north with water; this allows the “center” of the calendar to be associated with soil, in effect placing the court, or the empire, or the reader of the text, at the conceptual center of the system, on the very ground below us. This is crucial. The insistence on a center is not a mere convenience employed to solve the disparities in a system based both on the symbolism of four directions/four seasons and on the logic of the Five Phases and their many quintuple associations (the five internal organs, the five flavors, the five planets, etc.). Rather, construction and legitimation of a ritual center is the very goal of the system itself. That center is occupied by the ruler, the reader of the classic, the supplicant at a temple—it is the center of agency in any architectural, textual, political, or ritual setting governed by the spatial/temporal logic of the system.Footnote ¹⁹ Notice, however, that in his role as scholar, Zhang does not occupy the symbolic center of the system but rather, from the outside, interprets the history and dynamics of the system, thereby creating a sort of displacement of agency from the imperial center to the interpretive voice that simultaneously bolsters the system and the power at its center but also attenuates, critiques, and potentially even resists it.Footnote ²⁰ This logic is crucial to understanding how scholars, ritual specialists, geomancers, etc., stake claims to their mediating positions in society, and plays subtly into the way Zhang Peifang arrives at his conclusion.

Figure 1.

Spatial configuration of the seasons, Five Phases, Five Emperors, and the directions according to the “Yueling” chapter of the Record of Rites.

The arrows in the figure show the movement of time through the seasons, starting with Spring in the east, then moving to Summer, then to the center and then back out to the west for Autumn before completing the cycle with Winter in the north and starting over again. Understanding the precise arrangement of ancient emperors within this system is a complex matter, in large part because early textual traditions that name or discuss the “Five Emperors” do not always agree on which five emperors are to be accorded the honor of standing in each of the five ritually important positions portrayed in the diagram. In any case, given the names presented to us in the “Yueling,” we can discern some logical associations running through the system: the association between the east (where the sun’s daily cycle begins), spring (when the year’s cycle of life begins), and vegetation is intuitive, and makes the appearance of the Green Emperor (Qingdi 青帝) and Fuxi 伏羲 (often treated as the first of the many ancient sage rulers) in this position sensible. As we move to Figure 2 below, we can see the same logic determining the placement of Gaomei’s spirit image on the eastern side of the main altar; her association with childbirth is analogous to spring’s association with the start of the seasonal cycle of life.

Figure 2.

Reconstructed diagram of Zhang Peifang’s imagined canonical temple to Gaomei.

Beginning with the first month of spring, the “Yueling” lays out a long list of calendrical/celestial coordinates correlated with various natural and ritual phenomena, including such details as the portion of the sacrificial animal that is given precedence during offerings in a certain month and the physical place within a building where that sacrifice is to be performed. It also names the first of the five emperors, Taihao 太昊, the royal title of Fuxi. The text continues on to name the remaining ancient rulers, Yandi 炎帝 in the south/Summer, then the Yellow Emperor 黃帝, Shao Hao 少昊, and finally Zhuan Xu 顓頊, associated with the center, the west and thus Autumn, and finally the north/Winter, respectively.Footnote ²¹ Unlike many other early texts that mention five emperors in systems of elaborate correspondence, the “Yueling” never explicitly names these five ancient rulers, Taihao (Fuxi), Yandi, etc., as the “Five Emperors” wudi 五帝, but testament of both later commentaries and ritual practice, including of course the brief history Zhang gives in this description, demonstrates that this is in fact the stable and canonically most authoritative list of the Five Emperors.Footnote ²²

The canonical ritual system described in Figure 1 remained a productive force in guiding ideas about ritual and sacred space throughout the imperial period, and could be invoked or implicated when questions about or challenges to accepted ritual practice emerged. A closer look at imperial rites offered to Gaomei shows that this figure in particular was often at the center of ritual debates, clearly in part, at least, due to her gender. Qin Huitian 秦蕙田 (1702–1764), in his monumental Wuli tongkao 五禮通考 (Comprehensive Examination of the Five Categories of Rites), presents a detailed account of the textual evidence concerning rites to Gaomei from the reign of Emperor Wu of the Han down to the Qing. That evidence, as well as the analysis Qin provides, makes clear that there was never a stable, regularly implemented protocol for rites to Gaomei, and that whenever rites were revived, they proved “difficult.”Footnote ²³ Aside from the claim made in the “Yueling” that ancient rulers performed these rites annually, the first historically documented imperial rites to Gaomei were conducted under Emperor Wu of the Han, clearly on the inspiration of ritual specialists versed in (and promoting the status of?) the “Yueling.” Records of rites to Gaomei after that and down to the Tang are rare, and demonstrate a considerable lack of clarity concerning details of the rites in the courts that did perform them. Qin, citing the Sui shu 隋書, notes that rites to Gaomei only begin to gain stability starting in the Northern Qi (550–577)—thus Zhang Peifang’s citation of this period. Qin described those rites in this way:

The description ends with Qin’s own observation: “the details of rites offered to Gaomei before the Qi dynasty are unclear. It would appear that worship of Qingdi was established at this time” (齊以前祀高禖之禮不詳至是乃祀青帝似爲得之).Footnote ²⁵ Rites to Gaomei were only offered sporadically over the coming centuries. It seems that imperial rites offered to a female deity remained a problematic proposition, especially when those rites might include female supplicants as well, as described above.

The continued instability of and apparent anxiety over rites offered to Gaomei is demonstrated by a record of the relocation of the imperial altar to Gaomei in the Southern Song dynasty preserved in a gazetteer of the time. The document notes that sacrifices to Gaomei had for some time been offered not at a dedicated site but rather by utilizing space in another existing temple, but that, in 1142, the erudite Liu Rong 劉爃 requested an imperial investigation into the proper location for a permanent altar. After deliberations, a site was chosen in the southeastern suburbs of the capital, but in 1146 the Department of Rites reopened the matter for debate, and they eventually moved the altar to the east to better match understandings of past precedent. The text then tells us:

This matter is clearly the source of Zhang Peifang’s note that in the Song dynasty, the spirit image of Di Ku (that is, Gaoxin) was added to the altar next to that of Fuxi and above the spirit image of the Matchmaker on High (see Figure 2). Di Ku is almost universally accepted as the ancient ruler following after Zhuan Xu (even in Sima Qian’s quite different list of early rulers), so his placement on the east side of the main altar, alongside of Fuxi, is logical: the cycle of five starts again, with the next earthly emperor, Di Ku, in the same position that Fuxi occupies, but seated to his left (to the south of Fuxi; the north being the position of priority). As noted, the placement of Gaomei’s spirit image or tablet on the east wall of the temple, associated with spring, is not arbitrary; her role as a spirit of matchmaking and marriage aligns her with childbirth and creation, and thus the beginning of the life cycle. Her placement below the altar subordinates her and her creative powers to the male dynastic imperial line, represented by Fuxi.

Yet the record of events in the Southern Song dynasty complicates our understanding of the altar to Gaomei and its symbolism by naming two historical or legendary figures whose spirit tablets occupy the space that Zhang Peifang makes clear is the ritual position of the Matchmaker on High. Moreover, the record never specifically denotes the placement of a spirit image or spirit tablet to Gaomei herself. Was she replaced in this instance by Jiandi and Jiang Yuan, or were they simply placed next to her? These two figures have significant, if somewhat contentious, canonical status themselves. The ode “Xuanniao” 玄鳥 (Mao no. 303) of the Book of Odes, one of the Five Classics, describes the birth of Xie, the first ancestor of the Shang ruling line, after Heaven ordered a “dark bird,” the swallow, to descend and give birth to the House of Shang. Ancient tradition is unanimous in naming his mother Jiandi, the first appearance of the name coming in a line from the “Tianwen” poem in the Chuci. Footnote ²⁷ Sima Qian’s Shiji retells the story of this ode in detail, also naming Xie’s mother as Jiandi. Jiang Yuan is celebrated, by name, in the ode “Shengmin” 生民 (Mao no. 245) of the Book of Odes, as the mother of Houji, first ancestor of the Zhou ruling house. Both Xie and Houji are described in the Book of Odes as products of miraculous birth; if their fathers had been mere mortals, they could not themselves be founding ancestors (that is, their fathers would be founding ancestors). While in the mythic telling of their stories in the Odes, their mothers can be known but not their fathers, in the Shiji, a text dominated by the logic of both historical chronology and patriliny, their fathers must be named, as part of Sima Qian’s grand construction of the patrilineal line of all early sage rulers and dynastic founders. He makes clear that in fact both Jiandi and Jiang Yuan were wives of Di Ku, a position that is supported by many other early classical texts but never definitively settled in the commentarial tradition. There is a tension, then, in these earliest tellings and retellings of Jiandi and Jiang Yuan’s stories; the earliest mentions of these figures come within the high canon itself, the Odes, but introduce a theme that flouts orthodox patrilineal and patriarchal norms. This tension is never fully resolved in the commentarial tradition, and thus makes the appearance of these figures on the altar described in the passage above intriguing.Footnote ²⁸

It seems likely that the impetus to add Di Ku’s spirit tablet to the eastern side of the altar to Gaomei, next to Fuxi, was precisely to then allow the inclusion of his two wives, both of whom gave birth to sons who founded dynasties.Footnote ²⁹ If Jiandi and Jiang Yuan are now assumed to have made sacrificial offerings to Gaomei, with such success, then their inclusion at the altar bolsters the legitimacy of these rites. Given that Qingdi already occupies the preeminent placement facing south, and Fuxi is already seated on the eastern side of the altar as accompanying spirit, what other purpose can we imagine to seat yet another (redundant) accompanying ruler-spirit, Di Ku, at the altar? The original Han-era insertion of Jiandi and Jiang Yuan into the narrative of rites to Gaomei was itself surely an interpretive move; Qin Huitian’s study of rites to Gaomei offers us his analysis of the longstanding interpretive disagreement centered on Jiandi, Jiang Yuan, and the spirit Gaomei. Stated simply, the implication of “miraculous (fatherless) birth” for the founders of the Shang and Zhou ruling lines apparent in the Book of Odes did not sit well with many later scholars; it was likely redolent of the fantastical elements of the apocryphal chenwei 讖緯 texts that flourished in the Eastern Han. By asserting that Di Ku was the true father of Jiandi and Jiang Yuan’s sons, the mysterious spiritual interventions implied in the odes were now simply the interventions of proper sacrifices to the Divine Matchmaker to ensure the birth of a son. To a modern eye, this interpretation does nothing to demythologize the stories of the births of Xie and Houji, but that was not the real point. Rather, this interpretation replaces the mystery of divine birth with a staunch defense of the patrilineal order, at the same time projecting the fantasy of a single line of legitimate royal power extending back through the Zhou and Shang and all the way to the time of the Five Emperors. Placing Jiandi and Jiang Yuan’s spirit tablets below that of Di Ku at the altar to Gaomei bolstered the legitimacy of the rites by stressing their deep historical links to the founding of two glorious and long-lasting ruling houses while also reinforcing the patrilineal symbolism of a rite otherwise offered to a female figure (by women in the palace, no less). Jiandi and Jiang Yuan appear occasionally in later debates about proper offerings to Gaomei, but never become a stable part of the rites.Footnote ³⁰ Gaomei’s role in debates about the role of miraculous birth and the status of mothers of emperors would have been known to Zhang Peifang, and so these interpretive tensions would have added to the symbolic gendered context of Zhang’s inscription, a topic I return to below.

If, at the Guangyu Shrine in Zhang Peifang’s hometown, there had been only the single figure of Gaomei, seated on the back wall and facing south, perhaps Zhang would not have felt compelled to offer such a complex historical analysis of the canonically correct spatial configurations of spirit images for worship and sacrifice. That there was a pair of spirit images here, as we will see, opens up an additional layer of questions about this temple. Notice that, in answer to his local interlocuter’s first question, Zhang offers the idea that the main spirit of the temple is Gaomei, but he does not offer an explicit opinion on the identity of the male figure. Zhang likely understood that, at this temple in his own hometown, most visitors prayed to the goddess figure, not to whoever the male figure was meant to represent. In his initial description of the appearance of the two figures, we can sense a hesitancy to make any claims about the pair; he can only say that they have the appearance of ancient emperors (see Figure 3).

Figure 3.

Reconstruction of the actual placement of spirit images in the Guangyu Shrine according to Zhang Peifang.

Interpretive inversion and tolerance for local custom

Zhang’s hesitancy to name the male figure seems, ultimately, rooted in his discomfort at seeing the male and female figures seated next to each other on the main altar, a placement that in the next few lines of the inscription we see breaks all rules—we may rightly call this ritual configuration heterodox. Zhang’s unwillingness to settle definitively the issue of the male spirit image stems in part from the ambiguity of the early textual record itself. Textual evidence from the Western Han on describes a Qin tradition of regular sacrifices at altars near the capital to Qingdi, Baidi 白帝 (the White Emperor), Huangdi (the Yellow Emperor) and Chidi/Yandi 赤帝/炎帝 (the Red Emperor/ the Fire Emperor—the two names seem to be used interchangeably in the Shiji but we do not know if that was already the case in the Qin), with the Han founder adding sacrifices to Heidi 黑帝 (the Black Emperor) and thereby completing the paring of gods of the four directions and the center with the Five Phases. Sacrifices to these deities were adopted, though apparently never become completely stable, during the Western Han, but from these earliest mentions to later periods, there is not complete clarity about the relationship between the five deities named for their colors and the Five Emperors, e.g., Fuxi, Yan Di, etc.Footnote ³¹ They are sometimes treated as separate entities, sometimes as alternative names for the same figures. Zhang himself refers to Qingdi as a celestial emperor, and to Fuxi, and by extension, Di Ku also, as earthly emperors; and this solution to the problem of the ambiguous Han sources seems common, if imperfect. This solution adds a layer of explicit articulation between the historical and the cosmological, the human and the divine, emphasizing the legitimacy of the cycle of rulers across time and also their ties to heavenly authority. It also enables the dynamism needed to accommodate the shifting needs of ritual performance to the canonically fixed spatial associations; Qingdi, hierarchically the highest figure represented in the temple Zhang envisions through his reading of ancient sources, is given the most powerful position, facing the south (source of yang), while Fuxi and Gaomei can occupy the symbolically important position of the east, representing birth and the beginning of the cycle of life. But the lack of clarity on the Five Emperors in ancient sources does not fully explain Zhang’s reluctance to articulate clearly the identities of both spirit images in the temple. Zhang’s reference to historical precedents that place Fuxi together with Gaomei on the east side of the main altar makes fairly clear that, from Zhang’s own frame of reference, the two spirit images in the Guangyu temple should be understood as Gaomei and Fuxi. Openly naming Fuxi, however, might work against Zhang’s ultimate interpretive goals, since this figure is not regularly associated with Gaomei. This possibility can better be explored after addressing the rest of Zhang’s inscription.

We see that the ritual system correlating time/the seasons and space/geography was itself a dynamic model that allowed a certain flexibility. This flexibility, in turn, entailed ambiguity, allowing continuing opportunities for interpretation, innovation, and even improvisation. People such as Zhang Peifang could intervene in the operation of the system, and in fact were a necessary part of its functioning. Some, like Zhang, were agents of the state, negotiating between local practice and state orthodoxy, but the inscription mentions others who managed to find a place in the operation of the system: local figures Huo Mingxiang and Liu Rui, he notes, “headed up a group of ten people, along with the Daoist priest Xi Zhiyong” to spearhead the temple’s renovation a year earlier. In the actual maintenance of temples and management of activities there, through layers of local elites, religious leaders, and men like Zhang Peifang, village temples were—if only ever partially or just potentially—both theoretically and practically nested into the formal hierarchies and symbols of power, authority, cosmology, and orthodoxy. And here, in the interactions among such people and others (villagers, merchants, healers, etc.), the grand ritual system of canonically bolstered imperial power provided ample space for the exercise of agency (interpretation, innovation, and improvisation). This agency becomes conspicuously visible in the closing lines of the inscription, as immediately after critiquing the placement of the spirit images in the temple, Zhang surprisingly turns his own canonical critique of local practice completely on its head:

Here, in the remarkable closing lines of the inscription, is the true rhetorical point of the piece, a powerful inversion of the top-down logic of both imperial rule and canonical authority. Do we necessarily judge a system (an institution, a political arrangement, a solution to a social problem) by its conformity to orthodoxy? Or, rather, do we judge a system by its efficacy? The former is prescriptive, and may be taken as representative of the central state and its elite officers; the latter is descriptive, and may be taken as representative of the local and vernacular. But of course, once again, our dichotomies only show us the two poles, and we must turn our attention to the real social and political activity that happens in between. Local officials such as Zhang Peifang, working near the bottom of the bureaucracy, necessarily work at the intersection of the imperial state and local society. Zhang’s inscription begins in the prescriptive canonical mode, but in the end moves decisively to the descriptive, and to the middle ground, the local, and the ambiguous. By not condemning the local practices at his hometown temple, he is creating a discursive space for tolerance and mediation. In choosing historical examples carefully, he points generally to local folk traditions that earned imperial attention and, more specifically in the first example, to prayers offered to female folk deities for the birth of sons. Because of his position within the state, and because of the erudite form his essay takes, this discursive space, marked by both prescriptive canonical example and these apt descriptive historical precedents,Footnote ³³ has the potential to create opportunities for actual tolerant policies and relatively forgiving mediation practices.

This last claim, of a move from the discursive to the practical (that is, in terms of the opportunities noted above, the move from interpretation to problem-solving), requires more exploration. How might the language of an inscribed stele in a remote, rural village temple have any sort of real-world impact on policy or practice? There are at least two ways we need to explore: at the site itself, and in a readership that somehow extends beyond the site. The inscription presented above, which reads very much like an essay, is not at all an isolated or unusual example. In fact, we have a particularly rich record of inscriptions from Shanxi province, many of which appear to be composed with a complex array of potential readers in mind. Zhang Peifang is certainly writing, at least in part, with the sponsors of this inscription in mind: Huo Mingxiang, Liu Rui, Xi Zhiyong, and the temple managers. But to the extent that these specific figures represent types of readers, it seems likely that the first two stand for local, prominent families, and the last stands for local religious leaders (in this case, apparently also from a prominent local family); presumably, there would be other literate and interested people like these frequenting the temple, especially over time. Visits from another potential group of readers—local officials—might seem to modern scholars unlikely, but the record in Shanxi shows us otherwise. Temple festivals of various sorts in rural Shanxi during the Ming and Qing era were frequent, well-funded events, often drawing commoners, merchants, scholars, and low-ranking officials from outside the immediate locale.Footnote ³⁴ Even more fascinating is the demonstrable fact that local officials often authored inscriptions for steles at temples that were in conversation with previous inscriptions, and that certainly anticipated future scholar-official visitors and/or readers.Footnote ³⁵ As we will see below, a later inscription at this temple engages in the same discourse of canonical interpretation intended to resolve tensions between local practice and imperial orthodoxy. In such ways, we can discern how authors of local temple inscriptions such as Zhang Peifang would have been aware of addressing a socially and temporally broad audience.

In addition to a sense of one’s audience as visitors to the temple, authors such as Zhang would also have been able to imagine their inscriptions circulating and surviving through other means. Inscriptions at local temples were sometimes included in sections of local gazetteers devoted to temples or to local writings. It can be demonstrated in some cases that the texts of inscriptions were recorded into gazetteers not by repeated visits to temples, where, theoretically, it would have been possible to copy out the content of local inscriptions, but rather by reference to copies of these inscriptions held in official collections in local county administrative offices or perhaps the homes of prominent residents.Footnote ³⁶ When Zhang Peifang authored his elaborate inscription for the Guangyu Shrine in his hometown, he could imagine local prominent residents, religious figures, and visitors from beyond his village reading and making reference to the inscription in the temple, as well as readers of local documents and gazetteers in the county seat and beyond encountering and being inspired by its erudition and, perhaps, even the implications of its surprising conclusion.Footnote ³⁷

Those implications, as noted above, include canonical and historical license for a local inversion of the symbolic logic of imperial orthopraxy. Notably, this inversion is not intended or portrayed as revolutionary or even disruptive of social and political order, local or imperial. Rather, the inversion is presented as a benign accommodation to the local and customary. The beneficent, canonically informed imperial gaze lights upon local heteropraxy, and in its ordering and harmonizing power and wisdom, subsumes it, overwriting local practice with canonical precedent; the local female deity of fertility is transformed by Zhang Peifang into Gaomei, and by naming and contextualizing her, she is tamed. Zhang’s model of negotiation and accommodation is all the more striking when we keep in mind that it was by no means the only option for the official imperial representative encountering local social and religious diversity. At the other extreme from Zhang’s tolerance we find examples of high literary elite or imperial officials ordering the destruction of local temples and the suppression of local ritual practice.Footnote ³⁸

If Zhang was writing in part with later residents of the village and visitors to the temple in mind, there is some evidence that this aspiration was realized. Another inscribed stele at the Guangyu Shrine in Dayangquan village, erected in 1826 to commemorate the restoration of the opera stage in front of the temple, forty-four years after Zhang’s inscription, echoes Zhang’s tone of tolerance for local custom. The inscription was authored by a local literatus named Li Rimao 李日茂, who is listed in the Pingding Prefecture Gazetteer of 1883 (7.26b) as having passed the Shanxi provincial exam of 1835, nine years after he authored this inscription. The calligrapher of that inscription is listed as Zhang Jinxian 張晋暹 (1794–1830). Two years later, in 1837, Li Rimao—now awaiting assignment as a county magistrate—authored an epitaph for this Zhang Jinxian, who we learn from Li’s text was the grandson of Zhang Peifang and a classmate and lifelong friend of the epitaph’s author.Footnote ³⁹ Zhang Jinxian’s younger brother, Zhang Mu (1805–1849), who later in life became a fairly well-known intellectual himself, would have been twenty-one years old in the year this inscription was erected, but he is not mentioned. The list of participants in the creation of the inscribed stele includes another man surnamed Zhang on the list of temple managers, but it is unclear if he is from the same family.

Li Rimao’s inscription at the Guangyu Shrine relies heavily on passages from, or references to, the ritual classics, the Rites of Zhou and the Record of Rites, with the clear goal of demonstrating the commensurability of village customs and imperial rituals. The inscription begins with a paraphrase of a line appearing in Zhang’s inscription:Footnote ⁴⁰

The inscription continues to develop a series of associations and parallels—father/mother, Heaven/earth, imperial court/local community, beneficence/recompense—before invoking the ritual classics’ description of ancient ritual drumming and the performance of ritual hymns at state sacrifices. Then the author turns his attention, for the first time, to his own village:

Li’s approach here, suggesting both an organic historical link and a structural homology between the highest level of imperial ritual and the lowest level of community sacrifice, may look relatively innocuous, but its full implications are nearly as bold as Zhang Peifang’s inversion of canonical standards. What is really at stake in Li’s inscription is the permissibility of local opera performances and the various other associated ritual events that were ubiquitous in rural Shanxi communal religion. Li’s inscription, like Zhang’s before it, protects village custom and ritual performance by constructing a canonical defense for it, in this case an origin story. And like Zhang’s inscription, Li’s also offers a vision of full cultural integration, top to bottom, from imperial court all the way down to village she-altar 社. No matter how attractive this latter vision was, the notion that local opera was organically integrated with the high literary canon and imperially sanctioned ritual was highly controversial. To get a sense of what was happening at temples such as the Guangyu Shrine in Shanxi villages during the late imperial era, one should begin with David Johnson’s ambitious study of what he terms “ritual autarky” in Shanxi and Hebei. Local opera performance, in particular, was the target of official approbation throughout the late imperial era, condemned for its vulgarity. Johnson describes in great detail what we know about local opera and related rituals in Shanxi, which in some cases involved the acting out of ritual killings of scapegoats in costume, who were chased through the village by a crowd and then, theatrically, beheaded.Footnote ⁴³ While not all local opera was so violent, the genre was largely held in low regard by the educated elite.

The same Pingding Prefecture Gazetteer of 1771 that preserves much of our knowledge about Zhang Peifang describes these operas and the festivals they are a part of, typically termed sai 賽 or saishe 賽社 (sacrificial festival honoring the gods at the level of she-altar and its local community) in great detail, with supplemental examples included:

While Li Rimao’s inscription is certainly not the only voice of tolerance for local festivals and operas, among the educated elite reluctant acceptance or even open rejection of local customs, as seen in this passage, were both far more common than Li’s rigorous and erudite defense. Compare the following contemporaneous passage from the Xiaoyi County Gazetteer of 1770; Xiaoyi is also located in Shanxi, north of Linfen and roughly 200 kilometers southwest of Pingding. After presenting an overview of such orthodox annual ritual customs as the Lantern festival and Tomb Sweeping festival, the passage closes with these observations:

One suspects that at the local level, scholars such as Li Rimao and Zhang Peifang needed to be adept at navigating at least two very different registers, one evidenced in the tone of local gazetteers, and another attuned to customs quite prevalent at temples and shrines in villages across Shanxi and, in fact, much of China in the late imperial period.Footnote ⁴⁶

Gender in Chinese popular religion

The Pingding Prefecture Gazetteer passage above provides us not only a sense of what local festivals in Shanxi were like, but also a glimpse into the sorts of complaints the educated elite leveled at these events. The opening comments about the ubiquity and frequency of saishe should be understood as expressing shock and displeasure; similar mentions in other gazetteers make explicit the critique that the extravagance and frequency of these festivals exhausted the resources of villages and placed unnecessary burdens on village families.Footnote ⁴⁷ Following this, all subsequent critiques in the passage are highly gendered. The authors note the high number of female deities in these temples. This might seem an unremarkable observation, but there is no doubt from the tone here that the authors find this fact unpleasant. The fact remains true to this day: in Shanxi province, and throughout much of China, goddesses very often constitute the cultic center of popular temples.Footnote ⁴⁸ One can easily find temples dedicated to culturally important female deities in China, Guanyin being the most obvious. But we must posit an important distinction among types of temples if we are to better understand both the gender tensions visible in the materials discussed above and, it will turn out, the full discursive range of the two temple inscriptions under investigation here. The authors of the Pingding Prefecture Gazetteer passage above are concerned only with temples where saishe festivals and ritual operas were held. It is an almost universal rule across China that neither Buddhist temples nor Daoist institutions contained opera stages, and standard Buddhist and Daoist rituals do not include opera performances presented as offerings to the spirits.Footnote ⁴⁹ This latter point is unmistakably clear and needs to be kept in mind: opera performances at local temples throughout China were presented as offerings of thanks to the spirits, to be enjoyed by their human audience, but not nominally directed at them; this cultural logic of ritual performances as expressions of gratitude is common to state-sponsored seasonal rites and religious folk custom. While both Daoist and Buddhist imagery and language often mingle with popular belief at the village level in China, it would be possible to posit this distinction as one part of a definition of Chinese folk religion: temples with opera stages (or where temporary opera stages were seasonally constructed at festival time), upon which operas were performed and presented as offerings to the spirits, often as part of a larger temple festival including welcoming, feasting, and otherwise entertaining the spirits, are as a general rule neither Daoist nor Buddhist sites, but rather the institutional sites of Chinese folk religion.Footnote ⁵⁰ The rituals and other festival activities at these temples were largely viewed as vulgar not only by the educated elite, but often also by members of both the Buddhist and Daoist clergy. One additional set of proposals is worth considering: that worship at these local temples was far more likely to be openly centered on female deities than worship at either Daoist or Buddhist sites; and that ritual and managerial activities at such sites might have been far more open to participation by women than was common at official Daoist or Buddhist sites.Footnote ⁵¹

This gendered nature of Chinese folk or popular religion seems to have been woven deeply into elite uneasiness with local ritual and custom. The matter is explicit in the Pingding Prefecture Gazetteer passage above—female deities fill the local temples; women are active participants in the boisterous and questionable activities at temple festivals (e.g., conveying prayers to the spirit world through the burning of written wishes); and spectators at these rural events were of both genders, meaning men and women would have mingled freely in their crowds.Footnote ⁵² But the question of the appropriateness of any female presence in the temple and its activities haunts both inscriptions at the Guangyu Shrine as well. Returning to Zhang Peifang’s inscription, we see that he obscures the question by dismissing it early on—since canonical records from the early dynasties show that seasonal offerings were made to Gaomei, the Matchmaker on High, presumed to be a female spirit, he is able to set the issue aside. But notice that the narrative force of Zhang’s entire inscription relies on an ambiguity concerning the true identity of the female spirit image in the temple. The inscription begins with a local interlocutor inquiring about the spirit’s identity, and closes with Zhang’s admission that “even if it is said that this is not actually Gaomei,” the temple has nevertheless proven efficacious.

The narrative reversal in the closing lines of Zhang’s inscription is both deft and surprising, opening, as discussed above, an interesting space for acceptance of and, most importantly, imaginative negotiation with local beliefs and ritual practices. But it likely relies on an obfuscation. Were the spirit images in this temple really unnamed? Perceptive readers might restate the village member’s original question, and ask, if this is not Gaomei, then who is it? Recall Zhang’s description of the pair of spirit images that “resembled ancient emperors,” and the discussion of ritual seating rules that refer to husband and wife. Zhang himself leads us to the notion that the male figure ought to be Fuxi, but stops short of drawing the relatively obvious conclusion that the female figure may then, very likely, represent Nü Wa, who in early art, in folklore, and in some widely consulted texts is often presented as Fuxi’s wife and, sometimes, the ruler who succeeds Fuxi to the throne in the deep mythic past. She is of course also celebrated as a creator of humankind and a restorer of cosmic order. Temples to either Fuxi or Nü Wa, and often to both, are common in Shanxi and much of north China, and wherever she is found in China, north or south, Nü Wa is very often the recipient of prayers and sacrifices concerning childbirth.Footnote ⁵³

If the pair of spirit images in the Guangyu Shrine in Dayangquan village were originally meant to represent Fuxi and Nü Wa, why would Zhang Peifang want to hide the female spirit image’s identity by reinterpreting it as Gaomei? There is, I believe, an answer to this question, and perhaps surprisingly it seems to have less to do with the challenge to patriarchy and patriliny posed by some early configurations of the list of the ancient Three Sovereigns that include her as a sage emperor than with her exalted status within popular culture (with all its excesses).Footnote ⁵⁴ At some level, of course, the distinctions between these two explanations converge on and dissolve into her gender. Nü Wa the Empress (often explicitly so named, Wahuang 媧皇) is mentioned only in texts from the Han dynasty and later, and never with much frequency; Nü Wa the goddess of fertility is known to us, if at all, primarily from folk temples and folk traditions. Neither appears in the canon proper. Neither has the weight of canonical authority.Footnote ⁵⁵ Perhaps Zhang Peifang’s motivation for assigning the identity of the temple goddess to Gaomei was just that simple—provide the local fertility goddess with the best canonical explanation possible. But why, then, does he shy away from explicitly naming the male figure as Fuxi, if not to avoid the easy conclusion that the female figure was Nü Wa, not Gaomei? Regardless of Zhang’s motivations, which may never be fully clear to us, we must note in closing that his transformation of Nü Wa into Gaomei—if that is really what he has done here—turns out not to be an original interpretive move.

Nü Wa’s association with matchmaking is first visible to us in the work Lushi 路史 by Luo Bi 羅泌 (1131–1189?), a text well-known for its inclusion of unorthodox, arguably even fantastical information about ancient history. The massive text totals forty-seven juan, divided into five main sections. The first of these, “Qianji” 前紀 (Earlier Chronology) covers a time in the murky past before the Five Emperors, which then figure in the second section, “Houji” 後記 (Later Chronology), where we see the canonical list matching the figures of the “Yueling” followed by Di Ku, and then the sage rulers Yao and Shun, closing with Yu the Great and the dynasty he is purported to have established, the Xia. In presenting these figures, Luo’s text includes numerous otherwise unheard of or rarely mentioned “historical” figures and stories. The fact that such an unusual text survived and was widely read in the centuries after its appearance is telling, and suggests, I would argue, that it taps into a deep well of popular narratives usually neglected by the high literary tradition.Footnote ⁵⁶ Luo’s coverage of Fuxi, at the opening of the “Later Chronology,” provides rich material about Nü Wa, including this passage:

Despite the fact that no other reliable text from antiquity records this tradition, by the Qing dynasty no less an authoritative source than the Kangxi Dictionary cites the Lushi passage concerning Nü Wa in its definition of the term mei—matchmaker. Perhaps the original author of this story was engaged in a sort of textual “standardization,” giving Nü Wa, a goddess figure with a rich narrative tradition and clear popular support, a link to canonical status. Zhang Peifang, then, performs a sort of reversal of this strategy: he disguises the popular goddess Nü Wa in his hometown temple as the canonically acceptable Gaomei.

It was common practice throughout most of late imperial China for inscriptions such as these two at the Guangyu Shrine in Dayangquan village to include information concerning the erecting of the finished steles, a sort of unveiling ceremony, usually including a list of important participants. Both of these inscriptions indeed include such information, information that suggests not only that Zhang Peifang’s family was involved in the management of the temple (his father, Zhang Keju, is named), but also that local women played some prominent role in its regular functions. Immediately following the closing lines of the narrative portion of Zhang’s inscription translated above, the text goes on to provide the time of the unveiling ceremony (in the ninth month of 1782), and to list the names of the participants, grouped into categories:

The only straightforward way that Xi Zhiyong could have grandchildren and great-grandchildren of different surnames (Yang, Yao, and Li) is, of course, if they are related to him through his daughters or, in the final case, his grandchild. This list of ritual specialists participating in the ceremonies for erecting the stele with Zhang’s inscription at the temple is in part interesting to us in this regard: those studying under Xi include three members of two later generations of his own family. Moreover, the list of temple managers confirms our sense that leadership of community worship in Dayangquan in the eighteenth century involved a handful of closely knit powerful local families: there are three members of temple management who share the surname Xi, likely all relatives of Xi Zhiyong; and two other surnames of temple managers, Yao and Li, match descendants of Xi who are his disciples or acolytes. This means, for example, that temple manager Yao Chong’an could have been the father of the disciple Yao Jiayin. Certainly not every repeated surname in such a list had to be related, but in a small village such as this, and especially when encountering less-common family names such as Xi, it seems likely that they were. Far more interesting, however, are the names themselves: based on conventional gender associations with names, at least three of the named disciples of Xi Zhiyong are almost certain to have been women or girls: Xue Xinmei, Lü Xinqin, and Li Xiang’ai. The implications for our understanding of the role of women in religious activities at the Guangyu Shrine is clear enough.

Those implications are confirmed by the corresponding list of participants in erecting the second inscription discussed above, authored by Li Rimao in 1826. In this case, the ceremonies are presided over by a Shi Xinmei 史信枚, joined by disciples Wang Jiahuan 王嘉宦, Li Xiangfeng 李祥風, and Wang Xiangming 王祥鳴. Recall that in a previous generation, Xi Zhiyong had women as acolytes. In this case, the officiant at the ceremonies, Shi Xinmei, appears by the name to herself be a woman. The list of temple managers is again ten names long, suggesting that the leadership committee was regularly comprised of ten people. And in this case, the first name mentioned, Wang Ying 王瑛, also appears to be a woman. Finally, as further attestation to the relationships among powerful village families and formalized community religious practices, we cannot fail to note the clear naming conventions apparent to us in the lists of disciples: two of the five personal names of disciples in the first inscription begin with xin 信 (trustworthy), Xue Xinmei and Lü Xinqin; two more begin with jia 嘉 (beautiful, fine); and one (likely the youngest and certainly, generationally speaking, of the youngest generation) begins with xiang 祥 (auspicious). The personal name of the officiant in the second inscription also begins with xin, and of the three named disciples, the personal name of one also begins with jia and those of the other two also begin with xiang. Perhaps relevant here is that the personal name of Shi Xinmei is a perfect homophone of one of the disciples in the first inscription, Xue Xinmei. None of these personal names, Xin, Jia, or Xiang, appear anywhere else among names of temple leaders or other community members listed in either inscription.

The claim by authors of the Pingding Prefecture Gazetteer of 1771, cited above, that women play a prominent role in the popular religious activities at village temples in the region is confirmed by the inscriptions at the Guangyu Shrine in Dayangquan. It also appears that a small group of families in the village formed the dedicated core of temple management and ritual activities there, even cooperating in the naming or ritual renaming of their children. These features of village life are visible to us across the forty-four year period separating these two inscriptions, from 1782 to 1826; it is possible that these naming conventions were generational, such that the common name of disciples in 1782, Xin, is now part of the name of the main officiant of rites in 1826, Shi Xinmei. We see here evidence of institutionalized features of community worship, including official roles for women. Zhang Peifang must have known all this, as his father was one of the temple managers. If Zhang Peifang authored his inscription with, among other motivations, the goal of disguising Nü Wa as Gaomei, then he may have been consciously protecting the formal features of community worship, including the roles of women at his local temple, with the canonical cover Gaomei could provide.

The unexpected inversion in the closing lines of Zhang Peifang’s inscription at the Guangyu Shrine in Dayangquan village is the author at his most masterful. He dances adroitly through a rich body of textual evidence toward orthodoxy, only to pirouette in his closing lines and celebrate the productive power of the local and the mysterious. His erudite transformation of a supposedly unnamed temple statue into the canonically acceptable Matchmaker on High, Gaomei, may even have been presented with a nod and a wink, given the likelihood that both village resident and educated reader would have known the true identities of the two spirit images as Fuxi and Nü Wa, or been able to figure them out. Given his somewhat irreverent conclusion, and the fact that he is, after all, writing about his hometown, we might be justified in looking for more playfulness in his prose. How we understand such playfulness, when we find it (and even our ability to find it at all), will depend largely on our openness to that vast, grey middle ground between the canonically prescribed and the locally permissible, and to all its possibilities.

Conclusion: The possibility of organic links between state rituals and local religious practices

Zhang Peifang’s interpretive move in his inscription does seem to make a contribution, large or small, to orthopraxy. To explain this in the language of the “standardization process,” Zhang “standardizes” the Guangyu Shrine’s main deity by giving her canonical credentials that might very well have contributed to the temple’s longevity, yet oddly this standardization, he insists, requires absolutely no changes to local ritual custom. Can this even be called standardization? His contribution to orthopraxy is rather to expand its range to cover areas clearly interpretable as heteropraxy, not to effect change at the temple itself. We might alternatively say his contribution is to acceptable practice by local officials, making tolerance more canonically and historically justifiable. This perspective recognizes that his intended audience likely included other local officials and students (future officials), and also recognizes “standardization” as but one part of a broader set of interpretive strategies that Chinese social agents were consciously aware of participating in; Zhang was making an intervention not just into local religious practice, but into the processes and goals of standardization itself.Footnote ⁵⁹

Given that tolerance generally and a concern for the common populace specifically are easily understood as philosophical or ideological tenets of Confucianism, what can we conclude about Zhang in relation to Confucian self-cultivation? Did his “ritual, moral, and textual mastery” allow him to grasp “the moral, emotional, and ritual circumstances” of his encounter in his hometown temple and produce “the precisely suitable words and actions to meet the occasion” (as I describe above)? The narrative he employs to present his interpretive move is certainly contrived, and the interpretive move itself (portraying Nü Wa as Gaomei), as we have just seen, was not fully original. Moreover, his entire critique of the temple’s layout and placement of the spirit images turns out to come not from an encyclopedic knowledge of the rich and varied textual record of past dynasties, but rather from an actual encyclopedia, the comprehensive institutional history Tongdian 通典, compiled between 766 and 801 by Du You 杜佑 (735–812), including both the information about accompanying deities in imperial sacrifices and the ritual seating arrangement of husbands and wives.Footnote ⁶⁰ Perhaps we should deny that Zhang’s inscription was an example of virtuosic improvisation, but in the end such a judgement will depend more on how we understand or define improvisation than anything else. What the inscription certainly does show us, in great detail, is how classical training could in fact equip even low-level officials to navigate complex social and political realities. To the extent that men like Zhang Peifang and Li Rimao after him were able to proffer interpretations and spin narratives that indeed integrated their local temple and its rituals into the “formal hierarchies and symbols of power, authority, cosmology, and orthodoxy” (mentioned above and made exceptionally clear by both Zhang’s spatial analysis and Li’s historical explanation of local custom), their classical learning helped to produce and reinforce social cohesion/social order, unquestionably a central goal of both Confucian learning and, of course, the imperial court.

Finally, returning to Li Rimao’s inscription at the same temple provides one last opportunity for reflection on the interpretive work done at such sites. Li’s inscription is centered on the argument that ritual opera and sacrifice at the village level are organically related to canonical state sacrifices at the highest level. We may understand his explanations of local opera being homologous to the state performance of hymns, or spring and autumn festivals centered on praying for and then offering thanks for the blessings of the gods being homologous to the annual state sacrifices to Heaven and earth, all as strategic moves within the framework of “standardization,” but his premise is neither historically unfounded nor anthropologically unappealing. Terry Kleeman has argued persuasively that imperial sacrifices and popular worship of local deities are conceptually united in the importance they place on sacrificial meat offerings. This is in contrast to institutional Buddhism and Daoism, both of which have historically rejected meat victuals, at least in principle if not in all details of practice.Footnote ⁶¹ This conceptual link is certainly not coincidental. So how did it arise? Simply put, either folk practices consciously copied elite rituals, or, as Li Rimao’s inscription suggests, both emerged organically out of common traditions. In either scenario, the history of this relationship must be very deep. We know that the offering of meat was already a part of routine rituals practiced at relatively low levels of society from as early as the fourth century BCE, as testified to by rites recorded in the many mantic Daybooks, rishu 日書, unearthed over the last several decades.Footnote ⁶² Rich evidence of the importance of meat offerings in state sacrifices, of course, goes back as far as our earliest written and archaeological sources from the Shang and Zhou dynasties.

I close with this reference to Li’s inscription in part because it reminds us that the high literary elite of the late imperial era, while they may have been consciously engaging in what we call the politics of “standardization,” were never the less also capable of offering deeply held convictions based on both their classical learning and their own heartfelt sentiments and intuitive insights engendered from a life lived in close contact with customs and rituals at many levels of society, often including the village. Whatever their true thoughts and feelings about folk customs—and we should expect those to be complex and not necessarily all internally compatible—both the accomplished official Zhang Peifang and the candidate for county magistrate Li Rimao showed extraordinary tolerance for, and perhaps sentimental attachment to, the ritual practices of their hometown.

Moreover, Li Rimao’s suggestion of deep, organic links between popular and state sacrifices deserves our serious academic consideration. Phrased in my own terms, I would argue that the religious and ritual practices of commoners known to us from across imperial Chinese history suggest that official state rituals have always been involved in a complex and often synergistic relationship with popular beliefs, cults, and rituals, and the history of those connections clearly has roots as far back as the pre-imperial era.Footnote ⁶³ Moreover, the interactions were certainly bidirectional; recall the early Han imperial family’s fascination with a local cult to the Lady of Changling. They were also often fraught with deep-seated contradictions or tensions; no amount of tolerance for local custom seems capable of completely erasing the structural gender bias of a political system rooted in patriliny. Sangren’s argument that Chinese patriliny “obviates women as subjects” suggests that the gendered contradictions between court ritual and folk practice are not only built into the system, but that local religious and ritual practices that are centered on women could, at times, become a form of resistance to patriliny and an active and even subversive pathway to female agency.Footnote ⁶⁴ Under these circumstances, searching for the precise origins of this or that ritual in official or folk practice becomes far less important than exploring how, over many centuries, individuals or groups activated, explored or exploited, or sometimes denied aspects of this long and highly productive relationship to pursue their own goals, goals that might be social or political (external), but might also reflect their (internal) moral or sentimental convictions.Footnote ⁶⁵