Spatial Characteristics and the Non-Hierarchical Nature of Regional Religious Systems (RRSs)

Abstract

Based on the spatial analysis and GIS modeling of the distribution of religious sites in Greater China, we have developed the concept of regional religious systems (RRSs) as a novel way of understanding and studying the spatial distribution patterns of religious sites and their relationship with other social and cultural factors. This essay further explores theoretical issues such as its center–periphery relations in existing administrative and economical hierarchies. Drawing on our current project on RRSs in the Hangzhou region and various available studies about pre-modern Chinese religion, the author explains the spatial characteristics of RRSs, such as the role of transportation, trade and pilgrimage routes in the formation of RRSs. Using Chinese Buddhism as an example, the author argues that RRSs in Greater China should be treated as a spatial formation without an internal hierarchical structure because the political and administrative hierarchy prevents the formation of a strong religious hierarchy.

1. Introduction

The concept of regional religious systems (RRSs) is a novel way of understanding and studying the spatial distribution pattern of religious sites and their relationship with other social and cultural factors. In our previous studies (J. Wu et al. 2013; Wu 2022), our team has developed this concept and conducted research on various religious traditions in different regions based on the spatial analysis and GIS modeling of the distribution of religious sites in Greater China. To summarize, we determined a preliminary definition as follows:

A regional religious system is a type of spatial formation in which a group of related or unrelated religious institutions are conditioned by physical, geographical, administrative, cultural, or socioeconomic systems and are highly dependent on regionally and locally distributed variables such as population, economy, transportation, education, culture, ethnicity, language, etc.

(Wu et al. 2013, p. 182)

This tentative definition serves as a loosely defined working definition to accommodate various attempts to study the spatial formation of religious phenomena identified regionally and locally. It emphasizes the relations in which the RRS is situated. According to our study, RRSs are basically a spatial formation characterized by the geographical distribution of religious sites. The formation of RRSs is shaped by regional systems such as William Skinner’s macroregions. Additionally, the distribution of these sites relies on regionally and locally distributed factors. In our recent edited volume, The Formation of Regional Religious Systems in Greater China (J. Wu 2022), we tested and further explored the various factors that have an impact on the formation of RRS in regions influenced by the Chinese state and culture. However, many theoretical and methodological issues have not yet been discussed. For example, because the concept of a system often implies the structure of hierarchy, it is natural to assume that a regional religious system is also hierarchical in nature, similar to an administrative hierarchy and economic system.

There is a great need to clarify these essential theoretical issues, some of which are presented in the introduction to my recent RRS volume (Wu 2022). We are convinced that religious sites, such as Confucian temples, shrines, and academies; Buddhist and Daoist monasteries; popular cults; etc., are not randomly distributed in space and time. Rather, they form a loose system with internal organizational logic. The history of Chinese religions is extremely complicated, and the Chinese cultural area is vast; therefore, we intentionally defined RRSs as open and informal systems in relation to the more formal administrative and economical systems. Basically, RRSs are informal systems with an emphasis on horizontal, rather than vertical, connections among various religious sites. We used “spatial formation” to describe all possible scenarios that we encountered in our experiments through visualization and mapping. Such an open definition enables researchers to discover certain rules and principles by analyzing an array of sites and the characteristics of such spatial formations without imposing a fixed and pre-determined model on the datasets. The concept of region, which has been discussed extensively in our RRS volume (J. Wu 2022, pp. 2–6), can be understood from the geographical, administrative, economical, and cultural perspectives. The regional perspective we adopted did not exclude us from viewing our datasets from national and local angles. Rather, we often observed dynamic interactions among the state, region, and locality in the formation of RRSs.

The working definition we provided captures some of the major aspects and factors that we needed to consider in the study of regional religious systems. First, the study of RRSs is not limited to the sites because institutions here are an integral foundation of the function of sites. Sites are only structures and locations for institutions or functions. To limit the RRS study only to sites, we greatly simplified the complex relations of religious systems from a regional approach. Institutions are an important part of our research as well. In addition, the various factors listed in the definition, such as population, economy, transportation, education, culture, ethnicity, and language, are qualifiable and quantifiable variables for the purpose of generating evidence-based research results based on data collection and operation. The variables listed in the definition can have actual values in spatial analysis, regression studies, and quantitative applications.

This study explored some additional factors to be considered in RRS research. In this essay, I draw freely from existing scholarship as evidence and intend to cover the whole history and spectrum of Chinese religion. The brief comments and observations of specific religious phenomena need to be understood in their original contexts. Many topics covered in this paper reflect our tentative thoughts; therefore, further empirical studies need to be performed to finetune and develop some of the arguments.

The central question is this: does a regional religious system, labeled as “informal politics” or “local parapolitical structure” by William Skinner (Skinner 1977), exist besides the administrative and economical hierarchies in a Chinese region? If it does exist, does it form a hierarchical structure? Our inquiries into the RRS have largely been shaped by William Skinner’s thoughts on the macroregion theory and center–periphery relationship in regional economic systems (J. Wu 2022, pp. 10–14). In addition to basing my discussions on the research presented in our recent RRS volume, I draw from our current project on the RRS in Greater China (rrs.arizona.edu) and various available research, especially from socio-historical, anthropological, and sociological studies of Chinese religion, to explain the spatial characteristics of RRSs. The formation of RRSs is examined in relation to regional administrative hierarchy, which mostly represents the domain of political dominance, and to regional economic hierarchy, which shows the organization of economic “nodes” where social and cultural resources are concentrated. In addition, transportation and pilgrimage routes connect these “nodes” together and form a multilayered trans-local network, which has also shaped the formation of RRSs. In conclusion, RRSs in Greater China should be regarded as a spatial formation without an internal hierarchical structure. My point is that the political and administrative hierarchy prevents the formation of a strong religious hierarchy. By delineating these various factors, this study also provides a guide to conducting RRS research.

2. RRS in the Center–Periphery Relationship

2.1. Distinction of Center and Periphery

The distinction between center and periphery is important to discern the pattern of the spatial formation of RRSs because the mode of governance varies in the center and periphery. The state power in China and its influence on social, cultural, and religious life is strong and will remain strong in the foreseeable future. Its role in shaping RRS thus has to be considered and reflected in data collection and model-making. In pre-modern China, there were a variety of religious sites related to state ceremonial systems, which should be collected and studied. Moreover, the state influence that can be seen in the data collection process shows that the religious site data vary considerably from the state statistics and the data collected from local sources. In immediate areas, when the state and local government have exerted strong control, such as administrative centers, we see the formation of an urban RRS with strong state interference and officially patronized sites clustered and spread along administrative centers.

Islam and Christianity are no exception: during the Yuan dynasty (1271–1368), many Central Asians with an Islamic background served the Mongols as administrators; Islamic communities tended to spread in major administrative centers (Ryavec and Henderson 2015). Christianity in the late nineteenth and early twentieth centuries, under the aegis of state protection guaranteed by “unequal” treaties, spread along administrative structures (W. Wu 2008). In contemporary China, as Christian churches are targets of state control and surveillance, they respond by growing faster in peripheral border areas where the state power is less forceful (Yang et al. 2022).

This pattern of the state influence in the administrative centers can be clearly seen from Hangzhou during the Wuyue 吳越 kingdom (907–978) in today’s Zhejiang area. The Wuyue kings adopted a pro-Buddhist policy and built many stupas and monasteries in the area. Through our mapping of the sites built by the five kings (See Figure 1), we can tell that these Buddhist sites tend to cluster around the city, and they formed an RRS linked by the state power, especially in the West Lake area, whereas fewer sites were built in remote areas. This spatial pattern provides a template for building Buddhist temples in later periods (Welter 2022, pp. 43–45).

The political and administrative centers often attracted state-sanctioned religious sites, which conformed with the state ideology. However, in the peripheral areas, “illegal” religions and “illicit” cults often flourished because of the weak state control. Donald Sutton, for example, surveyed the Chinese shamans (wu 巫) in local gazetteers and determined that the great majority of shamans tend to concentrate at the periphery or the edge of the core area (Sutton 1981). Due to the sharp contrast in social control between core and periphery areas, some social historians of China have suggested that pre-modern rural China has achieved a certain level of “ritual autarky”, according to which the state seldom intervenes unless their practices threaten state power. According to David Johnson, rural China was largely insulated from the state and government, and also from a centralized ecclesiastical hierarchy, which rarely intruded into village life except for tax collection and moral indoctrination. Then, various religious sites and associations played a larger role in organizing village life (Johnson 2009). Due to the importance of this distinction, we should explore the formation of RRSs in central and peripheral areas more closely.

2.2. RRS in Cities and Regional Centers

The power of the Chinese state is particularly strong in cities and regional administrative centers in shaping the religious landscape. In pre-modern China, state-stipulated ceremonies were not only held in capital cities, but were also ordered to be held in regional and local administrative centers.1 In central metropolises and regional cities, city gods (chenghuang 城隍), revered protective territorial deities representing an otherworldly equivalent of the governing magistrate of that city, have been worshiped since the Tang dynasty. Accordingly, the rise of the cult in the Song reflects the rise of merchants and craftsman (Johnson 1985). However, as von Glahn points out, in the Song, the founding of City Gods Temples largely corresponded to the administrative hierarchy; no City Gods Temples were built in market towns. Rather, the “detached palaces” (Xinggong 行宫) of the Dongyue God 東岳 became the hallmark for promoting the stature of the market towns (see “The Song Transformation of Chinese Religious Culture”; von Glahn 2004, pp. 170–71; Goossaert 2011, vol. 1, pp. 192–93; Hamashima 1992).

Megacities such as Chang’an (now Xi’an) in the Tang dynasty (618–907), Hangzhou in the Southern Song dynasty (1127–1279), Beijing in the Ming (1368–1644) and Qing (1644–1912) dynasties, and Nanjing in the late imperial and the Republican periods are examples of fully developed religious ecologies.2 Not only was the state ceremonial system fully installed and marked the cityscape, private interests of the royal family, aristocrats, and their associates also became patrons of temple building activities. More importantly, city dwellers who engaged in trade and crafts patronized their own cults and created their own version of sacred space characterized by regional pilgrimage cults, such as the famed Lady of Azure Cloud (Bixia Yuanju 碧霞元君) in Mount Miaofeng 妙峰山, close to Beijing.3 These city dwellers and sojourners brought in their own gods and patronage deities for their guilds and same native place associations.

In late imperial China (1368–1911), as Kristofer Schipper and Shiba Yoshinobu observed, temples in cities such as Tainan 臺南and Ningbo 寧波 often served “one quarter or even a single ward within the city” (Schipper 1977; Shiba 1977; Skinner 1977, p. 264). Stephan Feuchtwang also pointed out that village temples in China were institutions for that village area, whereas religious cults in the city, such as city gods, were directly serving the city’s people without ritual jurisdiction over other similar cults within the city’s administrative boundary (Skinner 1977, p. 591). It appears that in most Chinese cities, two types of religious institutions were close to city nuclei and were the most emblematic official religious institutions: Confucian school-temples and city god temples, which were adjacent to government buildings (yamen 衙門) and market centers, respectively. In a typical Chinese city, Confucian school-temples represent the “culture” (wen 文), together with Guan Di, representing “martial art” (wu武). These cults were sanctioned by the state; therefore, their sites were usually closer to government offices in the center of the city. The city god temples, as well as other major popular religious cults, often reflected the bustling economic activities in city; thus, there were closer to the markets.

In Southern Song Hangzhou, the distribution of various social and economic functions followed the south–north axis created by the Grand Canal which led to Hangzhou. According to Shiba Yoshinobu’s 斯波信義 study (1988), three functional zones can be clearly delineated in the irregular city landscape shaped by the Western Lake. The economic zone is located in the center of the city; the official zone is located in the south; and the gentry zone is located in the north. The state religious facilities, i.e., the Confucian academy and Ministry of Rite’s School of Tribute Scholar (Gongyuan 貢院), are concentrated in the southern part of the city. A large number of people moved to Hangzhou after the fall of the north to the Jurchens in 1127; thus, immigrants brought their native cults with them to symbolize their regional identify. Buddhist temples were concentrated around the West Lake, especially in the north and south shores, where many Merit Cloisters (Gongdeyuan 功德院) and tombs of eminent generals and officials were located. On the west shore, large-scale monasteries such as the three Tianzhu monasteries 天竺寺 were located.4

2.3. RRSs in the Peripheral Area

In the hinterland rural areas, the state also exerted its influence by setting up local shrines and shaping communities according to the administrative structure, thus creating boundaries for local religious activities. It can be expected that state power has the strongest influence on religious sites close to the administrative centers; however, in peripheral hinterland rural areas, the state has less rigid control.

As early as the second century B.C., the Chinese state had installed a fully-fledged administrative order in rural areas. Below the county level, the Qin empire (221–206 B.C.) established a system of townships (xiang 鄉), cantons (li 里), pavilions (Ting 亭), Group of Ten (shi 什), and Group of Five (wu 伍). This system was inherited by the Han dynasty (202 BC to 9 AD; 25 to 220 AD). In the Northern Wei dynasty (386–585), a system of neighborhoods (lin 鄰), cantons (li 里), groups (dang 黨), and associations (bao 保) was established. In late imperial China, the administrative Baojia 保甲system, including cantons(里), wards(甲), sub-neighborhoods (圖), and precincts (境)was followed. More prominent was the rise of lineage organizations that played an essential role in building communities. In addition, some locally originated gods became regional due to the spread of the cults through merchant activities, festivals, pilgrimages, introductions from new host communities, and promotions by the clergy (von Glahn 2004, pp. 130–79).

The spatial characteristics of RRSs in peripheral rural areas were largely shaped by ritual activities organized by the locals. The religion in Taiwan, because of its continuity with tradition, has been most intensively researched by several generations of historians, sociologists, and anthropologists. Fan I-chun 笵毅軍 and his team showcased the most important step toward digitizing the results in various databases related to ceremonial processions in rural areas. Their studies show that the local cults and village alliances have clear spatial and territorial structures, which have been often referred to as a “ritual sphere” (jisiquan 祭祀圈), referring to a clear territory of relative ritual activities such as communal festivals and processions related to a given popular god in the community and neighborhood as an expression of social solidarity.5 These “ritual spheres” can be clearly demarcated by mapping the activities of “raojing” 繞境 or “precinct-circling”, using GPS trackers logging around the ritual boundary of a given local cult. Once every three to four years, these activities are often carried out simultaneously on different routes within one to four days. According to their studies, the processions are motivated by the same identity of belief and organized within a regional spatial structure (Fan and Allio 2014; Fan et al. 2022).

2.4. RRSs as Part of an “Informal” Territorial Structure

William Skinner emphasized the hierarchical nature of the regional economic system. However, he characterized religion, education, neighborhood councils, and lineage organization as “informal” social and cultural structures which paralleled the layered marketing systems (Skinner 1977, p. 336). His insight can be extended to RRS research, especially when we investigated temple-building activities and their relationship to the local economy.

With the exception of government patronage, construction activities of religious sites are largely spontaneous endeavors undertaken by local communities and thus have significant implications in regional economic systems. In pre-modern China, religious sites were built for various reasons and received patronage from different groups of people. There was no internal hierarchy within a given religious tradition to mandate the building of a religious institution in a local area; therefore, the initiatives of building a site largely depended on local circumstances. During their expansion, Buddhism and Daoism often took over shrines of local cults and incorporated them into the Buddhist or Daoist pantheon. Some religious institutions were shared by Buddhists, Daoists, and local ritual experts. Even those institutions which were sponsored by the state tended to draw significant support from the region for their maintenance and repair.

In late imperial China, the survival of religious institutions depended on the resources that a given temple was able to garner within a region. Very often, temple building activities require considerable investment drawn from local areas and were often supported by local elite, lineage organizations, and various voluntary religious associations. As Timothy Brook shows in his case studies of the Buddhist revival in the sixteenth and seventeenth centuries, temple revival around that time was largely sponsored by local gentry communities. Such a fervor in building temples was often linked to a changing relationship between the state and the local community and to the rise of local activism during a particular time (Brook 1993).

This means that religious sites in pre-modern China developed a high level of spatial dependency on regional and local economic resources. Immigration and population growth create the need for reallocating natural resources, and for organizing social activities such as self-defense, and especially, water reallocation. All major Chinese regions were based on drainage basins where rivers and their tributaries formed networks of irrigation and transportation. Religious activities on the regional level thus were closely associated with water work, irrigation, and the allocation of water resources in pre-modern China. This aspect was illustrated in Kenneth Dean and Zheng Zhenman’s study of the Putian 莆田area, which shows the centrality of irrigation to the Chinese state, the administrative division, and the organization of the local religious system (Dean and Zheng 2010).

In the north, the development of local cults was also closely related to land acclamation and water projects. The Fenshui 汾水 watershed in Shanxi 山西, for example, was a cultivated area that heavily depended on the fair distribution of water resources among different villages in the watershed.6 Its tributary, Jinshui 晉水 River, has been utilized for irrigation since the Han dynasty, and the allocation of its water to the northern and southern canals was historically divided according in the ratio 7:3. However, the distribution of the water resource was crucial to the livelihood of local people and often resulted in disputes. Temple systems, such as Jinci Temple晋祠, a Water God Shrine (Yuanshenmiao源神廟 or Shuishenmiao水神廟), where the spring emerged were thus erected as genius loci and dividing junctures of key irrigation projects and water control systems. These religious sites were maintained through moral obligations for local people, enabling them to engrave the rules of the water distribution onto religious symbolism and folklore related to the Holy Goddess of the Jin River (Jinshui yuanshen shengmu晋水源神聖母) (Harrison 2002, pp. 94–101; Miller 2007; Zhao 2002). Periodical sacrifices to the water god held on the sites were acknowledged by the state and local government; local resources were invested to maintain its symbolic function. The function of these irrigation temples was thus also economical. These hydraulic temple systems largely determine the establishment of administrative systems, which depend on the allocation of water resources.

Similar processes happened in other places where the allocation of water resources becomes an issue. Research into the relationship between religion and these economic activities suggests that Chinese religion has a spatial structure closely following the economic and demographic structures.

3. RRSs in Relation to Transportation, Commercial Routes, and Pilgrimages

3.1. Transportation

The rise of Chinese civilization rested on the development of a vast transportation system linking the entire Chinese ecumene. Roads and highways were built in the early stage; waterways were utilized for transporting goods, provisions, grains/rice, and human passengers; canals were dug to connect different water systems. Religion also followed certain construction projects. Shrines were often erected along the highways and waterways for the builders or road gods; travelers carried “travelling implements” such as wooden tablets with a deity’s name on it to sacrifice during road offerings (Nylan 2012). In addition to roads and waterways, the state developed extensive courier and postal systems to relay government documents and goods. Based on these official networks of transportation, commercial routes were developed and utilized by merchants to transport goods. These routes have no doubt facilitated the travels of religious personnel. As a result, certain religious traditions were often transmitted from one place to another along transportation routes.

Our research has placed great emphasis on the role of transportation, including its relation to commercial and pilgrimage routes. In our previous study, the importance of transportation in the formation of RRSs has been stressed (Wu et al. 2013). Through mapping, it has been identified that Buddhist institutions mostly spread along the transportation routes characterized by waterways. The role of transportation network in spreading a particular religion and cult has been also confirmed by social historians of religion. The construction of the Grand Canal was the most important project that connected the water systems of the Yellow River and the Yangzi River; both are east–west oriented. The so-called “Hydraulic religion” was thus developed along the waterways to offer religious protection. During the Ming and Qing dynasties, the state promoted the cult of the Fourth Son Golden Dragon Great King (Jinlong sidawang 金龍四大王) as the patron god of the Grand Canal and the Yellow River (Dodgen 1999). The Luo sect 羅教or Wuwei sect 無為教, popular in late imperial China, founded by a solider from Shandong called Luo Qing 罗清 (1442–1527), who developed his teachings when he was stationed in the Beijing area, was spread further to the south along the Grand Canal (ter Haar 2014; Overmyer 1978). The boatman population along the Grand Canal also developed complicated religious culture involving hydraulic gods, ritual opera, etc. (Lin 2012).

The spread of religion in the southeast coastal region, mostly the Fujian area, was particularly influenced by transportation. Despite the mountainous topography in this region, cities and towns were connected through waterways and highways, which played an important role in facilitating the exchange of goods and personnel. In a study on RRSs, I have identified a triangular area along the border of the three neighboring provinces, commonly known as northern Fujian (Minbei 闽北) and located at the border areas of today’s Zhejiang, Jiangxi, and Fujian. This “triangle area” used to be a transportation hub, but also a cluster of Buddhist temples (Wu et al. 2013, p. 188). Not only was Chan Buddhism developed along the transportation route; Christianity also spread along the waterways of the Min River from the coastal center of Fuzhou during the late Qing period (W. Wu 2008).

Often neglected is the role of the military system in transportation, communication, and internal immigration, especially after the Ming dynasty (1368–1644) when the system of military households and stations was established. The networks of garrisons and the movement of troops with their households provide additional channels for communication. In the Ming dynasty, military garrisons had their own territories and populations and were sometimes equivalent to local administrative units (Brook 1985; Dreyer 1988). Military systems even influenced immigration in the north and the spread of religion, such as the cult of Guan Yu 關羽, which was transmitted quickly through military garrisons in the early Ming dynasty (ter Haar 2017).

3.2. Trading Routes

Studies have shown that many regional cults derived from the local worship of baleful ghosts who often had insecure personal lives and suffered tragic death without descendants, but had miraculous power to protect local communities. These cults then spread regionally and even nationally. Similar processes occurred in the spread of eight popular deities in Fujian, as Barend ter Haar posited. One distribution pattern ter Haar identified is the spread along the trading route where merchants, monks, and migrants traveled. More interestingly, temples were built close to bridges. In later times, Buddhist monasteries often appeared to be the private sponsors of road- and bridge-building. The distribution pattern of these Fujian gods also fits in William Skinner’s macroregion of the Southeast Coast, as ter Haar pointed out, and can be further divided into three sub-regions for the cults: northern Fujian and southern Zhejiang, coastal southern Fujian and northern Guangdong, and western Fujian (ter Haar 1990).

Social historian Valerie Hansen also showed that the spread of regional cults such as the Five Manifestations (Wuxianshen 五顯神), Zitong God (Zitong Dijun梓潼帝君 or Wenchangjun文昌君), the Heavenly Consort (Mazu 媽祖), and Zhangwang God 張王神 were spread with the aid of merchants, following the waterways extending from one region to another. During the “commercial revolution” of the Song dynasty, these newer gods of “low birth” were patronized by new settlers as extra-local “regional deities” in newly developed areas, such as the lowland area in Huzhou 湖州 prefecture near Lake Tai 太湖 (Hansen 1990; Shi 2016). The imperial transportation system also became the origin of some local and regional cults. Japanese historian Hamashima Atsutoshi 浜島敦俊, for example, confirmed that after the Yuan dynasty, some of the local gods, all of whom were anthropomorphic with a human name, such as the various Zongguan 總管 (Superintendent) cults, derived from the late Yuan dynasty government office titles concerning tax grain transport. They became water gods and were later worshipped by peasants as well (Hamashima 2011).

Semi-religious organizations, such as ritual opera troupes, served local societies as possible links to the dissemination of religious symbolism through the regional network of transportation routes. According to Tanaka Issei 田仲一成, in late imperial China, opera, often supported by market towns, villages, and lineage organizations, had become an integral part of the life of Chinese lineage organizations, which often sponsored performances for the purpose of sacrifice to communal gods (waishen外神) and lineage gods/ancestors (neishen 內神) at the occasion of festivals and memorial days (Tanaka 1972, 1985). The spread of certain types of ritual opera, clearly based on the distribution of local dialects, has the most obvious regional features. In his study on the geographical distribution of the Mulian opera 目連戲, Tanaka Issei traced the regional origin and distribution of its “text system”. He found that the spread of the opera had a unique geographical aspect and was closely related to local transportation and dialects. The spread of the Mulian opera from Huizhou 徽州, where the standard version was finalized by the literatus Zheng Zhizhen 鄭之珍 (1518–1595), was closely associated with the travels of Huizhou merchants and the extension of a rice trading network centered on Huizhou (Guo 2005). The records of the Mulian opera performance match the key travel routes around Huizhou, as recorded in the travel guidebooks published during that time. To explain the origin and spread of the Mulian opera in Fujian and Zhejiang, Tanaka consciously borrowed insights from William Skinner. Within the framework of Skinner’s macroregion theory, Tanaka identified that within the region of the Southeast Coast, in which Fujian is the center, it was the vernacular version of the Mulian opera developed in Putian, which was spread to South Zhejiang because of the maritime transportation of merchants (Tanaka 2016, pp. 168–70, 747–57).

3.3. Pilgrimage Routes

Pilgrimage routes and the spatial pattern they create is an important characteristic of RRS in China. In spatial terms, pilgrimage, through imposing a spatial lineup of a group of sacred sites, created a new relationship of these sites which were often located in the periphery and remote places. It insinuates a unique spatial formation different from the center–periphery relationship dominated by demography and economic factors. Pilgrimage routes also have regional features and often function as a mechanism of spatial organization in a given region. These routes constitute a spatial structure at national, regional, and local levels. Therefore, pilgrimage connections are particularly important in China because there are no formal and structural links to connect the numerous religious sites together.

According to Steven Sangren’s study of Ta-ch’i in Taiwan, pilgrims are often organized at the grassroot level, such as the village territorial-cult level (Sangren 1987, pp. 88–89). Several recent studies (Bingenheimer 2022; Ouyang 2022; Zhang 2022), following Timothy Brook’s research, have successfully mapped pilgrimage routes and situated them within their regional contexts, such as Skinner’s macroregional systems, as suggested in Susan Naquin and Chunfang Yu’s book on pilgrimage and sacred sites in China (Naquin and Yü 1992). The Buddhist pilgrimage network, for example, largely overlapped with the commercial routes, as these authors identified. Marcus Bingenheimer, in particular, showed the existence of a “pilgrimage square”, a spatial pattern of spiritual traveling that connects sacred mountains. This interesting pattern in north China suggests that Chinese temples tended to cluster along a rectangle regional circumference in late imperial China, leaving the middle area relatively “hollow” (Bingenheimer 2022).7

The crucial issue here is whether pilgrimage routes and the organization of pilgrimage groups indeed provide an internal hierarchy for RRS. Zhang Weiran’s study (Zhang 2022) did allude to the multilevel order of pilgrimage routes in the lower Yangzi River region. However, this order did not have any actual subordinate relationship. Isabelle Charleux also noted that, among Mongolian pilgrims to Mount Wutai, there were a variety of social and economic motives (Charleux 2015, pp. 205–76). Some temples clustered around a famous site, such as Mount Wutai 五台山, may form a certain kind of local alliance without an obvious hierarchical structure (Charleux 2015, pp. 68–76). Moreover, as Sangren notes, in Taiwan, pilgrimage centers are often located in the periphery of political and economic hierarchy, but they depend on the same transportation routes (Sangren 1987, p. 122). Their observations confirm what Victor Turner has pointed out: “any region possessing a certain cultural, linguistic or ethnic unity, often corresponding also to an area of economic interdependence tended to become at once a political unit and a pilgrimage catchment area” (Turner 1974, p. 179). This means that although pilgrimage routes connect a multitude of religious sites together, they do not form an internal hierarchy among these sites which oriented these sites towards a center. These potential hierarchical structures may work well in local areas and small regions, but are usually suppressed when they tend to grow trans-locally. In Section 4, Chinese Buddhism is used as a case study to discuss this non-hierarchical aspect more thoroughly.

4. Hierarchical and Non-Hierarchical Aspects of RRSs

4.1. The Issue of Hierarchy

RRSs exhibit a great level of spatial dependence on existing administrative and economic systems. The question, however, is whether or not RRSs also form hierarchical structures closely mirroring the regional administrative and economic hierarchies. My answer is no, although RRSs indeed draw support from the two hierarchies. Although RRSs do not possess an internal hierarchy, they have served important social, economic, and cultural functions in Chinese society, paralleling the political and economic structures.

In our previous study, regional religious systems (RRSs) were described as being situated in the regional context defined by William Skinner’s macroregion theory in order to identify their relation to political and economic hierarchies in a given region (Wu et al. 2013; Wu 2022). This approach has been adopted by other scholars and was proven to be effective. Steven Sangren used Skinner’s regional systems method to study Chinese religion in Taiwan and described the rural territorial cult system as a “nested hierarchy”, showing “patterns of social interaction in spatial, behavioral terms” (Sangren 1987, p. 14). He noted that “hierarchies of ritual organization and social identity correspond closely in spatial terms to the nested hierarchies of economic regions, just as Skinner predicts” (Sangren 1987, p. 15). Although scholars have tried to describe the Chinese pantheon, especially Daoist and popular religion, as “structured hierarchy” with a “subordinated” lower status deity, such symbolic and the formal “universalization” of local deities plays little role in spreading and organizing local cults, even less influential than popular novels and folktales, as Paul Katz showed (Katz 1998, pp. 110–11). This approach gives readers the impression that RRSs also constitute hierarchical systems, which have their own structure to orient spiritual and institutional recourses towards a regional center. Although the mechanisms and tendencies of forming a spatial hierarchy do exist, they are limited in spatial scales and are often seen within village alliances in a small region. Most religious sites in pre-modern China remain establishments in their immediate locality, without the vertical connections that a hierarchy requires, except in the cases of religious sites with foreign origins (i.e., Christianity and Islam) and some short-lived attempts to form sectarian organizations.

This non-hierarchical characteristic has been observed by many scholars of Chinese religion. It is still useful to cite John Knight Shryock’s study on temples in Anking (Anqing 安慶). According to him, Chinese religion is remarkably “diffused” without a strong institutional base:

A Chinese temple must not be thought of as a kind of church, standing in the midst of a group of people who look upon it as their religious home. These temples are not built for worship by large bodies of people at one time. ……And it must be remembered that a Chinese does not belong to a temple and regularly attend worship there, as a Christian belongs to his parish church, for there is nothing in the Three Religions corresponding to the Christian congregation.

(Shryock [1931] 1973, p. 28)

Sidney Gamble reported a similar observation of religious institutions in the modern Ting county (Dingxian 定縣 or Dingzhou 定州) in Hebei:

The temples seemed to be relatively independent units. Most of them had been erected by the people living in the village. The building was financed sometimes by popular subscription, sometimes from village funds, occasionally by an individual who had acquired wealth and wanted to do something for his native village. When built, the temple was dedicated to the deity who offered the type of protection the people felt they needed or who personified the characteristics they honored and revered. Priests were in attendance only for special services. These priests had had, of course, the official Buddhist or Taoist training.

(Gamble 1954, p. 401; Yang 1961, pp. 324–25)

These observations suggest that (1) Chinese religious sites were independent (2), there was no hierarchical structure among them, and (3) there was no dominance of ordained clergy either. Of course, this conclusion was based on observations on popular sites devoted to popular cults, and established traditions such as Buddhism may give us the impression of having an internal hierarchy. As shown below, however, not only did Buddhist temples exhibit a non-hierarchical pattern, the attempts to establish a hierarchal structure across space also resulted in failure.

4.2. The Early Spread of Buddhism in China

In Chinese history, regional regimes and dynasties supported Buddhism as a national policy; a monastic bureaucracy was even established to manage Buddhist institutions. For example, among the Daoists in late imperial China, the office of the Heavenly Master appointed by the court in the Longhu mountain 龍虎山 in Jiangxi has often been regarded as a “Daoist pope” who dispatched his spiritual power of officiating Daoist ritual masters (Goossaert 2004). However, such a bureaucratic structure was only concentrated in major cities and administrative seats. Ideally, the central foci of these religious bureaucracies at each level of the administrative system would correspond to the spatial structure of standard, intermediate, and central market towns, as well as regional cities. Although these regimes patronized religious sites, some scholars of Chinese religion, especially scholars of Chinese Buddhism, tend to overemphasize the role of the state and the alliance between the Buddhist community and the state without an adequate consideration of Buddhist sites in regional and local settings. However, a hierarchical structure patronized by the state was never fully established.

As a matter of fact, the majority of Buddhist sites in China were not built by the state, but rather by local forces. According to an often-cited source in A Comprehensive History of Buddhism (Shishi tongjian 釋氏通鑒), in the Northern Wei dynasty (386–534), the state built approximately 47 great Buddhist monasteries, whereas the aristocrats built about 840 temples. However, the “commoners” had built approximately 30,000 temples.8 Little is known about the details of these 30,000 temples, such as their size and services provided. However, their existence must be explained from other perspectives, especially the mode of transmission and its relationship with the regional economic system.

The non-hierarchical nature of Buddhist institutions in China can be explained based on the early transmission mode of Buddhism in China. As Erik Zürcher posited, in the vast rural area of China, Buddhism adopted “a spontaneous mode” of transmission or “disorganized, polycentric infiltration”, which responded to each unique local situation differently in order to fit into each locality. Therefore, no centralized structure took form within Buddhist communities unless mandated by the government. This is in contrast to “contact diffusion” in India, as he describes below:

Once a local sīmā (‘alms circuit’) had been established and grown to its optimal size (corresponding to the number of mendicant monks that could be borne by a local productive community), monks would move on to establish new vihāras in adjoining territories. Thus Buddhism branched out from an ever-increasing number of centres, filling the territory in a homogeneous way. In China, sheer distance and physical geography combined to produce a completely different type of diffusion.

(Zürcher 2013, p. 340)

Erik Zürcher considered the geographical setting in the process of the “Buddhist conquest” of China, making insightful observations as if he had a huge database of Buddhist sites. He recognized the vast geographical differences in China by first identifying the diffusion of Buddhism as a slow process, taking roughly four centuries to spreading to all major regions in China and across all social spectrums. Unlike the mode of contact diffusion in India, such a process is “fragmented and piecemeal”, first being developed in the north and northwest under the aegis of non-Chinese rulers, whereas people in the south developed a more Sinicized version of Buddhism. The two types merged after China’s unification in the sixth century, giving rise to indigenous Chinese Buddhist traditions. Secondly, Chinese Buddhism originated from different sources representing different kinds of Buddhism. The great emphasis on unity and integrity in Chinese Buddhism exactly demonstrates its diversity and contradictions. Thirdly, the vast distance between China and India created a linguistic barrier which prevented the Chinese from directly communicating with Indian Buddhist centers, but enabled more freedom to translate and understand Buddhist teaching in classical and vernacular Chinese (Zürcher 2013, pp. 339–42; Liu 1988). In contrast, according to Zürcher, the spread of Christianity in late imperial China was a centralized and guided process commanded by the Vatican in Rome through the dispatch of capable foreign missionaries to build a hierarchical church network throughout China (Zürcher 2013, pp. 384–88). These three factors had a lingering effect on the formation of Chinese Buddhism in spatial terms.

4.3. The Spread of Buddhism in Medieval and Later China

The spread of Buddhist sites displays some interesting spatial patterns during the early medieval China. Yen Gengwang 嚴耕望, for example, based on his study of the origin of eminent monks, discovered a “vacant area in central plain” 中原空虛區 between the Yellow River and the Huai River 淮河 with no significant distribution of eminent Buddhist monks, as shown in Figure 2. Accordingly, the distribution of eminent monks was concentrated along the Taihang mountain 太行山 and in southern China along the Yangzi River (Yan and Li 2005, pp. 57–58). This left the central area between the Yellow River and the Yangzi River as a noticeably empty space, without significant Buddhist activities. The reason for this spatial pattern is not clear. This certainly does not mean that no religious activities occurred in this area. Rather, it shows that there might have been a different mode of religious dissemination in this vacant area.

In medieval China, Buddhist institutions often appeared as land grant estates with their own land, affiliated populations, and fortified compounds, similar to the great estate of larger clans. They competed with the state power, dispersed into the vast unclaimed land, and played a significant economic role in organized efforts for land reclamation, thus forming small local economic, social, and religious units along basins of short rivers or tributaries where water resources are available.9

The late Tang and Five Dynasties period witnessed a change in the spatial distribution mode of Buddhist sites in addition to the shift of China’s demographic, economic, and cultural center to the south (Liu 1989). In late imperial China, Buddhist institutions were even further relegated to the margin of the state power, and largely became individual local institutions without being integrated into any bureaucratic structure. This function of Buddhist monasteries in agrarian China gradually declined as such functions were taken away by lineage organizations after the late Tang and Five Dynasties periods, starting from the second half of the eighth century.10 The government did appoint clerical supervision agencies; however, these offices rarely functioned efficiently. Timothy Brook’s subsequent observation of Chinese monasteries in the Ming dynasty may be extended to all other periods and indicates that Buddhist institutions did not appear to be part of a large hierarchy or network:

Ming Buddhism existed as a congeries of little institutions dispersed randomly across the country, without hierarchy, internal organization, or any regulatory body other than what the state supplied. With the exception of limited ties among sister monasteries and linked pilgrimage sites, Buddhist institutions did not participate in a larger institutional framework at any level. Unlike European Christianity, Ming Buddhism was not woven into the net of secular power.

(Brook 1993, p. 29; J. Wu 2008, pp. 261–62)

4.4. The Development of a Hierarchical Lineage Model

Within Chinese religion, some mechanisms possibly enable the process of sectarian formation, which leads to the development of religious hierarchy. However, these mechanisms have not been fully developed. A bureaucratic model can be considered as the first mechanism, which mimics the state administrative system. Many popular Daoist and territorial cults tended to organize their pantheon around an administrative hierarchy model by naming and arranging their gods with official titles and ranks. However, this bureaucratic model was restricted to the symbolic realm, with extension to the actual organization of their religious sites.

The mechanism of “the division of the incense (fenxiang 分香)” among popular cults can be considered the second alternative. In south China, some temples of popular religions were organized through this mechanism of “the division of the incense (fenxiang 分香)”. This also created a potential multilevel structure of mother–daughter temples, which was formed by transferring incense burner ashes from an existing temple to a newly founded temple, creating an institution “under the incense burner” (luxia 爐下) of the older one (Skinner 1977, pp. 581–608).

In a given region, a group of temples can form a spatial structure with one prestigious temple as the center, resembling a hierarchical structure. Although this relationship makes the new temples tributaries to a place of origin, and on special ritual occasions, monetary contributions and pilgrimage to the senior temples are expected, this kind of relationship is subject to constant change and is interdigitated with its connection to various voluntary “liturgical associations”, as Kristofer Schipper identified (Skinner 1977, p. 653). This is rather an informal and loose type of association rather than forming a hierarchy with layered one-directional orientations. Moreover, many other local cults such as the Earth God 土地公, a neighborhood cult whose burners are usually filled with rice, never related to each other through fengxiang. Additionally, among the sectarian divination (bailuan 拜鸞) “phoenix halls” (luantang 鸞堂) in Taiwan, even though the “mother-and daughter temple” relationships do exist, they only indicate historical connections rather than “any thought of centralized control of a system of temples” (Jordan and Overmyer 1986, p. 80).

The traditional lineage model, however, is the most convenient and promising tool to organize clergy and religious sites in Buddhism such as Chan 禪, and Daoism such as the Longmen 龍門and Zhengyi 正一 traditions in late imperial China, and sectarian religions and salvationist movements such as the Luo sect 羅教 and its variants.

Within Buddhism, the Chan Buddhist tradition adopted the lineage model to its full extent through the method of dharma transmission after the eighth century. As detailed elsewhere, dharma transmission “bonds a group of monks with a special kind of spiritual relationship comparable to that of father and son in a secular Chinese lineage organization” (J. Wu 2008, p. 10). Moreover,

The Chan lineage is perhaps the largest and longest lasting lineage organization in China. Unlike lineages in the secular realm, the Chan lineage is maintained by an imagined form of reproduction. By means of dharma transmission, dharma heirs gain legitimacy to succeed to the patriarchal position in an imagined family. Therefore, the continuity of dharma transmission is central to the survival of Chan lineages.

(J. Wu 2008, p. 34)

This system of dharma transmission implies a hierarchy and has the potential to establish a spatial structure for Chan monasteries. As early as the seventh century, Chan Buddhists created a fictional patriarchal transmission lineage from Śākyamuni through an unbroken line from the twenty-eight Indian patriarchs to six Chinese patriarchs. The golden age of Chan Buddhism was the Five Dynasties and the Song Dynasty, when collateral lines of transmission from the Sixth Patriarch Huineng’s 慧能 (638–713) Southern School developed and Chan genealogies, the so-called “Record of Lamp Transmissions” (chuandenglu 傳燈錄), were compiled and sanctioned by the court. The old myth that Chan communities developed in remote mountainous areas might bear some truth; for the most part, the early Chan congregations were rural institutions in the south amid immigrant agrarian communities and developed a sense of self-reliance and identity through emphasizing their dharma lineage connection based on a master–disciple relationship. During the late Ming dynasty, the notion of dharma transmission and the compilation of such genealogies reached another height and then gradually declined again (J. Wu 2008).

In late imperial China, a typical Chinese temple operated on a lineage model closely resembling a secular lineage organization. Masters and disciples formed a lineage of dharma connections without blood ties. Buddhist communities were divided into public monasteries (Shifang conglin 十方叢林), hereditary monasteries (zisun miao 子孫廟), and dharma transmission monasteries (chuanfa conglin傳法叢林). Monastic lineages were divided into the tonsure family (tidu zongpai 剃度宗派) and dharma transmission family (chuanfa zongpai 傳法宗派), and they were most relevant to the abbot succession system. The public monasteries would allow heirs from all lineages to become abbots. The hereditary monasteries, however, only allowed the heirs of a given tonsure family to inherit the temple. Dharma transmission monasteries accepted candidates from a given dharma transmission lineage (Wu 2015, pp. 53–80). In large temples, monastic properties were divided into different “households” (fang 房) of monks where ownership was handed down along the lineage lines (Zhang 2015). Demonstrably, this lineage model was restricted to the abbot succession system and did not affect the independent status of a given Buddhist temple.

As revealed in this study of the seventeenth-century Buddhist revival, the fully revived Buddhist temples, although based locally, began to network along the hierarchical Chan dharma transmission lines cross-regionally. This attempt, however, was met with a severe response from the Qing government which would not allow any “vertical” structure paralleling to the state power in the central, regional, and local levels. Potentially, a group of religious sites, based on a coherent teaching and leadership, could have grown translocally based on a vertical power structure, thus posing threat to the existing governmental authority (J. Wu 2008, pp. 258–62). However, it seems that the existing hierarchies, both administrative and economical, prevent the development of an additional hierarchical structure. The potentials for developing a regional hierarchy, before they were fully grown, were suppressed by the Chinese state, which has serious concerns over any hierarchical parallel to the existing political structure. This also explains the concerns of the Chinese government about religion and the reason why potentially building a religious hierarchy should be suppressed from the state’s point of view.

4.5. The Government’s Role in Preventing the Formation of a Religious Hierarchy

RRS implies a power structure, a “spiritual bureaucracy” mimicking the government offices, which might be in parallel to the existing power structures represented by the Chinese government. In some areas, especially the peripheral regions where the Chinese state power does not have a strong presence, religion constitutes the function of the “second government”, as Kenneth Dean heavily emphasized (Dean 2001). Kristofer Schipper developed a similar concept and claimed that “Neighborhood religious associations formed the focus of a sort of local self-government” (Skinner 1977, p. 615). Thus, the political implication of a hierarchy-oriented RRS is even more threatening in some areas, especially in China’s ethnic regions, because a certain religious network with a structure could lead to political claims and function as quasi-political organizations trans-locally. The government would certainly be aware of the potential challenge of the RRS in a given region. The usual state strategy, which can be described as a “territorial principle” in legal terms, would be to suppress the transregional potential of this structure and to keep religious sites strictly local without trans-local hierarchies. Many factors influence and prevent RRSs from becoming hierarchical, and there is no benefit for the state to foster a hierarchal RRS. Instead, RRS is allowed to develop a horizontal relationship within a given region and locality rather than a vertical structure parallel to the existing political and economic hierarchies.

Official state cults and ceremonial systems implied a hierarchy because their establishments were closely paralleled with the administrative level and were supposed to cover the subdivisions within the region’s administrative boundary. However, a religious hierarchy delineated by an administrative boundary is very limited and not successfully established by the government. Many official state cults were local phenomena within a city or a given region, although the imperial government gave symbolic support to similar activities throughout its administrative region, although the popular temples were also territorially defined with its participants coming from a given territory in the city. In addition, numerous territorial neighborhood religious associations are affiliated with the temples or other civil organizations, such as guild and same-place associations (Tongxianghui 同鄉會), which operated outside the state system.

If the state were to exert such a strong influence on the religious system, it can be expected that the state would aim to establish a hierarchical system closely mirroring its administrative structure. Although the state has tried to impose certain administrative structures upon religious sites, most noticeably in Buddhism, such as the establishment monastic officer systems, the Five Mountain system in the Song dynasty, etc., these attempts were never successfully accomplished until the People’s Republic of China initiated nationwide bureaucracy to control religious affairs.

RRSs also represent heterogeneous forces foreign or even detrimental to homogeneous administrative systems. Thus, the RRS has to be suppressed and prevented from becoming a competing hierarchical structure. In this sense, the political hierarchy tends to “push” away the religious system. On the other hand, in relation to the economic system, each religious site relies heavily on local sources, as numerous studies show. The result is that these sites are “pulled” toward each “node” within the economic system, causing clustering along the existing regional central places and settlements and becoming a function of the economic hierarchy. This phenomenon also prevents the spread of a hierarchy within a religious system. The forces of “push” and “pull” from the two systems, one rejecting and one attracting, explains why a hierarchical RRS is not desirable and has to be controlled. At times of crises, these two hierarchical structures tend to encroach on religious resources, as seen in the late Qing and early Republican time period when a large number of religious sites were appropriated, leading to the collapse of stable RRSs (Katz 2014, pp. 17–67).

5. Conclusions

Unlike religious sites in pre-modern western Europe and Edo Japan, where parish systems and household registration systems were established through utilizing religious institutions, RRSs in Greater China were not absorbed into the administrative hierarchy because they did not have a function in Chinese administrations. In contrast, the Chinese administrative system tends to be comprehensive and has penetrated to the lowest level of the social spectrum, leaving little space for other social structures to grow. Even in areas where religion appears to undertake some functions of the “second government”, as Kenneth Dean describes, religious sites and the activities associated with them have to be “domesticated” through sponsorship from existing “legal” social entities such as lineage organizations and local elite groups.

The non-hierarchical nature of Chinese religious institutions was also noted by sociologists such as C. K. Yang, who investigated the social function of Chinese religion from the mid-nineteenth century to the 1960s and identified that the structural “weakness” of religious institutions is the lack of “any effective hierarchical structure”, except for in some sectarian religious organizations. As he observes, “The primary units, the temples and convents, were not only small but were also lacking in centralized organization of any significant size and any effective hierarchical structure. In actual operation, each temple or convent functioned largely as an autonomous unit” (Yang 1961, pp. 312–13).

However, RRSs are indeed spatial formations, not just an “ideal-type”, in addition to the political and economic hierarchies. More importantly, RRSs do not follow the confines of these two hierarchies. As Kenneth Dean showed, coastal areas such as Putian have often relied on overseas economic and cultural resources for revival, thus extending their domains outside China and forming more coherent unity with overseas Chinese communities (Dean 2022). Moreover, the locals created unique gods and beliefs only popular in one region such as the Three-in-One movement (Sanyijiao 三一教), suggesting that factors other than politics and economics, such as immigration, transportation, and local dialects, might have played bigger roles in Greater China (Dean 1998).

Whether or not the explanation presented here covers all religious aspects in every tradition, the kernel of truth is that no stable hierarchy exists among religious sites despite some potentials and initiatives on the local level, which tend to be thwarted by other exogenous forces. For the reasons suggested earlier, the non-hierarchal nature of RRS shows that we cannot assume that RRSs have structures similar to Skinner’s notion of hierarchal space, although administrative and economical forces may bring religious sites into close alignment with the levels in the two existing hierarchies. Rather, in addition to considering the role of administrative and economic hierarchies, the formation of RRSs in a given region exhibits more complicated organizational and network patterns fostered by various links such as transportation, trade, and pilgrimage, some of which are historical and can go back many centuries. It also shows, as one might suspect, an over-reliance on political and economic explanations of the spatial formation of RRS may be obscuring some regularities in religious communities, which routinely reoccur but can easily slip out of contemporary discourse.