LOCAL GODS (TUSHEN 土神) AND SPIRIT MEDIUMS (WUSHI 巫師)

During the late Qing, in the twenty-sixth year of Daoguang (1846), the peasants of Zhaowen County 昭文縣 in Suzhou Prefecture staged a large-scale riot, refusing to pay the rent their landlords demanded.Footnote ⁷ Before the tenant farmers mounted their rebellion, it is said that they performed divinations at their respective village shrines, and it was only after they had obtained the most auspicious results that they resolved to act. This massive uprising was eventually put down when forces from the Green Standard Army (Luying 綠營) were mobilized. More than twenty of the ringleaders were arrested and executed in the prefectural seat of Suzhou. The Prefect of Suzhou, Gui Chaowan 桂超萬, commanded the suppression, and at the same time that he meted out punishment for these denizens of the earthly realm, he also stated that “Malign forces have clearly taken possession of the deities and caused them to issue mistaken revelations and thus mislead the people.” He therefore seized the images of four deities (Zongguan, Zhoushen 周神, Mengjiang 猛將, and King Li 李王) that were enshrined in small temples located in the villages, hauled them to the county seat, and exposed them before the great gate at the City God temple (Chenghuangmiao 城隍廟) of Zhaowen County, punishing “the gods on the charge of confusing the populace with their oracles.”Footnote ⁸ In other words, on the occasion of this uprising against the rent that their landlords charged, the tenant farmers looked to the oracular judgments of these four deities as the basis for determining their course of action. Therefore, while the government authorities mounted a crackdown in the temporal world, going so far as to mobilize the army, they were compelled at the same time to tread also into the population's spiritual domain. In short, they needed to demonstrate concretely to the people that the Gewalt, or violent force of state authority in the temporal realm, was superior to the spiritual power of the deities. We must conclude that their actions attest to the profundity of the local peasants’ belief in these deities.

I imagine that this is the first time many readers will have encountered the four deities who figure in this episode: namely, Zongguan, Zhoushen, Mengjiang, and King Li. These are not the familiar “national gods” (quanguoshen 全國神) such as Guandi 闗帝, Mazu 媽祖, Dongyue 東嶽, Wenchangdijun 文昌帝君, and so on, who are universally worshipped across the Chinese world. They are all, instead, “local gods” (tushen 土神): regional gods who are limited to the Jiangnan Delta. The word tushen, which appears in historical documents, is sometimes confused with the “Earth God” (tudishen 土地神), the tutelary deity that protects a Chinese village, but this is a mistake. A tushen refers to a local god (difangshen 地方神) that arises in a particular area and therefore has a local legend associated with it and is worshipped in that area. To put it differently, “local god” (tushen) and “national god” (quanguoshen) are contrastive concepts. The scope of the region where local gods are worshipped varies widely; there are instances where a given local god is worshipped only in a single village, but the range also includes urban town (zhen 鎮)-level or rural township (xiang 鄉)-level local gods, county-level gods, and cases where the scope of the local god extends over a still-wider area, such as the entirety of the Jiangnan Delta. However, even though worship of these four gods extends over the whole Jiangnan Delta, from the perspective of the Chinese world as a whole they are little more than nameless local gods. Let us take a look at each of them individually.

Zongguan 總管.Footnote ⁹ Belief in the foremost of these four deities, Zongguan, had already taken shape in the late Yuan dynasty in Jiangyin County 江陰縣. People there worshipped Zongguan deities associated with the Chen 陳, Shen 沈, Wang 王, Li 李, and other families. The Jiangyin area had been an important base for the navy during the Southern Song dynasty, and even into the Yuan dynasty it continued to be a key center of maritime transport, a fact which was deeply connected to the name Zongguan, as we shall see.

In the Ming Dynasty, the existence of several more Zongguan deities can be ascertained, and of these, the Zongguan deity with the surname Jin 金, who emerged from the rural village of Jinjiatang 金家堂村 on the northern outskirts of Zhouzhuang Town 周莊鎮, Kunshan County 崑山縣, Suzhou Prefecture, became the most powerful. From then until the present day, the Jin Zongguan 金總管 deity was the paramount object of worship among all the Zongguan deities in the Jiangnan Delta region. The god was an anthropomorphic deity with the surname of Jin, and the given name of Yuanqi 元七. The powers of these Zongguan deities could be employed to protect against such threats as disease, flood, locusts, banditry, and external enemies. However, this range of practical benefits that worshipping Zongguan was said to yield gives no sense of the particular local characteristics of the Jiangnan Delta. Such benefits are sought by individuals from all regions within the Chinese agricultural world, where traditions of worshipping local gods in various districts inevitably entail stories of obtaining such practical benefits. But when we trace the various Zongguan deities from the late Yuan through the Jin Zonguan deity that appeared in the early Ming, we see that in addition to universal, generic stories of miraculous power, there emerged traditions unique to the Jiangnan Delta: legends that told of miraculous powers used to protect waterborne transport of grain to the capital (caoyun 漕運). The original meaning of the term caoyun is synonymous with shuiyun 水運 or “waterborne transport,” but from the Yuan dynasty onward, it came to be identified with state financial activity, namely the transport of taxes in kind that the state collected in the Yangtze River Valley to the political center and military bases in the North. These taxes were overwhelmingly in the form of rice and other grains, but also included domestic handicrafts such as cotton cloth, and, from the mid-Ming, also silver. In other words, while it was not unheard of to use the term caoyun to refer to other forms of waterborne transport carried out by private individuals or groups, such usage was extremely rare.

Ever since the Tang–Song transition, the most important source of revenue for state finances had been sought in the Jiangnan Delta of the Yangtze River Valley. The landholders of this region (called yehu 業戸) bore burdensome taxes, and powerful landlords were furthermore saddled with the responsibility for sending this revenue to the North. During the Yuan Dynasty, when the use of the Grand Canal had declined due to the longstanding division between North and South, the government turned to sea routes for caoyun, entrusting the duty to groups of maritime merchants (who were also pirates) in the Jiangnan Delta. We can conjecture that in carrying out seaborne transport, these squadrons of ships adopted an organization and command structure that resembled a naval fleet. In earlier times, back into the Tang Dynasty, high-level commanders within military organizations on a war footing were given titles such as “Great General Commander” (Da Zongguan 大總管), “General Commander of the Advancing Army” (Xingjun Zongguan 行軍總管), or simply “General Commander” (Zongguan 總管). The commanding officer in the Yuan dynasty ship squadron that carried out seaborne transport was thus also a Zongguan. In the Yuan shi 元史, men such as Zhu Qing 朱清 and Zhang Xuan 張瑄, who received commissions from the Yuan court to carry out seaborne transport, were called “Commander on the Ocean” (Shanghai Zongguan 上海總管). We can see here, in the commanding officers of these seaborne squadrons from the Yuan dynasty, the roots of the name “Zongguan” for various local gods of the Jiangnan Delta region who emerged in the late Yuan, with their unique, miraculous stories concerning protection of tax grain transport.Footnote ¹⁰

Zhoushen 周神. The second of the four gods, Zhoushen, has a narrower range of worship than the other three gods: fundamentally, he is a local god of Changshu County 常熟縣 (which naturally also included Zhaowen County). According to traditions passed down in the Ming and Qing periods, he is held to have originally been a low-ranking county government official during the Song Dynasty, with the surname Zhou 周 and the given name Rong 容. But by the Yuan Dynasty, he had already become a local god of Changshu. The crux of the legends surrounding him is that he was filial toward his mother, and that is why he is also known as “Filial Son Zhou” (Zhou xiaozi 周孝子). During the first years of the Hongwu (1368–1398) period, the administration of the first Ming emperor Zhu Yuanzhang 朱元璋 ordered the officials of prefectures, sub-prefectures, and counties to submit applications to the central authorities identifying gods who they thought should be entered in the register of sacrifices (sidian 祀典) and worshipped officially. Changshu County petitioned for Zhou Rong. After assessment by the Ministry of Rites, Zhoushen was formally enfeoffed as “the Filial Son Zhou Deity of Changshu” (Changshu Zhou xiaozi zhi shen 常熟周孝子之神), and was worshipped by the county magistrate twice annually in spring and autumn.Footnote ¹¹ However, though stories of filial piety are at the heart of traditions surrounding Zhoushen, stories have also accumulated about him providing protection for tax grain transport. In this way, Zhoushen manifests the shared distinctive feature of the local gods of the Jiangnan Delta.Footnote ¹²

Mengjiang 猛將. The third of the four local gods, Mengjiang (‘Fierce General’), also has a surname: in his case, Liu 劉. The circumstances surrounding this god are somewhat convoluted. According to local gazetteers, a god with the surname of Liu had come into existence in the Jiangnan area no later than the Hongzhi period (1488–1505) of the mid-Ming, and was worshipped by the populace. At the time, there were two gods with the surname of Liu: Liu Qi 劉錡 and Liu Rui 劉鋭, both of whom appear in the Song shi 宋史. The two deities rivaled each other in influence, but the populace created the legend that the two were brothers, thereby resolving the contradiction.

At the heart of Mengjiang's mysterious power are stories about the god driving away locusts, but in this case too, we also find the inclusion of Jiangnan-specific stories about the god's protection of tax grain transport.Footnote ¹³ However, during the Qing dynasty, a new and bothersome situation suddenly intruded. In the second year of Yongzheng (1724), an imperial order was issued to all prefectures, sub-prefectures, and counties in the realm, commanding local officials to build temples for the “Fierce General Liu” (Liu Mengjiangjun 劉猛將軍) and worship him as a deity able to drive off locusts. This deity was Liu Chengzhong 劉承忠. He was an ethnic Han general, but during the Yuan–Ming transition, he opposed Xu Da's 徐達 Northern Campaign and died in battle as a loyal servant of the Yuan Dynasty. This meant that a new deity with the surname Liu had suddenly burst onto the scene, one who in life had served with direct imperial authority no less. As subjects, the people had no choice but to comply. However, by this point the two gods Liu Qi and Liu Rui already had several hundred years of solid tradition behind them, and belief in these two gods continued to exist. From the Yongzheng period (1678–1735) onward, the various accounts written about gods with the surname Liu in the Jiangnan Delta could not conceal a certain puzzlement. On the one hand, “Liu Mengjiangjun or Liu Chengzhong” is a nationwide god produced by the state, and is thus not what we should properly call a “local god.” But on the other hand, when tenant farmers waged their rebellion in the late Qing against taxation policies, we may conjecture that the Mengjiang they believed in was in all likelihood Liu Rui or Liu Qi: that is, the unique local gods of the Jiangnan Delta.Footnote ¹⁴

King Li 李王. The last of the four gods, King Li, has the surname Li, the given name Lu 祿, and the courtesy name Fugong 福公. Like the other three gods, he is an anthropomorphic god with a human surname. Originally, King Li was nothing more than the Earth God (tudigong 土地公) worshipped in one village along the western bank of Lake Tai 太湖 in Changxing County 長興縣, Huzhou Prefecture. The legends that surrounded him as Earth God at first included such standard practical benefits as protection against disease, banditry, and natural disaster. He had no connection whatsoever with waterborne transport. At some point in the Yuan Dynasty, this god made his way across Lake Tai to the east, reaching Jiangyin on the bank of the Yangtze, where he was transformed into a protector of seaborne transport and acquired the titles Taiwei 太尉 (‘Grand Commandant’) and Zongguan.Footnote ¹⁵ He then came to Changshu, which borders the area to the south, and metamorphosed into the “Sea God King Li” (Haishen Liwang 海神李王), whose principal role was to protect seaborne transport. In Changshu at the time there lived an extremely prominent landlord named Cao 曹, whose yearly income from tenant farming was said to have exceeded 360,000 bushels (shi 石), and he was what we might call the primary patron of the King Li temple. We may safely surmise that this Cao of Changshu was one of those who bore responsibility for seaborne tax transport during the Yuan dynasty, and we may also deduce that his enormous wealth was intimately connected to this task. During the Yuan dynasty, the sub-prefecture of Changshu on the Yangtze River was a point of origin for transport by sea, and during the mid-Ming, a new sub-prefecture called Taicang 太倉州 (‘Grand Granary Sub-Prefecture’) was created by partitioning off parts of Changshu and other counties. The new sub-prefecture took its name from a group of large granaries for seaborne transport that existed there during the Yuan dynasty. In this way, a deity who had been simply the tutelary Earth God of Changxing in Huzhou Prefecture came to the banks of the Yangtze where he was transformed into a sea god who offered protection for tax grain transport.Footnote ¹⁶

* * *

In this way, the four gods of Changshu County, and the various gods of Jiangyin County, all have in common the following three points. First, each of them is an anthropomorphic god with a human surname and given name. Second, they all have descendants who bore the same surname. Third, they all have miraculous stories concerning the protection of tax grain transport associated with them. I will have more to say on this final point in the next section, but here I would like to consider the significance of the fact that these gods had descendants.

The fact that the various gods of Jiangyin County all have descendants bearing the same surname is discussed in detail by the author of the Zhengde (1506–1521) era Jiangyin xianzhi 江陰縣志, Huang Fu 黄傅, who was county magistrate during the Hongzhi era (1488–1505). It is highly likely that descendants have survived down to the present day, but I have not yet had the opportunity to carry out a field investigation in Jiangyin County.

Turning to the four gods of Changshu, it is first of all confirmed by various local gazetteers that descendants of Zhoushen existed in Changshu County. Presumably there are still descendants alive there today. In the course of several field investigations that I have undertaken in the last ten years, I was able to confirm with certainty that belief in Zhoushen has revived in Changshu.

It can be verified from various types of local gazetteers from the Ming and Qing period that the descendants of Jin Zongguan lived in the areas north of Zhouzhuang town, Kunshan County. From 1990 to the present, I have made repeated visits to Jinjiatang in Duibang village 对浜村 of this town, where I was able to confirm that his descendants are still living, though they are now impoverished farmers.

As for King Li, his descendants survive even today in Changxing County, where they live in Tongzhuangdian Village 童荘殿村, Houyang Township 后漾乡. In spite of official prohibitions against superstition, they have built a magnificent temple in his honor and are presently engaged in a game of cat-and-mouse with the authorities, who repeatedly destroy the temple only to have the locals rebuild it. I have determined that among residents with the surname of Li there are spirit mediums (wushi and shiniang 師娘), about whom I will have more to say later. In Ming-era Jiangyin County, there were also descendants of gods with the surname Li who had migrated from Changxing County. Presumably their descendants came to Changshu County where they propagated belief in gods with the surname Li. However, there is no record of their existence in written documents such as local gazetteers. I was unable to locate any descendants in the course of a brief field survey; whether any remain in Changshu is thus yet to be confirmed.

There are a variety of gods with the surname Liu, and it is therefore difficult to make any blanket statements about their descendants. At the very least, in the northern outskirts of Jiaxing, there is a King Liu 劉王 temple dedicated to Liu Mengjiang on the shore of Liansidang Lake 蓮泗蕩湖. Living there is a member of the Liu lineage who inherits the role of “god singer” (shengeshou 神歌手); passed down from generation to generation, the position entails inviting the god down and those who perform the task refer to themselves as “descendants of Liu Chengzhong.” The origins of the temple can be inferred to go all the way back to spirit mediums with the surname Liu who lived along Lake Poyang 鄱陽湖 in Jiangxi Province during the Song Dynasty. It is thought that the construction of the temple in Jiaxing took place during the Chongzhen era (1627–1644) of the late Ming, which means we can be certain that the god was already being worshipped on the shore of Liansidang Lake prior to the 1724 intrusion of Liu Chengzhong into the area. However, it remains unclear just when the Liu lineage religious practitioners created the legend of Liu Chengzhong and incorporated it into their genealogy.Footnote ¹⁷

Thus, the local gods of the Jiangnan Delta typically have descendants. Why should this be? But to frame the question in this manner is actually to get matters backwards. It is not that there is first a god (or a person who becomes a god), and thereafter come his descendants. Rather, first there are descendants and they create a god in the guise of their ancestor. To put it differently, without the descendants, the ancestor god would never have been produced.

Then what kind of people were these descendants who created the god? On this point as well, Huang Fu, the county magistrate of Jiangyin during the Hongzhi era who authored the Zhengde era (1506–1521) Jiangyin xianzhi, has provided us with precious information.Footnote ¹⁸ Born in Lanxi County 蘭渓縣 in Jinhua Prefecture 金華府, eastern Zhejiang, Huang Fu grew up as a direct successor to the faction of orthodox Zhu Xi Confucian thought that predominated in eastern Zhejiang, serving as a shidafu 士大夫 bureaucrat. He criticized superstition from the vantage point of orthodox thought on religious matters, and strove to exterminate improper religious cults (yinsi 淫祀). He investigated the origins of the local gods then worshipped in Jiangyin County, mincing no words as he leveled sharp criticism and denunciation. Huang Fu resolutely denied that these local gods were in fact “gods” (shen 神) at all. They were simply the spirits of the deceased, and thus no more than ghosts (gui 鬼). He therefore used the term “descendants of ghosts” (guisun 鬼孫) in reference to those people who called themselves descendants of gods, saying they were simply the descendants of perfectly ordinary dead people.

Who were these “descendants of ghosts”? Huang Fu puts the matter clearly by calling them “spirit-medium descendants” (wusun 巫孫), in other words, people who made their livelihood as shamans. Huang Fu provides a detailed account of these spirit mediums and their practices, one that enables us to discern that they were so-called “possession-type” shamans. When sought out by a client, they would enter a trance, and various gods and departed spirits would take possession of them and offer oracles.Footnote ¹⁹ In order to elevate the authority of the gods who took possession of them, the sorcerers would create miraculous legends about their own ancestors and relate these to the client. In other words, there were first of all shamans (or in some cases, female shamans known as shiniang) who served as professional religious workers, and the “gods” were created as they fashioned miraculous stories concerning their ancestors.

One shared characteristic of these “gods” is that the legends that surround them tell of righteous conduct during life ending in a tragic death. Among the “ghosts” who undergo this process, those that reveal miraculous power after death or provide benefits to people become “gods.” These two factors are all that is necessary for a “god” to come into being, but it is also desirable to secure the guaranty and recognition of human political authority. As a third factor, then, certification by the emperor, God's terrestrial agent, was sought.Footnote ²⁰ Ultimately, of course, not all gods could manage to satisfy this third condition, and thus it was not infrequent for titles and court ranks to be forged, in most cases by claiming that a title had been awarded by the preceding dynasty. The spirit medium was active in every step of this process. We can say that the kernel of the rural population's religious beliefs lay in this type of shamanism.

I have been able to verify the existence of Zongguan (Jin Zongguan), Mengjiang (Liu Mengjiang or King Liu), and King Li as three local gods unique to the Jiangnan Delta area. These three are worshipped throughout this area, but belief in them is almost non-existent outside the Delta. Though the associated sphere of belief in Zhoushen is more narrowly circumscribed within the Delta's interior, I have also been able to verify the existence of Zhoushen (also known as the “Filial Son Zhou”) as a local god unique to Changshu County in Suzhou Prefecture. In my research, I refer to belief in all these local gods collectively as “Belief in Zongguan,” taking the name of one representative god from their number.

This type of popular belief in local gods had its most basic support not in Daoism, nor in Buddhism, nor (needless to say) in Confucianism, but rather in its shamanistic component, for which the existence of a spirit medium (wushi) was indispensable. This most fundamental feature of belief in local gods was surely not limited only to the Jiangnan Delta region, but presumably is applicable to the entire Chinese world.Footnote ²¹

STORIES OF PROTECTING WATERBORNE TRANSPORT AND THE TAX CAPTAIN CLASS

On the premise that worship practices are one potential vector for creating a sense of social collectivity among residents of agricultural villages, I have examined the Jiangnan Delta, mainly during the Ming and Qing periods, and have confirmed a set of local god worship practices that can be collectively called “Belief in Zongguan.” The practices that fall under the “belief in Zongguan” label were born in the period from the late Yuan to the early Ming, and have one nearly universally shared characteristic: these local gods of the Jiangnan Delta invariably have associated with them some sort of miraculous story about providing protection for waterborne transport.

As mentioned earlier, the term caoyun refers to the “waterborne transport” by which northern dynastic authorities moved the taxes (particularly in the form of grain, as the term caomi 漕米, ‘transporting rice’ indicates) collected in the Yangtze River Valley (especially the Jiangnan Delta region) to the capital or military centers in the north.Footnote ²² During the Yuan Dynasty, when the function of the Grand Canal had deteriorated, the dynasty turned to seaborne transport of tax grain by depending upon groups of professional marine merchants who were also engaged in piracy. It is for this reason that among the patrons of local gods who have legends attesting to their protection of waterborne transport, we can find many great landlords whom we can infer accumulated their vast wealth through participation in seaborne transport. Seaborne transport entailed certain risks, and in order to carry it out smoothly and safely, these landlords sought the protection of the gods. To the individuals who made their living as spirit mediums, such large-scale landholders who also incurred huge risks were ideal clients. The spirit mediums needed to emphasize the efficacy of the gods who took possession of them. They therefore refashioned their ancestors into gods and created appropriate legends, in which the gods deployed their miraculous power to protect grain transport. For example, the ancestor god might figure in a narrative such as the following: late one night a grain transport ship encountered a wild storm at sea that tore apart its sail, swept away its paddles, and made it completely lose direction. But, amid a burst of brilliant light, the god appeared above the ship's masts, calmed the winds and waves, and guided the grain transport ship to its destination.

During the reign of the first Ming emperor Zhu Yuanzhang, when the capital was situated in the Prefecture of Nanjing Yingtian, grain transport from Jiangnan to Nanjing, which was known as “southern transport” (nanyun 南運), became much easier than the northward transport of the previous dynasty. According to Huang Fu, the author of the Zhengde-period (1506–1521) Jiangyin xianzhi, the various local gods in Jiangyin who had all boasted of their protection of seaborne grain transport during the Yuan dynasty, as well as the spirit mediums who had earned their livelihood from these same local gods, went into decline for a time at the beginning of the Ming period, only to regain their status in the Yongle (1402–1424) and Xuande (1425–1435) eras, after the transfer of the capital to Beijing. Needless to say, grain was once again transported from the Jiangnan Delta to the North, now via the Grand Canal. The social stratum of “tax captains” (liangzhang 糧長) in the Jiangnan Delta had already been put in charge of grain transport to Nanjing and other areas during the Hongwu era, and now the Ming court placed them in charge of grain transport by canal to the North.

The task of the tax captains was to provide a form of corvée labor unique to the Jiangnan Delta. In the beginning of the Hongwu period, their principal duty was to take charge of and lead the li 里 units and li heads (lizhang 里長) of several villages, assembling the tax grain that they had collected and transporting it to the county granaries. In addition, they were also occasionally given responsibility for the waterborne transport all the way to northern China. Naturally, this burden was extremely onerous, not something even mid-range landlords could manage, let alone typical peasants, and thus large-scale landlords assumed the responsibility. Presumably as recompense for their service, these large-scale landlords laid claim to a significant degree of de facto authority in rural society, including the right to exercise administrative jurisdiction. They constituted a ruling class within rural society, providing leadership to their communities on a wide range of matters, including water control and agriculture policy.Footnote ²³ In addition to leasing tenant farmers large tracts of land, these so-called “village-dwelling landlords” employed bonded servants (called jiaren 家人 in the Ming Dynasty), and also engaged in cultivation themselves.

These Jiangnan Delta village-dwelling landlords were required to bear an arduous allotment of work for the dynasty, but at the same time, because of this burden they were also able to strengthen their authority; they constituted a stratum that ruled and guided rural society. In the Jiangnan Delta, the rule of the Ming court was maintained because it functioned in a mutually complementary relationship with the control over rural villages by village-dwelling landlords of this type.

The Yuan-era Jiangnan Delta spirit mediums, in response to the demands and desires of their biggest clients, who were these large-scale landlords and maritime merchants who took charge of seaborne tax transport, created the tradition of worshipping the Zongguan Deity, which had at its center narratives of protecting waterborne transport. After the decline in the Hongwu period and the subsequent transfer of the capital to Beijing, the Jiangnan Delta's tax captain stratum would only have grown all the more keen for the smooth completion of their tax grain transport duties. Their fervor came first from the fact that the huge expenses required to carry out tax grain transport had to be borne by the tax captains themselves. Moreover, in the event that disaster struck, restitution for any loss was also the responsibility of the tax captain. It was by no means unusual for an unfortunate tax captain to lose his entire fortune and for his household to be completely devastated as a result.

From the Yongle period onward, the spirit mediums of the Jiangnan Delta followed in the footsteps of their predecessors in the Yuan Dynasty. In response to the fervent hopes of the village-dwelling landlords who were the leaders of the rural societies to which they belonged, many of these spirit mediums wove together miraculous stories involving the protection of tax grain transport. We can see this process clearly in the case of Jin Zongguan, the prototypical Zongguan god who was worshipped across the entire Jiangnan Delta area. The case of the spirit mediums of Kunshan with the surname Jin presents us with what is truly an extremely rare case where contemporary historical documentation survives concerning a member of the family who lived at the end of the Yuan. The reference appears in Kunshan junzhi 崑山郡志, a late-Yuan local gazetteer of Kunshan (a sub-prefecture in the Yuan Dynasty) compiled in the Zhizheng era (1341–1368). In the text's fifth section (on people, spirits, and temples), a contemporary living individual named Jin Yinglong 金應龍, a powerful spirit medium, is introduced as follows:

It is standard for local gazetteers to introduce living individuals in a given region, but unheard of for a “spirit” – a religious professional who acts as a medium – to be introduced in such concrete terms. I have examined countless local gazetteers over the course of the last fifty years and have never encountered another example; this local gazetteer of late Yuan Kunshan is exceptional. Clearly this is material concerning the Jin family of Kunshan and their associated Jin Zongguan deity, but what demands our attention is the fact that there is no mention whatsoever of waterborne transport (or seaborne transport, as it was called in the late Yuan) in the accounts of the deity's miraculous power. The spirit mediums of the Jin family in the late Yuan (presumably known as Yuanqi 元七 with the given name of Yinglong) transmitted oracles from the gods through spirit writing (shujiang and fuluan 扶鸞) and spirit possession (futuo), but the traditions of miraculous stories of protection for waterborne transport had not yet taken shape. The oldest mention of Jin Zongguan in an extant local gazetteer appears in Gusu zhi 姑蘇志, a Suzhou Prefecture gazetteer dating from the Zhengde era (1506–1521), which notes that a Jin-surnamed god (Jin Yuanqi) had already acquired the title Zongguan. By the Yongle and Xuande eras, the Jiangnan Delta social stratum consisting of tax captains and village-dwelling landlords was facing distinct hardships. We may conjecture that the Jin family shrewdly discerned the hopes of their rural society's ruling class, making up stories that featured their own ancestors manifesting miraculous power in the protection of tax grain transport during the Song and Yuan dynasties.

During the early Ming, Jiangnan Delta rural village society was under the control of the ruling class of village-dwelling landlords that was in charge of coordinating corvée labor as “tax captains.” Spirit mediums invented the Zongguan god in response to the needs and desires of this ruling class. The god of the ruling class became a ruling god. Belief in Zongguan spread across the entire Jiangnan Delta, transcending social class, until he became the local god of all residents of rural villages, the object of communal worship by the rural society. Still, though Zongguan became a god who was worshipped by residents of all social classes, he was essentially the god of the tax captain class of village-dwelling landlords. In other words, Zongguan was a god with a hierarchical dimension.

We must also bear in mind that belief in Zongguan, with its core of stories concerning the protection of tax grain transport, never took hold in other areas lacking the Jiangnan Delta system of tax captains and its attendant hardships, for the foundation was lacking. This is the reason that belief in Zongguan is limited to the five prefectures of Suzhou, Songjiang, Changzhou, Huzhou, and Jiaxing; with a very limited number of exceptions, it did not spread even to the three prefectures of the surrounding area (Yingtian [Nanjing], Zhenjiang, and Hangzhou). As the associated terms “tax captain” (liangzhang) and “tax grain transport” (caoyun) make clear, the gods of the Zongguan belief were characterized by both their hierarchical dimension and their strong regional identity; they were ultimately local gods of the Jiangnan Delta.

COMMERCIALIZATION AND THE REMOVAL OF LANDLORDS TO CITIES

The above discussion has demonstrated that, above all else, belief in Zongguan was established around a core of miraculous stories about the protection of tax grain transport. It responded to the religious needs of the tax captains or rural landlord stratum to whom responsibility was assigned for tax grain transport as a corvée obligation. Rural society at the time was controlled by these village-dwelling landlords, who in effect formed a ruling class there, and therefore the Zongguan cult spread throughout the Jiangnan Delta as a dominant religious practice in which the entire population took part.

From the Yuan dynasty through the mid-Ming, there were village-dwelling landlords who themselves engaged in agriculture in Jiangnan Delta rural society. They were the crux of the lijia 里甲 (‘hundreds and tithings’) system, meaning that they constituted the de facto resident ruling authority; at the same time they were obliged, in their capacity as “tax captains,” to serve state interests in overseeing the transport of tax grain. In other words, they were at once part of their own local society, which they led and controlled, while at the same time they bore a responsibility to state authorities to represent that society.

However, this social and economic structure underwent a massive transformation in the early to mid-sixteenth century. During this period, which corresponds to the Zhengde (1505–1521) and Jiajing (1521–1567) eras, the tax captain or village-dwelling landlord class collapsed and disappeared from rural village society. As a result, the de facto resident authority that they had constituted within the rural village disintegrated. Although elements of the wealthy stratum remained partially intact in some places, smallholding peasants began to have greater influence in Jiangnan Delta rural villages.

Let us first present an overview of the major transformation that took place in the mid-sixteenth century.Footnote ²⁶ If one were to find a single word to sum up this transformation, that word would be “commercialization” – the process of commercialization undergone by the rural villages. The causes of this transformation lay in a complex of internal factors originating within the Jiangnan Delta itself and external factors connected to global historical developments.

Internal factors. The internal dynamic driving economic development since the Tang–Song transition had been the opening up of lowlands lying south of the Yangtze River, especially the creation of lowland paddy (yutian) in the Jiangnan Delta. The process began by digging a meshed network of drainage canals called tangpu 塘浦 through areas that had been, as late as the Tang dynasty, uninhabited and undeveloped wetlands. In order to maintain good drainage, dredging was carried out every winter when water levels were low. The dredged river mud accumulated on both banks of the waterway and gradually an artificial landscape took shape there. The resulting land formation was a polder, or low-lying land surrounded by creeks and dikes, that was known in Chinese as yu 圩 or wei 圍, or sometimes as the binom yuwei 圩圍. The cultivated land created within the polder was called yutian 圩田 or weitian 圍田.Footnote ²⁷ This development of polder land was initiated by the Five Dynasties regime of Wuyue 呉越, and continued under the Northern and Southern Song, the Yuan, and then on into the Ming. From the tenth century onward, the lowlands of the Jiangnan region attracted large numbers of settlers, and the undeveloped lowlands gradually transformed into developed areas with settlements and cultivated lands. Before long, from about the end of the Northern Song period, it reached the point that the Jiangnan Delta had become the most fertile grain belt in the entire realm, and the most important source of revenue for state finances. The tax captains or village-dwelling landlord class who personally engaged in agriculture also sometimes doubled as landlords who spurred development by themselves creating polder land and irrigation systems.Footnote ²⁸

By the mid-Ming, however, there was less and less undeveloped lowland suitable for transformation into yutian. Lowland development approached a state of saturation as the potential profit became negligible in comparison to the costs of developing the land. We can see one sign of this in the phenomenon of “polder division” (fenyu 分圩): the splitting up of large tracts of polder (dayu 大圩) into small-scale tracts (xiaoyu 小圩). This practice aimed at more intensive land use, relying on engineering to compensate for the disappearance of suitable new land for reclamation. Very simply, it sought to drain the lowlands that remained within the heart of the polder, transforming them entirely into cultivated land. Polder division took place when the outward-extending development of polder reached its limit and land development turned inward, toward the interior of the polder enclosure. This division began as early as the mid-fifteenth century in the area around Suzhou, which was the region where lowland cultivation had first begun, and reached its conclusion in the mid-seventeenth century, when it was carried out in Qingpu County 青浦縣 in Songjiang Prefecture, the area of the Jiangnan Delta where the terrain lies lowest. In other words, polder division was the final stage in the development of polder land that had been carried out continuously in the Jiangnan Delta since the tenth century.

The end of polder division indicates that land reclamation in the lowlands had reached its limit. From the point of view of development history, this signified the disappearance of an “internal frontier” that had consistently attracted a large population to the area since the Five Dynasties era. The Jiangnan Delta thus now experienced an imbalance between resources and population. Historical documents make occasional references to land saturation and related population pressures, stating that “people are numerous while land is sparse” (ren duo di shao 人多地少). Inevitably, the land cultivated by each smallholder shrank. In general, in the Jiangnan Delta, 10 mu 畝 (about 60–70 ares or 1.5 acres), was the minimum size for a typical smallholding family to survive. By the late Ming and early Qing period, this became impossible, and the land area that smallholding farming families cultivated is estimated to have dwindled to between 5 and 7 mu. It therefore became necessary for them to supplement their livelihoods by some other means beyond agricultural cultivation. For affluent residents who maintained possession of wealth, by contrast, once this point of saturation was reached, they directed their investment to activities unconnected to land.

Skinner has identified regions in pre-modern China that turned to new forms of enterprise or investment once local land development ceased or reached saturation point.Footnote ²⁹ He termed this phenomenon “mobility strategy,” and provides examples of changes in the development and enterprises engaged in by residents of the coastal regions of eastern Zhejiang. This framework is extremely useful in considering the economy and society of the Jiangnan Delta in the sixteenth century. Once development reached saturation point and population pressures emerged, the residents of Jiangnan Delta rural villages were compelled to seek new outlets. This led to commercialization.

External causes. In addition to the above circumstances internal to the Jiangnan Delta itself, there was another major factor that the region faced at precisely the same time. It was just then that the Age of Discovery was getting underway, and trade activity was increasing in the maritime world of East Asia too. In the early stages, large amounts of Japanese silver began to enter China, followed by immense quantities of Mexican silver that flowed through Luzon into Fujian. The availability of silver constituted a significant foundation for commercialization in the Jiangnan Delta in the mid-sixteenth century and beyond.

* * *

Commercialization brought about a profound transformation in the economy and society of the sixteenth-century Jiangnan Delta.The livelihood of Jiangnan Delta rural village residents changed in the following manner.

In addition to maintaining their traditional wet rice cultivation, smallholding farmers began to engage in subsidiary occupations, pursuing various cottage industries in accordance with the particular handicrafts that had traditionally existed in their area. In the lowlands of the western Delta (known as dixiang 低鄉 or xixiang 西鄉), sericulture and silk reeling became the principal side occupations among the households of smallholding peasants, while in the eastern highlands (known as gaoxiang 高鄉 or dongxiang 東鄉) the dominant side occupations were cotton-based industries such as thread and cloth weaving. Unlike the traditional domestic handicrafts, these side industries were not intended to provide goods for domestic consumption; rather, they produced fully commercial products, intended for sale at market.

How then did the village-dwelling landlord stratum, which had ruled over Jiangnan rural village society since the Song dynasty, respond to this situation? In the western lowlands of the Jiangnan Delta, where sericulture and silk reeling were dominant, mulberry field cultivation required a minimum three years’ investment to realize a profit (from the initial planting of mulberry seedlings to the first harvest of mulberry leaves). This was not possible for smallholding peasants, who lacked sufficient land. Rather, it was common for wealthy landlords who engaged in cultivation to invest capital in mulberry field management and for the smallholding peasants to purchase mulberry leaves from them for sericulture. Moreover, the work of past scholars has shown that it was customary for smallholding peasants to make advance arrangements in the winter of the previous year to purchase mulberry leaves and thereby guard against a shortfall, a practice known as shaoye 梢葉 among other names. The Zhuang Zhongfu zazhu 莊忠甫雜著 (“Miscellaneous works by Zhuang Zhongfu”), a collection of writings by a late-Ming gentry resident of Huzhou now held in the National Library of China, provides us with insight into this practice. Upon passing the highest level of the civil service exam as a “presented scholar” (jinshi 進士), Zhuang Yuanchen 莊元臣 (that is, Zhongfu) left his village behind, transferring his residence to Nanxun 南潯鎮, a large market town in Huzhou. Though this shidafu had completely abandoned direct paddy field cultivation, he continued to engage in personal management of 20 mu (about 3.3 acres) of mulberry fields located around his former rural residence, hiring three long-term laborers (changgong 長工) under the supervision of an “overseer” (jigang 紀綱). Among the ledgers that Zhuang kept in connection with his household finances, I was able to confirm the existence of an “Advance leaf purchases ledger” (Shaoyebu 梢葉簿, that is, a ledger recording reservations for the advance sale of mulberry leaves) alongside the “Ledger of leaf sales” (Maiyebu 賣葉簿).Footnote ³⁰

In addition to their foray into commercialized agriculture, the rural landlords of the Jiangnan region during the mid-Ming, roughly during the Chenghua (1465–1487) to Zhengde (1505–1521) eras, engaged in other economic activities which merit our attention. In this period, we begin to see many village-dwelling landlords’ engraved epitaphs and other biographical accounts recording their wide-ranging travels to various areas. What this travel in fact represents is these landlords’ long-distance commercial activity.Footnote ³¹ Moreover, not only do we find isolated biographical materials recording such travel as the behavior of specific individuals, but there are also historical documents that attest to it as a widespread phenomenon. Earlier in this article, I mentioned Huang Fu, who served as magistrate of Jiangyin County (Changzhou Prefecture) during the Hongzhi era, and in Jiangyin xianzhi, which he compiled in the Zhengde era text, he addresses contemporary social phenomena in Jiangyin County, dividing merchants into three categories: wealthy merchants (fushang 富商), poor merchants (pinshang 貧商), and “brokers” (jingji 經紀 or yahang 牙行). Of these three, he writes the following about the “wealthy” merchants:

The times of year when these individuals traveled to Huguang 湖廣, Sichuan, Henan, and other places to engage in trade corresponded with the agricultural off-seasons in the Jiangnan Delta region, where a two-crop system was carried out with wet rice cultivation as the dominant crop. This demonstrates that the individuals who made these journeys were the very same resident landlords who were directly engaged in cultivation. Moreover, the particular commodities that they traded are a vivid testament to the commercialization and urbanization of the Jiangnan Delta, as the cotton grown provided the raw material for cottage industries, the charcoal served as a fuel for urban residents, and the sesame cake was fertilizer for commercial crops.

The traveling merchant activities that we see among the village-dwelling landlord class of the Jiangnan Delta in the mid-Ming were not seen at all in the early Ming. In the late Ming too, they disappeared once the commercial activities of Huizhou merchants expanded into the Jiangnan Delta. At this point, the only group of merchants from the Jiangnan Delta who still practiced traveling commercial activity were the Dongting merchants, introduced to us through the work of Fu Yiling.Footnote ³³ Why should all other Jiangnan merchants engaging in long-distance commercial activity have disappeared by the end of the Ming?

The answer is suggested by the location of the Dongting merchants. Their homeland, eastern and western Dongting, consisted of two islands in Lake Taihu. As topographical maps of the area show, there is little flat land there, for steep inclines rise up around the lake. The area is an exception within the rich Jiangnan Delta region, for there is no industry or enterprise to speak of beyond fishing on the lake and citrus cultivation on the steep hillsides.

The three most prominent groups of traveling merchants to rise up during the late Ming were known as the Huishang 徽商, Jinshang 晉商, and Minshang 閩商, names deriving from the regions whence they came, Huizhou, Shanxi, and Fujian, respectively, and, significantly, none of these regions is particularly well suited to agriculture. We might therefore say that these regions were pursuing a “mobility strategy” at a region-wide level when they began to specialize in traveling trade, and in the same way, the people of Dongting discovered a means of ensuring their livelihood in long-distance trade. By contrast, the wealthy across the rest of the Jiangnan Delta, weighing the costs and risks involved, would surely have seen no need to take the trouble to travel to faraway lands, with all the opportunities for profit-making and investment within the Delta itself: in finance, brokerage and wholesale, commodity storage, housing, or any of a variety of urban handicrafts. We may infer that even as the Huizhou merchants made inroads, the wealthy from Jiangnan simultaneously withdrew from long-distance trade activities, devoting themselves instead to safer and more lucrative activities within the Jiangnan Delta itself.

As a result of this kind of commercialization, various changes could be seen within Jiangnan Delta society, two of which are especially significant. First, there was a change in the landlords’ mode of existence, as they began to dwell in the cities. Second, there was a change in the everyday life space of the peasantry. Both of these changes were intimately connected to the main topic of this article: belief in Zongguan. I will address the second change in the following section, and would like to close this section by discussing the changes we see in the landlord class.

In reaction to the population saturation that accompanied the end of land reclamation, the tax captain class, who were the village-dwelling landlords that constituted the ruling class of rural society, began to engage in traveling commercial activity. However, as I have discussed, during the Jiajing period of the late Ming (1521–1567), such traveling commercial activity on the part of Jiangnan merchants disappeared again, with the minor exception of the Dongting merchants. In the lowlands of the western Delta, the landlords who remained in the rural villages and continued to engage directly in cultivation also discovered a lucrative livelihood in commercial agriculture through the cultivation of mulberry.Footnote ³⁴

As a whole, however, the fundamental trend was for village-dwelling landlord households either to fall into decline or to depart the villages.Footnote ³⁵ Naturally, this also meant the abandonment of direct cultivation. What emerged instead across the whole of the Jiangnan Delta was the city-dwelling landlord (chengju dizhu 城居地主), who held property on a large scale and did not personally engage in agriculture at all.

These city-dwelling landlords accumulated large amounts of land through purely economic activity such as the purchase or foreclosure of mortgaged paddy land when owners defaulted on high-interest loans. Yet what is important at the same time is that the majority of these city-dwelling landlords were “local gentry” (xiangshen 鄉紳) who enjoyed a variety of special privileges, notably exemption from corvée. In the state's household registries, they were designated “official households” (guanhu 官戸), distinguished from commoner households. In the five prefectures of Jiangnan, such official households enjoyed the special privilege of “favored exemption” (youmian 優免) which freed them from all corvée duties. It was standard practice after the promulgation of the “ancestral institution” (zuzhi 祖制) system of Emperor Hongwu for bureaucrats also to be assigned proper duties within the lijia system including responsibility for serving as heads of their li (lizhang) and as tax captains. Yet the five prefectures of Jiangnan were an exception, for there alone it was the local custom for the official households to be entirely relieved of all such responsibilities. The total amount of corvée due in a single county was fixed as the “ancestral amount” (zu'e 祖額), which meant that if one household obtained an exemption, then the responsibility to provide their labor would shift to other households. This practice also accelerated the collapse of commoner landlords unable to qualify for guanhu status.

Many of these urban-dwelling landlords also belonged to the social stratum of gentry (xiangshen). The term and concept of xiangshen developed naturally in Jiangnan civil society from the Jiajing era onward, signifying individuals or households who had qualified to hold office by passing the civil service exam or through other means. In this sense, there was no difference between the connotations of the term xiangshen and those of shidafu that had taken shape after the Tang–Song transition. Essentially, a social stratum that was nearly identical to the Song and later shidafu class widely came to be referred to by the new name xiangshen. Underlying this change in nomenclature was an emerging social and political problem: official households in the five prefectures of Jiangnan were deliberately exploiting their special privileges of exemption to amass land. More and more frequently, smallholding peasants who feared the burden of corvée duties “donated” their land to members of the gentry and became their tenant farmers (dianhu 佃戸). Some peasants even sold themselves into servitude, reducing their status to tenant-serfs (dianpu 佃僕) of the gentry. Another increasingly common phenomenon was called “fraudulent trusteeship” (guiji 詭寄) and involved members of the wealthy farmer and commoner landlord class nominally surrendering their lands to the gentry. By falsely registering their lands as guanhu land, they could escape corvée duties.Footnote ³⁶ This is one of the causes behind the Jiangnan Delta's “one field two owners” (yi tian liang zhu 一田兩主) pattern of land ownership.

The transition from village-dwelling to urban-dwelling landlords completely dismantled the function of the tax captain stratum within rural society.Footnote ³⁷ The onerous tax transport responsibilities that they had fulfilled for the state underwent reforms in the mid- to late Ming period, as “military transport” (junyun 軍運) carried out by the transport corps (yunjun 運軍) began to replace civil corvée labor, and the expenses associated with such military conveyance came to be collected as a surtax.Footnote ³⁸

In summary, as commercialization advanced in the late Ming, the rural-dwelling landlord class disappeared, the result of which was that legends of protecting tax grain transport became meaningless to the residents of the Jiangnan Delta.

MARKET TOWNS AND THEIR SURROUNDING VILLAGES

A major social transformation that resulted from the progress of commercialization in the late Ming was a transformation of the lifestyle patterns of smallholding farming households in Jiangnan Delta rural villages, a development that consequently expanded the sphere of these peasants’ daily lives. We have seen above that as land reclamation ceased and population reached saturation, the area of land cultivated by each household shrank; it became impossible for them to subsist solely on the basis of cultivation. Historically, the Jiangnan Delta had been a region where various traditional handicrafts had been produced, and this accumulated technical knowhow provided these families with the foundation to pursue a variety of cottage industries during the agricultural slack season, obtaining currency that they used to support their households. Research into the cottage industries of Jiangnan Delta rural villages, which began in the 1940s with the pioneering studies of Nishijima Sadao 西嶋定生, has shown that in the highlands of the eastern Delta, cotton came to be widely cultivated by smallholding peasant households, while in the lowlands of the western Delta, sericulture and silk reeling became dominant.

In Delta highlands such as Songjiang Prefecture and the eastern part of Suzhou Prefecture, cotton was cultivated in the summer along with wet rice. However, because of the prosperity of the Songjiang cotton industry, it was impossible to fulfill demand relying solely on locally grown cotton, so that large quantities of raw cotton, chiefly from northern China, were brought into the Jiangnan Delta by traveling merchants from Huizhou and elsewhere. Nishijima Sadao and others have elucidated the minutely divided sequence of steps by which the crop of raw cotton was processed. After the step of ginning (yahe 軋核) removes seeds from the raw cotton, the ginned cotton is then spun into cotton thread, which can then be woven into cloth. The smallholding peasant families of the Songjiang region specialized by taking charge of different steps in the process. The weaving of thread and the weaving of cloth were independent processes, each of which was divided into a sequence of discrete steps. Specialization in modern industry is a result of the deliberate rationalization of the production sequence, whereas the minute division of labor that we see in the Songjiang cotton industry stemmed instead from the scarcity of capital that these individual peasant families could invest. In this way, specialist families participated in only one particular step in the entire process that transformed raw cotton into products such as cotton thread and cotton cloth; each step in the process began with the purchase of materials, and the products of that step were immediately brought to market, in return for which the peasant received currency. And, although the cotton industry produced a simple commercial product, it could only be established once a distribution system was in place.

As the work of Tanaka Masatoshi 田中正俊 and Saeki Yūichi 佐伯有一 has proven, the existence of a distribution system was also an indispensable precondition for the establishment of the sericulture and silk reeling industries in the lowlands. Technical factors meant that sericulture as practiced by smallholding peasant families was inseparable from silk reeling, and thus cocoons did not become a discrete commodity. Sericulture began with the purchase of silkworm-egg cards, and required the constant purchase of fodder in the form of mulberry leaves. The physiology of mulberry trees necessitates their replanting every ten years or so, but a minimum of three years is required from the time of planting mulberry seedlings to the first harvest of leaves. It was impossible for peasant families who had little or no land to let their invested capital lie idle for three years before realizing any return. Therefore, as mentioned before, the stratum of wealthy farming families took charge of the management of mulberry fields, and the smallholding peasant families had to purchase leaves from them annually. As soon as the silkworms formed cocoons, the smallholding peasant families immediately began to reel the silk thread. It was extremely rare for them to weave silk cloth themselves; instead they took the silk thread directly to market.

In this way, cottage industries had become economic activities indispensable to the very survival of these peasant families. Characterized as the system was by the purchase of raw or partially processed materials and the sale of products at each step along the way, it depended on the existence of routes of distribution for its functioning. Furthermore, smallholding families obtained a minuscule harvest from the paddy fields they cultivated, from which they then had to pay their public taxes and land rent. The rice that remained was insufficient for their annual food needs. In particular, they needed to acquire rice from elsewhere during the pre-harvest months of early summer.Footnote ³⁹ This demand was met with Huguang rice cultivated in the plain of the middle Yangtze valley. An important commodity transported into the Jiangnan Delta by merchants from Huizhou, this rice became a significant food source not only for the lower stratum of urban residents of the Jiangnan Delta, but also for the Delta's smallholding peasant families who themselves carried out wet rice agriculture.

Research by many scholars has shown that market towns (市鎭 shizhen) proliferated explosively from the Jiajing period (1521–1567) onward, increasing steadily every year through the mid-Qing.Footnote ⁴⁰

We can discover in the formation of such market towns, as well as in the development of increasingly intricate networks linking them with villages, geographical phenomena that epitomize the Jiangnan Delta's commercialization during the late Ming. The cottage industries in which smallholding peasant families were engaged likewise developed in a mutually complementary manner with the market towns that acted as nodes in a distribution network. The household head of a smallholding peasant family would set out for the closest market town every day, carrying products that he would sell to a broker in exchange for currency.Footnote ⁴¹ Before turning home, he would use the currency he had obtained to purchase raw or partially processed materials and perhaps rice for his family's consumption. As it became routine for him to travel back and forth between his village of residence and the nearby market town, an individual smallholder's sphere of activity expanded beyond the confines of his village to include the market town.

Both historical documents and modern field investigations confirm that peasants of the Jiangnan Delta typically traveled back and forth to one specific market town; visits to multiple market towns were unusual. As a result, a coherent regional unit would take shape around a single market town and encompass the rural village settlements on its periphery. The following historical document provides eloquent testimony to this phenomenon:

Through the sales and purchases made by residents of rural village settlements, a fixed regional entity was established with a single market town at its core. Each such regional entity arose naturally through trade, that is through buying and selling, and as such we can think of it as a type of market area, one without any connection to governmental or administrative divisions.Footnote ⁴³

Fei Xiaotong has discussed these regional entities that took shape in Jiangnan Delta rural villages, pointing out the importance of market towns in the modern-day Jiangnan Delta. In his report, entitled “Xiao chengzhen, da wenti” 小城镇大问题 (“Small cities and towns, big problems”), he draws on his own experiences and observations of his Wujiang County 呉江縣 hometown, proposing an analogy in which the market area is a single cell, the market town forming its nucleus and the rural village settlements on the periphery forming the cytoplasm.Footnote ⁴⁴ According to the Republican era Xincheng zhenzhi quoted above, village settlements were identified in relation to the closest market town. According to Fei, such clusters of rural village settlements centered around a single market town were known in the dialect of Wujiang County as xiangjiao (鄉脚; lit. “country feet”). I will refer to the entire regional entity including the market town at its center as xiangjiao.Footnote ⁴⁵

Fei Xiatong, sociologist and anthropologist, did not explore the historical origins of these rural village units that still existed in the Jiangnan Delta in the twentieth century. However, we historians can connect Fei's analysis of twentieth-century conditions to historical research into the society and economy of Ming–Qing era Jiangnan. We cannot be mistaken in understanding the xiangjiao as a historical development which occurred in conjunction with the advance of commercialization in the late Ming. The sphere of activity for smallholding peasant families, whose very existence was inseparable from the circulation of goods in a network in which the market towns were nodes, expanded to the scope of the xiangjiao.

What was the sphere of activity like for the smallholding peasant families of the Jiangnan Delta prior to commercialization, that is to say, prior to the appearance of the xiangjiao as regional entities?

We must begin to approach this question by considering it first from the point of view of settlement geography: what sort of settlements did they live in? Unfortunately, there is a lamentable lack of research on settlement geography in mainland China from the late imperial to modern period. This gap is related to the surprising lack of interest in this question among mainland Chinese researchers. Human geographers elsewhere in East Asia, as in Japan, Taiwan, and Korea, have devoted considerable energy toward researching the settlement geography of rural villages, and socioeconomic historians have also paid much attention to this research, utilizing its findings in their own work.Footnote ⁴⁶ When it comes to mainland China, however, there is simply no body of work on settlement geography that socioeconomic historians would find useful.Footnote ⁴⁷ Settlement geography is a discipline in which we would naturally expect researchers from the country in question to have the advantage in elucidating phenomena. Faced, however, with the complete absence of any research by Chinese scholars on settlement geography in Jiangnan Delta rural villages, I have had no choice but to approach this topic myself, relying on extremely limited written documents and on-site field investigations.

In the eastern highlands, we see sparsely populated scattered villages (it is not at all uncommon to encounter villages with only a single isolated residence), while in the western lowlands, we can clearly perceive a tendency toward clustered villages.Footnote ⁴⁸ Even within the restricted area of the Jiangnan Delta, a slight topographical difference of a few meters in elevation can produce a great change in the form and scale of settlements. What sort of unified terms exist to refer collectively to living spaces in which such differences exist? On the basis of the scarce historical documentation available, we can infer that it is appropriate to call them she 社 or shecun 社村.Footnote ⁴⁹

The term she originally referred to the deities that protected the settlements of the Chinese people. As is well known, the protective deities of these settlements from primitive times through antiquity were these she that apotheosized natural forces. Because the she deities were natural gods, they existed alongside the qi 氣 of the natural world, and thus it was not necessary for them to fear natural phenomena such as wind and rain, sunlight, and thunder. The she deities were therefore enshrined at open altars (tan 壇; or lishetan 里社壇) completely exposed to the elements. In the period from the end of the Eastern Han into the Three Kingdoms era, the Daoist religion came into being, a process that took place under the influence of Buddhism. And moreover, under Daoist influence, a change occurred in the form of worship in settlements. While the term she was continued, its meaning completely changed. People came to worship a particular anthropomorphic god as the protective deity of the settlement. The she gods came to be called “earth gods” (tudishen), and since many of them were also imagined to be old men, they also came to be called tudigong (“Earth grandpas”). They were individuals who were believed to have once lived in the human realm. For this reason, they required protection from wind and rain, sunlight, and thunderstorms. The place of worship thus changed from the exposed altar (tan) to the temple (miao 廟; or tudimiao 土地廟, “Earth [god] temple”) with a roof and four walls. This custom is still alive in the twenty-first century; it has been continued as a nearly universal worship practice in settlements throughout the Chinese world down to the present day.

It is this geographic area with a she temple or “Earth God temple” at its center that has formed the most fundamental unit of social cohesiveness and sense of community in the Chinese world. In the Yuan period, the village social organization that functioned as the terminus of regional rule was the she. In the early Ming as well, records from the period of creation of the lijia system in the early Hongwu period show that the terms li 里 and lishe 里社 were used interchangeably. Both the she system of the Yuan and the lijia system of the Ming were clearly founded on social units with such Earth God temples at their center.

In this way, Earth God temples became a universal feature of life in the Chinese world from medieval times, but how was worship practiced in the Jiangnan Delta? As I have already discussed, there was a major difference in the form of rural village settlements in the Jiangnan Delta between the scattered villages of the highlands and the clustered villages of the lowlands. Written documents such as local gazetteers were edited from the point of view of elite intellectuals who lived in the cities and towns, and it was therefore rare for them to make any concrete reference to the structure of settlements in rural villages or to discuss the form of the she temples or Earth God temples where the lower-status commoners worshipped. Since the late 1980s, I have spent a total of about ten years carrying out field research, and I would like to offer my answer to this question below.

In the clustered village areas of the lowlands, the median number of houses is approximately one hundred, and typically a single Earth God temple is associated with one settlement or shared by multiple settlements.Footnote ⁵⁰ Fei Xiaotong's 1939 Peasant Life in China is a famous ethnography of rural village life in the Jiangnan Delta, and at the time of his field survey in the mid-1930s, the village that is the object of his analysis, Kaixiangong Village 開弦弓村 in Wujiang County, had two Earth God temples. However, with its population of 360 households, this village is exceptionally large for the Jiangnan Delta.Footnote ⁵¹ We cannot take it as typical or normative.Footnote ⁵² Each Earth God temple has a territory over which the god offers protection, and this means that the local residents who worship that Earth God are linked by their membership of the same “temple sphere” (miaojie 廟界). In the lowlands, the territory of a single Earth God temple is expressed as the aggregate of the villages that worship there.

In the scattered village areas of the highlands, by contrast, the scale of each settlement is extremely small; there are many settlements with only a few residences, or in some cases just a single residence. It is no wonder, then, that it would be impossible to establish the temple sphere of the Earth God temple on the basis of settlements, for “settlements” are not part of local consciousness. Instead, the unit referred to as a tu 圖 determines the temple sphere.Footnote ⁵³ The tu is a geographical subdivision, and in the Jiangnan Delta, before the township system (xiang and zhen) was implemented at the end of the Qing, each county was divided into regions called du 都 (or sometimes qu 區 or bao 保); the subdivision below the du was the tu.Footnote ⁵⁴

While there was thus a distinction between the western lowlands and eastern highlands in terms of whether the settlement or the tu was the basic unit that determined its extent, each Earth God temple (she temple) had a particular scope, and we may infer that the social unit that arose there was called a she. Before smallholding peasants came to have an intimate connection with market towns in the late Ming, the sphere of their daily lives was limited to the scope of this she. However, after this change, the development of cottage industries enlarged their sphere of actvity. The peasants headed out to the market town carrying the products of their cottage industry, and once they had sold them and made purchases, they would invariably stop by the teahouse (chaguan 茶館) where for a few copper coins they could drink tea, nibble on watermelon seeds, and enjoy conversing with other peasants like themselves who had come to the market town from all directions. The range of the smallholding peasants’ life activity and the sphere of their knowledge thus came to extend beyond the boundaries of the she, expanding to the bounds of the xiangjiao.

TOWN-LEVEL “CITY GOD TEMPLES”, THE “DELIVERING THE TAX” CUSTOM, AND LEGENDS OF RICE DISPENSATION

It is inconceivable that this transformation of the smallholding peasant household's economic structure and the enlargement of its members’ daily life sphere would have had no impact on their communal beliefs and religious practices. In this last section, I would like to analyze two major shifts: a change in the structure of their worship; and a transformation of the various local gods they worshipped (that is, a change in the content of their belief).

1. Changes in worship structure. From the late Ming into the Qing, a religious phenomenon that deserves our attention occurred in the market towns of the Jiangnan Delta: the appearance of town-level City God temples (zhen chenghuangmiao 鎮城隍廟). We might define the City God temple as an Earth God temple located in an urban settlement that is enclosed by city walls.Footnote ⁵⁵ It is impossible to pinpoint the precise period when they emerged, but they seem to have appeared at roughly the same time as the Earth God temples. That is, the City God temples that were first born in the period from the end of the Eastern Han to the Three Kingdoms era became increasingly widespread as time went on, and by the end of the Yuan, they were a universal feature of all walled cities in the Chinese world. The cheng 城 in Chenghuangmiao means ‘wall’, and the huang 隍 refers to the ‘dry moat’ that surrounds it (moats that are filled with water, by contrast, are called hao 濠). In other words, the “wall” and the “dry moat” were both originally elements of a settlement's defense apparatus. Because all Chinese cities were walled cities surrounded by moats, the word chenghuang took on the sense of “city.” From this point, the meaning expanded further, and came to also indicate the City God (Chenghuangshen 城隍神), who was: (1) the protective god of the city. Because almost all of the cities that have existed in the Chinese world over time were also centers of government administration, the scope of the City God was not limited to the city proper; the term designating the City God also acquired the meaning of (2) the god that manages the administrative district. An analogy with magistrates of the earthly realm allowed the City God to acquire the corresponding status of an administrator of the underworld or spirit world. And because administrators throughout Chinese history have also simultaneously served as judicial officers, the City God was also transformed into (3) an underworld judge. In the Chinese world today, the City God is occasionally worshipped outside City God temples, and in these cases, it is principally in this third capacity.Footnote ⁵⁶

The evolution of the City Gods, from their genesis to the expansion of their functions, arose through spontaneous processes from among the people. Occasionally we see instances in which court authorities awarded titles of nobility to individual City Gods, but through the end of the Yuan, they were not incorporated into the court's state system of ritual. In 1369 at the beginning of the Ming, for the first time in Chinese history, the City God system entered the state's ritual system when it was stipulated that each county seat would establish a City God temple where the county magistrate would worship twice yearly, in spring and in autumn.Footnote ⁵⁷

Of course, according to the established regulations of the state, the City God temple could only be established within the county seats. Strictly speaking, these “town-level City God temples” were thus a form of “improper worship” (yinsi) that violated the ritual systems of the state. However, the market towns had grown to the point where their scale (in terms of population, economic power, and other factors) was on a par with the county seats, or in some cases even exceeded them. They began to seek a symbol that befitted their status as central places.Footnote ⁵⁸ What the county seat had but the market town lacked was a City God temple. Ignoring state regulations, the people of the market town constructed a City God temple as a means of displaying the settlement's newly elevated status.Footnote ⁵⁹

It has already been more than thirty years since I first took notice of these town-level City God temples. At the time, I started to daydream: a longstanding debate in Chinese historical studies has existed, especially among Western scholars, about whether or not “cities” actually exist in China. The market towns had elevated the status of their settlements through economic activity, and ignored state restrictions in building City God temples. Within these market towns that proliferated in the late sixteenth century, I wondered, is there not something with the characteristics of a “city” as an organization (Verband) possessing both autonomy and autocephaly?Footnote ⁶⁰ Would this not serve as evidence that the “city” existed in China too? This was the biggest motivation that drove me to research City God temples at the town level. When I actually embarked upon concrete consideration of the issue, however, my romanticism was soon dashed. It turned out that in the vast majority of cases, the principal gods worshipped at the town-level City God temples were not independent gods of that town, but rather City Gods of the relevant county who had been invited to the town. In addition, there were numerous cases in which an area that would be regarded as a single market town from a geographic perspective in fact straddled the administrative divisions of several counties. In such situations, I invariably found that multiple town-level City God temples had been constructed, each corresponding to a single county, and in each the City God of that county was enshrined as the principal god. Such phenomena clearly demonstrated to me that these market towns lacked the consciousness of being a “city” as a single social organization. Rather than aspiring to independence, autonomy, and autocephaly as a “city,” their orientation was in the direction of fitting themselves into the bottom rung of an administrative hierarchy that descended from the capital (the emperor) down through sequential levels of provincial seat, prefectural seat, and county seat. This is not something that should be entirely surprising to us, considering the special hierarchical characteristics of Chinese bureaucratic society. Rather, we might instead say that such characteristics manifest themselves in the town-level City God's very nature.

While the town-level City God temple shows this sort of subordination to those in superior positions to itself, with regard to the rural villages on its periphery, it stands in a position of superiority, seeming to extend this hierarchy of control that descends from the emperor outward to the xiangjiao. This hierarchical relationship is evident in a custom known as “delivering the tax” (jie qianliang 解錢糧), which we can infer appeared at the same time that the town-level City God temples were established. As we have seen, the sphere of daily life for smallholding peasants underwent an expansion beyond the confines of the settlement or natural village in which they lived (or beyond the confines of the she of which the settlement was part), coming to be coextensive with the world of the xiangjiao that was centered around a market town. We can readily imagine that this new historical phenomenon of the expansion of the daily life sphere of the smallholding peasant family would exert some kind of influence upon the structure of their communal belief and worship practices. One local gazetteer reports the appearance of the new custom, which began during the Shunzhi era (1643–1661) of the early Qing:

As market towns in the Jiangnan Delta developed in the late Ming, town-level City God temples began to appear, and as this passage clearly demonstrates, they stood in a superior position to the various village Earth God temples, from which they received payments of paper coin and courtesy visits. I refer to the town-level City God temple that was the recipient of the courtesy visit as the “superior temple” and the various Earth God temples (she temples) that made such expressions of homage and payments of paper coin, the “subordinate temples.” The metaphoric significance of the Earth God temples’ submission of tribute to the town-level City God temple is worth noting, for these actions can be seen as analogous to state authorities’ levying of taxes: a process by which tax captains and lizhang collect taxes from taxpayers and then remit them to the county offices. Moreover, it can be determined that the beginning of this custom in Kunshan County was during the Shunzhi era: in other words the late Ming–early Qing period. This religious custom of “delivering the tax” was the spontaneous creation of the peasants.Footnote ⁶² It is possible to find many examples of these sorts of customs in the local gazetteers and other types of materials from this region during the Qing dynasty. I have also been able to confirm them through on-site field surveys that I have carried out from the 1980s into the 1990s.

My discussion above has shown how in response to changes in the lifestyle patterns of smallholding families that accompanied the commercialization of the late Ming, there were corresponding changes in the structure of worship, including the emergence of town-level City God temples and the custom of “delivering the tax”.

2. Transformation of the various gods. Zongguan and other local gods of the Jiangnan Delta were originally gods who responded to the desires of the stratum of tax captains and village-dwelling landlords, who had been put in charge of transporting tax grain (caoyun). As we have seen, protection of such waterborne transport lay at the heart of their miraculous powers. When tax grain transport passed from the hands of the citizenry to the military, the responsibility of the tax captains to carry out this form of corvée generally disappeared. Moreover, amid the commercialization of the time, village-dwelling landlords themselves disappeared from the villages. Some wealthy farmers remained and continued to have influence, but in general village society became one in which the peasants were dominant. Once this point had been reached, the residents of the rural village had no connection to tax grain transport, and we can surmise that legends of its miraculous protection would no longer inspire interest or any sense of devotion in them.

Nevertheless, the tenant-farmer stratum continued to believe in local gods such as Zongguan even into the nineteenth century, as the anti-rent riot that erupted in Zhaowen (Changshu) County during the Daoguang era affirms. Nor was it just the nineteenth century; it can be confirmed that belief in Jin Zongguan, King Li, Liu Mengjiang, and other gods still existed in the Jiangnan Delta during the latter half of the twentieth century. Gu Xijia 顾希佳 has produced a distinguished body of research that collects and analyzes the folklore of Zhejiang Province. In 1990, he published the results of his careful efforts to gather the “god songs” (shenge 神歌) of the Hangzhou-Jiaxing-Huzhou region.Footnote ⁶³ According to his research, during worship rituals in temples, the three highest-status gods – “The tathagata of the three ages” (Sanshi Rulai 三世如來), “The God of the Three Heavens” (Santian Shangdi 三天上帝), and “The Jade Emperor” (Yuhuang Shangdi 玉皇上帝) – are first summoned down by the “god songs” to join the “supreme banquet table” (gaoyan 高筵) at the altar. Next, the City God descends to the “main banquet table” (zhengyan 正筵), followed by the “Southern Dynasty Holies” (nanchao shengzhong 南朝聖衆) in second place, and then King Liu, Jinfu Zongguan 金府總管, and King Li descend in third, fourth, and fifth place. According to Gu's explanation, the “Southern Dynasty Holies” in the second position is a collective term for a variety of local gods; the term is unrelated to the “Southern and Northern Dynasties” (AD 420–589). The three gods of the “supreme banquet” are likely the supreme deities in Buddhism, Confucianism, and Daoism; they are not part of the category of “local gods.” Worth noting is that of the four gods of Changshu that I introduced at the outset as unique local gods of the Jiangnan Delta, we can see three of them appearing together here, the one exception being Zhoushen, who is a local god specific to Changshu County.Footnote ⁶⁴ We can see that Zongguan, King Li, and Liu Mengjiang survived the commercialization of the late Ming, then also survived the anti-superstition policies of the Chinese Communist Party in the mid-twentieth century, and continue to be vital parts of the local religious scene today. Even the Changshu County-specific fourth god, Zhoushen, who does not appear in Gu's description, is now undergoing a revival.

When the resident landlords who had controlled rural village society disappeared in the late Ming, it meant that the clients of belief in Zongguan also vanished. So why should Zongguan and other local gods have survived? Why should these gods have captured the devotion of smallholding peasant families and tenant farmers? Written materials reveal nothing on this question. By the mid-1980s, I had amassed a substantial store of documents concerning belief in Zongguan, bit by bit constructing a logical historical argument that would account for the shared beliefs and religious practices of peasants in the Jiangnan Delta. But this question was the one missing link in the chain of logic. Zongguan and the other local gods of the Jiangnan Delta weathered the massive transformations of the late Ming, but once legends of offering miraculous protection for the transport of tax grain became meaningless, what sort of miraculous stories were told about them?

Precious information leapt out at me from the work of Gu Xijia and other folklorists who participated in efforts to gather oral materials in Zhejiang.Footnote ⁶⁵ A study of Zhejiang customs was published as a volume in the “Brief Survey” series of Zhejiang Province, and it introduces several miraculous stories concerning Jin Zongguan and a variety of other local gods. These stories had survived long enough to be gathered as part of a fact-finding survey that was carried out in the 1980s.Footnote ⁶⁶ The miraculous stories that had become associated with Jin Zongguan and other local gods were mainly as follows:

Tax grain transport remains, in this account, a vestige of the original miraculous stories associated with grain transport protection. However, by this point Zongguan is no longer a god who suddenly descends from the clouds to rescue the tax captains who have encountered disaster. The god is now a low-ranking military officer, though still involved in the transport of tax grain, who extends a hand to his hometown in the Jiangnan Delta, offering smallholding peasant families help in securing their most urgent need: a supply of rice. These local gods have undergone a metamorphosis from guardians of tax grain transport to gods who provide rice (shimishen 施米神). As we have seen, smallholding peasant families of the Jiangnan Delta began to purchase rice for their own dietary needs in the wake of commercialization in the late Ming. The new belief in Zongguan had now acquired a legend of sacrificing oneself to release transported tax grain and thereby save the starving people. This is a god that exactly responds to the needs of the smallholding peasant. The disappearance of tax captains and village-residing landlords in the late Ming meant that the local spirit mediums, religious professionals who worked as possession-type shamans, lost their former clients. They had no choice but to transform the smallholding peasant households who were numerically dominant in the rural villages into their new clients. They were able to extend the life of the gods and their own work as spirit mediums by creating new miraculous stories about the dispensation of rice that responded to the deepest needs of these people. When the tenant farmers mounted an anti-rent riot against their landlords in 1846, the four gods whom they consulted before deciding to act were the very rice-dispensing gods who were responsive to these pressing needs of the common peasantry.