Why China Had No Autonomous CitiesThe Procedures by Which Urban and Rural Groups in Qing Hsinchu Obtained Standing

In comparing Chinese and Western society, Max Weber asked why modern capitalism and a burgher class emerged only in the West. The medieval Western city was one of the conditions he proposed: the city had fortifications, a market, its own court and at least partly its own law, and could act as a corporate body. Great cities such as Genoa and Cologne were originally founded when the burghers swore oaths to one another and formed a sworn brotherhood (Latin conjuratio). He asserted that Chinese cities lacked precisely these conditions. The residents of a Chinese city belonged legally to their lineage, and through the lineage to their native village; the city itself had no standing to conclude contracts or bring lawsuits in the name of a corporate body, whereas the village did, by virtue of the village temple. This judgment later came to be widely used to explain why China had no autonomous cities.

I test this judgment against yamen documents from Qing-dynasty Hsinchu. The documents show that the headmen of the city gates and the headmen of villages obtained their standing by exactly the same procedure: nomination by the community, verification by the authorities, and the issue of a seal stamp; nor was the village temple at the center of this procedure. Weber's statement that China had no urban commune of the Western kind holds. But his inference from it that groups in Chinese cities could not act in their own name, and his drawing of the line between those who could act and those who could not between city and village, do not hold. Whether a group could come forward in its own name depended on whether the authorities appointed it or accepted its documents, not on whether it was in the city or in the countryside.

In the twenty-second year of Daoguang, Zhou Bangzheng, headman (zongli 总理) of the East Gate of Hsinchu, returned to his native place on the mainland because his mother was ill. Each of the four gates of Hsinchu city, east, west, south, and north, had one headman. The headman was nominated locally and appointed by the Tamsui subprefecture (Danshui fenfu 淡水分府, the yamen that then governed Hsinchu); it was a local public office. When conducting business, the headman stamped his documents with a seal stamp (chuoji 戳记) issued by the subprefecture. A seal stamp was a seal granted by the authorities to a local post, and a document bearing it was thereby shown to issue from that post. After the East Gate post fell vacant, the headmen of the other three gates jointly submitted a petition to the subprefecture nominating Lin Yangfang as successor. The subprefecture endorsed it with three characters, "hou yan chong" (候验充), meaning that he was to await verification by the authorities before being permitted to take up the post. Lin Yangfang then submitted a declaration of acceptance (renchong zhuang 认充状), the document in which a new appointee states his willingness to take on the post; the three gate headmen submitted a bond of guarantee (baojie zhuang 保结状), the document in which the nominators stand surety for him; and the subprefecture summoned and verified him and issued him a letter of appointment (yutie 谕帖) and a seal stamp. The following year the North Gate headman died, and "Jin Changhe and others, guild and shop households of this city" jointly elected Guo Shangmao as successor, submitting a petition with a bond of guarantee attached; the subprefecture endorsed it: "already verified and appointed." A jiao 郊 (merchant guild) was a group formed by the city's merchants according to their line of trade; shop households (puhu 铺户) were the merchants who kept shops. In the same year Hong Deliang, the West Gate headman, was disabled and unable to carry out his duties, and the East Gate headman Lin Yangfang, together with twenty-five shop households including the Baoyuan firm and the Jichang firm, jointly stood guarantee for Lin Cheng'en as successor. In Daoguang 24 the North Gate post fell vacant again, and the headmen of the South, East, and West Gates, "together with the Changhe firm of all the guilds and shops," stood guarantee for Wang Lirang; the bond of guarantee bears the seal stamps of the three gate headmen and one seal reading "Changhe gongji" (长和公记). A gongji (公记) was a seal that a group carved for itself, not one issued by the authorities.

These documents are in the first collection of the administrative section of the Danxin archives (淡新档案, Tan-Hsin archives), items 325 to 339. The Danxin archives are the official papers preserved by the two yamen of Hsinchu and Tamsui. These items record the procedure by which the headmen of the county seat's gates were replaced: joint nomination (qianju 佥举), acceptance (renchong 认充), guarantee (baojie 保结), verification and appointment (yanchong 验充), and issue of the stamp (geichuo 给戳). In the same archives, the villages used the same procedure to replace their headmen. These documents can therefore be used to test, point by point, Weber's three judgments about Chinese cities and villages.

I. What Weber Judged about Chinese Cities and Villages

Weber held that Chinese cities could not act in the name of a corporate body, while villages could. He said so three times in the first volume of his Gesammelte Aufsätze zur Religionssoziologie (Collected Essays in the Sociology of Religion), in the study known in English as The Religion of China. The first is on page 293: the city did not possess, even formally, the capacity to conclude contracts, to bring lawsuits, or in general to come forward as a corporate body (korporativ aufzutreten), whereas the village did. The second is on pages 382 to 383: through the village temple, the village possessed the capacity to act as a community, both legally and in fact; the government recognized only the gentry as the village's representatives, and the gentry acted in the name of the temple. The third is on page 382: the government was unwilling to let any group, or any substitute for a group, acquire legal standing. [Note 1]

By "formally" he meant legal standing, not whether anyone in fact acted on the city's behalf. He said this more clearly elsewhere. On page 522 of the chapter The City, he lists five criteria for "the city community in the full sense": fortifications, a market, its own court and at least partly its own law, a corporate character, and at least partial autonomy together with authorities in whose appointment the burghers took part; the bearer of these rights was a distinct burgher estate. He then judges that Chinese cities possessed not one of these: city residents belonged legally to their lineage, and through the lineage to the native village where the ancestral temple stood; China had no citizenship in the ancient or medieval sense, and the corporate character of the city itself was unknown. [Note 2]

To test these judgments, one must see whether the urban and rural groups in the documents came forward to the authorities in the name of a group, under what name they did so, and how the authorities treated them.

II. Under What Names Sixty Group Documents in the Hsinchu Records Came Forward

In the first collection of the administrative section of the Danxin archives, one can count, item by item, how often urban and rural groups submitted documents to the authorities in their own name, the names used in those documents, and how the authorities endorsed them. The first collection of the administrative section contains 475 items, divided into 478 records; of the 383 records whose year can be determined, 328 date from the Tongzhi and Guangxu reigns, 53 from the Daoguang and Xianfeng reigns, and one each from the Jiaqing and Qianlong reigns. Each record is coded for seven fields: whether the document was submitted by the authorities, an individual, or a group; which kind of group it was, cheng 城 (walled town), jie 街 (street/market quarter), zhuang 庄 (village), jiao, pu 铺 (shop/ward), miao 庙 (temple), zu 族 (lineage), or ju 局 (bureau); the name written on the document; whether the document was a petition, a nomination and guarantee, a covenant, or an undertaking; whether the authorities approved it, left it pending, rejected it, or gave no endorsement; who the addressee was when the authorities issued a document; and whether the matter belonged to the level of the walled town, the street, the village, or the county. The coding, that is, reading each record and classifying it under these fields, was carried out independently twice, and each disagreement between the two passes was adjudicated individually. [Note 3]

Of the 478 records, 378 were issued by the authorities or submitted by yamen runners in their official capacity, 40 were submitted by individuals, and 60 by groups. The sixty are listed below by kind of group:

Because a single document may be assigned to two kinds of group, the total in the table exceeds sixty. Of the group documents, 48 received an endorsement: 22 approved, 22 pending, and 4 rejected; of the individual documents, 35 received an endorsement: 21 approved, 9 pending, and 5 rejected. Between groups coming forward and individuals coming forward, the authorities' endorsements show no clear difference.

There are 36 group documents from villages. The name written on the document is headman in 12; in 17 it is shengyuan, jiansheng, linsheng, or gongsheng (all scholars holding examination degrees or school status), functionaries (zhiyuan), or gentry, gentry managers, or gentry elders (local gentry and elders of standing); in 4 it is zhuangzheng 庄正 (a village-level local post); in 6 it is shop households, shop residents, or shop tenants; in 5 it is reclamation households, landholders, or proprietors (those who opened land for cultivation and those who owned it); in 4 it is village elders or rural elders; a single document often carries several names. The number giving a palace, temple, or shrine (gong, miao, ci) is zero. Of the 36, 24 nominate or guarantee someone to serve as headman, zhuangzheng, director, or lineage head; 5 are inter-village covenants, that is, agreements concluded jointly by several villages, signed jointly by the headman, the heads and deputy heads of each village, the gentry elders, and the shop tenants, with the provision "apart from one copy deposited in the official file, each village keeps one copy."

The two documents in which a group came forward in the name of a temple or a charitable hall were both from outside the villages. One is a petition by Lin Buyun and other directors of the Wenchang Shrine of Huwei street, submitted because garrison soldiers were obstructing repairs to the shrine; the other is a reply from the gentry managers of the Mingshan Hall in Hsinchu city, who, on the instruction of the subprefecture, had sold off surplus timber and applied the proceeds to repairing the earthen city wall, and reported afterwards.

When the authorities issued documents, the addressee was a group in 22 items, written as the headman of a given village, the headman of a given street and village, the guild households, the gentry of the Dengyun Society, or the lineage head; in 2 items the addressee was written as a temple or hall. There are two directives (yu 谕, the document by which the authorities issued orders) to the guild household Jin Changhe: one ordering that half of the guild's goods-carrying be handled by the head of the official porters, the other ordering the guild households and the guild clerk to elect a head of porters. There is also one to the gentry managers of the Mingshan Hall, and one to Lin Rumei and other gentry of the Dengyun Society, ordering them to pay the Dengyun Society's interest silver.

III. Results of Testing Weber's Three Judgments against the Hsinchu Documents

Setting Weber's three judgments against the Hsinchu documents sentence by sentence gives this result: that China had no urban commune holds; that groups in the city lacked the capacity to act, and that the village's capacity to act came from the village temple, do not hold; that the government was unwilling to let any group acquire legal standing, taken literally, also does not hold.

Weber's first sentence says that the city could not, even formally, act in the name of a corporate body. These documents do not support this statement, but two matters should be considered separately. The gate headman was a local public office appointed by the authorities, and joint nomination, acceptance, guarantee, verification and appointment, and issue of the stamp made up the same procedure used for village headmen; the guilds and shops within the city came forward with seals they had made themselves, such as the "Changhe gongji," and the authorities accepted these as well. Groups at the city level and at the village level both obtained the capacity to act through the same legal form; the capacity to act here means the ability to issue, in one's own name, documents valid before the authorities. The line Weber drew between city and village does not exist in these documents.

But Hsinchu had no urban commune in the sense of Weber's The City, that is, no city community in the full sense: it had no law or court of its own, no burgher estate, and no city-level body within the walls above the gate headmen; the government of the city was the county magistrate himself. Weber's judgment on "whether there was an urban commune" therefore holds; his judgment on "whether groups in the city had the capacity to act" does not. He conflated these two matters.

Weber's second sentence says that the village had the capacity to act as a group, by virtue of the village temple. The first half holds; the second half does not hold in these documents. Of the 36 group documents from villages, not one uses the temple as its name; those submitting them are headmen, gentry and degree holders (shenjin 绅衿, gentry and scholars with examination degrees or school status), zhuangzheng, reclamation households, tenants, and shop households, signing jointly. The two documents in the name of a temple come instead from a street and from the city. However, the administrative section contains documents on appointments to public posts and on public business; documents on temple repairs, offering rituals (communal rites of sacrifice to the gods), and temple property are mostly in other sections; items 1, 4, and 21 are petitions concerning temple repairs, which shows that such affairs did exist. When villages acted in the name of a group in administrative matters, they used the names of public posts and of gentry and elders signing jointly, not the temple.

Weber's third sentence says that the government was unwilling to let any group acquire legal standing. Taken literally, this sentence does not hold. The authorities recognized groups not through a single law laying down the standing of associations, but case by case, chiefly in two forms. For local posts, that is, posts such as headman, zhuangzheng, and lineage head that were nominated locally and recognized by the authorities, the authorities gave recognition through verification and appointment and through the grant of official seals (yinxin 印信, the seals of various grades issued by the authorities): they issued a directive and stamp to the lineage head Chen Zongqi, stamps to the heads and deputy heads of the twenty-four villages under Houlong, and directives and stamps to the headmen of the various streets and villages. For merchant groups such as the jiao, the authorities granted no seal, but they accepted documents bearing the group's own corporate seal and issued orders addressed to the group under its collective firm name. The former is legal appointment, the latter de facto recognition; both enabled groups to issue, in their own name, documents valid before the authorities.

Lin Yu-ju had already pointed out the difference between these two kinds of treatment in 1998. She used the same Danxin archives, and she too wrote about Zhuqian, that is, Hsinchu. On page 52 she writes that for merchant groups such as the jiao the authorities "permitted them to register the seals representing their organizations. However, fearing that these organizations would monopolize the market and extort from shops, the authorities refused to endorse their founding and their collection of public fees, or to give them legal recognition, namely the issue of stamps and proclamations"; on the same page, discussing local-governance personnel, she writes that the guild's censer master "did not, like the local-governance personnel of the streets and villages, such as baozhang, headmen, xiangzhang, and zhuangzheng, need to pass through the four steps of guaranteed petition for appointment (the guarantors submitting a bond of guarantee), approval of appointment (the authorities issuing a directive and stamp), acceptance (the new appointee submitting a declaration of acceptance), and official proclamation to the locality." On page 51 she writes how the corporate seal entered the official record: merchants submitted the wording of the seal to the authorities to be noted in the files, so that other firms could not use it fraudulently, a step she calls registration with the authorities. On page 73 she writes of the authorities' de facto recognition: "Local government often recognized the existence of the jiao in fact; commercial disputes concerning the locality were mostly ordered by directive to be mediated by the jiao." [Note 12]

I am not the first to draw this distinction. What these archives add to the distinction she had already explained is quantity: under what names the 60 group documents among the 478 records came forward and how they were distributed, and, point by point, where the procedures for local posts and for merchant groups agreed and where they differed. She writes about the general procedure for the local-governance personnel of streets and villages; she does not treat the gate headmen separately, nor does she compare city and village. That the gate headmen and the village headmen went through the same procedure is something I found by comparing these documents.

In the Qing system of official seals, the stamp was the lowest grade: principal officials (chief officials such as county magistrates and prefects) used yin seals and guanfang seals, assistant and miscellaneous officials (lower officials who assisted the principal officials) used qianji seals, and local posts and runners used stamps. [Note 4] Once a group had obtained a stamp, it could use it to submit petitions, conclude covenants, and issue bonds of guarantee.

Weber took the history of European cities as his reference and looked in China for a burgher estate and an urban commune. Not finding them, he turned "no urban commune" into "the city has no corporate capacity to act," and attributed the village's capacity to act to the ancestral temple he had read about in sinological works. In the Hsinchu documents, the capacity to act of groups in both village and city came from appointment or acceptance by the authorities, not from a commune, and not from a temple.

IV. Earlier Debates Tested Only Weber's Judgment about the City

For a hundred years, scholars who rebutted Weber's account of the Chinese city addressed only his sentence about the city. William Rowe's two-volume study of Hankou, published in 1984 and 1989, argued that in the nineteenth century the guilds and benevolent halls of Hankou in fact took on the city's public affairs, and that an urban identity formed in that century. The reviewer Bianco called the two volumes a "démolition" (demolition) of Weber's theses, and Bergère said that Rowe "went against him point by point." Frederic Wakeman in 1993 weakened Rowe's rebuttal, on the grounds that the founders of those guilds came from elsewhere and could not count as local self-government. [Note 5]

Both sides argued over how much autonomy Hankou's groups had in fact, but neither discussed whether Hankou, as a corporate body, had standing under Qing law to conclude contracts and bring lawsuits. Bergère even stresses that this power was only de facto: "Ce pouvoir de facto n'en demeure pas moins extrêmement vulnérable." (This de facto power nonetheless remains extremely vulnerable.) [Note 6]

David Faure's reassessment of Weber in 2013 identified more precisely where Weber went wrong. He confirmed half of Weber's judgment: Chinese cities were not independent of the state, did not develop their own law, and were in effect seats of government. He also overturned the other half: Weber did discuss the economic functions of the lineage, but held that these functions operated only in the village and not in the city; "Weber was right that capitalism in China did not take root in the city, but he was wrong in thinking that only in the city could it take root."

His judgment on the question of legal persons also agrees with the Hsinchu documents. A legal person means a group able to hold property and conclude contracts in its own name. In the Ming and Qing, groups formed on the basis of locality and kinship operated through three kinds of written instrument, account books, contracts, and genealogies; "Together, the instruments were sufficient to create legal persons," and property could be held in the name of an ancestor or of an association; but "Because no law protected the funds as such, they existed through the protection of local political powers." That is, these groups functioned as legal persons in practice, but the law did not establish their status as legal persons. The stamps and corporate seals in the Hsinchu documents are the concrete documentary form of this "protection of local political powers." [Note 7]

Faure holds that Weber judged wrongly because "he was fed the wrong facts," that is, the sinology of his day supplied him with wrong material. [Note 8] Weber's original text held that the village had a corporate capacity to act and explained this capacity by the village temple; this exception clause was lost in the English translation and in later restatements. The debate of the following hundred years thus concentrated on the sentence about the city, and no one tested the sentence about the village temple against evidence. None of the nine studies I had read earlier took the village temple as its subject. [Note 9] These nine studies were gathered under the topic of Chinese cities and merchant groups; none took the countryside or ritual as its topic, so none would have discussed the village temple in any case, and for four of them I had not, at the time, obtained a single page of the main text. Studies that take the village temple as their subject do exist: Okada Yuzuru's 1938 survey of the ritual spheres of Shilin street in Taipei prefecture, a ritual sphere being the area over which the residents who jointly worship the same principal deity are distributed; Shih Chen-min's 1973 survey of settlements on the Changhua plain; and Lin Mei-rong's 1987 survey of local organization in Caotun township. All three take the village temple as their subject, but none is among the nine, none cites Weber, and still less does any of them discuss under what name the village temple came forward in official documents.

Earlier scholars have already shown with evidence that the village temple could be a property-holding subject, that is, a party holding property in its own name. Drawing on the official temple registers of the Japanese colonial period, the religious registers, Lin Mei-rong writes that much temple property was registered in the name of the deity and that deities in Taiwan had the status of legal persons (p. 58; for the point about legal-person status she cites Tai Yen-hui's Local Governance in Qing Taiwan, pp. 193–202, a book of which I did not obtain a single page). [Note 13] Lin Yu-ju writes that the authorities "explicitly ordered the protection" of the jiao's "public property that took the form of temple property," and that the properties belonging to the Changhe Temple were recorded in the official archives (p. 66). The Hsinchu and Tamsui documents can test another matter: whether villages came forward to the authorities in the name of a temple. Among the 36 group documents from villages, not one is in the name of a temple.

Three mutually independent sources show that people in towns also acted in the name of temples. Faure writes that the town was not a self-governing municipality and did not develop a set of commercial institutions distinguishing it from the rest of society; rather, it did its best to present itself as a collection of villages; the governance of towns, and especially the weight that temples and their managing bodies carried in it, underwent over the centuries the same changes as the villages: "the governance of towns, especially the prominence that temples and temple committees played in them, was subject to the same changes over the centuries as were experienced by the villages" (p. 76). Hsu Chia-ming, surveying the Changhua plain in 1973, placed the village temple and the merchant guilds of the towns in the same class of institution: "Like the merchant guilds in the towns, it became a center of local defense and self-government" (p. 187). [Note 13] The third is the item Lin Yu-ju read in the Danxin archives: the authorities protected the public property that the jiao held in the form of temple property and recorded the properties of the Changhe Temple on file. The three sources use different methods: Faure writes an institutional-historical synthesis, Hsu Chia-ming does field research on the Changhua plain, and Lin Yu-ju does a case study of the Danxin archives. Weber held that the temple was what enabled the village to act as a community, and said that the city could not do so; all three sources show that holding property in the name of a temple and coming forward in the name of a temple were not peculiar to villages, and that groups in towns used them too. In Zhuqian, taking Lin Yu-ju's study and these Hsinchu documents together, the situation is exactly the reverse of Weber's account: those who held property in the name of a temple and left an official record of it were the jiao in the city. These three sources can overturn only one claim, namely that only villages could use the name of a temple; their periods, places, and materials all differ, and none of them counts how many cases there were in city and in village. Whether villages used the name of a temple more than cities did is a question of proportion, which none of the three answers; only this batch of Hsinchu documents can answer it, and it covers only Hsinchu, with the group documents dating from between Daoguang 22 and Guangxu 20.

V. Under What Names Groups Erected Inscribed Steles

Whether the conclusions drawn from the Hsinchu documents are merely the product of a bias in official documents as a kind of material can be checked against stele inscriptions. Official documents were kept in the yamen archive room, and groups appear in them only when they did business with the authorities. Stele inscriptions were carved at the expense of groups and set up in front of temples or at street entrances; those who contributed money and came forward are named with their titles, and the location of fields, the amounts of rent, and regulations are often carved as well; the authorities occupy only a secondary place in the inscriptions. Each body of material has its own bias: official documents contain only affairs that passed through the yamen, and stele inscriptions only groups able to erect steles. To use them to judge the same question, one must see whether the results from the two sides agree.

The three sources, the compendia of stele inscriptions of southern and central Taiwan and the education stele inscriptions of Taiwan, contain, after duplicate entries are removed, 567 inscriptions in all, dating from Kangxi 24 to Guangxu 21. Each inscription was coded by the same scheme for the erecting party, kind of collective, name, action, official involvement, group property-holding, and location, and was coded twice. Inscriptions are harder to code than official documents: an official document states at the outset who is submitting the petition, whereas for an inscription one often has to judge from three places together, the signatures at the end, the narrative in the middle, and the editor's notes, and one person is often at once a director, a member of the gentry, and a shop household of some street. The two codings agreed only 69 percent of the time in the kind-of-collective cell; a cell is one field in the coding table, so that kind of collective and location are each one cell. Across the six cells, the two codings disagreed in 520 cells, which I adjudicated cell by cell by reading the steles; in 59 of these both codings were wrong and I assigned a different code. The figures used below are all those after adjudication. [Note 11]

The distribution in the inscription corpus is the reverse of that in the document corpus. Who erected the stele: in the inscription corpus, officials 116, individuals 44, groups 377, undeterminable 30; in the document corpus, officials 378, individuals 40, groups 60. In the yamen archives groups make up one-eighth, on the steles two-thirds. This difference cannot show whether groups in the Qing were in fact many or few; it shows only that the kind of material used determines which groups can be seen. What kind of group (377 inscriptions; one inscription may belong to several kinds): miao 167, zhuang 87, shen 65, jie 41, ju 31, jiao 28, zhong 众 (the collective) 26, pu 15, fan 番 (indigenous peoples) 11, zu 10, cheng 8. How the authorities appear on the stele: no authorities on the stele 210; officials donating silver or presiding over construction in an official capacity 156; the authorities granting a proclamation, or registration on file and engraving in stone, 130; the authorities issuing their own proclamation 47; officials only composing the text or the heading 24.

Only eight steles were erected in the name of the city. Five are merit and good-governance steles praising local officials; those praised are three county magistrates, one of them a subprefect concurrently acting as county magistrate, Jiang Yunxun, the Taiwan prefect who repaired the embankments, and Shi Lang, the general who pacified Taiwan. Two are petitions by the gentry of the whole county about matters concerning their own class, such as false registration of native place for the examinations and extortion of gentry and degree holders by clerks and runners. One is the Tiangong Altar of the prefectural city. Only two steles state that property belonged to a city-level group, namely the Tiangong Altar and the regulations stele on false registration. Steles erected in the name of jie, jiao, and pu total 75; 25 are within the prefectural or county walled cities, and 44 are in port towns such as Yanshuigang, Lukang, Beidou, and Anping, with the prefectural city having the most. The result agrees with the official documents: the "city" as a whole came forward almost solely to praise officials or to petition on behalf of the gentry; those who came forward for a piece of property or a construction project were groups inside and outside the city organized by street and by trade, and the authorities accepted their petitions and had the outcome carved in stone. Weber treated the "city" as a possible actor and then denied that it had the capacity to act; but in these two bodies of material, the city level was not an actor in the first place.

Steles were mostly erected for temples, so if Weber's claim that villages acted by virtue of the village temple were true, it should be easiest to see in the inscriptions. Of the 87 steles erected in the name of a zhuang, only 25, under 30 percent, also used the name of a temple; the remaining 62 used the set of names village multitude, village elders, reclamation households, gentry managers within the bao, and headman. In the document corpus, the number of the 36 village documents that came forward in the name of a temple is zero. Official documents are biased toward yamen affairs and inscriptions toward groups able to erect steles; the two bodies of material have opposite biases, yet they yield the same result. In my judgment Weber was right that villages could act as groups, and wrong that this capacity came from the village temple: when villages came forward in the name of a group, they used the names of the village multitude, the village elders, and the headman, and the temple was only a name that some villages used.

The inscriptions also allow a fuller test of Weber's third sentence. Of the 130 steles on which the authorities granted a proclamation or registration and engraving in stone, the petitioning party was a group in 105, or 80 percent; among these 105, zhuang are the most numerous at 47, followed by shen 19, miao 17, and jie 16, so that what the authorities recognized was not limited to the name of a temple. Steles stating explicitly that property belonged to a particular temple, street, village, guild, academy, or lineage number 161, or 28 percent of the total; of these, 65 assign it to temples and 31 to villages. There are 36 steles on which a group petitioned the authorities, the authorities approved and had it carved in stone, and the stele also states what fields, shops, and houses the group held and the amounts of rent. The authorities not only accepted the documents of groups but recognized them again and again by granting proclamations, approving registration, and approving engraving in stone.

What the authorities approved was "this property belongs to this temple, or to the people of this street, and no one may encroach on it," not "this group has standing to conclude contracts and bring lawsuits." Nowhere on the steles is it written that a group as such possessed any standing; what is written is always a particular matter, a particular property, a particular petition. Faure judged that these groups functioned as legal persons in practice, but that the law did not establish their status as legal persons; these 567 inscriptions directly support this judgment: groups functioned as legal persons in one concrete matter after another, and the authorities never wrote the standing of groups into a generally applicable rule.

Counted in three periods, the share of all steles erected by groups was 55 percent in the Kangxi to Qianlong reigns, 72 percent in the Jiaqing and Daoguang reigns, and 70 percent in the Xianfeng to Guangxu reigns. The share of group-erected steles rose in the Jiaqing and Daoguang reigns and did not fall thereafter. Groups erecting steles in their own name was not something that began only at the end of the Qing: in the Kangxi to Qianlong reigns it already accounted for 55 percent. Weber spoke of China without restricting himself to a dynasty; the share was above 50 percent in all three periods, which shows that groups erecting steles under their own name was the norm throughout the whole period, not a practice that appeared only in the late Qing.

VI. How Much This Material Can Show

The conclusions above are limited by the following aspects of the material and method. First, what is used here is the first collection of the administrative section; contracts and litigation documents are in the civil and criminal sections, which are not in the corpus used here. Of the "concluding contracts and bringing lawsuits" that Weber speaks of, all that was counted here is five inter-village covenants and one complaint, which is not enough to judge these two kinds of action. Second, the location is only the Hsinchu and Tamsui area; of the 60 group documents, 59 bear a date, and apart from one from Jiaqing 25, all fall between Daoguang 22 and Guangxu 20; Hsinchu was a county seat, not a great city like Suzhou or Hankou. Third, four of the five city-level documents come from the same case file (the file covers items 324 to 343), so the frequency with which city-level groups acted cannot be inferred from this number. Fourth, the situation changed after Guangxu 29. Around 1904 the Qing court used the Brief Regulations for Chambers of Commerce to encourage commercial towns to establish chambers of commerce, and the Suzhou General Chamber of Commerce was established in 1905; the chamber of commerce was the first kind of urban group established under a dedicated set of regulations and recognized uniformly by the authorities as a representative body, and thereafter the form in which groups obtained standing differed from that in the period of this material. [Note 10] Fifth, the two codings were made by two instances of the same model; the agreement rate is high partly because both codings came from the same model. Sixth, the 520 disagreements in the inscription corpus were adjudicated by me alone, reading the steles, and no second person reviewed them; the changes made in adjudication mostly turned shen and jie into miao, so the temple's share is somewhat higher than in the two codings. But for the several key comparisons used in Section V, the conclusions that codings A and B each yield when read separately are the same as after adjudication: for steles in the name of a zhuang that use the name of a temple, the two separate codings give 0.28 and 0.25; for the share of groups among those approved, they give 0.83 and 0.81 (both pairs computed on the 589 rows before duplicates were removed). Seventh, groups unable to erect steles are absent from the inscription corpus, as are most groups unable to acquire property; what the inscription corpus can show is only under what names groups able to erect steles came forward and how the authorities treated them. Eighth, the earlier scholarship was only partly read. Of the twelve studies gathered on local governance, ritual spheres, and north China villages, only five were obtained in full text made public by the publisher and read: Okada 1938, Shih Chen-min 1973, Hsu Chia-ming 1973, Lin Mei-rong 1987, and Lin Yu-ju 1998; for five, only reviews or fragments were obtained: Duara 1988, Dean and Zheng Zhenman 2010, Zelin et al., eds., 2004, Mann 1987, and Wang Tay-sheng et al. 2004; for two, not a single page was obtained: Tai Yen-hui, Local Governance in Qing Taiwan, 1979, and Schluchter, ed., 1983. Wherever "none of them" or "no one" is written, the scope is limited to the five whose main text was read and cannot be extended to the other seven. The two most serious gaps are these: Tai Yen-hui's pp. 193–202 on the legal-person status of deities I know only from Lin Mei-rong's citation and have not checked against the original book; and the Schluchter 1983 volume of essays, the book in the scholarly history of Weber's China studies most likely to deal with the village exception clause, was not obtained.