The three books reviewed here deal with the complicated subject of the sale of land in China. Qing dynasty land deeds were brief documents that left a great deal of their content to interpretation. Qing dynasty land purchasers did not have the benefit of any effective land registry, and land ownership was poorly defined. Providing a practical guide on Chinese land law, in 1888, Rev. Peter Hoang, said, “if one knows well the ordinary forms [of the land deeds] of one district, it will be easy for him [sic] to ascertain the special usages and customs of other districts, if any there be, and to conform to them as may be required.”Footnote ¹ The many thousands of land deeds and multiple surveys of local customs all over China that have now been found and published lend support to the statement. Land deeds employ a common format, but local usages vary and local people seem to know how to deal with the variations.

In view of the variations between the common format and local practices, studies of Chinese land conveyancing have often taken the form of explication. For instance, they detail how the general practice of huomai (活卖 revocable sale) might be different from juemai (绝卖 irrevocable sale), how the terms of the deeds might relate to the Qing code, and how magistrates might pronounce on them drawing on their local knowledge and expertise on the code. The first title for this review, by Long Denggao, takes this approach. However, in the second title, Cao Shuji and Liu Shigu go a step further, relating the different manners of land holding and conveyancing to a unified argument. In the third title, Taisu Zhang goes further still, setting Chinese land practices into a comparative format with English practices. All three books give considerable coverage to the practice described as dianmai” (典卖 sale by pledging), which I shall refer to as “dian-sale.” In this review, I shall concentrate on their discussion on this subject.

Let me first state the general case for the practice of dian-sale. At the general level, dianmai may be considered a revocable sale. Moreover, the word dian suggests that money is handed over as a loan, so that land is surrendered as a pledge. The pledge terminates when the loan is repaid and the property is “redeemed” (赎 shu). By the same token, it may be said that if the loan is not repaid, the property continues to be subject to pledge. It pushes the question a step further to ask if any time limit might be imposed for the process of redemption beyond which it is considered that the pledge becomes forfeit and that the property so pledged accrues permanently to the creditor. Therefore, pledging could, and often did, lead to the outright sale of land, the pledger thus becoming the vendor, and the creditor becoming the purchaser. Qing dynasty law does not impose such a time limit for forfeiture. On that much of the dian-sale, there is general agreement among writers.

I turn now to Long Denggao's book, in which two of eight chapters deal specifically with the terms of dian-sale and provide examples of variations of that practice. Those two chapters help to set up some simple typologies. (1) In its simplest case, the dian-sale involves the creditor-purchaser paying money and the pledger-vendor offering land in pledge. No assumption is made on the relationship between the former and the latter. But (2) complexity begins if tenancy is introduced into their relationship. Consider two possibilities: when the creditor-purchaser is none other than the sitting tenant or when the creditor-purchaser leases the property pledged to the pledger-vendor.Footnote ² The two situations lead to diametrically opposing interpretations. Where the tenant is the creditor, rent paid would be offset against interest payment due, the tenant thereby enjoying security of tenure until the pledger, who is the landlord, annuls the pledge by repaying money received. Where the tenant is the pledger, he now pays interest on money received in addition to his rent. He becomes locked into a process of gradually losing title to the land. However, (3) because payment in dian-sale is not regarded as payment for the property at full market value, when land value appreciates, the pledger-vendor, rather than redeeming the property, can demand supplementary payments (找价 zhaojia), much to the annoyance of creditor-purchasers. (4) For another variation, citing examples from Taiwan, Long shows that in the frontier situation where the Qing imperial government did not permit Han Chinese settlers to register land beyond a designated boundary, natives pledged land to Han Chinese creditors who paid them the fan dazu (番大租 aboriginal big rent). This would have been a variation of (2) but it would have been a commonly employed roundabout way of purchasing land that should not have been sold. (5) Long also cites surveys conducted into local customs (民事习惯 minshi xiguan) in the last years of the Qing dynasty for the extraordinary arrangement in some parts of Zhejiang where vendors were committed by not only one, but two separate deeds. When they pledged their properties in return for loans, they also, at the same time signed a deed of irrevocable sale which was kept by the purchaser. In the event that debtors forfeited loan repayment, the deed of irrevocable sale would take effect and the creditor could take over the property. It is interesting that creditors required the irrevocable sale deed to ascertain the default guarantee of irrevocable transfer of title. That in itself says that local practices had never made totally certain the consequence of default when land was pledged. Long's examples do not exhaust all possibilities, but they are sufficient to show that behind common terms such as dian lurk many a varying social practice. Despite the title of his book, Long presents other examples that may be considered typologies in their own right, but he does not present a general explanation as to how dian evolved.

In the second title reviewed here, Cao Shuji and Liu Shigu present a general typology that relates not directly to the dian-sale, but to perpetual tenancy (永佃 yongdian). I shall show that the argument puts dian-sale in a new light. The typology is supported by twentieth-century local studies and enormous numbers of documents collected by Cao and his students. With reference to Jiangnan, they summarize the various routes to tenancy in Table 1.

Table 1.

Comparison of Types of Tenancy in Jiangnan

Source: Cao and Liu, p. 70.

To elaborate, the four types of tenancy of column (1) are placed on a continuum relating the manner by which tenancy is set up to the permanence of the tenancy relationship and the control tenants might have over their tenancy rights. In the first and second types, by agreement, tenancy may cover a short or a long period of time, but the tenant has no right to transfer tenancy to another tenant while the landlord retains the right to terminate tenancy. In the third type, where a pledge-rent (押租 yazu) is paid, the tenant holds a long-period tenancy and the right to transfer or resell tenancy rights, but the landlord retains the right to terminate the tenancy. By way of explanation, it may be said that the “pledge rent” is the rent paid by the tenant to the landlord when the landlord has pledged the property to the tenant in return for a sum of money (that is, the first possibility within type (2) in my description of Long's schema). Finally, where tenants have contributed substantially to reclamation, they hold surface rights recognized not by the landlord alone, but by the wider community (thus “publicly acknowledged”), they may transfer or resell their tenancy and the landlord may not terminate tenancy. In this typology, type 1 and 2 tenancies hold no “surface rights,” but types 3 and 4 do.

“Surface rights” is the term used by scholars to describe the situations when tenants (who pay a rent to the landlord) may sub-rent their holdings to other tenants. In other words, in Table 1, column (1) is defined by column (4). Cao and Liu believe that the “pledge rent” and the existence of “surface rights” are closely connected, but for them, whether “pledge rent” existed does not depend on the occurrence of the term in the land deed. Commenting on the abundant land documents they have collected from Shicang village in Zhejiang province, they say, “When we analyse in depth several thousand Shicang land deeds, we see no record at all of ‘pledge rent’, but we see almost daily the sale of land which is liable to redemption (回赎 huishu)” (69). Recall that a characteristic of the dian-sale is the right of redemption by the pledger-vendor, a right that might be translated into further demands for payment. Demands for supplementary payments are made not only for dian-sale but for all revocable sales. However, Cao and Liu continue, “We shall prove that pledge-rent and the price of surface land are not essentially different” (69). Demonstrating that pledge-rent is closely related to “surface rights” is the novelty of Cao and Liu's argument.

The argument is laid out in Chapter 10, using archival documents found in Jiangjin county in Sichuan province. There on the Chengdu plain, tenants are required to pay a deposit up front when renting (the payment being commonly referred to as the “pledge-rent”). The reason this arrangement is close to dian-sale is that landlords must pay interest on the deposit in the form of an annual rent deduction. The larger the deposit that is paid, therefore, the greater the rental deduction that follows, and, for that reason, when the deposit interest approximates the rent, no further rent is paid. Local people refer to such tenants who pay no rent as “big pledge-rent households” (大押户 daya hu). As tenants can enjoy security of tenure without paying a rent, Cao and Liu argue that the pledge-rent and the price of surface rights are essentially equal. As Cao and Liu note, the practice is not unique to the Chengdu plain. They cite no less an authority than Chen Zhengmo 陈正謨, who published a comparative study of provincial rent practices in 1936, for recalling that the practice was found in his home county of Zaoyang in Hubei. The same discovery was also made more recently by historian Gao Wangling through interviews and gazetteer reports in Hunan province.Footnote ³

Cao and Liu cite further evidence from the “reduce rent and return pledge rent” (减租退押 jianzu tuiya) movement in the early 1950s in support of their argument. Demanding that landlords return the pledge deposit posed considerable hardship to the landlords, many of whom were middle or poor peasants. The tenants themselves might likewise also be so poor that they had borrowed money in order to pay the deposit. The archival records of the county party committee show time and again guidance issued to the villages to take into consideration the economic conditions of landlords and tenants in demanding deposit reimbursement. The return of pledge rent deposit became such an intangible issue that by 1951, scores were being settled not by any calculation of the degree of exploitation, but by “the speaking of pain” (诉苦 suku), often leveraged upon landlords having been engaged in acts of cruelty or disloyalty that had no connection to how they held their land.

Cao and Liu's book is not likely to be the last word on the subject of dian-sale of land. However, it is an example of an attempt to reach a generalized understanding of land holding practices and to subject their generalizations to vigorous tests against abundant documentary sources. I shall not nitpick, but simply acknowledge that it is a fine example of that approach.

In the third title to be reviewed here, Taisu Zhang begins with the observation that Qing dynasty law imposed no time limit on the period allowed for redemption on dian-sale while English mortgage law did, and he argues that this difference can be explained by the fact that small holders in China were accorded higher social status, a feature of Chinese society that he relates to strong kinship ties.

He looks at dian-sale in its simplest form, that is to say, a debtor-vendor providing land as surety for a loan from a creditor-purchaser. That narrow definition convinces him that vendors of land must necessarily be under greater financial strain than purchasers. He believes, therefore, that “due to the economic primacy of land and the paucity of alternative employment opportunities, rural households likely had extremely strong incentives, in both societies [England and China], to retain or, in the case of collateralized borrowing [by dian-sale rather than irrevocable sale], redeem their land whenever possible” (96). That would seem to be an overly simplified view of a complex situation, in support of which he says, “nearly all existing contract archives for Qing and Republican China suggest that land mortgages and sales generally involved a few large householders accumulating land from dozens of poorer neighbours” (97). Obviously, one would have to have examined the archives to make that claim and as I simply do not believe that Zhang has done it, for me, that is just an empty claim.

Unfortunately, this book is full of empty claims. Let me focus on chapter 4, “the empirical centerpiece of this book,” as the author says (123). This chapter argues that, in China, “status and authority did not correlate strongly with wealth but did correlate strongly with generational seniority in kinship networks” (136). I find very troubling the documentation for this, which is cited separately for north China and Jiangnan.

The documentation for the north China case comes from some villages surveyed by the South Manchurian Railway. Zhang focuses on one village, in particular, Sibeichai in Hebei province. For this particular village, Zhang provides a table listing the names of the village elite (made up of village or division chiefs), how much land their family owned or rented, and their lineage positions. The object is to demonstrate that while these people did not own very much land, their lineage position was nonetheless quite high.

I have much difficulty following Zhang's arithmetic, and not for lack of clarity. He says that Sibeichai households owned, on average, 1.9 mu per capita and farmed 2.19 mu. By his reckoning, “12 of the 18 village leaders identified earlier were therefore at or below the village average in one category, whereas 9 were at or below average in both” (139). By my reckoning, 6 of them indeed owned less than 1.9 mu per capita, but 3 more owned 2 mu, 2 owned 2.5 mu, and 1 owned 2.75 mu. I do not see any household where renting land pushed them below the farmed average of 2.19 mu, only that, if rented land is included, 2 of households that owned less than 1.9 mu now farmed respectively 9.29 or 8 mu per capita, and 1 other is not certain because it co-farmed 76 mu with another household of some unknown size. I have no doubt that Sibeichai was a poor village at the time of the South Manchurian Railway survey, possibly around 1941, but the table shows that among these poor people, it was the slightly richer ones who became the village elite. Maybe being family heads also gave them an edge, but that is not Zhang's argument.

But let me not quibble, for we know much more about Sibeichai village than Zhang lets on. Both Philip C.C. Huang and Ramon H. Myers devoted a great deal of attention to this village. Huang says of Sibeichai:

Myers has more on village leadership and organization:

What did the village leadership do? Myers continues:

Myers sums up the work of the village head in the following terms:

By no stretch of imagination could Sibeichai have been run by poor people high on the generational count. But what I find truly bothering in Zhang's study is the total disregard of background. He dismisses Huang and Myers in one sentence: “Quite contrary to what previous and less statistically thorough studies of the Mantetsu surveys have suggested” (136). It does not matter to Zhang that the leadership structure might have changed in the few years before the South Manchurian Railway survey, that the economy had been transformed by a single rich family buying up a substantial portion of the village land, or by the outbreak of war (whatever might have happened to landlord Wang Zanzhou?), all that matters is that he comes up with a listing of eighteen names at a point in time; and he did not even count them correctly.

Zhang's Jiangnan documentation is no less opaque. I find it very hard to come to terms with throwaway remarks that Zhang thinks constitute his evidence. Referring to his first body of documentation, namely eight written genealogies purportedly held in the Shanghai Library, he says, “These lineages shared two basic organizational characteristics. First, none [sic] of the registries [Zhang's term for “genealogies”] positively identify personal wealth or landholding with higher internal status and authority … . Second, of the eight lineages that published selection criteria for leadership positions, all eight highlight the importance of generational seniority” (148). These statements that look as if they are founded on documentation are completely misleading.

To begin with, no genealogy is a registry. Calling it a registry suggests that there might have been mechanisms for registration. As anyone who has worked with Chinese genealogies knows, these records include substantial portions reproduced from earlier genealogies and additions that might be based on memory, grave records, spirit tablet entries, and numerous other fairly haphazard processes. Some families do register male births. I know of none that registers marriages and deaths. Very occasionally, one sees signs that the birth registries are reflected in genealogical entries. It is well-known that Chinese genealogies are full of gaps.

Indeed, generation and age are criteria of status, but not of authority. Moreover, generation and age are not the only criteria of status in the lineage, nor, indeed, are they criteria for filling managerial positions. As for highlighting generational seniority as opposed to wealth and land-holding as selection criteria for leadership positions, let me turn to the two genealogies that in a later paragraph Zhang refers to as indicative of lineages that “selected councilors on the basis not only of seniority, but also of intelligence and moral reputation” (149).

Zhang does not cite page references, and so I have to use my judgement to see where in the genealogies he thinks anything is said about selection criteria. I think in one of the two genealogies, the Bianshi zongpu 边氏宗谱 of Wuxi county, the statement is drawn from a document included that is described as the “lineage regulations” (宗规 zonggui). The first clause of that document says that lineage and sub-lineage heads (宗正 zongzheng and 族长 zuchang) are determined by seniority, but, “because they may not all be virtuous or capable (贤明 xianming), it is necessary to choose someone from the lineage whose heart is straight and fair in dealing to assist them as the ancestral hall manager (祠正 cizheng).”Footnote ⁶ That statement makes it quite obvious that seniority alone is not trusted as a criterion for any leadership position, except in ritual terms. But there is a more pressing issue: what does it mean to say that they should be assisted by an ancestral hall manager? Does it imply that they had an ancestral hall, or properties, to manage?

Like many genealogies, the Bianshi zongpu is made up of lines of descent of people who did not live within a single territorial community. The genealogy does not note that these people maintained any ancestral hall in common. From the biographies included, several branches built ancestral halls by the nineteenth century. We do not know if the “lineage regulations” were ever put to use in any of those ancestral halls, but it would be safe to conclude, for most of those people whose names were recorded in the genealogy, most of the time, no such regulation ever applied.

The other genealogy that is said to select “councilors” on the basis not only of seniority but also of intelligence and moral reputation is the Yaonan Dingshan Fangshi zupu 姚南丁山方氏宗谱 (1921) of Ningbo. I fail to see where it includes a statement that might reflect Zhang's conclusion, but the genealogy provides considerable documentation on the management of lineage properties which Zhang might have considered, if he had really read it. From 1695, the lineage had maintained an ancestral hall and had accepted donations of properties from lineage members, in return for which donations they placed spirit tablets, either of the donors themselves or of their forebears, at the hall to be sacrificed to collectively. Such contributary arrangements from lineage members were common practices and the Fang surname at Yaonan Dingshan would seem to have attracted substantial amounts of landed properties from them. Proceeds from the properties paid for regular sacrifices, at which seniority might be given privileged recognition. But did management of the hall or its properties depend on “councilors”? Not quite, for even though there must have been managers for what amounted to a substantial estate, oversight was provided not by any “council,” but by rotation of the sub-lineage heads. That, again, was common practice in Jiangnan as elsewhere. Was it likely that wealth would not have mattered in an estate established by contributions and maintained as a collective holding in which each sub-lineage could retain a share in management? I do not think there was the slightest possibility.

Zhang's other Jiangnan documentation includes a published collection of 412 land deeds from Ningbo and “a batch of 140 land sales from Shicang, Songyang county, in Zhejiang province.” As he says, “the argument here is that the great majority of Lower Yangtze dian transactions involved middlemen who were senior members of either the dian seller's or the dian buyer's kinship group” (155).

Zhang's description of the Ningbo documents beats all understanding. According to him, “412 were either dian sales contracts, redemption contracts, or conversions from dian to permanent conveyance” and “only three were direct permanent conveyances that had not previously been a dian sale” (155). Those three documents, he notes, are documents numbered 81, 167 and 169. If only one would read the book in which those documents are reprinted, item 81 is described not as permanent conveyance but as a “redemption deed” (立找契 li zhaoqi).Footnote ⁷ Item 167 is indeed described as “absolute sale” (juemai), a term that has the backing of the Qing code for indicating forfeiture of redemption. But, if one goes on to item 168, one sees the complications. In the ninth month of Daoguang 30 (1850), by item 167, Mao Rongchang and his brother Rongrui sold by “absolute sale” three plots of land and collected 35,000 cash from the purchaser, noting that the tax for those plots of land would be removed from the record of the vendor for transfer to the purchaser. However, by document 168, in the eleventh month of Daoguang 30, Mao Rongchang and Rongrui demanded and received for redemption of those plots of land another 20,300 cash.Footnote ⁸ Of the three documents which are meant to be exceptions to the rule, only 169 is, as Zhang says, a “direct permanent conveyance.”

My difficulty does not stop there. By my reckoning, the collection includes forty-two deeds specifically stated to be instrumental in redemption: three are deeds of absolute sale (without reference to dian practices) and the rest are deeds of “permanent sale” (永卖 yongmai), except for a few deeds where the wording is indeterminate. What makes Zhang so certain that these were dian sale contracts, redemption contracts or conversions from dian to permanent conveyance? Nothing in the wording of those deeds suggests that this was the case.

I think Zhang has relied on the interpretation of the preface to the collection, written by its compiler, Wang Wanying. Wang draws on local knowledge of Ningbo, to say that “permanent sale” in the eastern parts of Zhejiang referred to the sale of surface rights.Footnote ⁹ One has to fudge the argument by quite an extent to jump from there to saying that all land held under surface rights was subject to the principles of dian-sale.Footnote ¹⁰ Even if one might agree with Cao and Liu in understanding that payment of rent deposits for which interest is reckoned could give rise to the equivalence of dian-sold property with surface rights, it does not follow that all land held under surface rights had originated in dian-sale. There is a very wide gap between what the Ningbo deeds show and what Zhang claims they do.

Do any of those deeds show that “the great majority of Lower Yangtze dian transactions involved middlemen who were senior members of either the dian seller's or the dian buyer's kinship group”? (155) I fail to detect any indication of seniority or otherwise of middlemen in those deeds. Zhang also cites “a batch of 140 land sales from Shicang” for evidence that “of the 140 transactions, 130, or 93 percent, involved at least one middleman who was related to one of the contracting parties, and the great majority involved at least three” (157). Shicang has yielded thousands of deeds, and it is thus hard to understand which batch of 140 Zhang examined. The deeds do not give any indication of family connection or social position of the middlemen, and so I presume Zhang concludes from common surnames that some people are “related.” The status and connections of middlemen is an interesting question, and Shicang, with the very detailed work conducted by Cao Shuji and his team at the Shanghai Jiaotong University, offers an opportunity for investigating it. But I do not think a casual comparison of surnames is likely to be the methodology for solving that problem.

Zhang's chapter 5 proceeds to “permanent tenancy.” Again, whatever analysis is claimed is mired by casual reading of the sources. Let me cite one last case for discussion, a dispute settled by the magistrate of Zhuji county in Zhejiang province in 1896 (158 and 176). According to Zhang, a “relatively junior member of the local Zhou lineage attempted to ban a cousin-in-law from redeeming land her husband had conveyed to him five to six years ago.” (I struggled with this sentence: If Mrs Zhou conveyed land to Zhou Zulie, surely, it was Mrs Zhou or her husband's heirs, who had the right to redeem, not the purchaser.) Zhang continues, “the point of contention was whether the conveyance was a dian sale” (158)

I looked up the document cited.Footnote ¹¹ According to the text of the document, Zhou Zulie had brought charges against Feng Chufeng, for “colluding to receive land” (串受田亩 chuanshou tianmu). It states that Mrs Zhou was the widow of Zhou Zulie's xiong (兄 possibly “cousin”). She sold Zulie two separate properties that had belonged to her husband through writing a deed (or deeds) because she was in debt to Zulie, for which debt interest was reckoned annually. The manner by which the properties was conveyed is described as follows. Property A, consisting of 4.1 mu was “sold” (出卖 chumai), while property B, consisting of 2.8 mu, was “pawned” (出押 chuya). It was noted that the pawning of B occurred in Guangxu 17 (1891). Thereupon, Zulie established a sacrificial trust known as “succession trust” (继成祀 jicheng si). However, in resentment that interest had been charged and that the properties had been sold cheap, Mrs Zhou sold (chumai) property A to Feng Chufeng. Zulie objected to the middleman. A case of litigation ensued on the claim that gambling was involved. The magistrate set aside the gambling charge, examined the deeds, and summoned witnesses. Thereupon he found that Feng Chufeng had paid 65 silver dollars for 4.0 mu in property A, while Zhou Zulie had paid only 30 silver dollars and 15 strings of cash for both properties A and B. While the magistrate did refer in his judgement to the family relationship between the plaintiff Zhou Zulie and the widow Mrs Zhou, which Zhang cites, he really focused his attention on the differences in the payments. Surely, the magistrate said, the fact that Zhou Zulie paid so little was an indication that the land might be redeemed. For that reason, he ordered that the new purchaser, Feng Chufeng, provide 15 silver dollars, and Mrs Zhou to provide another 15 silver dollars, so that the sum of money that Zhou had paid might be returned to him. As for property B, which had not been sold to Feng Chufeng, he ruled that Mrs Zhou might continue to hold it or to sell it. So, indeed, the magistrate ruled on whether the sale amounted to dian, but whether a cousin-in-law had to be barred from redeeming had nothing to do with the case.

In short, I do not see the evidence that supports Zhang's claim that status and authority correlated strongly with generational seniority in kinship networks. Instead, I see deliberate attempts to misread the records so that they might be given that interpretation. I think there are always reasons for poor research. In Taisu Zhang's case, despite appearances, the reason is that he has no comparative framework that sets the Chinese village and the English village side by side. Not knowing what to look for, he seizes the most superficial reference for corroboration. As always, conclusions drawn from Chinese historical records without a thorough understanding are unsafe.

Nevertheless, with the other two titles being reviewed here, Zhang's aspiration adds to the scholarly work that might contribute to an understanding of Chinese land-holding practices. Gone are the days when historians of China should master Qing land practices for practical purposes, and so much documentation is now available, even from the comfort of one's own computer, that we historians should contemplate more systematic approaches to an unwieldy subject. The documents show such a great deal of variations that I am sure future studies will introduce more typologies than we already have. At some stage, we must also contemplate cross-cultural comparisons. For the moment, I advocate Cao and Liu's approach, involving detailed examinations of local practices and relating them to a broad framework. Their framework is not the only one that can be constructed, but when a few more such frameworks become available, they may be compared and contrasted and the study of Chinese land practices can move on.