宋代绝户的法律后果（英文原版）

宋代绝户的法律后果（英文原版） WHO GETS IT WHEN YOU GO: THE LEGAL CONSEQUENCES OF THE ENDING OF HOUSEHOLDS 絕戶,IN THE 1SONG DYNASTY (960-1279 C.E.)) BY BRIAN E. MCKNIGHT (The University of Arizona) Abstract In imperial China the normal pattern for the passing on of property across generations, giving assets to daughters as dowry at marriage and bequeathing the family estate in equal portions to sons upon the deaths of the parents, was distorted when both parents died without there being a living son, a situation called the extinguishing of the household (jeuhu). Under some dynasties the state attempted to deal with this phenomenon by mandating the posthumous adoption of sons, but during the Song (960-1279 C.E.) the government accepted the fact of the frequent ending of household (and family) lines, and established an elaborate set of laws and policies to assure the orderly transfer of assets. Song policy in this regard was prompted by the state?s desire to avoid the concentration of land in the hands of the rich and the potential loss of tax revenues. Song policies both reflected and supported two social trends, the increasing independence of smaller stem family units from the influence of larger kin groupings and the increasing control of property by women, because most of those receiving juehu assets were women. In societies based on private property, patterns of transmitting assets across generations have major effects on power relations, which in turn affect patterns of marriages, adoptions, gender roles, and the social context of family life. Changes in these distributing practices both betoken and influence changes in economic and social behavior. In imperial China, as in peasant-based societies all across Eurasia, the most common pattern of distribution involved diverging devolution, a vertical handing down of wealth through both men and women, often coupled with an abiding concern 2for the continuation of lineage.) The practice usually said to have 1 This essay was first presented in June of 1999 at the “International Symposium: New Directions in the Study of Late Imperial Literature and History”, Taibei, Taiwan. I wish to thank the organizers of that conference and my fellow participants, whose comments were useful in preparing a revised version of this work. 2 This pattern and its implications have been most fully explored by Jack Goody in a variety of publications. See for example Goody 1973; Goody 1969a; 1969b; 1969c. ? Koninklijke Brill NV, Leiden, 20? JESHO 43,3 Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 315 been standard in China involved passing assets to women through dowry and to men through equal post-mortem division of parental property among sons. This pattern, though dominant during the later history of the empire, always coexisted 3with variant practices such as pre-mortem distribution of assets,) the use of testamentary documents, and in some periods the transmitting of control over property to women not just through dowry, but also as recipients of pre- mortem distributions, as residual heirs, and even, during the Southern Song, as heirs outright in post-mortem property division. From within the society, the key units in this transfer process were composed of relatives by blood and marriage. The key consideration was the preservation of the stem family (jia ^), a unit centering on a married couple and their children but often ,including one or both of the husbands parents and possibly other relatives. In some periods larger agnatic kin-based units were able to exercise effective control over some of these resources. Variations in the shares of property passed to stem family members or retained by larger kin units reflect changes in the power relations involved, which in turn reflect changes in the economic systems in which both types of units were embedded. From an external, state perspective, although continuation of families or larger kin units was of some interest, the most immediate pressing concern often was the continuation of the principal tax paying unit, the household (hu 戶,.In actuality the membership in a household was often identical with the membership of a family, e.g. 3 Evidence from the Qin-Han into the Tang suggests that it was common practice, indeed possibly the modal pattern for transmitting property. The Qin law allowing and the practice of pre-mortem division continued into the Han. Although, in contrast with Qin law, Han law did not actively encourage, much less mandate, pre-mortem division, the practice was legal and approved by some segments of Han society. See for example the case of the Han figure,Lu Jia 陸贾,cited in the biography of Yao Chong 姚u崇 in Liu 1975,96.3027. One Han father, proposing such a division, says explicitly, the rites treat living separately as appropriate; families practice the way of dwelling apart.” Fan 1965, 76.2471. Although such division may have been common practice during the dynasty, orthodox Confucians seem to have become increasingly critical of any system that allowed sons to own property while their fathers still lived. During the Han itself these ideologues were unable to alter the state laws that allowed pre-mortem division but in the Wei state (220265) of the Three Kingdoms, the state “abolished the [Han] law that had allowed sons to live separately.” See the preface to the Wei statutes quoted in the Treatise of Law in Fang 1974, 30.925. This new policy, however, would seem to have been sharply at odds with local customary practices, at least in some parts of the empire. A literatus and official of the Jin (265-420), Shi Bao, was praised for seeking to ward off later disputes between his sons by dividing his property before his death. Yao Chong 姚祟 in the Tang cites both Lu Jia and Shi Bao as models for his own decision to divide his estate. See again Liu 1975, 96.3027. And under the later Liu-Song state (420-479), “in seven out of ten families, from the literati class on down, the sons lived separately even while the mothers and fathers were still alive.” Among the commoners themselves “in five families out of eight, fathers and sons had separate property.” See the biography of Zhou Lang 周朗 in Shen 1974,82.2096. This practice continued into the Tang (618-907). The biography of the eighth century high official Yao Chong reports that he “divided his agricultural properties pre-mortem and instructed his sons and nephews to preserve their shares. Then he wrote a testamentary instruction to serve as a warning for his descendants.” Liu 1975, 96.3027. Copyright ? 2000. All rights reserved. 316 BRIAN E. MCKNIGHT a set of parents and their children. In such circumstances whether to call the unit a “family” or “household” hinges on perspective rather than constitution. Looking outward from within the social group, the unit was a family; looking from outside, ,from the states legalistic viewpoint, the defining factor was the presence of a household head responsible for paying taxes. But state decisions concerning household units, based substantially on fiscal considerations, could and did affect family and other kin dynamics. Decisions by the authorities affected the patterns of property distribution across generations, and property distribution, by affecting the tax base, affected the state. Although clearly distinct in some ways, the interests of both the government authorities and the kin units within the society obviously overlapped and interacted, and both were affected by an intractable problem. The desire among the people to continue their families and among government officials to avoid having tax-paying units disappear ran headlong into one of the key demographic traits of traditional peasant societies. Historians and anthropologists, looking at such societies in various parts of Eurasia, have found that, with remarkable regularity, in twenty to thirty percent of couples the wife reached the end of her child-bearing years without there 4being a living natal male heir.) When that happened the state and the people specifically affected had immediate, but not always commensurable, interests. State authorities principal concern was that the properties of the household remain in productive use so that they could be taxed (though they also had some secondary agendas); the people who might into control of the properties had an immediate economic interest in how the properties were to be distributed. Both sides, to differing degrees, might use family-continuity arguments in justifying their actions. People might respond to the threat of the ending of a family (jia) line by adopting a male heir. They might respond by bringing in woman as mates of the household head in roles such as concubine or by using a device like uxorilocal marriage, where the husband married and lived in the family of his father-inlaw so that the elder ,couples daughter might give birth to the family heir. Or they might simply permit the family line to die out. In imperial China adoption was the most common way of creating heirs. Indeed, from the Ming dynasty (1368-1644) on the state legally favored the appointment of adopted heirs. The addition of new women to the household of the original family was a secondary approach, largely confined to the rich. As such it was probably the 4 For example, Collver 1963. Collver found that in contemporary Indian society, 22% of all parents reached the end of their reproductive period without there being a living son. Hollingworth 1964-65, p. 371, indicates that in ducal families, 16% of the males in the cohort born between 1330 and 1729 and 17% of the females in that cohort had childless marriages at completion of fertility. When all marriages are considered the figure for men rises to 27% and 23% for women. Peller 1947 indicates that 15% of the married men in his sample of the European nobility were childless. This pattern seems also to have held true in traditional China. See below, note 6. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 317 least common way of dealing with the lack of a male heir. The bringing in of uxorilocal husbands was more common than concubinage, though much work clearly remains to be done before we can arrive at a fair judgment about this practice, especially for periods prior to the late Qing (1644-1911) and the twentieth century. These three sorts of responses (adoptions, uxorilocal arrangements, or the adding to 5the original family of new women who might bear heirs) have been widely studied.) However, families (jia) and the households (hu) they often represented, might respond by doing nothing, so that the paternal lines of descent simply died out, severing the continuity of the jia, and eliminating at least temporarily a tax-paying unit (juehu 絕戶,,thus threatening the state with possible loss of revenues. In the case of a childless couple the mere death of a husband did not result in juehu. The surviving widow, now the head of a unit classified during the Song as a “female household”,nw/m 女戶),always retained the prerogative of naming a male heir. Thus, so long as she lived the household was not legally juehu. Because she was required to act as a trustee for the property, preserving if for her (potential) adopted son, her control of the property was, at least legally, less provocative of competition than the transfer of control that followed juehu itself. Juehu created a potentially destabilizing situation by releasing valuable properties from the control of kinship units led, at least potentially, by males, and transferring ownership predominantly into the hands of women who were not household heads. In a largely patriarchal society this process created the possibility of increased competition and struggle, motivated not simply by greed for more property but also by concerns about maintaining male dominance. Despite its obvious importance, the juehu phenomenon has not been of much concern to scholars. The discussions of scholars on these issues often give the impression that it was uncommon for lines of descent to die out in imperial China, a supposition that has 6been accepted with little attempt to see whether it was indeed true.) However, at least for some eras, there is persuasive evidence suggesting that the dying out of family lines was by no means uncommon. Even under the dynasty that first pushed for the appointment of heirs, the Ming, the legal rules also contain a regulation which 7deals with the distribution of the estates of extinguished households.) 5 See for example Goody 1973; Goody 1969a; 1969b; 1969c; Waltner 1990; Ebrey 1992. 6 Several reasons suggest themselves for this misconception about the frequency with which family lines died out. First, readers often mistake prescriptions of law for descriptions of social practice. Second, the misconception may reflect the reading back into the earlier Chinese record of situations known from more recent times. Third, it was in the interest of the late Confucian elite to stress continuity in ideas and institutions because it derived a substantial portion of its self-justification from the notion that it was descended intellectually and socially from those who lived millenia earlier. One scholar who did point out the apparently high rate of the extinction of family lines was Wolfram Eberhard. See Eberhard 1967, p. 39. Eberhard suggested, on the basis of his study of family genealogies, that 22% of the men in his sample died without male heirs. 7 See Nilda 1964, p. 839. Copyright ? 2000. All rights reserved. 318 BRIAN E. MCKNIGHT The root problem was always there. A sizeable minority of couples ended their child-producing years without having a living male heir, in a society that valued male line continuity. The potential ways of responding to this situation also remained the same (acquisition of heirs through adoption or through adding women who might bear heirs, or the dying out of the line and household). The preferred responses to this situation varied. Such evidence as there is suggests that the roots of variations in the choice of responses should be sought in changes in the socio-economic conditions of the society and, perhaps to a lesser extent, in the ideology of family-line continuity. For these same reasons, the relative percentages of families whose lines died out, often in the process eliminating a household (juehu), doubtless varied from era to era. The degree to which the extinction of households affected state interests also varied, depending on taxation policies, inheritance practices, and patterns of family solidarity. During the early Song, with decreasing solidarity within larger kin groupings, a highly fluid market for landed property, and a tax system heavily dependent on land taxes, the government was forced to pay special attention to the juehu phenomenon. The Song state did not attempt to solve the problem of juehu by encouraging families so threatened to create heirs; rather the state seems to have accepted frequent dying out of households as a fact, and tried to legislate a body of rules that would reduce protracted struggles over ownership. The resulting state enforceable rules on disposition of the property of juehu impinged forcefully on family success strategies. Control over property was seen as a necessary condition for family success. In a society dominated by private property, where, partly as a result of the general preference for equal post-mortem division of property, kin units were forever engaged in intense competition for scarce resources, the way in which the property of extinguished households would be distributed was 8of intense interest. Who was to benefit?) State laws on the distribution of juehu property provided defenses for legitimate recipients of juehu property but, when unclear, created opportunities for competition over resources. Such rules were involved, in intricate and mutually influential ways, in two of the most striking phenomenon of the Song, the increasing stress on stem family interests at the expense 9of those of larger kin units,) and the concentration of greater control over property in the hands of women. In a monetized economy with greatly increased opportunities for personal (and family) mobility, smaller kin units were presented with the possibility of lessening their dependence on larger kin structures. In the process the members of these units 8 Two recent works on the general topic of civil law in traditional China give succinct surveys of the materials discussed below. See Ye 1993 and Kong 1996. 9 The rise of lineage organizations from the late Northern Song on, a phenomenon that has been widely commented on, should be seen in this context and a moderately successful reactionary counterattack against the increasing tendencies for stem families to put their own interests first, and to do so in an effective way. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 319 came to view active participants in the struggle for stem family welfare as legitimate concerned parties in the economic life of the family, even if these people were not related by either blood or marriage. Even more importantly, the development of a complex set of rules for the distribution of juehu property favored the power of women. There were others who benefited from the rules; clearly women benefited the most. Women could receive the property under the law, and the law would defend their possession. Women’s increasing control over property, combined with a decreasing dependence of stem families on their agnatic kin, to promote growing awareness of the importance of creating solid affinal ties. Thus the study of changes in the patterns of distributing juehu property and in the laws that help shape this process contribute to our understanding of the evolution of Chinese society in the Song, especially the place of women in the society and the struggle between competing ideals of kin relationships. Inheritance of juehu Property: Tang and Early Song Although under the early empire, from Qin to Tang, the pattern of the distribution 10of property when family lines (and households) died out) is not wholly clear, evidence indicates that women sometimes received portions when households died 11out.) In this as in other ways, women during this era exercised considerable control 1213over property) and at times owned it in their own right.) The pattern of daughters 10 Division after the ending of a family line was only one facet of the larger pattern of household division in traditional China. For the general situation see Wakefield 1998. Note that the general popular term for such divisions was not division of the household, but fen- jia, “division of the family”, an understandable usage, since in general practice this was a matter handled without state interference as a question of dividing property among blood kin. 11 The Song dynasty (960-1279) work, The Imperial Encyclopedia Compiled in the Taiping Period (Taiping yulan), is the source for a story demonstrating that, during the Han, if a family line became extinct, daughters had the right to inherit. The story, taken from the Han work., General Explanation of Customs (Fengsu tongyi), is both complicated and colorful. A ninety year-old man, without a living son, married in his dotage the daughter of one of his tenants, and died almost immediately after the marriage was consumated. The woman later gave birth to a son whom she claimed was the posthumous scion of her deceased husband. The daughters, knowing that the estate would come to them in the absence of a male heir, refused to accept this claim and went to court. In the end the (purported) son?s rights were vindicated but the salient point in this discussion is the obvious recognition on the part of the daughters that, in the absence of a son, the property, all of it, would come to them. See Li Fang 1960, 639.2863-64. 12 For one example of a woman who exercised the power to distribute land to family members, and to take it back and redistribute when she thought that was important, see Bret Hinsch?s recent article describing two Han dynasty wills. Hinsch 1998. 13 If the property had already been divided up and parcelled out to new independent households at marriage, as seems to have been the common pattern in the Han, the property already given would legally be fully under the ownership of the new households. This is of special importance because, in the event that the husbands in these newly independent households died, the women would inherit full rights over the properties, to do with as they saw fit. This clearly contributed to the independence of widows. Their ownership of properties made widows attractive marriage partners. Copyright ? 2000. All rights reserved. 320 BRIAN E. MCKNIGHT inheriting the landed property of households whose patriline died out, present in the Han, continued to be the common practice into the Tang. The earliest clear Tang reference appears in a work called the Pervasive [Rules] for the Bang Family (Jishi Bangtong): When persons die so that a household line is extinguished, close agnatic relatives are to deduct the 14expenses of the obsequies. The remaining assets—the bound followers,) slaves, fields and buildings (tianzhai), and [other] wealth (zicai)—are to be distributed equally among the daughters living at home. If there are no such daughters living at home, they are to be distributed among close agnatic relatives. If the deceased, while alive, himself established a will which can be verified, 15this ordinance is not to be applied. (Italics added)) A Tang ordinance on mourning reproduces this passage almost exactly: When people die so that a household has become extinct, close relatives are to sell off the ,households bound followers, female bondservants, male and female slaves, buildings (zhai), and wealth (zicai). After the deduction of funeral expenses, the wealth that remains is to be given to the daughters. If there are no such daughters, it is to be equitably divided among close agnatic relatives. If there are no such relatives, it is to be managed by the state. If the deceased, while still 16living, personally drew up a will concerning division which proves on examination to be clear,) 17this ordinance is not to be applied.) 14 Bound followers (buqu) was the name of a special status of unfree persons, ranked higher than slaves. The status originated in the disordered times after the Han when men placed themselves under the protection of stronger families to do service. 15 This work is found in an eighth century Japanese work and is cited by Niida Noboru. See Niida 1946, p. 176. The Chinese work, the 九咖’ 发紀氏傍通,Niida takes to be a work from the reign of Xuanzong (r. 712-56) but gives no evidence for his opinion. 16 Wills had long been a key in transmitting property. See for example Hinsch 1998. Other evidence from the Han suggests that women not only would be in a position to inherit full control of their husband?s properties in the event of his death, but might also receive shares in the estates of their fathers. A work dating from near the end of the Eastern Han (25-220) records a story of a Han father who, fearing the possible ending of his line, distributed his property to his already married daughters though a will. See Ying 1980, p. 421. Another story about the Han tells a similar tale, of an astute and aged father, fearful that his young son would be disinherited and possibly harmed by his already married daughters after his death, wrote a will awarding all of his property save a sword to his daughters, with the sword being kept in trust for the son by the daughters until the boy reached his fifteenth year. When, at that time, the daughters refused to return the sword, the young man sued, and an astute judge redistributed the property according to his reading of the deceased man?s real intention to protect both his son and his property. See again Li Fang 1960, ch. 836, pp. 3736-37. Another example of inheritance by a married woman is also described, for the Han, in an incident reported in Zheng 1937, 3.27. 17 Dou 1984, 12.198. It is curious that this code passage, while reproducing the Pervasive [Rules] of the Bang Family passage almost exactly, deletes the word for landed property. The prudent explanation of this would seem to be that the officials who wrote the ordinance felt that the expression zicai, which has the broader meaning of wealth or property, could include landed property, without that category being specifically mentioned. The “ordinance” mentioned here is a specific type of law used under most dynasties. Such laws were prescriptive rather than proscriptive, and were not penal, in the sense that their texts do not mandate penalties for infraction. Unfortunately, because they were of less stature in the minds of the official elite, they have not be preserved with the same care as have the better-known codes of statutes. Where they have survived it is often, as it is in this case, because they are included in some other form of official compilation. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 321 Scholars differ on whether the rule embodied in this passage would have been 18applied differently among daughters on the basis of their marital status.) The great scholar of Chinese legal history, Niida Noboru, understood the passage to mean only 19unmarried daughters living at the home of the deceased.) Although the issue of whether the phrase “living at home” in this passage includes women who were physically living at home after divorce or widowhood remains unsettled, it seems more probable that the phrase “daughters at home” also covered daughters who had married but had returned to the home of their living parents as a result of divorce or widowhood. The original commentary to this sentence says that “[In cases where daughters were] under the same household registry but already held wealth (zicai) in separate accounts, this rule also applies.” It seems highly unlikely that this would mean that the family had already divided its property unofficially while maintaining a single registration. It would seem more plausible that the passage was referring to returned daughters who brought back to their natal families wealth derived from their dowries. It is, however, probably reasonable to assume that it did not include married women still living with their husbands’ families. This pattern for distributing juehu estates to daughters was broadened by a Tang edict promulgated some time after the basic rule had been adopted. The Song Legal conspectus, immediately following the Tang mourning ordinance quoted above, cites an edict from the late Tang (836 C.E.): From now on, if among the commoners and the various sorts of people, men die and are cut off who lack sons and have only married daughters [living with their husbands], these [married] 20daughters are by ordinance to share the property (zichan 資產).) The problems of understanding Tang and earlier policies with regard to juehu, complex as they are, are less complicated than is the case for the Song. The Song is the first era for which there is a substantial body of passages that speak specifically about the rules on juehu. Having so much material highlights one vexing problem, the broadness of the social situations covered by the term. Legally the situation was simple. A limited group of people, the husband and his wife, the husband’s parents, and possibly his grandparents, had the legal right to appoint an heir. Juehu meant that all members of this group were deceased. As soon 18 See Bernhardt 1995. 19 For Niida?s position on this issue see Niida 1959-64, vol. 3, p. 381. 20 Dou 1984, 12.198. Such approved edicts were themselves laws, in the same sense that statutes were laws, i.e. they prescribed patterns of action while also asserting the power of the state to intervene to assure the following of the patterns, when it might choose to do so. Edicts could of course be written in the form of ordinances, statutes, or the other pre-modern Chinese law forms. Edicts that have come down to us, for the middle period of Chinese history, have for the most part been preserved in collections of materials intended for use as sources for the compilation of dynastic histories. Each dynasty from the Tang on collected the sorts of materials that could eventually be used by the succeeding dynasty to write a history. During the Song the material so collected was compiled into reign histories. We are fortunate to have one important collection of such source materials, the Songhuiyao jigao. See Xu 1965. Copyright ? 2000. All rights reserved. 322 BRIAN E. MCKNIGHT as this happened the household was extinguished. The possible later appointment of a posthumous heir was legally (and socially) irrelevant to this situation. In practice, this deceptively simple definition conflated a bewildering variety of possible factual situations. At one extreme, it could mean an unfortunate couple who died, not simply without a male heir, but without close agnates, close affines, or daughters of any sort. At the other extreme, it could be a couple with not only close agnates and affi nes, but also daughters of four possible sorts—never married daughters living at home, daughters who had been married but who, in the aftermath of divorce or widowhood, had returned home prior to the deaths of their parents, daughters married, but living at their parents home with uxorilocal husbands, and daughters married out, and all the possible combinations of these sorts of relatives. This list does not even include non-related parties such as fostered sons and unrelated but co-dwelling dependents who, as we will see, did eventually have some inheritance rights in the Song. Add to this the fact that the existence of a valid will could fundamentally alter the treatment of a person’s estate. And to complicate matters further, state law at times differentiated inheritance rights depending on the kind of property involved—land, buildings, liquid assets, and so on. Finally, as concerned landed property, inheritance rights depended on the kind of tenure in which the land was held or on other sorts of rights to some actual or potential control of the land. Did the deceased hold the land in a tenure akin to our idea of fee simple? Or did he hold it through what was called conditional sale (dianmai ^M)? Or was he perhaps the original landowner who had forfeited rights to the use of the land in return for payment in a conditional sale? If the sources always indicated clearly the factual conditions necessary for a type of property to be affected by a particular piece of legislation, this forest of possible relevant factors would make the interpretation of juehu evidence complex but not difficult. Unfortunately, the sources rarely indicate the factual situation that led to or was covered by a piece of legislation. However, an analysis of Song legislation and such scattered descriptions of practice as remain extant, when examined in light of contemporary political and economic conditions, provides a baseline for our understanding of this extremely complicated situation. The Song dynasty came to power in 960. In keeping with the founding emperor’s pragmatic approach to rebuilding a stable empire, those in power during the early years maintained a facade of continuity. They retained offi ces and policies in name, while often greatly altering their actual content and practice. Legal compilations were no exception to this pattern. During the opening years of the dynasty a committee drew up a new code, called the Song Penal Conspectus (Song xingtong). However, this “new” code consisted of an almost verbatim repetition of much of the Tang code of 737 as recompiled by officials of the Later Zhou (951-960) a few years Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 323 21before the Song founding.) This new code differed from the Tang Code principally in that it included many rules that had been enacted after 737, either in the late Tang 22or during the Five Dynasties (907-960).) This compilation has served as the major source for comments on the early Song system of laws, but it is well to keep in mind that, in legal practice, the early Song made use of eight collections of laws. Four of these collections were form the Tang period, two from the Later Tang (923-36) and two from the Later Zhou. Four (possibly five) of these works were collections of 23edicts.) Unfortunately for our understanding of the actual state of the law in early Song China, all of these collections are now lost. As law worked in China at that time, the rules in the collections of edicts would have had precedence over the statutes found in the Song Penal Conspectus, but we must perforce depend on the latter work, 24except when we can correct it using other scattered material.) The Song Penal Conspectus adopted and reaffirmed the late Tang rules under which, when a household became extinct, daughters, that is daughters living at home, 25inherited all chattels, buildings, and other wealth, minus the costs of funeral rites.) In the absence of such daughters living at home, these sorts of assets were divided among close agnatic relatives. The Conspectus is also our source for the law, promulgated in 836, under which in juehu families with no daughters at home, the assets went to married daughters. 21 The compilation of codes in traditional China, at least from the Tang dynasty on, was overseen by a committee of high government officials. Such codes brought together the most fundamental laws (in the eyes of the compilers). A given statute in the code would have associated with it a penalty for infraction. In that sense codes were criminal law, though the matters with which statues dealt might involve issues we would consider civil, like marriage and divorce. Codes, for a variety of reasons, were held in special esteem by successive dynasties. As a result most have survived to the present. Pre-modern laws other other than codes have survived in less complete forms for dynasties prior to the Ming and Qing. This historiographical situation has led in the past to a tendency to focus very largely on the codes, in commenting on law in pre-modern China. However, it is important to bear in mind that in imperial China codes represented only a very small fraction of enforceable state rules. 22 These new codes differed from the earlier Tang models in that they included rules in the form of approved edicts and other rules that were in the form of laws I have called ordinances, prescriptive laws that did not contain in the body of the law any specification of penalties for infractions. 23 See McKnight 1987, p. 118. Even in the earliest imperial law compilations, from the Qin, books were compiled that included actual case decisions and/or questions for interpretation of the basic statutes. During the Song use of such “precedents” was very common. They were periodically compiled into books circulated to local offi ces. Some were divided by region of applicability, some by topic, and some were generally applicable. We do not have for the Song any of these collections of precedents. 24 Approved edicts were, by definition, suggestions made by others but approved by the emperor whose will was sovereign. Obviously, the emperor as sovereign was (in theory if not in practice) not to be bound by previous decisions. Otherwise he would not be sovereign. Thus, the most recent manifestation of his will, as reflected in an edict, would have precedence over a previous expression of his will, as included in a code. Of course in practice emperors, often at the insistence of their nominal subordinates, usually refrained from acting in open and exceptional defiance of older established laws, if only to preserve the stability that comes with recognized rules. 25 Daughters who had returned to live at home after being widowed or divorced may have been included in this group, but this is a problematic and as yet unresolved issue. Copyright ? 2000. All rights reserved. 324 BRIAN E. MCKNIGHT The Song inherited this Tang pattern, but the compilers, in the process of their work, already began to reshape the code by inserting commentaries and clarifications to the original text. The late Tang edict dated 836, quoted above, is immediately followed by a clarification, probably written by the head of the Song code compilation group, Dou Yi, or his close associates. The authors were seeking to 26clarify what was, and what not, to be given to such women.) The passage reads: We have carefully examined the situation and ask that from now on, in cases of extinguished households, [if there are no daughters at home] one-third of the building, livestock, and wealth, after the deduction of funeral expenses, be awarded to daughters already married out. The remainder [of such sorts of wealth] is to be confiscated by the authorities. If there is landed property [in this confiscated part], it is equitably to be awarded to close agnatic relatives who will be the renting tenants. If the daughters who had married out have been expelled from their 5husbands families or if their husbands, not having living sons, died without their wives having received a share of their property, and these daughters have returned to their natal homes prior to ?the house- holds being extinguished, they should be treated like unmarried daughters living at 27home.) This amendment is the earliest harbinger of one of the one of the principal traits of Song inheritance law, the increase in the size of the share taken by the state. We will return to this development in more detail below. This passage is also the first clear Song move in the process of reducing the rights of agnates. Under the original Tang mourning ordinance, such relatives had received juehu estates in the absence of daughters at home. This new rule deprived them of these inheritance rights, leaving them merely the right to be tenants on two thirds of 28the landed property.) Early in the eleventh century the authorities eliminated even this vestigial right. A memorial dated 1023 reports that the Song government, in 1015, 26 Professor Billy So, who provided me with a number of valuable corrections to my original draft, pointed out that Dou Yi at a number of points inserts such materials in the text of the Conspectus. Because of his position in this compilation process, such inserted materials would become effective guides to legal practice. 27 Dou 1984, 12.198. 28 Niida 1964, pp. 837-38 suggests that the change, the distribution of land to agnates as tenants, was an innovation first put into the law in a compilation of ordinances dating from the Chunhua period (990-995). However, he gives no basis for his suggestion. He argues that a statement appearing in a memorial of 1023 (which in turn refers to the 1015 revocation of agnates rights to tenancy of land), when it refers to “the household ordinances”, is referring to this Chunhua compilation, though there is no indication in the passage involved that this was the case. However, during the Song such compilations seem almost always to have been merely the bringing together of previous rules, followed by an attempt to iron out conflicts between such rules and an editing out of obviously no longer appropriate language (such as the use of terminology stemming from compilations under previous dynasties). During the Song the commissions that compiled rules seem rarely to have used these occasions as opportunities for ?making substantive revisions in existing laws, Dou Yis insertions in the Song Penal Conspectus being the most notable exceptions. Such rules carried over from dynasty to dynasty and were useable until they were replaced. Thus, even if it were true that the rule involved first entered a compilation of ordinances during the Chunhua period, this tells us nothing about the timing of the original enactment of the rule involved. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 325 ,had rescinded close agnates prior rights to rent confiscated juehu landed property 29from the state.) The memorialists in this piece says that they had “examined the edict of 1015, which ruled that the landed property of extinct households was no longer to be divided equitably among close agnatic relatives [as tenants]. The money from the 30sale of such land was to go to the authorities ... This is a change from the old law.”) No longer could agnates hope for even the limited privilege of access to the confiscated portion of juehu landed property. The law might also have had a significant effect on local government units, by giving them more opportunities to maximize their returns on confiscated lands. This law of 1015 was the cause of continuing controversy. Clarifications and amendments, added in 1019 and 1023, indicate that the new policy was being 31enforced.) Land, confiscated by the state from juehu families, was being sold or rented out, with the current tenants supposedly having first option, followed by 32neighbors, and then the deserving poor.) In 1026 an attempt was made by the government to widen the pattern of distribution: The Judicial Control Office said, “We have investigated and determined the articles concerning extinct households. From now on, if households that have died out have no unmarried daughters living at home, but do have daughters that have married, [these married daughters] should be awarded one-third of the wealth (zicai), land, buildings, and moveable goods, minus funeral expenses. If there are no living married daughters, then give one third of such assets to married aunts, sisters, or nieces. As to the remaining two thirds, if during the time that the deceased was still living, there were blood relatives or uxorilocal sons-in-law, foster sons, sons who had followed their mothers into the marriage [but who had not thereafter been adopted by the new husbands], who had previously been co-dwelling with them as tenants, supposing these sorts of people had been living with the deceased for three years or more prior to his death, then this two thirds of the buildings, wealth, and the landed estate should all be given to them as owners. If there are no living aunts, sisters or nieces, then the total of the assets should be given to these co-dwellers. If the co-dwellers had been living with the deceased for less than three years, or if the deceased had no co-dwellers, the whole should go to the state. Landed property [in the confiscated portion of the estate], in accordance with the ordinance, is all to be divided equitably among close relatives as 33leaseholders responsible for the tax rent.) If there are no close blood relatives then their rights should awarded equitably to old tenants or sharecroppers who, as the owners, will be responsible for payment of taxes. If the deceased has written a will, which on examination proves to be clear, this will is to be followed.” 34OK?d) 29 Xu 1965, shihuo 63.171b. 30 Xu 1965, shihuo 63.171b. 31 Xu 1965, shihuo 63.166b-167a, 171b. 32 Niida 1964. A useful description of this pattern is included in a good, basically descriptive, essay on the juehu system. The author and I were, unbeknownst to each other, doing research on the same topic. See Li 1999. 33 This phrase may indicate a return to the old policy of giving relatives (or in the absence of close relatives, members of the other groups mentioned) preference in renting state expropriated land, if there were no eligible nuclear family members or co-dwellers. 34 Xu 1965, shihuo 61.58a. Copyright ? 2000. All rights reserved. 326 BRIAN E. MCKNIGHT This passage appears to clarify some matters. By implication, if there were unmarried daughters at home, they inherited all the assets. In the absence of unmarried daughters, one third of the assets was to go to married daughters or, in the absence of daughters, to married aunts, sisters, or nieces. The remaining two thirds of the estate, including landed property, went to men who had helped support the household. Such men received the whole of the estate in the absence of qualified female heirs. If there were no daughters, aunts, or qualified co-dwellers, the whole of the estate was forfeit to the government. Presumably, if there were qualified married women or aunts but no qualified supporting males, the state would confiscate two thirds of the estate. The confiscated landed properly would then be turned over, probably in tenancy, to relatives (or in the absence of relatives to others who were not co-dwellers but had been closely associated with the deceased). There would, of course still have been situations in which there were no person so qualified to be tenants. The state would then have confiscated the land and either rented or sold it to outsiders. This ruling indicates that juehu policy had continued to be a subject of debate between 1015 and 1026. The attempt in 1026 to clarify distribution did not settle the issue. In 1027 a group of officials was appointed to conduct a comprehensive review of the numerous laws and regulations governing the distribution of the assets of juehu families. These offi cials submitted a report which analyzed the circumstances under which problems arose. Their report dealt with three distinct types of assets; buildings and liquid wealth, agricultural assets owned by the deceased, and land that had been conditionally sold. If the deceased had left a clear, coherent will, the distribution of assets was not problematic. Song law continued the earlier practice of allowing such wills to override the other rules on the distribution of juehu assets. This was the situation preferred by the government, since it obviated all the potential legal struggles that might follow intestate distribution. There was no apparent dissent among the reviewing committee about this. However, in the absence of a clear will, the offi cials saw the three categories of assets listed above as posing distinct legal and practical issues. There apparently were no perceived problems with distribution of the estate to daughters at home. Indeed, the complete distribution of such assets to daughters at home was so unproblematic that the writers did not even mention it in their recommendations. Nor was the rule calling for a one-third distribution of such wealth to married daughters (or married aunts, nieces, or sisters in the absence of daughters) a source of trouble. The matter of lands alienated through conditional sale (dianmai 典賣) is touched on only briefly in the report as it has come down to us. In conditional sale the owner could sell a piece of property while retaining the right to redeem the property, after a period of time specified in the contract, by repaying the amount of the original sale to Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 327 the person who had purchased use-rights. If the owner or his successors failed to make the appropriate payment at the time fixed, the land became the property of the buyer. The officials reviewing juehu legislation recommended that such properties not be treated according to the design suggested for other properties. Instead, another assessment of the value of these conditionally sold properties was to be made, so that 35a settlement could be reached.) The major provisions of the laws concerning the proper disposition of juehu lands when the property had been alienated through conditional sale, a matter of considerable concern to officials during the middle of Zhenzong’s reign, had been covered in rules promulgated in 1014 and 1018. In 1014 the authorities dealt with the problems created when the household to which the land had been mortgaged through such conditional sale died out. The original owners had already received money in return for the use of the land. What, then, was to happen to the land? Prior to 1014 the authorities apparently had been confiscating such properties, but there would seem to have been no proper provisions in law about the situation. Then, in 1014, it was ruled that in all such cases the officials who confiscated the properties were to assess the values of these properties, make clear records of these assessments, and permit the original owners to redeem the properties for the original value of the loan, with a due 36date extended half a year beyond the date set in the original contract.) Four years later, in 1018, the authorities ruled on the obverse situation in which, after having turned over landed property to others through conditional sale, the originally owning families died out. Two officials from Sichuan reported that in their areas there were numerous lawsuits that resulted from this situation. The crux of the problem was that, after the extinguishing of the original owning households, those who had merely secured conditional sales which gave them the use of the properties, did not report the situation to the government but instead tried to take over permanent ownership of the properties. This of course provoked lawsuits from some of the neighbors, who under Song law might have pre-emptive redemption rights to lands 37held through conditional sale.) In the end, the central authorities issued orders that holders of such lands from juehu households were obligated to report this to the authorities within one year. The current tenants, i.e. those who had created the 35 Xu 1965, shihuo 61.58b. 36 Xu 1965, shihuo 63.165b. This law did not apply if thirty years had pass without there being an extant contract or even in the presence of a contract if that document was unclear. 37 People who had redemption rights over land that changed hands through what might best be called conditional sale (dianmai 典買,were called parties of related concern (//az/en rew 連分人).Parties of related concern, a group that might include close relatives and neighbors, were, at most times in the Song, supposed to sign agreements of conditional sale and had the right to redeem the property at the end of the term specified in the contract. For an example of this see Minggong shupan qingming ji [hereafter, Qingming ji] 1987, 4.106. McKnight and Liu 1999, pp. 171-72. Such parties could not, however, assert a right of redemption if the land involved had been sold to a relative. See Qingming ji 1987, 6.197-98; McKnight and Liu 1999, p. 245ff. Copyright ? 2000. All rights reserved. 328 BRIAN E. MCKNIGHT contract of conditional sale, could then secure clear title by paying the authorities the 38current market price for the property. Failure to report could result in confisca- tion.) The policies adopted for agricultural land held in this sort of tenure seem to have worked reasonably well. The offi cial review group in 1027 apparently made no major suggestions for changing them. However, the reviewers were concerned about the provision of the 1026 law mandating distribution of the remaining two thirds of assets to various codwellers of the deceased. They feared that the ambiguities of this provision would give rise to legal troubles and lawsuits. (It is possible that the state had no rules on just how such assets were to be divided up among possible claimants though it is perhaps more likely that such rules existed but were internally contradictory or were inadequately known.) The officials asked that an order be issued clarifying the specifics of such distributions. Further clarifications were added, including a stipulation that when estates exceeded a certain value, an imperial decision was required for their distribution. This situation is reflected in a brief passage dated 1028: Xiung Prefecture reports that the household of the commoner wife, the lady nee Zhang, has died out. In law one-third of her landed estate (tianchan 田產,should be given to her married daughter. As to the other two-thirds, although there is a co-dwelling uxo- rilocal son-in-law, still the value assessed is more than 10,000 strings of cash. This needs to be memorialized for a decision. The emperor said, “This is wealth that has been managed by this household from morning til evening. 39Do not profit from their loss by confiscating it. It should all equitably be given out.. .”) These rules were further clarified after the Administrator of Kaifeng Prefecture in 1030 submitted a report critical of some sorts of state confiscations. He begins by citing the law, initially established in 1023, under which: there was to be an assessment [of the value of confiscated property] with the information being given to the current tenants who, in accord with the price set, might make a purchase of the property as their estate in perpetuity. If the current tenants did not have the wherewithal, then the neighbors were to be asked. If the neighbors did not want to make the purchase, then landless middle grade or lower grade households might do so. He then goes on to complain that this confiscation system had been unfair in cases in which there were close associates of the deceased who had been living with him and who had continued to farm the properties involved after his death. They had, moreover, continued to pay taxes and had “worked the land diligently”,which probably means had improved it through their effort. He then suggests that: 38 Xu 1965, shihuo 63.166a-b. 39 Li 1965, 106.6a. We must assume in this incident that the wife of the uxorilocal son- in-law was already deceased. Otherwise at that point in the dynasty the married-out daughter would not have received one third of the estate. The idea that officials should memorialize for imperial decisions before distributing estates of great value continued during the Southern Song. The Emperor Xiaozong (r. 1163-89) is reported to have reacted exactly like his Northern Song predecessor when the issue of distributing an estate worth more than 20,000 strings was brought to his attention. See Tuo 1977, 200.4993-4994. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 329 all foster sons, heel-following husbands [men who married widows and went to live at their homes], uxorilocal husbands, as well as relatives of the deceased and dependents who, prior to Jingde 1 (1004) had been living with [the extinct households as] tenants, these continuers of the extinct households who up to the present have continued to pay in [the taxes] ought to be allowed, at the offices, to submit statements, which shall be checked to verify the facts and, except in the cases of daughters married out, whose situation is governed by the original legislation, [the land] shall all be awarded (1^^) to the current tenants, [with the registration] being changed so that their households are listed as the owners. Where [the properties] have already been assessed, the 40original decree should be followed.) His suggestion was approved. This piece of legislation broadened the pattern of distributing assets by including heel-following husbands as potential heirs and by eliminating the three-year co-residency requirement, at least for those whose association with the deceased had occurred a quarter of a century earlier. Probably very few people would have been affected by this rule. Indeed, it suggests a specific item of legislation submitted by a local official in response to a specific local situation. However, it also indicates clearly that during this era the authorities, and almost certainly social opinion in general, felt that the immediate supporting group of a landowner was paramount, not agnatic relatives. Agnatic relatives (and for that matter affinal relatives) were not excluded from shares, but those shares came to them because they had been co-dwellers with the deceased and active participants in the working of his lands. As such, they had to share their legacy with others who were not blood relatives. It is also important to recognize that these people were not being offered an option to purchase the property; it was being given to them. They had become heirs in a full sense. The acceptance of this proposal reaffirmed the decisive disinheriting of agnates whose claim to be heirs rested solely on their kinship relationship. There were to be some later, relatively minor, modifications of these rules, but in their general pattern they continued into and perhaps to the end of the Southern Song. After 1030 some aspects of the rules and policies concerning juehu remained stable for many years. The system in 1030 contained the following elements: 1) If the husband whose family line died out had established a clear will prior to his death, this will took precedence over all rules of intestate succession. This rule was to remain in force throughout the rest of the dynasty. 2) If, at the time that the household became juehu, there were unmarried women living at home, a category which included divorcees and widows who had returned to their natal home as well as never-married girls, all of the assets including land were awarded to them. 3) If there were no such unmarried women at home, but there were married daughters, these married daughters received one-third of the complete estate. (If there were no married daughters, but there were married aunts, nieces or 40 Xu 1965, shihuo 61.57b-58a. Copyright ? 2000. All rights reserved. 330 BRIAN E. MCKNIGHT sisters, they would receive this third.) The other two thirds would be awarded to certain sorts of male co-dwellers, if they had lived with the deceased in his household for at least three years prior to his death. (A special rule covered the specific case of those whose control over land was tracable to having been co-residents during the opening years of the eleventh century.) 4) In the absence of daughters at home, married daughters, or married aunts, nieces, or sisters, such qualified co-dwellers received the whole estate. 5) If there were no qualified heirs as defined above, all assets were forfeit to the state. There is some uncertainty about possible disposition of such properties. It may be that the state distributed the tenancy of landed property to close relatives as residual last choice heirs. In the absence of qualified parties, the state might sell land at market price, if that were possible, or call in tenants if it proved impossible to dispose of it by sale. The proceeds supported regular government operations. 6) If the household that died out held lands through conditional sale, the original owners or their heirs had the right to redeem the lands according to the original terms of the contract. 7) If the family that died out had alienated its own lands through conditional sale to others, these current land holders were supposed to inform the authorities of the situation within a fixed period. The authorities would assess the value of the properties, and would deliver clear and permanent title to the current holders, in return for a payment of the current value. Later Song Modifications of the juehu Inheritance Pattern In the era that followed, from the early 1030’s until late in the Southern Song, there were sporadic disagreements about the policies governing juehu, the social 41impact of the policies,) the continuing problem of corruption and abuse of the 42system by government functionaries,) and the appropriate distribution of assets 41 The official Yu Jing, in an undated memorial, complained about the impact of the current policy of turning the land over to the current tenants. He was appealing for some way of keeping the land in the hands of members of the family whose line had been extinguished. To no avail apparently, at least in this era. See Yu 1976, 12.16a. Li Xin, who complained about the vexatiousness of the closely written rules, has no better answer than to suggest even more rules, designed supposedly to limit abuses. See Li 1935, 22.16a-17a. 42 The problem of corruption, usually involving officials abusing their positions to force confiscations when these were uncalled for, is referred to repeatedly in the sources. See for example at report from 1043 (Li 1965, 145.15b-16a) and an even more interesting incident from 1055. This latter case involved a foreign merchant who had made a fortune, adopted as his son a boy originally brought into his household as a slave, returned to his native country, and was there killed by the local ruler. The local officials tried to treat this as a case of juehu, which in turn led the young man to send two agents to the capital, apparently to plead his case. The brief report on this affair in the Xuchangbian says that the property was to be confiscated because the supposed heir could not produce documentation proving that Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 331 43confiscated by the state.) These concerns, of course, overlapped at times in legislative responses. The rule makers themselves seem to have been guided by predictable motives: (1) their assessment of state needs, (2) political pressures from those who differed on the policies, and (3) their views about what constituted a fair and equitable distribution of resources. The resulting rules concerning the distribution of estate portions to possible claimants seem to have involved the interaction of four underlying principles: (1) relationship, with more closely related relatives receiving larger portions, (2) residence, with those co-dwelling with the deceased in a privileged position, (3) contribution, with an express desire to recognize as heirs those who had contributed, economically or ritually, to the household enterprise, and (4) need, with those felt to need support most, being most 44benefitted.) Four types of persons, from groups that sometimes had overlapping memberships, were of special concern to the rule makers: 1) co-dwellers, including relatives, foster sons, uxorilocal sons-in-law, sons who had followed their mothers into marriages (without being thereafter adopted by their mothers’ new husbands), and co-dwelling tenants. 2) married-in sons-in-law, who were co-dwellers but were also felt to have claims in the property for additional reasons. 3) daughters, including daughters married out, uxorilocally married daughters, unmarried daughters living at home, and divorcees and widows who had returned to their natal homes prior to the deaths of their parents. 4) male heirs posthumously appointed by the government on the recommen- dation of a group of agnatic elders. An examination of these groups in Song juehu inheritance laws speaks volumes about the government’s views and intentions. Co-dwellers as Inheriting Shareholders The list of qualified co-dwellers who stood to inherit shares under certain conditions is remarkable. It includes not only relatives, but also various sorts of unrelated parties such as foster sons and tenants, who had contributed to the welfare of the juehu household prior to the deaths of the husband and wife, and who therefore were felt to deserve shares. The willingness on the part of the authorities to go beyond kinship in distributing resources is one of the most interesting facets of Song the merchant had paid certain due taxes. No such failure is mentioned in the longer description found in Su 1965, 5.4a-b. Corruption continued to plague the system. See an incident dated 1251. Tuo 1977, 173:4180. 43 Initially the property seems to have been used as government office lands. Later, after suggestions by some high officials, it came to be used, at least in part, to support local charitable functions. See for example the ideas of Han Qi, cited in Li 1965, 186.6a-b. 44 Xu 1965, shihuo 61.62a. Copyright ? 2000. All rights reserved. 332 BRIAN E. MCKNIGHT rules. Relationship by blood or marriage was the overwhelming determinant of inheritance during other dynasties. After the initiation of inheritance rights for co-dwellers in the law of 1026, their opportunity to benefit steadily increased. One interesting modification was the promulgation, at some point prior to 1086, of a set of rules specifying benefits to co-dwelling claimants who had not lived with the deceased long enough to qualify for a share by earlier rules, but who had, by their efforts, increased the value of the property by one thousand strings or more. This policy is first mentioned in a report from the period 1086-93. During the period 10981100 the policy was changed so that the claimant had to have doubled the value of the estate, but in 1101 it reverted to the 45pattern followed in 1086-93.) Policies giving co-dwellers rights to some inheritance shares under certain conditions continued into the last years of the Southern Song. A law code that dates from about the turn of the thirteenth century includes an ordinance specifying the amounts of juehu estates that might legally be used to cover the funeral expenses of the deceased. Fixed amounts to cover funeral costs would be made available to "close agnatic relatives or others qualified to receive parts 〇/从e 似此“應得財產者 ”46,Italics added—tr.)) A half century later, the text of an amnesty reflects even more clearly the continuing legal rights of codwellers: The executive officials in the prefectures and districts are not allowed, in violation of the law, to ?confiscate peoples properties. Warnings have been most severe, yet covetous offi cials and venal clerks often do not ask whether the guilt is serious or not [in cases of confiscating the lands of criminals], nor do they look to see if there are codwellers who have rights to shares in the 47property. . .) [italics added—tr.] Married-in Sons-in-law as Inheriting Shareholders Among co-dwellers, uxorilocally married sons-in-law occupied a unique position. They were not simply co-dwellers, but also relatives by marriage and, in addition, were the principal figures responsible for the welfare of daughters of the families. In the most obvious sense, such men would gain from inheritance through their wives, because in juehu families daughters at home inherited wealth. (A number of cases suggest that married daughters living at home were at times treated like unmarried daughters living at home, that is to say, from the state’s point of view, the critical factor seems to have been that they lived at home, not that they were married.) Uxorilocal sons-in-law also merited shares as contributing co-dwellers. The preceding section described the laws of 1026 and 1030, which mandated distribution of some household assets to a wide variety of co-dwelling relatives and dependents. 45 Xu 1965, shihuo 61.62a. 46 Xie 1976, 51.487. 47 Tuo 1977, 173:4180. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 333 48Married-in sons-in-law were specifically included as one of the categories.) These policies seem aimed at giving benefits to those who had contributed to household welfare; other state policies seem intended as rewards for those who made contributions to the general welfare. In 1083 the central authorities approved a proposal for Hebei circuit. Under this policy married-in sons-in-law, as well as fostered sons or grandsons, sons or grandsons who had followed their mothers or grandmothers into second marriages (without being thereafter adopted), or husbands called in by widows (“heel following husbands”), if they were currently fulfilling 49household baojia) local service obligations, were to be given half shares at the time 50of family division.) The phrasing of this memorial clearly connects it with the rulings of 1026 and 1030 which mandated the distribution of shares of juehu properties to certain categories of qualified co-dwellers. Married-in sons-in-law also came to have rights in juehu property even in the presence of heirs posthumously adopted for the extinguished households. A discussion dated 1162 says: The Acting Administrator of Yuan Prefecture, Li Yu, said, “A recent decree ordered that estates be equally shared between [posthumously] adopted sons and married-in sons-in-law . .. Suppose that there was an estate worth 1,500 strings. Under the current law someone inheriting under a will can only receive one-third [of an estate of that size], with an equal amount being given to the adopted son. If we were to give each person 750 strings [by applying the 1131 ruling on equal division] this would run counter to the law on bequests. I ask for a decision.” The Ministry of Revenue investigated to see if other local jurisdictions had faced this situation, where the amount of a bequest exceeded the amount under the regular rules. The Ministry recommended that, apart from the amount given to the legatee according to the stipulations, the remainder be given completely to the adopted son. If the amount of the property did not fill up the amount under the law on bequests, it was to be equally divided in accord with the recent decision. This meant that if the amount were 1,000 strings or more, they would each receive five hundred strings; if 1,500 strings or more they would receive one third. When the property reached 3,000 strings application 51of this rule was to be stopped. The remainder was completely to be awarded to the adopted son.) This concern, to provide for the uxorilocal sons-in-law in families for which heirs had been appointed after the family line had been extinguished, rested on at least two considerations. First, the estate would be shared by the couples composed of the married-in sons-in-law and their wives, who were the daughters of the deceased couples and who therefore deserved special consideration. Second, the adoption necessarily had occurred after the arrival of the married- in sons-in-law. They and their wives presumably had been engaged in caring for the parents before the arrival ,of posthumous adoptees. Given their contribution to the households welfare, in 48 Xu 1965, shihuo 61.58. 49 In traditional China many low-level administrative tasks, tax collection, road repair, and so on, were undertaken by groups of people selected from the local population to staff state mandated organizations. The baojia was perhaps the most famous of these. For more information see McKnight 1971. 50 Li 1965, 332.13a. 51 Xu 1965, shihuo 61.66. Copyright ? 2000. All rights reserved. 334 BRIAN E. MCKNIGHT fairness they deserved to share in the wealth. A complicated Southern Song case, decided by an official known as Wu Shuzhai, also reflects this concern about the legitimate rights of uxorilocal sons-in-law and their wives in the presence of posthumously appointed heirs. Under the division mandated by Wu, one half of the estate was to go to the confirmed heirs and one half ,to the uxorilocal sons-in-law, because “the married daughters are the familys own blood and the uxorilocal sons-in-law have been living in the families of their wives for many years.” Thus, they deserved to receive a major share of the assets on the grounds both of closeness of relationship and of contribution to the welfare of the 52family.) There is some evidence indicating that this practice of equal division 53continued to be a part of Song law.) This decision was by no means unusual. In another case from the Collection of Enlightened Judgements the ruling official even disinherits an acknowledged natal son of the deceased, giving the whole of the estate to the uxorilocal son- in-law and his wife, on the grounds that the natal son had behaved in a grossly unfilial way towards the father, who had been supported and cared for by the son-in-law and 54daughter.) Finally, in still another case, three uxorilocal sons-in-law brought a lawsuit seeking to have the estate of their deceased father-in-law divided equally among three married daughters and the one unmarried daughter still at home. The sons-in-law seem to have been quite aware of the laws that called for shares to be given to qualified co-dwellers who through their labors had increased the value of the estate. The judge is also aware of the rules. Indeed, he cites the specific law: “All married-in sons-in-law who, in the process of managing the property of their wives, have increased its value, are to be given three-tenths of the total as their share, if the household is extinguished.” However, he did not allow their case to stand. The sons-in-law had claimed that a son purportedly adopted by the deceased was in fact a mere foster child and in any case had been too old at the time he entered the family to have been legally adopted. (Under Song law, children of different or unknown sur-names were not to be adopted if they were older than three years.) The judge dismisses their charges. First, records proved that the deceased had during his lifetime adopted the son, so the son could not be dismissed as a mere foster child. Second, the young man, although originally of a different surname, had behaved in an exemplary manner as a son, eventually marrying and begetting a son. The living grandson meant that the household could in no way be said to have died out. In the end the judge shies away from issuing a final resolution on the distribution of the 52 Qingming ji 1987, 7.205-06. For a translation see McKnight and Liu 1999, pp. 222-25. 53 See a case in Qingming ji 1987, ch. 7 where, in a dispute between a married-in son- in-law and an adoptee, reference is made to the “old law” which mandated equal division. 54 Qingming ji 1987, 4.126-27; McKnight and Liu 1999, pp. 156-57. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 335 estate, passing that responsibility on to others, though it would seem he favored a suggestion made by an earlier trial judge, that the property be equitably divided 55among all the claimants.) The salient point here is a hypothetical one. Had the adopted boy proven to be a foster child (and thus not eligible to inherit fully), or had the judge ruled his adoption improper, the daughters would have stood to inherit all the assets. Many of these cases reflect another social fact. Whatever the state of the law might be at a given time, married-in sons-in-law often assumed that they had legitimate inheritance rights, to the point where they would bring suit in court to enforce them. In his collected works, Liu Zhi mentions an incident in which an upright man married into a family where there were several other married- in sons-in-law. When the parents died, and the family prepared to divide its estates, the other sons-in-law readied lawsuits to enforce their claims to the property coming to their wives. The subject of Liu Zhi’s eulogy refused to participate in the scramble for 56property, because it was not proper behavior for a scholar.) Inheritance by Posthumously Appointed Heirs Sons adopted pre-mortem had, at least in law, the same rights as natal sons; sons appointed posthumously did not. Tang law (and early Song law), with regard to certain categories of inherited goods, distinguished between pre- mortem and post-mortem heirs. Where Tang lawmakers could effectively intervene to discriminate against post-mortem adopted sons, in areas like official perquisites which were under the close and immediate control of the state, they did so. Tang and early Song law specifically excluded posthumously appointed heirs from succeeding to their fathers’ titles or other official perquisites. Thus, posthumously appointed heirs might succeed to some but not all of the valued possessions of their adoptive fathers. Later Song law extended this discrimination against posthumous heirs to new classes of inherited goods. Song lawmakers did not create the legal status of posthumous heir, but they were more concerned to distinguish clearly the statuses of pre-mortem and post-mortem appointed heirs than the lawmakers in the Tang and early Northern Song. By the early 1090’s the authorities had enacted detailed ordinances permitting closely related agnatic elders to appoint posthumous heirs for juehu families. In 1092 the state ordered a stop to this system and made the authorities themselves responsible for appointing heirs. Two years later the Department of State Affairs criticized this policy because “there are already detailed ordinances that permit close agnatic relatives who are agnatic elders to appoint posthumous heirs for extinct families. It is 55 Qingming ji 1987, 7.215-17; McKnight and Liu 1999, p. 281. 56 Liu 1935, 14.26a. Copyright ? 2000. All rights reserved. 336 BRIAN E. MCKNIGHT 57inappropriate for the authorities to be involved.”) It would seem that a compromise was reached under which the elders recommended candidates for appointment as posthumous heirs, but the authorities retained the prerogative of actually making the appointments. This position became and remained law from that time on. Why this evolution in the practice of posthumous appointment? It seems plausible that posthumous adoption of heirs was little dealt with in the laws of the Tang and early Song periods because it was not, in practice, very common. During the Tang, and possibly during the early Northern Song, in undivided families, the landed property of a married brother who died without a male heir passed back to the general pool, though presumably the income from some of it would have been used to support his widow and female children. Under such circumstances, the other brothers had a positive incentive not to encourage the appointment of a posthumous male heir for their deceased sibling. Even if the family property had been divided, the widow without living children inherited the usufruct of her late husband’s land only for her lifetime. If she were to die without living daughters and without having appointed an heir, then the household she had headed until her death would be classified as juehu, and the property would fall into the hands of close agnates, the very group with the 58supposed responsibility for selecting a posthumous heir.) Why on earth would close agnatic relatives have pushed for posthumous adoption of an heir? Such an action would have been directly counter to the self-interest of their own stem families. Historians studying China often seem to believe that the most common way of dealing with the property of a man who died without an heir was for his relatives to 59appoint one for him.) Perhaps this was true in Ming and Qing times. And no doubt this device was used frequently in the Song, especially the Southern Song, but there seems to be no evidence from the Southern Song, much less from the Northern Song, which demonstrates that posthumous appointment of an heir was the most common way of dealing with the dying out of a family line. What the evidence does suggest is that, during the Song, a great many family lines died out; some had posthumous heirs appointed and some did not. In contrast to later dynasties, the Song state did not press for posthumous appointment of an heir, a process called “ordering an heir”,m/ng/Y 命繼,or “continuing the extinguished”60 繼絕).) There were those in the government who 61favored government support of, if not strict mandating of, such adoptions,) but no such policy was adopted. No contemporary evidence is extant that would demonstrate clearly what proportion of families whose lines of descent had ended eventually had heirs appointed posthumously. 57 Xu 1965, li 36.16a-b. 58 See the Tang ordinance quoted above. 59 See for example the comments of Bernhardt 1995, p. 276. 60 Qingming ji 1987, 7.211, 8.247. For a discussion of these issues see Qu 1992. 61 Xu 1965, li 36.2a-b. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 337 The appointment of posthumous heirs on the recommendation of close agnatic kin probably became more common as the rights of such kin to assume control of property of the deceased became more problematic. As agnates rights weakened, their self-interest dictated that they press for increasing definition of the rules governing the appointment of posthumous heirs, and the rights of such heirs to at least some share in juehu estates. As the practice of posthumous adoption became more common during the late Northern Song, the rules concerning inheritance of household property by such heirs created considerable discontent among some of those affected. Some materials suggest that, at that time, the right of such posthumously appointed heirs to inherit was not fully established in practice. In 1132 the Judicial Commissioner for Jiangnan Circuit memorialized concerning complaints that heirs appointed posthumously on the recommendation of family elders were not inheriting the property of their posthumously adoptive fathers. The Commissioner had looked into the law and found that such posthumous adoptions were legal. However, people felt that posthumous sons selected by senior relatives ought not to have the same inheritance ,rights as natural sons or sons adopted during the parents lifetimes. Such posthumous heirs had not borne the burden of supporting the parents during their lifetimes, nor could they have acted ritually within the family during its existence. The Commissioner in this 1132 case suggested that the state apply to this situation the policy used for married- out women when a household was extinguished, that is to 62say the posthumous heirs would receive one third of the estates.) On the positive recommendation of the Ministry of Revenue, his suggestion was approved, though 63an upper limit of 3,000 strings was put of the size of the inheritances.) In seeking to understand the inheritance rules concerning posthumously appointed heirs, the most important point to be borne in mind is that, from a legal point of view, the families had in fact died out. Thus, the shares that might come to heirs named posthumously on the recommendation of close agnatic elders, were governed by juehu legislation. That legislation dealt necessarily with the rights of posthumously adopted heirs and daughters (and possibly co-dwellers). The late Southern Song texts go to some lengths to make quite clear when such posthumous choices by relatives were justified: If there was no adoption during the lifetimes of husband and wife and after they have died the near relatives discuss establishing [an heir], this is what is called "continuing the extinguished^ (jijue ). Even if the husband and wife are dead, so long as the grandmother or grandfather are alive and adopt a grandson, or if the wife is still alive and adopts a son, these cases do not fall into 62 Perhaps the official felt that married daughters living apart from their natal families were analogous to posthumous heirs in that they were not actively engaged in supporting the parents and thus it would be appropriate to award them the same inheritance shares. 63 Xu 1965, shihuo 61.64. As Ye 1993, p. 401, points out, such sons also could not inherit the property of their natal fathers, since they had left those family lines. Copyright ? 2000. All rights reserved. 338 BRIAN E. MCKNIGHT 64the category of jijue.) This same case contains another relevant rule: When there is a desire that [the family line of] a household which has been extinguished be 65continued, and the elder relatives of the extinguished household so order, this is allowed.) By the late 13th century, the rules on inheritance by heirs posthumously appointed upon the recommendation of family elders, had been further refined. According to a Household Ordinance: In cases of households that have been extinguished, when posthumous heirs are appointed, if in the extinguished households there are only unmarried girls living at home, then [the posthumously adopted heirs] are to receive one quarter [of the estates]. If there are also women who have returned to the household, then [the heirs] are to receive one 66fifth.) If there are only women who have returned to the household, [and no unmarried girls], then, in addition to the amounts distributed in accordance with the Law on Extinguished Households, half of the remainder should be given out [to the posthumous heirs], with the remainder forfeit to the government. If there are only women who have married out, the estates should be divided into thirds. Two thirds should be shared equally between the married out women [and the posthumously appointed heirs], and one third should be forfeit to the authorities. If there are no unmarried girls, returned women, or married out women, then the posthumous heirs are to receive one third of the estate, up to a limit on the awards to the posthumous heirs of 3,000 strings. If the value of the estate reaches 20,000 strings, they may receive an added 2,000 67strings.) These rules demonstrate that heirs appointed posthumously on the recom-mendation of family elders received relatively small portions of household estates. There were, however, instances in which they received more than would seem to have been proper under the law. In one Enlightened Judgements case, the posthumous heir is granted three-quarters of the estate, with only one quarter going to the daughter. The small portion given the daughter is probably due to her having been the child of a maid, but it is diffi cult on the basis of the rules now extant to justify 68giving three-fourths to the posthumously-appointed son.) Juehu Inheritance by Unmarried Daughters during the Southern Song The detailed statements of the law and the factual decisions to be found in the 64 Qingming ji 1987, 7.220. 65 Qingming ji 1987, 7.220. 66 Judging from the sense of the ensuing sentence we should interpret this to mean that the remaining four-fifths would be distributed according to the Law on Extinguished Households. 67 This rule appears, in substantially the same form, in two places, at Qingming ji 1987, 8.266-67 and Qingming ji 1987, 8.288. Xu 1965, shihuo 61.64a also refers to this rule. The latter occurrence is pointed out by Ye Xiaoxin in his recent survey of Chinese civil law (Ye 1993, pp. 400-402) and is noted again in a fine recent article on adoption in the Song published in 1998. See Liu 1998. The last provision here clearly indicates that the decision of 1132 about awarding one third of the estate to the posthumous heir still stood in the thirteenth century. 68 Qingming ji 1987, 7.238-39. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 339 Enlightened Judgements clearly indicate that, in the Southern Song, daughters continued the traditional pattern under which they inherited shares of the entire estate. This was true whether the daughters were unmarried or married. Several cases illustrate the legal position of unmarried daughters and grand-daughters. One case, probably dating from the late years of the first half of the thirteenth century, involved two girls caught up in the warfare along the Song northern border. The elder girl, captured by the Jin, eventually was able to return to the Song. The younger girl was rescued on the battlefield by a gallant Song officer. The judge says, in ruling on the distribution of the agricultural properties of the deceased family head, “since the two girls [a natal granddaughter and an adopted daughter of the deceased] are now here, they should inherit. It is not 69right for the government to collect the rental income.”) In another case involving inheritance of land by a daughter, the judge in his initial remarks says, “Xiung Zi died and his widow, the woman Gan, remarried, leaving behind a daughter. The household of the late Xiung Zi owned three hundred and fifty ba of land [one ba is one twenty-fifth of a mu]. At the time the land was worth less than three hundred strings, so it was by law left entirely to the daughter.” [italics 70added].) The judge in still another case defends the rights of two daughters (by a concubine) to inherit their father’s estate. The concubine, their mother, was by his order to have managerial control over the properties as a fund for supporting the family and raising the girls to adulthood. The assets would come to them when they 71became adults.) Finally, another case indicates that even daughters born by 72maidservants had inheritance rights to a part of their father’s agricultural lands.) Daughters Married Out Despite the views of some scholars to the contrary, the lot of women married out 73with sisters still at home improved in the late Northern Song.) It is true that, from the founding of the dynasty until the 1090’s, women married out could not benefit 69 Qingming ji 1987, 7.287-89; McKnight and Liu 1999, p. 353ff. We must assume in this case that the opinion that has come down to us in abbreviated form. The two girls had a posthumously named adoptive brother. Under Southern Song law in such circumstances the girls would have shared three-fourths of the property, with the posthumous adopted heir receiving one-fourth as his share. 70 Qingming ji 1987, 4.110-11; McKnight and Liu 1999, pp. 172-74. 71 Qingming ji 1987, 7.232-33; McKnight and Liu 1999, p. 306ff. 72 Qingming ji 1987, 7.238-39; McKnight and Liu 1999, pp. 315-16. 73 See for example K. Bernhardt, who believes, incorrectly, that married daughters shared the juehu inheritance with other sisters during the Tang and early Song. Thus she sees the rights of married and returned daughters being “severely curtailed” during the course of the Song (Bernhardt, 1995, p. 273). In her eyes, during the Southern Song, if married out sisters “had either unmarried or returned sisters, they would receive nothing” (Bernhardt 1995, p. 274). The Tang view that daughters married out ought not to share in the estate is also reflected in Tang theoretical judgments (pan). See Li Fang 1965. Copyright ? 2000. All rights reserved. 340 BRIAN E. MCKNIGHT from the juehu estate, if they had sisters still at home. However, in 1098, the rule that women married out could benefit only if they had no sisters at home was modified. An edict stated that, if the estate was worth more than 1,000 strings, the married daughters would receive a portion, size unspecified, even if there were sisters at home. If there were only returned women living at home, these returned women were to share among themselves two-thirds of the estate. Women married out were to share one-half of the remaining third. The text does not indicate what was to happen to the remaining one-sixth share, but it probably would have been confiscated by the state. If there were only married daughters, the sizes of their portions depended on the size of the estate. If it was worth less than one hundred strings, they received it all; from 101 to 300 strings they received 100 strings; if from three hundred strings to 20,000, they received one third (probably a reflection of the more general one-third 74rule inherited from earlier times),) up to a maximum of 2,000 strings per person; if more than 20,000 strings, they could petition the state for an exception to the 2,000 75string maximum.) The legal position of such women continued to improve during the Southern Song. Early in that period the maximum size of their share, when they 76had sisters at home, was raised to 3,000 strings.) If there were no sisters living at home, married out women would seem to have done far better. A decision found in the Enlightened Judgements awarded a married out daughter one-half of the entire estate, with the rest forfeit to the authorities, 77because there were no other daughters and no posthumously appointed heir.) If a posthumous heir had been appointed, but there were no other sisters, the estate by law was to be divided into three equal portions. The state was to confiscate one of these portions. The other two were to be divided equally among the posthumous heir and 74 Li 1965, 63.1068. This one-third rule became so much a part of the thinking about the shares of married women that an emperor used it in the extraordinary circumstance of dividing up special grants given to a deceased official. The one-third share rule for married out women continued in use at least into the late 13th century, with some important modifications which set upper limits on total awards. 75 Li 1965, 501.5b-6a. 76 Xu 1965, shihuo 61.64. 77 Qingming ji 1987, 8.280-82; McKnight and Liu 1999, p. 347. There is a little confusion among the judges in this case. Those in the lower courts had felt that the daughter should receive one-third, in light of the well-known one-third share rule for married daughters that went back to the Tang. The official who wrote the opinion given in the Enlightened Judgments overrules the decisions of the lower courts with a very lawyerly and persuasive argument. While it is true that married out women should receive one one-third as their share, in this specific incident the woman involved had still been unmarried and living at home when her parents died. The fact that she later married was legally irrelevant. She had to be given a share as a still unmarried girl. The disturbing element here is the fact that this girl did not inherit the whole estate. Perhaps the judge was thinking in terms of male inheritance. A sole male heir, natal or adopted pre-mortem, would have inherited the entire estate. If he had had sisters at home, they would have received shares one-half the size of his share. Thus a daughter at home, without a brother, should receive one-half of the share her (nonexistent) brother would have received if he had been the sole heir. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 341 78his adoptive married sisters.) (It is of course possible that the 3,000 string limit still applied though it is not mentioned in the text.) Women Who Had Returned to Their Natal Homes By contrast, the lot of women who had returned to their natal homes as a result of divorce or widowhood became worse off during the course of the Song (if it is assumed, as I do, that for inheritance purposes such women were included among “daughters at home” in the Tang code). Under the decree of 1098 described above, which specified shares for women still married out, married women who returned to the family home, either as widows or divorcees, were also included. Such returned women seem to have shared equally with unmarried daughters still at home, a 79continuation of the probable Tang and Northern Song proportions,) but their share would obviously have been decreased by the provisions for their married sisters. If there were only returned daughters and women married out, the rule of 1098 specifies that the returned women were to divide two-thirds of the estate. Because the state confiscated a one-sixth share, returned women fared less well than unmarried women would have done. Materials from the thirteenth century tell us that, if returned women had to share the estate with unmarried daughters living at home, the returned women only received half shares. “According to the Ordinances, all the wealth of households that are extinguished is to be given to the unmarried girls living at home. Women who have returned home [after having been married] are to receive shares half as large [as 80those given girls who had never married].”) If there were only returned women, they only received half the estate, down from the two-thirds of the late Northern Song. The posthumously named heir received one-quarter, and the remainder was forfeit to 81the authorities.) 78 Qingming ji 1987, 8.266-67. We should note that such married women inherited by law only if the household was extinguished. Even those women who had brought matrilocal husbands into the family did not always inherit if there was a legitimate heir. See Qingming ji 1987, 7.215-216. Yuan Li also interprets a comment in another decision, Qingming ji 1987, 8.280-282, to mean that married out women only receive one-third of the estate. See Yuan 1988, p. 274. 79 See Li 1965, 501.5101 (1098 C.E.) where it is said that “the property (caichan) of extinguished households is to be given equally to unmarried daughters living at home and returned women.” As indicated above, evidence about the inheritance rights of women who had returned home after divorce or widowhood during the late Tang and early Song are not clear. Given what is now extant, I am inclined to the position that the phrase “living at home” included such women. 80 Qingming ji 1987, 7.217, 8.287-89, 9.316. 81 An important case, which specifies the proportions to be awarded to various sorts of daughters and posthumous heirs, says that when there were only unmarried women and returned women, they were to divide four-fifths of the estate. However, it does not specify that their shares of this four-fifths were to be equal. See Qingming ji 1987, 8.287-89. Copyright ? 2000. All rights reserved. 342 BRIAN E. MCKNIGHT Expansion of State Claims The expansion of state claims was one of the most important changes dis-82tinguishing Song from Tang law.) However, the extent of these claims needs to be seen accurately, in context, and in all its complexity. First, it is important to think about the Song legal changes against the background of the Tang system. In theory at least, the Tang state laid claim to all agricultural property. In practice, even in those areas and times when there were serious efforts to enforce the equitable fields system, the state’s ability to control property affected mostly the personal share lands which 83were supposed to be periodically redivided and reassigned by the state,) not the so-called perpetual lands. Furthermore, in the late Tang, when the equitable fields system had collapsed in practice, state law had not yet adjusted fully to the reality of near universal private ownership of land. Thus the Song inherited a legal system that assumed (1) families would continue to live together, a situation in which, under most circumstances, land would remain within the extended family group when a member died, (2) if there were legitmate male heirs at the time the mother and father died, the entire estate (minus amounts of marriage expenses of unmarried sons and daughters) would go to the male heirs, leaving nothing for the state, (3) if the family line (and its household) died out, the assets, aside from the portion needed for funeral expenses, would go (a) entirely to daughters at home or, in the absence of daughters at home, (b) to married daughters. Moreover, in the extant sources there is no specific revocation of the Tang mourning rule mandating that, in the absence of daughters, property went to close agnates. It is therefore possible that at the time of the founding of the Song, such agnates would have inherited if there were no living daughters, married or unmarried. Thus the state would have confiscated the assets only under quite 84limited conditions.) Measured against this minimalist set of claims on property, there can be no doubt that over time state claims increased, beginning with emendation to Tang rules found in the Song Penal Conspectus that awarded married women only one third of the estate, with the state confiscating the other two thirds. The state further increased its control with the law of 1015 that reduced the rights of agnates to have first option to be tenants on juehu lands. However, the state’s penchant for confiscasting property, 82 The point is well made by Bernhardt 1995, p. 275. 83 The Tang regime had inherited from the preceding northern states a land policy under which the state laid claim to all land. The portion not reserved for the elite was supposedly to be parcelled out to farming households on the basis of the number, gender, age, and status of members. One category of such land, called personal share land, was to be used for annual crops. The size of this portion changed as the composition of the household changed. The other category, perpetual lands, intended for longer term crop commitments, remained with the family unit. By the late Tang this system was no longer enforced. 84 The impact of valid wills under these conditions is a bit problematic. Legally valid bequests could, in general, be used to distribute assets only to mourning degree agnatic and affinal kin, though in practice they were sometimes used to pass property to some other closely associated individuals. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 343 though of great importance, was modified in 1026 and 1030, so that the inheritance rights of a whole list of qualified codwellers, including not only relatives but also others, were established. In the absence of more detailed information on the household structures of juehu families and the effectiveness of state enforcement of the rules, it is impossible to know how much in practice the state’s share actually changed. It seems plausible that the state got much more, but without knowing how common it was to have qualified co-dwellers, we cannot estimate accurately how their shares would have reduced the state’s confiscation of such inheritances. Caps on Inheritance Scholars have correctly noted that state claims were also increased by the expedient of placing caps on inheritances. To understand the impact of these caps accurately, caps on testate inheritance and on intestate inheritance must be distinguished. The earliest caps were concerned solely with limits on inheritance through bequest. Tang law (and early Song law as embodied in the Song Penal Conspectus) seemingly placed no limits on the size of estates that could be distributed by bequest. However, at some time between 1064 and 1093, strict limits were placed on the value of property that could be distributed by will. Only one-third of estates worth more than 1,000 strings could be devised. For estates worth between 300 and 1,000 strings, three hundred could be distributed. For those worth less than three hundred strings, the entire estate could be distributed. A later request that these limits be repealed was approved by Zhezong (r. 1085-1100). In fact, they conti皿ed 85to be applied.) The remainder above these limits was, presumably, confiscated by the state. Such limits seem to have remained in place during the Southern Song, though the amounts were made somewhat more liberal. The incident from 1162 cited above 86indicates that the basic limit at that time was 500 strings.) Thus, without question, in the latter half of the eleventh century, limits were placed on the amounts that could be distributed by bequest, a practice that continued during the Southern Song. Beginning in the late Northern Song, caps were also placed on juehu intestate inheritance, at least for married women. As noted above, after 1098, if there were only married daughters in a juehu household, married women could inherit all of an estate worth less than one hundred strings; if the value was from one hundred to three hundred strings, they could received one hundred strings; if from 300 and up they could receive one third (probably a reflection of the more general one-third rule 87inherited from earlier times),) with a cap at 2,000 strings. If the estate was valued at 85 Li 1965, 383.2b. See also Xu 1965, shihuo 61.61. 86 Xu 1965, shihuo 61.66. 87 Li 1965, 63.1068. This one-third rule became so much a part of the thinking about the shares of married women that an emperor used it in the extraordinary circumstance of dividing up special grants Copyright ? 2000. All rights reserved. 344 BRIAN E. MCKNIGHT 20,000 or more, they could petition the state for an exception to the 2,000 string 88maximum.) Early in Southern Song the maximum size of their share, when they 89had sisters at home, was raised to 3,000 strings.) At some point, caps were also introduced for other categories of heirs. The most widely-known passages describing limits are to be found in the Enlightened Judgements. In one passage the magistrate involved speaks of a small landed estate being left in its entirety to the deceased’s only child, a daughter, because the total 90value of the estate was less than 300 strings of cash.) At the end of a more famous and far more important passage, the official writing the Enlightened Judgements opinion says “if there are no unmarried girls, returned women, or married out women, then the posthumous heir is to receive one third of the estate up to a limit of 3,000 strings. If the value of the estate reaches 20,000 strings, he may receive an added 2,000 strings.” In the following sentence, the author goes on to speak about the young daughter and grandaughter of the deceased, still living at home, who according to the rules on juehu inheritance, would equally share any inheritances up to a limit of 3,000 91strings each.) This passage seems very clear. If it were the only evidence available, it would be reasonable to conclude that the state limited each share in a juehu inheritance to 3,000 strings of cash, with a possible increase to 5,000 strings, if the estate exceeded 20,000 strings. However, there is other evidence inconsistent with this passage. The most extended description of a juehu incident occurs in the Southern Song work, Daily Notes of Master Huang (Huangshi richao). This long piece concerns an unfortunate district sheriff named Yue Yi. The infant son he had adopted to be his heir died at less than two years of age. Shortly after this he adopted the ten year old daughter of one of his agents, presumably to take care of him, since he had divorced his wife, who had remarried and moved elsewhere. When the adopted girl came of age she married, taking with her a handsome dowry of land and silver. Sheriff Yue never remarried nor did he adopt another heir. Thus, when he died, his household was classified as juehu. At that point, the story began to become complicated. A relative nominated his own son to serve as the posthumously adopted heir to the adopted infant who had died young. However, under Song rules, a child of less than two years could not stand in the line of succession and so could not itself have a posthumous successor. The adopted daughter apparently claimed that she was in fact the birth daughter of the deceased man, but the court found her claim unacceptable. Neither the district court nor the prefectural court wanted anything to do with what was given to a deceased official. The one-third share rule for married out women continued in use at least into the late 13th century, with some important modifications which set upper limits on total awards. 88 Li 1965, 501.5b-6a. 89 Xu 1965, shihuo 61.64. 90 Qingming ji 1987, 4.110-11; Liu and McKnight 1999, p. 173. 91 Qingming ji 1987, 8.287-89. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 345 clearly a messy case. However, those concerned caused so much tumult at the prefecture that, in the end, the authorities there dispatched a functionary to investigate. The investigator appointed was deeply fearful that a fortune built up over many years might be the lost, if an acceptable heir were not quickly appointed. If the girl was not to be considered an appropriate daughter, and if the establishing of an heir for the deceased child was illegal, then the letter of the law called for the confiscation of the entire estate by the government. The most disturbing aspect of the situation was the distinguished lineage involved. Sheriff Yue was the descendant of a well-known Song official. Although he himself had been a mere sheriff, still it seemed intolerable for the state to confiscate his fortune. The investigator recommended the appointment of a suitable heir. Following the rules on the establishing of posthumous heirs, the authorities consulted with Sheriff Yue’s relatives and drew up a genealogy. This genealogy showed that there were three descent lines stemming from the well-known progenitor. One consisted only of the deceased sheriff and a man who had been a district registrar but whose line had also died out. In the second line, at the appropriate generation to provide a posthumous heir, there were only two sons. The elder, by Song law, could not desert the line of his own father to become the heir of someone else and the younger had already been appointed as the posthumous heir of the deceased district registrar. That left only the third line, in which fortunately there were three sons. The eldest was, as just noted, not allowed to leave his own line. The youngest was a young man of no learning who had opened up a shop to sell dyes. That left only the middle son of this third line. He also was not particularly learned, but at least seems to have been liked by his relatives, and thus could serve as an acceptable, if not an ideal, heir. All of this throws interesting light on the authorities willingness, under some circumstances, to intervene actively in seeking the appointment of a posthumous heir; more pertinently for our focus here, the essay goes on to spell out in some detail just how the estate of the deceased sheriff was to be divided up. The writer says that the young man, Wenbing, was to be established as the posthumous heir of the sheriff and so “according to the rules he is to be given one-third [of the estate] up to a limit of twenty-thousand [strings].” (italics added) This is a long way from the 3,000 strings spoken of in the piece from the Enlightened Judgements. The writer then goes on to distribute other shares of the deceased’s estate, giving 1,000 strings to another member of the agnatic kin group, and 1,000 strings to cover funeral expenses. He says that all of these sums, under the “law on the equitable giving out of juehu [assets]”,are contained “within the additional 20,000 strings that can be given away.” (italics added) The adopted daughter, of whom the adoptive father had been Copyright ? 2000. All rights reserved. 346 BRIAN E. MCKNIGHT 92very fond, was to benefit to a sum of 10,000 strings.) Because the deceased had already provided her with agricultural fields and gifts of silver, these prior gifts were to be considered in reaching this amount. The young man who had gone to serve as the posthumous heir of the deceased registrar, had at one point lived with the sheriff. To him the authorities were to award a sum of 5,000 strings. To the young man who had been nominated to serve as the posthumous heir of the deceased infant boy, they awarded 3,000 strings. The two brothers of the successful candidate were to be awarded a thousand strings each. After all these distributions, whatever was left over 93was to be confiscated by the authorities, to be used for worthy purposes.) All these figures are vastly beyond what might have been expected, given the limit of 5,000 strings per share to be awarded for estates worth in excess of 20,000 strings according to the Enlightened Judgments case. What is to be made of this? One might, at first thought, assume that the authorities in this incident were simply setting aside the rules as a special favor to an official of distinguished lineage. But that simply does not fit with the way the piece is worded. The text, in awarding such a huge sum to the named posthumous heir, says literally, “now we have established Wenping to be the successor to the extinguished household of the district sheriff. According to the stipulations, he is to be given one-third [of the estate] up to a limit of 20,000 strings”,今立文炳以繼縣尉之絕, 照條三分绐一,至二萬贾而止).In describing the awarding of the other amounts, the text says that “all this, under the law on the equitable giving out of jwe/zM [property], falls within the extra giving out of 20,000 strings”,凡此於戶絕均給之法多給二萬 貢內).Several possible solutions to this puzzle suggest themselves. Perhaps each of these laws, the 3,000 (or 5,000) string limit law and the 20,000 plus 20,000 string limit law, were practiced but at somewhat different times. There is, however, no indication in other sources of any such extraordinary alteration in the limits. It also seems possible that, although the basic rule was one which set a limit of 3-5,000 strings, there was also a separate rule, not cited in the Enlightened Judgments case because it did not come into play, which covered the distribution of truly enormous estates. Until further evidence surfaces this conundrum must be left an unsolved puzzle. Thus, the basic observation that the Song state set upper limits on amounts of juehu inheritance seems clearly true. However, it should be pointed out that even the lowest limit, 3,000 strings, was the rough equivalent of the price of from two to three hundred mu of land in the early thirteenth century. The 20,000 string limit was the rough equivalent of fifteen hundred to two thousand mu of land. These limits thus 92 This is a curious amount. If the girl had been properly adopted by the Sheriff, then under the law described earlier, she ought to have been awarded one third of his estate. See for example Qingming ji 1987, 8.266-67. 93 Huang 1971, 78.36a-40b. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 347 affected either the quite well off or the incredibly rich. This puzzle is only one part of the larger difficulty of understanding an extremely complicated legal and social situation. Unmarried daughters, returned daughters, married-out daughters, uxorilocally married daughters, close agnatic and affinal relatives, uxorilocal sons-in-law, posthumously adopted male heirs, concubines, maids, foster sons, other sorts of co-dwellers, holders of property through conditional sale, and neighbors with redemption claims on conditionally sold property—the mere list of the sorts of persons who might be involved in an actual juehu situation suggests something of the number of possible combinations of parties. And, in real life, the situation was even more complicated. Whatever the law might say, in practice people, including most pertinently officials, felt differently about daughters born of maids, concubines, and legitimate wives. Throw into this mixture the fact that the children involved might or might not be minors. Add the differences between testate and intestate inheritance. And, in addition, add the fact that at times the kinds of assets inherited might affect their potential distribution. The end product was perhaps the most complicated legal situation in Song China. Small wonder that it has not been clearly understood by scholars looking back across a millenium. Who Benefits? We can cut through some of this forest of complications, and begin to understand what happened during the Song, if we first focus on a simple question— cut bono? The juehu inheritance situation as it existed in 1030 was summarized at the end of the first section of this paper. What was the situation some two centuries later? And most importantly, who benefitted? What parties had become stronger, and what parties weaker, in the centuries since the early Song? 1) Will makers —lost some of their old powers during these centuries, because fairly restrictive caps were placed on the amounts they could dispose of through a testament. 2) Agnatic relatives—lost their rights to inheritance and could usually hope for a share of property only if they were qualified co-dwellers with the deceased or if their sons or grandsons were posthumously adopted as heirs. 3) Affinal relatives—benefitted only as co-dwellers, as recipients of assets given in wills, and because heirs seem quite often to have been chosen from the families of the wife or the mother-in-law, despite rules discouraging cross-surname adoptions. 4) Unmarried daughters—saw their possibilities for inheritance reduced to an unknowable degree by caps on the amounts of inheritances. Their shares were also reduced to an unknown degree by the legislation of 1098, which allocated shares of an estate to married women, even if there were sisters still at home. 5) Returned women— may have seen their potential shares decline. The leg- islation of 1098 seems to imply that they may have received smaller shares than their unmarried sisters at home. Otherwise it is difficult to see why, when Copyright ? 2000. All rights reserved. 348 BRIAN E. MCKNIGHT there were only returned women, they were to inherit only two-thirds of the estate. This would clarify the situation in the Southern Song, when returned women received shares half the size of their unmarried sisters. 6) Married daughters at home—benefitted as married women, and also seem to have been, in practice, aided by the tendency for property to come either to them or to their uxorilocally married husbands through bequest. 7) Married women—saw their situation improve in the 1090’s when they became eligible for shares, even though they had unmarried sisters. 8) Uxorilocal sons-in-law —could benefit in practice as qualified codwellers or as the beneficiaries of special legislation such as that aiding baojia workers. They also might benefit from rules rewarding them for having increased the family’s fortunes. But in practice probably they were most often beneficiaries through wills. They figure prominently as a group given shares in testaments. 9) Unrelated co-dwellers (such as foster sons) —might receive shares of the estate, if they were qualified either through extended residence with the deceased or because they had made significant contributions to increasing the value of the estate (though we should bear in mind the several Enlightened Judgments cases in which magistrates send such men back to their natal kin groups, emptyhanded). 10) Holders of land gained through conditional sale from families that thereafter ,died out probably on balance lost out. They gained, in the 1020s, the right to make their possession permanent and their title secure by paying the market value of the estate to the authorities. On the other hand, previously they might have hoped to gain such control without paying anything, through the failure of the selling family to redeem the property. 11) Posthumous heirs — occupied an odd position. It has been maintained, on the basis of no evidence that I have seen, that, in Tang and early Song law, they were treated like pre-mortem adoptees in the distribution of assets. Setting this unresolved question aside for the moment, I have argued that such adoptions were relatively rare during the first part of the Song. After agnates were cut out of the inheritance pie, the posthumous adoption of heirs became more common. However, documents suggest that, at times, they were not even given shares in the inheritances. If these reports reflected practice in some regions, then posthumously adopted heirs may have actually benefitted under Southern Song rules, which accorded them at least some, albeit rather modest, legal claims on the juehu estate. 12) The state—obviously was one of the biggest winners in the division of juehu estates, in law and practice. At the beginning of the Song the state received property only if there was no clear will, and the couple had no daughters at home or married out and no close agnates. With the promulgation of the Song Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 349 Penal Conspectus the authorities began to confiscate two thirds of juehu estates where there were only married daughters. Then, in 1015, the state moved formally to increase the flexibility of its control over confiscated property by stripping agnatic relatives of their prior rights to tenancies. Although juehu policies were modified in 1026 and 1030, still the state eventually came to control a larger share of the juehu pie. The authorities increased their take with caps on testate inheritance (inaugurated in the late 11th century) and then on intestate inheritance (in the late Northern and early Southern Song). We should, however, in trying to assess how these laws worked in practice, remember that the old debate between then confiscators and those in favor of leaving land in private hands did not stop with the promulgation of rules favoring the state. Cases recorded in the Enlightened Judgments make it abundantly clear that at least some judges were loathe to confiscate property if some reasonable way could be found to leave it in private hands. Still, the amount of revenue involved could be extremely large. As early as the middle Northern Song confiscated estates could in total be worth huge amounts. The official Wen Tong reports that five prefectures in 94Sichuan, in a few months, confiscated property worth four million strings.) In noting the gains for the state it is important to think about what this meant in terms of the eventual winners and losers. The state is, at least in part, a redistributive device and a service organization. Our sources indicate that the assets accruing to the state from the confiscation of juehu wealth, were sold or rented and thus served state units as sources of ongoing revenues. All other things being equal, to the degree that ongoing state organizations—like local government offices—were financed by juehu office lands, the winners were local taxpayers. The income derived from juehu properties was also used by the state to finance welfare and relief activities such as local schools and relief funds. In such cases the winners were the indigent and suffering. No doubt a substantial portion of the wealth thus gained went into the pockets of officials and underlings. However unsavory this might appear, it also would have decreased the demands of such people on their long suffering subjects. In short, we ought not to think of confiscation by the state as some sort of fiscal black hole into which assets disappeared without being of benefit to society. This sort of analysis needs to be pursued further. What are the implications of limiting the right of people to devise their property? We know that, under Tang and early Song law, a man whose family line was in danger of dying out could himself draw up a will which would be the ruling document for the distribution of all of his property after death. Under Song law he could distribute this property to any mourning degree affinal or agnatic relatives. What this situation reflects is the freedom of a property owner to distribute property with limited interference from the 94 Wen 1973, 39.7b. Copyright ? 2000. All rights reserved. 350 BRIAN E. MCKNIGHT state. In theory, it would seem, a man could use this power to disinherit his daughters, to give his entire estate to affinal relatives, or to award his estate to one favorite—a daughter, an affine, or an agnate. Did such things happen? Probably not. Wills were valid only if their makers were held to have been of sound mind at the time when they ,were drawn up. Wills that ran blatently contrary to peoples innate sense of justice 人95情,would probably have been thrown out.) We do not know of such a case from Northern Song times, but there is a case in the Enlightened Judgments in which a magistrate, while feeling himself compelled by the written law to uphold a will that altered the intestate distribution scheme in such a way as to reduce the shares of the daughters, expresses his outrage with the results. Had the girls been totally disinherited, he would have been quite within law and practice to overturn the will. In short, wills supervened the rules on the intestate distribution of property, provided that the wills were clearly drawn and their terms tolerable. Thus, the state’s creation of caps on the amounts that might be devised by will was a very significant reduction in the freedom of individuals to distribute their wealth. The greatly weakened position of agnatic relatives in the scheme of inheritance needs to be matched with the increases in the position of co-dwellers, a classic example of a shift from an emphasis on consanguinity, blood relationship, towards associational solidarity, a shift made so famous by Sir Henry Maine in his study of ancient law. Maine saw this movement as the defining change that separated the ancient from the modern. Maine’s thesis need not be accepted in its pure form for us to recognize that he was pointing to a real phenomenon. This change, as it occurred in eleventh century China, represented primarily a recognition in law of common attitudes, rather than a simple reflection of state interests. Song people felt that the stem family was the center of their lives, and that those individuals who had contributed to the welfare of that stem family had more legitimate claims on its resources than mere agnatic relatives, an attitude perhaps related to the increasing social, geographical, and economic mobility of Song times. In trying to understand this change in attitudes it is especially important to bear in mind the impact of the economic revolution though which China was passing at this time. The enormous increase in the market, the almost total monetization of the economy, the much greater fluidity of society that resulted from this contemporary transformation, made it possible for smaller family units to escape from the domination of larger extended kin groups and to pursue their own limited self interest, a phenomenon noted elsewhere and in other historic eras, when money begins to dissolve the power base of groups like lineages (though in China such organizations 95 Liu 1998, p. 106, makes the same point. The term renqing, which is used in rather a different sense in documents in late imperial China, as it appears in the documents of the Qingming ji, clearly refers to an innate moral sense, a sense of justice as fairness, that the magistrates felt was present in reasonable adults. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 351 were later to counterattack, with some limited success, against this transformation). The attitude that those most closely involved in promoting the interests of the stem family should have the greatest share in its wealth also affected popular attitudes towards daughters. They were the children of the deceased couple. Un-married daughters, who had not yet benefited from the family wealth through dowry, were of course to be especially privileged, but other sorts of daughters, those returned home and even those still married, did not by the fact of their having been married and having already taken some of the family’s wealth, cut themselves off from further help. Better them by far than agnates. Posthumously adopted heirs were in a difficult middle position. It seems abundantly clear that Song public opinion clearly saw them as non-contributing latecomers who in no sense ought to have the same rights as pre-mortem adoptees. They had not worked with the family. They had not suffered with the family. They had not mourned with the family. And yet, they did serve as ritual continuers of the family line. In that sense they certainly deserved something, and something they got, 96not a great deal in terms of percentage, but a sufficiency.) The legal provisions for their lot can be viewed as a partial reassertion of the power of agnates, because close agnates were supposed to be the group from which candidates for adoption were drawn. The Roots of Change The preceding discussion has suggested that one important impetus to changes in juehu law was popular opinion, not the opinion of the poor and disenfranchised, but the attitudes of the landowning groups in Song society, both rich and middling, the kind of people who could make their attitudes and feelings about what was fair and appropriate, their renqing, known to people of power. Law in Song China, as elsewhere, often reflected, in a tardy fashion, the alteration of popular attitudes. And yet, the evolution of juehu law cannot be wholly described as a reflection of popular sentiments. State interests, if not dominant, were also ineluctably involved. One line of argument, adopted by a number of Chinese and Japanese scholars writing on the Song, and repeated by Kathryn Bernhardt, is that the insatiable need for revenues fueled the drive for legislative changes. The Song state was chronically short of funds during the last half of the eleventh century. This fiscal problem was even more pressing in the twelfth and thirteenth centuries. During the late Northern Song and the Southern Song this lack of funds was in substantial part caused by military and other costs associated with the very real threats from beyond the borders. One can certainly argue cogently that the placing of caps on inheritance, testate and intestate, and the continuing problem of abusive confiscation of estates, were both in 96 On this general point see also Liu 1998, pp. 98-99. Copyright ? 2000. All rights reserved. 352 BRIAN E. MCKNIGHT part traceable to this crying need for revenues. However, this simply will not do as the explanation for the great spate of juehu reforming legislation issued between 1009 and 1030. These years span the last part of the reign of Zhenzong (997-1022) and the opening years of the reign of Renzong (1022-1063). During these years, marked by the enactment of the bulk of juehu legislation, the state was in the best fiscal shape ever to occur during the dynasty. Indeed, the state was running revenue surpluses. However important the need for revenues may have been in shaping later legislation concerning the inheritance of wealth, it cannot be used as an explanation for the creation of the bulk of juehu 97rules.) I suggest that three major factors were involved in this flurry of law making. Firstly, the over-riding original concerns of the Song government, external security and internal control, had been successfully dealt with. After decades of expansion to the south and continuous warfare with the sino-foreign states along the northern and western frontiers under the first two emperors, the Song state had, by the early eleventh century, reached its basic geographical extent. It remained for Zhenzong, early in his reign, to signal Song acceptance of this situation by negotiating the Peace of Shanyuan in 1005 with the Khitan forces along the northern borders. Internal control had also been stabilized by the continuing reduction in the powers of the military and the creation of an effective civil service. Thus, secondly, the years following the Peace of Shanyuan, with external peace, internal control, and abundant revenues, allowed policy-makers the luxury of expanding and refining internal policies, especially those concerned with social and 98economic legislation.) There were, of course, important internal problems. The focusing of state attention on critical foreign policy issues, and the need for the central authorities to negotiate internal political controls during the period from 960 to 1005 had necessitated a relatively weak state response to low level local control problems, so long as these problems did not threaten to grow into major internal unrest. In the absence of immediate government pressure, locally influential families predictably used this opportunity to increase their landholdings. Thus, it was the problem that Song writers called engorgement that fueled the push for reform in this era. When, in the aftermath of the Peace of Shanyuan, the authorities at last had the leisure to turn their attention to local economic issues, they began the long attempt to limit the trend towards the growth of landholdings so large that their owners could significantly thwart government control. This was a struggle which was to continue at varying levels of intensity for the remainder of the 97 As one last note on the importance of the need for revenues as a factor in defining state policy, we should point out that the sizes of maximum shares for certain people, like married women, actually increased during the early Southern Song, the revenue-hungry era par excellence. 98 For more on the reign of Zhenzong, the era which was the main period of juehu legislation see Olsson 1974, especially chs. 1 and 2. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 353 99dynasty.) The other local economic issues of concern to the authorities at this time was the abandonment of land. This problem affected state policy in at least two ways, the deserting families were no longer contributing tax revenues to state coffers, and the engorging families were often in a position to assert control over the lands thus left unused. Both of these issues appear repeatedly in the memorials and reports submitted concerning juehu legislation. Confiscation and redistribution of land by the state thus could serve two purposes, reducing the opportunities for the rich to become far richer, and allowing the state to bring vagrant farmers back into the producing classes. The third factor was the perceived frequency of juehu. This may mean that family lines actually died out more frequently in the early Song. Or it might reflect a change in the relationship between stem families and larger kin units. As the rise of the money economy made it more feasible for stem families to become more independent of and physically separated from larger kin units, the dying out of family lines became both more obvious and more problematic. Thus, the evolution of Song juehu policies was shaped by several factors —government revenue needs, popular attitudes about who ought to get what, a state concern about the dangers of engorgement by the rich, a related concern about vagrancy and landlessness, and concern about abuses of the inheritance system by both government functionaries and unscrupulous commoners, coupled with the perception that the juehu phenomenon was common and might give rise to increased conflict over property unless appropriate policies were adopted to demonstrate the position that the authorities might take in disputes. The motives of policymakers in designing responses to this situation were probably always somewhat mixed. Moreover, we do the officials of that time a disservice if we forget that they were often engaged in spirited debates about policies, including those concerned with juehu legislation. The legal outcomes were then, as they are here and now and indeed have always been everywhere, very much the product of struggle and often of compromise. What are we to say about ideology, and more specifically about the great intellectual movement that we associate with the great philosophical synthe- sist Zhu Xi (1130-1200)? Zhu Xi and his partisans had important, principled, views on the appropriate relations of women to property. While they applauded women as the appropriate managers of family property, they abhored the prospect of women actually owning property on their own and not as representatives of a male derived lineage line. Much has been said about the attempts by people associated with this increasingly vocal strand of opinion to shape public policy during the Southern Song, 99 See for example Southern Song comments on rich families engorging themselves by occupying juehu lands. Xu 1965, shihuo 61.2a, 10b-11a. Copyright ? 2000. All rights reserved. 354 BRIAN E. MCKNIGHT especially as it affected women and property. Those Song officials who espoused this line of thought did indeed influence the public discourse late in the Southern Song. In later times such views were to rise triumphant, at least in law, on issues of women 100and prop- erty.) Already in the Southern Song their more extreme partisans, men like Huang Gan, were knowingly issuing decisions in conflict with the written law, in attempting to weaken women’s claims to property. However, none of the evidence indicates that this line of thinking had any effect on Song juehu legislation. Perhaps these partisans, understanding better than we can the bewilderingly complexity of the issues and policies involved, had the good sense to avoid attempts to reform this particular set of laws, even though it clearly favored women. Only when, during the Yuan dynasty (1279-1368), this ideology that promoted the interests of the newly resurgent lineage organizations coincided with the political interests of the foreign ruling elite in buying the support of the dominant class, did these ideas come to dominate social legislation. This Yuan revival of lineage organizations, which entailed a weakening of the economic position of women and the wider adoption of the socially conservative ideas of the Zhu Xi school, reflects an enduring pattern of imperial Chinese life. Even within a system dominated by a pattern of diverging devolution wherein the power of women derived in part from their control over property, the co-existence of variant patterns of the transmission of property across generations always left room for modifications and reforms. Gains made by one group in one era were always subject to possible attrition in the next. In the never-ending negotiations about power, among groups and with the state authorities, the relative positions of social groupings rose and fell. The Song, well-known as a period when women exercised a remarkable degree of control over property, reflected especially in their inheritance rights during the Southern Song, was also a period in which the interests of stem families became paramount. Family success strategies were often in part built on solidifying affinal connections and building alliances with unrelated persons who were economically connected with the family enterprise. During the period of economic growth and change that began in the late Tang and only slowed in the mid- Southern Song, this social pattern was in the ascendant. The slowdown of the Song growth curve coincided with the increasing influence of a way of thinking that stressed lineage connections and downplayed the ownership role of women. These developments set 101the stage for the reactionary social movements of the succeeding Yuan dynasty.) The great concern about and the complex pattern of rules created for settling juehu 100 These arguments are taken from Birge 2000. 101 I am using reactionary here in a rather literal sense, simply suggesting that the men who led the movement to revive lineages and reduce women?s ownership of property thought of themselves as (and in part were) revivers of an earlier, and in their eyes more desirable, pattern of kin relationships. Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 355 property issues, so characteristic of the Song, reflect both the focus on family and the importance of women as owners. The reversing of this pattern is reflected in the greater social and legal insistence on the appointment of heirs, even if posthumously, that characterized late imperial China. Never after the Song was there to be such intense concern about the problem of juehu property, because the problem itself had been at least in part dealt with successfully. Afternote The clearest evidence on the disparity between law and practice appears in a document dated 1029. In it the Administrator of Mi Prefecture (密州,in Jingdong ,East Circuit, in commenting on problems in the states system of disposing of confiscated juehu landed estates, says that in his prefecture, during the fourteen years between the order calling for state expropriation of juehu lands in 1015 and the time when he was writing, a total of seventy-seven juehu estates had come into government hands. See Xu 1965, shihou 63.178a-b. No matter how you look at this figure, it seems extraordinarily small. According to Yue n.d., Mi Prefecture in the 980’s had approximately thirty-eight thousand households. See Yue n.d. A later geographic compendium reports that Mi Prefecture in the 1080’s had approximately 150,000 households. See Wang n.d. A little rough division would suggest that the population might have been increasing at roughly one thousand households per year. So, as a first order guess, we might put the population in 1029 at roughly 80,000 households. Most people dealing with the demography of pre-modern peasant societies have suggested that somewhere between twenty and thirty per cent of the couples in such societies would complete their reproductive years without having a living male heir. Let us say for the sake of discussion that about 15,000 of the households in Mi Prefecture at the time we are describing would have not had a living male heir. Even if we eliminate “guest households” (众e/m 客 戶,households who at this date probably owned little if any land) from these calculations, and consider only “master households”,z/m/m 主戶,a term that was at this time applied to landowners), at a guess perhaps 38,000 households, some 7,600 of the cadre would have had no living heir. How to explain the gross differential between this figure and the seventy-two cases of juehu confiscation spread over fourteen years? Many ideas suggest themselves. We need to keep in mind some aspects of the definition of juehu. First, all households of unremarried widows, who would no longer bear heirs for their deceased husbands but who did inherit his property at least until they died, would not be classified as juehu, because until their deaths they had the right to appoint an heir. Second, and of equal or greater importance, daughters inherited juehu property. Even married daughters received one third of the estate. Thus, the state stood to appropriate in full only estates of juehu families that left Copyright ? 2000. All rights reserved. 356 BRIAN E. MCKNIGHT behind no daughters. It might also be true that adoption was far more common than we have other reasons to believe it to have been. State inefficiency no doubt limited the authorities’ knowledge of, or willingness to become involved in, juehu cases. We should probably add also those undoubted cases in which the juehu families had mortgaged land which the receiving households managed to hide from the authorities and claim as their own. Despite all these partial explanations, the number of confiscations still seems very small. In short, there are still some serious puzzles connected with this term and its social meaning but, before we attempt the difficult task of trying to get closer to the practice of Song China, it is necessary that we clarify the legal framework that people obeyed or evaded. BIBLIOGRAPHY Bernhardt, Kathryn 1995 “The Inheritance Rights of Daughters: the Song Anomaly?” Modern China 21.3: 269-309. Birge, Bettine 2000 Women, Property, and Confucian Reaction in Song and Yuan China (Cambridge: Cambridge University Press). Collver, Andrew 1963 “The Family Cycle in India and the United States.” American Sociological Review 28.1: 86-96. Dou Yi寳慊 1984 S洲琴 xiw执?/极朱刑統,Song Penal Conspectus) (Beijing: Zhonghua shuju). Eberhard, Wolfram 1967 Settlement and Social Change in Asia (Hong Kong: Hong Kong University Press). Ebrey, Patricia, 1992 “Property Law and Uxorilocal Marriage in the Sung Period.” In the Proceedings of the Conference on Family Process and Political Process in Modern Chinese Society (Taipei: Institute of Modern History, Academic Sinica), pp. 33-66. Fan Ye,范曄 Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 357 1965 Hou Han shu (History of the Later Han) (Beijing: Zhonghua shuju) Fang Qiao 房喬 77" S/ZM 晉畨,History of the Jin) (Beijing: Zhonghua shuju). 1974 Goody, Jack 1969a “Adoption in Cross-cultural Perspective.” Comparative Studies in Society and History, 11: 55-78. 1969b “Inheritance, Property and Marriage in Africa and Eurasia.” Sociology 3: 55-76. 1969c “Strategies of Heirship.” Comparative Studies in Society and History 15.1: 3-20. 1973 “Bridewealth and Dowry in Africa and Eurasia” In Jack Goody and S.J. Tambiah, eds., Bridewealth and Dowry (Cambridge: Cambridge University Press), pp. 1-59. Hinsch, Bret 1998 “Women, Kinship, and Property as Seen in a Han Dynasty Will.” Toung Pao 84: 1-20. Hollingsworth, Thomas H. 1964-65 “The Demography of the British Peerage.” supplement to Population Studies 18: 1-20. Huang Zhen 黃展 1971 //waw取/zi rie/zac* 黄氏日抄,Daily Notes of Master Huang) (Jinding siku quanshu ed.). Kong Qingming 孔慶明 et al. 1996 Z/z洲狀/nin/a s/zi 中國民法史,A History of Chinese Civil Law) (Jilin: Jilin renmin chuban she). Li Fang李防 1960 Taiping yulan 太平御覽(The Imperial Encyclopedia Compiled in the Taiping Period) (Beijing: Zhonghua shuju). 1965 ying/zMa 文苑英華,Brilliant Flowers of the Literary Garden) (Taibei: Huawen shuju). Li Shuyuan李淑媛 1999 “Tang-Song hujue caichan jicheng zhi fenpei ji qi guishu 唐朱戶絕財產繼承. 之分配及其歸擓,The Division and Ownership of the Inherited Property of Copyright ? 2000. All rights reserved. 358 BRIAN E. MCKNIGHT Extinguished Households in the Tang and Song), which was delivered at the Di wu ji chuanguo lishi xuewen taohm hui 第五展全國歷史學文M論會 in Taibei, in late 1999. Li Tao李燾 1965 XMZ/Z/W_ 鑕資治通趣長編,Rough Draft for a Continuation of the Comprehensive Mirror) (Taibei: Shijie shuju). Li Xin李新 1935 Kuaao ji 跨S集,Bestriding the Turtle Collection) (Siku quanshu zhenben chubian edition). Li Xinchuan 李心傅 1956 Jianyan yilai xinian yaolu (A record of Important Events from the Jianyan Period and After) (Beijing: Zhonghua shuju). Liu Liyan柳立言 1998 “Ymger fmglao 養兒房老•” Zhongguo jiating ji qi lunli yantaohui lunwenji 中國家庭及其倫理研討會論文集,Taibei: Hanxue yanjiu zhongxin),no. 4: 85-lli Liu Xu劉昀 1975 力_M Tang Aw 舊唐眘,Old Tang History) (Beijing: Zhonghua shuju). Liu Zhi劉挚 忠肅集,Collected Writings of Liu Zhi) (Siku quanshu zhenben chubian 1935 ed.). McKnight, Brian E. and James T.C. Liu 1971 Village and Bureaucracy in Southern Sung China (Chicago: University of Chicago Press). 1987 “From Statute to Precedent: the Transformation of the Sung System of Laws.” In Brian E. McKnight, ed., Law and the State in Traditional East Asia, Asian Studies at Hawaii Series, (Honolulu: University of Hawaii Press), pp. 111-131. 1999 The Collection of Enlightened Judgments (Albany: SUNY Press). Mfwgg價g x/iMpaw 而名公畨判清明集 1987 (The Collection of Enlightened Judgments) (Beijing: Zhonghua shuju). “Tang-Song zhi jiazu tongchan ji yizhufa” 唐宋之家族同產及逢囑法,Common Family Property in the Tang and Song and the Laws on its Transmission) (Wang Jianshan, tr.) Shihuo 1.5: 175-190. 1959-64 中國法制史研究,Studies in Chinese Legal History) (Tokyo: Tokyo University Press). 1964 Torei shui (Collection of Tang Ordinances) (Tokyo: Tokyo daigaku shuppankai). Olsson, Karl F. 1974 “The Structure of Power under the Third Emperor of Song China: The Shifting Balance After the Peace of Shan-yuan” (University of Chicago Ph.D. Dissertation). Niida Nobora 1946 Peller, S. 1947 Copyright ? 2000. All rights reserved. WHO GETS IT WHEN YOU GO 359 Shen Yue 沈約 1974 S洲琴 S/ZM 朱害,History of the Liu-Song) (Beijing: Zhonghua shuju). Su Zhe蘇轍 1965 SM /zwang men /洲狀/zwaw /wez/zi 蘇黄門龍川路志,The Brief Record of the Dragon River) (Bai chuan xuehai edition). Tuo Tuo,脱脱 et al. 1977 Song shi (History of the Song) (Beijing: Zhonghua shuju). Wakefield, David 1998 Fenjia: Household Division and Inheritance in Qing and Republican China (Honolulu: University of Hawaii Press). Waltner, Ann 1990 Getting an Heir: Adoption and the Construction of Kinship in Late Imperial China (Honolulu: University of Hawaii Press, 1990). Wang Cun 王存 n.d. 以伽/?琴z/zi 元豐九域忠,Record of the Nine Regions from the Yuanfeng Period) (Taibei: Wenhai chubanshe), 1.7a. Wen Tong 文同 1973 也叩脱咕新刻石室先生丹淵集,Newly Carved Collected Works of Mr. Stone Cottage) (Taibei: Xuesheng shuzhu edition). Xie Shenfu謝深甫 1976 Qingyuan tiaofa shilei 慶元條法事頼,Classified Legal Items compiled in the Qingyuan Period) (Taibei: Xinwenfeng chubangongsi). Xu Song徐松 1965 Song huiyao jigao 宋會要liH (Collection of Important Song Documents) (Taibei: Shijie shuju). Ye Xiaoxin葉孝信 1993 Z/z洲狀min/a s/zi 中國民法史,A History of Chinese Civil Law) (Shanghai: Shanghai renmin chuban she). Ying Zhao 應助,comp.),Wu Shuping 吳樹平,ed.) 1980 琴;yi 力„aov/n•風俗通袭4釋,An Annotated Critical Edition of the “General Explanation of Customs”,,Tianjin: Tianjin renmin chubanshe). Yu Jing余靖 1976 斯叫/ 力武溪集,Collected Writings of Yu Jing) (Siku quanshu zhenben 6 ed+). Yuan Li袁俐 1988 “Songdai nuxing caichan quan shulun” 朱代女性財產權述論,A Discussion of Women?s Property Rights during the Song Dynasty), Songshi yanjiu jikan, vol. 2 (Hangzhou daxue ishixi Song shi yanjiu shi), pp. 271-309. Yue Shi樂史 n.d. /zwa/tyM力太平環宇記,Universal Geography compiled in the Taiping Period) (Taibei: Wenhai chubanshe), 24.207. Zheng Ke 鄭充 1937 Z/zeyw 折獄龜鑑,The Magic Mirror for Solving Cases) (Congshu jicheng edition).44 5Mans Reproductive Activity.” Bulletin of the History of Medicine 21: 51-65. Qu Chaoli屈超立 1992 “Songdai shouyang zhidu yanjiu” 宋代收養 制度 关于办公室下班关闭电源制度矿山事故隐患举报和奖励制度制度下载人事管理制度doc盘点制度下载 研究,A Study of the System of Adoptions in the Song Dynasty), Songdai wenhua yanjiu, no. 2, (Chengdu: Sichuan daxue chuban she), pp. 31-47. Copyright ? 2000. All rights reserved.