1. Yen Chih-t'ui had spent his youth in the south but was taken to the north in 554; in his "Family Instructions," he frequently mentioned differences in the customs of north and south. That he mentioned no such difference in regard to bargaining about betrothal gifts may mean this custom was common in both areas.

2. These were specified lines of the Po-ling and Ch'ing-ho Ts'ui, the Chao-chün and Lung-hsi Li, the Fan-yang Lu, the Ying-yang Cheng, and the T'ai-yuan Wang. Each family was defined by specifiying its focal ancestor, generally two or three centuries earlier, and included several hundred geographically scattered households.

3. There are also indications that in the T'ang betrothal gifts could be substantial even when both families had eminent pedigrees. In a short story by Chiang Fang (early ninth century), a Lung-hsi Li who had recently passed the chin-shih examination found that his mother had engaged him to a cousin, a Lu of a top-rank family ( chia-tsu ). The Lu family expected a betrothal gift of one million cash, even though the groom's family was a relative and social equal. As his family was not rich, Li had to borrow from friends and relatives to collect it all ( TPKC 487:4008). If indeed this sum would be returned as dowry, then Li was collecting money for his own wife and children, and the bride's family, by insisting on the sum, was stipulating the money that was to be earmarked for their daughter's conjugal unit.

4. In fact, when his first wife died, Feng Ching married another of Fu Pi's daughters ( STPC 6:90). Feng Ching may even have lived with Fu Pi, as SSWCL 9:94 says Fu's two daughters and their husbands and children all lived in his house.

5. The contrast between these two approaches is reminiscent of differences between England and continental Europe in early modern times as described by Macfarlane (1986:251-62). In Europe marriages were much more strictly confined to status equals. In England, there was a more flexible process of negotiation in which yeomen's daughters might marry younger sons of gentry, with wealth as much a factor in decisions as "blood."

6. Some forms for T'ang marriage letters ( hun-shu ) survive in Tun-huang. See Ebrey 1985.

7. Such engagement letters are often found in Sung literati's collected works (e.g., CLC 18:247; KKC 68:916-19; LTLWC 2:38-40; PCWC 64:8b-15b).

8. A distinction between land accompanying a bride in marriage and trousseau items was also made by Huang Kan (1152-1221), who argued that land became the property of the husband's family, while a widow could take her trousseau with her on remarriage ( MCC 33:31a-b). Other writers, however, seem to have been quite willing to lump together land and hairpins.

9. I have not seen the term ti-yin elsewhere. Perhaps the eleven thousand is money or goods that will remain under the control of the wife and the five thousand money or goods that will go to the husband or to his larger family. The Sung imperial family similarly seems to have divided marriage payments into two categories, though the terms used are not the same ( SS 115:2732).

10. This legal protection did not last into the Ch'ing period and thus contrasts with the situation described in Ocko's chapter here. Some of the forces leading toward weakening of women's property rights after the Sung are discussed at the end of this chapter.

11. This change had already taken place by the time the Sung hsing t'ung was issued in 963. See SHT 12:14a-b.

12. This generalization is based on reading forty-two epitaphs by nine authors in the CTW , an exercise that did not uncover any references to their dowries.

13. This is based on a study of 161 Sung epitaphs for women, of which 23 (about 15 percent) referred to their dowries.

14. I do not have space here to analyze the legal status of married women's control over their dowry. Suffice it to say that many Confucians—particularly those I have identified as holding a "descent line" ( tsung ) orientation (Ebrey 1984a)—seem to have opposed women's treating their dowries as their private property, though their arguments reveal that such an attitude was common.

15. One story even suggests that the bride's dowry should correspond to her husband's family property, not to the betrothal gifts he sent. In "The Honest Clerk," an official confides to a matchmaker: "My family has property worth 100,000 strings of cash. For a mate I need someone who will bring a dowry of 100,000 strings" ( CPTSHS 13:44; cf. Yang and Yang 1957:29). This story is set in the Sung, but its exact date is unknown and may be early Ming.

16. A few other items of evidence are worth mentioning. Po Chü-i (772-846) wrote some model legal decisions for marriage disputes. Several make mention of the return of betrothal gifts when an engagement was broken (e.g., CTW 672:14b; 673:1b-2a, 18b-19a), but none refers to return of dowry or disputes over dowry, even in cases of divorce (e.g., CTW 672:17a; 673:6a-b, 11a-b). A T'ang guide to marriage ceremonies preserved in Tun-huang gives elaborate details for the ceremony of delivering the betrothal gifts, but no ceremony is outlined for delivering the dowry (Chao 1963:185-86). Yet in Sung times delivering the dowry was seen as an occasion for colorful display ( SMSSI 3:33). Moreover, three studies of marriage in the T'ang (Chang 1976; Wong 1979; and Niu 1987) uncovered no evidence that provision of dowry was a major expense for girls' families.

17. There recently has been considerable debate on how "open" the Sung elite was (Hartwell 1982; Chaffee 1985; Lee 1985; Davis 1986; Hymes 1986). My own understanding of the situation is shaped especially by Chaffee's statistics on the rapid growth in the educated class (as measured by participation in the examination system); by Hymes's demonstration (from a study of one prefecture) that those who entered the national elite were usually already members of the local elite, and his evidence that the local elite of this particular prefecture was almost entirely new in the Sung, having virtually no descent ties to T'ang officials; and by Davis's demonstration of how a relatively minor family, once it gained prominence in the national government, could smooth the way for dozens of descendants. See also Ebrey 1988.

18. Sung writers did not necessarily indicate whether someone of nonofficial background was of merchant background because merchants did not share the esteem of officials or educated landlords. But some references to marriages between families otherwise quite unequal raise one's suspicions. For instance, the epitaph for a woman, née Ch'en, whose father and grandfather had not been officials remarks that her father was too fond of her to marry her to someone in his own town, so he moved to the capital to choose a husband for her ( WLCC 99:629-30). The one he chose was Ch'eng Lin (988-1056), whose father, brothers, uncles, great-uncles, and so on had all been officials ( OYHCC 21:151-54). One can well imagine that the Ch'ens were merchants and offered a handsome dowry to facilitate this marriage into an official family.

19. For instance, Chao Li, a rich magistrate, in about 1030 met a young man who had passed the examinations three years earlier at age seventeen. This man's family seems to have been undistinguished: it had moved to its current location about a century earlier during the Five Dynasties and none of his three direct ancestors had been officials. Chao Li gave him his daughter as well as a handsome dowry ( KSC 53:64b; PCC 37:489).

20. No Sung source I know makes this argument in full, but there are anecdotes of people thinking along these lines. For instance, the epitaph for a woman from the Li family in P'u-yang descended from the aristocratic Chao-chün Li family of the T'ang described her natal family as one that had produced officials for generations and had accumulated a great fortune. While she and her sisters were young, their dowries were prepared and set aside. When the woman in question reached marriageable age, the estate had declined somewhat because her brother Li Ti (971-1047) had repeatedly failed the chin-shih examination. At this point the girl emptied out her dowry chests to give everything to her brother. The brother accepted the property, then out of a sense of obligation to her secretly selected ''a scholar of integrity'' to be her husband, presumably because she could no longer marry someone rich. Li Ti made a good choice, however, for his sister's husband also passed the examination the same year he did ( HNHS 4:3b-4a). In this case, then, property set aside for dowry was diverted to educational expenses, but the great success of the brother in his career made this in retrospect a wise choice and one that they could plausibly claim did not hurt the sister.

21. The lingering interests of a natal family in the disposition of dowry can be seen in legal cases in which the woman died without heirs. A woman's family was explicitly excluded as heirs to her dowry in the Sung code ( SHT 12:12b). If an heir was posthumously set up for the woman and her husband, he would be the appropriate recipient of the dowry, but the wife's family could object if the husband's family tried to keep the dowry without establishing an heir. In a case recorded in Ch'ing-ming chi , a woman, née Ch'en, had brought a dowry of land worth 120 chung when she married into the Yü family. She and her husband both died without children. Her husband's father, an official, did not want to set up an heir (apparently wanting to keep all the property for a young son he had had by a concubine). Ch'en's father, also an official, sued, and Yü was forced to select an heir who would inherit the dowry fields and continue the sacrifices for the couple ( CMC 8:248-49). There are also several other cases in which the existence of dowry property underlies the continuing involvement of a woman's natal family in the affairs of her husband's family (e.g., CMC 6:191-92).