Five Ch'ing Imperial Marriage and Problems of Rulership
1. Here and throughout this chapter, "consorts" refers to the empress and concubines of an emperor.
2. I am grateful to Pamela Kyle Crossley, who generously read an earlier draft of this paper and corrected my references to Manchu names and institutions, and to James L. Watson, who lent me books on Mongol society. I should also record my gratitude to the staff of the Number One Historical Archives in Peking for their help on my research in 1987 and 1988.
3. On Nurgaci's competition with his brother and his eldest son, see Yen Ch'ung-nien 1983:281, 283; on sibling rivalry in Hung Taiji's generation, see Sun and Li 1982:416-17.
4. But in the case of the T'ung-chih emperor, there was no choice because he was an only son. Ch'ing empresses, like their counterparts in earlier dynasties studied by Holmgren, were relatively infertile. Low fertility and high infant mortality meant that frequently there was no empress's son living at an emperor's death, so the deliberate rejection of such a candidate occurred only once in the whole dynasty.
5. For information on these Manchu rulers and emperors, see their individual biographies in Hummel, ed., 1943-44.
6. On the foundation myths, see Crossley 1987. The Aisin Gioro lineage conforms to the definitions proposed in J. Watson 1982; by contrast, the other Manchu descent groups that evolved out of the mukun of the pre-1644 era are called "clans" throughout this paper. See n. 8 below.
7. The 378 members of the main line (1660) are estimated to have increased to more than 29,292 persons by 1915 (Harrell et al. 1985:38).
8. The term "clan" is deliberately used in preference to "lineage" in an effort to differentiate the Manchu mukun from the Chinese lineage, which it eventually came to resemble. The Manchu mukun was a loosely organized kinship group based on the pastoral economy that could fission or combine according to economic circumstances. As Crossley (1987) notes, the production of written genealogies in the eighteenth century artificially fixed the clan as the historical center of Manchu culture and systematized clan identities for the first time in Manchu history.
9. This event is omitted from the TCYT, but appears in the Niohuru genealogy (Huang 1986:636) and has been "filled in" from the shih-lu account by Li Feng-min 1984.
10. According to Shirokogoroff (1973:70), the marriage of a male with father's sister's daughter was rare among Manchus in the twentieth century; this was the "disapproved" form of cross-cousin marriage among Han Chinese (Gallin 1963:108).
11. T'ung-chih and Kuang-hsu were childless; P'u-i, the last emperor, abdicated as a child. Here and elsewhere in this paper, the ranks given to Nurgaci's consorts are anachronistic; the first Manchu ruler to adopt the Chinese ranks was Hung Taiji. See Wang Shu-ch'ing 1986:57-61.
12. Not to be confused with the "T'ieh-mao-tzu wang" (princes of the iron cap), also known as the Eight Great Houses, who were all Aisin Gioro. For sketches of the founders of the Manchu aristocracy, see Yang and Chou 1986:46-63.
13. The marital relationships did not preclude execution of these imperial affines at the time of the San-fan uprisings: see Kessler 1976, chap. 4.
14. The imperial genealogy indicates that the daughter of a Chinese senior vice-president of the Ministry of Personnel, Shih Shen, joined the Shun-chih emperor's harem.
15. This was certainly the case when the Ch'ing is compared to the Northern Sung; Chung 1981:18-20 states that the largest number of consorts recorded in the Northern Sung was for the Hui-tsung emperor, who had nineteen; the average for the nine Sung emperors was ten, as compared with the Ch'ing average of seventeen.
16. During the reigns of the Shun-chih and K'ang-hsi emperors, the lowest rank was the shu-fei (ordinary consort).
17. Princes were in theory allowed to have a wife ( ti fujin ) and four concubines ( ts'e fujin ), but the genealogies also list lower-ranking concubines, called shu fujin, and, lower still, ying-ch'ieh. When a prince became emperor, the titles of his concubines were changed to accord with the imperial ranks.
18. That the Manchus closely studied the late Ming precedents in preparation for the selection of an empress for the first Ch'ing emperor is clear from a Grand Secretariat memorial dated November 25, 1653, in which Jirgalang, Prince Cheng, reported on the procedures used in 1577 to select the Wan-li emperor's empress: see LYYS. My thanks to Susan Naquin for this reference.
19. Wang Tao-ch'eng 1985a:306 argues that the distinction between the hsiu-nü and kung-nü draft came about after the Yung-cheng reign.
20. The horses and saddles that were included were not a Manchu addition: see Soullière 1987:194-95 for the comparable Ming gift exchanges. Horses, sheep, camels, and weapons were also part of the gift exchanges for the marriages of princes and princesses.
21. In the ta-hun section of the TCHTSL, c. 324, the term for dowry ( chuang-lien ) is used to describe items sent by the court to the bride. The illustration of the Kuang-hsu emperor's ta-hun ( CTKTSH :256, illus. 399), which depicts the delivery of the empress's dowry to the palace, is erroneous in its assumption that the chuang-lien being delivered to the palace was provided by the bride's family: archival documents in Beijing include the chuang-lien lists for the T'ung-chih and Kuang-hsu ta-hun (KCTC), with notes on the subagencies within the Nei-wu-fu responsible for supplying each item.
22. This ritual distinction was thus granted only five times in the course of the dynasty.
23. Analysis of the entries in the various sources that supplement the imperial genealogy ( AHCL; T'ang 1923) shows that there are eight imperial concubines (from the Shun-chih emperor on) whose surnames are unrecorded in the TCYT; all were childless. Three additional concubines of the Hsien-feng emperor whose surnames are unrecorded have been identified from archival sources as Yehe Nara women (Wang Tao-ch'eng 1985b:53), and they are not included in the figure cited. There are twenty-seven other concubines whose clan names are known, but whose father's names are unrecorded: of these eleven had children and are recorded in the TCYT, which also records only three of the remaining sixteen childless concubines. All of these twenty-seven concubines occupied the rank of shu-fei or kuei-jen. Imperial Household Department materials held in the Number One Historical Archives suggest that there were even greater omissions of concubines from the TCYT. Comparisons of the TCYT with lists of recipients of the annual palace distribution of gifts, held at the New Year, show that childless kuei-jen, ch'ang-tsai, and ta-ying were most likely to be dropped from the records because of their infertility and low status (Yü 1985b:57 for 1853-54 lists; PAPS nos. 1251, 1254, 642, for palace lists of 1751, 1756, and 1767; PAPS no. 1246, for palace lists of 1830, 1832, 1833, 1846; all included names of ranked concubines who were not listed in the TCYT or AHCL. Another indication of the extent of underreporting is provided by comparison of the numbers of empresses and concubines recorded in the TCYT for the Ch'ien-lung emperor (twenty-nine) with the forty-one women buried with him: Wang Lai-yin 1983:25.