Sungkyun Journal of East Asian Studies Linlan WANG Peking University James LEE Hong Kong University of Science and Technology Cameron CAMPBELL UCLA You may use content in the SJEAS back issues only for your personal, non-commercial use. Contents of each article do not represent opinions of SJEAS. Institutions and Inequality: Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage Sungkyun Journal of East Asian Studies. Vol.10, No.1. 2010 Academy of East Asian Studies. pp.33-61
Transcript
Sungkyun Journal of East Asian Studies
Linlan WANG Peking UniversityJames LEE Hong Kong University of Science and Technology
Cameron CAMPBELL UCLA
You may use content in the SJEAS back issues only for your personal, non-commercial use. Contents of each article do not represent opinions of SJEAS.
Institutions and Inequality:Comparing the Zongshi and the Jueluo in
the Qing Imperial Lineage
Sungkyun Journal of East Asian Studies. Vol.10, No.1.2010 Academy of East Asian Studies. pp.33-61
ABSTRACT
To gain insight into the implications of institutional affiliation for inequality in historical China,
we examine differences in the demographic behavior and social outcomes between the main and
collateral lines of the Qing imperial lineage. The former consisted of direct descendants of Takeshi,
grandfather of the Qing founder Huang Taiji, while the latter consisted of descendants of Takeshi’s
uncles and brothers. State regulations mandated differences in privileges and opportunities
between the two lines, but previous examinations of inequality in the lineage focused only on
differences within the Zongshi. By analysis of a newly expanded dataset constructed from the most
recent edition of the Aixinjueluo Genealogy that also includes Jueluo records, we compare the
quality of data, socioeconomic attainment, and demographic behavior of the Zongshi and Jueluo.
The result is a complex picture of inequality between the two lines that reflects differences in their
treatment by the state.
Keywords: imperial lineage, Zongshi, Jueluo, data quality, socioeconomic and demographic
outcomes
IntroductionThroughout Chinese history, it has been a common practice for the state to definegroups of elite with different levels of entitlements.1 Such state-mandatedinequality is embedded in a long tradition of institutions that redistribute politicaland economical resources to maintain elite status. The most prominent and bestdocumented examples of such populations and institutions in imperial China arethe imperial lineage and the Office of the Imperial Lineage who are welldocumented for at least the last thousand years (Chafee 1999). This paperexamines the implications of such state-mandated inequality within one such elite
Institutions and Inequality:Comparing the Zongshi and the Jueluo
in the Qing Imperial Lineage
Linlan WANGPeking University
James LEEHong Kong University of Science and Technology
Cameron CAMPBELLUCLA
Sungkyun Journal of East Asian Studies. Vol.10 No.1Ⓒ 2010 Academy of East Asian Studies. pp.33-61
34
Linlan WANG, James LEE, Cameron CAMPBELL
population by comparing socioeconomic and demographic outcomes of twodistinct branches of the Qing imperial lineage, the Zongshi, or so-called Main Line,who were descended from Takeshi, the grandfather of the Qing founder HuangTaiji, and the Jueluo, or the collateral lines, who were descendents of Takeshi’sbrothers and uncles.
Comparisons of the Zongshi and Jueluo, qualitative or quantitative, are rare.Historians typically mention the Jueluo in passing as a counterpart to the Zongshiand note that they received less favorable treatment both in terms of money andcareer opportunities (Jing1993; Rawski 1998). One of the few direct examinationsof which we are aware is Hu Xiangyu (2010), which compares legal cases formembers of the two lines. Almost nothing is known about the demographicbehavior and social attainment of the Jueluo because previous relevant work onthe Qing imperial lineage focused almost exclusively on the Zongshi, as they alonehad their social and demographic data entered in machine readable form (Lee,Campbell, and Wang 1993; Lee and Guo 1994; Lai 1997). Published studies of theZongshi examined their infant and child mortality (Lee, Wang, and Campbell1994), fertility (Wang, Lee and Campbell 1995), adoption (Wang and Lee 1998),and marriage (Lee and Wang 2000; Lee, Wang, and Ruan 2001), but were unableto make comparisons with the Jueluo.
We divide the paper into four parts. In part one, we provide background onthe institutional contexts of the Zongshi and Jueluo, emphasizing differences thatshould have implications for their social and demographic outcomes. In part two,we introduce the dataset, which is based on a more recent edition of the genealogythan Lee, Campbell, and Wang (1993) and which also includes newly entered datafor the Jueluo. In part three, we present results on data completeness, and alsotrends and patterns in socioeconomic and demographic outcomes in the Zongshiand Jueluo. These results illuminate strengths and weaknesses of the expandeddatabase as a source for analysis, and yield preliminary findings on differences inthe social and demographic outcomes of the Zongshi and Jueluo. We conclude inpart four with a summary of key results and an assessment of their implications forour understanding of the imperial lineage.
Our paper contributes to the study of stratification in China by highlightingthe importance of the differentiated treatment of the kin of the emperors. Weprovide another example of how population categorization, in this case accordingto degree of relationship to the dynastic founder, and consequent differentialtreatment led to different socioeconomic and demographic outcomes. Theseresults are valid for about 80,000 men in these two lineage branches and byimplication for equal size populations of their daughters and wives, perhaps aquarter million people in all who comprised part of the pinnacle of the Qing social elite.
* This research was supported by NICHD 1R01HD045695-A2 (Lee PI), “DemographicResponses to Community and Family Context.” We would also like to thank Mr. Ju Deyuan forcomments and suggestions based on his expertise on the Qing Imperial Lineage and relevant historicalsources.
1 Chen (2009) elaborates this idea by analyzing inequality in land distribution in 120 villagesfrom Shuangcheng County in northeast China.
BackgroundAt the beginning of the Qing dynasty, the Imperial Lineage was a small elite group,among whom the emperors, princes and dukes (wanggong), were recognized asthe highest strata of Qing society. But as the population grew, a steadily largerproportion of kinsmen who were distant relatives of the emperors had neither titlenor position and lived only on a monthly stipend from the state (Jing 1993). Thisinternal hierarchy based on status, political participation and financial wellbeinghas been discussed previously (Guo 1994; Du 1997; Lai 1997). However, a morefundamental but less studied distinction was made in 1636 (Rawski 1998, 74), inwhich members of the imperial lineage were differentiated according to theirkinship relations to Takeshi, the grandfather of the founder of the Qing dynasty,Huang Taiji. Zongshi, defined as the direct patrilineal descendents of Takeshienjoyed numerous material benefits and career opportunities throughout much ofthe Qing Dynasty. Jueluo, being the descendents of Takeshi’s brothers and uncles,were not accorded the same privileges as their Zongshi counterparts, but had someadvantages over regular bannermen.2 Being Jueluo was more honorable thanpractical, with more implications for status than for material wellbeing. Table 1summarizes some of the salient differences in entitlements and opportunitiesbetween Zongshi and Jueluo. Indeed, in terms of material stipends, the Jueluowere at best only slightly better off than most other Manchu bannermen (Zheng1985). While imperial kinsmen who held noble titles were much better off thanregular Manchu bannermen, such titles were largely restricted to the Zongshi.
This internal hierarchy within the imperial lineage is reflected in thecomparative detail of coverage in such official compendia of regulations andprecedents about the imperial family as the Imperially Commissioned Regulations ofthe Office of the Imperial Lineage (qinding zongrenfu zeli). Most of the text deals withthe Zongshi. References to Jueluo appear mainly in three contexts: 1) familyregulations such as rules related to the naming of a newborn, appointment of thehead of a family branch, and educational opportunities; 2) economic stipends foradult Jueluo males, orphans, widows; and small subsidies for wedding and funeral;and 3) restrictions on their social behavior.3 There is no specific provision in theregulations for the Jueluo regarding the allocation of the most important resourcessuch as princely titles and government positions.4 While Zongshi dominated theprincely titles and were eligible for special job opportunities referred to as“vacancies for Zongshi” (zongshique), Jueluo only had access to the same titles andposition opportunities as other bannermen (Du 1997), especially Manchu bannermen.
35
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
2 We are still in the process of locating materials on regular bannermen and non-bannermen thatwill illuminate differences between them and the Zongshi and Jueluo.
3 For example, Jueluo were like the Zongshi forbidden from marrying brides from outside theBanners (The Palace Museum 2000, volume 2, 349), and were not allowed to leave the capital withoutpermission (The Palace Museum 2000, volume 2, 269).
4 Although the system of titles or positions were the same for Jueluo as those for other Manchubannermen, Jueluo did have special access to titles or positions awarded or assigned according to militarytalent and kinship identity together. See the name list of title and position holders where a reason ofgiving a title or a position is stated briefly in the Imperially Commissioned General History of the EightBanners (Li et al. 2002, 5165-6219). We will explore this issue in more detail in our future work.
The implications of this hierarchy for differences between the institutionaland socioeconomic context of Zongshi and Jueluo, and their differences fromregular bannermen and even non-bannermen, are complex. The researchmentioned earlier by Xiangyu Hu (2010) on the treatment of members of theimperial lineage by the criminal courts illustrates this complexity. His findingssuggest that in some cases, Zongshi and Jueluo were treated better than ordinarybannermen or non-bannermen. In other cases, they were treated much the same.Moreover, there was a strong temporal trend towards ignoring such statusdifferentiation to the point that eventually, even the state treated certain subgroupsof the Zongshi and the Jueluo alike.
Alongside the differentiation between Zongshi and Jueluo, distinction byplace of residence was also important. Lineage members were transferred from
Table 1. Differences in the Institutional Context of the Zongshi and Jueluo
Zongshi Jueluo
Blood Descendents of TakeshiDescendents of people other than Takeshi
Title Princely titles The same to Bannermen
HonoraryFourth Rank in the official No honorary rank for
Statushierarchy for those with no those with noother title other title
PositionImportant posts for princes and The same as Manchuriandukes, “Imperial Clans’ vacancies” bannermen
Having a title: Having a title: 110 to 10,000 taels of silvera 45 to 700 taels of silverb
Stipend Having no title, age 20 and above: Having no title, age 18 and36 taels of silver, above: 24 taels of silver, 45 hu of grain a yearc 21.2 shi of grain a yeard
Subsidy100 taels of silver for wedding, 20 taels of silver for wedding,120 taels for funeral expensese 30 taels for funeral expenses
Privilege over ordinary menPrivilege over ordinary men
Law Special restrictions, more thanSpecial restrictionsJueluog
Notes: a Guo (1994, 121). Either adjusted or abated when it was granted, the actual amount was notthe same throughout the dynasty, which also applies in notes b, c, d, e, and f (Lai 1997;Rawski 1998)b Anonymous online article, Eight Banner and Position, Salary and Title in Qing History.http://bbs.yunjian.com/284933p1 We will refer to official historical materials later.c The Palace Museum eds. (2000, volume 2, 145-6).d The Palace Museum eds. (2000, volume 2, 159).e The Palace Museum eds. (2000, volume 2, 163).f The Palace Museum eds. (2000, volume 2, 168). g There are ritual and political bans especially for princes and dukes, for example, princes(qinwang and junwang), were not allowed to associate with officials expect for business(The Palace Museum eds. 2000, volume 2, 281)
36
Linlan WANG, James LEE, Cameron CAMPBELL
Beijing to the secondary capital, Shengjing, on a sporadic basis throughout theQing dynasty (Du 1998, 400; Rawski 1998, 90; DQSL, Xuanzong, Vol. 199, 1129).5
Most transfers occurred for one of two reasons. Some lineage members weretransferred to Shengjing for punishment and rehabilitation when they committedcrimes or broke rules (DQSL, Renzong, Vol.222, 992-993). Some lineage members,usually ordinary ones without property, were sent to Shengjing to start a new lifeunder special arrangements (DQSL, Renzong, Vol. 270, 652-653; Vol. 273, 711-712). Given the sporadic nature of movement from Beijing to Shengjing, themixture of reasons for movement, and the generally small numbers of lineagemembers living in Shengjing, we expect patterns and trends in the demographicand social outcomes of Shengjing residents to be more erratic and unpredictablethan those for Beijing.
In this paper, we focus on three basic sets of questions about theimplications of differences in the institutional context of the Zongshi and Jueluo.The first set of questions centers on differences in the quality of the data for theZongshi and the Jueluo. Even though the official genealogy was supposed torecord members of both branches, we expect recording to have been morecomplete and detailed for the Zongshi because they had more privileges andrestrictions, and the genealogy was the most important source for confirmation ofeligibility. The state and the members of the Zongshi all had a strong interest in thedata being accurate and complete. With the Jueluo, there was less at stake for thestate and possibly for individuals if they were omitted completely, or were includedbut some details were omitted.
The second set of questions revolves around differences in socioeconomicoutcomes. We investigate whether the privileging of the Zongshi by the state led tothem to have an advantage over the Jueluo in socioeconomic attainment. We focuson differences in the attainment of titles and official positions. While titles weremostly bestowed, many positions were at least in principle awarded based onmeritocratic principles. We expect Zongshi to have been more likely to be giventitles. For positions, we will examine whether the material privileges enjoyed bythe Zongshi also made them more competitive for positions, perhaps by allowingthem to acquire better education or other forms of human capital that wouldadvantage them in the competition for titles. Alternatively, their privileges mayhave reduced their incentives to compete.
The third set of questions revolves around demographic behavior. Weexpect male Zongshi to have been more likely to marry, and more likely to havechildren. In historical China, socioeconomic privilege generally translated intoimproved marriage chances for males and higher fertility within marriage (Lee andWang 1999). We will compare the marriage behavior and reproduction of theZongshi and Jueluo to assess whether the Zongshi experienced an advantage. Wedo not have strong expectations about mortality differences. In China before thetwentieth century, the relationship between socioeconomic status and mortality,
37
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
5 Shengjing is now Shenyang, capital of the province of Liaoning in northeast China.
however, was inconsistent. Privilege was sometimes associated with elevatedmortality (Lee and Campbell 1997; Campbell and Lee 2000, 2004).
Data The Zongshi and Jueluo datasets we compare were created based on the mostrecent publicly available edition of the Aixinjueluo Genealogy (aixinjueluo zongpu).This edition has several advantages over the Jade Records (yudie) used in Lee,Campbell, and Wang (1993) and other studies of the imperial lineage but suffersone major disadvantage. Whereas the Jade Records used in previous studies ends in1922 and only records Zongshi, the Aixinjueluo Genealogy used here includesJueluo as well as Zongshi and records both populations to 1935. Because theGenealogy covers the Zongshi and Jueluo up to 1935, it contains more entries formales than the Jade Records: the Genealogy records 46,194 male Zongshi, 3,151more than in the Jade Records.6 It records 36,112 male Jueluo. The limitation of theAixinjueluo Genealogy is that like most other Chinese lineage genealogies, it onlyrecords male descendants. By contrast, the Jade Records also records daughters.
The details recorded about males in the Genealogy are largely the same as inthe Jade Records. For each male, the Genealogy provides name, generation,positions, titles, lunar dates of birth and death, age at death in sui, father’s name,mother’s name and status, and the number of sons. The lunar dates of other keyevents such as attainment of certain titles or positions are also provided in somecases. Information about marriage includes the number of wives, their names, andthe names, titles, and positions of fathers-in-law. For men who were adopted, theGenealogy provides the name and rank of the adoptive father and the lunar date ofthe adoption. Because brothers are listed together and the names of their father andmother are provided, reconstruction of family trees via automated record linkagewill be as straightforward as in the Jade Records, and future analysis will takeadvantage of linked data.
The difference between titles and positions recorded in our data is that theformer were honorary statuses that came with stipends but may or may not haveentailed duties or responsibilities, while the latter were actual salaried posts in thecivil or military bureaucracy. In some cases the latter also included students whowere preparing for posts or higher level of civil exams. There were actually twotypes of titles, princely and regular, though in this analysis we do not distinguishbetween them. As we mentioned before, the Zongshi had exclusive access to theprincely titles, which had 14 levels (Lai 1994, 137; Du 1997), while the 8-level titlesystem for Jueluo was the same as that for other bannermen (Du 1997; Lei 2003).As for the acquisition of a position, there was no formal distinction betweenZongshi and Jueluo as rigorous and explicit as the one for the assignment ofprincely titles.
38
Linlan WANG, James LEE, Cameron CAMPBELL
6 Surprisingly, 922 men included in the Jade Records are missing in the Genealogy. According toour preliminary review, the missing men are scattered through the Jade Records, not concentrated in aparticular branch in the family. In future analysis we will assess whether there was any particular patternin the omission of men in the Genealogy or whether their omission was simply the product of randomerror.
Table 2 summarizes and compares basic features of the data for Zongshi andJueluo in the Genealogy. From this table, we can see three important features. First,recording is more complete for Zongshi than for Jueluo, especially when it comesto age at death, which may reflect the higher percentage of Zongshi living inBeijing and other factors. Second, in terms of demographic outcomes, thepercentage of those who married and had at least one son are almost the sameacross the two groups. Third, as for socioeconomic attainment, a larger percentageof Zongshi have noble titles, but Jueluo and Zongshi were nearly equal in theirattainment of position. Later in the paper, we will discuss these features in more detail.
The data have some important limitations relevant to the analysis. First, likethe Jade Records, they do not record information about time of marriage. As will beseen later, we have to infer age patterns of male marriage from the proportionmarried by age at death. Moreover, when we examine men recorded with multiplewives, at this point we cannot distinguish between widowers who remarried andmen who were genuinely polygynous, in the sense that they had more than onewife at the same time7. Nevertheless, widower remarriage and polygyny were bothindicative of privilege, thus patterns and trends in the chances of having more than
Table 2. Social and Demographic Results in the Aixinjueluo Genealogy: A Comparison of Zongshi and Jueluo
Data Zongshi(%) Jueluo(%)Located in Beijing 87.4 71.8
Demographic DataDate of birth 99.4 98.3Date of deatha 59.4 44.1Had at least one wife 45.8 43.1Had at least one son 31.8 37.4Adopted 4.8 2.5
Socioeconomic DataNoble title 4.8 2.0Government office 7.5 6.1Mother’s status 99.8 99.4Wives’ status 12.1 8.9
N 44143 35197
Notes: a People born in the early twentieth century were too young to have married, hadchildren, or died by the 1930s, which is when the edition of the Genealogy used herewas compiled, but they are included in the denominators here to show the generalpicture.
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
39
7 Among the 11,255 people who had at least two wives, only 1,381 of the wives had theannotation that was a secondary wife that would help identify her as a concubine, and distinguish herfrom the wife of a remarried widower.
8 A more robust test of hypotheses about underreporting will be done in the future aftercompletion of record linkage and reconstruction of pedigrees. Family lines and individual characteristicswill be checked in order to see the substantial patterns underlining the increase or decrease of the numberof new births.
one wife still help illuminate inequality. Second, as noted earlier, even though theJade Records record daughters, the Genealogy does not. Because the emphasis hereis on comparison between the Zongshi and the Jueluo, we do not make use of theinformation on daughters in the Jade Records. All of the analysis is based on recordsof males in the Zongshi and Jueluo. The information on daughters has alreadybeen used in numerous other publications such as Lee, Wang and Campbell(1994), Wang, Lee, and Campbell (1995), Lee and Wang (1999), and Lee, Wangand Ruan (2001).
Data CompletenessNumber of births by yearThe number of births of males in Beijing Zongshi, Shengjing Zongshi, BeijingJueluo, and the Shengjing Jueluo reflect differences in their growth, especially fromthe eighteenth century onward. According to Figure 1, before 1710, almost alllineage members were born in Beijing, and the majority of them were Jueluo. Thisis the natural result of the original differentiation between Zongshi and Jueluo.Zongshi were defined narrowly as the main line, thus their number of births wassmaller than that of the more broadly defined Jueluo. The growth patterns of thetwo lines diverged early in the eighteenth century. After 1710, annual numbers ofbirths in the Beijing Zongshi increased rapidly until 1850, fell slightly, then reacheda peak again in 1890. In contrast, Beijing Jueluo first experienced stagnation innumbers of births starting around 1725, and then a decline in recorded new birthsafter around 1770. When their numbers of births bottomed out in 1850, it wasonly one third that of Beijing Zongshi. In Shengjing, the story is reversed. Jueluobirths outnumbered Zongshi births over the entire period, and the gap began toincrease in the beginning of the nineteenth century.
In light of the steady growth in the numbers of births for the Zongshi inBeijing, the stagnation and eventual decline in numbers of births to Jueluo inBeijing likely reflects some combination of factors. Candidates include: increasedunderreporting that resulted in the omission of steadily larger numbers of births,lower living standards that led to reduced marriage chances or lower fertility, anddeliberate institutional arrangements. The complex picture of the differencesbetween Zongshi and Jueluo suggests there is no single rule governing the realityor recording of births for Zongshi and Jueluo. Differences in Beijing that emergedin eighteenth century may have been caused by a deterioration in the recording ofthe less privileged group, the Jueluo, but if that is the case, it is unclear why therewas not a similar deterioration in Shengjing, where surveillance and regulationwere looser.8 If the differences in the numbers of births were real, it suggests that inBeijing, less favorable institutional treatment of Jueluo led to a lower livingstandard, reduced marriage chances, lower fertility, and possibly higher mortality.In Shengjing, the fact that Jueluo outnumbered Zongshi, especially in later years,
40
Linlan WANG, James LEE, Cameron CAMPBELL
may indicate that the Qing government dispatched more Jueluo than Zongshi toShengjing. It may also reflect that the Jueluo were somehow able to secure a betterlivelihood than Zongshi that allowed more rapid growth.9
Figure 1. Male Births in the Imperial Lineage by Year
Recording of mortalityInformation on the timing of death is always a key criterion for assessing the qualityand completeness of demographic data. Information about the timing of death iscrucial for analysis of a variety of socioeconomic outcomes and demographicbehavior. From the time of death, we can infer whether an individual survived longenough to be eligible for other outcomes of interest, including marriage,reproduction, and attainment of position or title. Regression analyses ofdemographic behavior and socioeconomic attainment later in this paper all includeage at death as a control variable to account for differences in opportunity by age. Insome descriptive results, we use information on age at death to exclude individualswho died before they were old enough to experience the outcome of interest.Completeness of recording of death is important in its own right, of course, becausethe imperial lineage records are an important source for the study of mortality (Lee,Campbell, and Wang 1993; Lee, Wang, and Campbell 1994). Here, if the record fora person provides the lunar date of death, usually by listing the reign name, reignyear, lunar month and day, but in some cases by indicating that death occurred onthe first day of life, we define their record as having death information.
41
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
9 In the future, we will test these hypotheses by detailed examination of the data and relevanthistorical documents.
Completeness of death recording depended heavily on whether lineagemembers resided in Beijing or Shengjing. These two cities were the primary andsecondary imperial capitals and lineage members were required to reside in onecity or the other. According to Figure 2, which plots the proportions of lineagemembers with death information by year of birth, the Zongshi in Beijing had themost complete information about the timing of death, followed by the Jueluo inBeijing, the Zongshi in Shengjing, and finally the Jueluo in Shengjing.
Figure 2. Proportion Having Death Date by Year of Birth
Death recording was nearly complete for both Beijing Zongshi and Jueluofrom the beginning of the Qing dynasty to the middle of the eighteenth century,but less complete afterwards. At that point, completeness of death recording in theZongshi and Jueluo began to diverge. While over 80 percent of the Zongshi bornbetween the mid-eighteenth and mid-nineteenth century had information aboutthe timing of their deaths recorded, the proportion of Jueluo with complete deathinformation declined to 60 percent in the early nineteenth century and then 30percent in the middle of the nineteenth century. In Shengjing, this decline wasapparent for both Zongshi and Jueluo and began in the very early years of the Qianlongreign (1736-1795).
To assess whether changes in the availability of death information reflectomissions in the recording of deaths that occurred at young ages, Figures 3 and 4plot infant and child mortality by year of birth.10 In historical Chinese sources, and
42
Linlan WANG, James LEE, Cameron CAMPBELL
10 Figure 3 excludes Shengjing Zongshi born before 1730 and Shengjing Jueluo born before1660 because there were too few observations. The same restriction is applied in Figure 4, 6, 7, and 9.
indeed most historical demographic sources, infant and child deaths are the onesmost likely to be omitted, and such omission is typically reflected in implausiblylow levels of infant and child mortality. According to Figures 3 and 4, infant andchild mortality for the Beijing Jueluo and Zongshi were similar until around 1770,suggesting that at least until then, missing death information among the Jueluo(about 20 percent, see Figure 2) was not concentrated among infants and children.By the standards of other historical demographic sources for China, this is anunusual pattern of omission. This may also indicate that until then, the recordingof births was relatively complete.
Figure 3. Infant Mortality Rate by Year of Birth
43
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
Figure 4. Child Mortality Rate by Year of Birth
Infant and child mortality rates for the Zongshi and Jueluo diverged in thelate eighteenth century. Infant mortality in Beijing Zongshi remained stablethrough the middle of the nineteenth century, while child mortality dipped andthen recovered in the late eighteenth and early nineteenth centuries.11 However,according to Figures 3 and 4, recorded infant and child mortality for the BeijingJueluo first declined at the end of the eighteenth century, during the Qianlong andJiaqing (1796-1820) reigns and then fell even further, to implausibly low levels,after the beginning of the Daoguang (1821-1850) reign. The sharp decline ofinfant mortality for Beijing Jueluo after 1780 and child mortality after 1800,coincides with their overall decline in the proportions of individuals withinformation about timing of death in Figure 2, suggesting that in this time period,omission of information about the timing of infant and child deaths contributed tothe reduction in the completeness of death recording for Jueluo. This occurs muchlater than the divergence between Beijing Zongshi and Jueluo in numbers of birthsthat according to Figure 1 began in 1725, suggesting that complete omission ofBeijing Jueluo sons who died in infancy and childhood played at most a small rolein that divergence.
In the period before the end of the eighteenth century when infant andchild mortality for the Zongshi and Jueluo in Beijing appeared to be reliable, theBeijing Jueluo experienced lower infant mortality but higher child mortality.
44
Linlan WANG, James LEE, Cameron CAMPBELL
11 According to Figure 2, there is no sudden deterioration in the record of death date for BeijingZongshi in the beginning of the nineteenth century, so the dramatic increase of child mortality in thisperiod might not be caused by recording, but reflect something real in the social context. We willexamine this issue later in our future analysis of infant and child mortality.
According to Figure 3, from 1650 until 1770, infant mortality among the Jueluoand Zongshi were similar, but child mortality was consistently higher among theJueluo. Studies in contemporary developing countries generally suggest thatdifferences in infant mortality tend to reflect differences in infant care practices thatmay have nothing to do with socioeconomic status. Differences in child mortality,by contrast, are more likely to reflect differences in socioeconomic status that inturn affect nutritional intake and exposure to infection. Accordingly, higher childmortality among the Jueluo may reflect their lower socioeconomic standing.
Comparison of proportions of men surviving to different ages in Figure 5seem to confirm that especially among the Shengjing Jueluo, many of the men whohad no information about timing of death died at later ages. In Figure 5, ShengjingJueluo exhibit a substantial survival disadvantage at younger ages. Later in life,however, they actually appear to survive in much higher proportions than anyoneelse. If this pattern were true, it would suggest an unusually pronounced mortalitycrossover, in which Shengjing Jueluo experienced higher mortality risk in early lifethan other groups, but then experienced lower mortality later in life if they werefortunate enough to survive that long. Given the circumstances, however, wesuspect that among the Shengjing Jueluo, it was men who died at later ages whowere most likely to have missing information about age at death.
Figure 5. Proportion Surviving by Age at Death, Males Born 1644-1849
Determinants of the reporting of death dateIn this section we examine the influence of individual characteristics on thechances of having time of death recorded for insight into possible patterns ofunderrecording. Because we have already assessed recording infant and childmortality through inspection of levels and trends in rates, we are particularly
45
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
interested in clues about the quality of death recording in adulthood and old age.To assess how social and demographic characteristics influenced the completenessof recording of death information in the four combinations of residence andaffiliation, we estimate separate logistic regressions for Beijing Zongshi, BeijingJueluo, Shengjing Zongshi, and the Shengjing Jueluo.
According to the results in Table 3 next page, different mechanismsinfluenced the completeness of death record for Zongshi and Jueluo. For example,in Beijing, although the deterioration in reporting over time is similar, theinfluence of demographic and socioeconomic features differs substantially forZongshi and Jueluo. Having a title, a position, a wife or a son makes Zongshi morelikely to have a death date reported. In other words, the Zongshi whose death dateswere missing were more likely to be at the margin, either because they died young,or grew up but never married or had children, or never held a title or position.Among the Jueluo in Beijing, the underreporting of death date did not exhibit aclear pattern, except that men with at least one son were more likely to have a dateof death. Having a wife, a title or a position didn’ t increase the chances of having adeath date reported, suggesting that omission of information about the timing ofdeath was either influenced by other, unobserved characteristics, or else wassimply random.
In Shengjing, differences between the patterns of death recording forZongshi and Jueluo were much less apparent. As in Beijing, there was a downwardtrend in completeness of death recording over time. In contrast with Beijing, therewas an inverse relationship between the chances of having a wife recorded and thelikelihood of age of death being recorded. We interpret this to indicate that inShengjing, older males are more likely to be missing dates of death than youngermales. As suggested in the discussion of the survival curves in Figure 5, it may bethat in Shengjing, some men were lost to follow up in the registration system afterthey reached adulthood and married. Men who held a position were clearly anexception. In Shengjing, men who held a position were more likely to have a dateof death recorded. Zongshi in Shengjing who held a title were also more likely tohave a death of death recorded.
Socioeconomic OutcomesIn this part we present preliminary results on stratification in the lineage, focusingon trends and patterns in the attainment of noble titles and official positions. Table4 summarizes the percentage of Zongshi and Jueluo who held a noble title or agovernment position. According to Table 4, a much higher percentage of Zongshihad noble titles, especially those in Beijing. This reflected the domination of theprincely title system by the Zongshi. However, in spite of the less favorabletreatment accorded the Jueluo, the proportions of Zongshi and Jueluo in Beijingwho held position were similar. In Shengjing, Jueluo actually outperformedZongshi.
46
Linlan WANG, James LEE, Cameron CAMPBELL
Trends in the proportion of men holding titles reveal that Jueluo andZongshi all had opportunities to hold titles in the second half of seventeenthcentury, but later, only Zongshi held them. Figure 6 and 7 present the proportionsof men born in each year who held a title or position by the time they died,restricted to men who survived at least to age 30. Because the number of titles didnot keep pace with the growth of the lineage, the proportion of Zongshi who heldtitles declined over time. According to Figure 6, the proportion of men who heldtitles started out high for those born in the very beginning of the dynasty, but fellrapidly in the following fifty years. For Zongshi, the proportion declined from 0.58 to 0.25, while for Jueluo, it fell like an avalanche, from nearly 0.6 to 0.07. Forthe following 150 years, the proportion of Beijing Zongshi who held a title fell onlygradually and kept the level above 0.1 in most time. The proportion of Jueluo whoheld titles remained low and never reached 0.05. In Shengjing, the proportion ofZongshi bounced above and below 0.1, but was continuously higher than Jueluoin Shengjing, though not always higher than Jueluo in Beijing.
47
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
Table 3. Logistic Regression of Recording of a Death Date,Imperial Lineage Men Born 1644-1849
Title 6.95 2.56 1.39 0.27Position 10.69 10.21 2.61 5.02
N 23338 16623 1800 3325
Figure 6. Proportion Ever Having at LeastOne Title by Year of Birth, Males Surviving Age 30
Zongshi and Jueluo living in Beijing had similar chances of acquiringpositions, especially after the seventeenth century. According to Figure 7, around70 percent of the Jueluo and more than 40 percent of the Zongshi born in Beijingin the earliest years of the dynasty held a position. The advantage of the Jueluo wasshortlived. For those born between 1675 and 1725, the Jueluo were less likely tohold a position. Afterward, for the next one hundred years, proportions weresimilar for the Zongshi and Jueluo, although the Jueluo often had a slightadvantage. The most interesting finding in Figure 7 is that even in the nineteenthcentury, about 20 percent of Zongshi and Jueluo in Beijing held positions. Thissuggests that even though the dynasty may have been in decline and conditions forthe lineage were deteriorating (Guo 1994: 126-30; Lai 1997: 277-87), men still hada reasonable chance of acquiring a position.12
48
Linlan WANG, James LEE, Cameron CAMPBELL
12 This argument is confidently made for Beijing Zongshi, but not Jueluo, since the deteriorationof data for Beijing Jueluo in the nineteenth century can’t be ignored according to Figure 2.
Figure 7. Proportion Ever Having at Least One Position by Year of Birth
In Shengjing, trends for the Zongshi and Jueluo were erratic. As mentionedearlier, the number of lineage members in Shengjing reflected movement fromBeijing as well as natural growth. Changes from year to year and decade to decadein the number and type of lineage members relocated from Beijing affected thecomposition of the population in Shengjing. Moreover, the overall number of menwho survived at least to age 30 in Shengjing and were included in the calculationwas relatively small, typically a few per year, giving rise to some of the fluctuationsapparent in the figure. As Table 4 suggests, Zongshi in Shengjing were actually lesslikely to hold position than Jueluo. Accounting for the patterns and trends forShengjing in Figure 7 will require additional analysis as well as detailedexamination of official documents related to the awarding of positions, and will bea focus of our future work on stratification and inequality in the lineage.
Demographic outcomesIn this section we examine the implications of state-mandated inequalities betweenthe Zongshi and the Jueluo for trends and patterns in marriage and reproduction.We discussed trends and patterns in mortality earlier during the assessment of dataquality. Our motives for examining marriage and reproduction are two-fold. First,differences in marriage and reproduction between Beijing and Shengjing, andbetween Zongshi and Jueluo, illuminate the concrete implications of differences instatus. Second, peculiar or counterintuitive patterns by age or time may revealshortcomings in the data that need to be accounted for in future analysis.
49
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
NuptialityThe clearest differences between Zongshi and Jueluo were in the timing ofmarriage, not the overall chances of marrying. Figure 8 compares age patterns ofmale marriage. Because the original data do not record age at marriage, we have toinfer marriage patterns from the proportions married by age at death. This requiresan assumption that men who die at each at age were as likely to be married as themen who survived. In Figure 8, there were substantial differences in proportionsmarried between ages 15 and 25, with Shengjing Zongshi being most likely tomarry early, followed by Beijing Zongshi, and then the Jueluo. Nearly all men wholived long enough married, except for Beijing Jueluo, who were slightly more likelyto not marry. In both Beijing and Shengjing, Zongshi had higher proportionsmarried than Jueluo at almost every age.
Figure 8. Proportion Ever Married at Time of Death, Males Born 1644-1849
The advantage of the Zongshi persisted over time, even though there weresubstantial fluctuations in the overall chances of marrying, as well as differencesbetween Beijing and Shengjing. Figure 9 presents the proportion of men born ineach year who married by the time they died, restricting to men who survived to atleast age 30. According to the Figure 9, marriage was nearly universal for Zongshiand Jueluo well into the eighteenth century, with over 90 percent of men marrying.Afterward, marriage chances for Beijing Jueluo began to decline, and the gapbetween the Zongshi and the Jueluo in Beijing widened.
Linlan WANG, James LEE, Cameron CAMPBELL
50
Figure 9. Proportion of Males Ever Married by Year of Birth
There are two unexpected results in Figure 9 that we mention here, but willstill need to address in detail in future work. The first is the higher proportionsmarried in Shengjing. Our general assumption was that because Shengjing was thesecondary capital, the members who lived there had lower status, and possiblylower standards of living, than the ones who lived in Beijing. We saw earlier thatthey were less likely to hold a title or position. It may be that there was a recordingproblem in Shengjing such that men who didn’t marry were more likely to beomitted from the data, but it may also be that lineage members there had higherstatus relative to the local population than their counterparts in Beijing, and facedless competition for brides from other prominent families.
The second unexpected result is the extraordinarily low proportions ofBeijing Jueluo born in the first half of the nineteenth century who married. Again,this might have been real, or may have been a problem with the data. It may bethat as the lineage grew, and its status became diluted, Jueluo were disadvantagedenough that they encountered serious problems securing brides. Alternatively, itmay be that Jueluo were marrying, but not all their wives we being recorded,perhaps because they were childless. This is plausible, but would be hard toreconcile with the pattern of incomplete registration proposed above to account forthe higher proportions married in Shengjing. In future work we will examine thesephenomena in detail with more refined tests of data quality.
We also examined whether holding a position or title translated intoincreased marriage chances. Table 5 presents results from a logistic regression ofmarriage for males. The outcome is an indicator of whether a male had at least onewife. Right-hand side variables include controls for age at death and time period ofbirth in the form of categorical variables. Indicator variables distinguish between
51
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
Zongshi and Jueluo, and between Beijing and Shengjing. Interactions assessdifferences between Zongshi and Jueluo in the effects of living in Shengjing, havinga title, or having a position.
Results are largely as expected given the general tendency in historicalChina for higher status and socioeconomic advantage to increase the chances ofmale marriage. According to Table 5, Zongshi had odds of marrying about 70percent higher than Jueluo. Results for main effects indicate that for Jueluo,residence in Shengjing multiplied the odds of marriage by 3.67, holding a titlemultiplied the odds of marriage by 3.08, and having a position multiplied the oddsof marriage by 16.8. Coefficients on interaction terms indicate that the effects ofliving in Shengjing and holding a title were roughly similar for the Zongshi and theJueluo, but that the effects of having a position were much smaller for the Zongshi.
Differences by residence, status, and time period are also apparent in thechances of having more than one wife. Figure 10 presents trends over time in theproportion of married men who were recorded as having more than one wife.13 As
52
Linlan WANG, James LEE, Cameron CAMPBELL
Table 5. Logistic Regression of Having at Least One Wife, Imperial Lineage Males Born 1644-1849
Factors Mean Odds RatioZongshi 0.557 1.736 **Shengjing 0.114 3.671 **Title 0.046 3.070 **Position 0.098 16.800 **Age at Death
13 Not all concubines were recorded (Ju 1989), so the prevalence of polygyny may be underestimated.
noted earlier, at this point we cannot distinguish between widowers who remarriedand men who were polygynous. Men in Beijing were most likely to have more thanone wife if they were born at the beginning of the Qing, and their chances declinedsteadily afterward, reflecting the deteriorating living standards as the lineageexpanded and privilege was diluted. Zongshi in Beijing were more likely to havetwo or more wives than Jueluo, and this advantage persisted from the beginning ofthe Qing at least until the end of the nineteenth century. Zongshi in Shengjingwere initially much more likely than Jueluo to have more than one wife, but theiradvantage declined quickly, so that by the beginning of the nineteenth century,they were only slightly more likely to have two or more wives than Jueluo in bothBeijing and Shengjing.
Figure 10. Proportion With at Least Two Wives by Year of Birth, Married Males
Individual characteristics such as position and title also influenced thechances of being recorded with more than one wife. Table 6 presents the results ofa multinomial logistic regression that examines the determinants of the odds ofhaving 2 wives, 3 or more wives, or no wives, relative to the odds of having onewife. According to the results in Table 6 next page, position, title, and Zongshistatus have roughly the same influences on the chances of having more than onewife as they do on the chances of being married in Table 5. Effects were larger forthe odds of having three or more wives than they were for the odds of having twowives, suggesting that socioeconomic differentiation was more pronounced as thenumber of wives increased. Thus holding other variables constant, Zongshi had1.36 times higher odds of having two wives than Jueluo, but 2.70 times higherodds of having three or move wives. Similarly, men with titles had 1.31 times the
53
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
odds of having two wives of men without titles, but 2.05 times higher odds ofhaving three or more wives.
The results for men with missing age at death, meanwhile, suggest thepossibility that their actual ages at death were bimodal. On the one hand, menwhose age at death was missing were twice as likely as men who died in theirtwenties to not have any wife at all recorded, but much less likely than men whodied in their teens to not have a wife. This suggests that the men without deathdates included one cluster whose ages were concentrated in the late teens and earlytwenties. However, they were as likely to have two wives, or three or more wives,as men who died when they were in their forties. Accordingly, men who had noage at death recorded may have been a mixture of one group of who died atrelatively young ages, and another group who died much later in life.
The finding that other things being equal, Zongshi were not only morelikely to marry than Jueluo, but were also more likely to have more than one wife,is consistent with expectations based on official regulations and practices related tomarriage. Table 1 demonstrated the substantial gap of between Zongshi and Jueluoin the wedding subsidy that they received. Scholars have also discussed thepolitical purpose of arranging the marriage of Zongshi with Mongolian and Hannoble families (Du 1990). In light of the general similarity of mortality levels of theZongshi and Jueluo, the higher proportions of Zongshi who married or who hadmore than one wife, may have made a major contribution to the divergence in thegrowth rates of the Zongshi and Jueluo that according to Figure 1 began in themiddle of the eighteenth century.
54
Linlan WANG, James LEE, Cameron CAMPBELL
Table 6. Multinomial Logistic Regression of Number of Wives, Imperial Lineage Males Born 1644-1849a
No wife 2 wives 3 or more wivesOdds Ratio Odds Ratio Odds Ratio
Zongshi 0.642 ** 1.360 ** 2.707 **Shengjing 0.278 ** 1.075 1.182 Title 0.357 ** 1.312 * 2.051 **Position 0.074 ** 1.780 ** 3.910 **Age at Death
ReproductionPrevious work has shown that in historical China, socioeconomic and otherprivilege not only conditioned access to marriage, but also the reproductivebehavior of married couples (Wang, Lee and Campbell 1995; Lee and Campbell1997; Lee and Wang 1999). To assess trends and patterns in the reproductivebehavior of the Zongshi and Jueluo, Figure 11 presents the mean number of sonsby father’s year of birth. According to Figure 11, men born in the earliest decadesof the Qing had by far the most sons. At this time, the lineage was still small, andthe members were relatively privileged. This high fertility accounts for the rapidgrowth in the numbers of lineage members apparent in Figure 1. Initially, theBeijing Zongshi had higher fertility than the Jueluo, but they converged later. Thegeneral similarity in the fertility of married Zongshi and Jueluo after 1720 suggeststhat within the lineage, access to marriage was the key source of difference in thereproductive success of Zongshi and Jueluo, and may have played a major role inthe divergence of the growth rates of the Zongshi and Jueluo in the middle of theeighteenth century apparent in Figure 1.
The similarity in the mean numbers of sons for married Zongshi and Jueluoborn in first half of the nineteenth century would seem to rule out the possibilitysuggested earlier that the wives of Jueluo who had no sons were being omitted,since such a pattern of omission should have led married Jueluo to have unusuallyhigh fertility.
55
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
N 45009a The comparison category for the outcome variable is having one wife. *.p<.05, **.p<.01
Figure 11. Mean Number of Sons by Year of Birth, Restricted to Married Males
We also estimated a Poisson regression to examine the determinants of thetotal number of sons for married men. Table 7 presents the results. According tothese results, Zongshi on average had slightly fewer sons than Jueluo. Men whoresided in Shengjing or held title or position had slightly more sons than men whodid not. According to the results for the interaction terms, Zongshi had even moreof a fertility disadvantage in Shengjing than they did in Beijing. Having a titlebenefitted Zongshi more than it did Jueluo. Once more, the men who were missingan age at death had roughly as many children as men who died in their forties, againsuggesting that on average, they actually tended to be men who died at later ages.
Linlan WANG, James LEE, Cameron CAMPBELL
Table 7. Poisson Regression of the Number of Sons,Imperial Genealogy Married Males Born 1644-1849
Mean Incidence Rate RatioZongshi 0.597 0.956 **Shengjing 0.132 1.353 **Title 0.083 1.154 **Position 0.185 1.225 **Age Group
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
ConclusionThe expanded dataset based on the 1935 edition of the Aixinjueluo Genealogyyields a more complete and detailed picture of the Qing imperial lineage. Whereasanalysis of the Zongshi from the Jade Records was by definition limited to an elitesubset of the lineage, the newly created dataset allows for comparison between thetwo branches by also providing data on Jueluo. This preliminary analysis here hasbegun to illuminate the potential of the expanded dataset for future analysis thatfocuses on comparison between and within the branches of the imperial lineage.
Differences in the institutional context of the Zongshi and Jueluo led todifferences in the quality of their data, their socioeconomic outcomes, and thedemographic behavior. Overall, Jueluo have less complete data than Zongshi. Amuch larger share of the Jueluo lived in Shengjing, which may account for at leastpart of the difference in their data quality from the Zongshi. Even among theZongshi, those who resided in Shengjing were less likely to be recorded completely.Nevertheless, even when comparing only the Zongshi and Jueluo in Beijing, salientdifferences in recording quality are still apparent. Also, by examining the infantand child mortality, we find that unlike what usually occurred to Chinesegenealogies, the omission of report were not likely to happen to people who diedin infancy or childhood when the overall data are not too bad.
The pattern of differences between Jueluo and Zongshi in socioeconomicoutcomes and demographic behavior is more complex. Zongshi were notuniformly advantaged. For ordinary Jueluo and Zongshi, who account for amajority of the both populations, there is no substantial difference in the majordemographic outcomes, getting married and bearing a son. Even though Zongshiwere more likely to marry, at this stage it appears that other things being equal,Jueluo had roughly as many sons. Even though Zongshi were advantaged inobtaining titles, Jueluo managed to be equal as Zongshi in the attainment ofposition.
Finally, some of our preliminary analyses point to an interestingphenomenon where the recording, socioeconomic attainment and demographicbehavior of Zongshi and Jueluo appeared to vary from reign to reign with marked
changes even in the completeness of demographic recording at the beginning ofthe Qianlong, Jiaqing, and Daoguang periods. In that regard, this study, like Chen(2009), provides a better understanding of how categories work in China to createinequalities that were not only durable - at least by dynasty - but also mutable fromemperor to emperor. Categorical differences between Zongshi and Jueluo, and formany other institutions and populations in China, were and are as much afunction of transitory policy as durable institutions, of people as well asregulations.
This is but a first step in the analysis of the newly expanded dataset, and welook forward to more detailed study of the socioeconomic and demographicimplications of the many differences in institutional context within and betweenthe Zongshi and Jueluo, and between them and regular bannermen and non-bannermen. We will also extend on the analysis here by examining in detail theirtitle or position holding in the political context of and nobility construction in theempire. Eventually, linkage of records of sons to fathers and reconstruction ofreproductive histories and pedigrees for individuals will allow for studies ofdifferentiation in socioeconomic attainment and demographic behavior among thedescent lines within the Zongshi and Jueluo.
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
REFERENCES
Primary SourcesAixin jueluo zongpu (Genealogy of the Aixinjueluo). ed. Jin Songqiao
et al. re-ed. Aixinjueluo Chang Lin et al. Beijing:Xueyuan chubanshe . 1998.
Qinding zongrenfu zeli (Imperially CommissionedRegulations of the Imperial Clan Court: Two Editions). re-ed. The PalaceMuseum Haikou: Hainan chubanshe . 2000.
Qinding baqi tongzhi (Imperially Commissioned General History ofthe Eight Banners). ed. E ertai et al., col. Li Xun, Zhao Degui, ZhouYufang et al. Changchun: Jilin wenshi chubanshe
. 2002. (This edition is based on the block-printed edition by theHall of Martial Valour during Emperor Jiaqing’s reign collated against theedition in the Four Treasuries by the Hall of Literary Profundity.)
Da qing shilu (The Veritable Records of the Qing Dynasty), Renzong,Xuanzong. Beijing: Zhonghua Shuju . 1986.
Secondary Sources
Campbell, Cameron and James Lee. 2000. “Price fluctuations, family structure andmortality in two rural Chinese populations: Household responses toeconomic stress in eighteenth and nineteenth century Liaoning.” InPopulation and the economy: From hunger to modern economic growth. eds.Bengtsson, Tommy and Osamu Saito, 371-420. Oxford: Oxford UniversityPress.
___________. 2004. “Mortality and Household in Seven Liaodong Populations,1749-1909” In Life Under Pressure: Mortality and Living Standards in Europe andAsia, 1700-1900. eds. Bengtsson, Tommy, Cameron Campbell, James Lee, et al.Cambridge: MIT Press.
___________. 2006. “State views and local views of population: Linking andcomparing genealogies and household registers in Liaoning, 1749-1909.”History and Computing 14(1+2): 9-29.
Chaffee, John W. 1999. Branches of Heaven: A History of the Imperial Clan of SungChina. Cambridge: Harvard East Asian Monographs.
Chen, Shuang. 2009. Where Urban Migrants Met Rural Settlers: State Categories,Social Boundaries, and Wealth Stratification in Northeast China, 1815-1913.”Ph.D dissertation, University of Michigan.
Ding, Yizhuang . 1998. Qingdai manzu de qie yu qie zhi tanxi”(A Study of Concubines and Concubine System in Qing
Dynasty). Jindai zhongguo funvshi yanjiu (Journal of theHistory of Modern Chinese Women) 6: 75-108.
__________. 1999. Manzu de funv shenghuo yu hunyin zhidu yanjiu(A Study of Manchurian Women s Life and Marriage).
Beijing: Beijing daxue chubanshe .
Du, Jiaji . . Qingdai de zongshi fengjue jiqi dengji chabie de teshuxing”(The Awarding and Hierarchy of
Titles for the Zongshi in Qing Dynasty). Manzu yanjiu (Man Ethnicity Study)1: 41-45.
_________. 1998. Qing huangzu yu guozheng guanxi yanjiu(A Study of the Qing Imperial Lineage and Qing Politics). Taipei: Wunantushu chuban youxiangongsi .
__________. 2003. Qingchao man meng lianyin yanjiu (A Study ofthe Political Alliance by Marriage between the Manchurians and theMongols). Beijing: Renmin chubanshe .
Guo, Songyi . 1994. Qing Zongshi de dengji jiegou ji jingji diwei”(Class Structures and Economic Situation of Qing Zongshi).
In Qingdai huangzu renkou xingwei yu shehui huanjing(The Qing Imperial Lineage: Social Structure and Population
Behavior). eds. Li and Guo. Beijing: Beijing daxue chubanshe . Hu Xiangyu . 2010. Sex, Status and Normalization of the Law: Illicit Sex Cases
Involving Imperial Clansmen in the Late Qing. Unpublished Manuscript.Jing Junjian . 1993. Qingdai shehui de jianmin dengji
(Social Outcasts in Qing Soceity). Hangzhou: Hangzhou renmin chubanshe.
Ju Deyuan . 1989. Qing huangzu duoqizhi he nvzi laiyuan”(Polygyny and the Origins of Women among the Qing Imperial
Lineage). Zijincheng (The Forbidden City) 3: 42-43. Lai Huimin . 1994. Qingdai huangzu de fengjue yu renguan yanjiu
(Nobility and Office among the Qing Imperial Lineage)In Qingdai huangzu renkou xingwei yu shehui huanjing
(The Qing Imperial Lineage: Social Structure and PopulationBehavior), eds. Li and Guo. Beijing: Beijing daxue chubanshe .
___________. 1997. Tian huang gui zhou: Qing huangzu de jieceng jiegou yu jingjishenghuo (The Heavenly Branch:Class Structures and Economic Life of Qing Imperial Clans). Taipei:Zhongyang yanjiuyuan jindaishi suo .
Lee, James, Cameron Campbell and Wang Feng. 1993. The last emperors: anintroduction to the demography of the Qing (1644-1911) imperial lineage. InNew and Old Methods in Historical Demography. eds. Roger Schofield andDavid Reher, 361-82. Oxford: Oxford University Press.
Lee, James and Guo Songyi eds. 1994. Qingdai huangzu renkouxingwei yu shehui huanjing (The Qing ImperialLineage: Social Structure and Population Behavior). Beijing: Beijing daxuechubanshe .
Lee, James and Wang Feng. 1999. One Quarter of Humanity: Malthusian Mythologiesand Chinese Reality. Cambridge: Harvard University Press.
________________. 2000. Male nuptiality among the Qing nobility: polygyny orserial monogamy. In Fertility and the Male Life Cycle, eds. Caroline Bledsoe,Susana Lerner, and Jane Guyer, 188-206. Oxford: Oxford University Press.
60
Linlan WANG, James LEE, Cameron CAMPBELL
Lee, James, Wang Feng, and Cameron Campbell. 1994. Infant and child mortalityamong the Qing nobility: implications for two types of positive check.Population Studies 48 (3): 1-17.
Lee, James, Wang Feng, and Danching Ruan. 2001. Nuptiality among the Qingnobility: 1600-1900. In Asian Historical Demography, eds. Ts ui-jung Liu,James Lee, David Reher, Osamu Saito, and Wang Feng, 353-73. Oxford:Oxford University Press.
Lei, Bingyan . 2003. Qingdai qiren fengjue luelun (AStudy of the Title System for Bannermen in Qing Dynasty). Beijing shehuikexue (Beijing Social Science) 4: 143-51.
Rawski, Evelyn S. 1998. The Last Emperors: A Social History of Qing ImperialInstitutions. Berkeley, Los Angeles and London: University of California Press.
Wang Feng, James Lee, and Cameron Campbell. 1995. Marital fertility controlamong the Qing nobility: implications for two types of preventive check.Population Studies 49 (3): 383-400.
Wang Feng and James Lee. 1998. Adoption among the Qing nobility and itsimplications for Chinese demographic behavior. History of the Family 3 (3):411-28.
Zheng, Chuanshui . 1985. Lun qingchao de qi xiang zhengce jiqi yingxiang”(A Study of the Salary Policy for Bannermen and
Its Influence in Qing Dynasty). Liaoning daxue xuebao (Journalof Liaoning University) 2: 76-80.
61
Institutions and Inequality:
Comparing the Zongshi and the Jueluo in the Qing Imperial Lineage
![Vol.19 No.1 APR. 20191].pdf · 1 Journal of East Asian Studies Vol.19 No.1 APR. 2019 Sungkyun ISSN 1598-2661 (Print) ISSN 2586-0380 (Online) Academy of East Asian Studies Sungkyunkwan](https://static.documents.pub/doc/80x56/600bed9b0e7f4417d263fda7/vol19-no1-apr-2019-1pdf-1-journal-of-east-asian-studies-vol19-no1-apr.jpg)
![Vol.19 No.1 APR. 2019sjeas.skku.edu/upload/201904/SJEAS_Yuniu_LI_0419_hp[1].pdf · 1 Journal of East Asian Studies Vol.19 No.1 APR. 2019 Sungkyun ISSN 1598-2661 (Print) ISSN 2586-0380](https://static.documents.pub/doc/80x56/5f0dbcaa7e708231d43bd57c/vol19-no1-apr-1pdf-1-journal-of-east-asian-studies-vol19-no1-apr-2019.jpg)
