Pursuing rights and getting justice on China's ethnic frontier, 1949-1966

A Legal Controversy in Rural China, 1956

It seemed like just another ordinary fall day in a village of the Yi minority community in Yunnan Province, on China's southwestern border. As the sun rose over the glistening fields, families woke up and began their chores. Sometime later, their tasks completed, several young women, laden with recently harvested produce, met in the village square to go to the market town, as was their custom and duty. This was not an easy trip: Yunnan was a poor, mountainous province with few paved roads, railways, or motorized transport.

After walking for several hours, they arrived at their destination. The market town, located in a valley, was bustling with peasants who had streamed down from mountain villages. The sounds and smells of donkeys, goats, squealing pigs, and terrified chickens filled the air as the women took up their positions on the roadside to hawk their produce. By early afternoon they were finished. Their responsibility fulfilled, and with some free time on their hands before their return trip, the women decided to wander a bit. Passing through the town's main thoroughfare, they saw the gated building of the county's political administration. Nearby was the court. Noticing a crowd gathered inside, the women decided to see what was going on. This activity, a cultural practice known as kan renao, or "getting in on the fun," was not entirely novel, as rural folk in the area (as well as in others) had been known to "gather around" any event that broke the monotony of life and promised free entertainment for a while (Perry 1993). On this particular occasion, the Yi women were witnessing several divorce cases. Taking in this scene, some of the Yi women who had come to town to market their produce decided that they, too, wanted a divorce. They approached the clerk and the judge and demanded a divorce on the basis of the People's Republic of China's 1950 "Marriage Law," which allowed divorces if marriages were "arranged" or "coerced" by parents or other members of the community. The Yi women testified that indeed their marriages had been coerced, and this gave them the right to divorce

The court granted their requests, despite the fact that their husbands were not present in court. Their divorce papers in one hand and the money they received from selling their produce in the other, they made their way back to their village. Upon arrival, they presented their husbands with the divorce papers. Aghast that their wives had divorced them, the aggrieved husbands marched off to complain to their local officials. The local officials were also incensed, and worried that if village women could divorce so easily, soon the village would be left without any young marriageable women. The courts, they complained, decided cases in a "rash and careless" manner. This problem was compounded, they argued, because the Yi minority is particularly litigious and prone to divorce. Courts, however, were not convinced that this was correct, and proposed a political explanation instead: Yi women were divorcing because their "consciousness had been raised." A court also noticed that "the more we intervene, the more divorces there are" (CXA 16-27-A1:2).

Statistics from the early 1950s bear out this observation. Court-granted divorces were a rarity before the 1949 Communist Revolution in most areas of rural China (Huang 1996). In 1950, the Chuxiong Prefectural Court handled 510 divorce cases, but this number shot up to 6,600 by 1953, a twelvefold rise.' CalcuIMAGE FORMULA 13

fated on a household basis, as many as 5% to 8% percent of families experienced a divorce per annum between 1950 and 1953.2 Most divorce cases were settled in women's favor (women represented well over 80% of plaintiffs). In some areas of Yunnan, 100% of cases were adjudicated in the woman's favor (CXA 44Al:38). Minorities in the province-not only the Yi-divorced at the same rate as the Han Chinese (CXA 16-75-BI:89). Moreover, evidence shows that the rate of rural divorce in Yunnan Province continued well into the 1950s and early 1960s. In one county, there were as many divorces in the early 1960s as there had been in the early 1950s (Chuxiong yizu zizhi zhou zhi 1993, vol. 2:255). This finding is particularly striking considering that after 1953, when the government mounted a vast (and last) national campaign to implement its new Marriage Law, the state ceased the law's proactive implementation. According to some China scholars, the absence of active state intervention resulted in the near absence of divorce in rural China after 1953. The state, they have argued, left rural women in a legal lurch. "Modern" urban women, however, were able to take advantage of the law and obtain divorces, thanks to their education, relative freedom from lineage pressures, and proximity to legal institutions (Johnson 1983).

This case and the statistics attesting to the increasing divorce rate in Yunnan (in addition to its relative ease) were unearthed in a rural Chinese archive while I was conducing field research on the impact of the People's Republic of China's "Marriage Law" (hunyinfa) of 1950. Taken together, these data raise important questions concerning the role of law in social change, community, gender, ethnicity, and legal culture, questions that are relevant both to the study of law in China and to larger debates in the law and society literature. Just what led these women to court, and what factors led to a favorable outcome (from their perspective, not their husbands')? Were township officials on the mark in pointing to a certain cultural proclivity to sue for divorce among the Yi minority, or was it the case that rural Chinese, or even minorities in China, irrespective of their particular ethnicity or area of residence, were somehow prone to divorce? Since most of these women were young, was generation an important variable? Moreover, what significance should we assign to the fact that IMAGE FORMULA 16

these Yi women went to market and court in groups? Was there an underlying legal culture that led people to court? Finally, how should we understand the role of the state in legal change when one layer of its apparatus-mainly lower-level officials-opposed using state law to settle a family dispute, while a different onecourts-appeared to use state law rather haphazardly and in a way that violated the interests of poor male peasants, the Chinese regime's primary "constituency"? Were courts correct to attribute Yi divorces to their emancipation by the regime? In this article I aim to figure out this puzzle. In doing so, I will rely upon a treasure trove of archival documents chronicling how the Chinese Marriage Law of 1950 was implemented and interpreted in rural and in urban China.3 Does this evidence point to the importance of the state, or to ethnic, gender, cultural, or generational variables, or some combination of these?

Although both the puzzle and the empirical evidence emerge from rural China, the answers to the questions posed above should be of considerable interest to scholars working on comparative law and law and society. In China, after all, we have a case in which a revolutionary state introduces a law in the hope that women would "emancipate" themselves from the shackles of the "feudal marriage system." The law's ideals were highly modernist, if not blatantly "bourgeois." The law's framers (about whom we know very little) argued (or assumed) that relationships formed out of individual free will after an extended period of courtship were happier, more productive, and would contribute to a better, more productive society and state than the unhappy, violence-ridden marriages that were arranged by parents, family elders, or community members. Yet, the society into which this law was introduced was "traditional," at least in the sense that the overwhelming majority of Yunnan Province's residents were peasants, and peasants are typically considered to be traditional (particularly when compared to "modern" urbanites).

This story, then, might easily be framed around the categories of modernization theory: To what extent can "modern" laws change the way rural cua, "traditional," people interact in family matters? Previous studies on family law in traditional societies published in the Law & Society Review provide some clues. For instance, June Starr's (1989) excellent study of the impact of secular law in Turkey noted that Ataturk's Swiss-inspired Civil Code, introduced in 1926, affected rural areas only 30 years later, but its impact on urban society was immediate. Similarly, Gregory IMAGE FORMULA 18

Massell's (1968) analysis of the Soviet Marriage Law in Muslim regions in the former Soviet Union (also published in the LSR), suggested an inherent tension between state and periphery, modern state law and local customary law, and between rural women struggling for greater freedoms and young men intent on preserving the perks of patriarchy. In fact, every available account on the implementation of family laws by revolutionary states (France, the Soviet Union) clearly shows their ineffectiveness in rural areas and relative effectiveness in cities, particularly among the educated elite (Farnsworth 1985:253-4; Madsen 1977:311; Philips 1980:93-96) .4

A brief glance at ancillary literature on mediation and legal access lends support to this view. According to William Felstiner (1974), people who know each other-such as villagers living in small, relatively isolated rural communities such as the Yi in Yunnan-tend to prefer mediation to adjudication. In cities, however, where more anonymity prevails, the use of courts increases. For instance, David Engel's (1978) research on a Thai provincial court has found that residents of rural areas rarely use the court system to pursue private claims. In addition, claims of "Asian" or "Eastern" non-litigiousness abound in the literature, often bolstered by Asian elites who use this supposedly timid legal culture to justify limits on popular political participation and authoritarian rule (Diamant 2000b; Upham 1998). Furthermore, if Charles Epp's (1998) comparative study of the conditions for a "rights revolution" has any bearing on the China case, we would not expect Chinese peasants to successfully mobilize and act upon their rights. Epp found that in India many well-intentioned laws have little impact on the ground, largely owing to people's inability to mobilize financial resources rather than to def ciencies in the laws or the apathy of high courts. In Yunnan, people were generally quite poor when they were granted new rights, but Chinese women, despite their poverty, managed to construct what Epp aptly calls a "support structure for legal mobilization" (1998:3) and use the state legal system to divorce their husbands.5 IMAGE FORMULA 21

To be sure, this story is not only one involving "modern" law and "traditional" peasants. These "traditional" people did not just belong to the "generic" category of "peasant": They were women, and they were members of an ethnic minority group. How women generally, and minority women specifically, interact with law and legal institutions are questions about which there is very little consensus in the comparative law literature. Consensus does exist in the China field, however. Those who have looked at the impact of the Marriage Law in China have emphasized, almost without exception, the law's toothlessness in rural areas and generally positive reception in more modern urban areas. Most rural Chinese women were too illiterate to understand the law, too inexperienced in using law, or too deeply embedded in tightknit, patriarchal communities to even get to court. Once in court, they were confronted by a patriarchal state hostile to women. As a prize-winning book on rural China noted of the period when our Yi women went to court: "There was no law" (Friedman et al. 1991:153). Others have claimed that rural divorce was "almost non-existent" as a result of the state's indifference to the plight of women and lack of sustained enforcement of the Marriage Law (Honig & Hershatter 1988:206).

Such a one-sided conclusion, to be sure, does not square with all of the literature on women's and other marginal groups' relationship to law. While radical legal scholars such as MacKinnon (1993:109-10) argue that "to most women the law is a foreign country with an unintelligible tongue, alien mores, obscure but rigid dogmas, barbaric and draconian rituals, and consequences as scary as they are incomprehensible," others note that, in some circumstances, law can be used by marginalized citizens in their battles for greater political space and voice. As Lazarus-Black and Hirsch suggest in their collection Contested States, law can allow marginal actors to "appropriate the terms, constructs, and procedures of law in formulating opposition" (1994:10; also see Comaroff 1994). In a related vein, Michael McCann has noted (1994, 1998) that law's impact often can be indirect and even unintentional. That is, even if judicial decisions do not favor plaintiffs, the social mobilization and learning processes that accompany filing a suit can, in the long term, significantly impact both the participants and the legal process. Given this controversy, the case of the Marriage Law in Yunnan presents an excellent opportunity to assess the role of law in rural women's lives. If many officials opposed the Marriage Law owing to patriarchal views, exactly how did it have any impact? Was the social change taking place after 1949 the direct consequence of the law, or rather the byproduct of the political and social mobilization that accompanied it?

In this article I argue that much of the social change generated by the Marriage Law among minorities in southwest China can largely be attributed to the way the historically unusual revolutionary dynamics of the Chinese Revolution dramatically altered previously existing opportunity and support structures for divorce throughout rural China. This rural China-wide dynamic, I suggest, was more important than any particular regional political configuration or ethnic-based cultural tendency among the Yi, although these certainly played a role in shaping how cases were brought to court and the causes for divorce. While the Marriage Law of the People's Republic of China was similar in both intent and goals to the laws of other modernizing or revolutionary states, such as France, Turkey, Thailand, and the former Soviet Union, and while rural folk in other societies (including Egypt, Nepal, Japan, and Russia) have been willing to engage the state apparatus (whether native or controlled by colonizers) to pursue private claims (Bennett 1983; Hill 1979; Farnsworth 1986; Engelstein 1992; Czap 1967; Galanter 1989; Haley 1978), only in China was rural divorce a significant outcome of legal implementation.

Throughout the 1950s and early 1960s (until 1966, when the Cultural Revolution closed the courts), divorce was not an infrequent occurrence throughout rural China. Whether among peasants in the rural areas surrounding Shanghai or in one of the most culturally "conservative" provinces (according to conventional wisdom in China), such as Shandong Province in the northeast, divorce rates increased dramatically after the 1949 Revolution, persisted at a lower level until the late 1950s, and then spiked to their previously high levels in the early 1960s (Songjiang County Archives [SJA] 8-1-32:32; Zhongguo renkou-Shandong fence 1990). In one village in north China, for instance, 25% of all married couples under the age of 25 were either divorced or in the process of divorcing within a 16-month period; in the township (an administrative unit amalgamating several villages), 11 out of 329 households experienced divorce between April 1953 and January 1955 (3%) (QPA 48-2-59:78). At the same time, however, young urban ethnic minorities who expressed interest in divorce were often stymied by their communities. Furthermore, urban divorce rates-particularly among those whose identity derived directly from their urban status (white-collar professionals, business people, artists, full-time factory workers)-were lower than those in rural areas, particularly in the early and late 1950s.6 This happened, I argue, because only China had a revolution that moved from the countryside to IMAGE FORMULA 26

the city and which later staffed the state with people who had firsthand experience with rural life and revolution. In contrast, the Turkish, French, and Russian revolutions were centered in cities, only moving to rural areas years later. Their revolutionaries were usually urbanites and gave priority to urban social change. In China, however, peasants were assigned a leading role in national political and economic development. This is seen most concretely in the active and often violent role they took in land reform and later during the mobilization of the Great Leap Forward (1958-1960). Furthermore, only in China did the state, largely because of its rural experience and success, glorify the peasantry and discriminate against the same urban elites the literature identifies as the people most inclined to use modern state law.

Revolutionary politics in rural China thus had direct consequences for rural women in Yunnan. In what follows I argue that the substantial increase in divorce in Yunnan was, in many ways, the by-product of rural revolutionary politics in the area and not the direct consequence of the Marriage Law itself. The state, I suggest, did not intend to radicalize ethnic minority women; instead, this happened because other revolutionary policies, particularly land reform, led to new demands on the state. The central state did not want minority women to bypass village mediation and proceed to court. However, local officials who witnessed summary executions of various "enemies of the people" and harsh disciplinary punishments against erring cadres preferred to send marriage disputes to court-where women received a more sympathetic hearing-than risk committing what might be misconstrued as a "political" error.7 Furthermore, the state did not intend judges to award divorces at a high rate; this happened because the courts were operating in a dangerous, uncertain, and unprecedented political context, and they preferred to abide by what they understood to be the "general spirit" of the law rather than follow its more concrete articles or officially published guidebooks. It was also largely coincidental that the state superimposed new laws upon both a rural legal culture that was already quite litigious and militarized and upon certain community cultural norms and structures (particularly dense social networks among women) that ended up providing the social support necessary for ethnic minorities in Yunnan to divorce. That these norms and structures existed prior to the Revolution but did not result in divorce suggests that the shift in the opportunity IMAGE FORMULA 29

structure resulting from the state's rural policies was absolutely crucial. Local political, economic, and social features in Yunnan, which I discuss shortly, undoubtedly shaped the manner in which people took advantage of the new opportunities presented to them, but given the extent to which rural society was shaken up by massive political and social change at the time, perhaps it was somewhat inevitable, or at least not very surprising, that changes would also take place in the family.

This article will begin by first taking a closer look at the Marriage Law itself and the features of the area upon which it was imposed after the success of the Communist Revolution. I then focus on four dimensions where state politics intersected with rural southwestern society in ways that help explain the anomalous outcome of the Marriage Law in rural China: (1) the impact of the state (in law and other policies) upon minorities in China, in Yunnan and elsewhere; (2) the role of the state in empowering Chinese youth; (3) the impact of the Revolution on community relations and mediation; and (4) rural culture and law. As the following discussion will show, while the impact of the state law and policies was critical, the state did not create de novo the legal culture, local customs, and political circumstances that made the law a vehicle for change in rural Yunnan Province, nor were its effects on society the result of deliberate, conscious decisions made by state leaders.

The Marriage Law and Southwest China

Before I examine some of the explanations of the law's impact on Yunnan society, it is worthwhile to briefly discuss its original goals and intentions. Notwithstanding its modest-sounding title, the Marriage Law went well beyond marriage-related matters. It not only banned arranged marriages, bigamy, underage and minor marriages,8 and allowed easier access to divorce, it also advanced a decidedly modernist vision of "proper" family structure. Under the Marriage Law and its accompanying propaganda, families in revolutionary China should not be based on community involvement in the selection of marriage partners, nor upon "crass" concerns such as social status and wealth. Instead, families should be based upon individualism, monogamy, "love," and "free" choice in the selection of marriage partners. Although the People's Republic of China (PRC) has made much of its anticolonial and anti-Western revolutionary history, voices arguing that a modern state could in fact coexist with the Chinese traditions of concubinage and arranged marriages were virIMAGE FORMULA 32

tually nonexistent. Arranged marriages were simply assumed to violate individual will; community oversight over marriage could not possibly lead to a satisfying and happy marriage (and, by implication, unhappy marriages would result in an unproductive citizenry).

In addition to marriage, relations between generations also came under the law's purview: Children, parents, and grandparents were called upon to be "harmonious" and to care for one another in time of need. The law also broke new, crucial ground in the role of the state vis-a-vis the family. Although both Imperial and Republican Chinese states had extensive family laws dealing with marriage, sexuality, divorce, property, and inheritance (Huang 1996; Sommer 2000; Bernhardt 1999), only the Communists required prospective couples to register their marriage or divorce at a state institution.10 The Communists also gave previously informal legal procedures (such as mediation) a formal role in the state judicial structure. According to the 1950 law, couples were allowed to divorce only after "mediation was unsuccessful" (Johnson 1983).

In its vision, its rearticulation of state-family relations and its scope of enforcement, the Marriage Law of 1950 dramatically departed from past patterns. Whereas some Imperial regimes promulgated family laws, and courts heard a wide range of cases, ordinary people still had to initiate legal proceedings. In contrast, the Chinese Communist Party (CCP) demanded mass official and popular participation on a nationwide scale. This requirement placed many officials in a difficult predicament, since Chinese legal tradition offered them very few specific guidelines for how to deal with problems that until recently were defined as personal, family, or community affairs (Li 1951).

As noted earlier, most of the scholarship on the impact of the Marriage Law has reflected the law's own assumptions about the coercive power of "feudalism" and its main victims. Traditionalist social forces prevailed in rural areas, but particularly those in areas such as Yunnan, which was far removed from the cosmopolitanism radiating from major urban centers. Rural villages there were too insular, collectivist, poor, and patriarchal to become the soil where a new family ideology and law could take root (Johnson 1983:117). State policy made family change even less likely in ethnic minority communities. Officially, in the early 1950s, minority communities were exempted from implementing socially transformative policies, such as land reform and the Marriage IMAGE FORMULA 35

Law (Haberer 1987:9). Han Chinese landlords were subject to execution and political persecution, but state policy banned the ousting of local minority elite. If one precondition of family change is said to be the weakening of traditional social structure, either through modernization or political revolution, then it was precisely in remote minority areas such as central Yunnan Province where we would not expect much change to take place. But if this is the case, how might we explain evidence of peasant women taking advantage of the law, not only during the Marriage Law campaigns in the early 1950s but later on as well, when the state's legal attack on the "feudal family" was supposedly in full retreat in light of massive resistance by poor male peasants?

Indeed, if we look more closely at the political history, economy, and society of southwest China in light of the prevailing theories of law-induced change we can find ample reasons the law would not take hold in rural minority communities.?I Politically, in Yunnan and other provinces in the Southwest, separatism, ethnic conflict, and militarization, often led by local elites, was widespread. From the Han dynasty (206 B.C.-A.D. 220) to the Qing (1644-1911), many areas in southwestern China were indirectly governed by the enfeoffing local tribal chieftains as either civilian administrators (known as tu guan) or military chief tains (tu si), or through military colonization; one estimate suggests that roughly one-third of Yunnan Province, over threequarters of Sichuan, and at least one-half of Guizhou were "under native chieftainship control" in the 16th and 17th centuries (Lee 1982:715; Chang 1956:27-28).

Both of these systems of administration proved difficult to enforce over the long term: Tu si in mountainous regions often established quasi-independent fiefdoms that often served as a rallying point against the central state (Chuxiong zhou zhi 1993, vol. 1:14-15; Herman 1997:50-52; Barnett 1985:283; Goodman 1986:35-36.) Military colonization ultimately contributed to tension between central state authorities and local leaders by forcing local communities out to more mountainous and poorer regions, which left Han garrison leaders and troops the more fertile valley basins (forcing the Yi women to come down from the mountains to the market town). During the Qing dynasty, tensions between Han military administrators and tu si resulted in not a few instances of armed conflict (Ford 1997:10; McCord 1990:165). During the late Qing, the southwest, like other areas of China in this period of dynastic decline, was racked with local rebellions, leaving many areas virtually ungovernable (Chuxiong zhou zhi 1993, vol. 1:11, 145). As in many areas of China during the late Imperial period, local elites relied primarily on military forces to mainIMAGE FORMULA 37

tain a modicum of control over an increasingly restive, armed population (Kuhn 1980; McCarthey 1879:499). In the Republican period (1911-1949), bandits, often led by ethnic minority elites, proliferated, finding refuge and bases in the mountains from which they would plunder both Han and minority communities (Hosie 1897:66; Clarke n.d.:116). By the end of World War II and the outbreak of the Chinese Civil War (1945-49), political control in much of the southwest had almost completely devolved into the hands of local Han and ethnic minority elites. Although Chiang Kai-shek managed to establish some authority in the urban basins of Chongqing and Kunming after he retreated to Sichuan during the war, he could not control rural areas and was forced to compete with secret societies and local warlords (Solinger 1977:61-63).

Chiang's retreat into the southwest during the war also had implications for the Communists. Rural areas in Yunnan, Sichuan, and Guizhou were the last bastions Nationalist of resistance against the Communists. In contrast to north and central China, where the Communists had established wartime base areas during the 1930s and 1940s, in the southwest the CCP did not have any significant legal, military, or administrative presence. Since virtually all CCP officials were outsiders, hailing predominandy from the older revolutionary base areas in northwest China (An xian zhi 1990:460, 506), local units of the CCP and People's Liberation Army (PLA) had little legitimacy with the populace (Skinner 1951:61-63).

From the standpoint of political history it would be hard to find a less hospitable environment for implementing a new family law than southwest China. This situation is also true in terms of minority groups' economic resources and ability to access state legal institutions. As Epp (1998) has noted, resources have often proved crucial both in gaining access to and receiving favorable outcomes from the legal system. In Yunnan, however, the 10% of land that was arable (Richardson 1940:123) was not equally distributed among ethnic minority groups. Ethnic minorities (with the exception of Muslim, or Hui) lived in mountainous areas, where good land was particularly scarce and transportation cumbersome. In contrast, Han Chinese lived in the more fertile valley basins, where most industries and government institutions were also located (Chuxiong zhou zhi 1993, vol. 1:152). Inequitable land distribution, poverty, and rural unemployment during the slack season encouraged peasants to engage in high-risk but potentially lucrative nonfarming activities such as growing opium. During the late 19th and early 20th century, opium was often a staple of southwestern regional and inter-regional trade (McCarthey 1879:498; McCord 1990:165; Hosie 1897:37; Fei & Chang 1945:46).

Rivalry over scarce and inequitably distributed resources shaped the nature of social interaction with legal authorities as conflicts over resources pitted one community against another, fostering a fairly rigid hierarchy of power among them. This rivalry also encouraged a highly militarized and often litigious political culture of a type that is usually said to be inimical to the implementation of central state law. Because many communities were based on ethnic ties, conflict frequently assumed ethnic forms. Tu si who were unable to muster military power to defend themselves often sought protection and redress in government courts or at the county seat (Ford 1977:21). One 19th-century commentator noted that "the lairds (to si) . . . are constantly at law with one another in the Chinese courts . . . an opportunity for the magistrates to enrich themselves (Clarke n.d.:122).

In addition to poverty and a hierarchical social structure, cultural practices did not appear to have encouraged the use of state courts for divorces. Although minorities in Yunnan were often described as more promiscuous and open about sexuality than the more prudish Han Chinese, with evidence of rather bawdy courting rituals and public nudity among some groups (Yunnan Provincial Archives [YPA] 89-1-36:75; Gao 1990:39; Carey 1900:490, 494, 509; Ford 1977:7), there is little evidence of women going to Han-dominated courts to divorce, as the Yi women did in the 1950s. The record suggests that marriage and divorce were usually community, rather than state, affairs. Sometimes, however, marriage was explicitly political, as when community leaders forged alliances with other group leaders through the marriage of their offspring. Not surprisingly, marriage disputes could erupt into full-scale community feuds. Among the Yi who lived on the Yunnan/Sichuan border, for example, there was a saying, "If something happens to a person, the whole clan has to take responsibility" (Bamo Qubu 1992:69-70). Marriages also involved the community, as many were arranged by matchmakers. Conflicts and separation negotiations were handled by someone with high social or political status in the community (Chuxiong zhou zhi 1991, vol. 1:373). For instance, couples among the Yi, Hani, Dai, and Bulang minority groups would arrange to meet with the village head, who would handle issues of custody and property division. In Yongren, a county in Chuxiong with Yi, Han, and other minority groups, the process was quite simple: The woman would return to her natal home. To formalize the separation, however, she was required to compensate her husband for his wedding expenditures (Gao 1990:46; Shi 1992:156).

Community involvement in marriage and divorce is important to our initial puzzle in several ways. In the West, divorce is usually conceived as an individualistic enterprise, a breaking away from community or family ties. In Yunnan, however, it is quite clear that the individual was inextricably tied to the larger collecfive. Because of this tie, most of the scholarship saw the Marriage Law as incompatible with Chinese rural life. Yet, it is equally clear that minority women in Yunnan Province did divorce, and did so without moving to cities and acquiring a more individualistic identity. Moreover, they went to Han-dominated courts to divorce in groups, and divorces were granted to them by a Han Chinese state that historically had usurped their land and had considered minorities to be "barbarians" in need of superior Han culture. I now examine the role played by the state in shaping minorities' reaction to the Marriage Law in China.

State-Minority Relations

As much as it is easier to speak of the relationship of "minorities" to the Han Chinese, it is clear from southwestern history that minority groups rarely presented a united front against Han Chinese conquest. There were struggles over scarce resources within and between groups, which were often exploited by central state leaders to pacify the local population. Arrangements between the state and minority leaders were but one part of a package of policies intended to secure the southwestern region of the empire and bring about the inclusion of minorities into what was considered the culturally superior Han Chinese civilization (Harrell 1995). At least according to official policy, the PRC state was as sensitive as its Imperial counterpart to differences between Han and "minorities" and was careful not to intensify preexisting conflicts and cleavages within groups. More than this, the PRC state, facing security threats along some of its long borders during the early 1950s, was hell-bent on pacifying areas that had long been prone to extensive militarization. Yunnan, with its warring minority groups, history of separatism, and antagonism to the central state, was certainly such an area.

Communist officials who had only recently arrived in the southwest from Communist-held territories in the northwest (where they had a political presence since the mid-1930s, a decade and half before they established control over the southwest) were often unprepared for the complexities of minority conflicts and people's historical grievances against one another. In the southwest, the CCP was forced to deal with a dearth of trained personnel (none were very knowledgeable about minority customs or languages), long-standing animosities among ethnic groups, and widespread resistance to the Communists led by local minority elites (Chuxiong zhou zhi 1993, vol. 2:106-8). In this hostile environment, the task of pacification fell largely to military and political officers who had begun their careers with the CCP in the northwest, where their bases were located prior to 1949 (Chuxiong zhou zhi 1993, vol. 2:23). Schooled by military struggles against the Japanese and Nationalists and by harsh intraparty struggles, Communist soldiers overwhelmed these minority-led "armies" (Solinger 1977:96).

During military pacification campaigns, soldiers and political officials had great difficulty differentiating between leaders of resistance movements and rank-and-file soldiers, and between Han and ethnic minorities, who often looked alike, lived together, and spoke similar dialects. Officials whose primary political experience was in the northwest were accustomed to associating all "feudal remnants"-as these groups were labeled-with "landlords" who should be either "struggled" against or summarily executed. Despite official policy that called for conciliation with local elites and respect for minority customs, many Han officials were reportedly unable or unwilling to distinguish between former Nationalist officials, Han, and minorities, or between Han and ethnic minority landlords. This problem was deemed "experiencism" (jingyanzhuyi): the "blind" application of a political modus operandi that is suitable in one area to a different area, without sufficient consideration of the differences between them (Wang 1989:45, 125, 133). In a Chuxiong District, for instance, the local government convened a meeting of 28 Yi leaders, and then, in direct violation of central state policy, executed them (CXA 11-1-14B-1:134). Other reports from around the province found local minority elites in panic, afraid that they too would be executed (Wang 1989:16).

CCP and PLA counterattacks targeted not only local minority elite military power but the sources of their economic strength as well. As noted earlier, local elites in Yunnan (and Sichuan) often enriched themselves and their communities through the opium trade. However, both the Nationalist government and the CCP viewed the opium trade as a symbol of China's humiliation before the West, and consequently enacted laws banning its purchase and sale. The opium problem was particularly vexing in Yunnan, given that ethnic minorities, who were now theoretically a "protected" group, were heavily involved in its trade and consumption. According to some reports, it appears that state officials in Yunnan usually sided with the victims of the opium trade, and took a very firm stance against the use of arms to resolve intra-ethnic disputes. In one case in northern Yao'an County, for instance, several White Yi serfs escaped from their Black Yi masters and sought refuge at the district government offices. The district government freed the serfs and then confiscated the Black Yi's opium and warned them to no longer "use force to resolve their community conflicts" (CXA 11-1-14B-1:103-12). During the initial stages of the Revolution, weaker minority groups thus learned that district governments, despite being Han, unfamiliar, and new to the area, could be used to solve all sorts of community and resource-based grievances. Memoirs penned by formal officials in this area also indicate that district governments resolved disputes over water supplies, public land, and forest resources among a variety of ethnic groups in the region. These disputes came to the state's attention only when the weaker community petitioned the district government for redress, and they were amicably settled by the district only after the conflict was on the verge of erupting into a full-scale feud (Duan 1993:271-73, 278-81). Whether because of its coercive power or its ability to serve as relatively objective outsiders to interethnic conflict, government institutions associated with central state power came to be seen by the region's residents as legitimate, powerful, and helpful organizations.

This perception of the state intensified when land reform began in late 1951. According to official land reform policy, local officials were required to study discrepancies in land-holding patterns in different areas of the province (such as mountains and valley basins), as well as the ethnic composition of each area. Land reform was to commence only after proper investigation and if there were a sufficient number of poor peasant officials. Moreover, land reform was not to be implemented in a way that would increase the hostility among ethnic groups or strike a "crippling blow" against minority elites; cadres should "emphasize each group's common interests" (Chuxiong zhou zhi 1993, vol. 2:117; Wang 1993:273-74). Unfortunately, policymakers were largely unfamiliar with the history of ethnic tensions in the area and thus seem to have been unaware that minority groups had few "common interests" to begin with. Land reform unintentionally intensified these disputes by making very tangible rewards and punishments available to the peasants.

In many cases the violent and revolutionary struggle over land widened preexistent cleavages in minority villages by terrorizing political and economic elites and older people. For instance, when a minority group (the Miao) installed a rich peasant as chairman of their village's peasant association, northwestern cadres, as was their habit, threatened a purge. The new chair, terrified, committed suicide. Other Miao then panicked, fleeing the village (Wang 1989:89). In a Yi community, a 40-year-old Yi woman was murdered by hanging (CXA 11-1-14&1:134). Elsewhere, there were complaints of cadres beating minority village headmen, secretly organizing unauthorized peasant associations to encourage peasants to withhold rent payments, and itching to attack tu si, calling them by the same term used in Han areas"village bullies." Some headmen, fearing struggle, abrogated their traditional responsibilities in the village completely, while many others showed passive resistance. Some headmen told peasants, "If you have a problem, go to your government, not me!"(CXA 11-1-14B-1:164, 121-22).

Land reform did not resolve longstanding land conflicts either. Political radicalism raised villagers' expectations of what they might gain from the Revolution. All over the province, there were reports of landless peasants making "new demands" and "taking initiative," with a concomitant rise in the number of land disputes handled in district governments and courts (Wang 1989:73, 141). These demands were difficult to fulfill, howeverparticularly in the minority-dominated mountains, where there was very little arable land. One Yi complained to senior Han officials, "In Luquan County the minorities are mostly in the high mountains. There are only a few landlords and not a lot of land. During land reform, the fruits of struggle were also few" (CXA 11-1-14B-1:140). Officials who rejected these proposals as too radical could still not escape peasants' new demands. Because land reform intensified disputes over land, the state had no choice but to play an increasingly critical role in resolving them.

This pattern of state-minority relations in Yunnan from the late 1940s to the early 1950s clearly shaped the outcome of the Marriage Law. Because of previous intercommunity conflicts among minority groups, the post-1949 state was forced to expand its mediatory and adjudicative roles, which later spilled over to the domestic sphere. Conflicts that minority elites would have handled in the pre-1949 period were transferred to the Communist state, as minority elites abandoned their traditional roles when they thought about the possibility of facing CCP persecution. At the same time, when the state intervened in disputes and favored those at the lower rungs of the local status hierarchy, other powerless groups learned that the power of the state might be invoked for their benefit as well.

The evidence concerning the impact of land reform and early 1950s politics in this region supports such an interpretation, as does some comparative evidence. In India, for example, Marc Galanter notes that the establishment of British courts in rural areas resulted in a "flood" of litigation, as individuals who could not get justice in their village owing to low caste status saw an opportunity for legal relief (1989:19-20). In colonial Africa, political scientist Crawford Young notes that foreign "district officers found little difficulty in attracting as many disputes as they wanted to solve; those who believed themselves unlikely to prevail through African juridical institutions . . . viewed the district officer as an alternative opportunity for litigative action" (Young 1994:115). This was also the case in China, where ethnic minorities who had suffered at the hands of Han Chinese (particularly in the distribution of land) or of other minorities welcomed an outsider's presence. This role clearly carried over to the Marriage Law, as poor minority women went to courts and districts to sue for divorce. Yet, even with this evidence, several questions remain: Should we place explanatory weight on the ethnicity of the divorce petitioners, on the fact that they were minorities per se, or that they were rural? Was the dramatic increase in legal suits presented in Chinese courts the product of a regional dynamic, or of new opportunities created by the state for minority groups more generally?

These questions can be partially answered when we take a look at how minority groups in a different region interacted with officials implementing the Marriage Law. Were ethnic minorities in a major city such as Shanghai as inclined to invoke state law in their private affairs as their rural counterparts? Here the evidence is more ambiguous, in part because I have not been able to find any statistical data on urban minority divorce and am thus forced to rely upon more general statements of ideological support or resistance to the law. Nonetheless, the evidence suggests that community elders in the urban areas played a greater role in preventing divorce among youth than did their rural counterparts. The state, moreover, was not seen as an emancipatory organization, and it did not play a very important role in settling minority-majority disputes.

The available evidence suggests that reactions to the Marriage Law were mixed. Some female Buddhists in Shanghai were said to enthusiastically support the law, calling it "necessary" (SMA C32-2-3:24-25). At the same time, however, many others were worried. The Marriage Law attacked various feudal beliefs and superstitions, and Shanghai Buddhists were concerned that the CCP's conceptualization of "religion" and religion-based marriage practices would be conflated with "feudal superstition" and then suppressed (SMA C32-2-3:24-25). Among Muslims (Hui) there were also mixed reactions. Elders and even officials expressed support for the legal principle of "marriage freedom," but worried that granting children too much freedom would result in assimilation into the Han community and the gradual decimation of the Hui community. Some wanted the government to issue a legal regulation barring Han-Hui marriages. As one said,

The essential factors constituting ethnicity are language, literature, customs, habits and bloodline. If the bloodline becomes diluted, the Hui will no longer exist .This is an invisible destruction of the Hui. We support the Marriage Law, but there should be freedom only among those with the same faith. (CMA C32-2-3:36)

Given that Muslims were only 0.3% of the Shanghai population and that Han and Hui frequently mixed in the workplace, it was assured that intermarriage would take place. Elders recognized this and hoped that the non-Hui spouse would convert or that he or she would at least observe various religious rituals and prohibitions, especially not eating pork.

Muslim youth, however, much like their Yi counterparts, had a far more favorable reaction to the Marriage Law than their elders. One man surnamed Wang argued, "In the Hui marriage system Hui cannot marry non-Hui, and as a result parents arrange and coerce their children into marriage." The "feudal marriage system" thus "must be destroyed" in order to "increase the level of productivity." While others did not take such a radical view, it was generally felt that the Marriage Law should be used to empower youth vis-a-vis their elders, even if this came at the expense of their viability as a separate ethnicity in the city: "Marriage is young person's matter; young people need not accommodate themselves to older people's problems" (SMA 032-23:38).

Uigur and Mongol youth in Shanghai supported this stance, along with another urban minority-Catholics-who believed that marriage should be "free," at least in the sense that parents should not coerce a couple into marriage. However, this support did not extend to other key provisions of the Marriage Law, particularly the right to divorce, "equality" between men and women, and inter-faith marriages. The CCP in Shanghai was hostile to Catholics and accused them of harboring "feudal beliefs," of "backwardness," and of serving as the "running dogs of imperialism" during the years of Western domination of the city (SMA C32-2-3:40). Higher-level officials advised their subordinates to avoid conducting Marriage Law propaganda among Catholics, given the extent of their opposition to several key provisions. There were, to be sure, some Catholics who hoped to use the law's divorce provision to escape abusive relationships. Often this was to no avail, as fellow Catholics threatened them with excommunication (SMA C32-2-3:40).

These accounts reveal several things. First, it is clear that there was no China-wide "minority" or "ethnic" support for the Marriage Law. While the opportunities the law provided were the same for each group regardless of region, urban minorities, owing to their demographic situation (in the case of the Hui), community structure (which was not radically altered after 1949 owing to the lack of a political movement similar to land reform), religion, and pre-1949 politics (in the case of Catholics), appeared to have been somewhat less interested or less able to take advantage of new legal opportunities than their rural counterparts. This finding helps strengthen my argument that the legal dynamics I witnessed in Yunnan Province were most probably based on rural political developments, such as land reform and military pacification, that were superimposed upon preexisting, ethnic-centered, community conflicts needing state intervention.

Another interesting finding is the role that generations played in the implementation of the Marriage Law. Among both Hui and Mongols in Shanghai, youth wanted to mobilize the law to contest parental power. Given that those taking advantage of the Marriage Law in Yunnan were also young and that land reform inadvertently targeted elderly Chinese, it is worth exploring a bit further how generational dynamics intersected with the Marriage Law. A generational explanation might be more persuasive to some than the spatial one I articulated. After all, if rural and urban youth both responded positively to the Marriage Law, what was more important-their age or locale? As I argue later, rural youth (both minority and Han Chinese) were far more radicalized and were in a far better structural position to take advantage of the law than were their urban counterparts, even when the latter expressed support for it. China's Communist Revolution did not affect rural and urban youth in quite the same way.

Generational Conflict and the Marriage Law

It is interesting that not one of the 19 articles in the 1950 Marriage Law specifically accused older Chinese of violating the rights of the younger generation; neither do accompanying policy documents. The generational issue snuck into the Marriage Law unintentionally, as the result of a much older elite discourse on the solutions for China's difficult predicament in the 19th century, a time when it faced numerous encroachments upon its territorial integrity and national sovereignty. For many intellectuals, Confucianism was to blame, and China's salvation hinged upon the overturning of many of its key tenets, particularly "filial piety" (Chou 1960). The traditional family structure, it was argued, oppressed the young by forcing them into relationships not of their choosing and gave them no role whatsoever in the management of state affairs. The widespread tropes of establishing a "new China" and a "new society" after 1949 indicated both political and generational transitions. Youth, no longer bound by the restrictions of Confucianism, would take the lead. The Marriage Law, by explicitly attacking the values and practices of "feudal" China, was therefore an implicit attack on the people who, knowingly or unwittingly, supported that system. Most frequently, elderly Chinese were accused of perpetuating these values and practices. Such attacks on the elderly occurred largely in rural areas where a social revolution took place, but not in cities.

When officials returned from Marriage Law training sessions and began to teach villagers about the law, they drew upon methods from the campaign that immediately preceded it: land reform. They convened separate meetings for young men and women, and older men and women, encouraging them to "speak bitterness" (suku) about their travails in the "old" society. Young women-the main target of mobilization of the Marriage Lawwere not always easy to mobilize, however. In some cases, husbands' families locked the women inside their homes, fearful they would file for divorce after a meeting (CXA 16-14-AI:2). Rumors about the campaign's objectives further complicated matters. Some believed that women who were "not married by 18 or 21 would be taxed or fined," and that women who were 25 and still childless would be "kicked in the ass by the government" (CXA 4-2-Al:97; CXA 16-16-131:75).

Lacking the time to conduct proper investigations and generally unfamiliar with the new political lingo, officials frequently selected targets they believed to represent the feudal family order; one report complained that officials "are not clear on what behavior is Teudal'; they associate feudal behavior with the landlord class" (CXA 16-3-AI:20). When propagandizing the law in villages, they did not bother to recite specific articles of the law, but instead adopted simplistic rules of thumb, such as "men are bad, women are good," or "old people are wrong, daughters-inlaw are right." Young women were said to have been particularly radical, using the land reform discourse of "revenging old debts." Others even said, "Men and old ladies who abused women should all be executed" (YNA 89-1-22:27-28; CXA 16-15-111:176).

Young women's thirst for revenge was sometimes quenched. In Yao'an County, reports noted that "very many daughters-in-law spoke about their past bitterness, and are now requesting that the government educate their mothers-in-law and husbands" (CXA 11-4-1411:54). When officials located older people who had been particularly abusive toward women, they were subjected to public trials and in some cases were sentenced to hard labor for 2 to 8 years (YNA 89-1-15:42). Witnesses to these events panicked, fearful that they, too, would be selected as targets of struggle. Men-including local officials, one report noted-were terrified when they saw village women congregating in separate rooms to "speak bitterness" (CXA 16-3-AI:18; CXA 16-15-BI:89).

As it turned out, their fears were justified. Some rural young women took the Marriage Law's promise of freedom quite literally. As one report about young minority women complained:

Women understand [marriage] "freedom" as allowing them to behave rashly in sexual relations. If they are criticized by the court or by other peasants, they feel that their "freedom" is being infringed upon. Others who genuinely seek freedom [to leave a coerced marriage] might encounter the resistance of their family or the Peasant Association. If they do, they are incensed that their freedom is being limited and resist. Some even commit suicide or abuse and kill their own children to get revenge. (CXA 11-14B-1:126)

As might be expected, such a sense of freedom destabilized family relationships, and not a few young rural women, convinced that the state would now support them, flocked to state courts to petition for divorce. Reports show that the number of divorce cases received by state institutions doubled, and sometimes tripled, after land reform and the Marriage Law (CXA 16-8-B1:69; CXA 16-3-A1:180). Courts frequently granted divorces based on a woman's unverified claim that the marriage had been coerced or arranged. Nor did judges differentiate between Han and minority marriages. When they handle cases, a report noted, "most judges use the Marriage Law, and grant divorces if the woman claims the marriage was arranged." Some reports accused judges of "rash judgments" that resulted in men's suicide after hearing of the divorce (CXA 11-14B-1:126).

Such a sense of rural generational empowerment persisted well into the 1950s. By the late 1950s, there were signs that the Marriage Law and other changes (such as collectivization) had brought about enough instances of abuse and neglect of elders to warrant special state investigations. These investigations revealed that there were daughters-in-law who used the Marriage Law and the language of land reform to abandon their in-laws, abuse them, or drive them to suicide. As one 1958 investigation found:

Most abuse cases are the result of children hating that their parents are old and cannot work. Usually the son and daughterin-law collaborate. In other cases, the daughter-in-law initiates the beating and the son joins in, fearing that if he goes against her, she'll petition for divorce. (CXA 16-65-131:171)

Reports that reached the archives give little indication that the state stepped in to alleviate the plight of aging parents. Long ofter the revolutionary political campaigns were over, young rural women were able to use official state language to intimidate officials, husbands, and in-laws alike. In Yunnan, for example, a woman surnamed Hu despised her mother-in-law and refused to feed her. At the dinner table, Hu's mother-in-law pleaded for some food, but Hu refused, and told her, "Even if this food was shit, I still wouldn't give it to you!" Close to starvation, Hu's mother-in-law brought the matter to the attention of the township government, but during a meeting Hu confounded the officials by accusing them of "supporting the old and oppressing the young." (CXA 16-65-111:172).

Generational empowerment thus appears to be an attractive explanation for the behavior of the Yi women described at the outset, as well as for the spike in divorce cases after the Communist Revolution and the implementation of several key policies. Because rural social structure in the pre-1949 period made divorce difficult to obtain, it makes intuitive sense that when politics shook up and "loosened" that structure, it would be easier for women to access state institutions to divorce. Yi women, therefore, had more opportunities to escape from their families than did their urban counterparts. In cities, the state did not disrupt family structure. Instead, it focused on the restoration of political, social, and economic order, rather than pitting one generation against another.

While analytically satisfying, generational empowerment resulting from the Revolution probably does not explain everything. After all, worldwide, most divorce occurs among the young. Moreover, a general sense of empowerment among youth does not necessarily explain how women actually acted upon their desires and how they overcame numerous obstacles in their path. Many rural men, as we have seen, resisted divorce, often violently (YNA 103-145:149). To further explain the rash of family-related lawsuits we need to take the argument about the impact of rural politics down one notch and look at the "micro-dynamics" of legal processes in rural Yunnan. One clue in particular seems worth pursuing further: that the Yi women I discussed previously divorced in groups without going through mediation prior to going to court. If we can explain this, we can then differentiate between causes leading to a general sense of generational empowerment and actual outcomes of divorce cases.

Law, Community, and Mediation

In much of the literature on rural society, peasants' embeddedness in community has been said to militate against their willingness and ability to use state law. Over the centuries, many rural communities developed myriad lineage regulations governing marriage and property, mostly reflecting male concerns with patrilineal continuity and power. This was the case in China as well: Many disputes were resolved according to customary law (Fei 1997). The Marriage Law, with its liberal divorce provisions, equitable post-divorce property arrangements, and modern emphasis on love as a necessary component of happiness in marriage, threatened traditional village laws and power structure. After all, if village women could easily divorce and take their property with them, village men would lose two resources critical to maintaining their patriarchal domination. Indeed, village-level resistance to the Marriage Law has been cited as one of the key causes of its early demise as a force for change in rural China (Johnson 1983; Wolf 1985; Stacey 1983). Such an outcome would certainly not surprise social scientists working on state-building in developing countries, who have tended to portray semi-autonomous rural villages and local "strongmen" as antagonistic to legal centralization (Migdal 1988). Even in the United States, it took several decades before the federal government was able to successfully prosecute Mormons in Utah Territory for violating laws prohibiting polygamy (Cresswell 1991). Still, we are left with the same puzzle presented earlier: In China, rural women went to court, avoided mediation, and were often successful at securing a divorce. I now examine two elements of community that worked to facilitate family change in rural China.

Community and Mediation

As noted previously, the secondary literature suggests a close connection between the "density" of social ties and people's opportunities to access state legal institutions. The general theoretical proposition is as follows: Where social ties are dense, there is less access to the state; where ties are thinner, there is greater access. In rural Yunnan, women were clearly enmeshed in overlapping networks of family, ethnicity, kin, and community, "dense" ties that would seem to work against their ability to access the state's legal apparatus. In several instances, this certainly was the case, as disgruntled husbands mobilized villages' political and security establishment to arrest and even rape divorce petitioners. In other instances, however, community had the exact opposite effect: Precisely because village ties were dense everyone knew about each others' personal weaknesses and peccadilloes, and this made making supra-village political and legal institutions the preferred address for people embroiled in family and community conflicts. What made such state institutions attractive, as well as legitimate, in the eyes of ordinary villagers was the rural revolution's replacement of older village elites who had "face" or status with inexperienced, "face-less," young men. Even though state elites promoted mediation as a cost- and labor-saving method of dispute resolution synergic with "Chinese tradition"a view echoed by some Western scholars (Wall & Blum 1991)village officials and petitioners alike preferred an outsider to handle domestic disputes in adjudicative forums. (See Huang 1996, and Bernhardt & Huang 1994, for evidence of such patterns in the 17th and 18th century China.) In China, peasants were, and continue to be, quite litigious, much to the state's dismay.

Archival sources clearly demonstrate that rural officials, precisely because they were embedded in social networks and were inexperienced in handling domestic conflicts, pushed divorce cases to courts and other higher-level authorities, despite their patriarchal interests. Many local officials recognized the limitations imposed by the local knowledge generated by dense community settings. In a township in Yunnan Province, for instance, mediators said, "Mediating is just like whipping a mule: You whip the mule and it farts, stinking you up. You try to mediate and everyone will be angry with you." Another complained, "When you mediate a marriage dispute both sides threaten you, and each other, with murder or suicide. How can we possibly find a solution?" Others quipped, "Divorce cases give us a headache." Unable to reconcile the parties, village mediators adopted the pragmatic strategy of sending cases to higher-level institutions and letting them deal with the hassle. Before long, district governments and courts were complaining that village mediators were sending cases their way "as soon as there's a conflict in the village" (CXA 11-4-14B-1:141; CXA 11-4-14B-1:126).

In addition to their inexperience in handling family disputes, there were other reasons local Chinese officials would send family disputes to higher-level political authorities. Many of them were firsthand witnesses to the wave of executions and confiscations of property and land during land reform. When village officials heard (more often than "read," as many officials were barely literate) that the Party had promulgated a Marriage Law and that its basic "spirit" was "the protection of women's rights," many feared that if they botched a mediation session and something terrible happened, they would pay a heavy price. Indeed, their concerns were justified, since one well-worn method of revenge in Chinese culture (particularly among young women) involved committing suicide on the premises of those who offended them (Wolf 1975; Lee & Kleinman 2000).12 During the Marriage Law campaign, some women feigned suicide by leaping into rivers and screaming, "I'm drowning, I'm drowning!"-even though the water level only reached their knees (SMA A71-2-1856:4).

This coincidental constellation of traditional cultural practices and revolutionary rural politics made distant legal institutions both accessible and politically desirable. According to one report, mediators feared "committing an error and being criticized," and "being held responsible if someone dies" (CXA 11-414B-1:141; CXA 11-11-14B-1:46). Action flowed from these considerations, with mediators instructing petitioners to "go to court" and let judges deal with the disputes, rather than addressing them in the village.

But even if local officials were tempted to get involved in settling villagers' marital disputes, there is little evidence suggesting that villagers preferred to consult with them anyway. After all, officials' personal quirks prior to their recent elevation to the political class were well known. A typical report noted that "women are not willing to go to village or township officials for mediation, and are not willing to accept village mediation at all" (CXA 11-11-14B-1:46). The women preferred to go to those institutions that were not only more authoritative in the state hierarchy but also more legitimate, since large geographic distances between the village and the state did not allow for the generation of local knowledge. In short, women in rural China practiced "comparative law." Reports from several areas of rural China indicated that supra-village political and legal institutions, not the more familiar villages and townships, were the main focus of petitioners' legal strategies. Inundated with lawsuits, court and Women's Federation reports complained that "women go rushing off to court, IMAGE FORMULA 67

traveling for days, just because of a trivial dispute" (CXA 11-1114B-1:104-5; SMA C32-2-5:5).

These tactics, it must be emphasized, did not originate with the advent of the Communist Revolution and the Marriage Law. Recent path-breaking research on law and society in the Ming (1368-1644) and Qing (1644-1911) periods clearly shows that, contrary to the popular idea that Chinese eschewed the formal legal system, villagers-men and women alike-used village mediation and the District Magistrate's office in their pursuit of justice. In one detailed study of lawsuits in Fujian Province, Melissa Macauley found that "when Chinese peasants could not resolve their disputes over land, debts, marriage, and other, primarily property, issues, they turned to the . . . courts for a settlement. Far from abhorring the state court system . . . they increasingly relied on state structures to resolve disputes which were beyond the powers of local people of influence to address satisfactorily" (1994:84). What made courts more important in the Communist period, however, was the impact of the rural revolution on local elites, many of whom had performed mediation in the past but were now under a political cloud, if they were even still alive. As a result, the opportunity structure for divorce expanded.

Community and Collective Action

In addition to facilitating a legal strategy focusing on appeals to supra-village institutions, community also made it easier for some peasants to sue for divorce because it provided the organizational basis for collective action. In Yunnan, divorce could be a collectivist enterprise, particularly among young women who shared roughly similar lifecycle experiences, such as the early death of parents, marrying out of one's village, and suffering the high-handedness of their mothers-in-law. Well before 1949, village women belonged to groups with mixed-sex membership, such as secret societies and lineage organizations, as well gendersegregated, ad-hoc communities, such as during marriage and mourning rituals. On other occasions, village women went as a group to sell their produce in rural China's periodic markets (Fei 1945:166; Vassal 1922:60). Ironically, the Marriage Law, which was based on individualist premises, provided ample opportunity for this collective consciousness to translate into legal action.

When village officials were instructed to implement the Marriage Law, many immediately realized that it would be extremely unpopular among men, as they faced the prospect of losing their wife should she successfully petition for divorce. In many cases, the same officials who were expected to enforce state policy resisted it most violently, sometimes using male-dominated village institutions, such as militias and peasant associations, to beat and lock up women who "dared" to petition for divorce. While such events were certainly dramatic, they obscured something extremely important: Often, village officials brutalized the first woman who petitioned for divorce, believing that such action would deter others from following her precedent. Although this sometimes succeeded, all it took was one case of success-a particularly feisty woman willing to take a risk and bear the "start-up costs" of collective action-to establish a precedent for other women and to lower the risk of subsequent challenges to village authorities.

Village officials thus found themselves in a pickle: They wanted to prevent divorce, but they knew they could not lock up or kill all of the village's young women. Nor could they prevent women from discussing marriage and divorce with one another. Take for instance the case of Zhang Chaoxiu in Yunnan. Zhang had an arranged marriage to a man named Zhu Rongsheng. In 1951, Zhang demanded a divorce. Encountering obstacles at the township, she reportedly "cried all the way from the township to the county court." When the court granted the divorce, she returned to her village. The report noted that the other village women "had been concerned about her marriage problems, so when they heard the government had granted her a divorce, over twenty women went over to see her" (YNA 89-1-15:40).

One can only guess at what was discussed in these informal gatherings, but it would not be unreasonable to suspect that Zhang's divorce encouraged other women to follow suit. One report noted that in villages "women want to divorce, but cannot decide who will take the lead." They would go when they "saw others" go to court (CXA 16-14-AI:2-3). Indeed, courts complained about women coming in groups to divorce, as well about as the ruckus caused by women loudly talking about their marriages (CXA 4-2-A1:97; CXA 11-4-14B-1:126).

In Yunnan, it appears, community could contribute to divorce by ruining the prospects for successful mediation and by creating the conditions for successful mobilization: trust, communication, and common experiences. The behavior of Yi women who went to court in 1956 might very well be explained by the fact that they were probably from the same village, and thus knew each other well. When they returned to their village they were able to present a united front against their detractors, giving the women strength in numbers. In other words, divorce in rural China should not be conceptualized as the end product of individualization or modernization but rather as the result of relatively high levels of trust, solidarity, and willingness to take risks within group settings.

This analysis suggests that we might need to shift the theoretical framework from one that places the individual front and center to one that focuses on the conditions for successful mobilization of law. Whereas modernization theorists would not predict much rural divorce in Yunnan, collective action theorists such as Charles Tilly (1978), Doug McAdam (1982), Sidney Tarrow (1994), and Elizabeth Perry (1980)-all of whom stress the role of networks ("traditional" or otherwise) in social mobilization-could have easily hypothesized this outcome. In a similar vein, June Starr's (1989:518-19) article in the LSR persuasively demonstrates that rural Turkish women's participation in gender-segregated social activities-hardly the stuff of modern feminism's ideal of equality between the sexes-facilitated their ability to take advantage of modern law. In Turkey as in Yunnan, we can argue that community networks were very important to the actual methods by which women sought a divorce.

While accounting for the role of community in explaining Yi women's divorce is very important, it is still debatable whether community was the decisive contributing factor. This is because there is compelling evidence that rural Han women in other regions of China also used numbers and solidarity to pave their way to courts. For instance, in Fujian Province, a district chief recalled in a meeting, "Ten women came for a divorce and all ten were granted divorces. We were afraid that if one died we would be held responsible." And a township official said, "One day four women came to the office and said, `We want a divorce. All of us or none of us.' What could I do? I granted them all a divorce" (SMA E81-2-117:57).

Cases such as those in Fujian appear to suggest that there was nothing unique to Yunnan or to the Yi that created the conditions for successful collective action. Instead, what seems to be the case is that there was something in rural Chinese culture, both in Yunnan and elsewhere, which led rural women into the arms of the state and provided a sufficient support structure for them to secure divorce. So far, however, the evidence for this is fragmentary and mixed: In the legal case that opened up this article, court officials in Yunnan hinted that there was something culturally specific to the Yi women that contributed to divorce (by claiming that the Yi "loved suing for divorce"). Moreover, contemporary Chinese are still inclined to attribute greater promiscuity to Yunnan minorities. Nonetheless, the evidence previously presented from other areas of rural China hints at a more generic pattern. I now turn to the question of rural culture and law in the southwest and elsewhere: In what ways did culture contribute to rising divorce rates in rural China?

Culture and Law

According to the secondary literature, there are simply no cultural causes for divorce in rural China. In part, this is because scholars have largely refused to admit the possibility of rural divorce in the first place; "rural divorce" is an empty category. Equally important is the widespread notion that peasant culture, absent modernizing forces imposed from above, is too patriarchal and all-encompassing to allow for successful challenges to it. It has been argued that during a brief but shining moment in the 1950s, when the state actively enforced the Marriage Law, there was some hope for dramatically improving women's status in Chinese society, which was dashed when the state halted all-out legal mobilization in 1953 and patriarchy again raised its ugly head. As Friedman et al. (1991:122) noted of a conservative Han region of north China, "No practical challenge was mounted to the values, practices, and institutions of male supremacy." Such a conclusion, unfortunately, may not be very relevant to our puzzle because of the particularities of ethnic minority culture in Yunnan, and in the southwest more generally. As noted earlier, there is ample literature demonstrating differences between Han and minority cultures, particularly regarding sexuality, marriage, and divorce practices. Indeed, archival evidence from Chuxiong Prefecture in Yunnan does point to several distinctive features that contributed to divorce there, and perhaps to the court visit of the five Yi women.

One key feature that sets Yunnan apart from many other areas of rural China is its striking beauty. The province beckons residents and visitors to scale its mountains and walk through the countryside; the weather is mild all year. Not surprisingly, many cultural activities involve commingling with nature, such as the "splash water" festival and communal bathing in mountain gorges among the Dai minority. The Yi minority had similar rituals, but they involved more trips to mountains than to rivers. One such ritual was called "drink mountain wine" (chi shan jiu), which involved drinking, singing, and dancing around a bonfire in the mountains. While chi shan jiu did not lead to divorce prior to 1949, Communist policies that provided rural women new institutions to handle family conflict after the Revolution led to not a few chi shan jiu-inspired divorces. In the early 1950s, when the CCP established political control during intense political campaigns, chi shan jiu was temporarily suspended. A bit later, however, as the intensity of campaigns subsided, young Yi men and women resumed their mountain excursions. During peasant association meetings, one report complained, men would whistle to the women, the women would holler in reply, and the group would then run off into the mountains. During chi shan jiu couples had sex. After their moment of passion, some became dissatisfied with their original spouses, and soon after-ward pressed for divorce (CXA 11-4-141:125).

Chi shan jiu continued to shape interpersonal relations, as well as court activity, well into the mid-1950s, when our Yi women arrived in court. In Chuxiong, courtship, sex, and adultery during chi shan jiu led some couples to hastily marry, while others went to court to file for divorce. In an analysis of 80 divorce cases from 1956 in Luquan County, 13 were caused by adultery during chi shan jiu, and 13 by "rash marriages." One Yi, for instance, a man surnamed Chen, met his wife (surnamed Yang) during chi shan jiu, had sex, and married soon afterward. Arguments over "household matters" led Yang to file for divorce. Chi shan jiu, however, was not only Yi youth's most popular venue for premarital romps; it was also a source of postmarital adultery. Some Yi were notorious "home wreckers" during drinking sessions. A Yi man surnamed Zhang was said to have "destroyed marriages" by his persistent forays into the mountains. "Even after accusations led to his arrest and reform," the court complained, "Zhang and his friends continued to go with other women to the mountains, sometimes as many as four men and three women at a time." Some Yi became alcoholics after one-too-many drinking sessions and sold their bed quilts to make money to purchase wine. When spouses discovered the theft, they petitioned for divorce (CXA 16-27-Al :4).

In addition to chi shan jiu, Yi marriage practices may have contributed to a rising number of divorce cases and to a rather casual attitude toward divorce. In many Yi communities there was a tendency toward "delayed" marriages-a woman would formally marry in her husband's village, but would not live with him until up to a year or two later. In the interim, she occasionally visited her husband, usually to offer assistance with the harvest. Yi women, therefore, were quite mobile and knew the local terrain because of their travels from their natal homes to their husbands' village (Chuxiong zhou zhi 1993, vol. 1:373-74). Moreover, marriage itself did not signify a break in their relationship with their parents.

Yi women's postmarriage connection to their natal homes and the mobility this encouraged frequently appears in court cases and government reports from the southwest. Oftentimes, those officials responsible for chronicling family disputes in Yunnan expressed a certain degree of exasperation when confronted with this fluidity in social relations. In a township in Yongren County, for instance, officials complained that a woman named He Lanzhi was married five times before she turned 18 years old. On one occasion, she married one Xu Jiazhu one evening, but on the following morning "ran back to her natal home" (YNA 891-82:40).

The divorce case of Chen Faying and Tao Yingxian (of Mai Diao village, Luquan County) is another, better-documented example. This case began when Chen fell in love with another man, Yang Chunfu, and wanted to marry him. Chen's father refused the match and quickly married her off to a man from another village, Tao Yingxian. After this coerced marriage, Chen moved into her new husband's home, but, dissatisfied with the match, she "frequently returned to her natal home." Several months later, Chen petitioned the Luquan County Court for divorce. The court tried to mediate the case three times, but when she refused to reconcile, it granted her a divorce (CXA 44 Al:39-40).

Taken together, these cases suggest that divorce among the Yi was directly related to several aspects of their culture: drinking practices, a rather fluid marital situation in the early years, the support they received from their natal homes, and their mobility through rural space.13 Perhaps this is why Han officials believed that the Yi loved suing for divorce. My argument, however, is that such cultural features were not sufficient to cause the spike in divorce rates in the early 1950s; these same features, after all, were present prior to 1949, and yet divorce was rare in Yunnan. A more plausible explanation is that the Communists, by providing a counter-weight to local power (usually in district governments) and by proactively engaging the rural family in politics, provided the opportunity for these Yi cultural traits to become legally and politically salient. That said, were these cultural traits more prominent among the Yi minority, thus making the Yi more likely to turn to the law than would the purportedly more "rigid" Han Chinese? My argument about the crucial role of rural politics (with culture playing a facilitating role) in producing social change would be greatly strengthened if I were able to show that (1) in Han rural areas similar dynamics occurred; and (2) that divorce in China was less likely to occur in urban areas.

On both of these accounts the evidence is unambiguous. Contrary to the view of many Han Chinese who think that ethnic minorities in Yunnan are more "open" about sexuality, marriage, and divorce than themselves, it turns out that many of the cultural traits that contributed to divorce in Yunnan made their appearance is other, less-ethnic, provinces as well. As Chinese anthropologist Ellen Judd's (1989) research on several North China villages has shown, during the first years of marriage it was not uncommon for Han women to return to their natal homes, where their parents accepted them. For many women, this support structure was crucial when they sought to leave their husband and his family. Moreover, in rural areas near Shanghai, women took advantage of close networks to protest injustices suffered by other women, congregate around courthouses to "listen in" on cases, and frankly discuss other people's marital affairs (SMA A71-2-1958:57-60; QPA 48-1-13:100, 102). Nor was legal feistiness the sole province of women in Yunnan. In north China, supposedly the bastion of Chinese patriarchy, an investigation report from the mid-1950s noted that "men are afraid their wives IMAGE FORMULA 82

will divorce them, so they give into their every whim. Men buy them whatever they want. If their wives do not feel like working, they don't dare say anything" (QPA 48-2-59:78). Village women, the report complained, "loaf around at home and don't cook," but instead "demand that their husbands cook and serve them" even after the men return tired from work. Complaints were ineffective: When husbands "tried to reason with them," the women reportedly "threatened to divorce." Some threats were carried out: Marital relations in the township were described as "unstable," and "rash divorces," usually initiated by women, were said to be "relatively common." (QPA 48-2-59:78-79).

Such willingness to use law in "conservative" areas of rural North China stands in marked contrast to the behavior of the most "modern" segment of Chinese society-educated and welloff urban elites. Many of these elites did not experience the Revolution as "emancipation" but as repression, since the Communists were a rural-based party that suspected urban denizens of political disloyalty. After 1949, the Party made a concerted effort to extract resources from the urban middle and upperclasses. Another source of conflict was that elite family structure was patently unacceptable to the new state, particularly the presence of concubines, second wives, and mistresses. In Shanghai, for instance, a daughter of a well-known Shanghai capitalist said, "The Marriage Law campaign's real objective is to confiscate the capitalist class's concubines" (SMA C32-2-3:42). In Beijing, a merchant who was rounded up by local officials said, "My money was [already] confiscated, now, with the Marriage Law, one of my wives will be taken away too!" (BMA 84-1-32:38). Other reports noted that businessmen were "very worried, unwilling to talk," and "panicked." Other wealthy men, reports found, were unhappy with their marriages, but did not request divorces because they were worried about how Communist courts (whose judges were often military officers) would handle property, housing, and custody arrangements (DCA 11-7-17:14; DCA 11-2-16:2). Fearful of how the Communist regime would deal with their family affairs, such families typically resolved marriage difficulties through mediation leading to reconciliation, rather than through the same courts that handled political and criminal accusations during previous campaigns UAA 23-2-2:11-13).

Even though the winds of the Marriage Law eventually died down after the early 1950s, there is little evidence that the CCP allowed urban residents more freedom. In fact, statistics from one major city-Tianjin-reveal that the overall number of people getting divorced declined year by year until 1966, whereas in many rural areas there were as many divorces in the early 1960s as during the height of the Marriage Law campaign in 1953 (Xu 1994).

Although the Communist Party arrived late to cities, it quickly learned the best ways to dominate the key sectors and populations of urban society. The mid-to-late 1950s witnessed the consolidation of the "work unit" (danwei), which soon came to structure urbanites' lives by controlling the distribution of scarce resources and fulfilling certain welfare functions as well as public security (Li & Perry 1997). This was particularly the case for white-collar professionals, scholars in universities and research institutions, and workers in large state-run factories, all places where the presence of the party-state was dominant at all levels. In these units, every person had a dossier documenting their political reliability, political history, and personal details. Someone seeking a divorce, for example, would first need permission from his or her immediate superior and the Party Secretary well before he or she could proceed to court. Using modern communication infrastructure, courts called work units to verify testimonies, and any discrepancies resulted in divorce cases being rejected. In other cases, urban density made it easy for courts and work units to summon plaintiffs for multiple hearings and send investigation teams to rummage through their personal belongings, a highly risky proposition (SMA C31-2-1064:1-4). But even if these modern obstacles were overcome, courts were more inclined to tell plaintiffs to "rely on their organizations" than to offer a remedial solution (SMA C31-2-1064:1-4).

All of these urban features significantly increased the risk of seeking a divorce using the Marriage Law and deterred many from even trying. The evidence regarding the urban working class is more mixed, however. Those employed in large state-run factories faced obstacles similar to cultural and economic elites. However, less-privileged workers, usually working in smaller factories or self-employed, had an easier time getting to court due to a less-dense state presence in their workplace. Investigators repeatedly complained that workers' (many of whom were transplanted peasants) excessively contentious legal culture manifested itself in excessive trips to court. Such litigiousness, however, did not result in many divorces, since the state was able to quash many suits before judges could render a verdict (SMA C31-2-368:9).

Nevertheless, there were more than political and financial reasons to avoid the state. Culture-based claims to high status were also at issue. Reports emphasized that elites not only feared the state (not surprising given that the Revolution moved from the countryside to the city) but also were unwilling to expose their private difficulties in a public state forum. For the urban elite, keeping "private matters private" was a matter of maintaining class status (or face), and differentiating themselves from the "let it all hang out" peasants (and workers). As one intellectual told me, it simply would not do for someone of his class to be seen screaming at his wife in public, and certainly not in a state forum (Interview with SG, Shanghai 1994). Among this group, the thick line between "public" and "private," together with the dense state presence, prevented many of them from even discussing their difficulties with friends or state agencies. In a case in the city of Hangzhou (circa 1953), for instance, the wife of the manager of a barbershop was having an affair. After her husband discovered her and her lover in bed, she threw herself in a river, committing suicide. According to the investigation report, no one was even aware of her difficulties. "In order to save face," the report noted, "family conflicts such as these are usually not spoken of The women don't tell [their problems to] fellow employees, relatives, or friends, nor do they go to court to ask for assistance" (SMA E81-2-115:18).

Officials in Shanghai and Beijing reported similar experiences. In Shanghai, a Women's Federation report from 1951 found that "among the bourgeoisie, the women suffer largely from psychological abuse; on the outside everything is tranquil, but inside they are suffering . . . they are quite vain and concerned with face" (XHA:9-10). And in a meeting with female teachers in a well-heeled district in the city, Women's Federation officials complained that it was difficult to help the teachers because they "kept their problems to themselves" (SMA CI-2611:37-38). For these urbanites, the new Communist state and its laws were, in some sense, culturally foreign even if they were promulgated by their legally and ethnically sovereign state. Not only were local state institutions (such as mediation and residence committees) dominated by illiterate workers and former People's Liberation Army officers, its new Marriage Law demanded that they break down the cultural barriers that allowed them to claim elite class status in the first place. Not surprisingly, many avoided the state.

Truth be told, very little in the previous statements would surprise social historians of Russia, France, or Japan. In all these cases there is a good deal of evidence demonstrating that the class differences in China regarding the "thin" or "thick" line between private and public, as well as the bold or passive way of interacting with the state, crosses national boundaries. In Imperial Russia, for instance, historian Laura Engelstein notes that peasants "saw nothing mysterious in sex; everything was natural, out in the open, and therefore also completely social" (Engelstein 1992:118-19). There is also evidence that Russian peasant women, like their Chinese counterparts, could also be quite bold in seeking assistance from the state, particularly from courts (Farnsworth 1986:63-64; Lewin 1968:35). In France, Martine Segalen reports that when peasant women gathered in groups, they spent much time "criticizing, denouncing, insulting, slandering, [and] relating family histories" (1983:129, 139). In Japan as well, anthropological studies in the 1930s found Japanese peasant women to be quite bawdy and willing to leave their husbands. However, as universal education spread, such behavior was increasingly considered to be uncivilized; granddaughters of these women in the 1980s were embarrassed to hear about their behavior (Smith 1981).

As for urban elites, the same sense of unease experienced by those in China was shared by elites in many other countries in both the East and West. In Russia, Engelstein cites one observer of the Russian urban scene who noted, "An educated person does not like and tries not to permit outside eyes to penetrate his strictly individual life" (1992:118). Likewise in Japan, family court mediators and judges-all of who come from a narrow strata of the elite-actively discourage divorce since it does not conform to what they understand to be "Japanese" social values (Upham 1987:12-13).

Given these similarities between peasants and elites in China and their respective counterparts in the former Soviet Union and Europe, how can we explain the different patterns of reacting to new family laws? Japan is explained relatively easily by the absence of a revolutionary family law after the Meiji Restoration of 1868. There, family policy was quite conservative, changing only after the United State's occupation after 1945. Of the other, more similar, revolutionary cases we would expect that Russian and French peasants would be those best able to adopt new legal strategies to leave unhappy or abusive relationships; given that a new law was on the books, the line dividing their private and public lives was rather thin and litigation in their cultures was not uncommon. Nevertheless, this was not the case: Evidence shows that peasants in these countries were barely touched by new family laws, whereas in China peasants mobilized the law to their advantage.

Here I must return to my original thesis regarding the distinctive characteristic of the Chinese Revolution: It was peasantbased and moved from the countryside to the city. The state in rural areas of China had the institutional presence, power, and legitimacy that urban-based revolutions lack. This distinctive revolutionary dynamic combined with peasants (including the Yi's) rather open way of dealing with family matters and bold way of interacting with the state to produce an outcome different from those countries whose revolutions were based in urban areas and only then moved to rural areas.

Conclusion

This perspective on law in China allows me to reconcile my findings with the literature that would probably not anticipate peasants going to court in groups to divorce. Peasants in rural Yunnan were not modern individuals. State officials, moreover, were often unskilled in applying law and unsympathetic to female petitioners. In short, they had few of the resources necessary for marginal groups to mobilize the law on their behalf. As we have seen, however, law was used-and in ways that did not always please those charged with implementing it.

My argument in this article is straightforward: This use of the courts would not have happened without the rural-to-urban dynamic of the Chinese Revolution. The Revolution, by decimating the pre-1949 political structure, encouraging the young (particu-Lilarly women) to empower themselves vis-a-vis their elders, and causing local officials to think twice about infringing on new rights, dramatically changed the opportunity structure for divorce. Whereas prior to the Communist Revolution a marital spat would have been resolved locally through mediation or through informal separation, political and social changes after the Revolution led to top-down involvement in domestic affairs (in the form of state mediation or adjudication) and to a sense among women that they had new rights and institutions that would help them in times of distress. Women who wanted to divorce could now approach the state (in one form or another), and the state, unlike before, had to deal with them. This was quite new for many local officials, and, not surprisingly, many preferred to pass the buck upward. There, in courts and district governments, higher level officials often gave them what they wanted, either to follow the "spirit" if not the letter of the law ("liberate women") or, far more prosaically, to get them off their backs.

But while the rural nature of the Chinese Revolution was critically important in changing family relations in rural China, it was not enough. To explain the puzzle I presented at the outset, I have also invoked the dual notions of both a feisty legal culture and a support structure. Whereas the stereotypical depiction of Asians is of passivity vis-a-vis legal institutions and a cultural preference for mediation, this article has shown, and recent historical literature supports, that peasants in Yunnan Province and other areas of rural China could be quite aggressive when pursuing their interests, even in lawsuits. This feistiness certainly contributed to their willingness and ability to take advantage of the Marriage Law. When they did so, however, they acted both as individuals and, more importantly, as members of families and communities. Before the Revolution and after, young women in rural China seeking divorce (and perhaps in Yunnan, to a greater degree owing to the relative insularity of villages) found support among other young women and were able to use their natal homes as refuges from abusive or unsatisfactory marriages. These women were certainly not modern in the sense of believing in equality or having a clear consciousness of their rights, but this did not stop them from taking advantage of a highly modernist law. In fact, it might be argued that it was their organized inequal ity, their participation in gender-segregated activities that promoted trust and solidarity, facilitating the organization of collective action for divorce.

In both of these respects-feistiness and social networks-I do not believe that Chinese women are very different from women in some other rural societies. As Arturo Escobar (1995) correctly points out, much of the feminist literature on "development" emphasizes women's weakness and passivity, despite very convincing evidence that rural women in the third world are not always passive, and can act politically without "modern" identity and consciousness. Even in India, not a country stereotypically known for proactive rural women, rural women have taken advantage of close ties to one another to protest injustice. In the early 1990s, for example, some women who objected to their husbands' drinking (which caused domestic violence) banded together and "began attacking the liquor shops, pouring the booze into the streets and shaving the heads of men found drunk there ... groups of women seized drunken patrons [of liquor shops], wrapped the skirts around them and paraded them through villages on donkeys to humiliate them" (Moore 1993).

If this is the case, if rural women in the developing world can exhibit a degree of feistiness vis-a-vis legal institutions and the state, as well as the capacity to mount collective action, then why is rural divorce a rarity in many of these countries? This issue is extremely complex and involves religious, cultural, and historical factors. In this article, however, I have suggested that an important part of the answer is in the way opportunity and support structures interact in time and space. The Marriage Law was backed by a simultaneous political and social revolution, and these changed (in ways intended and unintended) the opportunity structure for divorce in rural China in ways quite unlike other revolutions and states. This political explanation, I argue, best explains the legal dynamics in rural Yunnan.