Taiwan’s Intersectional Cosmopolitanism: Local Women in Their Communities
Over the course of hundreds of structured life-history interviews with rural women—mostly elderly women—in Taiwan and the People’s Republic of China during the past thirty years, I have found, not only that women over the past century have routinely made substantial economic contributions to their families, both natal and marital; not only that these contributions are routinely undervalued in household-level social crediting; but also that there is a qualitative difference between Taiwan and China in whether social credit is conferred at all. Taiwanese women’s contributions might often be credited at less than their actual value, but they were never unspeakable in the way that I often found Chinese women’s contributions to be. For example, in rural Hubei in 2010, after interviewing a woman in her nineties and learning how she had singly supported her six children during the war years by weaving, I was astonished when her sixty-something son came home and proceeded to tell me that I was wasting my time asking about women’s work because women cannot contribute very much, certainly not enough to make any difference to a family. But when I began to say what his own mother had done, she cut me off and shook her head—she did not want me to give social recognition of her contributions, even all those years later. Similarly, poor women in Beijing avoided social credit for their contributions to prevent their husbands from losing face (Evans 2017). But in Taiwan, both men and women talked very matter-of-factly about what women actually did—whether working in fields or factories—even though they might often value women’s contributions less than men’s (cf. Harrell  2015, 180). Why this difference in what is speakable about women’s contributions in China and Taiwan?
I suggest that the answer to this question lies in a larger historical pattern that differentiates Taiwan from China: acceptance of local women in common public space and the relationship of women’s public-sphere participation to cosmopolitanism. This relationship is intersectional—with interwoven and mutually reinforcing influences between gender and, in this case, indigeneity. Taiwan’s multiple historical colonizations render indigeneity a relative and historically shifting concept: each new wave of colonizers took the local peoples already in place as indigenous, heedless of lumping together—for example, as Bendi 本地 (local)—descendants of earlier colonizers (Hoklo and Hakka) with descendants of earlier subalterns (Austronesians) (cf. Brown 2004, 179, 181–185) (see table 1). In this article, I illustrate this pattern, examining three moments across Taiwan’s history—in the seventeenth, nineteenth, and twentieth centuries—when a populous but subordinated (and ethnicized) category of local women were themselves cosmopolitan actors—contributing importantly to a larger communitarian (community-building) engagement across diversity. Various masking processes can make it difficult to recognize women’s contributions, even when they are present in the historical and ethnographic records. But those intersectional contributions show that cosmopolitanism has been a fragile but recurrent social niche across Taiwan’s history.
Table 1. Ethnic groups in Taiwan’s History
* Other terms for this ethnic group include Hua 華 and Tang 唐.
** Many Qing documents refer to Austronesians as barbarians (fan 番), often distinguishing between “civilized” (shu 熟, lit., “cooked”) peoples, who lived on the western plains, and “wild” (sheng 生, lit., “raw”) peoples, who lived in the mountains and on the eastern plains (Brown 2004, 2010; Teng 2004). Japanese and Mainlander colonial governments treated mountain Austronesians (J. takasago高砂; gaoshanzu 高山族) differently than Bendi.
*** Sanyi was an alliance of people from Jinjiang 晋江, Nan’an 南安, and Hui’an 惠安.
“Cosmopolitan” is a word like “culture,” whose polysemic range includes contradictions. Each term has a general meaning implying the sophisticated worldliness associated with elite levels of wealth that facilitate education and travel. These popular meanings may apply to individuals, who are “cosmopolitan” and “cultured,” or to places, which are “urbane” (assumed to be urban as well as central). These meanings imply a contrasting barbaric condition or periphery, labeled with a variety of derisive terms including “provincial,” “hick,” and “rustic.” In short, these popular meanings are attributions of civilization formulated by metropole elites.
By contrast, in a social-science framework, all humans have culture (shared meaningful ideas that shape people’s understanding of their world) and all humans are potentially cosmopolitan. Philosophers talk about cosmopolitanism as an ethical orientation or moral principle that embraces an identity as a “citizen of the world” (from the Greek kosmopolitês) over other collective identities, especially those linked to membership in a family or clan, or membership in a city-state or nation-state (Appiah 2006). Citizenship implies community, one that is neither exclusionary nor elitist—cosmopolitanism as “the idea that all human beings belong non-exclusively to a single community” (Duara 2015, 21). This view emphasizes (a) individuality, by taking individuals as the basic unit of moral concern; (b) universality, by viewing every individual as having equal moral value; and (c) generality, by holding all individuals—not merely those with whom we share ties—as worthy of moral concern to every one of us (Taraborrelli 2015, xiv). First one takes individuals as the basic unit, then one views all individuals with whom one has personal connections to be equally valuable, and finally one grants value to strangers. But classification of human beings by gender does not fit neatly into this framework. Although every society necessarily has ties between men and women—and across gender differences, more broadly—many patriarchal societies readily accept the moral equivalence of men, despite differences that make them strangers, while refusing to recognize women (and other feminized genders) as morally equal to men. Theorizations of cosmopolitanism need to include all human beings present in a community.
In defining the community, however—especially within the universalist view of a single, global community—lies a tension between the high levels of face-to-face social engagement, cooperation, and mutual support implicit in the concept of community (Halperin 1998) and the necessarily imagined character of social interactions on a global scale (Anderson 1983; cf. Appiah 2006, 135). Circulation of written works creates the potential for imagining connections to distant places and peoples—a “cosmopolis,” or sphere of soft-power influence. Women and commoners are not excluded per se from such a cosmopolis, but literacy is necessarily limited to those classes who can afford education and, in many times and places, to boys and men. For empires that used a phonetic alphabet and widely promoted literacy, such connections were possible for a range of economic classes, via newspapers and other print media (e.g., Lewis 2009). But fewer commoners and elite women accessed the literary “great traditions” found in South Asia using Sanskrit (Pollock 2006), in maritime South and Southeast Asia using Arabic (Ricci 2011), or in East and Northeast Asia using classical Chinese (Park 2017). Moreover, colonial subjects, with restricted access to the imperial metropole and its knowledge, were precluded from cosmopolitan status because of their subaltern identity (S. Shih 2001, 97). Such exclusionary imagined communities—regardless of whether the inequality is imagined via gender, class, religion, race, or some other concept—necessarily attribute cosmopolitanism in terms of purported civilization.
By contrast, approaches to cosmopolitanism that examine whether empirical interactions among social actors are communitarian—whether they “practice” community (Halperin 1998) and whether these linkages cross lines of difference (Appiah 2006, 79; Delanty 2012; Nascimento 2013, 124–132)—allow for the potential to find cosmopolitanism within conditions of inequality. In these communitarian approaches, cosmopolitanism is experienced at the level of ordinary individuals, not merely among the elite; everyone’s first potential exposure to cosmopolitanism is local; and a person growing up in a cosmopolitan place may be cosmopolitan without ever having left that place.
There are significant differences between conceptualizing cosmopolitanism in terms of such a person engaged in diversity-crossing local social networks, on the one hand, and an individual who uses literature to leapfrog “provincial” attitudes and connect to a larger cosmopolis, on the other. These conceptualizations differ in the type of underlying linkages. Viewing cosmopolitanism as a specific connection to the cosmopolis (or the global) relies on imagined, often literary, linkages. By contrast, viewing cosmopolitanism as diversely engaged local networks relies on face-to-face linkages that build community.
In gendering cosmopolitanism, we need to recognize how gender influences these different linkages. Because women in patriarchal societies often are—or are supposed to be—cloistered (to varying degrees), individual women rarely achieve cosmopolitanism by leapfrogging local conditions and usually do so only with extreme wealth. Elite women’s knowledge of international fashion, however, often signals their entire household’s “cosmopolitan” connections to the colonial or global metropole. At the local level, even in patriarchal societies, face-to-face interactions between men and women are not usually perceived as engaging across diversity—perhaps because women are often viewed as “helpers,” not as fully autonomous and contributing social agents (Brown 2017). Moreover, women’s ability to practice communitarian social interactions across recognized diversity in their quotidian lives is often class-related, with women at the more modest end of the wealth spectrum often having more freedom for diverse social engagement.
For a local society to be cosmopolitan in this communitarian sense, diversity is not sufficient, as the nationalistic and racist violence of contemporary identity politics reminds us only too urgently. But engagement across that diversity—the kind of engagement that builds and maintains community, the kind necessary for conversations and cooperative ventures, the kind that is the antithesis of identity-based violence—may be sufficient. Engagement requires treating others with civility and taking them seriously as social actors—what we might consider respect. Thus, communitarian approaches examine the networked quotidian social interactions of all the people who actually live in a place, and these approaches consider a place cosmopolitan because a significant proportion of people living there have regular social interactions that engage across differences (e.g., Lewis 2009; Fewkes 2014).
Documenting emically salient diversity is therefore only the first step of identifying historical cases of communitarian cosmopolitanism. We must consider evidence of engagement across that diversity—such as interethnic marriage and economic cooperation—in contrast to evidence of exclusionary discrimination and violence. We cannot expect every person in any population to have cosmopolitan engagements; rather, we must consider whether the evidence of quotidian engagements is such that most people in the population experienced it. I argue that it is crucial to include women—and gender more broadly—in the diversity across which tolerance and cooperation are measured because social interactions are necessarily operationalized through gendered categories and all self-replicating communities necessarily include women as well as men.
Austronesian Women and Seventeenth-Century Marital Networks
Written records suggest that southwestern Taiwan was a cosmopolitan place during the mid-seventeenth century, and historians have examined cooperation as well as tensions across European, Austronesian, and Han ethnolinguistic boundaries. But Austronesian women—who are mentioned repeatedly in the Dutch records in particular—receive little credit. I argue that Austronesian women made essential contributions to interethnic cooperation. Moreover, I suggest that Austronesian women’s public visibility laid the social and cultural foundation for later generations of women to work in public.
When the Dutch arrived during the early seventeenth century, Taiwan’s indigenous peoples spoke dozens of distinct, mutually unintelligible Austronesian languages. Trade and colonization brought at least a half dozen more languages. From the 1620s through the 1660s, the Dutch East India Company (Vereenigde Oost-Indische Compagnie, hereafter VOC) worked to consolidate colonial control over Taiwan. Although the Dutch in Taiwan never numbered more than twelve hundred, they suppressed endemic Austronesian warfare on the plains, first controlling and later relying on military assistance from the Siraya and other Austronesians living nearest to the Dutch forts (in present-day Tainan). Missionaries converted Austronesians to Christianity, overthrowing the indigenous woman-led religious practices and pushing newlyweds to settle in their own (neolocal) home and start raising children immediately. These colonial efforts led to increases in the Siraya population overall and in the number of women available for marriage during the 1640s and 1650s, a crucial Han immigration period (Brown 2004, 136–153).
In 1650, the plains Aborigine population was about twenty-six thousand. The Han population grew from a few hundred in 1622 to several thousand in 1640, fifteen thousand in 1650, and an estimated twenty-five thousand men by 1661. Han women were rare, certainly less than 10 percent of the population; European women were even rarer. Yet Dutch sources report that many Han and European men married and had families. By 1650, almost half the population under Dutch control was either of mixed Han-Austronesian ancestry or living in mixed households (Brown 2004, 141). The male-biased population sex ratio meant that Austronesian women were necessary to local social dynamics, but it did not in itself give women power.
Women were important members of Siraya communities in their own right. Gender division of labor upon Dutch contact made farming—swidden, dry-field agriculture—the province of women, and their matrilineality (calculating inheritance through one’s mother), meant that Siraya women owned the agricultural fields. Young men left their mothers’ homes to live collectively in a men’s house, moving through a militarized age-grade system that focused on warfare and hunting; Siraya men only moved to their wives’ homes to do agricultural work and raise children after retirement, when they were in their forties. Siraya women chose their own husbands and shocked both Dutch and Chinese with their willingness to engage in sexual activity, whether married or not. Perhaps their attitude was due to the fact that they had access to a reliable means of fertility control that preserved not only the life of the woman but her future fertility. Austronesian women’s local agricultural expertise and kin networks were probably viewed positively by Han men seeking marriage partners. Austronesian women’s presence in publicly visible spaces was striking to seventeenth-century Chinese observers—no doubt for their flagrant disregard of neo-Confucian ideals. Not only did Austronesian women work in agricultural fields with men but they also danced publicly during festivals (figures 1 and 2).
Also crucial to plains Aborigine women’s social importance was the fact that both Dutch and Han invested in learning to speak Siraya and other Austronesian languages. The VOC economically and politically supported missionaries and schoolteachers in many Austronesian villages to learn the local languages as a means of both proselytization and political control. This continued investment despite clear tensions throughout the Dutch records between VOC officials and missionaries suggests how fundamental Austronesian peoples were to the Dutch colony—for the deer trade and for military control of the plains. Many Han learned Austronesian languages sufficiently to become middlemen—collecting the tribal tax imposed by the Dutch and holding VOC-licensed monopoly rights to purchase specified goods, such as deerskins or sugarcane. The Han and European men who married Austronesian women enhanced bilingualism, for husband and wife and for their children.
The network of marital ties in Dutch Taiwan had implications for access to resources and implementation of policies. Many Han middlemen had plains Aborigine wives. These middlemen purchased from not only Austronesians but also any Han agriculturalists living in their licensed territories, many of whom also had Austronesian wives. Moreover,
it was not unusual for [VOC] Company employees, especially schoolteachers who as a rule were living in the midst of their pupils, to marry a converted [Christian] village girl.… [T]he fact that Governor Coyett mentions [that German schoolteacher Hendrick Noorden’s wife is a “native woman” when recording Noorden’s 1657 promotion to political administrator over the southern territories under Dutch control] leads us to suspect that Noorden’s wife, [baptized as] Maria, somehow played an active part in the network of relations between the Company personnel and the indigenous population. Was she also the one who taught her husband to speak the language of her people so well? (Everts and Milde 2003, 245)
Remembering Han intermarriage in these same villages, it is not so surprising that—despite orders from their VOC superiors in Batavia to remove all Chinese from Austronesian villages—in 1642, the Taiwan VOC allowed Han to remain in six communities near the Dutch stronghold as well as in the village of Favorlang (some twenty-five miles north, in central Taiwan), “because the Dutch civil administration stationed in these villages could keep an eye on their activities” (Heyns 2003, 182). The Batavia VOC did not accept these accommodations, and further movement occurred in 1644. The Dutch administrators and the Han men living in the villages were married into the same matrilineal Austronesian network—they were marital kin—something the Taiwan VOC knew well but probably thought better of telling their superiors in Batavia. Marital ties were always local and thus capable of micro-level feedback to larger social pressures. Consequently, Han-Austronesian intermarriage could cause trouble for the Dutch, as these ties in Favorlang did from about 1635 until 1645 or so. “Favorlang Chinese”— married to local Austronesian women and with their own ties to seafaring Han—diverted deer and crops to Han smugglers, circumventing VOC licensing and fees; Favorlang Chinese attacked “Dutch Chinese,” with VOC licenses, and encouraged Austronesian resistance to Dutch control (Andrade 2005).
Marital ties gave Austronesians access to both Dutch and Han networks. VOC Governor Van Nuyts, who bungled diplomacy with the Tokugawa shogunate and left his position in disgrace, was criticized for a long-term liaison with an Austronesian woman (he also had a wife in the Netherlands) (Blussé 2003; Clulow 2014). Such matrilineal connections—to the VOC administration, Dutch and German missionaries, Han middlemen, and Han agriculturalists—may explain how some plains Aborigines became registered landlords, sponsoring the clearing of forests and the establishment of paddy fields by Han in exchange for tenancy rights; the Taiwan VOC allowed and even recorded several instances of such registration, which the Batavia VOC later rejected (Heyns 2003). Plains Aborigine marital ties, however, appear not to have extended to Han merchants in town near the Dutch forts, for these men were wealthy enough to bring wives from China (or elsewhere).
Opportunities for Taiwan residents to thrive existed, but they required the ability to engage networks across ethnolinguistic differences. Taiwan’s natural resource wealth lay not only in existing deer herds (hunted to extinction by about 1730) but also in the land’s lush potential for agriculture (and later in sulfur and camphor extraction). Han merchants and the VOC recruited Han men to migrate from China (mostly Fujian) and both Han and Dutch—and later Austronesians—sponsored Han laborers and farmers to clear forestland for commercial production of rice (in paddy fields) and sugarcane (in dry fields). Interethnic cooperation was necessary to identify appropriate land; ensure peaceful access; organize and supply work teams; plant, protect, harvest, and process crops; pay VOC duties; and arrange shipping to the global market. In this context, the identity politics of the day hinged on taxation, trade, and language, all of which marital ties could potentially mediate.
The documented practices of Austronesian women must have disrupted Han and Dutch gendered notions of public space and roles, but neither Dutch nor Han colonizers could prescribe all women’s public activities (figures 1 and 2). Surely the demography—a population sex-ratio imbalance, much larger proportions of Austronesians to Europeans, and, initially at least, larger proportions of Austronesians to Han—combined with knowledge of the local environment and regional customs to give Austronesian women leverage to maintain many public practices, especially those related to marital networks and resource management. It is difficult to say precisely what women’s status was, relative to that of men during the Dutch period, given that the records we have were written by men (European and Chinese) from patriarchal societies. It seems likely that the status of women declined—not only from the Dutch measures to disempower Siraya women who were religious leaders, but also as the proportion of Han men in the local population grew. However, because the skewed demographic context obtained through both the Zheng and early Qing periods, the visible participation of women in the public sphere very likely continued for more than a century. I suggest that this lengthy visibility established a historically important social foundation, one that may have survived in cultural memory.
Bendi Women Processing Tea, 1860–1910
Qing Taiwan (1683–1895) was by all accounts distinctly not cosmopolitan—notorious as it was for ethnic and rebellious violence. Notably, cosmopolitanism ebbed and then reemerged in correlation with women’s quotidian visibility in public. Because it is easier to see the connection of women’s visibility to cosmopolitanism with their resurgence than with their diminution, I focus here on women’s role in the global tea trade and its implications.
Qing administrative management has been criticized as chronically understaffed, thus failing to prevent Han settlers’ encroachment on tax-paying Austronesians’ land rights and fomenting the growth of Han strongmen with ethnic-based militia (Meskill 1979; Shepherd 1993; C. Chen 1999), but it could be interpreted instead as a deliberate strategy of settler colonialism (cf. Herman 2018). After 1788, when the ban on women settlers was permanently lifted, most Han settlers could marry and work endogamously, even within county-level native-place groups (table 1), creating few engagements across ethnolinguistic distinctions (Brown 2004, 156). Faced with 159 major uprisings and feuds between 1684 and 1895, Qing officials exploited ethnic tensions, supplementing troops with rival militias (Lamley 1981; Harrell 1990; Shepherd 1993; C. Chen 1999). By 1860, cosmopolitanism reached a nadir.
After 1860, Taiwan’s early nineteenth-century insular focus on the eco-engineering and ethnic competition of settler colonialism shifted to global markets. European unequal treaties opened Taiwan’s ports to international trade (figure 3). After 1878, 20 percent of Taiwan’s population was involved in export production of tea, sugar, or camphor (Ka 1995, 35). By 1881, tea had surpassed sugar to become Taiwan’s leading global export (Gardella 1999, 167). Global trade interests necessitated cooperative interactions for resource extraction, refinement, and shipping. Incursions into the forested territories of mountain Austronesians to cut trees, set up camphor stills, and, later, plant tea increased. Qing official oversight of developing the tea and camphor trade lay with the Settler’s Annexation Bureau (Fuken ju 撫墾局)—an office that profited Han militia leaders. In 1888, for example, the Lins of Wufeng received a land grant of several thousand hectares with virgin camphor stands in the hills near Zhanghua, after Lin militia suppressed three mountain Aborigine uprisings (Meskill 1979, 186–189, 191, 240–242; Gardella 1999, 181). Were these uprisings strictly about resource exploitation, or were they like the 1731 uprising in nearby Dajiaxi 大甲西, where Han abuse of Austronesian women doing corvée labor work (Shepherd 1999, 117)—that is, public-sphere labor—led to violence? Similar to Dutch-era strategies, some Han (and later Japanese) men married mountain Aborigine women sharing profits with their in-laws, who then supported these resource-extraction practices to other Austronesians (e.g., P. Barclay 2005). When such mediation failed, as it did at least nine times between 1875 and 1892, Qing authorities used plains Aborigine militia to quell the armed resistance.
During their rule (1895–1945), Japanese colonizers’ immediate encounters with armed resistance, disease, and poor infrastructure led to substantial investment in public-health measures, a policing system, railways, and roads (Ts’ai 2009). Cognizant of Qing-era feuding, officials initially tracked Hoklo, Hakka, and plains Aborigine identity, but after 1915—with armed resistance by Han and plains Aborigines ended and ongoing Austronesian resistance contained in the mountains—only distinctions between Japanese, mountain Aborigines, and an amalgamated “locals” category, which I refer to as “Bendi,” remained salient (Brown 2004, 2010) (see table 1). Focusing on exports to offset Japan’s trade imbalances, the colonial government monopolized opium (1897), camphor and salt (1899), sugar (1905), and tobacco (1906), thereby eliminating Western involvement in these trades and substantially increasing revenues (Ka 1995, 52–55). To entice Japanese capitalists to invest in sugar production, colonial administrations invested in infrastructure (dams and irrigation systems, industrial mills), technology (seedlings, fertilizer, crushing machines), even direct cash subsidies, and finally monopoly zones. Before Japanese capitalists arrived in 1906, sugar production declined: the amount produced in 1900 was only 63 percent of what had been produced in 1895 (Ka 1995, 75). Tea continued to be an important export through 1910.
The post-1860 tea trade changed the face of northern Taiwan, not only deforesting the hills for tea plantations and making towns into tea-processing hubs but also bringing many Bendi women into the public-sphere side of wage labor. Han women’s economic contributions to their households have long been legion—as well as underestimated and undervalued—but in rural areas, such contributions, often in the form of homemade handicrafts and textiles, did not previously necessitate leaving the household (Brown 2016, 2017; Brown and Satterthwaite-Phillips 2018; cf. Bray 1997). As tea became Qing-era Taiwan’s leading export, Westerners built warehouses and hired Han middlemen to deal with cultivation and processing. Americans and Europeans retreated to handling export shipping when their former compradors outcompeted them at the local level. The river port of Dadaocheng became a boom town where tea harvested in the hills around the Taipei basin was brought for sorting, processing, and shipment to the coastal port of Danshui (figure 3). Moreover, the tea industry apparently smoothed ethnic tensions among the various local Han.
The post-1860 demand for female labor ran from the hills to the river ports because picking tea (leaf by leaf), sorting, and processing were all traditionally women’s work (e.g., Chien 2015), though men also processed tea (figure 4). Women could earn good wages from tea—enough to support themselves and another person or two—and by 1900, thousands were doing so in Dadaocheng on a daily basis (Davidson  1988, 385–386; A. Wolf and Gates 2005, 121). This economic organization has major implications for not only household economics and power dynamics but also communitarian cosmopolitanism. Japanese-period demographic records show that almost all Han women in towns (91 percent) and rural areas (98 percent) of Taiwan married, but in Taipei city almost 23 percent of women never married (A. Wolf and Gates 2005, 119). This significantly lower marriage rate may be due to urban life generally (A. Wolf and Gates 2005, 125, 128), but not all urban wage labor brought women into common public spaces with the same frequency.
The wealthiest families claimed achievement of the Han (purportedly neo-Confucian) patriarchal ideal of keeping women cloistered within their household, except for occasional visits to temples. The daily work of women in non-elite households surely took them frequently outside the house—to stores or markets (whether for buying supplies or selling homemade handicrafts). The employment supported by the sex trade in Mengjia—not only prostitutes and brothel owners, but also seamstresses and silver platers—was carried out largely behind closed doors. By contrast, the tea industry brought local women into Taiwan’s public sphere to such an extraordinary degree that it was noteworthy to contemporaneous observers. James Davidson ( 1988, 385), an American who reported on Japanese colonization first as a journalist and later a diplomat, wrote, “During the summer months, nothing is more striking than the crowds of girls”—he estimated a daily average of more than twelve thousand—“who at noon and night simply overrun [Dadaocheng]” (cf. Lin 2001, 987).
The growth of the tea industry correlates to the broad acceptance of Bendi women’s highly visible expansion in public-sphere wage labor. The global tea trade pervaded northern Taiwan society during the late Qing period, extending into smaller towns and the rural hinterland. The socioeconomic and ecological shifts of the tea industry are, like other socioeconomic and ecological shifts, not simply top-down, externally driven processes. Rather, Bendi women and their families interacted and engaged with the larger political-economic (colonial and imperialist) forces that confronted them, and their micro-level agency—the niche-constructing actions of individuals and households—reconstituted these macro-level processes (cf. Brown 2016, esp. 519–527). More and more households—including its female members—thus had quotidian public engagements with strangers and people of different ethnic groups that are so fundamental to communitarian cosmopolitanism.
Japanese colonial policies continued to move Bendi women into public-sphere labor. In 1915, the Japanese government added its ban on footbinding to the list of items to be checked during semiannual police visits—when it understood that women with bound feet would not work in the paddy fields (Brown 2004, 95, 265n45). Implementation of the ban, as intended, expanded women’s participation in agriculture, moved men into off-farm wage labor, and resulted in more women working off farms as well. Other policies helped keep women in the public-sphere workforce. For example, in the tobacco monopoly, women were granted pregnancy leave and the right to continue working after marriage (Brown 2010, 470). Bendi women were visible in public-sphere employment throughout the Japanese period—in agricultural production (of rice, tea, and even sugarcane), manufacturing, clerical work, education, and medicine. Although explicitly aimed at boosting productivity, this visibility also affected civil society—not only expanding women’s quotidian experience in the public sphere across classes and regions of Taiwan well beyond what the nineteenth-century tea industry had produced, but also promoting such participation as both proper and “modern” (cf. Wang W. 2014).
Bendi Women’s Late Twentieth-Century Living Rooms as Factories
As we come closer to the present day, more information is available about women’s thoughts and understandings of their relationship to work, the public sphere, community, and even cosmopolitanism. I continue to examine evidence of communitarian cosmopolitanism in terms of the quotidian engagements of most women—that is, women who were Bendi, not elite, not highly educated, and (during the immediate postwar period) predominantly rural—not merely to maintain comparability to earlier centuries lacking this information, but also to avoid the predominance of elite women’s views even during the twentieth century (e.g., Ku 1988, 1989; Chiang 1999; Rubinstein 2004; Lewis 2009; Wang W. 2014).
The postwar transition to Nationalist (Kuomintang 國民黨, KMT) rule in Taiwan stymied cooperation across diversity. As the KMT increasingly lost ground in the Chinese Civil War on the mainland (1945–1949), it viewed control of Taiwan as more crucial to its survival, and Bendi viewed being drawn into the lost cause with greater alarm. The KMT was unable to administer Taiwan competently or fairly. For example, there were outbreaks of malaria when the KMT loosened Japanese public-health measures and also massive confiscations of Bendi property by KMT officials and soldiers at all levels (e.g., G. Barclay 1954; Gates 1987). These failures exacerbated the huge economic challenges of rebuilding Taiwanese infrastructure and economic institutions damaged by the war as well as the challenges of resettling, feeding, and employing more than a million Mainlander refugees. Attitudes quickly polarized, and the rapid deterioration to violence galvanized as ethnic the sociopolitical distinction between Bendi (bensheng ren) and Mainlanders (waisheng ren) (table 1; Brown 2004, 2008, 2010). Bendi rebelled beginning on February 28, 1947, and after KMT troop reinforcements arrived on March 8, Mainlanders suppressed Bendi-armed resistance as brutally as the Japanese had decades earlier, extinguishing not only the rebellion but all expressions of dissent. Conservative estimates suggest that tens of thousands of Taiwanese lost their lives or were jailed, and many more were blacklisted (T. Lai, Myers, and Wu 1991). But the firm martial-law grip on Taiwan society did not stop the violence so antithetical to cosmopolitanism. In the aftermath of its decisive civil war loss, the KMT expanded its authoritarian grip on the peoples in Taiwan during the early 1950s with the White Terror (Baise kongbu白色恐怖) purges of KMT ranks, Mainlander refugees, and some Taiwanese. Neighbors informed on neighbors and people disappeared without explanation, their fates unknown sometimes for decades. In such a context, in which one cannot trust neighbors, engagement across diversity is all but impossible. Cosmopolitanism was not merely at a nadir; it was entirely extinguished.
Taiwan’s economic development from 1950 through 1987 has been hailed as miraculous (e.g., Gold 1986; Rubinstein 1999). But we need to recognize this development as colonial in order to understand its social impacts—on the communitarian social engagements necessary for local cosmopolitanism and on transnational engagements fostering global cosmopolitanism—and how these influences were gendered. Taiwan’s economic miracle occurred under authoritarian (martial-law) rule by a Mainlander-dominated KMT that sought to use Taiwan’s population, infrastructure, and natural resources to serve Mainlander goals and needs. This Mainlander colonialism was neither the full-on settler colonialism of the Qing period, which displaced indigenous Austronesians and transformed ecosystems across Taiwan, nor the imperial colonial model of the Japanese period, which largely limited the long-term migrants from Japan to the number needed to extract desired resources. It took time for the KMT to reorient to the permanency of their settlement in Taiwan. The one to two million Mainlander officials, troops, and refugees were outnumbered by the six million people who had been Japan’s colonials, but with Mainlanders representing 20 to 25 percent of the resulting population, the discrimination that favored Mainlanders in almost every sphere was widely experienced. Moreover, Mainlander policies recognized little of the ethnic variation within Taiwan, lumping together as Bendi all but mountain Austronesians, inadvertently promoting interethnic cooperation among Bendi, and contributing to the social experience of a distinct Taiwanese identity (Brown 2004, 2008, 2010).
It was not Nationalist policies that first brought women into Taiwan’s public sphere, though it may have seemed that way to planners unaware of Taiwan’s history. Although women retreated from outside-the-household wage employment during the inflationary and incendiary context of the Cold War, White Terror period, Taiwanese women had been visible in the public sphere during the Japanese period—in agricultural production (of rice, tea, and even sugarcane), manufacturing, clerical work, education, and medicine. Although little scholarship addresses women’s labor-force participation in Taiwan from 1945 to 1960, there must have been strong pressures—including widespread rhetoric about women serving the nation primarily by being good wives and wise mothers—pushing women, Bendi and Mainlander alike, out of the workforce. But KMT policies did rely crucially on women’s labor for economic development. And from 1956 to 1966, women in the labor force rose from 20 to 34 percent; by 1973, it was 40 percent; and by 1986, 45 percent (Greenhalgh 1985, 273; Gallin 1989, 374). Their contributions on farms and in manufacturing are undervalued (but see Greenhalgh 1985, 272–73) at least in part because they came mostly from ordinary, not-highly-educated Bendi women.
Beginning around 1950, Mainlander policies moved men into off-farm wage labor and feminized agriculture. Land reform, carried out in the midst of the White Terror purges, first (in 1949) reduced rents then (in 1951) put ownership of small plots of land into the hands of the families who actually farmed it, something for which I heard farmers still expressing gratitude in the early 1990s. Landholding limits shifted Bendi elites’ investments to industry and also meant that farming families needed additional off-farm wage income to make ends meet (e.g., Gates 1987; P. Wang 1999, 324–325). Because these policies effectively feminized agricultural labor, the rice-crop yield increases (of 37 percent!) that occurred over the decade following land reform (Rigger 2014, 47) have to be understood as relying not only on the land-to-the-tiller policy and green-revolution technology but also on Bendi women’s labor.
Around 1960, rural Bendi women made visible contributions to their households, but it did not bring most of them high status or power in household dynamics (e.g., Harrell  2015; Lu Y. 1984; however, rural women’s status has changed, e.g., Lai E. 1997). Although families readily made women the managers of household finances, when they had the best skills (Cohen 1976), women were still largely dependent on the loyalty of their sons to gain substantial authority (M. Wolf 1968, 1972; A. Wolf and Huang 1980). The commonly expected pattern was for unmarried daughters to help with farmwork, housework, and childcare through their years of compulsory education. Then, between completion of schooling and marriage, they would work where there was the greatest familial need or the highest available pay. A well-paying job had long been recognized as a way to stave off an unwanted marriage (e.g., A. Wolf and Huang, 1980).
In the 1960s, national-level economic plans to shift from import substitution to export-oriented manufacturing drew unmarried young women, mostly Bendi and many rural, into factory work, leaving agriculture to married women. The majority of never-married women (age 15 and older) participated in wage labor: 57.3 percent in 1967, 62.3 percent in 1973, 58.5 percent in 1984, and 54.6 percent in 1988 (Hsiung 1996, 40). The participation of teenage girls (15–19 years old) was particularly high during the 1970s, rising from 34.9 percent in 1970, to 55.2 percent in 1975, and 67.9 percent in 1980 (Hsiung 1996, 38; cf. Harrell 1981, 36). These jobs constituted hard, exhausting work (e.g., Kung 1976, 1983; Hsiung 1996). And although newspapers, television, and textbooks suggested that such work was a temporary phase of women’s lives, to be ended by marriage (Harrell  2015, 63; Greenhalgh 1985, 272), the fact that many women pursued, even paid on their own for, education and training beyond the minimum necessary for their jobs (Kung 1976, 1983; Greenhalgh 1985, 299–300) indicates women’s recognition that their work lives would extend beyond marriage. This pursuit demonstrates their agency in seeking to improve the circumstances of that future employment.
Among those working “out,” the ability to go home varied by region, factory, and time period. In the early 1960s, there were few large-scale factories, and they were primarily in urban areas, so although some workers might live at home, most lived in dormitories—returning home weekly perhaps, if they lived within an easy bus distance, but only a few times a year if their homes were more distant (Kung 1976, 39; Greenhalgh 1985, 294). Starting in 1984, workers were legally entitled to four Sundays off each month, which large factories with multiple shifts honored, but smaller factories often required twenty-eight days of work per month (Hsiung 1996, 113, 119; Gallin 1989, 380). In 1987, when time off commonly began Saturday afternoon, I saw bus and train stations regularly congested with workers going home as well as an entire street of living-room factories with workers assembling light bulbs through Saturday evenings. Workers might take extra days off during harvests, but they risked being laid off or fined (Kung 1976, 54–55n14; Hsiung 1996, 115–118). In 1991, when many workers had motorcycles, I saw one small rural village’s population expand considerably on weekends with returning workers.
Thus, the feminization of agriculture did not just affect farmwork. Bendi rural women ran households and conducted whatever business could not wait for returning workers—engaging daily with marketplace and itinerant vendors, schoolteachers, and postal workers, in addition to making the many routine decisions required in the daily operation of a farm and household (e.g., Harrell  2015). Bendi women had no choice but to be active social participants in rural communities once most rural men had found off-farm wage employment.
Such local, rural engagements do not constitute the kind of social interactions that we usually consider cosmopolitan—they are communitarian engagements, but we do not usually think of cross-gender interactions as representing diversity. Nevertheless, we can see that making the village-level public sphere a place where most local women routinely enact communitarian social engagements while creating household connections to a larger regional (or even national) public-sphere wage-labor workforce creates at least the potential for communitarian cosmopolitanism writ large. As these daily interactions expand—with industrialized transportation, for example—they cross lines of emically salient difference. Taiwan’s economic miracle built further on that potential.
The 1968 Living Rooms as Factories (Keting ji gongchang 客廳即工廠) program explicitly targeted married women, urban as well as rural, whom planners blindly considered “idle,” as surplus labor (Hsiung 1996, 52). There may have also been a presumption that Bendi women were docile, easy “victims” of exploitation (cf. Greenhalgh 1985, 301, 303; Gallin 2012: 163–164). But the government took no chances—strikes and labor protests were illegal under Mainlander colonial rule, and labor-protection laws, including the right to unionize, did not apply to factories with fewer than thirty workers (Gallin 1989, 381; Hsiung 1996, esp. chap. 2)—that is, factories small enough to operate out of living rooms. Married women’s participation in Taiwan’s labor force grew from 27.2 percent in 1967 to 35.3 percent in 1973, 39 percent in 1984, and 42.7 percent in 1987 (Hsiung 1996, 40). By 1987, 56.5 percent of married women—in contrast to 34.3 percent of single women—working in manufacturing were employed at satellite factories with fewer than thirty workers (Hsiung 1996, 77; cf. Harrell 1981, 36–38; Harrell  2015, 66–71; Gallin 1989, 379–381). This “living-room” piecework system continued through the 1990s, even after the end of colonial rule (Hsiung 1996; Skoggard 1996; Simon 2005).
Government statistics undercount married women’s labor in satellite factories because they undercount female family members working without pay in family-run living-room factories, erroneously counting them as housewives (Hsiung 1996, 43) and considering their labor to be merely “helping” with work credited to the household head. Although men usually held legal title to a living-room operation, women regularly acted as proprietors—setting the production pace by their own work or interspersing workers paid by the piece among workers paid a daily wage to ensure a faster pace (Hsiung 1996, 104–107, 121–123). Women proprietors also conducted necessary “outside” interactions—for example, with factory representatives to negotiate schedules, orders, and wages (Harrell 1981, 37–38; Harrell  2015, 67) (see figure 5). This agency does not deny that women faced discrimination. There was a clear gender gap in pay; for example, women’s wages were lower than those of men in the same position, and women who belonged to the satellite factory owner’s household often were not paid wages at all (Hsiung 1996, 41, 42–43, 45, 104). But Bendi women’s labor-force participation expanded women’s experience and social engagement and made living rooms into common public spaces.
It was not only proprietors who became cosmopolitan agents by expanding their social-engagement skills. Wage workers also had to learn these skills, not least because women in general (and especially older, married women) were perceived as “docile” (Simon 2005, 106–110; cf. Greenhalgh 1985, 305). Whether satellite employees were owners’ kin or not, they had to learn to negotiate their work as waged or unwaged (often within the idiom of “helping”), and the implications for their status in the household, where non-kin or extended family might board (Hsiung 1996, 131–132). Women workers were hardly docile (cf. Simon 2005, 109). They learned to “wrangle”—use barbed conversation, sometimes cooperatively, to improve their pay, resist exploitative requests, or embarrass the boss (Hsiung 1996, 134–35). Difficult work conditions contributed to women’s outspokenness and, in a favorable labor market, they were sometimes willing to risk being fired (Gallin 1989, 380; cf. Kung 1976, 48). These kinds of social-engagement skills, learned in wage labor, are at least marginally cosmopolitan in their own right because they require sufficient cooperation to result in a marketable product. In addition, these engagements set up a potential for greater cosmopolitanism if workers use these skills elsewhere.
Non-elite, mostly Bendi women simultaneously made significant economic contributions and became cosmopolitan actors by earning wages in common public spaces—and by creating common public spaces in living rooms. Bendi women managing farms, working “out” in their youth, and bringing factories into their living rooms contributed importantly, albeit on a very local scale in terms of the individuals involved, to a much larger communitarian engagement that enabled Taiwan’s postwar economic and cosmopolitan development.
One extraordinary example of Bendi women using social-engagement skills to foster a larger communitarian cosmopolitanism is the Tzu Chi Foundation (Ciji jijinghui 慈濟基金會, Buddhist compassion relief foundation). This now-international philanthropic organization was founded in 1966 by Buddhist Dharma Master Cheng Yen 證嚴, a Bendi woman, with a membership of thirty “housewives” in Hualian. Tzu Chi reports its origins as fundamentally cosmopolitan: an encounter between Master Cheng Yen and some Catholic nuns who accepted Buddhism’s high principles but challenged her to recount organized Buddhist compassion for the poor and needy, and the death of an Austronesian woman who was turned away from a hospital in eastern Taiwan because of her inability to pay for care. Tzu Chi provides aid and medical care across differences of all kinds. It grew to eight thousand members by 1986, four million by 1994, and it now claims more than ten million members worldwide in fifty countries, operating a university, multiple hospitals and clinics, disaster-relief teams, and environmental protection and recycling centers (http://www.tzuchi.org/). Its membership remains primarily women (about 70 percent), and its disaster relief and other volunteer work bring ordinary Taiwanese women into the public sphere in ways both large and small, even as it presents its work as gender-neutral (Weller 1999; H. Lu 2016). This organization illustrates how Taiwan’s women—especially Bendi—have been at the heart, not only of Taiwan’s postwar economic development and concurrent globalization, but also of cosmopolitanizing these processes to link local communities across differences and also across national borders.
Across these three examples, we see women—considered local or indigenous by the globally connected economic planners—working actively in public common spaces in ways that foster communitarian ties. These are not the only possible examples of indigenous women’s contributions to cosmopolitanism in Taiwan, but they capture three critical periods and illustrate how women’s ties constitute linkages in a social structure made dynamic at least as much by the aggregated micro-level agency of these local women as by the macro-level colonial projects of Dutch, Qing, Japanese, and Mainlander authorities. Austronesian women’s marital networks facilitated seventeenth-century economic and military cooperation. Bendi women’s public-sphere work processing tea fueled the late nineteenth-century’s largest export and expanded into other, “modern” export production through 1945. Bendi women’s farm management and manufacturing work—even making living rooms into factories—powered Taiwan’s late twentieth-century economic miracle. Interacting with colonial forces, indigenous women recurrently fostered local communitarian, cooperative ties.
As a social niche—an inherited social structure that individuals themselves modify by their actions—communitarian cosmopolitanism requires people working across diversity in ways that create and maintain community. It is fragile, because ties across difference require effort—practice—in order to overcome the social discomfort, or sometimes even moral disapproval, that may come with unexpected customs or unexpected responses in cross-cutting social interactions.
But the historically contingent processes that both maintain and modify social systems grant large impacts to highly visible precedents. The visibility of Austronesian women in Taiwan’s public sphere probably lasted until at least 1788—and perhaps much longer if nineteenth-century Austronesian uprisings, which were surely widely discussed, were also associated with Austronesian women’s public-sphere labor. Bendi women’s post-1860 visibility began in northern Taiwan with tea processing but expanded throughout Taiwan during Japanese rule. That highly visible public-sphere labor expansion was within living memory of postwar Mainlander demands on Bendi women’s contributions to economic development. These illustrations of Taiwan’s intersectional cosmopolitanism—Austronesian and Bendi women’s repeated contributions to communitarian cosmopolitanism in Taiwan—provide fertile grounds for further empirical and theoretical research.
Anderson, Benedict. 1983. Imagined Communities: Reflections on the Origins and Spread of Nationalism. London: Verso.
Andrade, Tonio. 2005. “Pirates, Pelts, and Promises: The Sino-Dutch Colony of Seventeenth-Century Taiwan and the Aboriginal Village of Favorolang.” Journal of Asian Studies 64 (2): 295–321.
———. 2007. How Taiwan Became Chinese: Dutch, Spanish, and Han Colonization in the Seventeenth Century. New York: Columbia University Press.
———. 2012. “Koxinga’s Conquest of Taiwan in Global History: Reflections on the Occasion of the 350th Anniversary.” Late Imperial China 33 (1): 122–140.
Appiah, Kwame Anthony. 2006. Cosmopolitanism: Ethics in a World of Strangers. New York: W. W. Norton.
Arendt, Hannah.  1998. The Human Condition. 2nd ed. Chicago, IL: University of Chicago Press.
Barnard, Alan. 2006. “Kalahari Revisionism, Vienna and the ‘Indigenous Peoples’ Debate.” Social Anthropology 14 (1): 1–16.
Barclay, George W. 1954. Colonial Development and Population in Taiwan. Princeton, NJ: Princeton University Press.
Barclay, Paul D. 2005. “Cultural Brokerage and Interethnic Marriage in Colonial Taiwan: Japanese Subalterns and Their Aborigine Wives, 1895–1930.” Journal of Asian Studies 64 (2): 323–360.
Blussé, Leonard. 1984. “Dutch Protestant Missionaries as Protagonists of the Territorial Expansion of the VOC on Formosa.” In Conversion, Competition and Conflict, edited by Dick Kooiman, Otto van den Muijzenberg, and Peter van der Veer, 155–184. Amsterdam: Free University Press.
———. 1990. “Minnan-jen or Cosmopolitan? The Rise of Chen Chih-lung alias Nicolas Iquan.” In Development and Declines of Fukien Province in the Seventeenth and Eighteenth Centuries, edited E. B. Vermeer, 245–264. Leiden: E. J. Brill.
———. 2003. “Bull in a China Shop: Pieter Nuyts in China and Japan (1627–1636).” In Around and About Formosa: Essays in Honor of Professor Ts’ao Yung-ho, edited by Leonard Blussé, 95–110. Taipei: Ts’ao Yung-ho Foundation for Culture and Education.
Blussé, Leonard, and Marius P. H. Roessingh.1984. “A Visit to the Past: Soulang, a Formosan Village Anno 1623.” Archipel 27: 63–80.
Bray, Francesca. 1997. Technology and Gender: Fabrics of Power in Late Imperial China. Berkeley: University of California Press.
Brown, Melissa J. 1996. “On Becoming Chinese.” In Negotiating Ethnicities in China and Taiwan, edited by Melissa J. Brown, 37–74. Berkeley: Institute for East Asian Studies, University of California.
———. 2004. Is Taiwan Chinese? The Impact of Culture, Power, and Migration on Changing Identities. Berkeley: University of California Press.
———. 2008. “When Culture Does Not Affect Behavior: The Structural Basis of Ethnic Identity.” In Explaining Culture Scientifically, edited by Melissa J. Brown, 162–183. Seattle: University of Washington Press.
———. 2010. “Changing Authentic Identities: Evidence from Taiwan and China.” Journal of the Royal Anthropological Institute 16: 459–479.
———. 2016. “Footbinding, Industrialization, and Evolutionary Explanation: An Empirical Illustration of Niche Construction and Social Inheritance.” Human Nature 27: 501–532.
———. 2017. “Dutiful Help: Masking Rural Women’s Economic Contributions.” In Transforming Patriarchy: Chinese Families in the Twenty-First Century, edited by Gonçalo Santos and Stevan Harrell, 39–58. Seattle: University of Washington Press.
———. 2020. “Tigers on the Mountain: Assessing Is Taiwan Chinese? in 2018.” In Taiwan Studies Revisited, edited by Dafydd Fell and Hsin-Huang Michael Hsiao, 112–140. London: Routledge.
Brown, Melissa J., and Marcus W. Feldman. 2009. “Sociocultural Epistasis and Cultural Exaptation in Footbinding, Marriage Form, and Religious Practices in Early 20th-Century Taiwan.” Proceedings of the National Academy of Sciences USA 106 (52): 22139–22144. https://doi.org/10.1073/pnas.0907520106.
Brown, Melissa J., and Damian Sattherthwaite-Phillips. 2018. “Economic Correlates of Footbinding: Implications for the Importance of Chinese Daughters’ Labor.” PLoS ONE 13 (9): e0201337. https://doi.org/10.1371/journal.pone.0201337.
Campbell, William. 1896. “The Island of Formosa: Its Past and Future.” Scottish Geographical Magazine 12: facing p. 398; available from the Reed College digital archive, Formosa: Nineteenth-Century Images, https://rdc.reed.edu/c/formosa/s/r?_pp=20&s=8aacb88ffe734e9f95ef48bb65c9043aaf09b189&p=10&pp=1.
———. 1903. Formosa Under the Dutch. London: Kegan Paul.
Chan Su-chuan [Zhan Sujuan] 詹素娟 and Chang Su-bing [Zhang Subin] 張素玢. 2001. Taiwan yuanzhumin shi: Pingpuzu shi pian (bei) 臺灣原住民史: 平埔族史篇(北) [The history of Formosan Aborigines: Pepo tribes (northern)]. Nantou, Taiwan: Taiwan sheng wenxian hui.
Chen, Chiukun. 1999. “From Landlords to Local Strongmen: The Transformation of Local Elites in Mid-Ch’ing Taiwan, 1780–1862.” In Taiwan: A New History, edited by Murray A. Rubinstein, 133–162. Armonk, NY: M. E. Sharpe.
Chen, Shu-Juo. 2008. “How Han Are Taiwanese Han? Genetic Inference of Plains Indigenous Ancestry among Taiwanese Han and Its Implications for Taiwan Identity.” PhD diss., Stanford University.
Chiang, Lan-hung Nora [Jiang Lanhong] 姜蘭虹. 1999. “A Personal Thought on Women’s Studies in Taiwan.” Funü yu liangxing yanjiu tongxun 婦女與兩性研究通訊 [Bulletin of women and gender studies] 51: 21–28.
Chien, Mei-Ling. 2015. “Leisure, Work, and Constituted Everydayness: Mountain Songs of Hakka Women in Colonized Northern Taiwan (1930–1955).” Asian Ethnology 74 (1): 37–62.
Clulow, Adam. 2014. The Company and the Shogun: The Dutch Encounter with Tokugawa Japan. New York: Columbia University Press.
Cohen, Myron L. 1976. House United, House Divided: The Chinese Family in Taiwan. New York: Columbia University Press.
Davidson, James W.  1988. The Island of Formosa Past and Present. Taipei: Southern Materials Center.
Delanty, Gerard. 2012. “A Cosmopolitan Approach to the Explanation of Social Change: Social Mechanisms, Processes, Modernity.” Sociological Review 60 (2): 333–354.
d’Entreves, Maurizio Passerin. 2019. “Hannah Arendt.” In Stanford Encyclopedia of Philosophy (Spring 2019 edition), edited by Edward N. Zalta. https://plato.stanford.edu/archives/spr2019/entries/arendt.
Duara, Prasenjit. 2015. The Crisis of Global Modernity: Asian Traditions and a Sustainable Future. Cambridge: Cambridge University Press.
Evans, Harriet. 2017. “Patriarchal Investments: Expectations of Male Authority and Support in a Poor Beijing Neighborhood.” In Transforming Patriarchy: Chinese Families in the Twenty-First Century, edited by Gonçalo Santos and Stevan Harrell, 182–198. Seattle: University of Washington Press.
Everts, Natalie, and Wouter Milde. 2003. “‘We Thanked God for Submitting Us to Such Sore but Tolerable Trials’: Hendrick Noorden and His Long Road to Freedom.” In Around and About Formosa: Essays in Honor of Professor Ts’ao Yung-ho, edited by Leonard Blussé, 243–272. Taipei: Ts’ao Yung-ho Foundation for Culture and Education.
Fewkes, Jacqueline H. 2014. “Living in the Material World: Cosmopolitanism and Trade in Early Twentieth Century Ladakh.” In Sites of Asian Interaction: Ideas, Networks and Mobility, edited by Tim Harper and Sunil Amrith, 38–59. Cambridge: Cambridge University Press.
Fortunato, Laura, Clare Holden, and Ruth Mace. 2006. “From Bridewealth to Dowry? A Bayesian Estimation of Ancestral States of Marriage Transfers in Indo-European Groups.” Human Nature 17 (4): 355–376.
Gallin, Rita S. 1989. “Women and Work in Rural Taiwan: Building a Contextual Model Linking Employment and Health.” Journal of Health and Social Behavior 30: 374–385.
———. 2012. “Writing Women/Reading the Self: An ‘Itinerary of Thinking’ about Hsin Hsing Village, Taiwan.” Taiwan Journal of Anthropology 臺灣人類學刊 [Taiwan renlaixue kan] 10 (1): 159–173.
Gardella, Robert. 1999. “From Treaty Ports to Provincial Status, 1860–1894.” In Taiwan: A New History, edited by Murray A. Rubinstein, 163–200. Armonk, NY: M. E. Sharpe.
Gates, Hill. 1987. Chinese Working-Class Lives: Getting by in Taiwan. Ithaca, NY: Cornell University Press.
Gold, Thomas. 1986. State and Society in the Taiwan Miracle. Armonk, NY: M. E. Sharpe.
Greey, Edward (Sung-tie). 1871. “‘Tai-wan’—Formosa.” Frank Leslie’s Illustrated Newspaper 33 (834) (September 23): 28. Available from Reed College digital archive, Formosa: Nineteenth-Century Images. https://rdc.reed.edu/c/formosa/s/r?_pp=20&query=tea&s=5ded2bcd5b8394afbf0e4c2b755ac5bf70bcc1bf&p=1&pp=1.
Greenhalgh, Susan. 1985. “Sexual Stratification: The Other Side of ‘Growth with Equity’ in East Asia.” Population and Development Review 11 (2): 265–314.
Guenther, Mathias, Justin Kenrick, Adam Kuper, Evie Plaice, Trond Thuen, Patrick Wolfe, Werner Zips, and Alan Barnard. 2006. “Discussion” (of Alan Barnard’s “Kalahari Revisionism, Vienna and the ‘Indigenous Peoples’ Debate”). Social Anthropology 14 (1): 17–32.
Halperin, Rhoda H. 1998. Practicing Community: Class Culture and Power in an Urban Neighborhood. Austin: University of Texas Press.
Harrell, Stevan. 1981. “Effects of Economic Change on Two Taiwanese Villages.” Modern China 7 (1): 31–54.
———. 1990. “From Xiedou to Yijun: The Decline of Ethnicity in Northern Taiwan, 1885–1895.” Late Imperial China 11 (1): 99–127.
———.  2015. Ploughshare Village: Culture and Context in Taiwan. Seattle: University of Washington Press.
Herman, John E. 2018. “From Land Reclamation to Land Grab: Settler Colonialism in Southwest China, 1680–1735.” Harvard Journal of Asiatic Studies 78 (1): 91–133.
Heyns, Pol. 2003. “Land Rights in Dutch Formosa.” In Around and About Formosa: Essays in Honor of Professor Ts’ao Yung-ho, edited by Leonard Blussé, 175–207. Taipei: Ts’ao Yung-ho Foundation for Culture and Education.
Hsieh, Jolan. 2018. “The Changing Identities of Taiwan’s Plains Indigenous Peoples.” In Changing Taiwanese Identities, edited by J. Bruce Jacobs and Peter Kang, 12–26. Abingdon, UK: Routledge.
Hsiung, Ping-Chun. 1996. Living Rooms as Factories: Class, Gender, and the Satellite Factory System in Taiwan. Philadelphia, PA: Temple University Press.
Huang Shujing 黃叔璥.  1957. Taihai shicha lu 臺海使槎錄 [Record of a tour of duty in the Taiwan Strait]. Taiwan wenxian congkan 臺灣文獻叢刊 [Taiwan literature series] 4. Taipei: Taiwan Yinghang.
Huber, Johannes. 1990. “Chinese Settlers against the Dutch East India Company: The Rebellion Led by Kuo Huai-i on Taiwan in 1652.” In Development and Declines of Fukien Province in the Seventeenth and Eighteenth Centuries, edited by E. B. Vermeer, 265–296. Leiden: E. J. Brill.
Ka, Chih-ming. 1995. Japanese Colonialism in Taiwan: Land Tenure, Development, and Dependency, 1895–1945. Boulder, CO: Westview Press.
Ku, Yenlin. 1988. “The Changing Status of Women in Taiwan: A Conscious and Collective Struggle toward Equality.” Women’s Studies International Forum 11 (3): 179–186.
———. 1989. “The Feminist Movement in Taiwan, 1972–87.” Bulletin of Concerned Asian Scholars 21 (1): 12–22.
Kung, Lydia. 1976. “Factory Work and Women in Taiwan: Changes in Self-Image and Status.” Signs 2 (1): 35–58.
———. 1983. Factory Women in Taiwan. Ann Arbor, MI: UMI Research Press.
Lai Erh-Rou [Lai Errou] 賴爾柔. 1997. “Taiwan nongjia funü de jiating diwei yu juese” 台灣農家婦女的家庭地位與角色 [Family status and roles of farm women in Taiwan]. Nongye tueguang xuebao 農業推廣學報 [Review of agricultural extension science] 13: 55–81. https://doi.org/10.29788/RAES.199712.0003.
Lai, Tse-han, Ramon H. Myers, and O. Wu. 1991. A Tragic Beginning: The Taiwan Uprising of February 28, 1947. Stanford, CA: Stanford University Press.
Lamley, Harry J. 1981. “Subethnic Rivalry in the Chʼing Period.” In The Anthropology of Taiwanese Society, edited by Emily M. Ahern and Hill Gates, 282–318. Stanford, CA: Stanford University Press.
Lee Jen-Der [Li Zhende] 李貞德. 1996. “Chaoyue fuxi jiazu de fanli – Taiwan diqu ‘Zhongguo funüshi yanjiu’ (1945–1995)’” 超越父系家族的藩籬－台灣地區「中國婦女史研究」(1945–1995) [Research on Chinese women’s history in postwar Taiwan (1945–1995)]. Xinshixue 新史學 [New history] 7 (2): 139–79. https://doi.org/10.6756/NH.199606.0139.
Lewis, Su Lin. 2009. “Cosmopolitanism and the Modern Girl: A Cross-Cultural Discourse in 1930s Penang.” Modern Asian Studies 43 (6): 1385–1419.
Lin, Man-houng. 2001. “Overseas Chinese Merchants and Multiple Nationality: A Means for Reducing Commercial Risk (1895–1935).” Modern Asian Studies 35 (4): 985–1009.
Lu Yu-Hsia [Lu Yuxia] 呂玉瑕. 1984. “Funü jiuye yu jiating juese, quanli jiegou zhi guanxi” 婦女就業與家庭角色、權力結構之關係 [Women’s labor-force participation and family power structure in Taiwan]. Zhongyang yanjiuyuan Minzuxue yanjiusuo jikan 中央研究院民族學研究所集刊 [Bulletin of the Institute of Ethnology, Academia Sinica] 56: 111– 143. https://doi.org/10.7116/BIEAS.198412.0111.
Lu, Hwei-Syin. 2016. “‘The Bodhisattva’s Path’ as Gender-Neutral Practices: A Case Study of the Buddhist Tzu Chi Community in Taiwan.” In The Bloomsbury Research Handbook of Chinese Philosophy and Gender, edited by Ann A. Pang-White, 357–376. London: Bloomsbury Academic.
Mao Zedong. 1966. Quotations from Chairman Mao Tse-tung. Bilingual ed. Beijing: Foreign Languages Press.
Meskill, Johanna Menzel. 1979. A Chinese Pioneer Family: The Lins of Wu-Feng, Taiwan, 1729–1895. Princeton, NJ: Princeton University Press.
Nascimento, Amos. 2013. Building Cosmopolitan Communities: A Critical and Multidimensional Approach. Basingstoke, UK: Palgrave Macmillan.
Pan, Inghai. 2000. “Pingpu” Consciousness in Today’s Taiwan: On History and Ethnicity. China Perspectives 28 (March–April): 82–88.
Pan Ying 潘英. 1992. “Tan Taiwan pingpuzu de qianxi” 談台灣平埔族的遷徙 [A discussion of the migrations of Taiwan plains tribes]. Taiwan wenxian 台灣文獻 [Taiwan historica] 43 (4): 277–301.
———. 1993. “Pingpuzu shilue” 平埔族史略 [A short history of plains tribes]. Taiwan wenxian 台灣文獻 [Taiwan historica] 44 (1): 113–153.
Park, Saeyoung. 2017. “Me, Myself, and My Hegemony: The Work of Making the Chinese World Order a Reality.” Harvard Journal of Asiatic Studies 77 (1): 47–72.
Pollock, Sheldon. 2006. The Language of the Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India. Berkeley: University of California Press.
Ricci, Ronit. 2011. Islam Translated: Literature, Conversion, and the Arabic Cosmopolis of South and Southeast Asia. Chicago, IL: University of Chicago Press.
Rigger, Shelley. 2014. Taiwan Matters: Small Island, Global Powerhouse. Rev. ed. Lanham, MD: Rowman & Littlefield.
Rubinstein, Murray A. 1999. “Taiwan’s Socioeconomic Modernization, 1971–1996.” In Taiwan: A New History, edited by Murray A. Rubinstein, 366–402. Armonk, NY: M. E. Sharpe.
———. 2004. “Lu Hsiu-lien and the Origins of Taiwanese Feminism, 1944–1977.” In Women in the New Taiwan: Gender Roles and Gender Consciousness in a Changing Society, edited by Catherine S. P. Farris, Anru Lee, and Murray A. Rubinstein, 244–277. Armonk, NY: M. E. Sharpe.
Shepherd, John Robert. 1993. Statecraft and Political Economy on the Taiwan Frontier, 1600–1800. Stanford, CA: Stanford University Press.
———. 1995. Marriage and Mandatory Abortion among the 17th-Century Siraya. Arlington, VA: American Anthropological Association.
———. 1999. “The Island Frontier of the Ch’ing, 1684–1780.” In Taiwan: A New History, edited by Murray A. Rubinstein, 107–132. Armonk, NY: M. E. Sharpe.
Shi Lang 施琅.  1958. Jinghai jishi 靖海紀事 [Record of pacifying the seas]. Taiwan wenxian congkan 臺灣文獻叢刊 [Taiwan literature series] 13. Taipei: Taiwan Yinghang.
Shih, Shu-Mei. 2001. The Lure of the Modern: Writing Modernism in Semicolonial China, 1917–1937. Berkeley: University of California Press.
Shih T’ien-fu [Shi Tianfu] 施添福. 2013. “Cong ‘Kejia’ dao kejia (yi): Zhongguo lishi shang benguanzhuyi huji zhidu xia de ‘Kejia’” 從「客家」到客家（一）：中國歷史上本貫主義戶籍制度下的「客家」[From “the guest” to the Hakka (1): “The guest” under the principle of bon-gwan in the household registration system in the history of China]. Quanqiu Kejia yanjiu 全球客家研究 [Global Hakka studies] 1: 1–56.
Simon, Scott, 2005. Tanners of Taiwan: Life Strategies and National Culture. Cambridge, MA: Westview Press.
Skoggard, Ian. 1996. The Indigenous Dynamic in Taiwan’s Postwar Development. Armonk, NY: M. E. Sharpe.
Taraborrelli, Angela. 2015. Contemporary Cosmopolitanism. Translated by Ian McGilvray. London: Bloomsbury.
Tehrani, Jamshid J. 2013. “The Phylogeny of Little Red Riding Hood.” PLoS ONE 8 (11): e78871. https://doi.org/10.1371/journal.pone.0078871.
Teng, Emma Jinhua. 1998. “An Island of Women: The Discourse of Gender in Qing Travel Writing about Taiwan.” International History Review 20 (2): 353–370.
———. 2004. Taiwan’s Imagined Geography: Chinese Colonial Travel Writing and Pictures, 1683–1895. Cambridge, MA: Harvard University Asia Center.
Thompson, Laurence G.1964. “The Earliest Chinese Eyewitness Accounts of the Formosan Aborigines.” Monumenta Serica 23: 163–204.
———. 1969. “Formosan Aborigines in the Early Eighteenth Century: Huang Shuching’s Fan-suliu-k’ao.” Monumenta Serica 28: 41–147.
Ts’ai, Hui-yu Caroline. 2009. Taiwan in Japan’s Empire Building: An Institutional Approach to Colonial Engineering. London: Routledge.
Tu, Weiming. 1996. “Cultural Identity and the Politics of Recognition in Contemporary Taiwan.” China Quarterly 148: 1115–1140.
Wang Fu-chang [Wang Fuchang] 王甫昌. 2013. “Taiwan ruoshi zuqun yishi fazhan zhi lishi guocheng kaocha” 台灣弱勢族群意識發展之歷史過程考察 [The development of Taiwan’s ethnic consciousness: A historical examination]. Taiwan wenxue yanjiu 台灣文學研究 [Taiwan literature studies] 4: 60–83.
Wang, Lianmao. 1995. “Migration in Two Minnan Lineages in the Ming and Qing Periods.” Translated by Steven Harrell. In Chinese Historical Micro-Demography, edited by Stevan Harrell, 183–213. Berkeley: University of California Press.
Wang, Peter Chen-main. 1999. “A Bastion Created, a Regime Reformed, an Economy Reengineered, 1949–1970.” In Taiwan: A New History, edited by Murray A. Rubinstein, 320–338. Armonk, NY: M. E. Sharpe.
Wang Wan-Ting [Wang Wanting] 王琬葶. 2014. “Shijie de shengxiang: Rizhi shiqi Taiwan nüxing zazhi de nüxingzhuyi yuedu (1919–1939)” 世界的聲響：日治時期台灣女性雜誌的女性主義閱讀 (1919–1939) [The voices of the world: A feminist reading of women’s magazines in colonial Taiwan (1919–1939)]. Nüxue xuezhi: Funü yu xingbie yanjiu 女學學誌：婦女與性別研究 [Journal of women’s and gender studies] 34: 37–75. https://doi.org/10.6255/JWGS.2014.34.37.
Weller, Robert P. 1999. “Identity and Social Change in Taiwanese Religion.” In Taiwan: A New History, edited by Murray A. Rubinstein, 339–363. Armonk, NY: M. E. Sharpe.
Wolf, Arthur P. 1985. “Fertility in Prerevolutionary Rural China.” In Family and Population in East Asian History, edited by Susan B. Hanley and Arthur P. Wolf, 154–185. Stanford, CA: Stanford University Press.
———. 1989. “The Origins and Explanations of Variation in the Chinese Kinship System.” In Anthropological Studies of the Taiwan Area, edited by K. C. Chang, K. C. Li, A. P. Wolf, and A. Yin, 241–260. Taipei: National Taiwan University, Department of Anthropology.
Wolf, Arthur P., and Hill Gates. 2005. “Marriage in Taipei City: Reasons for Rethinking Chinese Demography.” International Journal of Asian Studies 2 (1): 111–133.
Wolf, Arthur P., and Chieh-shan Huang. 1980. Marriage and Adoption in China, 1845–1945. Stanford, CA: Stanford University Press.
Wolf, Margery. 1968. The House of Lim: A Study of a Chinese Farm Family. Englewood Cliffs, NJ: Prentice-Hall.
———. 1972. Women and the Family in Rural Taiwan. Stanford, CA: Stanford University Press.
Zhuluo Xianzhi 諸羅縣志 [Zhuluo gazetteer], comp. Zhou Zhongxuan 周鍾瑄.  1962. 12 juan 卷 in 2 vols. Taiwan wenxian congkan 臺灣文獻叢刊 [Taiwan literature series] 141. Taipei: Taiwan Yinghang.
About the Author
Melissa J. Brown is the managing editor of the Harvard Journal of Asiatic Studies at Harvard-Yenching Institute.
 Rather than a Habermasian bourgeois- and discourse-based conception, I use an Arendtian “common” conception of the public sphere—a socially constructed, shared space, outside the household, of human material culture and institutions where people meet and engage with one another (Arendt  1998, esp. chap. 2; d’Entreves 2019, sec. 6.1).
 From the Qing period through the twentieth century, all the variants of Bendi (in table 1) have been used to refer to Hoklo only, especially in contrast to Hakka (e.g., Tu [1996, 1132] and Shih , who uses the term Tuzhu 土著 interchangeably with Bendi). But the same terms have also been used for Hoklo and Hakka combined (e.g., Wang F. 2013, 66). The combined term also includes plains Aborigine peoples who had assimilated into a Han identity; many of these people have reasserted an Austronesian identity since the mid-1990s (e.g., Brown 2004, 2010, 2020). I use the term “Bendi” in this combined sense when discussing the Qing, Japanese colonial, and postwar periods.
 Anthropological debate questions whether the following criteria for identifying indigenous people (used by the United Nations, activists, and others) reinstate racist, nineteenth-century notions of primitive survivals (e.g., Barnard 2006; Guenther et al. 2006): descended from people who historically preceded other settlers, politically subordinated, culturally different from the majority population, and self-ascriptive claimants of the term “indigenous.” In Taiwan’s history, these criteria are complicated by ethnic intermarriage, changes in political dominance and alliances, acculturation, and assimilation (Brown 2004). I use “Bendi” as a category of indigeneity in the broad sense of “locals,” recognizing that it derives from a context of multiple colonizations and refers to peoples who have an ascribed identity of the earliest Han colonists (Hoklo and Hakka)—not an ascribed identity of any of Taiwan’s original inhabitants (Austronesians), regardless of some Bendi having some Austronesian ancestry.
 Delanty (2012, 341) does not assume progression.
 These approaches cut across the social sciences, philosophy, and legal studies.
 See Nascimento (2013, 211–223) on multiple cosmopolitanisms.
 For theoretical discussion of social linkages, see Brown and Feldman (2009).
 Sometimes also perceived as “modern” (e.g., Lewis 2009), such signaling can be viewed quite differently in colonial spaces versus in the metropole (S. Shih 2001).
 See, for example, Shi ( 1958), Huang ( 1957), Campbell (1896, 1903), Thompson (1964, 1969), Blussé (1984, 1990, 2003), Blussé and Roessingh (1984), Huber (1990), Shepherd (1993, 1995), Pan Y. (1992, 1993), L. Wang (1995), Brown (1996, 2004, 2008), Teng (1998, 2004), Chan and Chang (2001), Everts and Milde (2003), Heyns (2003), Andrade (2005). Of these, only Brown, Teng, and Everts and Milde emphasize the importance of Austronesian women. Notably, Andrade (esp. 2007, 2012) consistently ignores women’s contributions in his analyses of Taiwan’s historical plurality. For more recent views of Austronesian identity, see I. Pan (2000) and Hsieh (2018).
 This estimate is consistent with genetic research (S. Chen 2008).
 Shepherd (1995) discusses seventeenth-century Dutch reports that Siraya externally manipulated a woman’s abdomen to cause abortion.
 The woodblock print images in these figures are from the eighteenth century; thus, the Austronesians depicted here were already subject to a hundred years of colonization.
 Centuries later, anthropologists noted the greater willingness of people in Taiwan and southern China than in northern China to allow women to “stand in” for men when circumstances or skill called for it (A. Wolf 1985; A. Wolf and Huang 1980; Cohen 1976). This proclivity may constitute cultural substrate influences of Austronesian and Tai ancestors (A. Wolf 1989; Brown 2004; S. Chen 2008). For evidence of such deep-history cultural substrate influences elsewhere in the world, see Fortunato, Holden, and Mace (2006), and Tehrani (2013).
 The common translation “Pacification and Reclamation Bureau” is socially inaccurate. Although “pacification” conveys the ambiguity of fu 撫 as both mollification and military subjugation, “reclamation” fails to capture the imperialist mindset that virgin forests are “wastelands” appropriately annexed for agriculture (kaiken 開墾, lit., “opening wasteland”).
 By 1920, sugar was more than 70 percent of Taiwan’s total industrial output (Ka 1995, 75). Meanwhile, by 1915, camphor fell below 4 percent and tea below 7 percent (Ka 1995, 64).
 Dadaocheng’s founding resulted from Quanzhou-descended Han driving Tongan-descended Han out of Mengjia; they joined Zhangzhou-descended Han downstream in what became Dadaocheng (A. Wolf and Gates 2005, 115).
 Accounts of such cloistering often rely on patrilineage genealogies. Such family records may expunge not only women (showing fathers giving birth to sons) but also unsavory or criminal elements not to their credit (Meskill 1979, 65). It is unlikely that they recorded women departing from neo-Confucian ideals unless it served the lineage. Such occasional reports raise questions about where the women learned the necessary skills—for example, recruiting and leading troops, or pursuing four legal suits (Meskill 1979, 162–175, 188).
 I refer to this colonial effort as “Mainlander colonialism” because it favored Mainlanders in almost every sphere, regardless of whether they were KMT members. This term reminds us of the tensions that could accrue in quotidian social interactions between Mainlanders and Bendi.
 In the meantime, both Taiwan and Japan were effectively within the U.S. imperial sphere. Taiwan imported textiles and fertilizer from Japan, for example, in 1949–1951, when Taiwan was still reeling from the flood of Mainlander refugees, and during the KMT’s 1945–1949 focus on the Chinese Civil War (P. Wang, 1999, 338n24).
 Non-elite Bendi had little if any interaction with the peoples of Qing and Republican China during the Japanese period. (Chinese sojourners to Taiwan were strictly regulated, though many elite Bendi had businesses in China [e.g., Lin 2001].) But there was even less interaction between Taiwan and China under Mainlander colonialism.
 Thus, much postwar women’s-studies scholarship in and about Taiwan focuses on women’s roles within the family or efforts to overcome the limitations of these roles (Lee 1996).
 The decrease in percentage from 1973 to the 1980s probably reflects more teenage girls attending high school.
 Other mainland-origin leaders also failed to credit women’s contributions: in the inscription for the first (July 20, 1949) issue of Women of New China (Xin Zhongguo funü 新中国妇女), Mao Zedong (1966, 554–555) exhorted women to “unite and take part in production” (tuanjie qilai, canjia shengchan... huodong 团结起来，参加生产。。。 活动) as if women’s labor had never before produced anything significant.
 The idiom of “help” (bangmang 幫忙) is often used, even by women themselves, in ways that mask and undervalue women’s work (Brown 2017). The use of “help” to describe women’s work in Taiwan probably reflects Taiwanese women’s own descriptions (e.g., Kung 1976, 46; Gallin 1989, 377, 379–380; Chien 2015, 44, 45, 53).