INTRODUCTION
As has been well recognized by the existing literature on Korean nationalism (Shin Reference Shin2006), colonial occupation was crucial for the rise of nationalism in Korea. Yet the literature has long failed to theoretically grapple with the fact that, in the Korean case, the emergence of national categories and nationalism was coeval with the unprecedented level of population movement within the ambit of the expanding empire. A few recent, pioneering studies have started to criticize this sedentary bias, recognizing the critical importance of migration, diaspora networks, and the experience of exile to the construction of nationhood in Korea (Schmid Reference Schmid2002: 199–252; Lie Reference Lie2008: 168–94; R. S. Kim Reference Kim2011). Dispersal, collective suffering, homeland orientation, troubled relations with host societies, and identification with co-ethnic communities elsewhere (Cohen Reference Cohen1997) have all been central to the imagination of the Korean nation. These studies, however, have given no systematic, sustained attention to the concrete practices of the colonial state, which enabled, channeled, and challenged these emerging cross-border networks, national categories, and diasporic imaginations.
This article examines the role played by the colonial state in the formation of Korean diasporic nationhood, defined broadly as the institutionalization and the imagination of the Korean nation as a community of common descent, history, and destiny, whose primordial identity was ruptured neither by the extinction of “its own” state nor by its separation from an ancestral homeland.Footnote 1 Instead of analyzing Korean diasporic nationhood as a reactive formation to colonial occupation, I begin by noting the striking “isomorphism” (Anderson Reference Anderson1991: 114) between the Japanese colonial state and Korean anti-colonial nationalism in terms of their definitions of national collectivity. Studies of European overseas empires have employed a fruitful line of inquiry focusing on this isomorphism, and analyzing the dialectic role that colonial states played in the formation of national “geo-bodies” (Winichakul Reference Winichakul1994) and anti-colonial territorial nationalisms (Breuilly Reference Breuilly1993). The Korean case is different in that both the colonial state and anti-colonial nationalism defined the Korean nation in trans-territorial terms. From various possible sources of this distinctive phenomenon, I focus on the development of the Japanese colonial state's official classification practices concerning its Korean subjects who were on the move.
The population classification practices of colonial states have attracted much attention in studies of European colonialism. Scholars have examined how these states, as “taxonomic states” par excellence (Stoler Reference Stoler2002), contributed to the making, unmaking, and remaking of various “peoplehood” identities (Lie Reference Lie2004) of their heterogeneous populations (see also Steinmetz Reference Steinmetz2007).Footnote 2 The institutionalization of customary courts run by indigenous power-holders, for instance, involved transforming multiple, overlapping, and fluid local identities into mutually exclusive, exhaustive, and hereditary legal identities (Vail Reference Vail1989; Appiah Reference Appiah1993; Benton Reference Benton1999). The institutionalization of settler privileges constructed a sharp boundary between European settlers and native populations (Elkins and Pedersen Reference Elkins and Pedersen2005) and obliged colonial states to deal with various categorical “abnormalities,” such as offspring of sexual unions between settlers and natives, which blurred this boundary (Stoler Reference Stoler1992; Young Reference Young1995). The development of and debates over these classification practices were therefore deeply entwined with the vicissitudes of the plural legal orders that characterized European colonial rule. These practices also left a deep and lasting imprint on postcolonial nation-building processes (Young Reference Young1976; Mamdani Reference Mamdani1996; Longman Reference Longman, Caplan and Torpey2001). The central question that informed these studies was how heterogeneous colonial populations, once “encaged” (see Mann Reference Mann1986) within a single imperial administrative unit, could gradually be molded into a nation in some contexts, yet not in others. The literature has limited its analytic scope mostly to the encounters and ethnic mixings that took place within the territorial boundaries of the colonies they examine.
Such an analytic framework cannot adequately explain the legacy of colonialism in Korea, the largest and most important colony of Japan from 1910 until 1945. Unlike most European colonizers who encountered a patch of overlapping rules and a dizzying array of diverse human groupings in their conquered territories, the annexation of Korea brought Japan a relatively well-bounded territory that had been governed by a centralized monarchy (if one with limited bureaucratic power; Palais Reference Palais1996). The population was relatively homogeneous in terms of phenotype, language, religion, and cultural practices, despite the rigid status system in place (Duncan Reference Duncan, Lee and Park1998; Hwang Reference Hwang2007).
Further, the question of “mixed-blood” people posed no legal, administrative, or political quandaries comparable to those that arose in European empires, for several reasons: the short duration of imperial expansion limited the number of interethnic unions; the phenotypical similarity between colonizers and colonized, and the dominant (albeit debated; see Oguma Reference Oguma and Askew2002 and Fujitani Reference Fujitani2011) racial ideology that underscored their affinity, made the presence of these unions and their offspring relatively invisible and less disturbing; and broadly shared belief in patrilineality helped minimize ambiguity in sorting out their offspring. Therefore, the sites for population classification struggles that were so crucial in European overseas colonies were relatively unimportant in Korea.
But another vexing question did arise: how to deal with those who left the colony. In stark contrast to the absence of any large-scale, cross-border migration in premodern Korea, the incorporation of Korea into the Japanese Empire set in motion a massive outward migration of the colonized population. By the end of the Pacific War, “Koreans” outside the Korean peninsula comprised approximately 15 percent of the entire “Korean” population.Footnote 3 They became the focus of disagreements over who should be allowed to cross the internal and external borders of the empire, with what rights and obligations and what consequences for their ethno-national membership, and how they should be identified. These conflicts helped to construct, police, naturalize, and challenge the meaning and the boundary of “Korea” and the “Koreans.” Therefore, to understand population classification struggles in the Japanese Empire, and the long-term consequences these had for ethnic and national boundary making, we must look beyond the territorial unit of the colony.
In theorizing the importance of these classification struggles for the construction of Korean diasporic nationhood, I draw on a burgeoning interdisciplinary literature that has highlighted the relationship between migration control, state-building, and nation-building. This body of scholarship has shown how the global expansion of capitalism, long-distance migration, and the modern interstate system drove a “révolution identificatoire” (Noiriel Reference Noiriel and de Laforcade1996) beginning in the mid-nineteenth century. Scholars have examined how states' efforts to define and identify the set of persons entitled to privileged access to their territory, the labor market, and citizenship led to the development of their infrastructural power. This, in turn, enabled modern states to monopolize the legitimate means of movement (Torpey Reference Torpey2000) and engage with mobile people more effectively than ever before (Noiriel Reference Noiriel and de Laforcade1996; Zolberg Reference Zolberg, Fahmeir, Faron and Weil2003; McKeown Reference McKeown2008; on “infrastructural power” see Mann Reference Mann1986). The literature has also started to extend its analytic focus beyond the context of immigration by documenting a similar development in emigration states, which sought to monitor their growing emigrant populations, access their resources, and secure their loyalty (Green and Weil Reference Green and Weil2007; Choate Reference Choate2008; Fitzgerald Reference Fitzgerald2009). How such state practices shaped the idioms of self-identification by migrants has also attracted scholarly attention (Caplan and Torpey Reference Caplan and Torpey2001; Gabaccia and Ottanelli Reference Gabaccia and Ottanelli2005).
Three characteristics of this body of work are important for this study. First, these studies identify specific innovations in institutional arrangements and bureaucratic techniques that provided—intentionally or not—a critical institutional skeleton for the imagination of nationhood; they do not consider these arrangements and techniques merely as a background for intellectual discourses and political movements that animated this imagination. As John Torpey briskly put it, “national communities must be codified in documents rather than merely ‘imagined’” (2000: 6). Second, these studies have not treated the boundaries of nation-states as the ultimate horizon of their analyses; instead, they place the development of the official classification systems targeting mobile populations within the context of multiple and mutually intersecting border-spanning practices such as international protocols, bilateral treaties, consular practices, cross-border migrations, and various transnational networks. These practices disseminated particular ways of controlling borders and defining national boundaries across the globe, nationalizing not only territory but also identity (McKeown Reference McKeown2008). Third, this body of work has shown that the globalization of border control, with its heavy reliance on the rise of the “documentary regime of verification” (Robertson Reference Robertson2009), was a product of long-term, uneven, and tumultuous historical processes that varied across regions. A comparative and historical approach to this topic is essential.
Up to now, the empirical focus of these studies has largely been limited to the intra-European and transatlantic migrations of Europeans—and to a lesser extent Chinese or Indian migrations to either Southeast Asia or white settler colonies of the British Empire—which peaked at the turn of the twentieth century. This theoretical perspective has rarely been applied to processes that occurred in northeast Asia (present-day Japan, Korea, northeast China, and maritime Russia) that were analogous but took place in a different macro-regional environment (see Park Reference Park2009 for an exception).
From the mid-nineteenth century through the first half of the twentieth, northeast Asia experienced the dissolution of the historic, sinocentric regional order, the rapid rise and expansion of the Japanese Empire, and an unprecedented level of intra-regional movement and ethnic mixing. Over three million people migrated from the Japanese archipelago to Korea and Manchuria as colonial settlers (Watt Reference Watt2009: 19–39), while at least thirty million moved from mainland China to Manchuria, sinicizing this historic frontier (Gottschang and Lary Reference Gottschang and Lary2000). The competing state- and nation-building processes in the region profoundly shaped the regulation of these migration flows. These processes were further complicated by the migration of more than four million colonial subjects from the Korean peninsula to Japan, Manchuria, and maritime Russia.Footnote 4
In what follows, I will focus on Korean migration to Manchuria, the disputed borderland between China and the Japanese Empire, where conflicts over the membership status of Korean migrants (about two million by 1945) were particularly vehement. The colonial state's effort to bring these Korean migrants under its control was undermined by the sparse presence of any state in Manchuria, by China's competing jurisdictional claim over Korean migrants, and by the region's complex ethno-political landscape. All of these factors stood in sharp contrast with the situation in the tightly controlled Japanese archipelago, which was the destination for a comparable number of Korean migrants (see Weiner Reference Weiner1994; Lie Reference Lie2008; Kawashima Reference Kawashima2009). Yet this historic frontier, which had been integrated only loosely and unevenly into the Chinese nation-building project, also gave the Japanese considerable room to maneuver, and they envisioned Manchuria as a potential component of their expanding empire. These conditions were of course absent in the United States (and to some extent in maritime Russia) where Japan's influence, including its jurisdictional claim over Koreans, was fairly limited (Park Reference Park2009; R. S. Kim Reference Kim2011).
I will show how the state's efforts to both control and comprehend colonized populations moving beyond its territorial reach, often carried out against the migrant preferences or claims of rival states, provided an impetus for the growth of its infrastructural power. The ensuing development of various legal, organizational, and bureaucratic infrastructures, in turn, further helped the state to enumerate, identify, and penetrate its increasingly mobile population, if with varying degrees of success. Moreover, such state efforts contributed to the construction, solidification, and naturalization of the category of “Koreans” by making the “Korean nation” a community that was legally codified, pervasively institutionalized, and enduringly documented. As migrants engaged with various state practices that regularly sorted, re-sorted, and treated them as “Koreans,” they also came to experience their Korean identity as tangible and consequential. Their chances for migration, education, or employment, their rights to property, subsidies, or protection, and their subjection to the state's extraction or violence all came increasingly to depend on how state agencies defined and identified them. By the mid-twentieth century, “nation” (minjok) had become an easily understandable category of identity, and “Korean” (Chosŏn) an unchallengeable ethnonym for people of peninsular origin. “Korean nation” (Chosŏn minjok) came to be a well-bounded, pervasively institutionalized ethno-national group, not only within the colony but also outside of it.
I argue, in a nutshell, that the state's engagement with migrants outside the peninsula provided, albeit unwittingly, a critical institutional scaffolding for the imagined community of the Korean nation, eventually conceived as transcending the colony's geographical boundary. Without considering this trans-border dimension of Japanese colonial state-building, which occurred in the context of imperial expansion, massive migration, and competing nationalist projects, we cannot explain the genesis of Korean diasporic nationhood, not only as a core of the subjective self-understandings of exiled intellectuals, but as an objectified legal and administrative identity imposed categorically on migrants from all walks of life.
I will begin by examining the Japanese Empire's nested membership structure and its bureaucratic manifestations, focusing on the family registration system. This will help explain why migrants became the primary locus for the empire's population classification struggles. This will be followed by two sections that analyze how these legal and bureaucratic infrastructures shaped, and were shaped by, an unprecedented level of Korean migration to Manchuria, before and after the 1932 establishment of Manchukuo (the Japanese puppet state). In each section, I ask which ethno-national boundary—the internal or external boundary of Japanese nationals—became the focus of dispute; what challenges the state faced from colonial subjects, indigenous populations, and other states; and to what extent and on what grounds migrants could negotiate their identity vis-à-vis the states involved. I conclude by summarizing my arguments, and discussing the generalizability of my findings and the article's broader contributions.
I draw primarily on a broad range of secondary literature in English, Korean, and Japanese, and complementarily, archival materials including laws, inter-state treaties, administrative regulations, newspaper articles, memoirs, and fictional accounts. My primary topic is contestations of the colonial state's identification practices, so I will only discuss materials that show a range of nationalist thoughts, or the kaleidoscopic reconfiguration of migrants' subjective identities, when they are pertinent to my focus.
THE PLURAL LEGAL ORDER AND FAMILY REGISTRATION SYSTEM
Like many of its Western counterparts, the Japanese Empire did not constitute a single, unified national space. Before the empire's expansion in the late 1930s, it consisted of six different “regions” (chiiki, 地域), one of which was Korea (Chen Reference Chen, Myers and Peattie1984). In the Japanese legal lexicon, “region” signified, not simply a geographically circumscribed space, but also a legal and administrative unit governed by a separate set of laws.Footnote 5 Korea, like Japan's other formal colonies of Karafuto and Taiwan, was governed by the Government General, which enjoyed broad legislative power and was fairly independent of the Imperial Diet or the metropolitan government (ibid.).
Corresponding to these plural legal spaces, Japanese nationality came to be made up of different “regional” categories, including Japanese (narrowly defined), Taiwanese, Korean, and others. This plural legal system engendered a series of questions: On what basis should an individual's “regional” membership be determined—by birthplace, residence, descent, culture, or some other characteristic? How should the state identify which individuals were Japanese, Taiwanese, or Koreans? Should such regional memberships be alterable, and if so, under what conditions? These questions of classification, identification, and boundary-crossing all converged on the empire's basic civil registration scheme—the family registration system (koseki, 戶籍). The distinctive and intricate ways in which this system came to buttress the empire's nested membership structure help explain why migrants became the primary locus for population classification struggles.
The family registration system was first enacted in 1871 in the Japanese archipelago, and served to delineate the boundary of the “Japanese” (nihonjin), which had not been clarified by either the constitution or the nationality law (both of 1889) (Chapman Reference Chapman2011).Footnote 6 Every Japanese subject belonged to an ie (family, 家), the unit of registration, and each ie had its own honseki (place of origin, 本籍). Honseki indicated the address at which a family was originally registered, and linked each ie to a concrete point in the newly nationalized space. This space was reconceived through cadastral surveying, the reorganization of administrative units, and postal systems. Unless a person's entire family relocated its registry, or a family member was adopted or married into a different family, their official “place of origin” (inherited patrilineally and patrilocally) was supposed to remain the same regardless of any change of residence. And having one's “place of origin” within Japanese territory validated a person's Japanese nationality regardless of where they actually lived.
Japan's imperial venture raised the question of whether this rule should be equally applied to colonial subjects residing in the newly annexed Japanese territory. One of the first actions the Japanese government took in its colonies was to institute family registration systems. These emulated the system in the metropole yet were also distinct from it. In this way they attested that both colonizers and the colonized shared membership in the Japanese Empire, but also served to mark and maintain distinctions between them.
In Korea, where the family registration system was first introduced in 1909, a year before the annexation, this was done in two ways. First, the system helped the state impose unified practices of naming (Cho Reference Cho1981; Oh and Ryu Reference Oh and Ryu2009) and kinship (Yang Reference Yang1998; Kim Reference Kim2007) onto the entire “Korean” population regardless of status, class, or gender differences.Footnote 7 The system thus, to some degree, homogenized and equalized Koreans.Footnote 8 This was quite different from many European colonies in Asia and Africa where the state established multiple customary courts that helped consolidate, institutionalize, or even create tribal, ethnic, or religious identities and fissures.
What I want to underscore here is how family registration institutionalized and thereby consolidated differences between Koreans and Japanese proper. Given the negligible phenotypical differences between them and the deepening, if slow, cultural and linguistic assimilation taking place under colonial rule, Korean-style names served as the clearest diacritical marker for distinguishing Koreans from Japanese. Different kinship practices, buttressed by the separate civil and customary legal systems, made it difficult for them to form families together through marriage or adoption (Yang Reference Yang1998). These “different customs” the state sometimes ignored, at other times carefully maintained, and at still others actively created, though it claimed all the while to be merely recognizing extralegal realities (Kim Reference Kim2007). As postcolonial studies have documented, these shifting stances highlight a contradiction inherent in imperial rule between the demand for assimilation and the insistence on difference (Cohn Reference Cohn1996; Chatterjee Reference Chatterjee1993; Bennett Reference Bennett2004; Hoffman Reference Hoffman2010). This tension is obvious in a 1911 ordinance that forbade Koreans from taking Japanese-style names, and a 1939 one that forced them to adopt such names (along with Japanese kinship practices) (Miyata, Kim, and Yang Reference Miyata, Kim and Yang1992; Mizuno Reference Mizuno and T'ae-jŏng2008). Even in 1939, the colonial state encouraged Koreans to adopt Japanese-style surnames that were rare among the Japanese, or to at least keep their Korean-style proper names (Mizuno Reference Mizuno and T'ae-jŏng2008). Clearly, the Japanese continued to worry about boundary-maintenance.
Yet naming and kinship practices were not the most important grounds for distinguishing “Japanese” from “Koreans”; what ultimately determined whether a person was Japanese (narrowly defined), Korean, or Taiwanese was their “place of origin” in the family registry (Endō Reference Endō2010; Chapman Reference Chapman2011). As mentioned already, within the empire's nested membership structure, to be recognized as a Japanese national one had to belong to a certain “family,” which in turn belonged to a certain “region” depending on the family's “place of origin.” No one, except royal family members, could become a Japanese national without the mediation of this nested membership structure. The family registry, therefore, came to serve as the ultimate documentary foundation that defined an individual's regional membership. It also made that identity visible to anyone engaged in transactions involving official papers, which facilitated the regulation of the external and internal boundaries of Japanese imperial membership. The family registration document, in other words, was the final barrier against “passing.”
There were conditions under which Japanese imperial subjects were allowed to change their place of origin, and by extension their regional membership. After reviewing several alternatives, by the late 1910s the government established that Japanese nationals could change their regional membership only by changing the “family” (ie) to which they belonged by means of “inter-regional” marriage, divorce, or adoption. Otherwise, neither long-term residence, birth in other regions of the empire, or cultural assimilation were considered legitimate grounds for changing one's “place of origin,” and by extension, regional membership.Footnote 9 Through such legislation, “regional” membership in the Japanese Empire effectively became ethno-national membership, determined and inherited based on patrilineal and patrilocal principles.Footnote 10
Yet, the establishment of these legal principles did not guarantee their implementation on the ground. To document the entire population in a uniform registry required tremendous infrastructural power (Loveman Reference Loveman2005; Sadiq Reference Sadiq2008). That is, to effectively maintain the horizontal boundaries of membership, the state would have to transform the vertical transactions between itself and the population. This was easier planned than done, and the protectorate government had only limited success in inaugurating the first family registration system in 1909. The first survey, conducted by the police, did manage to count some three million hitherto undocumented Koreans, showing that the protectorate government's administrative capacity was superior to the previous Korean government's (Hwang Reference Hwang2004: 367–68). However, the omission rate remained high in the early 1910s, at an estimated 20 to 25 percent of the Korean population (Oh and Ryu Reference Oh and Ryu2009). Updates of the family registry ultimately depended on voluntary reports by civilians, who seldom saw reason to cooperate. To many Koreans, traveling to a remote police office to register seemed not just bothersome but terrifying, and the reward for doing so remained obscure.
The intense colonial state-building process in the following decades did improve the situation. Increasing interactions with lower-level bureaucrats, who swelled in both numbers and types (Uchida Reference Uchida2011: 23, 68; Fujitani Reference Fujitani2011: 42–43, 63), made the family registration document and the official identity it established necessary for conducting many important transactions, such as marriage, divorce, adoption, inheritance, or admission into schools. But as Takashi Fujitani observes, even as late as the early 1940s, “the colonial government still found itself scrambling to put Korean [family] registers in order” (2011: 38). In 1942, it was estimated that over 1.1 million Koreans (approximately 4 percent of the entire Korean population throughout the empire) remained unregistered, and the accuracy of even the existing registry was in doubt.
This “large, fluid segment of the population [that] remained invisible to the state” (ibid.: 69) was partly a product of high rates of home birth and infant mortality among Koreans, and their low education levels, meaning incompetence in the empire's lingua franca, Japanese (ibid.: 37). But another significant obstacle was a dramatic rise in migrations, especially to Manchuria. It is much more difficult for a state to track a mobile population beyond its territorial governance when that group's ties to its ancestral homeland, and its political allegiance to the state, are weak. Moreover, the state in which migrants settle may challenge the trans-territorial jurisdictional claim of their state of origin, and claim them as its own. In this way, Korean migrants were not only a source of ambiguity in the conception of the empire's regionally demarcated, plural legal system; they also posed a tremendous practical challenge to that system's implementation.
MAKING LOYAL COLONIAL SUBJECTS IN THE CONTESTED BORDERLAND: MIGRATION TO MANCHURIA BEFORE 1931
If we take no action in response to the incomparable treatment and contempt to which [Koreans] are subjected in adjacent Manchuria, what else would it mean other than the annihilation of the Korean nation (Chosŏn minjok)?… The Government General of Korea is simply looking the other way, and the Chinese, while looking askance at this bystander, are whipping us on our heads and committing heinous atrocities against our brethren (tongp'o)…. We are twenty million (i-ch'ŏn-man) in number, our own rivers and mountains reach 3,000 (sam-ch'ŏn) li [a traditional unit of distance], and we boast five thousand years (o-ch'ŏn-nyŏn) of history.… My co-ethnic brothers! We must take an action!
———Saengjon'gwŏne Kwanhan Chawich'aek (A measure of self-defense for survival), Tonga ilbo, 9 Dec. 1927What is often called Manchuria constituted the northeastern frontier of imperial China. The Qing court carefully maintained the separation of its homeland-cum-hinterland from China proper by, among other things, banning Han Chinese settlement in Manchuria (Elliott Reference Elliott2000).Footnote 11 Ethnic Koreans occasionally crossed the border for hunting, herb collecting, or seasonal farming, and actively engaged in cross-border trade, yet they did not set up concentrated, enduring settlements there until the mid-nineteenth century. State control of irregular border-crossers and private merchants remained selective and sporadic (Kim Reference Kim2006).
These patterns began to change after the mid-nineteenth century. Deepening economic and political crises in Korea gave the first boost to Korean settlement in eastern Manchuria, beginning in the 1860s. The Qing court, increasingly wary of Russian expansion, sought to nationalize this sparsely populated frontier by promoting Han Chinese settlement there. Korean settlement, by contrast, was tolerated rather than encouraged. Koreans were granted ownership of the newly opened lands only if they changed their hairstyles and outfits to Manchu styles and paid a land tax to the provincial government. In the absence of the modern nationality law, the Qing state considered these actions markers of their naturalization (Liu Reference Liu2002: 33–52).
The Korean state initially refrained from confronting the Qing government's new policies. In fact, the precolonial Chosŏn court (1392–1897) did not clarify in principle how its subjects' migration and settlement outside the Korean peninsula, or acquiring membership in another state, would influence their relation to the Korean state (Hwang Reference Hwang2007). This began to change after the historic tributary relationship was dissolved in 1895. The Korean government (the short-lived successor of the Chosŏn court, the Taehan Empire) installed a police station near the northern border, dispatched officials to carry out a census of settler communities, discouraged Korean settlers from naturalizing as Qing subjects, and insisted on revoking completed naturalizations (Liu Reference Liu2002: 58–64). The government also revived a long-dormant territorial dispute, claiming that eastern Manchuria, historically called Kando by Koreans, belonged to Korea. This new assertiveness was informed by the thriving nationalist imagination of the day, which rediscovered Manchuria as the birthplace of the Korean nation and demanded its “recovery” as a way to redeem the nation's vigor, autonomy, and self-confidence (Schmid Reference Schmid2002: 199–236).
This inchoate dispute did not end after Korea became Japan's protectorate. Japan had its own interest in bringing Koreans in Manchuria under its control: apart from its concern with the growing anti-Japanese nationalist militias in Manchuria, Japan considered Korean settlers as a bridgehead for its advance into the area, whose rich resources and strategic value seemed indispensable to the empire's survival (Brooks Reference Brooks2000). The protectorate government (1905–1910) began to claim eastern Manchuria and Korean settlers as “its own” just as the Korean state had, only more aggressively and systematically. In addition to opening a police office in Kando, conducting a census, banning Korean naturalization, and posing a territorial claim to Kando, Japan took a further step in clarifying the nationality status of Korean settlers in the language of the modern interstate system; for the first time, the horizontal boundary of “Koreans” was demarcated legally and administratively. The protectorate government announced an ordinance in 1908 that forbade Koreans adopting a foreign nationality under any circumstances (Liu Reference Liu2002: 134–35; Seo Reference Seo2006: 182–87). This was an exceptional move, for the Japanese nationality law allowed voluntary renunciation of Japanese nationality.
A 1909 treaty between the Qing and Japan resolved the dispute over territory, but not that over jurisdiction. Japan withdrew its territorial claim, but obtained the right to establish consular offices in eastern Manchuria, which were put under the joint control of the Ministry of Foreign Affairs of Japan and the Government General of Korea (GGK).Footnote 12 By this time there were almost two hundred thousand Korean settlers in eastern Manchuria (Liu Reference Liu2002). Although the Qing court tried to place them under its own jurisdiction, the treaty neither clarified their nationality status nor made any clear distinction between naturalized and non-naturalized Koreans. Japan continued to claim that ethnic Korean migrants were all Japanese nationals and expanded its zone of influence by asserting its right to protect “its own” nationals. China's policies concerning Korean migrants came to be guided by its wish to neutralize such claims. Disputes therefore focused, not on the boundary between Koreans and Japanese proper (i.e., the internal boundary within the category of Japanese nationals), but instead on the boundary between Japanese nationals and non-nationals (the external boundary of Japanese imperial subjects). This Japan-China competition shaped the basic contours of the membership politics concerning Korean migrants in Manchuria throughout the 1910s and the 1920s. I cannot present here the complexities of how this politics unfolded, but for our purposes I need only highlight its four distinctive characteristics.
First, the state authorities involved, rather than cooperating, competed to make Korean migrants their own loyal nationals. For this purpose, both Japan and China increasingly predicated migrants' access to valued goods and services on their national affiliation: landownership, land leases, tenancies, residencies, poverty relief, tax reductions, financial aid, higher education, and consular protection. As a corollary, both governments punished those whose loyalty they presumed lay with the other side with measures such as heavier taxation, police harassment, closing of schools, cancellations of tenancy, confiscations of land, and deportations (Kim Reference Kim2002; Sin Reference Sin2002).
The shifting Chinese policies toward Korean migrants reveal a dilemma of the Chinese government. It could not drive Koreans out of Manchuria because their communities were already well established and their rice production brought considerable economic benefits. Nor could it embrace Korean migrants unconditionally, since they tended to bring Japanese influence in the form of aggressive, consular jurisdictional claims. This double-bind shaped Chinese naturalization policies vis-à-vis Koreans. Despite local variations (Park Reference Park2005: 64–123), China moved generally from encouraging to restricting the naturalization of Koreans. While non-naturalized Koreans became subject to severe discrimination, naturalization also became more difficult. This was because the Chinese authorities suspected that the Japanese government or capital was behind naturalized Koreans surreptitiously trying to take possession of land through various backstage agreements. The shift displays China's frustration that Korean naturalization did little to block Japan's encroachment into Manchuria.Footnote 13
Increased exclusion, in turn, made Korean settlers ever more reliant on whatever protection and benefits that Japanese consular offices could offer. The 1919 mass uprising of Koreans initially jolted Japan into taking a more conciliatory approach toward its colonial population in the name of the “Cultural Rule.” This policy change was also applied to Koreans in Manchuria, where Korean nationalist militias were seeking co-ethnic support and alliance with Chinese or Soviet communists. The GGK made “protection and control” the double goals of its new policies vis-à-vis Koreans in Manchuria (Kim Reference Kim1992). On one hand, strict control continued: the GGK and the consular police jointly carried out a series of brutal crackdown campaigns that targeted nationalist guerillas and (presumably) complicit civilians. They also, through the 1925 Mitsuya Agreement, provoked and allowed the Chinese authorities to harass Korean communities and arrest subversive Koreans, an attempt to deny the GGK an excuse to send “counter-bandit” troops to Manchuria (Esselstrom Reference Esselstrom2009). However, the Japanese authorities also tried to appease Koreans' hostility toward the colonial power by providing them with low-interest loans, poverty relief, disaster relief, primary education, subsidies to private schools, vaccines, advanced farming techniques, irrigation facilities, and other forms of support (Kim Reference Kim2002). The Japanese consular offices also began to intervene more directly and resolutely in favor of Koreans in a variety of disputes with Chinese, especially those concerning land contracts, which only aggravated Chinese resentment and hostility toward both Japanese and Koreans.
A second distinctive characteristic of this situation was that the entry of Korean peasants into Manchuria was left almost free of control. The primary site of struggle was not the control of Korean entry into Manchuria, as was the case in Korean migration to Japan (Weiner Reference Weiner1994), but rather the control of the terms of their settlement. The Chinese warlord regime that had emerged in Manchuria by the late 1910s wanted to populate the resource-rich and sparsely populated borderland for development and state-building, whereas in Japan proper emigration was promoted to ease high population pressure (Lynn Reference Lynn, Elkins and Pedersen2005). Nonetheless, even after the number of Han Chinese settling in Manchuria had soared, and tensions between China and Japan increased, the infrastructurally weak Chinese state could not seal the Korea-China border as effectively as it wanted to. In addition, the GGK did not cooperate with what Zolberg (Reference Zolberg, Fahmeir, Faron and Weil2003) dubbed the “remote control” of immigrants—that is, stopping Korean migrants from leaving in the first place. The Korea-China border was an international border in the formal sense of the term, but long remained porous to Koreans seeking entry into Manchuria.
The limited involvement of the GGK in Korean migration to Manchuria hampered its effort to define, identify, and thereby produce “its own” subjects out of migrants in the borderland, and this is the third distinctive characteristic of the situation. We have seen that the Japanese colonial state had legally and administratively defined the boundary of “Koreans” for the first time, and made this identity an almost immutable one that neither settlement nor birth outside the empire could dissolve. The success of a consular jurisdictional claim like this is predicated on a state's capacity to render “legible” (Scott Reference Scott1998) “its own” members among the opaque masses beyond its territorial reach; in other words, to clarify the horizontal boundary of its population, the state must transform the vertical relationship between the state and the population. This was a particularly daunting task in Manchuria, for several reasons. Many migrants had left their native hometowns before the 1909 introduction or the mid-1920s consolidation of the family registration system in Korea. A significant number did not migrate there directly from Korea, but rather from Russia/the Soviet Union where they had initially settled. And few Koreans took the trouble to register their second- or third-generation children in their native hometowns. Given all of these complexities, Korean migrants in Manchuria rarely had left solid documentary traces that could substantiate their peninsular origin and nationality status (Kwon Reference Kwon1933). The vast and unwieldy geography of the region, the unorganized nature of migration to Manchuria, the indifference or resistance of many Koreans to the colonial power, and the opposition of the Chinese authorities combined to cripple the colonial state's capacity to follow, penetrate, or take in Korean populations in Manchuria.
The state tried to overcome these adverse circumstances through various registration and documentation practices of a quasi-governmental organization called the Korean Association (Chōsenjin Minkai). Organized and sponsored by the GGK since the mid-1910s, this association was increasingly delegated a variety of administrative tasks, including registering Korean migrants' civil status, residence, and property, and issuing them travel certificates or, for those who had been arrested and released during the Japanese “counter-bandit” operations, certificates of defection/absolution (Kim Reference Kim2002: 59–62). These registration and documentation practices were crucial if the state was to identify and produce (docile) “Koreans.” Japan was able to use these administrative tools to smother Korean nationalist resistance and subject Korean residents to everyday surveillance and control (Esselstrom Reference Esselstrom2009: 65–91). The same practices, however, also provided migrants with a degree of stability and security in their everyday lives, and validated their entitlement to whatever meager protection and benefits the Japanese authorities could offer. These benefits encouraged migrants to register with the Korean Association, and in the late 1920s, some forty-six thousand Korean households in eastern Manchuria, about 60 percent of the estimated total, were reportedly affiliated with it (Kim Reference Kim2002: 61–64).
The preceding discussion brings us to the last characteristic that made this context distinctive: the degree of migrants' agency regarding the state's efforts to nationalize both territory and identity. Korean migrants in Manchuria initially had considerable room to negotiate their categorical identities vis-à-vis the claims of the competing states. The sparse and uneven state presence in the region allowed them to keep the state at a distance unimaginable in Japan or the United States. A peculiar land ownership arrangement prevalent at the turn of the century is a case in point. Although the Qing court formally banned non-naturalized Koreans from owning land, many jointly purchased land, but included one naturalized Korean, appropriately called “the owner on paper,” who served as the nominal owner on occasions when local Qing officials had to be dealt with (Son Reference Son2001: 98–102).
Also important was that China vigorously asserted a competing jurisdictional claim. For instance, when confronting Japan's ban on Korean naturalization, Chinese local governments argued that Korean migrants, especially those who had migrated before 1910, had become stateless as of the annexation and thus did not need Japan's permission for their naturalization (ibid.: 150–81). China's competing claim enabled Korean migrants to resist the jurisdictional claims of the colonial state, which followed them all the way to Manchuria, by offering an “exit” option, in Albert Hirschman's sense (Reference Hirschman1970).Footnote 14 Immediately after the annexation of Korea, thousands of Korean settlers jointly submitted to the provincial Qing government an application for naturalization. The establishment of the Republican regime in China in 1912 motivated some Korean intellectuals in Manchuria to launch a naturalization campaign in 1913 (Son Reference Son2001: 111). Chŏngŭibu (正義部), one of the anti-Japanese military organizations established in Manchuria in 1924, encouraged Koreans to naturalize as a way of avoiding Japanese surveillance and persecution of nationalist activities (Kim Reference Kim2002: 150).
When migrants made decisions about naturalization, considerations of practical benefits or sanctions were as important as ideological concerns. To illustrate, when a Chinese soldier murdered an innocent young Korean man in a crowded Longjing marketplace, that became a lightening rod for early-1923 Korean protests against insufficient protection provided by the Japanese consular office. The incident even provoked the otherwise pro-Japanese Korean Association to organize mass protests, several days of business closures, and region-wide assemblies. It produced a resolution encouraging the renunciation of Japanese nationality and promoting self-rule to be submitted to the Imperial Diet in the metropole (Ch'oech'angho 1923).
Toward the end of the 1920s, however, intensification of competitive nation-building and state-building in Manchuria gradually narrowed the migrants' room to maneuver. As I have said, the Chinese government took an increasingly prohibitive and punitive approach toward Korean migrants, who were now seen primarily as a vanguard of Japanese imperialism, and excluded them from its nation-in-the-making on its northeastern frontier. Korean migrants, in turn, turned to the Japanese consular office to file grievances and solicit protection. A series of editorials in Tonga ilbo—a Korean-language newspaper in colonial Korea that provided a forum for right-wing nationalists—and petitions submitted by the Korean Association in the late 1920s struck a surprisingly similar chord, despite differences in tone (solemnly critical versus ruefully pleading) and in the remedies they proposed (self-defense versus a tighter integration into the empire). They attributed the Japanese authorities' lukewarm responses to Korean sufferings to their status as “an adopted son of the Emperor” (Kantō Kyokukogai 1971 [1930]),Footnote 15 and lamented, “If Japanese suffered … the persecution to which Koreans are subjected in Manchuria, the troops would have already been dispatched” (Chaeman Chosŏnin Munje 1927).
The unfolding of what is known as “the Wanbaoshan Affair” in 1931 dramatically illustrates how political alliances and patronage structures in northeast Asia were being reconfigured on the eve of Japan's occupation of Manchuria. A conflict between Korean and Chinese peasants over an irrigation canal led to several days of bloody clashes and interventions by both Chinese and Japanese authorities. The repercussions were not limited to Manchuria, since Koreans in Korea expressed their solidarity with their “brethren” (tongp'o) by organizing fund-raising groups, dispatching civilian investigation teams, boycotting Chinese merchandise, and rioting against Chinese residents in Korea (Lee Reference Lee2009). As this section's epigram conveys, the transnationalization of this particular episode of ethno-political conflict made clear that co-ethnics in Manchuria had come to be considered part of the imagined community of the Korean nation (Chosŏn minjok). This imagined community was often expressed with the rhymed set phrase, “i-ch'ŏn-man tongp'o (twenty million brethren), sam-ch'ŏn-li kangsan (the territory that reaches about 733 miles), and o-ch'ŏn-nyŏn yŏksa (five thousand years of history).” By the early 1930s, the colonial state was constantly being criticized for “looking the other way,” and a range of Koreans of different ideological stripes called upon it to rescue and protect the “Korean brethren in Manchuria.” The colonial state did not reject this demand outright, but rather played along with it, portraying itself as the custodian of Koreans both inside and outside the colony (whether or not its actual performance stood up to this claim).Footnote 16
MANCHUKUOANS, JAPANESE, OR KOREANS? KOREAN MIGRATION TO MANCHURIA AFTER 1931
Since he naturalized [as Chinese], old man Park had carried his [Chinese] registration document on his person at all times. But after Manchukuo was established, he stashed the paper away in a chest at home.… Although he himself could barely speak a few Japanese words, with his son Myŏnghwan's help, old man Park [proficient in Chinese] was able to earn some money by brokering land transactions between Manchurians [Manin] and Japanese. There was no point in carrying the evidence that he had in fact naturalized as Chinese before.
———Mansŏn Kim, Ijung Kukchŏk (Dual nationality) (Reference Kim1989 [1948])As implied by old man Park's change of attitude in this short story, Japan's 1931 occupation of Manchuria transformed the political context within which disputes over the membership of Korean migrants played out. For complex geopolitical reasons, instead of making Manchuria its fourth formal colony, Japan established “Manchukuo” (Manzhouguo, in Chinese) and stylized it as an independent, multiethnic state, a product of self-determination of “the people of Manchuria” (Duara Reference Duara2003). Manchukuo's relationship to the Japanese Empire was thus ambiguous. In its substance, it was arguably Japan's “informal colony,” yet its form was carefully modeled on the modern nation-state (Han Reference Han1995), and its ideology was rooted in pan-Asianism (Duara Reference Duara2003; Yamamuro Reference Yamamuro and Fogel2006).
Though Manchukuo was eloquently defended in the language of nationalism and splendidly adorned with the paraphernalia of a nation-state, it was no offspring of national self-determination. The historical, social, and cultural unity of its population could not have been weaker. Only slightly over 10 percent of the population could have been considered natives of Manchuria, and most of the Han Chinese, who made up more than 80 percent (Watt Reference Watt2009: 31), had migrated from neighboring provinces within the previous decade or so (Gottschang and Lary Reference Gottschang and Lary2000). Given the cyclical migration pattern of the Han Chinese migrants, their enduring ties with their communities of origin, and their brushes with Japanese authorities over the previous decade, it is no surprise that they saw Japanese more as colonizers than as co-nationals. Migration from neighboring Chinese provinces continued well into the 1940s, in the context of intensifying China-Japan conflict, and the government would continue to wrestle with the questions of how to regulate their immigration and membership status, and how to instill in them a sense of belonging, until Manchukuo's collapse (Endō Reference Endō2010: 231–36).
Just as tricky was how to incorporate growing numbers of Japanese migrants into the Manchukuo national citizenry. Their membership status could not be resolved in the same way as it had been in Japan's other colonies, where plural legal systems persisted. To maintain the façade that Manchukuo was an independent nation-state, Japanese migrants had to be molded into Manchukuo nationals, yet without jeopardizing their de facto colonizer status. This thorny question was further complicated by the presence of so many Koreans in Manchuria—in 1932 they outnumbered Japanese three to one—who had been fiercely claimed as “Japanese nationals” and strictly forbidden from taking Chinese citizenship. This raised the question as to whether Koreans should be allowed to become Manchukuo nationals on the same terms as Japanese proper. And how should their previous relationships to Chinese and Japanese proper be readjusted? What required retuning was not just the external boundary of imperial membership, which separated Japanese (broadly defined) from non-Japanese; also up for reconsideration was the internal boundary that separated Japanese from Koreans.
No definitive answers to these questions were offered before this short-lived state collapsed. The state elites, unable to reconcile multiple contradictions that beset Manchukuo from its inception, failed to enact a consistent body of nationality and citizenship laws (ibid.: 197–329), which would have provided the “ethnotaxonomic orthodoxy” of Manchukuo, to borrow a concept from Thomas Mullaney (Reference Mullaney2011). Manchukuo thus remained saddled with multiple classification schemes that were ambiguous and mutually contradictory (Tamanoi Reference Tamanoi2000). Nevertheless, the late 1930s brought a general shift in policy from a focus on independent nation-building to an emphasis on the unity of the Japanese Empire, Manchukuo included, and this affected which population classification scheme gained primacy.
The founding spirit of Manchukuo as articulated in 1932—“the cooperation and harmony of five ethnic groups (gozoku kyōwa)”—represents the zeitgeist of the early years. In this classification scheme, Koreans were officially named as one of these five groups, along with Han, Manchu, Mongol, and Japanese. Koreans were considered separate from Japanese, and both Koreans and Japanese were declared to be integral members of Manchukuo, on equal terms (at least in principle) with the other three groups (Han Reference Han1995; see also Manshū no Chōsenjin Shidō Hōan 1933). Similar classification schemes were institutionalized in agrarian cooperatives, settlement colonies, various state-controlled associations (including the Concordia Associations [Kyōwakai] in the early years), special military units, and the official census. Many of these were organized along ethnic lines, with Japanese and Koreans in separate groups (Sin Reference Sin2002: 86–153; Park Reference Park2005: 124–61). The 1936 forfeiture of extraterritorial rights previously accorded to Japanese nationals (Koreans included) exemplifies the major trend of the early to the mid-1930s: the gradual, if sometimes controversial, progression toward the integration of Japanese and Koreans into Manchukuo in the name of independent nation-state-building (DuBois Reference DuBois2010).
Koreans responded to this early trend in varying ways. Some settlers expressed hope that their new affiliation with this ostensibly independent multiethnic state would enable them finally to sever their legal ties to the Japanese Empire, in which they were perennially relegated to the status of colonial subjects. They hoped it would secure them equal rights to land, tenancy, education, and high-ranking government positions, and even allow them to realize a form of self-rule, a decades-old demand that had been suppressed by both the Chinese and Japanese authorities (Cho Reference Cho1933: 12–13). Other Koreans were less pleased, and in fact distressed by the prospect of this possible separation and their incorporation into Manchukuo. To them, the establishment of Manchukuo had meant an opportunity to obtain recognition as full-fledged Japanese imperial subjects who would be the dominant ethno-national group in Manchuria (Kim Reference Kim1935: 3–4). The forfeiture of extraterritorial rights, the dissolution of the Korean Association into the Concordia Association, and the educational reform that put Korean schools under the supervision of Manchukuo (instead of the GGK) (Tanaka Reference Tanaka2004: 93) all provoked profound anxiety that Koreans, as a minority without political power or now even Japanese patronage, would be inevitably marginalized in Manchukuo and, worse, exposed defenselessly to Chinese hostility (Son Reference Son2001: 281–88).
Whether welcomed or feared, the complete absorption of Koreans into Manchukuo and the dissolution of their extant membership in the Japanese Empire turned out to be short-lived illusions. On one hand, Koreans were a small minority at 3 percent of the 1940 population, with less status and power, and it soon became clear that their integration into nation-building was not the priority for the Manchukuo state-builders. The Korean Association was dissolved, Japanese and Chinese were adopted as the lingua francas, and Chinese soon dominated the lower-level bureaucracy (Yoon Reference Yoon2012: 242–43). These changes inevitably resulted in the marginalization of Koreans in Manchukuo's administration, educational system, and cultural life (Yoon Reference Yoon2001).Footnote 17 Many Koreans remained illiterate peasants living in remote villages. On the other hand, as Japan sank into the mire of the ever-expanding war from the late 1930s onward, Japanese authorities ceased to worry about maintaining the façade that Manchukuo was an independent, multiethnic nation-state. They instead began to stress the leadership of the “Japanese” in building the Greater East Asia Co-Prosperity Sphere, of which Manchukuo was an integral part. In a similar vein, it was asserted that becoming a good Japanese imperial subject was not incompatible with, but rather essential to becoming a good Manchukuo national. Unlike other members of Manchukuo, Japanese were recognized as having dual nationalities, Japanese and Manchukuoan (Endō Reference Endō2010: 201–39).
How did Koreans fit into this category of “Japanese”? Various dichotomous classification schemes, which proliferated in many institutional arenas, did frequently group Koreans with Japanese, juxtaposed to other ethnic groups combined. These included: “Manchurians versus Japanese,” “natives versus Japanese,” or “persons of Manchurian origin versus persons of Japanese origin” (Tamanoi Reference Tamanoi2000). What tied Koreans to Japanese in these binary schemes was not ethno-cultural commonality or a comparable economic or political standing, but rather their common legal status as Japanese nationals, broadly defined. The sharp dichotomy between Japanese and non-Japanese was the product of the membership politics of earlier years through which all Korean migrants were made Japanese nationals despite the persistent hierarchy between Koreans and Japanese and the differences among Koreans. That is, Korean settlers born in Manchuria in the second or third generations, who were culturally sinified or who acquired Chinese citizenship (like old man Park), were still classified as Japanese (immigrants), whereas Han Chinese settlers who migrated from neighboring provinces in the 1920s were classified as “natives” or “persons of Manchurian origin.”
These dichotomous classification schemes solidified the external boundary of the imperial membership, but one should not deduce from their prevalence that the internal distinction within the category of “Japanese” (broadly defined) lost its political import or institutional grounding. After all, given the nested membership structure of the Japanese Empire, maintaining this internal distinction was consistent with the claim that Koreans were Japanese nationals via their first-order belonging to the colony, Korea. The establishment of Manchukuo did not lead to a discarding of this principle or its related legal and bureaucratic edifices. In fact, if one looks at its implementation, the establishment of Manchukuo had quite the opposite effect, reinforcing rather than weakening the compulsory nature of Koreans' belonging to the colonial state. The establishment of Manchukuo removed many restraining factors that had in previous decades crippled the state's trans-border reach, and this enabled the state not only to nominally claim Koreans in Manchuria as its own, but also to keep track of them on the ground.
There are two aspects of this process that I want to draw out. First, the establishment of Manchukuo radically transformed Korean migration to Manchuria, not only in terms of its numbers, composition, and geographic distribution (Kim Reference Kim1992), but also in the degree and means of the state's control over it. With the removal of the antagonistic Chinese authorities, Manchuria seemed to provide a vast, unencumbered space where the surplus populations in Japan and Korea could be discharged (Lynn Reference Lynn, Elkins and Pedersen2005). For security reasons, it was considered necessary to encourage Korean immigration while controlling Han Chinese immigration. Nevertheless, the ensuing “Manchurian boom” in Japan and Korea decreased rather than increased chances for unregulated migration.
From 1937, two “Colonization Companies,” one sponsored by the GGK and the other by the Manchukuo government, conducted all of the recruitment of, pre-departure training for, and settlement of Korean migrants. Migrants without immigration certificates and valid family registration documents could be blocked or deported at any time on their way to Manchuria.Footnote 18 Newcomers were thus more likely to be properly documented than were their predecessors. In addition, as part of an effort to uproot anti-Japanese guerillas, many early migrants were forcibly removed from their historic villages, relocated to state-controlled agrarian cooperatives, and issued residential certificates (Son Reference Son2001: 415). In this way the colonial state belatedly incorporated these early migrants into its documentary circuit.
A second aspect of Manchukuo' establishment that I want to highlight is that the quickening and deepening state-building there was important for strengthening the colonial state's long-distance control over hitherto undocumented Koreans. By the mid-1930s, an estimated 40 to 50 percent of Manchukuo's Korean population possessed no proper family registration documents (ibid.: 267–68). This began to change when the Manchukuo state required Koreans to provide valid family registration documents in order to, among other things, establish their rights to land (ibid.: 249–93) or gain admission to public schools (Sin Reference Sin2002: 181). This pushed many undocumented Koreans to complete their belated registrations through a simplified procedure, guided by the Korean Association (before its dissolution) and government officials dispatched from Korea.Footnote 19
The establishment in Manchukuo of the family registration system in 1940, and the residential registration system in 1943, further increased the documentation of Korean immigrants.Footnote 20 The Manchukuo government provided the resulting new information on its Korean population to the GGK and allowed it to carry out the registration of Koreans directly in Manchuria. With such assistance, the colonial government between 1941 and 1944 was able to newly register 115,721 Koreans and to update the family registries of 108,180 (ibid.: 186). In sum, the increased involvement of the colonial state in migration process, the intensive state-building in Manchukuo, and the cooperation between Manchukuo and the GGK all rendered Korean migrants in the borderland more visible and legible than had previously been possible. Migrants of peninsular origin in Manchuria came to be sorted and re-sorted as Koreans more effectively than ever before.
To be sure, some Koreans tried to capitalize on the state's ever-more-intensive efforts to “embrace” (Torpey Reference Torpey2000) Korean migrants, in order to challenge the hierarchical distinction between “Japanese” and “Koreans” (Caprio Reference Caprio2009; Fujitani Reference Fujitani2011). Rallying under the banner of naisen ittai (“the mainland [Japan] and Chōsen [Korea] are one body”) and embracing radical assimilationist policies also seemed the most effective strategy for enhancing Koreans' standing in Manchukuo. For instance, many Korean leaders petitioned for the equal application to Koreans in Manchukuo of the Korean Educational Ordinance of 1938, which, for the sake of assimilation, substantially reduced discrimination and segregation in the Korean school system (Piao Reference Piao2005: 272). By restoring the supervisory role and the patronage of the GGK (called the “motherland” in these petitions), the leaders hoped to improve the quality of Korean education, which in the Manchukuo educational system had been marginalized and under-funded. Other assimilation policies such as name changes, and voluntary enlistment or drafting into the military, were welcomed for similar reasons (Yoon Reference Yoon2001).
After all, in a society like Manchukuo in which a rigid ethnic hierarchy permeated everyday life, elevating the status of Koreans to that of Japanese might well have offended and alienated the much more numerous Han Chinese. Sensitive to the volatility of this issue, the Manchukuo government, in cooperation with the Kwantung Army, the Ministry of Foreign Affairs of Japan, and the GGK, ensured the maintenance of hierarchical distinctions between Koreans and Japanese proper in a range of institutional domains such as wages, taxation, education, and legal jurisdiction, and the food rationing that began in the early 1940s (ibid.).
Discrimination in food rationing—a blue card for Japanese, yellow for Koreans, and red for others—became the most visceral source of frustration and resentment, and left a vivid imprint on both the retrospective accounts (Chungguk Chosŏnjok Ch'ŏngnyŏn Hakhoe 1992) and fictional portrayals of the period. Novelist Mansŏn Kim's short story Kwigukcha (Repatriates), for example, describes a Korean woman, Young-ae, who liked being called Mrs. Kimura (her “Japanicized” name) and gave an impassioned speech on “Drafting and Women's Duty.” The character feels humiliated and betrayed by the discriminatory treatment she experienced as a “Korean” (Senjin/Xiaren) at a ration station. However, what concerned the authorities more than the disappointment of pro-Japanese Koreans like Young-ae was the resentment of Chinese, who reportedly insisted that Koreans use their old Korean-style names on their food ration cards to prevent them from enjoying the same privileges as Japanese (Sin Reference Sin2002: 184).
In the short story quoted at the beginning of this section, old man Park decided to stay put even after Japan's defeat, scoffing at other Koreans who, fearing Chinese riots, hurriedly left for Korea. He was fluent in Chinese, always wore Chinese-style clothes, had a Chinese concubine, and maintained a friendly relationship with his Chinese neighbors. More than anything else, Park believed that his certificate of naturalization would keep him safe by proving his full-fledged Chinese (Chunggugin) identity. But the document did not prevent a defected Manchukuo soldier from summarily killing him; it actually confirmed the soldier's suspicion that Park was a Korean who had simply obtained naturalization. Old man Park's hybrid identity as a long-time migrant in Manchuria was demonstrated by his occupation as a broker, his social networks, and his linguistic and cultural practices. But it was his categorization as a Korean, confirmed, ironically, by his certificate of naturalization, which turned out to be fatal at the end of the Pacific War.
CONCLUSION
The relationship between the colonial state and the formation of post-colonial nations has been one of the most important questions in the social scientific study of colonialism. This article has addressed this question by examining a hitherto understudied phenomenon, the colonial state's trans-border engagement with its subjects who left the territorial unit of the colony, and its impact on the formation of diasporic nationhood. My empirical focus has been the struggles over how to classify Korean migrants in Manchuria during Japan's occupation of Korea.
I began by identifying the migration of colonial subjects as an important source of population classification struggles in the Japanese Empire. These colonial migrants not only confounded the underlying principle of the plural legal order of the empire, but also hampered the implementation of this principle on the ground. In analyzing membership politics that unfolded within the context of Korean migration to pre- and post-1931 Manchuria, I have focused on identifying the tumultuous and uneven development of specific legal, organizational, and bureaucratic infrastructures that helped the colonial state extend its trans-border reach, if with varying success, in different geopolitical and ethno-political contexts. These included the implementation of the family registration system, the development of border policing and consular practices, and the formation of quasi-governmental ethnic organizations. These various infrastructures enabled the Japanese colonial state to follow its population on the move and forcibly define and identify these migrants as its own, often against suspicion, sabotage, hostility, and resistance on the part of other states, indigenous populations, or migrants themselves.
The extensive and intensive engagement of the colonial state with migrants in Manchuria contributed to the formation of Korean diasporic nationhood in two ways. First, it made the Korean nation a trans-territorial community that was legally codified, pervasively institutionalized, and enduringly documented. While the modern notion of “nation” gained traction among Korean elites prior to the annexation (Schmid Reference Schmid2002), it was the colonial state that institutionally grounded this imagined community in a distinctively trans-territorial form. The colonial state defined the boundary of “Koreans” for the first time in the language of the modern interstate system. It restricted the crossing of this categorical boundary even at the height of cross-border movement. It further made a person's Korean identity almost immutable so that neither settlement nor birth even outside the empire could dissolve it. Finally, the state succeeded to varying degrees in forcibly incorporating people on the move into this documented community of Koreans. Through dynamic and contentious interactions with Koreans and authorities of other involved states, the Japanese colonial state emerged as a public and autonomous agent that assumed paternalistic responsibility, if not political accountability, for the colonial society—one of the critical conditions for the dialectic emergence of anti-colonial nationalism (Breuilly Reference Breuilly1993: 218–29)—the boundary of which transcended the territorial borders of the colony on both the institutional and the imaginary levels. Moreover, this infrastructural scaffolding of the Korean nation—its legal definition of “Koreans,” the bureaucratic techniques it used to establish this legal identity, and the entrenched organizational practices through which it engaged with trans-border Koreans—left an enduring imprint on the nationality laws, consular practices, and minority, repatriation, and immigration policies of the region's subsequent states. This shaped the complex tapestries of diaspora politics during the Cold War and post-Cold War periods (J. Kim Reference Kim2011).
The colonial state's engagements with migrants of peninsular origin shaped Korean diasporic nationhood in a second way: they rendered abstract official categories, including ethno-national taxonomies, experientially tangible, consequential, and resonant for a broad range of migrants. This included even migrants who had left the peninsula before the onset of colonial rule, were born outside the peninsula, or were culturally sinified. To be clear, I am not arguing that migrants of peninsular origin all came to deeply identify themselves primarily as “Koreans,” or reached consensus about the desirable future of the Korean nation. Categories do not automatically produce identities (Brubaker Reference Brubaker2004). Any in-depth study of subjective identities of Korean migrants would require a more systematic examination of the cultural world they inhabited than has been possible here.
My argument has been that migrants were constantly made aware of and forced to negotiate their identification as “Koreans” by state authorities in novel ways. This was so regardless of their subjective identification with this particular national category, let alone any commitment to one of the many, and mutually competing, forms of anti-colonial, nationalist movements. Establishing identity, including national identity, by presenting papers to state agents became part of the life experience of those who participated in these migration flows. The identification document, the sine qua non of the grammar of the modern state (Caplan and Torpey Reference Caplan and Torpey2001), became a primary, and often the most authoritative medium through which they conceived, represented, and performed their identities and belonging, including their national identities. These migrants represented a broader social base than did the bilingual colonial elite or exiled nationalist intellectuals, and they circulated within the field of the empire as a whole rather than being limited to the colony's territorial unit. To reprise and extend Benedict Anderson's celebrated metaphor (1991: 113–16), we can call them colonial pilgrims writ large, people who became nationals through complex interactions with state apparatuses outside of the territorial boundaries of the colony.
What I have done here is bring the colonial state back into the explanation of nation-formation in colonial Korea, by focusing on a colonial state's quintessentially modern effort to monopolize the legitimate movement of people and render its population legible (Torpey Reference Torpey2000). This has been given little attention in the literature on Korean nationalism or that on colonial nationalism more broadly. The temporal and spatial parameters of the works of Andre Schmid (focusing on the period prior to Korea's annexation) and Richard Kim (studying Koreans in the United States) partly explain why the Japanese colonial state receded to the background in their work. John Lie, in his sophisticated discussion of diasporic nationalism, does note how colonial migrants “encountered a world of passports and identity papers” that “reinforced the sense of national borders and belongings” (2008: 172). But he quickly stresses that the complex reality of diasporic Koreans cannot be reduced to these national categories.
To construct a theory of the coeval “spread of peoplehood identification” and “the consciousness of diasporic identity” (ibid.: 172), I would argue, it is essential to examine comparatively and historically how and why colonial states developed varying classification and documentation practices vis-à-vis their mobile subjects in different contexts. We must understand how successful these practices were in mediating the reiterated encounters between migrants and the lower echelon of the colonial bureaucracy. And we must grasp the sorts of long-lasting legacies these practices left for nation-state-building and ethno-national boundary making at a regional level.
This leads to a final point—the generalizability of my empirical findings and theoretical arguments beyond the Korean case. The existing studies of migration within the realm of the empire have focused on the migration of colonizers, first from the metropole to the colony, and then in the opposite direction. They have explored how this migration contributed to empire-building, decolonization, and nation-building in both the metropole and the colony (Elkins and Pedersen Reference Elkins and Pedersen2005). Analyses of the Japanese Empire have developed a similar line of inquiry. They have examined how the migration of Japanese contributed to the mutual constitution of metropole and colony (or the colonizers and the colonized), and how it shaped the shifting meaning and the boundary of the Japanese nation in the era of imperialism and decolonization (Duara Reference Duara2003; Tamanoi Reference Tamanoi2005; Watt Reference Watt2009; Uchida Reference Uchida2011). By contrast, the migration of colonial subjects and its ramifications have remained marginal in the literature.
The Japanese Empire may stand out in terms of the prevalence of the migration of colonial subjects, the intensity of the state's involvement in their cross-border migrations, and the political salience of the jurisdictional struggles over their membership status.Footnote 21 Indeed, the colonial state's aggressive trans-border engagement with its subjects abroad—analogous to the actions of the inchoate Italian nation-state as described by Choate (Reference Choate2008)—sheds new light on the claim that the Japanese Empire should be theorized as a nationalizing empire, with all its contradictions and tensions (Caprio Reference Caprio2009; Fujitani Reference Fujitani2011). Nevertheless, the circulation of colonial subjects within empires was more common than our conventional analytic framework would lead us to believe. The incorporation of the colony into the global economy, or the influx of colonizers from the metropole, led elsewhere to large-scale displacements of colonial populations, which subsequently migrated to the metropole, other parts of the empire, or outside the empire. The migration from the Indian subcontinent to Britain and other British colonies is only the best-known example (van der Veer Reference van der Veer1995; Bates Reference Bates2001; Harper and Constantine Reference Harper and Constantine2010). The entrenched Eurocentrism in the study of the first wave of global migration and the rise of nationalism (McKeown Reference McKeown2008; Sadiq Reference Sadiq2008; and Amith Reference Amith2011)—explainable in part by a lack of reliable data—may be partly responsible for the lack of attention to this type of migration.Footnote 22
Here I have suggested that we comparatively examine the political dynamics created, not only by metropolitan settlers in the colony, but also by colonial migrants from the colony to the metropole, other parts of the empire, or outside the empire. We need to ask how plural legal orders of the empire were challenged and transformed as the colonial population crossed its internal and external borders. How did colonial states differ in the degree and modes of their involvement in migrations of their populations? How did such states define and identify the membership status of migrant populations? Which category of membership—racial, ethno-religious, national, or imperial—was given primacy over the others in particular official classification schemes, and how effectively could migrants challenge or undermine these official practices? And lastly, how did different trajectories of colonial-era membership politics shape the formation of diasporic nationhood, or lack thereof, in the colonial and postcolonial periods? The comparative study of the migration of colonial subjects and its ramifications will contribute to reconstituting the study of empire, migration, and nationhood formation in a way more sensitive to the wider, global relations in which the present nation-state is embedded (Go Reference Go2009). This article is an invitation to such a collective endeavor.
