Land Tenure and Class Relations in Colonial Korea

Land Tenure and Class Relations in Colonial Korea Clark W. Sorensen JLand tenure and the nature of rural class stratification in colonial Korea (1910-45) has long been a subject of considerable discussion. Controversies began during the colonial period itself as Japanese and Koreans used land tenure studies and class analysis to back up differential understandings of the nature of Korean society and Japanese colonialism. Outside observers—from those assessing missionary programs and appropriate Christian commitment (Brunner 1928) to foreign policy analysts trying to assess the prospects for an independent postwar Korea (Grajdanzev 1944)—also entered into the discussion . And controversies about colonial land tenure and class relations continue to the present day as Koreans, Americans, and Japanese of various political stripes have frequently found land tenure and class analysis of colonial Korea an indispensable foundation for understanding postliberâtion political events, or for justifying differing postwar policies on the Korean peninsula. From the time of the Korean Protectorate of 1905—that developed into a full-fledged colony in 1910—the Japanese, of course, had pragmatic, political, and ideological reasons for studying land tenure and class relations in Korea. Pragmatically, they needed to know how to control land and labor to develop Korea for the benefit of the 35 36Journal ofKorean Studies Japanese Empire as a whole. Politically, they became concerned with rising numbers of tenancy disputes and their potential for giving rise to anti-Japanese rebellion. Ideologically their view of the "backwardness " of traditional Korean land tenure and class relations and the "modernizing" nature of colonial policy provided important legitimation for a regime whose international image was always a matter of concern to authorities (Chen 1973:253) intent on building an empire at a time when imperialism had begun to become unfashionable elsewhere (Cumings 1984:484). The views of Hisama Kenichi, one of the Japanese leading scholars of colonial agriculture, can best be understood in this light. He characterized rural Korea as being dominated by static yangban landlords who had inherited their land from the past and who "simply consumed the rents they received year after year, generation after generation, were not pioneers who by themselves created the dynamic activity of positive agricultural development , but were simply persons who hibernated in the world of a static economy of the pastjust as it was, and were incapable of something like taking charge of the productive activity of positive agriculture " (Hisama 1939:337). THE JAPANESE AND KOREAN LAND TENURE From the beginning of colonization, at least some Japanese had had grand plans for Korea, and Koreans early on became concerned about the effects of Japanese colonization on land tenure. A 1904 Japanese demand that uncultivated land in Korea be made available to Japanese colonists caused a storm of protest and was withdrawn (K. B. Lee 1984:308). Early plans to raise living standards in rural Japan by exporting surplus population to Korea for large-scale agricultural colonization—a task for which the Oriental Development Company (Töyö takushoku kaisha) was organized in 1008—were never realized on the scale they had originally been conceived as Japanese became aware that virtually all of Korea's potential agricultural land was already occupied—and densely—by Koreans (Moskowitz 1974). The similarity of Korea's environment and agricultural economy to that of Japan, however, quickly convinced the Japanese that technical improvements being made in Japan could easily be adapted to the indigenous Korean agricultural system that could then be made a source of cheap raw agricultural materials for Sorensen: Land Tenure and Class Retenons in Colonial Korea37 the rapidly industrializing métropole,l so even apart from the Oriental Development Company, Japanese ownership of Korean agricultural land had reached almost 90,000 hectares by 1910 (Ko 1977: 322). To confirm these titles and to provide a basis for the general commercialization of the Korean agricultural economy, theJapanese began their rule with a complete cadastral survey in which all land holdings were surveyed, graded, and registered between 1911 and 1918. < By 1927 Japanese holdings of agricultural land had quadrupled to some 6 percent of Korea's agricultural land (H. K. Lee 1936:147), and tenancy rates were on the rise. Ideologically uncommitted scholars saw the exposure of this situation primarily as a key to unmasking the hollowness of Japan's ideology of modernization in Korea, and exposing to a complacent world the fundamentally exploitative nature ofthe colonial regime. In his 1936 study done for the Institute of Pacific Relations, thus, Hoon K. Lee (Yi Hun-gu) noted that Korea was used "as a profitable investment colony for large capitalists in Japan ... in opposition to the national sentiment and racial feeling of the Koreans" (1936: 144), and that this policy had led to the "death" of rural Korea (ibid.: 274). Many Koreans had became convinced, following Pak Mun-gyu (1933), that the cadastral survey had been the basic cause of this. In Chöng-sik argued, for example, "Even for what could be called small-scale private land, the peasants, considering the customary conditions of the time, would have had no reason to know what private property rights were, and furthermore would have been totally incapable of carrying out sufficient reporting procedures in the allotted time. In truth, this "reportism" was a most wickedly plotted trap established by plan in order to ruthlessly, and on a large scale, strip away the small private plots of the peasants and the village common lands" (1949:52). THE MARXIST VIEW By the late thirties a strong Marxist influence on the independence movement had led many Koreans to think not simply about Japanese takeover of Korean land, but of the class analysis of rural Korea in general. The very fact that Korea had fallen prey to Jap1 . For example, Korea's rice, being ajapónica variety, was considered more palatable to the Japanese than imported indica varieties from Thailand, or even Taiwan. 38Journal ofKorean Studies anese colonialism proved there had been a basic weakness in Korean society. Features of Korea's class system—such as high tenancy rates and "parasitic" landlordism—might well account for Korea's demise, but such analyses tended at the same time to confirm at least some aspects of traditional Korean class relations that the Japanese were using tojustify their rule in Korea (cf. Hisama 1939). Fully committed Marxists, like In Chöng-sik, went further and sought in the analysis of colonial Korea as a semifeudal2 society into which capitalist relations of production had been introduced, a way of identifying progressive and backward elements in Korean society, as well as clues on how to promote the revolution they saw as the best hope for Korean liberation and development. Thus, in the classical fashion of Kautsky and Lenin, In saw the penetration of capitalism into a "natural" "feudal" rural economy creating large-scale proletarianization and concentration of landholdings in the hands ofJapanese and Korean landlords, eventually resulting in a system of "semi-feudal debt peonage" (han högenteki saimu nödotaiteki nödosei). Grajdanzev in less strident tones also characterized rural Korea as dominated by parasitic large landlords largely responsible for the growing misery of the peasants. No Marxist, in this Grajdanzev was simply following the analysis of the Japanese economist Takahashi Kamekichi who blamed the "resourcepoor ," "technologically undeveloped," "crudely managed," "precapitalist " condition of Korean agriculture largely on the "extremely vicious parasitic nature" of the tenancy system dominated by landlords developed during the Chosön period when officials had turned prebends into privately owned estates (1935:187). Takahashi's analysis of Korean rural conditions made the old society look so bad that the Japanese could only been seen as doing the Koreans a favor by taking them over. POSTWAR CONCERN WITH RURAL CLASS RELATIONS Korean, Japanese, and outside investigators of rural class relations in colonial Korea never came to full consensus, but they all tended to focus on backward landlords and miserable tenants. These 2. He defines a feudal social formation (högenteki kôsei no shaL·^ as "the inhabitants main production being agricultural, the society's basic classes are differentiated into feudal lords and subordinate peasants, and the general form of appropriating surplus labor is through high rates of labor rent (corvée labor) and rent in kind (tribute)" (1939:1). Sorensen: Land Tenure and Class Relations in Colonial Korea39 concerns turned out to be more than academic. At the time of liberation fromJapanese rule, Korea was occupied by the Soviet Union in the north and the United States in the south: two powers who held decisively different views on what type of land tenure regime and society was desirable for postwar Korea. Amongst the Korean leadership , too, there was disagreement. Although some of the most conservative postwar leaders, such as Syngman Rhee, viewed existing land tenure and social structure as unproblematic, most people assumed that, at the very least, the most serious of the social problems created during the colonial period—among which rural poverty and high agricultural tenancy rates were the most prominent—would be addressed through reform. Many on the Left felt only radical changes in Korean society could make the Korean economy viable and lead to a strong, stable society. Controversy and agitation over questions of land reform and class composition broke out immediately upon liberation in 1945. Many of the same people who had taken an interest in the social structure of rural Korea in the thirties—including Hoon K. Lee and In Chöng-sik—became spokesmen for various postwar policy positions. The political maneuvering and agitation of postwar Korea is not the theme of this paper, but it is important to note that the problems of class structure and land tenure were important and intractable enough in rural Korea that they were brought to a temporary resolution in the north and south3 only after Korea made a spectacular debut onto the world stage in a bloody, internationalized civil war. Attempts to understand this civil war have often focused on Cold War tensions and foreign relations—which indeed were important—but recent scholarship has refocused attention on the domestic political conflicts in the period immediately after liberation (Cumings 1981), and questions of colonial period land tenure and class structure were at the center of this conflict. In addition to this, postwar South Korean scholars have often traced the effectiveness, or lack thereof, of the land reform of 1950 (P. T. Kim 1981) and subsequent agricultural policy pursued by the South Korean government from their analysis of class relations and land tenure during the colonial period. 3. In the north, agriculture was collectivized after the Korean War and has remained within a basically unchanged structure since 1960. In the south, land reform was completed after the Korean War, and agriculture has developed steadily there on the basis of small-holding peasants. These solutions must be regarded as "temporary" simply because the division of the Korean nation is a solution which over the long run cannot be maintained. 40Journal ofKorean Studies POSTWAR APPRAISAL A few postwar scholars, most prominendy Kim Yong-söp, have argued that an indigenous movement toward capitalist farming that had begun in the late Chosön dynasty among "managerial-type rich farmers" (kyöngyonghyöng punong) continued during the colonial period. Gragert (1982), in the only empirical study of land tenure before and after the cadastral survey, has found no evidence that the cadastral survey itself led to large-scale transfer of land to Japanese (although the economic conditions of the thirties did seem to have had that effect). Most recent scholars in South Korea who have taken an interest in class stratification in rural colonial Korea, however, have, like the Japanese, emphasized the "backward" nature of Korean landlords and conflict between these large landlords and tenants. Kim Yöng-mo, relying in part on Hisama Kenichi's work, characterizes prewar ethnic Korean landlords as parasitic exploiters of the peasantry through high rents and usury who lived on "the expropriation of surplus labor value and the penetration into the villages of usurious merchant capital" (1971:111-12). An Pyöng-jik makes part of the definition of landlord that they are "simple appropriators of rent" (1975:2). Kim Pyöng-t'ae, preliminary to criticizing the land reform of 1950, characterizes Korean landlords as absentees who condoned the arrogant actions of their agents (P. T. Kim 1981:3). Cumings, too, though willing to concede that both "backward" and "progressive" landlords existed in colonial Korea emphasizes continuity of the colonial landlord class with the yangban gentry of the past, and the contradiction ("lord" and "peasant" ) between large landlords and tenant farmers being proletarianized due to a rising market economy. The tendency in the literature to emphasize the concentration of land ownership in the hands of a few, and the contradiction between these landlords and the tenants who worked the land certainly captures a central aspect of rural class relations in colonial Korea, but it does so at the cost of subtlety, balance, and, ultimately, to an overassessment of the revolutionary potential of postwar rural Korea, and an underassessment of the potential of the system for reform. Although there is no question that landholding was quite concentrated, that the typical peasant of the colonial period was, in fact, a tenant farmer, that more and more Korean agricultural land was coming underJapanese ownership, and that the living standard of the peasantry was falling, these trends do not exhaust the structurally important fea- Sorensen: Land Tenure and Class Relations in Colonial Korea41 tures of rural Korean class structure. In fact, by accepting at face value the self-interested Japanese assessment of rural Korea, and uncritically usingJapanese statistics on the "class" characteristics ofrural Korea,4 crucially important aspects ofthe colonial Korean class system are overlooked. Small landlords tend to be assimilated to large landlords , even though they have more in common socially with ownercultivators . The "people" tend to be equated with tenant farmers, even though farmers who rented all or some of their land were, as is described in detail below, a very diverse group with much in common with owner-cultivators and even small landlords. Very significant regional differences in the distribution of land and the type of land management are glossed over (Miyajima 1982). And owner-cultivators , landlord's agents, and agricultural laborers—three classes absolutely crucial to the social and political structure of colonial rural Korea—tend to get ignored altogether. It seems implausible that villages that today have a variegated class structure were reduced to a basic two-class structure before 1945—especially since I have met and interviewed old people who give me a much more complex picture of class relations during that time. I have come across people in rural villages who inherited no land, yet were able to begin buying land before 1945. And if the Korean agricultural system was characterized only by parasitic landlords who did not invest in agricultural improvements and tenants who could not invest in agricultural improvements, we are left with no explanation for the rapid increase in agricultural productivity—both absolute and per capita—during the colonial period , and the rapid growth of small-holding peasant agriculture in South Korea after 1955. LANDLORDS AND TENANTS RECONSIDERED Although they differ in details, as mentioned above, most of the analyses of rural colonial Korea revolve around the contradictions between landlords and tenants, who are presented as having implacably opposed interests and no significant classes mediating between 4. These classifications were not uniform throughout the colonial period. In 1932, for example, the classification was altered to cease distinguishing between Japanese and Korean landowners, between cultivating and noncultivating landlords (landlords A and B), and to add the category of agricultural laborer. The discontinuance of distinguishing ethnicity and landlord type was clearly a response to increasing political action among die peasantry from 1927 on. 42Journal ofKorean Studies them. Leaving aside the Japanese, who comprised some 9 percent of the landlords holding five chô or more (Miyajima 1982:174) in 1942, Korean landlords (chiju) are presented as a parasitic leisure class closely cooperating with the colonial state to extract agricultural commodities from the peasants. Tenant farmers (sojagin) are presented as an exploited proletarian class living on the edge of subsistence with no possibility ofaccumulating a surplus. The abject, and worsening, poverty ofthe peasants during the colonial period is seen as a result ofthe land tenure system, and since for neither class is accumulation possible —the one because of its parasitic life style, the other because its entire surplus is appropriated by the landlords, there is no hope for indigenous investment in agriculture and the development of a productive capitalist agriculture. There is no question that parasitic landlords existed in colonial Korea, or that most tenants lived in dire and worsening poverty. Analyses of rural Korea that stop with this observation, however, rely on too-simple assumptions about the uniformity of "classes," and are possible only by lumping together distinct categories and then assuming the "worst case" gives the "essence" ofall the categories so lumped together. Parasitic landlords, of course, are those who live off their rental income without making a contribution to the agricultural economy —through direct cultivation, management and improvement of their land, or other productive activity. This parasitism is most clear in the case of large, absentee landlords who have inherited their land from past generations, yet relatively few of the colonial Korean landlords were parasitic in this full sense. If one distinguishes, as the Japanese did until 1932, between landlords who rented out all their land (jinushi ko), and those who rented out most of their land but cultivated a portion of it themselves (jinushi otsu), in most years the second type of landlord was three to four times more common than the former.5 Although it is noted by many writers that Korean landlords were often absentee, preferring to live in Seoul or county seats (up) rather than on their land, this was true primarily of the larger landlords (Zenshö 1929:235). In 1933, 69 percent of landlords were living in the county where their land was situated (Kobayakawa 1944 1:63), and Kim Yöng-mo, in a "reputational" study of 233 large colonial landlords, classified 72 percent ofthem as village resident. This is consistent with 40 percent of the tenanted land being managed by agents in 1935 (a good sign of absentee landlordism) (Nörinkyoku 5. After 1932 this distinction was no longer made by the Government General. Sorensen: Land Tenure and Class Rehtions in Colonial Korea43 1937:90). As for whether a landlord's lands had been inherited, Kim Yöng-mo's study shows fewer than half of the large landlords had formed their estates before 1920. Miyajima Hiroshi, in fact, after viewing reports of the 1930 Tenancy Custom investigation on the time of formation of landlords' estates concludes: ... as for the colonial period ethnic Korean large landlords, rather than being yangban landlords from the Yi dynasty who are thought of as going way back in tradition, the mainstream has been those who appeared from the opening of the ports [1876] to the 1920s. Those seen as large landlords inherited from their ancestors in parts of Kyönggi, South Cholla, North Kyöngsang, South P'yöngan, North P'yöngan, and Anbyön County, South Hamgyöng, [mentioned in the Tenancy Custom investigation] on the contrary are the minority, and the aforementioned three Korean writers' [Kim Yöng-mo, An Pyöng-jik, and Kim Yang-sap] conclusions [that most large landlords appeared after 1876] are supported by the results ofthe 1930 Tenancy Custom investigation (1982:174). Far from being large, absentee, yangban landowners who inherited their land from the distant past, the great majority of landlords, in fact, were small, village-resident, and had acquired their land within the past few decades. Among rural people I have interviewed on the colonial period, the terms "landlord" (chiju) and "tenant" (sojagin ), while they are known to everyone, are not spontaneously used to characterize prewar class strata. Rather people use terms that classify households by their approximate annual incomes in rice: a landowner of no particular pretensions might be called a "hundred bagger " (paeksokkun), but great landlords were "thousand baggers" (ch'onsokkun) or even "ten thousand baggers" (mansokkun). Miyajima estimates that a thousand bagger needed at least 1006 hectares of land, a condition met in 1942 by only 1,378 of more than 100,000 landlord families (Miyajima 1982:170; Kobayakawa 1960 3:93). The Government General considered a landlord owning five chö or less a small landlord. Such a landlord would have a rice income of less than 50 bags a year and have had a small enough holding that, ifhe wanted to, he could manage it himselfwithout using agents—even if he used hired labor rather than tenants. In 1942 these small landlords owned some 60 percent of the tenanted land (Miyajima 166—67; Kobayakawa 1960 3:99) or about 38 percent of the agricultural land in Korea, while large landlords owned only about 40 percent of the 6. Kim Söng-mo estimates a thousand bagger needed only 30 chöngbo or so of land, but Miyajima's estimate corresponds better with known productivity during the Japanese colonial period. It could be that die "fold" definition of "thousand bagger" didn't require an actual annual income of 1000 bags of rice. 44Journal ofKorean Studies tenanted land, or about 25 percent of the agricultural land in Korea.7 And while these large landlords generally had agents manage their land for them and to this extent must be considered "parasitic," only half of them inherited their land from before 1920. Small landlords were not necessarily less exploitative than large landlords. In fact the opposite may well have been true, but socially they were enmeshed in village society in a way not much different from other village agriculturalists. Unlike the large landlords, who tended to live in cities and leave the management of their lands to agents, the small landlords tended to live in the villages and manage their own land. Their living standard was not particularly high. Figures from 1925 indicate that the income of these small landlords was comparable to, and often lower than, that of middle and upper owner-cultivators and even some of the owner-tenants (Zenshö 1929: 35-37). Although they lived from the management rather than the cultivation of their land, and controlled access to land for some tenants , in terms of kinship and life style, they would have been little different from other villagers, and would have participated in the same moral community. It seems likely, in fact, that after 1937 when large amounts of labor began to be impressed from rural Korea to support the Japanese war effort, a substantial proportion of the new tenancy represented short-term adaptations to labor shortages by middling landed peasants rather than simply proletarianization and the further development of a large-scale parasitic landlord class. Landlords, then, were a diverse group. Although perhaps a quarter of the agricultural land was owned by classic large-scale absentee landlords, only half of these large-scale landlords seem to have been yangban landlords who inherited their landlord status from the Yi dynasty. The amount of land controlled by "traditional, absentee, large yangban landlords" is likely to have amounted to some 10 to 15 percent ofthe agricultural land—a significant amount, to be sure, but hardly enough to account by itself for the main features of rural class structure. Small landlords enmeshed in ordinary village life were the vast majority of landlords, and the available evidence seems to indicate that most of the tenanted land was under their—rather than the large landlords'—control. 7. Because the Government General quit publishing statistics on landlords after 1932, 1 have had to estimate die total amount of tenanted land from the total amount of agricultural land in 1936 and the estimated tenancy rate of 1945 made by Pak Ki-hyuk. Sorensen: Land Tenure and Ctess Retenons in Colonial Korea45 Those who cultivated the landlords' land were even more diverse than the landlords. Tenants, those who maintain an independent household economy based on rented land, have been seen as basic in colonial Korea for three reasons: (1) they represent the labor component ofagriculture and thus complement the position ofthe landlord/ owners; (2) in the statistics collected by the Japanese they are always the largest group; (3) there was a secular trend throughout the colonial period both for the proportion of tenants in the population to grow at the expense of all groups except the landlord, and for more and more land to become owned by ethnic Japanese. These trends were so pronounced that it is estimated that by 1945 at least 10 percent of Korea's agricultural land was owned by Japanese, and more than 60 percent of all the agricultural land in Korea was tenanted (Pak Ki-hyuk 1966:88). Under these conditions, most writers have feltjustified in following the procedure of Hoon K. Lee (Yi Hun-gu) (1936) and lumping owner-tenants together with pure tenants so that 80 percent of the rural population would be considered members of the tenant class. More recently, Cho Tong-göl, writing on the peasant movement in colonial Korea (1978), after identifying the Korean nation with the peasantry (nongmin), uses this procedure to then identify the peasantry with tenant farmers. Since many of the landlords wereJapanese (though they controlled only about 20 percent of the tenanted land), the entire class of ethnic Korean landlords—large and small—are assimilated with the Japanese, and so excluded from the peasantry. Owner-cultivators and owner-tenants are ignored, so that, in effect, the class of tenant farmers are elevated to the essence of the Korean nation during the colonial period. As a class term, however, "tenant" has the same characteristics as "landlord": that is, it is an agglomeration of miscellaneous groups of people who have distinctive economic positions and strategies. As a term, "tenancy" (sojak) refers simply to land having been rented out (sojak-sik'ida) or in (sojakhada), and only secondarily refers to the status of one who rents such land. Making the distinction between land tenure classifications and class classifications is not merely a linguistic nicety; it reflects the reality that in rural areas there is no one-to-one correlation between land tenure, economic strategy, and class consciousness . The word "tenant" (sojagin) on the one hand can refer to the classical poor sharecropper who rents all of his land and, having no surplus, lives at the margin as a debt peon. But these were not the only tenant farmers. As recognized by In Chöng-sik, there were oth- 46Journal ofKorean Studies ers who managed tenant farming on a large enough scale to live as well or better than most owner-cultivators (In Chöng-sik 1939:112— 19). Some 40 percent of the tenanted land in colonial Korea, moreover , was not cultivated by pure tenants, butby owner-tenants.8 These are an even more miscellaneous classification than pure tenants, because along with thefact of tenancy, the level of rent, and the scale of production determines one's economic position and thus class interest . The smaller owner-tenants for the most part were poor peasants who did not have enough land of their own on which to subsist, and thus had to rent more to get by. They were often unable to make ends meet, highly liable to proletarianization, and indeed were not much different from pure tenants. The larger owner-tenants, on the other hand, were in a much better situation. Many of them were owners of substantial amounts of land who rented even more to expand their scale of production with hired labor and accumulate a surplus. They often lived as well as, or better than, the smaller owner-cultivators (Zenshö 1929:38). Far from renting land due to proletarianization, in fact, these larger owner-tenants were more likely attempting to accumulate a surplus through increasing the efficiency ofhousehold labor by enlarging the scale of production. This strategy should be seen as forward-looking rather than defensive. It is clear, then, that though parasitic landlords and miserable tenants were commonly found in colonial Korea, emphasizing only the conflict between the parasitic landlords and exploited tenants seriously oversimplifies rural class structure. Demographically and economically important intermediate groups—not only the ownercultivators , farming barely half the agricultural land and getting smaller each year,9 but also small landlords, and expanding owner tenants—each with a distinctive relationship to the means of produc8 .No statistics are available on the proportion of tenanted lands cultivated by pure tenants and and owner/tenants. Estimates were made based on the scale of production of the two groups in 1925 (Zenshö 1929:33). The number of pure tenant households was multiplied by the midpoint of each of the farm-scale categories to get an estimate of the total amount of land cultivated by pure tenants. The same was done for owner-tenants, owner-cultivators, and landlord-cultivators. The total amount of land accounted for in this was 85 percent, so each category was arbitrarily increased by 15 percent. Then the amount of land cultivated by the pure tenants was subtracted from the total amount of tenanted land in that year to get the amount of tenanted land cultivated by owner-tenants. 9.Self-cultivated land fell from a high of 49.8 percent of die land in 1920 to a low of 36.7 percent of the land in 1945. Sorensen: Land Tenure and Class Relations in Colonial Korea47 tion, played important roles in the colonial Korean countryside. These five groupings, moreover, do not exhaust the classes of rural Korea. LANDLORD'S AGENTS In addition to the categories based on land tenure noted above there are other economic groups that villagers mention as having existed during the colonial period, but which, since they are rarely mentioned in government statistical sources, seldom figure prominently in class analyses of colonial Korea: marüm (landlord's agents), nonggam (agents, especially of the Oriental Development Company), mösüm (agricultural laborers who do not maintain an independent household), koyonggun (agricultural laborers who do maintain an independent household), kojigi (keepers of rich persons' storehouses), sanjigi or myojigi (those who maintain yangban families' tombs and prepare seasonal ancestor worship rites for these families). These last three are especially interesting, because informants tell me that even children used panmal—language unmarked for respect—with adults of this status. For reasons of space all of these interesting social categories cannot be investigated here, but I would like to discuss in more detail one that has especially significant implications for the class analysis of rural colonial Korea—the landlords' agents. There are various terms for these in Korean: the term used in documents and writings was normally saüm. The Oriental Development Company called its managers nonggam, but the colloquial term by which the majority of landlords ' agents in Korea were known by peasants is marüm, and that is the term I shall use here. The institution of the marüm is very old in Korea, predating the colonial period by centuries. Their role was to manage tenanted land for landlords. They were not used by all landlords who rented out land, but only those who had too much land to personally manage, were engaged in other occupations and thus took no interest in the management oftheir land, were absentee landlords, or were incapable for personal or other reasons of dealing directly with tenants (Zenshö 1929:236)— that is, the "parasitic" type of landlord . Although the responsibilities of the traditional marüm could vary with the whims of the landlord, they typically took care of all aspects ofmanaging tenants: selecting and removing tenants; setting, inspecting , and taking custody of sharecrops; inspecting tenanted land; investigating bad crop yield and making appropriate rent reductions; 48Journal ofKorean Studies remitting taxes for the landlord; supervising repairs and improvements on the tenanted land; and dissolving, if need be, tenancy contracts (ibid.: 236-37). As the marüm was the person tenant farmers direcdy dealt with, he was a powerful, often feared person. Old residents of Moünalli, a yangban village I have investigated in South Ch'ungch'öng Province whose land was nevertheless almost entirely owned during the colonial period by a very large absentee landlord, recall that it was necessary to curry favor with the marüm, whose name and address they still remember, by providing him with a banquet whenever he dropped by. Nobody, in fact, has anything good to say about the marüm. Japanese works on Korean agriculture during the colonial period usually place discussions of marüm in a section devoted to the problems and abuses of the tenancy system. One of these works introduces marüm, for example, by noting that they "have traditionally been called in Korea a cancer among tenancy practices" (kosaku kankö naka no gan) (Nörimkyoku 1937:90). The reason for this, as noted by In Chöng-sik (1939:114) is that, although they represent the landlord to the tenants , they also represent themselves and are in a position to profit, at least surreptitiously, from their position. Although the people I have interviewed in Moúnalli reported that their marüm simply and openly added a charge of 5 percent on top of the rent—that is, charged tenants five toe of rice for every sok paid to die landlord—a method that is said to have been typical of South Ch'ungch'öng Province (Zenshö 1929:242), others obtained their compensation less direcdy. Zenshö notes that in Kyönggi Province where, because ofproximity to the capital, marüm managed an especially large proportion of the land, some were allowed to choose sufficient fields for themselves to cultivate rent-free in lieu of payment. This method of payment, in fact, was widely found throughout Korea (ibid.:241—45). Although the demand for payments on top of already high rents is one reason marüm were considered a cancer, the main reason was their potential for corruption. Zenshö notes "their power over the tenants being really very strong, they aren't any different from actual landlords. And since to a tenant a single frown or smile from a marüm translates directly into a big influence on his welfare, they are in a situation where they must diligendy complying with his wishes" (1929:241). "Sayings about them abound: "In one generation, or in three years, a marüm can become rich and powerful" (Nörinkyaku 1937:613), or "They give in pecks, but receive in bushels" (P. J. An 1975: 1 13). They typically seem to have extorted gifts and services out of tenants because of explicit or implied threats to raise rents or Sorensen: Land Tenure and Class Relations in Colonial Korea49 dissolve tenancies. They are accused of exchanging for the landlord's good quality rice poor quality rice that they have bought at the market , of dispossessing old tenants in favor of relatives or those who have given them gifts. Zenshö notes that "Agents constandy give completely or mostly over to themselves or their relatives the best land, and keep special books with lowered rents and proportions . . ." (1929:242). We know, of course, that there was not enough land available during the colonial period to absorb all labor available in rural areas. As population grew and urbanization took place at only a modest rate, landless farmers were desperate for tenancies. Though tenancy normally provided only a bare living, the alternatives—starvation, begging , fire-field farming, emigration to Manchuria or Japan—were worse. Under these circumstances, the potential of marüm to arbitrarily play around with tenancy arrangements could be a threat to one's family's very survival. It is no wonder, then, that the general consensus was that they were considered "a cancer" in the tenancy system. There is also another less well-known story to tell about marüm, however. Absentee landlords, despite their "parasitic" relationship to agriculture, may not have been the most serious exploiters of the land hunger in rural Korea. We have already noted that their concern seems to have been primarily for a fixed and assured income. Some writers, in fact, claim that rents of absentee landlords frequently were lower than those of the other, managerial landlords. Zenshö notes as one of the abuses of absentee landlordism that "... absentee landlords have no experience in agriculture, are occupied with commerce or other activities in the cities, and when they do not engage an appropriate manager, it frequently results that the tenants, who know the owner is unable to cultivate the land himself, request lower rents, and as these people's rents tend to be lower than others, their influence reaches to other landlords and entails great harm" (Zenshö 1929: 237). If we put this together with what we have already said about the importance of high rents, the scale and tenancy and the position of marüm we can identify people with a chance for accumulation that is not apparent from the statistics. For most tenants, the high rents and the inability to get enough land to fully use all of the available family labor resulted in virtually all of the surplus they produced being appropriated by landlords, landlord's agents, and the state. Marüm, and judging from the sources, in many cases relatives and/or friends of marüm seem, on the other hand, to have been in a position where 50Journal ofKorean Studies they could, corruptly perhaps, overcome these defects of the agricultural system in colonial Korea and accumulate capital for investment in rural improvement. Absentee landlords were dependent upon the marüm for their information about rents, and may not have been able to maximize them to the same extent as managerial landlords, while marüm seem regularly to be in a position to increase the scale of their agriculture to the point where a surplus might be accumulated. In Chöng-sik, in fact, argues that even marüm that had no land of their own often had a better income than small landlords. In addition to fees for his services, he often cultivated with hired labor, or even uncompensated labor service by tenants, land that he got free or rented at a nominal rate from the absentee landlord. His family labor, rather than being used in agriculture, could be used in side occupations such as sericulture or handicrafts. In addition, as he frequently had occasion to visit the city to remit sharecrops or consult with his landlord, he could be familiar enough with market conditions to take advantage of commercial opportunities and successfully engage in truck farming, fruit orchards, or other forms of purely commercial agriculture (In 1939:116-17). CONCLUSION This paper has had the purpose of beginning a more nuanced and complex notion of the rural class structure ofcolonial Korea than those conventional accounts that stress only the exploitation of landlords , the misery of tenant farmers, that equate tenants with the peasantry (nongmin) as a whole, and that only note the proletarianization ofthe middle peasants through low crop prices, high rents, and usury introduced with the commercialization of agriculture introduced by the Japanese. My purpose is not to exonerate the Japanese from exploitation, or to deny that these processes were important elements in rural, colonial Korea, but to see what other facets there were to rural stratification during the colonial period. I find the above characterization of rural colonial Korea inadequate not because I think it totally inaccurate (it is not), but because it overestimates the potential of the countryside for revolution, and underestimates the potential of the system for reform and development . A long line of scholars (Stinchcombe 1961; Moore 1966; Zagoria 1974) has fruitfully argued that small-holding tenancy systems with "parasitic" landlords are likely to lead to revolutionary move- Sorensen: Land Tenure and Class Relations in Colonial Korea51 ments because of the exploitative nature of landlord-tenant relations. It is only natural, then, when analyzing the concrete conditions of a particular society; such as Korea, to focus on the two classes locked in such an explosive relationship. If other important elements of the rural class situation that mediate between parasitic landlords and proletarianized tenants are thereby ignored, and if the relative size and importance of the various classes is not taken into account, however, only two political alternatives—revolution or repression—seem possible in a situation of disorder, when a more complex class analysis might reveal rapidly shifting class-based coalitions combining and splitting around concrete issues for which satisfactory compromises might possibly be found. This political question, of course, is closely related to the economic one, for if class situation in rural Korea had been dominated entirely by parasitic landlords and marginally productive peasants, then little potential for economic development would have existed in the system. This potential can be inferred for colonial Korea, however , not only because of the agricultural growth (both absolute and per capita) during the colonial period, which took place despite falling living standards for the majority of peasants. It is also implied by the capacity of South Korean agriculture for development after Liberation and land reform in the south. According to the conventional picture of rural, colonial Korea, capitalist elements could not have developed in the countryside because there was no class, with the possible exception of certain Japanese , capable of accumulating a surplus and willing to invest it in agricultural improvement. In general this seems to be correct assessment for the majority of Korean peasants. On general principles, however, when new capitalist elements emerge in a predominantly precapitalist society, one would expect these elements to emerge in the cracks. That is, it is not among the major classes of landlord or tenant confronting each other in the countryside where one should look for the emergence of capitalist, commercial agriculture, but in intermediate groups that may not yet have developed a full-fledged class nature. For this reason the statistics on land tenure published by the Japanese, unless subjected to clear-headed criticism, are less useful than they appear on the surface. Considerable ink has already been spilled on the question of whether entrepreneurial landlords were important in colonial Korea, and I have not dealt directly with this question here. Rather I have attempted to identify another rural group, marüm, who would be in a 52Journal ofKorean Studies position to accumulate capital and invest it. Their numbers should not be underestimated. Classified Government General reports that became available after Liberation indicate that in 1935, marüm numbered more than 66,624 persons (Nörimkyoku 1937:93), which, if we assume each agent resided in a separate household, represented about 2 percent of the agricultural households at that time—a percentage of the population quite comparable to that of managerial landlords (jinushi otsu) who in 1932, the last year in which they were listed separately from noncultivating landlords, numbered slightly less than 72,000 persons. The land the marüm managed comprised some 40 percent of all tenanted agricultural land (or 20.8 percent of all agricultural land) in Korea. This land, moreover, was concentrated primarily in Kyönggi, North Cholla, and South Kyöngsang provinces —precisely those areas near major cities where strategies of truck farming, and/or export-oriented commercial agriculture were most highly developed. It seems likely they were an incipient commercial agricultural class. In addition to these marüm, managerial landlords who cultivated on a relatively large scale with hired labor (mösüm) rather than renting out their land to tenants should also have been able to accumulate capital and invest it in agriculture. They may, thus, also be added to those who were developing, if only incipiently, a productive capitalist form of agriculture, though for want of space I. have not developed this theme here. EPILOGUE The story of the marüm and managerial landlords does not end with Liberation and land reform. It is said that at the time of land reform in South Korea, it was often the marüm rather than the tenant who was able to buy absentee landlords' land. In addition, landowners who cultivated their land through hired labor, the managerial landlords , were allowed to keep up to three hectares of agricultural land, while those who rented it out lost most of their land. It seems quite likely, then, that the relatively large-scale farmers that can been found in rural Korea today are the descendants of the colonial marüm and managerial landlords. But this is a different story. BIBLIOGRAPHY An Pyöng-jik. 1975. "Singminjiha Chosönin taejiju úi yön'gu" [A study of large Korean landlords during die colonial period]. Kyongje Nonjip 14, no. 1:1—18. Sorensen: Land Tenure and Class Relations in Colonial Korea53 Brunner, E. Schweinitz de. 1928. "Rural Korea." In The Jerusalem Meeting of the International Missionary Council, March 24—April 8 1928. Vol. o! Chen, Edward I-te. 1973. "Japan: Oppressor or Modernizer." In Andrew Nahm, ed., Korea underJapanese Colonial Rule, pp. 251—60. Kalamazoo Center for Korean Studies, Western Michigan University . Cho Tong-göl. 1978. ïlcheha Han'guk nongmin undong sa [A history of the Korean peasant movement under the Japanese Empire]. Seoul: Han'gilsa. Cumings, Bruce. 1981. The Origins ofthe Korean War: Liberation and the Emergence of Separate Regimes, 1945—47. Princeton, N.J.: Princeton University Press. _____ 1984. "The Legacy ofJapanese Colonialism in Korea." In Ramon Myers and Mark Peattie, eds., TheJapanese Colonial Empire, 1895—1945, pp. 478-96. Princeton: Princeton University Press. Gragert, Edwin H. 1982. "Landownership Change in Korea under Japanese Colonial Rule: 1900-1935." Ph.D. diss. Columbia University . Grajdanzev, Andrew J. 1944. Modern Korea. New York: John Day Co. for Institute of Pacific Relations. Hisama Kenichi. 1939. "Chosen ni okeru chinushiteki shokunö no chösei" [The regulation oflandlord performance in Korea]. Shakai Seisaku Jibö [Social Policy Bulletin], no. 229, October 1939. (Reprinted 1943 in Chösen nösei no kadai [Themes in Korean agricultural administration]). Tokyo: Seibidö Shoten. In Chöng-sik. 1939. Chösen no nögyö kikö [Korea's agricultural institutions ]. Tokyo: Hyakuyösha. _____ 1949. Chosön nongöp kyöngje non [The Korean agricultural economy ]. Seoul: Pangmun Ch'ulp'ansa. Kim Pyöng-t'ae. 1981. "Nongji kaehyök úi p'yöngga wa pansöng" [Reflections on and criticisms of land reform]. In Han'guk kyöngje üi chön'gae kwajöng—haebang ihu eso 70 yöndae kkaji, pp. 35—65. Seoul: Toi Pegae. Kim Yöng-mo. 1971. "Ilchesidae chiju úi sahoejök paegyöng kwa idong" [The social background and mobility ofbig landlords during the Japanese colonial period]. Asea Yön'gu 14, no. 2:107-25. Ko Süng-je. 1977. Han'guk ch'ollak sahoesa yön'gu [Studies in the social history of rural Korea]. Seoul: Ilji Sa. Kobayakawa Küro. 1944. Chösen nögyö hattatsu shi [A history of Korean agriculture]. 3 vols. Keijö: Chösen nökai. 54Journal ofKorean Studies _____ 1960. Chösen nögyö hattatsu shi [A history ofKorean agriculture]. Rev. ed. Tokyo: Yuhö Kyökai. Lee, Hoon K. [Yi Hun-gu]. 1936. Land Utilization andRuralEconomy in Korea. Chicago: University of Chicago Press. Lee, Ki-baik. 1984. A New History ofKorea. Cambridge, Harvard University Press. Miyajima Hiroshi. 1982. "Shokuminchi ka Chösenjin öjinushi no sonzai keitai ni kansuru shiron" [Comments on the existing types of large ethnic Korean landlords during the colonial period]. Chösen shisö 5/6: 1 5 1—235. Moore, Barrington, Jr. 1966. Social Origins ofDictatorship and Democracy : Lord and Peasant in the Making of the Modern World. Boston: Beacon Press. Moskowitz, Karl. 1974. "The Creation of the Oriental Development Company: Japanese Illusions Meet Korean Reality." Occasional Papers on Korea, no. 2, pp. 73—121. Nörinkyoku. 1937. Chösen kosaku nempo [Korean tenancy yearbook]. Keijö: Chösen Sötokufu. Pak Ki-hyuk. 1966. A Study of Land Tenure System in Korea. Seoul: Korea Land Economics Research Center. Pak Mun-gyu. 1933. "Nöson shakai bunka no kiten to shite no tochi chösajigyö ni tsuite" [On the land survey as the starting point for rural culture]. In Chösen shakai keizaishi kenkyü. Keijö: Keijö Daigaku Höbungakkai. Stinchcombe, Arthur. 1961. "Agricultural Enterprise and Rural Class Relations." AmericanJournal of Sociology 5, no. 67:165—76. Suzuki Eitarö. 1961. Han'guk nongch'on tapsagi [A field survey of Korean villages]. Seoul: Ehwa University Press. [Japanese original 1944] Takahashi Kamekichi. 1935. Gendai Chösen keizairon [The contemporary Korean economy]. Tokyo: Chigura shobö. Zagoria, Donald S. 1974. "Asian Tenancy Systems and Communist Mobilization of the Peasantry." In John W. Lewis, ed. Peasant Rebellion and Communist Revolution in Asia. Stanford, Calif.: Stanford University Press. Zenshö Eisuke. 1929. Chösen no kosaku kanshü [Korean tenancy customs ]. Keijö: Chösen Sötokufu. Journal ofKorean Studies 7 (1990-91): 35-54 ...