Kokugo and Colonial Education in Taiwan

Kokugo and Colonial Education in Taiwan
Eika Tai
"positions: east asia cultures critique 7.2 (1999) 503-540 "
--------------------------------------------------------------------------------

Problematizing "Teaching Japanese"
Under Japanese colonial rule, the Taiwanese were compelled to study and to speak Kokugo, the national language of Japan. 1 Yet when Japan acquired Taiwan in 1895 as a result of the Sino-Japanese War, the concept of Kokugo had only begun to emerge into Japanese political discourse and had yet to reach public consciousness. Japanese leaders and intellectuals were still struggling to establish a common language that would both standardize language practice and consolidate the numerous (local) vernaculars, as a means to establishing a common consciousness among a populace yet to be unified into a nation. Only gradually did the concept of Kokugo develop amid contentious political and pedagogical debate. In this process, Lee Yeounsuk contends, the Sino-Japanese War played the role of an "initial explosive," linking the language problem that shook Meiji Japan to national and imperialist consciousness. Central to this emerging discourse was Ueda [End Page 503] Kazutoshi's version of Kokugo. 2 In an important 1894 speech, this educational theorist who had come under the influence of European linguistics established Kokugo as the national language of Japan. Soon he was envisioning the spread of Japanese to the rest of Asia. In other words, even before its particular structures were completely set forth, the new Japanese national language took on a "double status: 'universal-imperial' and 'particular-national.'" 3

Sometimes referred to as an ideology (ideorogi), Kokugo provided a scheme for the linguistic assimilation of subjugated people into the Japanese nation. It reached people in the home islands of Japan through the nascent public school system, as it was concurrently being applied to Japan's first formal colony. 4

As an institutionally sanctioned doctrine of colonial education, Kokugo ideology implied universalism in the sense of its potential application; it suggested that any person who mastered Kokugo could become Japanese. Its actual implementation, however, was particularistic, since the colonized could never become Japanese regardless of their level of competence in the Japanese language. The distinction between the colonizers and the colonized was strictly maintained through the family-registry (koseki) system. Further, the discriminatory legal system in general ultimately ensured that the cultural assimilation of the colonized would fail despite the efforts of scholars and teachers to put the doctrine of Kokugo into practice. Their fruitless struggle suggests not simply a limit to the effects of language in making imperial subjects; it also underlines the tension at work between colonization and assimilation. The imposition of the Kokugo ideology ultimately failed because of these legal bars. Other reasons include the extremely qualified success of the long effort to transform the Taiwanese into speakers of Japanese and the resistance of the colonized to the Kokugo ideology.

In what follows, I look at how the concept of Kokugo emerged, migrated to Taiwan, and was, for political and economic reasons, applied to colonial education. I then turn to the debates in the 1940s over the reconceptualization of Kokugo in relation to the emergence of the concept of Nihongo (Japanese language) 5 and discuss how language teachers applied the doctrine of Kokugo. Finally I suggest how ordinary people living in Taiwan under Japanese rule received Kokugo, in light of the fact that it is now part of the [End Page 504] present-day, postcolonial cognitive framework of the majority of Japanese people. In examining the ways in which Kokugo ideology was applied to the project of assimilating the colonized, I want to shed light on what I see as a contradiction between colonization and assimilation, as I inquire into the role that colonial education ultimately played in the process of Japan's linguistic unification.

I am personally motivated to tackle this topic because I carry the legacy of colonial education. My father learned Japanese by fiat in Taiwan. He met my Japanese mother when he moved to mainland Japan for the college education he could not get at home. My interest in the topic also derives from my current occupation as a teacher of Japanese to nonnative speakers of the language, many of whom are of Asian descent. Since the 1980s the spread of Japanese overseas has been underwritten by the Japanese government, through both budgetary support and the involvement of governmental organizations. 6 As the child of a colonized Taiwanese now teaching Japanese myself, I feel a responsibility to consider the implications of this second tide of the spread of Japanese. The examination of the past, I believe, may offer us a way to shed light on the social significance of the new phenomenon. 7

The Creation of Kokugo
The concept of a Japanese national language, Kokugo, and its legitimization as the standard language of Japan were the products of Western-trained linguists and educators. Initially, Japanese thinkers found it difficult even to conceive of such a language. Yet the unity of the Japanese language needed to be assumed before language and nation could be linked. 8 Both leaders and intellectuals felt the lack of a unified language, particularly as political circumstances brought pressure to build a modern state. The Japanese language had yet to be "resurrected." 9 Indeed the suggestion that Japan adopt English made by Mori Arinori, who had studied in London and visited the United States, symbolized the pessimism of many Meiji intellectuals toward their language:

The commercial power of the English-speaking race which now rules the world drives our people into some knowledge of their commercial [End Page 505] ways and habits. The absolute necessity of mastering the English language is thus forced upon us. It is a requisite of our independence in the community of nations. Under the circumstances, our meager language, which can never be any use outside of our islands, is doomed to yield to the domination of the English tongue, especially when the power of steam and electricity shall have pervaded the land. 10
When Mori suggested the adoption of simplified English several years after the Meiji Restoration, the concept of Kokugo had yet to be born (thus he was not denigrating Kokugo or the Japanese state). 11 The language Mori thought English should replace was actually neither Kokugo nor Nihongo, but Nihon no gengo (the language of Japan), a sum of speech forms that existed in Japan which to his mind posed problems. Breaching the enormous gap between the spoken and written language also seemed impossible. Mori further believed that the strong Chinese influence in the writing system posed the biggest obstacle to modernizing the language of Japan. The significance of Mori's statement, therefore, lies in his inability to conceptualize a unified linguistic system, the Japanese language, which would be compatible with a unified political system, the Japanese empire. 12

The word Kokugo had been used since the Edo era in various ways. For example, it could refer to a Japanese dialect or to Japanese as distinct from foreign languages. Gradually it came to be used in association with the concepts of state (kokka) and nation (kokumin), 13 in a large part due to Ueda, who having witnessed the linguistic nationalism of Prussia, brought back to Japan the idea of a primordial unity of language, nation, and state.

In an 1894 speech titled "Kokugo to kokka to" [A national language and a state], Ueda asserted an indivisible link between the Japanese language and the Japanese state. Japan, he argued, was not a multiracial state. 14 In line with the notion of a family state (kazoku kokka) propagated by political leaders, Ueda narrated a history of the Japanese people in which one family had developed into a unitary people (jinmin), which evolved into one nation (kokumin). He argued further that what made cooperation within the Japanese nation possible was the Yamato minzoku (Japanese ethnic nation) possessing the Yamato damashii (Japanese soul), characterized by chukun aikoku (loyalty to the emperor and love of the country) and a [End Page 506] shared language. He likened this language to the spiritual blood of the Japanese from which the national polity of Japan (kokutai), "a moral concept that constituted the very essence of the state," 15 drew its primary sustenance. He declared that language not only symbolized kokutai but acted like an educator or a benevolent mother who could teach the national (kokumin-teki) ways of thinking and feeling. "Kokugo is the 'loyal retainer' of the Imperial Household," 16 Ueda argued in a formulation that implicitly supported the emperor system as it dignified the language, assigning it an official role within that system.

Unequivocal though he was in weaving the concept of Kokugo as the vital link between the state and its people, Ueda was also keenly aware that a unified linguistic system had yet to be constructed. He immediately began to implement his theory and turned to constructing a standard language based on a Tokyo vernacular. The construction of Kokugo took place around the issue known as the Kokugo-Kokuji mondai (the problem of the national language and the national script), which involved two major questions: Which kana syllabary, traditional or pronunciation-based, should be used? To what degree should Chinese characters be allowed? Those who, like Ueda, espoused Western linguistics advocated the pronunciation-based kana syllabary and the reduction or abolition of Chinese characters. In contrast, ultranationalists argued for preserving the traditional orthography, on the grounds that meddling with the language constituted the denigration of traditional values and kokutai.

Ueda's initiative was influential, first, in rehabilitating the academic debate over Kokugo-Kokuji mondai and, second, in intervening in the facilitation of the formation of a standardized Japanese, Kokugo, through the power of the state. 17 Kokugo was first disseminated among the Japanese in 1900, when the Elementary School Law integrated it into the school curriculum, officially subsuming local vernaculars under the national language as dialects.

For Ueda, Kokugo possessed the power to protect kokutai and produce the kokumin character. Kokumin, the putative native speakers of Kokugo, were implicitly assumed to lack all subjectivity and thus to be open to internalizing kokumin loyalty to the emperor and love of the country. In Ueda's view, language is the voice of kokutai, "the voice of absolute order and subjugation"; [End Page 507] indeed, individuals could become speaking subjects by internalizing kokutai. 18 Ueda's narrative thus delineated the constitution of national subjects in relation to the state and its official language.

But the question arises of what role a national language can actually play in the formation of subjects. Under certain conditions, a shared language can mobilize its speakers to form a new state. 19 Conversely, a state may search for national unity through inventing a national language. E. J. Hobsbawm sheds light on the invented nature of national languages, which are "almost always semi-artificial constructs" and "the opposite of what nationalist mythology supposes them to be, namely the primordial foundations of national culture and the matrices of the national mind." 20 Yet Hobsbawm's emphasis on state and nationalist ideology overlooks other forces that may support or undermine the state power. Kevin M. Doak points to the importance of analyzing not merely national narratives but "alternative and competing narratives of the ethnic nation." 21 Prasenjit Duara speaks of "a polyphony of voices" that negotiate their views of the nation; for instance, provincial narratives may play a significant role in constituting national identity. 22 The symbolic domination of a national language is further subject to economic prosperity or local pride, and local vernaculars may survive in spite of the insistence of statist education. 23 Other forces besides state power clearly affected the reception of Kokugo. Despite the severe punishments sometimes imposed for using dialects, common people in Japan proved neither as willing to discard their language habits nor as enchanted by the myths of Kokugo as nationalist intellectuals and teachers had hoped. 24

In colonial Taiwan the effect of the Kokugo ideology on the formation of subjects was likewise limited, both because the Taiwanese people insisted on maintaining their speech habits and because of Kokugo's contradictory role in the colonial context. Besides advocating language standardization for homogenizing nationals and building a modern state, Ueda also anticipated the spread of Japanese overseas. 25 In "Kokugo kenkyu nitsuite" [About the study of Kokugo], he expressed his interest in creating a common language of Asia as the Japanese national language developed. 26 Discussing "kokumin education and Kokugo education" in 1902, he stated even more explicitly that the Kokugo problem was relevant both to the formation of the Japanese nation and to the spread of the Japanese language to Asia. 27 In colonized [End Page 508] Taiwan, I will argue, these two objectives led to the contradiction Benedict Anderson calls "the inner incompatibility of empire and nation." 28

Ueda's famous formulation of the nation (kokumin) coinciding with the Japanese ethnic nation (minzoku) cannot be taken for granted. Rather, a distinction between the two must be made in order to uncover what Doak finds largely absent in discussions on Japanese nationalism, namely the "history of imagining the nation itself as an ethnic body." 29 Moreover, in the contemporary academic and political arenas, the question of what constituted the Japanese ethnic nation in conjunction with imperial expansion and nation-building was heatedly debated. 30 The application of Kokugo ideology in colonial education raised the question of whether the colonized could be assimilated by teaching the Japanese the language of empire. It presented an even more fundamental question: What was it, exactly, into which colonials were to be assimilated: the Japanese nation, Japanese ethnic nation, or something else?

Kokugo Education in Taiwan
Kokugo was central to colonial education in Taiwan from the very start. The patriotic educator Izawa Shuji initiated the teaching of Kokugo to the colonized. As an educator who worked with him later recalled, Izawa did not merely envision the Taiwanese mastering Kokugo but was determined to transform them into Japanese through Kokugo education. 31 Izawa's idea of Japanization was based on his own idea of kokka-shugi kyoiku (state education), and the goal of Kokugo education was to produce loyal shinmin (imperial subjects).

According to Yun Kwon-cha, Izawa played an important role in nation-building in the early Meiji period. Working closely in the Ministry of Education with his powerful patron, Mori Arinori, Izawa developed a theory of education based on the idea of perfect human beings whose core characteristic was their belief in chukun aikoku. For him, then, the goal of schooling was to transform pupils into loyal subjects of the emperor, which ultimately prepared them to die for the state. In his vision, even gymnastics and music (subjects that he adopted from Western education) could be used to develop loyalty to the emperor. 32 After Mori's assassination in 1889, Izawa left the [End Page 509] Ministry of Education, establishing the Society for State Education in 1890 to promote his theory of education. At the society's opening ceremony, he extolled Kokugo and insisted on the importance of fostering love and respect for it as a way of developing pride for and love of country. He was exhilarated to find that his ideas on education were in line with the Imperial Rescript on Education, the principles of moral education promulgated in that same year. 33

It was this idea of education in the service of the state that Izawa brought to Taiwan. While political leaders were still pondering how to control the colony, Izawa visited Kabayama Sukenori, who was to be the first governor-general. Impressed with Izawa's detailed proposals and his knowledge of Chinese, Kabayama decided to entrust to him the education of the people in the new territory. 34 In June 1895, as Taiwanese guerrillas were still vehemently resisting the Japanese, Izawa arrived in Taiwan and immediately began to put his plan into practice. Appointed chief of the Bureau of Education Affairs in the government-general, by July 1895 he had opened an experimental Japanese course for Taiwanese at Shizangan (Zhshanyán), 35 in a suburb of Taipei. At his invitation, some of the Taiwanese gentry and intellectuals in the area sent their children to this first Japanese course in Taiwan. In 1896 Izawa established fourteen Kokugo institutes in various locations for the study of Japanese by Taiwanese students and constructed the Kokugo School as the central institution to provide training for interpreters and language teachers. 36

Izawa did not, however, dismiss the Taiwanese language altogether. His approach to colonial education, which drew on his experiences as editor of a Japanese-Chinese dictionary, adopted the idea of konwa, or the integration of "us and them." In practice this meant he not only encouraged the natives to study Kokugo but also urged Japanese people to learn Taiwanese (Minnan). This approach was pragmatic during the early colonial period, when communication between the Japanese and the natives was extremely difficult. Yet as Iwamoto Yumiko argues, it is clear from examining the wording of education ordinances that Izawa's ultimate goal was to transform the colonized into imperial subjects. For example, the Kokugo Institute Ordinance stated that the objective was teaching the language and fostering the spirit of the Japanese homeland. Izawa's belief in state education was even [End Page 510] more clearly articulated in the Kokugo School Ordinance, promulgated a few months later, which emphasized chukun aikoku. 37

What was significant in Izawa's accomplishment in Taiwan was that he attempted to inject his belief in state education not only into school facilities and language policies but also into the minds of teachers. By capitalizing on an 1896 guerrilla attack, he successfully established an ideal image of a teacher calculated to inspire devotion to education. On New Year's Day 1896, Taiwanese guerrillas attacked Shizangan and killed six education officers. Izawa was behind the construction of a monument at Shizangan commemorating the dead officers' dedication and their sacrifice, designating it "the spirit of Shizangan." 38 The so-called Shizangan incident enabled Izawa to add inspirational symbolism to his groundwork for colonial education. The memorial ceremonies, held annually and on special occasions, were staged as spectacular shows. At the first anniversary, Izawa himself continually wept as he gave a speech to the audience of teacher-trainees he had recruited from mainland Japan, exhorting them to adopt the spirit of the six officers. Succeeding annual ceremonies were attended by politicians, educators, and thousands of pupils, both Japanese and Taiwanese. The incident was also told as a story in Kokugo textbooks for Taiwanese children and made into a song. Among the many monuments commemorating the spirit was one displaying the engraved brush of Premier Ito Hirobumi, demonstrating the home government's endorsement. Significantly, the death of the six educators was mystified for ideological purposes in the Kokugo campaign. 39 Constructing a model for teachers, Izawa planted the value of devotion to Kokugo education in the soil of Taiwan. No doubt he foresaw that teachers would be important agents in the imperialization of their colonized Taiwanese pupils.

Importantly, Izawa's aim was to transform people into loyal subjects, regardless of their descent. He approached Han people in Taiwan as i-minzoku (ethnic nation different from the Yamoto minzoku), but for him, competence in Kokugo and loyalty to the Japanese empire would be enough to qualify a Taiwanese as a Japanese subject. He thus aligned himself with the school of thought that insisted on the mixed-blood origin of the Japanese race and attributed the supremacy of the race to this heritage rather than to the supposed purity of Japanese blood, proclaiming that the "impartiality and [End Page 511] equal benevolence of the emperor" (isshi dojin) was granted not only to the Yamoto minzoku but to all subjects. 40 Summarizing Izawa's work, E. Patricia Tsurumi contends that in contrast to his successors Izawa dreamed of the establishment of schools that "would one day bring Japanese language and culture to every native islander"; he "imagin[ed] young Taiwanese rising to the top of the Japanese educational pyramid and making their way into the ranks of the elite that governed the empire." 41 Izawa's scheme assumed that Kokugo constituted the essence of the national character, reproduced that character in its putative native speakers (as Ueda had suggested it would), and furthermore could foster that essence in those outside this ethnic nation. In other words, as it moved to colonize Taiwan, the Kokugo ideology mutated; it was now interpreted as universal, since any person could become assimilated as a Japanese. It was this universalist version of the ideology that became an institutionally sanctioned doctrine in colonial education.

Izawa paved the way for Kokugo education in Taiwan, setting a precedent for education in the territories Japan would later rule. But his commitment to teaching Kokugo to all Taiwanese was not compatible with the policies of early colonial leaders, who did not consider the assimilation of the colonized to be an urgent goal. Izawa's proposal for expanding the Japanese language program was rejected, and he was forced to resign his position in 1897, two years after his arrival in Taiwan. 42 His enthusiasm for Kokugo education, however, was inherited by the teachers he recruited and trained, and his idea of transforming Taiwanese into Japanese imperial subjects through Kokugo education survived throughout the half-century history of colonial education in Taiwan.

Early Colonial Policy
Taiwan confronted Japanese leaders with the question of how to integrate this first colony into the existing empire. Having no clear vision of how to govern it, they consulted Western advisers working in the Ministry of Justice, who recommended two different approaches. The first was an assimilationist policy, as in Algeria under French rule. The second model was that of the British crown colony, as in the cases of India and Hong Kong. Correspondingly, [End Page 512] Japanese leaders were divided into two camps. A leading polemicist for the first approach was Hara Kei, representing the Ministry of Foreign Affairs; he wanted to apply the Meiji Constitution and other Japanese laws to Taiwan. The other camp, including Goto Shinpei (who was to play an important role in Taiwan), insisted on adopting a "two-systems-one-country" policy, which would dictate control of Taiwan independently from mainland Japan. 43

With this controversy yet to be settled, in 1896 Law 63 granted the government-general the authority to exercise legislative power over Taiwan independent of the home government. First passed with a time limit of three years, it was later extended as Goto succeeded in stabilizing colonial power in Taiwan. The law was essentially maintained throughout the colonial period, although with several nominal changes; the final version, Law 3 of 1921, emphasized the application of Japanese laws to Taiwan and clearly marked a change toward assimilationism. Contradictorily, however, it substantially preserved the legislative power of the government-general and thus continued to hinder Taiwanese legal assimilation. 44

In conjunction with Law 63 and its successors, the family registry (koseki) system played a crucial role in maintaining the distinction between the colonizers and the colonized. 45 When the old Nationality Law was promulgated in 1899 and applied to Taiwan, the Taiwanese, who had been proclaimed to be shinmin of the Japanese empire, nominally acquired Japanese nationality. 46 Yet their colonized status was maintained through the family registry system. Established in 1905, it proved especially useful for the military police in suppressing political movements and maintaining security. 47 In contrast, immigrants from the home islands of Japan filed their koseki within that domain. (Taiwanese could not transfer their koseki to the home islands.) With various privileges over the colonized, Japanese immigrants--called Naichijin (inner-land people)--were considered part of the Yamato minzoku. 48 Thus the family registry system tightly guarded ethnic Japanese boundaries. 49 In part because of this legal discrimination, the assimilationist policy that gradually took effect in Taiwan was doomed to failure. Japanese leaders boasted that Japan would succeed where Western colonizers had failed, because Japanese rule was based on dobun doshu (same script and same race), the racial and cultural affinity of the Japanese [End Page 513] and the Taiwanese, as well as on isshi dojin. But in reality these two ideas proved hollow.

It was the team of Kodama Gentaro, the fourth governor-general, and Goto, director of civil administration, that first laid the groundwork for establishing colonial power in Taiwan. During their regime between 1898 and 1906, they succeeded in improving social conditions in Taiwan in terms of public security, sanitation, and the economy, while quelling anti-Japanese guerrillas militarily. Advocating the British model of colonialism, Goto insisted that Taiwan, as part of Chinese civilization, should be ruled according to the habits and customs of the natives; assimilation would be not only impossible but harmful. 50 With a passionate faith in "biological principles," he argued that social habits and systems evolved over a long period of time, and that any attempt to apply "civilized" systems to an "uncivilized" society was doomed to fail. 51 But rather than entirely dismissing the idea of assimilation, he instead conceived of a "hundred-year plan" for the gradual evolution of Taiwanese society. 52 As the Bureau of Educational Affairs came under Goto's control, Izawa's idea of teaching Japanese to all Taiwanese and thereby Japanizing them was discarded, and "gradualism" was adopted in its place. 53

In response to the demand from local administrators for an increase in the number of schools, the common school (kogakko) system was established for Taiwanese children in 1898. The national budget for the system was very low, however; all school expenses except the salaries of Japanese teachers were charged to the local school district, which collected money from taxes and donations from the well-to-do. 54 Hence, only those districts that were affluent enough to support operations were able to establish schools. In 1900 the chief of educational affairs was forced to resign within a year of calling for the implementation of compulsory education. In 1903 school principals echoed the call, but their request was rejected for financial reasons. 55 As a result, in 1910 the school attendance rate of Taiwanese children was less than 6 percent. In contrast, more than 90 percent of the children newly arrived from Japan proper attended shogakko (elementary school), which was under separate administration. 56

Another impediment to the spread of Kokugo education was the Han Taiwanese themselves, who had their own education system. Both private [End Page 514] and public schools prepared students for Chinese government employment examinations and aimed to enhance the Han Chinese spirit among students. The Japanese attempted to attract students from Chinese schools (shobo) by offering at common schools some training in the Confucian classics, an appeal to wealthy Taiwanese. At the same time, the colonial government placed various forms of pressure on the operation of shobo, including forcing instruction in Japanese history, language, and the Imperial Rescript on Education, translated into Chinese. Yet having historically regarded the Japanese as barbarians, the Taiwanese elites resisted these requirements of the colonial government. Even so, by 1904 the number of Chinese pupils enrolled in Japanese schools actually exceeded those attending shobo. 57

Although Goto had no particular vision for education, the general direction of colonial education did not diverge much from Izawa's time. The Common School Regulations promulgated in 1898 embraced two general objectives: the mastery of Kokugo and the development of the Japanese national character, by implication the Kokugo ideology. Some Japanese educators protested that teaching the Kokugo language was hardly enough for achieving the goal of assimilating the Taiwanese; rather, they emphasized moral education. Yet keenly aware of the difficulty of convincing the Taiwanese--who historically recognized the model of a Chinese emperor ruling on the basis of his virtue--to respect the sacred blood of the Japanese emperor, educators often modified the content of their moral education to reflect the value system of the Taiwanese. 58

Assimilationist Policy
After 1910, in response to political upheavals such as the revolution in China, the tide of nationalism in Western colonies, the 1 March movement in Korea, and the nationalist movements in Taiwan, the Japanese government finally adopted an assimilationist approach to the control of its colonies. Hara Kei, a longtime advocate of naichi-encho shugi (the idea of controlling colonies as extensions of Japan proper), changed the direction of Japanese colonialism after becoming prime minister in 1918. Discontinuing military control of the colonial administration, he appointed Den Kenjiro to be the first civilian governor-general in Taiwan in 1919. Seeking to complete [End Page 515] the integration of Taiwan into the jurisdiction of the Meiji Constitution, Hara passed Law 3, referred to above, which in great measure limited the legislative power of the colonial government. 59 With Den's declaration of a policy to transform the Taiwanese into the Yamato minzoku, changes such as hiring Taiwanese as public officials and permitting intermarriage between Taiwanese and Japanese were instituted. 60 Den was in fact attempting acculturation (kyoka), a process aimed at civilizing the colonized to be loyal to the emperor while maintaining them in their "proper" low status.

Besides extolling the same two goals, the mastery of Kokugo and the development of the Japanese national character, the New Taiwan Education Rescript promulgated in 1922 "allowed" the coeducation of Taiwanese and Japanese. But this new policy brought about neither the improvement of education nor equal opportunity for Taiwanese, contrary to the propaganda of the colonial government. The law contained an important proviso dictating that only those who used Japanese regularly in everyday life could go to schools intended for Japanese. Such people remained rare throughout the colonial period, and the Taiwanese pupils who participated in this form of coeducation never exceeded 1 percent of those Taiwanese who received elementary education. 61 Observing the change brought about by this educational reform, Yanaihara Tadao, a contemporary critic of Japanese imperialism, argued that this system actually deprived Taiwanese students of higher education, as spots originally intended for Taiwanese were filled by Japanese students. 62 Even the medical school established for Taiwanese, the only institution offering them a higher education, was absorbed into the primarily Japanese Taihoku (Taipei) Imperial University. 63 Thus, Taiwanese seeking higher education had to go to mainland Japan.

By the 1920s and 1930s the overall school attendance rate among Taiwanese children was increasing steadily from about 25 to a little over 50 percent. 64 Yet, as Tsurumi suggests, "the common school was definitely meant to assimilate Taiwanese but only at the bottom of the Japanese social order." 65 Kokugo education for adults also expanded in this period, as the colonial government emphasized post-elementary vocational education for Taiwanese, along the principle of "agriculture in Taiwan and industry in Japan." 66

This coincided with the expansion of Taiwanese political participation in [End Page 516] the 1930s. In 1934 the first Taiwanese, Ko Ken'ei, was elected as a member of the national House of Peers, and in 1935 certain qualified Taiwanese were given both suffrage and eligibility for election to local seats. But theory and practice diverged: in actual practice local assemblies had little political power. 67

The Kominka Policy of Imperializing Taiwan
With the start of full-scale war against China in 1937 and the later advances into the South Pacific, the assimilation of the Taiwanese people became an urgent concern for Japanese leaders. In this period assimilation policy was consolidated under the banner of kominka (imperialization of subject people). Kondo Masami argues that kominka policy emerged during the wartime regime because of the military necessity to mobilize non-Japanese masses for the war. With the war, the strategic importance of Taiwan itself had changed, from a place for producing sugar and rice to a source of matériel and personnel needed in the South Pacific. Initially exempted from military conscription, beginning in 1937 Taiwanese were sent to war as military servants and, after 1942, as volunteer soldiers. Yet, colonial leaders, especially members of the Taiwan army, feared that most Han Taiwanese hoped for a Chinese victory. 68 Thus, it became imperative to inculcate the Taiwanese with the ideology of emperor worship.

As the base for Japan's expansion, Taiwan was increasingly treated more and more as part of Japan proper. Ohe Shinobu describes Japanese colonial expansion in terms of concentric circles, with mainland Japan as the center; the two colonies, Taiwan and Korea, comprised the next, inner circle; and the mainland China territories formed an exterior circle, to be eventually encompassed by the projected Greater East Asia Co-Prosperity Sphere. As this exterior circle expanded, control of the inner circle of colonies was increasingly tightened, and the assimilation of the colonized more and more demanded. 69 As Mark R. Peattie states, "Now there were only the naichi--'the inner area' (the Japanese home islands)--and the gaichi--'the outer area' (Japan overseas)--which were held to comprise a single expanding, Pan-Asian bloc." 70 As this bloc expanded, Taiwan was becoming more and more naichi-like. [End Page 517]

Meanwhile, military control of the colonial administration was resumed. When appointed governor-general in 1936, Kobayashi Seizo proclaimed as his threefold policy advancement to the South Pacific, industrialization in Taiwan, and kominka. But while his approach to assimilation resembled his predecessors', it differed in scale. Envisioning the diffusion of Kokugo and emperor worship through the entire Taiwanese society, he designed a plan for compulsory education for the colonized, something never seriously considered by previous governments. 71 The government-general in turn drew up a ten-year plan to promote the spread of Kokugo, and in 1940 stated the goal of enforcing Kokugo study for everyone under seventy. To achieve this goal, the government opened Kokugo schools for all ages, from preschool children to the elderly.

Kokugo education for girls and women also became important, as they were seen as agents for disseminating Kokugo into family life. Under the slogan "the use of Kokugo at home," the "Kokugo home" in which family members spoke Kokugo and lived in the Japanese way came to be officially certified, with door plates designating this special honor. Such families were granted material and social privileges. 72 Simultaneously, the use of Taiwanese vernaculars was strictly prohibited at school. 73

At this final stage, Kokugo was seen as the prime mover and barometer of the kominka movement, as Kokugo education was implemented to promote the consolidation of military power. A case study of a small Taiwanese village by Kondo reveals this concatenation. Asserting that the core of kominka was the everyday use of the Kokugo language, the village administration imposed the use of Kokugo on the villagers not merely as a language for school but as seikatsugo (a language for everyday life). Among the agents of the kominka movement, the leaders of the Young Men's Associations, in which Taiwanese youths were obliged to participate, played a significant role in mediating between the battlefield and village life. Recruited from among graduates of common schools, the leaders were first sent to acculturation training provided by both the village administration and the colonial government, and were then expected to disseminate Kokugo and the doctrine of kominka in their community. These young leaders were subsequently sent to war as military servants and volunteer soldiers. 74

During the final years of the war, as the empire's imminent defeat became [End Page 518] clear, the desperate struggle of military officers and soldiers on the battlefield resonated with the frantic efforts of administrators and educators at home to assimilate the colonized. Long-awaited compulsory education for Taiwanese people began in 1943 with subsidies from the homeland government. 75 Tsuchiya Tadao observed at the time that compulsory education seemed to have been implemented in accordance with a plan to conscript the colonized. 76 Now Kokugo education clearly had the dual aim of enabling Taiwanese to communicate with Japanese comrades on the battlefield and making them willing to fight for the emperor. By 1942, 65.8 percent of Taiwanese school-aged children attended school, and the rate increased to 71.3 percent in 1944. 77 Yet even in this stage of colonial education, the distinction between the Japanese and the colonized remained. Although the Japanese created the facade of integration by renaming both elementary schools for Japanese and common schools for Taiwanese "national schools" (kokumin gakko), they employed a ranking system that maintained the old hierarchy among schools. 78

As Tsuchiya points out, prior to the kominka movement, most Japanese officials were content to teach language skills without Japanizing the Taiwanese. 79 First implemented mainly for its economic benefit to the homeland and then accelerated according to military necessity, Kokugo education gradually gained the commitment of both the home and colonial government. Not surprisingly, Izawa made a comeback in the 1940s, when he was recognized as "the father of education in Taiwan." 80 In a 1944 book, the Association of Taiwan Education praised his efforts to apply the ideology of state education to Japan's new territory. The book also revived the ideologically loaded fame of the Shizangan incident. 81

To sum up, the educational doctrine of the Kokugo ideology was central to the colonial aim of assimilating the Taiwanese people. Interpreted as universal, in its implementation, in conjunction with other apparatuses of state control, Kokugo remained particularistic. Thus the colonized were prevented from entering the Japanese nation (kokumin) on the same footing as the ethnic Japanese. Throughout the colonial period, Taiwanese people continued to be controlled as i-minzoku, an ethnic nation different from the Yomato minzoku. The objectives of education rescripts and regulations stressed the development of Kokugo language skills as well as the kokumin [End Page 519] character and spirit, and such documents often referred to the Taiwanese as kokumin. But they were so designated only when their being kokumin would benefit the home country. When their assimilation was emphasized, on the other hand, they were instead called imperial subjects, shinmin or komin, which indicated a closer link to the emperor. Yet both naming strategies belied legal bars to genuine assimilation, as well as resistance on the part of the colonized.

The Emergence of Nihongo Education
In the advancement (shinshutsu) to the South Pacific that changed the direction of colonial policy, then, Taiwan was part of the larger strategic scheme to build a Greater East Asia Co-Prosperity Sphere. In this scheme, Japan would be the political leader "responsible for governance and guidance . . . of those peoples who lacked the capacity for independence." 82 The educational corollary was the shinshutsu of the Japanese language throughout Asia and eventually the world. It was against this backdrop that the concept of Nihongo emerged specifically to mean Japanese taught to non-Japanese people as a foreign language. 83

The prospect of a long-range "total" war prompted the homeland government to increase its direct involvement in the administration of Japan's colonies and occupied territories. The field of education was no exception. When the Association for Promoting Japanese Language Education (Nihongo Kyoiku Shinkokai) was established in 1941, colonial education took a new turn under the increasing control of the central government. 84 The establishment of this organization and its journal, Nihongo, marked the official start of the new field of Nihongo education. Matsuo Chozo, the association's leader, expounded on the orientation of this new field in the prologue to the journal's first issue, stating the mission of building a Greater East Asia Co-Prosperity Sphere based on Japanese national culture (kokumin bunka) in order to bring about a new world order; the spread of Japanese would be the key to this effort. 85

The word Nihongo had been used previously, but with the Co-Prosperity Sphere its usage was now juxtaposed with that of Kokugo. Koyasu Nobukuni observes that Ando Masatsugu, a scholar and leading policymaker in [End Page 520] Kokugo education, avoided confronting the conceptual problems of reconstituting the concept of Kokugo vis-à-vis that of Nihongo. Instead, Ando distinguished between the Kokugo problem, which he determined should be handled by intrastate measures, and the Nihongo problem, which he confined to extrastate measures. 86

In a special issue of Nihongo subtitled "The Advancement of Nihongo and its Policy," Ando argued that the formerly domestic problem of the national language and national script (Kokugo Kokuji), once domestic in nature, had begun to be discussed as encompassing an extrastate aspect. Separating the polemic on the overseas spread of Nihongo from the Kokugo Kokuji problem and defining Kokugo as the language that the Japanese nation had inherited from its ancestors, he argued that the reform of both Kokugo and Kokuji should conform to the needs of the Japanese nation; if pursued to promote the spread of Nihongo, reform would only denigrate the sacredness of Kokugo. Nihongo should therefore be spread as a foreign language to Asian countries outside the Japanese state, where Nihongo education would inevitably remain secondary to education in the native languages in question. 87 The impact of this construction of Kokugo and Nihongo continues even to this day. 88

In a subsequent issue of Nihongo, Ando further suggested simplifying the writing system of Japanese for foreign learners to facilitate the spread of the language and argued that teaching Japanese as an embodiment of Japanese culture was idealistic but impractical. 89 Non-Japanese Asians should study Nihongo, which Ando viewed as essentially different from Kokugo, the language that constituted the essence of the Japanese nation.

In contrast, as Koyasu points out, Tokieda Motoki reexamined the concept of Kokugo against the background of the contemporary language situation. 90 Observing that the "Kokugo problem" crossed the boundaries of the state and thus developed into the "Nihongo problem," Tokieda problematized an issue that stood between those two, namely the language situation in Japan's colonies. 91 Ando, in contrast, sidestepped this issue by arguing that Kokugo would eventually become the native language for those non-Japanese ethnic groups in Japan's colonies. 92

Tokieda revisited the concept of Kokugo established four decades earlier by Ueda Kazutoshi, his former teacher. Tokieda observed four phases in the [End Page 521] trajectory of language policy dating from the early Meiji era. The first phase, which lasted for two decades, was characterized by the language pessimism expressed by people such as Mori Arinori. The second phase was the period of ultranationalism, reaching its climax at the time of Ueda's 1894 speech. The third opened when Japan came to include non-Japanese ethnic groups, with the acquisition of Taiwan and Korea. The fourth phase, which began as Japan assumed the leading role in the establishment of the Co-Prosperity Sphere, was the period when the question arose of how to spread Nihongo as a common language in this sphere. 93

In Tokieda's analysis, the contradiction inherent in the Kokugo problem in Korea (a feature of the third phase) had been overlooked because the ethnonationalistic (minzoku-shugi-teki) view of the second phase still prevailed in Japan. In other words, Ueda's theory was developed when the Japanese state, ethnic nation, and language coincided; it was thus valid for precolonial Japan, which Tokieda likened to a happy family with no in-laws. But Koreans--whom he called igo-minzoku (ethnic nation speaking a language other than Japanese)--presented a problem for Ueda's theory. The love for one's native tongue, Tokieda argued, would naturally evoke in Koreans the love for their own language, since it was their mother tongue, their language of everyday life, and in Ueda's terms, their "spiritual blood." 94

The question was how to resolve the conflict between the situation in Korea and a colonial policy stressing the expansion of Kokugo. Tokieda tackled this question, beginning from the assumption that language has no existence apart from the language activities of speaking subjects. Thus, in his discussion of Kokugo's supremacy, he began from a consideration of "subjective value consciousness" (shutai-tekina kachi ishiki). Significantly, for him such consciousness was constituted by the state. Thus he defined Kokugo as a special language laden with value from the point of view of the Japanese state, while Nihongo was merely an object for linguistic study, independent of the Kokugo--that is, the state's value system. The recognition of Kokugo's supremacy was thus predicated on the structure of the modern Japanese state. In summary, Tokieda explicitly stated that Kokugo was no longer the monopoly of people in the home islands of Japan, but the property of this state. 95

This endorsement of state power amounted to the dismissal of the subjective [End Page 522] positions of both colonized peoples and the Japanese who spoke regional vernaculars. But in returning to Ueda's concept of the mother tongue and recognizing Koreans' attachment to their language, Tokieda's solution was for Kokugo to be used in Korean homes, where family members could take pleasure in speaking it. For this reason, Tokieda stressed Kokugo education for girls as future mothers. 96 Thus in the end, Tokieda effectively came to agree with Ando that Kokugo should become the native speech of the emperor's subjects. 97 Unlike Ando, who sought to preserve Kokugo as the particular language of the Japanese nation, Tokieda thought that this form of Japanese should be open to foreigners without any modification. He envisaged that Kokugo would acquire a universal character as an international, potentially universal language to be shared with foreigners. 98

Yet the Kokugo versus Nihongo debate underlines the intention of leaders and intellectuals such as Ando to preserve Kokugo for the Japanese nation. The Nihongo/Kokugo distinction further suggests their desire to set a limit in assimilating non-Japanese to Japanese culture, to close the boundaries of the Japanese nation linguistically. To the extent that others, like Tokieda, argued that Japanese should become a universal language, the debate reflected the tension inherent in the double status of the Japanese language as both universal-imperial and particular-national. 99 In suggesting that, in the context of a colony, this tension could be resolved by making Kokugo the mother tongue of the colonized, Tokieda conformed to the colonial language policy of the 1940s. 100 Yet the fact that Kokugo education in Taiwan came to be discussed in the journal Nihongo also marked the exclusion of the Taiwanese from the status of native speakers of Japanese. This position reflected the actual language situation of the colony, where most people spoke their local vernaculars.

Pedagogy
While those working in and around the central government debated the reconceptualization of the Japanese language, educators were concerned with how to teach Kokugo effectively in the actual classroom. Their goal was the development of native-like competence in Kokugo. Yamaguchi Kiichiro, the leading colonial scholar of pedagogy, made a substantial contribution [End Page 523] to developing a methodology for learners to achieve such competence. Yamaguchi was originally recruited by Izawa to teach in colonized Taiwan (he was in attendance at the first Shinzangan ceremony), where he had arrived in 1896 as an experienced Kokugo teacher. He stayed until 1911, when he moved to newly colonized Korea. Throughout his forty-eight-year career of teaching Japanese in Japanese colonies and occupied territories, Yamaguchi stayed very close to actual classroom teaching, while at the same time pursuing research and editing textbooks. 101

The significance of Yamaguchi's work lay in his contribution to the development of chokusetsu-ho (direct method). In his book on this method, he delineated his theory of teaching foreign language and explained the effectiveness of this technique. His inspiration was the method that a French educator, François Gouin, had developed for teaching German as a foreign language, based on his observation of young children learning their native tongues. 102 In line with Gouin's method, Yamaguchi argued that it was crucial to have students learn through authentic language activity, recognizing and mastering the form of the language along with the particular and concrete meanings of the material. 103 The study of language should take place in the context of everyday life, with the "study life" also conducted in the language. 104 Thus the process of learning a foreign language was to approximate as nearly as possible the acquisition of a native language.

Yamaguchi firmly believed that the mastery of a language would lead to the construction of the personality associated with the language. 105 Indeed, the connection between language and other cultural elements had been one of his core arguments since the early days of his career. As early as 1904 he made the claim that "the language of a nation contained the whole metaphysical possession of the nation and that the knowledge, emotion, and quality of the nation, as well as the people's activities and growth, all resided in the language." 106 As Komagome Takeshi points out, this assertion supported the belief in the Kokugo ideology being constructed by Ueda and others at the time. 107

Yamaguchi later elaborated on a methodology for fostering Japanese culture and spirit, leading the pedagogical polemic on Nihongo. In one article, he undermined the very distinction between jikokugo (native language) and [End Page 524] gaikokugo (foreign language) in regard to the learning processes. He argued that the psychological state in which one performed language activities would be the same for both, once one had mastered the new language. The difference between the psychological states involved in speaking gaikokugo and jikokugo, he argued, was probably quantitative, not qualitative. 108 In another article, he denounced the widely used "translation method," arguing that it allowed learners to comprehend thoughts and phenomena in their own native tongues without developing any real understanding of Japanese culture and spirit. His direct method, in contrast, would help learners expand their scope of understanding in the target language; they would ultimately be able to remember, imagine, and think about things and phenomena completely in Japanese, and at this stage they would be able to feel kotodama (word spirit), to understand Japanese culture, and to internalize Nihon seishin (the Japanese spirit). 109

In Yamaguchi's theory, learners come to master Kokugo in two distinct stages: first they begin to think in Japanese, and in the second stage they internalize Japanese culture and spirit. 110 Like many of his contemporaries, Yamaguchi's ultimate goal was the Japanese spirit. Yet, what is the Japanese spirit? Komagome points out that despite its centrality in colonial education, the concept was used ambiguously and differently by those educating non-Japanese Asians. Consistent, however, was the consensus that this education was ultimately in the service of Japanese imperialism. 111 One contemporary believed the essence of the Japanese spirit was faith in the emperor. 112 Although Yamaguchi himself did not offer a definition, his usage of the word was probably close to this interpretation.

Also problematic is Yamaguchi's linking of stage two to the attainment of the Japanese spirit merely by reference to the concept of kotodama (word spirit). 113 This belief that words possessed spirits of their own that, when spoken, could cause things to happen, documented in folklore since the seventh century, had been revitalized in order to hallow Kokugo. It resonated well with the idea implicit in Ueda's narrative that Kokugo had power in itself. Without resorting to this concept, Yamaguchi could not have completed his assertion that the direct method would foster the Japanese spirit.

Recent theory indicates that the mastery of a language cannot be immediately linked to the internalization of the ideology associated with the language; [End Page 525] for "linguistic practice" including "knowledge and proficiency" is distinct from "linguistic ideology." 114 And it is the latter to which the concept of kotodama belongs. Even if, through the direct method, students did learn to think in the language, it remains doubtful whether the consequence of their mastery was their internalizing of the Japanese spirit and becoming faithful to the emperor. Such would have required the internalization of certain ideological assumptions--both those supposedly embedded in the language and those contingent on changing political circumstances facing the empire.

The Japanese Teachers
In reality, moreover, the heated ideological arguments over the language, its orthography, 115 and the assumptions underlying pedagogy had little apparent impact on the daily practice of classroom teachers in Taiwan. It was clear to teachers that Kokugo was a second language to their students. In Taiwan, where the pronunciation-based orthography was replaced by the traditional one in the early 1910s, even this loaded debate failed to become a major pedagogical issue. 116 Besides the direct method originally invented in Taiwan, other approaches were also developed and adopted in the colony. 117 The impetus for each was practical. Yet Japanese classroom teachers were agents of assimilation; with the most direct impact on how the colonized perceived the Japanese empire, their responsibility to help their students to master Kokugo also involved--whether implicitly or explicitly--inculcating the Japanese national character along the lines of Kokugo doctrine. 118

Some teachers did take seriously the possibility of assimilating the Taiwanese people. 119 Kimura Masuo, a colonial educator in Taiwan for fifteen years, declared without hesitation that the Japanese education of colonized students was successful. The seventeen former teachers of Japanese in Taiwan who responded to his questionnaires attributed this success to factors such as the assimilationist policy that treated Taiwanese people as Japanese, the educational policy that aimed at the development of Japanese ways of thinking as well as language skills, the superiority of Japanese culture, the sense of trust in Japan held by the colonized, the mild and generous character of the Taiwanese people, and dobun doshu. Kimura himself emphasized [End Page 526] the devotion of the teachers to their occupation and their students, and documented that some appreciative Taiwanese continued to maintain contact with their former teachers. 120 Sai Moho relates the case of a Kokugo teacher who had his Taiwanese students stay at his house and showed them how to perform daily routines, such as eating, brushing their teeth, making the bed, and cleaning the toilet, all in the Japanese way, in order to teach them the delicate nuances of the Japanese language in a real-life context. 121 This recalls Izawa's ideal of devotion, the Shizangan spirit.

Yet it is necessary to interrogate the very notion of this success. For one thing, however sincere Japanese language teachers might have been in their dealings with their students, their teaching practice--and thus, whatever success they achieved--was never independent of police and military power. Whereas policemen, outwardly symbolizing colonial power, served to remind the colonized of their subjugated position within the Japanese empire, teachers must have presented a less obviously coercive face. But in fact this division of labor was often blurred. Until the first civilian governor-general arrived in Taiwan in 1919, Japanese teachers, as public officers, bore swords at ceremonies and kept sabers in their offices that they carried when going out. 122 During the kominka movement, public school teachers also provided military training to Taiwanese youths. 123 In aborigine districts, furthermore, education was usually provided by policemen. Some teachers believed that assimilation would end discrimination against the Taiwanese. 124 But because the aim of assimilation was to foster voluntary submission to the emperor, under the historical circumstances in which their daily practices were situated, even teachers' devotion and good intentions were destined to serve the goal of Japanese imperialism. And of course, cultural chauvinism and patronizing attitudes are also evident in Kimura's survey.

After colonialism, the Taiwanese condemned the notion of the Shizangan spirit and the aim of acculturating the colonized to the detriment of their native languages and cultures. 125 But it is also true that more than a half-century after the end of colonial education, elderly Taiwanese (including some of my relatives) still visit their former Japanese teachers to pay their continued respects. In the final analysis, the meaning of what teachers did in the scene of colonial education is multifold. On the micro level of personal [End Page 527] connections, a colonized people such as the Taiwanese may appreciate the tutelage they receive, yet they may be critical when examining such education against the backdrop of the macro-level social process of colonialism.

The Japanese Immigrants and the Colonized
In the end, the abstract discourse on the concepts of Kokugo and Nihongo and language pedagogy, as well as the best efforts of the most enthusiastic teachers, simply failed in predicting the language behavior of real people in the streets of Taiwan, both Japanese and Taiwanese. 126 Their language practices took their own trajectories, diverging from the route drawn by colonial scholars and policymakers. Japanese immigrants, in their dealing with everyday chores, let practicality take precedence over Kokugo. As Tsurumi documents, when schools for Japanese in Taiwan started, the children, who came from different parts of Japan and spoke different dialects, invented for their communication in the classroom and on the playing field a common language or "a kind of Taiwan-style abbreviated Japanese without proper word endings." 127 In the 1940s even the level of Kokugo competence of Taiwanese-born Japanese children was found to be considerably lower than on the mainland. 128 Adults, too, varied in their native tongues and infused their familiar, dialectal phrases into the Kokugo they spoke among themselves. Further, in the streets of Taiwanese cities, where Japanese needed to communicate with local Taiwanese who spoke little Japanese, these two groups of people together invented a pidgin Japanese in which Japanese words were put together in a Taiwanese order. 129 Predictably, this patois was severely criticized by officials as denigrating the sacredness of Kokugo. 130

If Japanese people maintained diverse speech habits despite the state's imposition of Kokugo, so indeed did the Taiwanese. Statistics show that after the kominka movement began, the number of people who understood Japanese increased rapidly; comprising less than 40 percent of the Taiwanese population before 1937, their numbers increased to 57 percent by 1941. 131 These official statistics are impressive. But as Murakami Yoshihide points out, the actual Japanese-language competence of persons counted in such statistics is questionable. With little opportunity to use the language, former [End Page 528] students were apt to forget it completely after leaving school. Sixty percent of the Taiwanese lived in rural areas where they met few Japanese, and even those who lived in cities had little contact with the immigrants. In addition, the statistics included those who studied Japanese for only several months on breaks from farming. 132 As Robert L. Cheng states, "monolingual Japanese children existed even in Taiwanese families," but such cases were restricted to special situations, such as "in government housing areas where the Japanese population was proportionally higher" or areas where different groups of Chinese descent lived together with Kokugo as their common language. 133

More importantly, language practice aside, colonial education must still be seen as far from successful when evaluated against its ultimate goal, that is, the transformation of the colonized into Japanese. The development of Kokugo competence did not necessarily lead to the cultivation of the Japanese spirit on the part of the colonized, as Izawa, Yamaguchi, and many teachers had believed it would. The majority of Taiwanese people simply never internalized the Kokugo ideology. Although they might maintain Shinto altars or Japanese flags in their houses, they never internalized the values associated with these Japanese national symbols but, instead, continued their own cultural practices. 134

To be sure, many Taiwanese elites did learn to use Japanese superbly. Yet, even for those people, whether as a result they were Japanized is a different issue. They studied Kokugo as a language of civilization so that they could acquire modern knowledge and succeed in colonial society to the fullest extent possible. But there were also Taiwanese who used their competence in the colonizer's language to liberate themselves and to challenge the goal of assimilation that Kokugo education was supposed to achieve. 135

Just as many educated Taiwanese people refused to speak Japanese, there were also those who excelled in the language yet refused to be honored as members of the "Kokugo home" by the colonial government. 136 Those adopting the customs of the Kokugo home faced criticism from their countrypeople for the betrayal of their culture, and those who spoke Japanese without a Taiwanese accent aroused antipathy. 137 A story passed down among locals tells of a Taiwanese who walked all the hilly way from Hualián to Taipei just to avoid speaking Kokugo at a train station. 138 In his [End Page 529] analysis of a Taiwanese village, Kondo concludes that a compulsory draft was impossible because of the presence of a group who stubbornly refused to speak Japanese, thus demonstrating the failure of the kominka movement to penetrate Taiwanese communities. 139

In conclusion, the Kokugo ideology had little impact on the majority of Taiwanese people. They were far from becoming Japanese, either in their language behaviors or, even more certainly, in their mental attitudes. As Albert Memmi taught us more than thirty years ago, "under the contemporary conditions of colonization, assimilation and colonization are contradictory." 140 While the colonizers struggled to theorize and put into practice the Japanization of non-Japanese people, the colonized knew very well that they were in reality precluded from becoming Japanese, however well they were able to speak the language. Nor did they fail to notice the social and legal discrimination that they suffered by virtue of being colonized. It is true that for a small number of Taiwanese, Kokugo did not remain a mere tool but began to penetrate their value system as well. 141 Their sought-after identification with the Japanese was, however, destined to fail. As Memmi rightly points out, "in order to be assimilated, it is not enough to leave one's group, but one must enter another; [at this point] he meets with the colonizer's rejection." 142

The assimilationist policy of imperial Japan in Taiwan was never implemented apart from the legal distinction between the colonized and the colonizers, which coincided with ethnic difference, pervaded the lives of both Japanese and Taiwanese, and in turn probably helped them to consolidate mutually the respective ethnic boundaries that separated them. Assimilation was in fact popular among neither Japanese immigrants nor the Taiwanese themselves. The former resisted because they wanted to maintain their privilege and because of their chauvinistic attitudes. 143 Kokugo, which scholars and administrators assumed would foster loyalty to the state, was clearly perceived by both groups to be the language of ethnic Japanese. For the people in the streets of Taiwan, Kokugo was the language of the Yamato minzoku, just as Ueda argued it should be, not the language of all the people under the Japanese empire, as Tokieda had advocated. [End Page 530]

Conclusion
In the context of Japanese colonial education, the Kokugo ideology provided a scheme for the assimilation of the colonized. Policymakers variously played down or stressed the goal of cultural assimilation of the colonized, depending on the changing direction of colonial policy, itself shifting according to changing strategic, economic, and military aims, while continuing to maintain legal discrimination against the Taiwanese. Certain scholars of pedagogy, such as Izawa and Yamaguchi, seriously sought to Japanize the colonized through Kokugo education. Following their lead were Japanese classroom teachers, the first-line disseminators of the language and its ideology. Despite policy, theory, and local efforts, however, the mass of Japanese and Taiwanese continued to speak their native languages, acquiring Kokugo language skills to varying degrees without internalizing the ideology. Further, most Taiwanese people either ignored or rejected the idea of becoming Japanese.

Nevertheless, with modifications of its arguments, emphasis, and interpretations, the Kokugo ideology thrived throughout the period of Japanese imperialism and survived into postwar Japan. After the war's end the universal application of the Kokugo ideology was abandoned as a project, and the ideology instead became an essential component of the theory of Japan as a single-ethnic-nation state (tan'itsu minzoku kokka-ron), which has dominated the minds of the majority of Japanese in the postwar era, underwriting their strong sense of ethnic national identity. 144 Interestingly, Ueda's equation of kokumin and minzoku, which arguably posed problems when the narrative of assimilation was applied to Japan's colonies, has been a principal thesis in the postwar development of Kokugo theory. 145 In short, Kokugo has come to be regarded as the tightly guarded, exclusive property of Nihonjin. In the postwar era, the nationalist narrative on Kokugo has become a self-evident component of the cognitive frameworks of the Japanese. 146

Ironically, colonialism may have been crucial to the formation of this postwar Japanese concept of Kokugo. By continuously posing questions about where to draw the boundaries of a Kokugo speech community or the boundaries of the Japanese nation and by persistently separating the colonized [End Page 531] from ethnic Japanese, colonial education (which purported to stretch Kokugo toward the universal-imperial) actually helped to consolidate it as a particular-domestic language. The assimilationist colonial education, which required the idea of openness to Japanese culture, paradoxically served to facilitate the process of exclusion that helped to enable the establishment of "Japaneseness."

The past decade has witnessed the second tide of the spread of the Japanese language to non-Japanese peoples. While, in this context, Japanese has been taught as Nihongo, there has also emerged a growing tendency to use the term Nihongo in place of Kokugo domestically, for example, in the names of academic subjects and departments. Koyasu criticizes the scholars debating the special topic "'Kokugo' or 'Nihongo'" in the June 1994 issue of the journal Nihongo-ron, on the grounds that the discussion neglects the political and historical significance of the term Nihongo. 147 As the term Nihongo reemerges, many scholars engaged in Japanese education seem to be indifferent to its colonial history. In this they risk approximating the behavior of the politicians who have angered Japan's neighbors by their refusal to acknowledge the outrages of imperialism.

This being the case, classroom teachers of Japanese must be especially alert to the political significance of our work. This requires an awareness of the evolution of the concept of Nihongo in modern Japanese history, lest we devote ourselves blindly to an unrecognized political goal superimposed on our personal objectives.

At this historical moment, when the isomorphism between the Japanese blood and the Japanese language is a thesis often taken for granted by many Japanese, 148 I join with those Koreans residing in Japan who call themselves Nihongojin (Japanese-speaking person). In this implicit assertion of hybridity, I attempt both to suspend the Nihonjin-Kokugo thesis and to be cautious about my own strategic position as it unfolds.

Eika Tai is associate professor of Japanese at San Francisco State University. She is the author of Tabunkashugi to Diasupora: Voices of San Francisco (1999).

Notes
*The initial stage of this project was supported by an Affirmative Action Award from San Francisco State University. I thank the anonymous reviewers for their comments. I also thank Nancy Kool and Janet Tallman for editing the article.

1. "The Taiwanese" here include the aborigines and the Han, both culturally and linguistically diverse. According to the Japanese national census, of the Taiwanese population of 2.8 million in 1897, more than 90 percent were Han Chinese. This article focuses on education for the Han; I plan a separate essay on education for the aborigines. Kokugo can be translated differently, depending on how one interprets the first character, koku, which can mean country, state, nation, or nation-state. Since this term does not specify any country, it discourages its users from perceiving the language vis-à-vis other national languages; rather, it encourages them to see a unity of the language, the state, and themselves as natural.

2. Lee Yeounsuk, "Kokugo" toiu shiso: Kindai Nihon no gengo ninshiki [The thought of Kokugo: Language awareness in modern Japan] (Tokyo: Iwanami, 1996), 72.

3. Benedict Anderson, Imagined Communities: Reflections on the Origin and Spread of Nationalism (London: Verso, 1983), 81. Anderson uses this phrase to describe the role of German in Austro-Hungary after the mid-nineteenth century.

4. Prior to the acquisition of Taiwan, Japan had colonized Hokkaido (including Chishima), Okinawa, and Ogasawara but treated them as part of Japan proper under the Meiji Constitution. See Tamura Sadao, "Naikoku shokuminchi toshite no Hokkaido" [Hokkaido as an internal colony], in Kindai Nihon to shokuminchi [Modern Japan and its colonies] (Tokyo: Iwanami, 1992), 1:87-99; Tomiyama Ichiro, Kindai Nihon shakai to "Okinawajin": "Nihonjin" ni naru toiu koto [Modern Japanese society and "the Okinawans": Becoming "Japanese"] (Tokyo: Nihon Hyoron, 1990); and Alan S. Christy, "The Making of Imperial Subjects in Okinawa," positions 1, no. 3 (winter 1993): 607-639. For the differences between internal and formal colonization, see Matayoshi Seikiyo, Nihon shokuminchika no Taiwan to Okinawa [Taiwan and Okinawa under Japanese colonial rule] (Okinawa: Aki Shobo, 1990).

5. The word Nihongo was often pronounced as Nippongo during this period; I spell it as Nihongo throughout the text for consistency and to follow current custom in Japan.

6. Hirataka Fumiya, "Language-spread Policy of Japan," International Journal of the Sociology of Language 95 (1992): 93-108.

7. Of the literature on Japanese colonial education written in English, the work of E. Patricia Tsurumi has been by far the most important. See her Japanese Colonial Education in Taiwan, 1895-1945 (Cambridge, Mass.: Harvard University Press, 1997).

8. Naoki Sakai argues that eighteenth-century discourse projected an idealized unity of the Japanese language onto the ancient world in such a way that "the present was analyzed as a lack" or a loss of a primordial unity. Japanese was "stillborn" into this discourse. See Naoki Sakai, Voices of the Past: The Status of Language in Eighteenth-Century Japanese Discourse (Ithaca, N.Y.: Cornell University Press, 1992), 310-311.

9. To use Sakai's term, fukkatsu.

10. Ivan Parker Hall, Mori Arinori (Cambridge, Mass.: Harvard University Press, 1973), 189.

11. Lee argues that later critics such as Yamada Yoshio, Ohno Susumu, and Tokieda Motoki neglect this fact when they accuse Mori of attempting to abolish Kokugo in favor of English.

12. Lee Yeounsuk, "Mori Arinori to Baba Tatsui no Nihongo-ron" [Theories of the Japanese language by Mori Arinori and Baba Tatsui], Shiso [Thought], no. 795 (1990): 49-64.

13. Furuta Tosaku, "'Kokugo' ishiki no seiritsu" [Establishment of Kokugo consciousness], in Nihon no gengo bunka [Language culture in Japan], ed. Furuta (Tokyo: Nihon Hoso, 1989), 61-73.

14. From this context Ueda appears to use the term jinshu (race) to mean ethnicity.

15. Irokawa Daikichi, The Culture of the Meiji Period, trans. and ed. Marius B. Jansen (Princeton, N.J.: Princeton University Press, 1985), 247.

16. Ueda Kazutoshi, "Kokugo to kokka to," in Meiji bungaku zenshu [The complete works of Meiji literature], vol. 44 (Tokyo: Chikuma, 1968), 108-113.

17. Nakamura Tetsuya, "The Nation-State-Building and National Language Discussions in the Meiji Era," Bulletin of the Faculty of Education, the University of Tokyo 27 (1987): 207-216.

18. Lee, "Kokugo" toiu shiso, 122-123.

19. For example, see Monsur Musa, "Politics of Language Planning in Pakistan and the Birth of a New State," International Journal of the Sociology of Language 118 (1996): 63-80.

20. E. J. Hobsbawm, Nations and Nationalism since 1780: Programme, Myth, Reality (Cambridge: Cambridge University Press, 1990), 54.

21. Kevin M. Doak, "What Is a Nation and Who Belongs? National Narratives and the Ethnic Imagination in Twentieth-Century Japan," American Historical Review 102, no.2 (April 1997): 283-309, 284-285. In this he concurs with those stressing the ethnic foundation of national identity.

22. Prasenjit Duara, Rescuing History from the Nation: Questioning Narratives of Modern China (Chicago: University of Chicago Press, 1995), 10.

23. For example, see Mary S. Erbaugh, "Southern Chinese Dialects as a Medium for Reconciliation within Greater China," Language in Society 24, no. 1 (March 1995): 79-94. See also Kathryn A. Woolard, "Language Variation and Cultural Hegemony: Toward an Integration of Sociolinguistic and Social Theory," American Ethnologist 12, no. 4 (November 1985): 738-748.

24. Before the start of radio broadcasting in 1925, a standard language in its spoken form was far from established in Japan. See Tessa Carroll, "NHK and Japanese Language Policy," Language Problems and Language Planning 19, no. 3 (fall 1995): 271-293. In 1942 Kieda Masuichi, a professor at a teacher-training school, deplored the fact that many Japanese made light of Kokugo. See Kieda Masuichi, Kokugo no michi [Way of Kokugo] (Osaka: Dekijima, 1942), 90-102.

25. Lee, "Kokugo" toiu shiso, 152-153.

26. Ueda Kazutoshi, "Kokugo kenkyu nitsuite," in Meiji bungaku zenshu, 44: 114-118.

27. Ueda Kazutoshi, "Kokumin kyoiku to Kokugo kyoiku" [Kokumin education and Kokugo education], in ibid., 146-155.

28. Anderson, Imagined Communities, 88.

29. Doak, "National Narratives and the Ethnic Imagination," 284.

30. See Oguma Eiji, Tan'itsu minzoku shinwa no kigen [The myth of the homogeneous nation] (Tokyo: Shinyo-sha, 1995).

31. Kokufu Tanetake, Taiwan niokeru Kokugo kyoiku no tenkai [Development of Kokugo education in Taiwan] (Taipei: Daiichi Kyoiku-sha, 1931; rpt.; Tokyo: Toji Shobo, 1988).

32. Yun Kwon-cha, "Meiji-zenki kokumin keisei-ron no tenkai" [Development of nation-building theory in early Meiji], Kyoikugaku kenkyu [Studies on education] 49, no. 2 (1982): 195-204.

33. Sai Moho, Taiwan ni okeru Nihongo kyoiku no shiteki kenkyu [A historical study of the teaching of Japanese in Taiwan] (Taipei: Dongwu University, 1989), 568-569.

34. Ibid., 567-568.

35. .

36. Hirotani Takio and Hirokawa Toshiko, "A Comparative Study of the Policy of Colonial Education in Formosa and Korea under the Rule of Japan," Bulletin of the Faculty of Education, Hokkaidô University 22 (1973): 19-92.

37. Iwamoto Yumiko, "Shuji Izawa and Japanese Language Education: The Groundwork for Japanese Language Education in Taiwan," Journal of Japanese Language Teaching 60 (1986): 11-41.

38. Sai, Nihongo kyoiku no shiteki kenkyu, 547-549.

39. Kondo Sumiko, "Shizangan Spirit," Journal of Japanese Language Teaching 60 (1986): 42-53.

40. Oguma, Tan'itsu minzoku shinwa no kigen, 65-71.

41. Tsurumi, Japanese Colonial Education in Taiwan, 43.

42. Sai, Nihongo kyoiku no shiteki kenkyu, 573-581.

43. Haruyama Meitetsu, "Meiji kenpo taisei to Taiwan tochi" [Meiji constitutional system and colonial rule in Taiwan], in Kindai Nihon to shokuminchi (Tokyo: Iwanami, 1993), 4:31-50.

44. Tanaka Hiroshi, "A Historical Survey on the Voting Right and the Compulsory Military Service of the Formosan and Korean Peoples under the Japanese Colonization," Journal of the Faculty of Foreign Studies, Aichi Prefectural University 9 (1974): 61-96.

45. Officially, this was the koko chosabo system.

46. Ibid., 68.

47. Ubukata Naokichi, "Tan'itsu minzoku kokka no shiso to kino" [Thought and function of a single-ethnic-nation state], Shiso, no. 656 (1979): 23-37.

48. The Han were called Taiwanjin (Taiwanese) or Hontojin (island people).

49. Intermarriage between Naichijin and Taiwanese was not legalized until 1917; then, typically, Taiwanese men married Japanese women, with the latter registered in their husbands' koseki.

50. Haruyama, "Meiji kenpo taisei to Taiwan tochi."

51. Kitaoka Shin'ichi, Goto Shinpei (Tokyo: Chuko Shinsho, 1988), 40.

52. Mark R. Peattie, "Japanese Attitudes toward Colonialism, 1895-1945," in The Japanese Colonial Empire, 1895-1945, ed. Ramon H. Myers and Peattie (Princeton, N.J.: Princeton University Press, 1984), 80-127.

53. Concerned only with developing a Japanese-speaking network among local elites, Goto told his education personnel that "they must take care to see that Taiwanese did not become educated above their stations in life." See Tsurumi, Japanese Colonial Education in Taiwan, 23.

54. Hirotani and Hirokawa, "Comparative Study of the Policy of Colonial Education," 21-22.

55. Sai, Nihongo kyoiku no shiteki kenkyu, 49, 75-76.

56. Sho Seikan, Nihon shokuminchi-ka niokeru Taiwan kyoiku-shi [A history of education in Taiwan under Japanese colonial rule] (Tokyo: Taga, 1993), 113.

57. Ibid., 116-117.

58. Komagome Takeshi, "I-minzoku shihai no 'kyogi'" [Doctrine for the control of other ethnic nations], in Kindai Nihon to shokuminchi 4:137-155.

59. Haruyama, "Meiji kenpo taisei to Taiwan tochi," 46.

60. Sho, Taiwan kyoiku-shi, 90.

61. Sai, Nihongo kyoiku no shiteki kenkyu, 172.

62. Yanaihara Tadao, Teikokushugi-ka no Taiwan [Taiwan under imperial rule] (Tokyo: Iwanami, 1934; rpt., Iwanami, 1988), 158-159.

63. Sho, Taiwan kyoiku-shi, 194.

64. Sai, Nihongo kyoiku no shiteki kenkyu, 159.

65. Tsurumi, Japanese Colonial Education in Taiwan, 145.

66. Sho, Taiwan kyoiku-shi, 184.

67. Ng Yuzin Chiautong, Taiwan sotokufu [Colonial government in Taiwan] (Tokyo: Kyoikusha, 1981), 135-137, 155-158. These reforms were no doubt efforts to appease activists of the "movement for the establishment of a Taiwan parliament" (Taiwan gikai secchi undo), which was founded in 1921. Their fifteenth and final petition was turned down in 1934.

68. Kondo Masami, "The Mobilization of War Service Laborers and the Policy of Complete Japanization in Wartime Taiwan," in Historical Studies of Taiwan in Modern Times (Tokyo: Ryokuin, 1988), 6:115-164.

69. Ohe Shinobu, "Higashi Ajia shinkyu teikoku no kotai" [Change from the old to the new empire in East Asia], in Kindai Nihon to shokuminchi 1:3-31.

70. Peattie, "Japanese Attitudes toward Colonialism," 124.

71. Kondo, "Mobilization of War Service Laborers," 121.

72. Sho, Taiwan kyoiku-shi, 205-207.

73. Sai, Nihongo kyoiku no shiteki kenkyu, 169.

74. Kondo, "Mobilization of War Service Laborers," 142-148.

75. Sai, Nihongo kyoiku no shiteki kenkyu, 79.

76. Tsuchiya Tadao, "Taiwan Hontojin no kominka to gimu kyoiku no shiko" [Imperialization of the Han Taiwanese and the implementation of compulsory education], in Kokumin kyoiku no doko [Direction of education for the nationals], ed. Kyoiku shicho kenkyukai (Tokyo: Meguro, 1943), 166-186. Conscription of the Taiwanese began in 1945.

77. Sai, Nihongo kyoiku no shiteki kenkyu, 163.

78. Ibid., 75.

79. Tsuchiya Tadao, "Taiwan niokeru 'kominka' kyoiku" [Education for imperialization in Taiwan], in Gendai kyoikugaku [Modern education] (Tokyo: Iwanami, 1962), 5:342-347.

80. Sai, Nihongo kyoiku no shiteki kenkyu, 583.

81. Taiwan kyoikukai, Izawa Shuji sensei to Taiwan kyoiku [Izawa Shuji and education in Taiwan] (Taipei: Taiwan kyoikukai, 1944), 1-2.

82. Ienaga Saburo, The Pacific War, 1931-1945 (New York: Pantheon, 1978), 154.

83. In contrast to the term Kokugo, whose first character, koku (state), means "the Japanese state" only to its members, the term Nihongo identifies the language as Japanese (Nihon) and thus opens up a comparative perspective on the language.

84. Komagome Takeshi, "The Policy of Japanese Language Education during the Sino-Japanese War: Plans of the Ministry of Education and the Asia Development Board," Bulletin of the Faculty of Education, the University of Tokyo 29 (1990): 179-188.

85. Matsuo Chozo, "Hakkan no ji" [A remark on new publication], Nihongo 1, no. 1 (1941; rpt., Tokyo: Toji Shobo, 1985): 4-5.

86. Koyasu Nobukuni, "'Kokugo' wa shishite 'Nihongo' wa umareta ka" [Has "Kokugo" died? And has "Nihongo" been born?], Gendai shiso [Review of contemporary thought] 22, no. 9 (August 1994): 45-57.

87. Ando Masatsugu, "Nihongo no shinshutsu to Nihongo no kyoiku" [Advancement of Japanese and Japanese education], Nihongo 1, no. 2 (1941; rpt., Tokyo: Toji Shobo, 1985): 4-10.

88. Koyasu, "'Kokugo' wa shishite 'Nihongo' wa umareta ka," 49.

89. Ando Masatsugu, "Nihongo no muzukashisa" [Difficulty of Japanese], Nihongo 2, no. 3 (1942; rpt., Tokyo: Toji Shobo, 1985): 4-11.

90. Koyasu, "'Kokugo' wa shishite 'Nihongo' wa umareta ka," 45-57.

91. Tokieda Motoki, "Chosen niokeru Kokugo seisaku oyobi Kokugo kyoiku no shorai" [The future of Kokugo policy and Kokugo education in Korea], Nihongo 2, no. 8 (1942; rpt., Tokyo: Toji Shobo, 1985): 54-63.

92. Ando, "Nihongo no muzukashisa."

93. Tokieda, "Chosen niokeru Kokugo seisaku."

94. Ibid., 58-59.

95. Ibid. 56, 60-61.

96. Ibid., 62. Here Tokieda recuperates the concept of mother tongue by placing it in relation to the state rather than to the ethnic nation. This shift neutralizes the contradiction in Ueda's narrative, in which Kokugo had been represented as the mother tongue of the ethnic Japanese, in spite of the fact that most of them spoke regional vernaculars.

97. By using the supremacy of the Japanese state to justify both the spread of Kokugo among Japanese, in preference to dialects, and the spread of Kokugo among Koreans, Tokieda avoided any consideration of the potentially different effects of Kokugo education on the respective ethnic consciousnesses of these peoples.

98. Ibid., 61.

99. For a discussion of still other views, see Lee, "Kokugo" toiu shiso, 282-310.

100. In Taiwan, Terakawa Kishio, a prominent scholar of phonology, wrote in 1942 that the Japanese language the Han Taiwanese spoke could be called a Taiwan dialect of Kokugo, based on the accent with which certain Japanese sounds were pronounced; the Han Taiwanese were thus native speakers of a dialect of Kokugo. See Terakawa Kishio, Taiwan niokeru Kokugo on'in-ron [Phonology of Kokugo in Taiwan] (Taipei: Taiwan Gakugei, 1942), 659.

101. Sai, Nihongo kyoiku no shiteki kenkyu, 480-481.

102. Yamaguchi Kiichiro, Gaikokugo toshite no waga Kokugo kyoju-ho [Methodology of teaching our national language as a foreign language] (1933; rpt., Tokyo: Toji Shobo, 1988), 472-474.

103. Ibid., 121.

104. Ibid., 154-165.

105. Ibid., 161-163.