Tropics of Savagery (1) by Robert Thomas Tierney

 ASIA PACIFIC MODERN

Takashi Fujitani, Series Editor

i. Erotic Grotesque Nonsense: The Mass Culture of Japanese Modern Times, by Miriam Silverberg

2. Visuality and Identity: Sinophone Articulations across the Pacific, by Shu-mei Shih

3. The Politics of Gender in Colonial Korea: Education, Labor, and Health, 1910-1945, by Theodore Jun Yoo

4. Frontier Constitutions: Christianity and Colonial Empire in the Nineteenth Century, by John D. Blanco

5. Tropics of Savagery: The Culture of Japanese Empire in Comparative Frame, by Robert Thomas Tierney

====

Tropics of Savagery The Culture of Japanese Empire in Comparative Frame

Robert Thomas Tierney

CONTENTS

Acknowledgments ix

Introduction 1

1. From Taming Savages to Going Native: Self and Other on the Taiwan Aboriginal Frontier 38

1. Ethnography and Literature: SatO Haruo's Colonial Journey to Taiwan 78

3. The Adventures of MomotarO in the South Seas: Folklore, Colonial Policy, Parody 110

4. The Colonial Eyeglasses of Nakajima Atsushi 147

Conclusion: Cannibalism in Postwar Literature 182

Notes 199

Glossary of Japanese Terms 249

Bibliography 255

Index 287

ACKNOWLEDGMENTS

A work of scholarship is inevitably a collaborative endeavor; here, I would like to acknowledge the people who made the most significant contributions to my research over the past decade.

I am most indebted to Jim Reichert, my dissertation adviser at Stanford Uni­versity, who was a wonderful mentor throughout my graduate studies. I would also like to thank Peter Duus and Miyako Inoue, who challenged me with the broader perspectives of their respective disciplines and helped me to approach my project from an interdisciplinary viewpoint. Susan Matisoff and Torn Hare were both ex­acting teachers who inspired in me a deep love for Japanese literature and for the craft of translation. During my last years at Stanford, I was greatly stimulated by the critical comments from Karen Wigen and Steven Carter. Throughout my years as a graduate student, I had the good fortune to be surrounded by friends with whom I have shared both ideas and laughs. I would especially like to thank Julia Bullock, Claire Cuccio, Michael Foster, Mark Gibeau, David Gundry, Shu Kuge, Miri Nakamura, Christopher Scott, Ethan Segal, Roberta Strippoli, Daniel Sullivan, and Michiko Suzuki.

In Japan, I had the honor of working with Kawamura Minato, a scholar at HO-sei University who has almost single-handedly created the field of Japanese colo­nial literature studies. Professor. Kawamura offered me access to his book collec­tion from the colonial period and shared with me his extensive understanding of the field. I would also like to express my appreciation to Professors Araki Masazumi and Yoshihara Yukari for inviting me to take part in the Critical Culture Research Group at Tsukuba University. Through my regular participation in this group, I met and exchanged ideas with many young Japanese, Taiwanese, and Korean scholars

ix

X ACKNOWLEDGMENTS ACKNOWLEDGMENTS Xi

and had a chance to present my own research to this exacting group. I would par­ticularly like to thank Hibi Yoshitaka, SaitO Hajime, Washitani Hana, and Wu Pei-chen for their encouragement and advice.

As a graduate student at Stanford University, I received financial support from the Department of Asian Languages, the Center for East Asian Studies, and the In­stitute for International Studies. I began to conceive of my dissertation project dur­ing one year at Shizuoka University on a generous fellowship from the Shizuoka Prefecture. I conducted most of my archival research at HOsei University with sup­port from the Department of Education and the Fuibright Foundation. At the Uni­versity of Illinois, I received a Mellon Foundation Grant for Junior Faculty in 2oo6-7, which released me from teaching responsibilities for one semester and enabled me to turn my dissertation into a book manuscript. From 2008 to 2009, the Japan Society for the Promotion of Science (JSPS) provided financial support for me to spend a year in Japan. I would like to thank the Department of Comparative Cul­tures at the University of Tsukuba for hosting me during the period that I was re­vising the final manuscript.

Since I began to teach at the University of Illinois at Urbana-Champaign, I have been fortunate to be surrounded by colleagues with whom I can exchange ideas. I would especially like to thank Nancy Abelman, Nancy Blake, Marilyn Booth, David Goodman, Wail Hassan, Karen Kelsky, Sho Konishi, Elizabeth Oyler, Dan Shao, Ron Toby, and Hairong Yang for their warm support and friendship. I would also like to express my gratitude to the following individuals who commented on earlier versions of my work, whether conference presentations or drafts of articles: Paul Barclay, Leo Ching, Kim Kono, Helen Lee, Michele Mason, Ann Sokoisky, and Mariko Tamanoi. I would like to thank Naomi Kotake at the Stanford University East Asian Library and Setsuko Noguchi at the University of Illinois for their as­sistance in finding my way through Japanese source materials.

I must thank Reed Malcolm, editor at the University of California Press, and Tak Fujitani, editor of its Asian Pacific Modern series, for their faith in my project and their patience; two anonymous readers for UC Press provided invaluable sugges­tions to improve my manuscript. I am deeply indebted to Jacqueline Volin and An­drew Frisardi, whose informed and careful editing of my manuscript has made my prose more readable. I am also grateful to Daniel Sullivan, 'who proofread the book. I would also like to express my appreciation to Deng Xiang Yang, who provided me the photograph of the second Musha Incident that appears on this book's cover. Chapters 2 and 4 of this book are expanded versions of my articles "Ethnography, Borders, and Violence: Reading between the Lines in SatO Haruo's Demon Bird," in Japan Forum 191 (2007), pp. 89-110; and "The Colonial Eyeglasses of Nakajima Atsushi," in Japan Review 17 (2005), 149-96. I would like to express my gratitude to Ann Waswo, editor of the Japan Forum, James Baxter, editor of the Japan Review, and to the anonymous readers for both publications for their many helpful com­ments and suggestions.

Last but not least, I would like to thank Hiromi Matsushita, my partner over the past twenty-five years. While I often doubted my ability to bring this project to com­pletion, Hiromi never wavered in his confidence in me. I dedicate this work—and others to follow—to him.

Introduction

Tropics of Savagery looks at the culture of imperial Japan, the most important non-Western colonizer of modern times. It consists of a series of historically situated studies of literary works and of colonial tropes focusing on the theme of "savagery" in Japanese imperial culture.' Borrowing from Hayden White's Tropics ofDiscourse, the title plays on the dual meaning of the word "tropics" to refer to two related as­pects of Japanese imperialism. On the one hand, Japan ruled over colonies situated in the "tropics' although this fact has not seemed especially important to most his­torians of the Japanese empire. On the other, Japan exercised domination over its colonies through the deployment of "tropes' that is, figures of speech, as well as through military conquest, political control, and economic exploitation. As Nicholas Thomas writes of the English empire: "Colonial culture includes not only official reports and texts related directly to the process of governing colonies and extract­ing wealth, but also a variety of travelers' accounts, representations produced by other colonial actors such as missionaries and collectors of ethnographic specimens, and fictional, artistic, photographic, cinematic and decorative appropriations. 112 In this book I focus on tropes of savagery in Japanese literature and representations that Japanese writers made of the tropics during the colonial period.

Many Japanese writers traveled to Japan's tropical colonies and wrote fictional works, travelogues, popular articles, and a vast variety of other texts about savage or primitive societies In this book I consider works by Oshika Taku, SatO Haruo, Akutagawa Ryunosuke, Nitobe InazO, Hijikata Hisakatsu, and Nakajima Atsushi that appeared during Japan's colonial period (1895-1945). These writers made the sav­age a foil against which the Japanese constituted themselves as members of a mod­ern, civilized nation. At different stages of Japan's colonial trajectory, the savages

1

2 INTRODUCTION INTRODUCTION 3

were headhunters to be eradicated, primitive societies to be studied "noble savages" who had escaped from the blight of Japan's industrial modernity; or hybrid subjects expected to conform to Japanese cultural norms. Just as images of savagery changed over time, the Japanese defined themselves in various ways in relation to thosethey colonized: as conquerors bearing the gifts of civilization to backward others, as ethnographers studying these others to find the hidden order of their society, as nostalgic romanticists in flight from civilization; and as colonial officials promot­ing assimilation policies. As I will show, savagery in these literary texts is less are-alistic description than a polyvalent trope that writers have employed to depict their encounters with other cultures and histories, as well as to tell about their own am­bivalent experiences as Japanese colonizers. As a result, these works open an espe­cially important window onto both the images of the societies that Japan ruled. and the lived experience of Japanese colonizers.

Alongside literary works, I also examine other discourses in Japan's colonial archives: ethnography, anthropology, colonial policy studies, folklore, and eugen­ics. In the late nineteenth century, Japanese scholars imported new academic dis­ciplines from the West, but they quickly established autonomous branches of these disciplines within Japan. These disciplines offered paradigms of knowledge that were sometimes applied to solving the practical problems of ruling an empire. In return, the realities of ruling an empire informed the conceptual framework of these dis­ciplines and shaped representations that the Japanese made of themselves and their nation. To consider only the case of the human sciences, Japanese anthropologists, ethnologists, and archaeologists—all pioneers in newly created fields of knowl-edge—formulated new ideas of Japan that situated it in the new spaces of empire and defined its relationships to the colonies. First, they conceived of Japan not as constituting an isolated archipelago but rather as forming a continuum with the continent of Asia and the Pacific region. Second, they stressed resemblances and analogies, rather than differences, between the people whom Japan ruled and the Japanese people. These new paradigms of knowledge justified empire as a territo­rial unification of areas originally one but later divided, and they legitimized colo­nialism as a system of rule over people who were related to the Japanese. In this book, I pay close attention to the development of these new paradigms of knowl­edge in order to situate literary texts in their context and link these texts to the wider sociopolitical nexus in which they were embedded.

In my studies of Japanese colonial writers, I highlight the connections between literary texts and social and intellectual contexts and ignore the boundaries that traditionally demarcate literature, popular culture, social science, and history. While I approach each literary text as a singular work to be understood on its own terms, I also believe that literature is in dialogue with other social discourses, in­cluding legal writings, scientific reports, Western colonial narratives, popular songs, and school textbooks. In my view, the fiction writer is not the architect who designs the habitation of social discourses or the builder who constructs it. Rather, he or she is the tenant who has little choice but to make a home there, much as he or she inhabits the language in which he or she writes. Nor is the writer's relationship to environing discourses predetermined, simple, or unidirectional. A writer may sometimes parrot, paraphrase, reproduce, and disseminate the tropes or topoi that circulate within society at any moment, but he or she may also invert, ridicule, and transform dominant discourses by relativizing or parodying them, or simply by in­scribing them in a fictional world. Indeed, even when a literary work restricts itself to reproducing these social discourses, it inevitably distances itself from them by the playful nature of the literary text. In letter on Art in Reply to Andre Daspre' Althusser argues that literary works are significant because they make us perceive "from the inside, by an internal distance, the very ideology in which they are held" In short, Althusser treats ideology as the raw material that literature transforms and puts on display.'

Furthermore, the works that I study mirror the shifts in Japanese views toward imperialism and the trajectory of Japan's imperial expansion over time. Nitobe's es­says are products of the early formation of the Japanese empire and of Japan's in­corporation into a global imperialist order. The stories of SatO and Akutagawa res­onate with the liberal critique of colonialism in the 1920s. Oshika's work offers an interesting counterpoint to the "return to Japan" movement in Japanese letters in the 1930S, and Nakajima's South Seas fiction intersects with 194Os rhetoric of the Greater East Asian Co-Prosperity Sphere. In my studies of specific texts, I locate literature within changing configurations of Japan's empire and highlight the in­stability of imperial discourse over time.

In addition to studying literary works and their relationship with social dis­courses, I use these studies to complicate the application of Western-oriented par­adigms to Japanese colonial literature. Japan is largely neglected in the major works of postcolonial studies. Yet its place in the history of modern empires is paradoxi­cal. Modern Japan was never colonized; nevertheless it was the product of a semi-colonial collision between an Asian society and the expanding West, After Com­modore Perry forced the Tokugawa shogunate to open its ports to international commerce, Japan signed "unequal treaties" with Western powers according to which Western nationals residing in Japan enjoyed the privilege of extraterritoriality. Japan remained under this system until the early twentieth century.4 At the same time that Japan was renegotiating with Western powers to replace these one-sided treaties with reciprocal agreements, it was quickly moving to establish its own "un­equal treaties" with its Asian neighbors. Indeed, Japan emerged as a colonizer in its own right by 1895, when it took over Taiwan, and it later went on to build a vast empire in East Asia. In fact, Japan was the paramount imperial power in East Asia during the first half of the twentieth century.

Tropics of Savagery studies the dynamics of a hybrid imperialism that was dif‑

4 INTRODUCTION INTRODUCTION 5

ferent from, but also mimetic of, Western imperialism. It challenges the dyadic mod­els that underlie most postcolonial theories, proposing alternate models with which to understand Japan. By focusing on the case of Japan, I also hope to illuminate the imperial cultures of other mimetic, late-developing empires, and of postcolonial nations for which Japan often provided a template and an example.

Finally, I will note that Tropics of Savagery is in close conversation with two re­cent studies of Japanese imperial literature: Faye Yuan Kleeman's Under an Impe­rial Sun and Leo Ching's Becoming Japanese. Kleeman's work is an excellent study of literary writing by Taiwanese and Japanese writers during the colonial period.

⦁ Like Kleeman, I center my attention on literature, but unlike her, I focus on the theme of savagery in literary works and situate this theme in the context of broader social discourses and new disciplines of knowledge. In addition, I approach Japa­nese colonial literature from a comparative vantage point and seek to uncover both similarities and differences between it and the literature of Western empires. Rather than a study of literature, Tropics of Savagery is a work of cultural studies that ex­plores the links between Japanese literature and discursive disciplines that in­fluenced it within the global context of competitive imperialism.

Leo Ching's Becoming Japanese, a pioneering work on Japan's colonial culture, defines Japanese imperialism as a non-Western imperialism that both resembles and differs from Western imperialism. Building on Ching's analysis, I problema-tize the relationship between Japan and the West by showing how Japanese discourse is always produced in relation to, and in a certain sense, refracted through, West­ern imperialism. Indeed, I show that Japanese imperialism (and its colonial litera­ture) has a mimetic relationship to Western models. For example, Nakajima Atsushi modeled himself on Robert Louis Stevenson, although his literary works differ in fundamental ways from those of his model. In Tropics of Savagery, I also offer read­ings of texts that go beyond Taiwan (the focus of Ching's book) and extend to the South Seas. I am interested in the figure of the savage, a theme that is peripheral to Ching's study of Taiwanese intellectuals.

ALLEGORIES OF THE SELF

Japanese writers used savagery as a polyvalent trope to write about their encoun­ters with the different cultures and societies that Japan colonized. At the same time, they wielded this trope to tell their own stories as modern Japanese people. Ac­cordingly, the savage in the works I study is often highly romanticized, an alter ego who has preserved what modern Japan has lost, even the mirror of its authentic culture. This archaic, estranged "other" sometimes offers the writer access to hid­den desires or his or her primordial self. By writing about savages, modern writers sometimes critique their own society in a veiled way, confess their own ambiva­lences, explore their experiences of modernity, or otherwise construct an imagi‑

nary terrain to write about themselves. Just as savages became alter egos of the Japa­nese, the islands of the South Seas were reconfigured as a space with important meanings. Far from being peripheral to Japan, this region came to be seen as cen­tral to its culture and as the birthplace of the Japanese people.

As writings about cultural "otherness one can characterize the literary works I consider in this book as "ethnographic" in the broadest sense. Yet, as James Clifford argues in "On Ethnographic Allegory," every ethnography about another culture is also "an allegory" and "an extended metaphor" that contains "additional meanings" about the ethnographer's own society. Ethnographic writing "is allegorical at the level both of its contents (what it says about other cultures and their histories) and of its form (what is implied by its mode of textualization)'5 In the individual chap­ters of this book, I explore these additional meanings and metaphors in specific works of colonial literature, but here I will comment briefly on the significance of the allegorical form in Japanese colonial texts.

The term "allegory" covers a vast field with a wide array of types: "Within the boundaries of literature, we find a kind of sliding scale, ranging from the most ex­plicitly allegorical, consistent with being literature at all, to the most elusive, anti-explicit and anti-allegorical at the other."6 Some of the texts studied in this book, such as Nitobe's essay on MomotarO, are straightforward, "naïve" allegories. In his study of this folktale, Nitobe proposes to interpret Momotarö's conquest of the is­land of ogres as an allegory for Japan's colonial expansion into the South Seas. De­veloping a multitiered reading of MomotarO that leaves little to the reader's imag­ination, Nitobe causes this seemingly inconsequential folktale to stagger under the crushing weight of its allegorical significance. By contrast; SatO Haruo subtly weaves an allegory of contemporary history into the very fabric of his "Demon Bird," a work published in October 1923, based on a Taiwanese aboriginal legend and of­fering Sat O's contemporaries a lens through which to interpret their own recent his­tory. Manifestly a story set in colonial Taiwan, "Demon Bird" alludes to the mas­sacre of Koreans in the streets of Tokyo during the great Great Kanto Earthquake of September 1, 1923. In Nakajima's 1942 stories about the South Seas, the narra­tor views a reflection of his own hybridity and self-colonization in the people he meets in Micronesia. In this case, the reader hesitates between the literal sense of the text and its rich allegorical overtones.7

While literary historians since the romantic period disparaged allegory as an in­ferior form of symbolism, modern critics have done much to restore its critical po-tential.8 Allegory provides a way for writers to juxtapose different levels 'of mean­ing within a single text and to comment intertextually on other literary works. I argue that Japanese writers use allegory to bracket their colonial narratives in or­der to tell their own stories of cultural colonization and hybrid identity. Subtle lay-erings of meanings that neither fully cohere nor fully merge, these allegories man­ifest the ambivalences of Japanese imperial culture. In formal terms, these texts are

6 INTRODUCTION INTRODUCTION

often hybrid products that combine other discourses or comment on other texts in a pastiche.

In his '935 story "The Savage," Oshika Taku depicts a Japanese hero who "goes native" in the wilds of Taiwan, effectively inverting Japan's colonial mission of civ‑

ilizing savages. In his story "MomotarO," Akutagawa inverts Japan's most famous

folktale by turning its hero into a villain, parodying this model of imperialist youth held up as a paragon of virtue by Japan's leading scholar of colonial policy studies.

In "Demon Bird," SatO Haruo deconstructs the discourse of colonial ethnography

and pens a radical critique of Japan's empire. Although I emphasize the critical po­tential of allegory, I also point out the limits of allegorical critique. An allegorical

text arranges its different levels of meaning in an implicit hierarchy, one that mir­rors the power relationships of the colonial relationship itself, thereby raising ques­tions of an ethical and political nature.9 The form of the allegory that allows the Writer to speak to his or her readers about their own contemporary reality also serves to mask the colonial domination that victimizes the colonized.

If the works of Japanese colonial writers studied in this book are allegorical in form, perhaps my study of Japanese colonial literature is too. Northrop Frye notes,

"The instant that any critic permits himself to make a comment about a literary

text, he has begun to allegorize '1° If Frye is right, then my bOok is no less allegor­ical than the works that it studies. When I began to research Japanese constructions

of savagery during the colonial period, I thought I was investigating a somber but—

in the final analysis—closed chapter of human history. As I approached the end of my work on this book, I realized that I was also telling my own story as an Amer‑

ican in the first decade of the twenty-first century. Certainly, since the 1990s, the

discourse of empire has made a dramatic comeback, and with it, the rhetoric of civ­ilizing missions, dehumanizing tropes of savagery, and even the "savage wars of

peace" of which Kipling spoke. The only piece missing from this picture is any his­torical awareness that we are resurrecting the demons of the past. For the fact of the matter is that there have never been any savages, only tropes of savagery, al­though these tropes possess their own logic and generate fateful consequences.

When we refer to our present as the postcolonial era, we often assume that we have transcended the age of empire. Indeed, it is true that some empires (the Japa‑

nese, for example) have vanished from the face of the earth. In addition, the forms

of imperial power have changed markedly since the great decolonization movement of the 19505 and 196os. However, to speak of our present as "postcolonial" maybe—

at least in part—an expression of wishful thinking. Much as the "post" of postwar does not signify the arrival of a period of peace on earth, I would argue that the "post" of postcolonialism is perhaps misleading and, if nothing else, premature. In this study, I am guided by the conviction that we can deepen our awareness of his­tory and better understand our present times by examining the rhetoric of a de­funct empire.

VIOLENT SAVAGES AND CHILDREN OF PARADISE

But who were the savages of the Japanese empire, what were they called, and where did they live? And why make them the focus of a book on Japanese colonialism? Let me first repeat that I do not believe that there actually were savages. I am con­cerned in this book only with rhetorical constructions of savages and with their representations in literature and social dicourses. I employ "savage" as a word with­out a referent or indeed without any basis in reality—a word that functions, nev­ertheless, as a polyvalent signifier. As a nominal, the word "savages" refers to ficti­tious creatures imagined by the Japanese, whose imaginations were shaped by discourses on civilization, race, ideology, and literature. These savages populate the pages of Japan's colonial writings rather than the spaces of its empire. Though the savage was an imaginary creature, he or she acquired a virtual reality by the body of statements, made by writers who cited and accepted similar claims by their pred­ecessors, attesting to his or her existence. Nevertheless, this figure of discourse was an unstable entity that assumed different forms as the boundaries of empire changed and policies shifted.

How did writers during the colonial period refer to the savages about whom they wrote? In the first place, they often used nonspecific, abstract terms such as yaban (barbaric), genshi (primitive), or mikai (unenlightened) to denote the savage. Of these three terms, yaban is an ancient Sino-Japanese term for those who lie outside the pale of Sinic civilization and control, whereas both mikai and genshi take on their modern meanings in the translated discourse of enlightenment and evolu­tionary civilization that entered Japan during the Meiji period. If kaika meant the process of evolutionary civilization, then mikai (unenlightened)denoted the state that preceded civilization. Genshi, a term signifying "origin" or "beginning," took on the added meaning of the earliest and most primitive stage of society when the­ories of evolutionary civilization entered Japan.

One encounters these terms throughout the colonial archive, but the words that writers used most frequently to speak of savages in the Japanese empire tended to have a more limited and concrete specificity. Adopting older Qing terms to refer to Taiwanese aborigines, Japanese spoke of the seiban (raw savage) or juku ban (cooked savage), terms that classified the aborigines in accordance with their acculturation to Chinese norms. In addition, particularly during the first years of Japanese colo­nial rule, they divided the aborigines geographically into nanban and hokuban, re­spectively, savages in the south and north of Taiwan. During Japan's decades-long war of conquest in the Taiwan highlands, the aborigines were often categorized sim­ply as allies (mikata ban) or enemies (tekiban). The kanji for ban, common to all these terms, differed from its homonym in the term yaban, which signified bar­barian, and it served as the root for many related terms, whether spatial (banchi, aboriginal land in Taiwan), gendered (banfu, or savage women), abstract (banjO, or

8 INTRODUCTION INTRODUCTION 9

the condition of the savages), or moral (kyo ban, or evil, violent savages who resis­ted Japanese policies). After the visit of the crown prince to Taiwan in 1924, the government began to replace this deprecatory terminology with the neutral term takasagozoku, based on an ancient Japanese appellation for Taiwan. In the late 1930S, the government general of Taiwan launched a campaign to enforce usage of this term in all official discourse.

The Japanese also divided the Micronesians between the Chamorro (numerous in the Marianas Islands, often racially mixed descendants of Spanish) and the kanaka (the residents of the Caroline and Marshall Islands). The term kanaka was also a general, derogatory term used to describe all South Seas islanders and not specifi­cally Micronesians, a term which derived from the Polynesian word for "human being:' More commonly, Japanese referred to islanders collectively as dojin, domin, and, most frequently, nan'yö dojin, an omnibus term for the indigenous people of the South Seas. The term dojin (a native person) originally had the neutral mean­ing of a "resident of a particular locality' but it acquired the pejorative sense of "back­ward, primitive person" during the colonial period. The earliest use of dojin in this new sense was the so-called 1879 law on "former natives" of Hokkaido (the Ainu). The appellation "former native" (kyüdojin), like the term "new commoner" (shin-heimin), applied to former outcasts, belonged to a lexicon of terms that replaced the status system of the Tokugawa period, which was abolished at the start of the Meiji period. In fact, these terms tended to perpetuate prejudice against minority groups in Japan and to reinforce discriminatory practices against them.

Since savages inhabited only the imaginations of the Japanese, there is no way that one can locate them on the map: they belong to the nonplace of discursive pro­duction. Indeed, since the savage is the product of the discourse that names him, the same writer could use the same words to represent other spaces and to refer to other groups than the ones named above. Consequently, the actual usage of tropes for savagery exceeds the delimited spaces of the Japanese imperium and its South Seas colonies, and can be used for writing about the home islands, the Japanese past, or the human condition in general.

Nevertheless, this study will focus predominantly on representations of Taiwan aboriginal society and the South Seas, the richest sites for the production of this discourse, and will cover the years 1895-1945. After Taiwan was ceded to Japan under the Shimonoseki Treaty, which ended the Sino-Japanese War, the Japanese fought a long colonial war to extend their control Over the new territory. At first the aboriginal tribes in the mountainous interior of Taiwan were the allies of the Japanese in their struggle to crush Taiwan Chinese who had launched a guerilla war against Japanese colonial rule. After the guerillas were vanquished, the colo­nial authorities turned their attention to the mountainous interior of Taiwan and to its rich resources. Since the aboriginal population resisted the encroachments of settlers and soldiers, the regime launched a large-scale military offensive that cul­ minated in the genocidal five-year pacification campaigns waged by Governor-General Sakuma Samata (1909-14). While the colonial regime succeeded in ex­tending its control to most of the Taiwan highlands by 114, the resulting order was periodically broken by aboriginal rebellions that punctuated the subsequent decades of colonial rule. During the Musha Incident of 1930, Sedeq aborigines from six vil­lages attacked a school athletic event and murdered some 134 Japanese, precipitat­ing a major crisis in colonial rule over the aboriginal territories.

The Japanese practiced a distinctive form of colonization in the highlands of Taiwan—what one could refer to as "expropriation by dispossession:' They subju­gated the aboriginal lands primarily in order to exploit their resources, but had lit-tie use for the indigenous population. Deprived of their control of vast territories, the aborigines chafed under a colonial system that differed from the one imposed on the Han Chinese in the plains. Indeed, the Japanese drew a clear border that separated aboriginal territories from the rest of Taiwan, and ruled the aborigines under a special administrative system in which the police assumed the leading role; until the 1920S, travelers wishing to visit these lands needed special permits issued by the authorities. Such special treatment both reflected and legitimated the stereo­types of aborigines as fearsome savages, stereotypes that dominated Japanese pop­ular culture and literary works.

By contrast with the brutal subjugation campaigns and the periodic upheavals that characterized colonial rule of Taiwan, the acquisition of Micronesia, Japan's sec­ond tropical colony, was a bloodless affair. Ruled by Germany from the late nine­teenth century, Japan seized the islands of Micronesia after a short naval campaign during the First World War. Passing from German to Japanese control, the islands were later entrusted to Japan as a trust territory under a mandate of the League of Nations. Mandated territories were divided into three classifications, A, B, and C, determined by the level of their cultural and political development. Territories clas­sified as "A' were considered closest to independence, and, as in the case of Iraq (a British mandate that became formally independent in 1932), could receive sover­eignty after the required period of "tutelage". Mandates given a "C" designation—such as Japan's Pacific allotment—were judged to be furthest from sovereignty; they "were regarded as being on such a low level of political development as to be suit­able for treatment as integral parts of the mandatory Power's territory, almost, that is, as annexed domains:" Much like the Taiwanese aborigines, Micronesians were not seen as a labor force to be exploited. Unlike Taiwan, however, the Micronesian territory was exiguous and poor in resources. As a result, the process of expropri­ation was far less brutal than in the case of Taiwan. On islands such as Saipan, where the Japanese developed a plantation economy requiring an abundant labor force, the most important Japanese enterprises imported their workers from outside, es­pecially from Okinawa. During the three decades that Japanese ruled Micronesia, the Japanese settler population eventually came to outnumber the indigenous is‑

10 INTRODUCTION INTRODUCTION 11

landers, notably in the islands in the Mariana chain. While some Micronesians re­sisted Japanese rule andcolonial policies, the Japanese authorities did not confront widespread and violent opposition, as they did in Korea and Taiwan.

Reflecting these different modalities of conquest and imperial rule, colonial lit­erature Set in Taiwan and the South Seas often offers diametrically opposed images of indigenous peoples. In general, the dominant trope for representing the Taiwan aborigines during the colonial period was that of the violent irrational headhunter. Reflecting the ferocity of the Japanese conquest period, narratives set in Taiwan re­count cycles of rebellion and repression that punctuated the period of colonial rule. By contrast, writers about the South Seas often draw on figures of a "soft" primi­tivism to idealize the islanders as innocent, happy primitives and to depict their is­lands as a tropical paradise. 12 As Japanese immigrants came to outnumber South Seas islanders, Japanese writers often evinced "imperial nostalgia" for the vanish­ing Micronesians, victims of colonial modernity and of rapid demographic changes, a theme common in ethnographic writings about both Taiwan and Micronesia.

This diptych of contrasting figures of the savage—noble savage and violent head-hunter—constitutes the light and shadow of colonial literature. Scholars of West­ern colonialism have often noted the ambivalence of the trope of the savage in Western discourse. From the time of the "discovery" of America by Europeans, the indigenous native has been portrayed as the frightening and violent cannibal but also as the innocent and happy child, the epitome of all that is pure and untainted.' It is no historical accident that a similar dichotomy can be found in the works of Japanese literature. After all, Japan was only one in a series of colonial powers that ruled Taiwan and Micronesia, and it borrowed liberally from its predecessors even as it strove to distinguish its rule from theirs. In addition, Japanese proponents of empire were quick to adapt figures of the savage along with the entire panoply of colonial discourses and tropes that had accumulated during several centuries of Western exploration and colonization of non-Western parts of the world. 14

Why make savages central to study of Japanese colonial culture? In general, his‑

⦁ torians stress that the Japanese were culturally close to people that they colonized. One scholar writes: "[Japan's] most important territories, Korea and Taiwan, were well populated lands, whose inhabitants were racially akin to their Japanese rulers with whom they shared a common cultural heritage. This sense of cultural affinity with its subject peoples made Japan unique among the colonial powers of modern times.

"5 Because of these purported cultural and racial similarities with the colo­nized, this author points out, the Japanese were less likely to feel an exotic fascina­tion with their colonies than Western colonizers did. While I will not attempt to refute this observation, I will note that the Japanese empire included societies with­out writing or complex state organization that were culturally remote from the home islands. The aboriginal population of Taiwan and the indigenous population of Mi­cronesia may have made up a tiny fraction of the population of the Japanese em­ pire, but they played a major role in the imagination of Japanese writers and loomed surprisingly large in colonial literature. Japanese writers devoted a disproportion­ately large number of fictional works to the Taiwanese aborigines, a mere 2 percent of the island's total population, while they wrote relatively few works about the nu­merically preponderant Taiwan Chinese. Indeed, Japanese writers wrote no fewer than fifty works that were inspired by the aforementioned 1930 Musha Incident, in which Ataiyal aborigines rebelled against the Japanese colonial regime.'6 Many Japa­nese writers and artists traveled to Micronesia, after it became a Japanese colony in 1914, in search of inspiration, producing an abundant and varied body of work.'7 Beyond their quantitative importance, such literary works are qualitatively signifi­cant because they intersect in manifold and complex ways with the different strands of Japan's colonial discourse and with the different phases of Japan's empire.

The Taiwanese aborigines were the targets of Japan's earliest colonial endeavor, the 1874 Taiwan Expedition, undertaken to chastise savages who had murdered the crew of a ship from the RyUkyU kingdom. Later, in encyclopedias of geography, ethics textbooks, travel books, and colonial policy studies, they were often depicted as the "headhunters" of the Japanese empire. Indeed, the subjugation of "head­hunters" was adduced by Nitobe InazO, Japan's first professor of colonial policy stud­ies, who started his career as a colonial official in Taiwan, as the paradigmatic case of Japan's civilizing mission. These same "headhunters" captured the imagination of early Japanese ethnographers who went to Taiwan in 1895 to survey the aborig­ines and to establish a system of classification of their major groups. Pioneers of colonial ethnography wrote primarily for like-minded experts, but they also pro­foundly influenced writers like Satö Haruo and Nakajima Atsushi, who based their stories on passages from ethnographic studies. Finally, whereas Meiji-era Japanese celebrated the subjugation of savages as an achievement of a modern and civilized Japan, their descendants in the twentieth century fled modern civilization and re­settled in the Taiwan highlands to recover their inner savage.

In Japan, the earliest depictions of primitive South Seas islanders appeared in political novels from the 188os, written by authors who had never traveled to the places they wrote about. These South Seas novels, which sparked great interest in the Pacific region among Meiji youth, often feature young officers of the Japanese navy who explore the South Seas and claim uncharted territory for Japan.'8 In Yano RyUkei's Ukishiro monogatari (Tale of the Floating Castle), published in 1890, a group of Japanese adventurers take over a British warship, set off to explore the South Seas, and eventually help a small island nation in its independence struggle against the Dutch. In Oshikawa ShunrO's Kaiteigunkan (Battleship at the Bottom of the Sea), written in 1900, a Japanese naval officer establishes a secret base on a desert island in the South Seas where he develops new weapons and fights Western empires to win respect for his nation.19

By the turn of the twentieth century, these fantasies started to turn into reality

12 INTRODUCTION INTRODUCTION 13

as Japanese established strong commercial ties in the South Seas and eventually ac­quired their first South Seas colony. Nitobe InazO saw Japanese expansion to. the south as a manifestation of Japan's national character. He also interpreted the pop­ular folk hero MomotarO as a metaphor for the nation's expansion and as a model for prospective colonizers to emulate, much as Robinson Crusoe had been an in­spiration to generations of English schoolboys. In fiction from the first decade of the twentieth century, young Japanese frequently travel to the South Seas in search of adventure or riches. With the vast expansion of Japanese media and economic ties in the South Seas in the 192os, primitive South Seas islanders begin to enter

Japan's burgeoning popular culture, giving rise to what Kawamura Minato refers to as a "popular [taishü] orientalism" Finally, South Seas islanders occupied the bot‑

tom rung of the racial and environmental hierarchies that the Japanese constructed

to legitimate their dominant position in an autarchic empire, particularly at the time of the Greater East Asian Co-Prosperity Sphere. To use Michel-Rolph Trouillot's

term, they filled the "savage slot" in the imperial hierarchy of power and knowl­edge as the empire came to encompass broad swathes of Southeast Asia and the Pacific Islands.20

JAPAN IN POSTCOLONIAL STUDIES

This book sets out to comprehend Japanese colonial culture by analyzing the tropes of savagery in literary works and exploring links between them and wider social discourses. An ancillary aim of the book is to place the Japanese empire within the comparative context of global imperial discourses and to initiate a dialogue with current postcolonial theory, which is so deeply informed by the study of European empire. Gyan Prakash offers a succinct statement of the aims of postcolonial the­ory in his essay "Postcolonial Criticism and Indian Historiography": "One of the distinct effects of the recent emergence of postcolonial criticism has been, to force a radical rethinking and reformulation of forms of knowledge authored and au­thorized by colonialism and Western domination.... Recent post-colonial criti­cism seeks to undo the Euro-centrism produced by the institution of the West's tra­jectory, its appropriation of the other as history, If the condition of the postcolonial world is the legacy of colonialism and of Western domination, then Prakash be­lieves that the central task of postcolonial critique is to undo this legacy and to de­construct "Eurocentrism'

Prakash thus uses the term "postcolonial" to refer to non-Western parts of the world (the "third world") once colonized by the West, a geographical notion that underlies other major postcolonial theoretical statements. In his seminal work on French and British domination of the Middle East, Edward Said calls Orientalism "a Western style of dominating, restructuring, and having authority over the Ori-ent."22 Homi Bhabha and Gayatri Spivak honed their theories to explain the British colonization of India, although Spivak frankly cautions her reader against treating India as emblematic of all colonial relationships: "The Indian case cannot be taken as representative of all countries, nations, cultures, and the like that may be invoked as the Other of Europe as Self."23 Notwithstanding this cautionary note, postinde-pendence India has a centrality in the literature of postcolonial studies not inferior to India's erstwhile position as the jewel in the British crown.24

Even when they go beyond this single case, however, postcolonial theories are generally based on a limited empirical perspective, cover a narrow geographical range, and tend to generalize excessively. As a consequence, their theoretical for­mulations tend to neglect the historical and linguistic features of colonial empires outside the Anglo-American framework.25 Nevertheless, forms of colonial domi­nation varied greatly depending upon the location of colony (Africa, Asia, Latin America) and the time period (sixteenth, eighteenth, or twentieth century) con­sidered. In addition, there are national differences in the style of imperial domina­tion: England's way of being imperial was not the same as that of France, Portugal, the United States, or Japan. If the British empire looms especially large in post-colonial studies, Japan is perhaps the most neglected of the world's major empires. 26 Japan controlled the most extensive territorial empire in East Asia, yet it is "un­marked as a colonizer in Euro-American, but not in East Asian eyes. 1127

That Japan, the only non-Western colonial power in the modern period, hardly figures in the annals of postcolonialism reflects the traditional disciplinary division between East Asian studies and studies of Western societies and mirrors a division of labor between their respective practitioners. If postcolonial scholars research the global aftermath of Western empires, East Asian area specialists tend to restrict themselves to their country or region of specialization. The development of East Asian studies in the United States, which came about as a, result of the need during the Cold War for specialists with cultural and language training in East Asia, has led to the segregation of this field in area studies programs in American'universi-ties. This segregation of East Asian specialists in area studies has had the perverse effect of cutting off East Asian specialists from broader trends of intellectual life, thereby fostering intellectual provincialism .21 By contrast, postcolonial theorists have tended to treat imperialism as a "Western" rather than a global problem, not­withstanding their interest in recovering the agency of the non-West. In fact, by ne­glecting to include the only non-Western colonial experience in the general 'critique of imperialism, they have forfeited a magnificent opportunity to broaden the scope Of postcolonial studies. Instead, they have tended to reproduce the Eurocentrism that they claim to combat.29

Even though postcolonial theorists have ignored Japan, students seeking to un­derstand Japan's empire can often apply postcolonial paradigms fruitfully to an analysis of its culture of empire. In other cases, however, these paradigms do not offer much insight into Japanese colonial relationships. Without proposing an over‑

14 INTRODUCTION INTRODUCTION 15

arching theory to explain Japan's imperial culture as a whole, I will highlight some distinctive traits of Japan's imperial culture and suggest lines of approach that in­form my own studies of colonial period literature. Suzanne Zantop's study of Ger­man colonial culture, which offers some interesting parallels with Japan, is a prece­dent for the type of empirically grounded and comparative study that can widen our cultural understanding of modern empires.3°

COLONIAL MIMICRY AND MIMETIC IMPERIALISM

Homi Bhabha's notion of "colonial mimicry" has proven to be extremely fruitful in analyses of colonial relationships. Colonial mimicry, an "ironic compromise:' is "the desire for a reformed, recognizable other as a subject of difference that is almost the same but not quite."" Colonialism is haunted by ambivalence toward this colonial mimicry: on the one hand, the colonizer demands that the colonized resemble him­self or herself through a process of "narcissistic identification:' but on the other, he or she also disavows this resemblance and even regards it as a "menace:' Thus the colonizer both requires and rejects colonial mimics; they are, impossibly, required to be "almost the same but not quite."

At first glance, this notion of a "desire for a reformed, recognizable other" seems especially pertinent to the assimilation policy (dOka) that Japan adopted toward the colonized. For the colonized subjects under Japan's control were placed in a classic double bind. On the one hand, they were encouraged to become "like the Japanese" by speaking Japanese and adopting Japanese customs; on the other, contradicting this rhetoric of assimilation, they were never accorded the economic and political rights granted to Japanese subjects. Like children exposed to contradictory mes­sages from adults, the colonized confronted by the insoluble double bind of the col­onizer responded with resistance and mimicry, or even with subversion through mimicry As Bhabha notes, mimicry is not only "resemblance" but also "menace" that the colonizer must "disavow: since mimicry "destroys narcissistic authority through the repetitious slippage of difference and desire. '32

Yet colonial mimicry can only deconstruct the authority of a colonial power that posits itself as original in the first place. But should we apply such a conceptual framework of original and copy to the case of Japan? And if so, where do we draw the line between imperial "original" and colonial "copy"? Before Japan insisted that the people it colonized mimic the Japanese, did it not imitate the West, including late nineteenth-century Western imperialism? It is well known that the architects of Japan's modernization studied Western societies and adapted Western institu­tions and techniques to remake Japan as a modern nation-state. At the same time, they made a careful study of Western empires and colonial systems of rule. In the nineteenth century, the control of colonial territories was as important a measure of national strength as industrial development and military readiness. Determined to rescind the "unequal treaties" that formalized its subordination to the West, Japan early on resolved to acquire and rule its own overseas territories. Peter Duus writes that "imperialism, like so many other aspects of Meiji development, was an act of mimesis) 1131 adding that "what ultimately enabled the Japanese to mimic Western imperialism was their simultaneous mimesis of other aspects of Western 'wealth and power' . . . Although the Meiji Japanese mimicked the imperialist culture sys­tem that developed in the West, Meiji imperialism was the imperialism of a back­ward or follower country. It was characterized by a psychology of inferiority vis-à-vis the West, a desire to catch up with the more advanced economies, limited foreign contacts, dependency on the import of capital goods, a lack of political leverage over the advanced powers, and a high degree of state involvement in economic devel-opment."34

Interestingly, Japan mimicked the tactics Western powers had earlier used to de­prive Japan of its sovereignty. Now, however, the Japanese were the active agents who mobilized these same tactics against their neighbors rather than the passive recipients of Western intrusions. In this regard, Japan's first imperialist move to­ward Korea is exemplary. In 1876, the Meiji government dispatched warships to the Korean coast and forced the ChOson court to sign the unequal Kanghwa Treaty, in a replay of Commodore Perry's gunboat diplomacy of 1853. The Japanese did not simply mimic the actions of the Western imperialists; they also imitated West­ern justifications of empire to legitimate their own expansionism. In the early years of the twentieth century, the liberal politician Takekoshi Yosaburö appealed for Japan's admission into the club of Western empires in a rhetoric clearly fashioned after Western models of the civilizing mission and the "white man's burden": "West­ern nations have long believed that on their shoulders alone rested the responsi­bility of colonizing the yet unopened portions of the globe and extending to the in­habitants the benefits of civilization; but now we Japanese, rising from the ocean in the extreme Orient, wish as a nation to take part in this great and glorious work. Some people, however, are inclined to question whether we possess the ability req­uisite to this task ."35

The mimetic character of Japanese imperialism is readily apparent in Japan's first foray into empire building. In 1874, the Japanese government invaded Taiwan to punish aborigines of the Botan tribe who had massacred fifty-four Ryiikyüan sailors a few years earlier.36 Japanese commercial newspapers vied with one another to cover this military expedition and depicted its major episodes in lavishly colored wood­block prints. In their media reports, reporters argued that Japan must place the abo­rigines under firm Japanese control, in part to secure Western recognition and to "underscore Japan's identification as a civilized nation. 1137 In temporal terms, the i 874 expedition coincided with the Meiji government's imposition of "civilization" on recalcitrant Japanese peasants and samurai. Japanese media reports often described the aborigines as cave dwellers and "cannibals," images that point to mimicry of Eu‑

INTRODUCTION INTRODUCTION 17

ropean discourse on savagery, since the cannibal had been the central European marker of lack of civilization from as early as the sixteenth century. However, they also deployed explicit visual analogies with Perry's expedition to Japan to explain the hierarchical relationship between the aborigines and the Japanese, associating Japanese dominance over the aborigines with the dominance of the government over the Japanese commoners and with the Western dichotomy of civilization and sav­agery. In addition to displacing Japan's subordination to the West onto the aborig­ines, such representations "of the aborigines' subordination foreclosed the possibility of solidarity between the Japanese and the aborigines in favor of asserting a hierar­chical relationship between them patterned in part after Western imperialism. "38

As the above examples suggest, one needs to supplement Bhabha's idea of colo­nial mimicry with the subsidiary notion of imperial mimicry—the mimicry by one imperial power of another—in order to understand Japan's imperial culture. Japan's imperial mimicry structured the public discourses in which the Japanese justified their empire, but it also inflected their intimate feelings as colonial subjects. It assumed a multitude of forms—historical, political, ideological, psychological, artistic—but I will be concerned primarily with three varieties of mimicry. In the first place, I will consider psychological mimicry, which often was accompanied by a sense of cultural colonization. Natsume SOseki provides an excellent example of this mimicry in his 1.911 lecture 'The Civilization of Modern-day Japan." In this lec­ture, he contrasts the "internally motivated" development of European countries with Japan's "externally imposed" development. If the former "develops naturally from within, as a flower opens: the latter has "to assume a certain form as the result of pressure applied from the outside" Japan's modernization is a "forced march" in which Japan had to "compress ioo years of development that went into the enlightenment of the West into a span of 10 years:' The end result of this external imposition and temporal compression is that the Japanese "feel that we are dressing up in borrowed clothing, putting up a false front:' To be sure, Japan's rapid modernization was a vol­untary "simulation" rather than the result of forced colonization. Yet SOseki treats this process as a form of self-colonization that alienates the Japanese people from their traditions and subordinates them to Western culture: 'All we can do is me­chanically memorize Western manners—manners which, on us, look ridiculous" As "ridiculous" as they may look, given the realities of geopolitics the Japanese have no choice but to "imitate the West": "Just look at how we socialize with Westerners: always according to their rules, never ours. Why, then, do we not just stop socializ­ing with them? Sadly enough we have no choice in the matter. . . . When two un­equal partners socialize, they do so according to the customs of the stronger:'39

Whereas SOseki stressed his sense of cultural colonization when confronted with European civilization, he readily adopted a superior position toward other Asians when he visited Japanese outposts, in Korea and Manchuria in 1908. Standing on the deck of a ship approaching Dairen (Dalian), he describes the crowd of coolies gathered along the shore as "buzzing and swarming like angry wasps" and producing in him an "impression of dirt:' The same writer who had felt the piercing gaze of European civilization on himself when he wrote about Japan's modern civilization in a global context appropriated this same gaze to look at other Asians from a po­sition of assumed superiority.40 This "complex" in which feelings .of superiority and inferiority form such an unstable, contradictory mix, was not a psychological pe­culiarity of SOseki; it can be found, in a thousand guises, in the works of many other writers. Bhabha sets his colonial mimicry in opposition to a narcissistic authority and mastery. The constellation of feelings expressed by SOseki stands in marked con­trast to the sense of ontological priority and psychological groundedness that Bhabha finds in Western colonial writings. He more closely resembles the colonized "mimic men" depicted in Naipaul's famous book of the same title.

I will also be concerned with literary imitation, the mimesis of specific works or writers or more generally of tropes of Western literature. If modern Japanese liter­ature as a whole is mimetic with respect to the West '41 this fact would apply a for­tiori to its colonial literature. As I have already noted, Japanese colonial writers had no choice but to copy Western models when they set out to write about savages. Besides borrowing tropes from Western writers, Japanese writers actually modeled themselves on Western writers and on Western artists. For example, Nakajima At-sushi, a key figure in late imperial literature, vicariously identifies with Robert Louis Stevenson in his 1942 novel Light, Wind, and Dreams. This novel, a "fictional auto­biography," is based on Robert Louis Stevenson's writings on his final years in Samoa. Shortly after entrusting the manuscript with an editor, Nakajima followed in the footsteps of Stevenson and set off on his own South Seas journey to work as the ed­itor of Japanese-language textbooks in Palau. For Nakajima, Stevenson provided a model of the romantic artist fleeing modern civilization—a model that he appro­priated to construct his own imperial subjectivity and define his standpoint toward the South Seas.

In the third place, colonial administrators and propagandists appropriated Western discourses to justify their rule of colonized Asian and Pacific peoples while scholars imported scientific disciplines which they used to construct the colonized as an object of knowledge: I will treat both as forms of discursive mimicry. Just as Western nations had earlier deployed international law to impose a semicolonial status on Japan, Japanese imperialists used the same instrument to legalize their domination of their Asian neighbors.42 International law assumed a hierarchy of states that were divided into the civilized, those who were full participants in the juridical order, and the less civilized, those who would be dealt with by means of unequal treaties. Japan renegotiated its unequal treaties with Western countries and replaced them with more equitable arrangements, but it also had recourse to the same law to justify colonization in Asia.43

To start with, foreign terms and concepts—international law, race, civilization,

18 INTRODUCTION INTRODUCTION

and so on—needed to be translated into Japanese. But Japan's acculturation of mod­ern discursive formations went beyond linguistic appropriation: it also involved us­ing these terms as instruments of power to redefine its position within a global or­der. Yet discursive mimicry offered Japan more than merely an idiom with which it could justify its imperialism. Japanese intellectuals also appropriated authorita­tive disciplines such as colonial policy studies and anthropology, veritable technol­ogies of colonial knowledge, and later employed them in the laboratory of the Japa­nese empire. When they constructed the colonized as objects of knowledge, the Japanese proved that they were knowledgably building their empire and that they controlled their colonies through knowledge.

In arguing that the Japanese empire was mimetic, I do not wish to underestimate Japanese agency or to depreciate Japanese responsibility for its colonization. Even a cursory examination of Japanese history shows that Japanese imperialism also had indigenous roots in the nation's premodern past. Under Hideyoshi, Japan invaded Korea in 1592 and 1597, centuries before it started to copy the West. Motoori Non-naga, the National Studies scholar, already placed Japan at the top of a global hier­archy in the eighteenth century.44 Nevertheless, Hideyoshi's empire came to an end in a few years and Motoori's speculations were written at a time when the Japanese were prohibited from leaving the country. Without the collision with Western im­perialism and the mimicry of its forms, it is hard to conceive that Japan would have followed the expansionist and imperialist course that it did in the modern period.

- THE AMBIVALENCES OF COLORED IMPERIALISM

At a crucial turning point in his argument in "Of Mimicry and Men," Bhabha iden­tifies the narcissistic authority of the colonizer as "western" and "white." The colo­nial mimic is always betrayed by a visible, racial difference or, to quote Bhabha, he or she must be "almost the same but not white: the visibility of mimicry is always produced at the site of the interdiction. Bhabha's verbal slippage from "quite" to "white" may seem natural for a Western imperial order in which whiteness and em­pire are synonymous terms. By contrast, the position of Japan as empire is closer to that of the colonial mimic who is "almost the same but not white" and whose mim­icry is betrayed by a visible difference. Indeed, Japan is the first example, in the mod­ern period, of "colored imperialism. "461 used the term "colored" not in a biologi­cal sense to mean a darker shade of skin but rather to refer to a discursive system in which the West stood for "white, civilized, and colonizer" and the East was "col­ored, barbaric, and colonized." Prior to their encounter with the West, the Japanese did not think of themselves as "colored." Indeed, one prominent intellectual objected to the notion that the Japanese could be classified as a "colored" or even an Asian, nation .41 Nevertheless, by the late nineteenth century, the Japanese had little choice but to accept the position assigned to Japan in this Western discursive system.

Japanese learned the Western codes of racial color by their direct experience of the West. Natsume SOseki discovered his "colored" man status as a student in fin de siècle London: "In the street, I saw a short and shabby looking fellow coming to­ward me: on second thought, it was my own reflection in a mirror. Since I arrived here, I realized for the first time that we are yellow. 48 Japanese-born Takao Ozawa learned that he was "colored" in the course of his lengthy battles with the Ameri­can legal system to obtain U.S. citizenship. In 1914, Ozawa's application for U.S. cit­izenship was rejected on the grounds that he was not white. He later appealed to the Supreme Court, arguing that he was indeed a "white person." In Ozawa v. United States (1922), the Supreme Court rejected his application, but it recognized that the discursive boundaries of race are not based only on skin pigmentation. As a Supreme Court justice put it, "Manifestly, the test [of race] afforded by the mere color of the skin of each individual is impracticable as that differs greatly among persons of the same race, even among Anglo-Saxons, ranging by imperceptible gra­dations from the fair blond to the swarthy brunette, the latter being darker than many of the lighter-hued persons of the brown or yellow races. "49

As "colored" imperialists, the Japanese harbored ambivalent feelings toward the Western powers: they admired Western wealth and civilization but they resented Western racism. After defeating "white" Russia in the Russo-Japanese War of 1904­5, Japan was elevated to the position of one of the major world powers, but it re­mained a minority among a concert of white and Christian nations. Indeed, Japan's military victory over Russia exacerbated Yellow Peril fears that a strong Japan could lead all of Asia in repulsing Western power.5° Within a year of the Russo-Japanese War, the city of San Francisco issued an ordinance that prohibited Japanese immi­grants from attending public schools, and, in 1907, the United States forced Japan to sign the so-called Gentleman's Agreement, effectively curtailing the immigra­tion of Japanese laborers to the United States51 Allied with Britain during the First World War, Japan participated on the side of the war victors in the Versailles Peace conference in 1919, but the victorious Western nations rejected Japan's proposal that a racial-equality clause be inserted into the charter of the newly established League of Nations. In 1924, the United States adopted the Immigration Act, which barred Japanese immigration to the United States, prompting one middle-aged Japa­nese man to commit ritual seppuku in front of the American embassy in Tokyo with a letter of protest addressed to the U.S. president lodged in his chest. 12 The passage of the 1924 Immigration Act and the rejection of the racial-equality clause were taken as crowning insults that strengthened Japan's perception of itself as the vic­tim of Western imperialism even as the nation was emerging as the dominant im­perial power in East Asia.

Yokomitsu Riichi's novel Shanghai captures the complicated psychology of Japan's "colored imperialism" in the i 920s, in which a feeling of superiority toward Asians was conjoined with a sense of inferiority to Westerners. Throughout the

20 INTRODUCTION INTRODUCTION 21

novel, KOya, a Japanese businessman who speaks French and German fluently and prides himself on the fact that women say he is "just like a foreigner:' fruitlessly pursues the club dancer Miyako, who prefers Western lovers and rejects her com­patriot "on account of the pigment of his skin." The pair enter a municipal park where the "only legs not permitted to enter. . . were those of the Chinese:' While KOya re­sents the fact that he is "scorned" by the "sturdy long-legged foreigners" strolling through the park, he takes pride in the fact that at least he is not Chinese; indeed, he prefers not to think about the Chinese barred from a park in their own city. As a citizen of a world power, KOya enjoys the exorbitant privileges that extraterritorial status accorded to all members of the colonial powers that dominated Shanghai. His position was slightly lower on the totem pole than that of Westerners, but it was far superior to that of the Chinese. Indeed, the position of the Japanese characters in the novel is articulated in terms of the two major ideological formations that bat­tle in it: colonialism and pan-Asianism.53 On the one hand, Japan was a foreign oc­cupier of Shanghai aligned with the United States, England, France, and Germany—its racial others. On the other hand, it sought to align itself with the colonized peoples of Asia by reason of racial similarities.54

Japan's position as a "colored" empire shaped the ambivalent, fractured sense of identity of the Japanese colonizer. 15 Even as it actively colonized its neighbors, Japa­nese imperialists presented themselves as "friends" and "brothers" that would lib­erate them from the threat of Western domination. Viewing Japan as a victim that had "suffered at the hands of the white race: they stressed their identification with the people they colonized. Nevertheless, Japan was not only "victim of the West" but also the victimizer of Asia. Just as Japanese diplomats were tabling a racial-equality proposal at the conference of Versailles, the Japanese army was brutally re­pressing March First demonstrations in Seoul, an ironic juxtaposition of events that did not escape the notice of contemporary observers. 16 Colonial ambivalences notwithstanding, the Japanese probably treated their colonized no better than any other empire did, as indeed the victims of Japan's rule were fully aware: "In some cases, this sense of victimization will cause the ruler to mitigate some of the harsh­ness of its domination, but in general this same sense may exacerbate its sadism to­ward the dominated, particularly when its sense of identity is in violent flux" Japan's "colored imperialism" was a "parasite that grew in the spaces between the binaries of 'Europe' and 'Asia: 'victimizer' and 'victim: 'colored' and empire' 1157

AN INTERSTITIAL IMPERIALISM:

THE TRIANGLE OF JAPANESE COLONIAL DISCOURSE

To summarize, Japanese imperialism was a form of imperialism that explicitly mim­icked the Western variety even as it was also a "colored imperialism" that differen­tiated itself from the West and promoted close identification between colonizer and

colonized. In noting that the Japanese were "mimics:' I do not mean to suggest that they were unique in this respect. In the late nineteenth century, rising empires such as Germany, the United States, and Russia clearly imitated the older imperialisms of Britain and France, who in turn followed the precedents of Spain or even ancient Greece and Rome.58 Nor do I wish to imply that their imperialism was nothing but a "copy' In fact, mimicry is never a simple matter ofcopying.59 Even the perfect mimic never produces a perfect "copy" of the original he or she seeks to imitate, if only because a copy differs ontologically from the original it reproduces. Nor is mim­icry a passive matter of duplicating a preexisting model, as a mirror gives a reflected image of the object standing before it. To the contrary, mimicry is an active, cre­ative process that requires resourcefulness and implies an ironic remove on the part of the actor producing a replica. Precisely because Japan was creative in its mim­icry, its mimesis of the West produced in the final analysis less a merging with its model than a distinctive type of imperialism that set Japan apart from the coun­tries it imitated. This distinctiveness resulted from the historical context in which Japanese imperialism evolved and from the psychology of colored imperialism, in which Japanese colonizers had a kind of colonized consciousness toward the West.

In general, Japanese imperial culture differed from Western paradigms of em­pire in two key respects: it had a triangular structure and it tended to promote close identification between the Japanese and the people they colonized. Typically, post-colonial critics treat the colonial relationship as a dyad composed of colonizer and colonized. However, in the case of Japanese imperialism the West was always the (implicit) third party. Due to this triangular structure, Japanese colonial discourse was always produced in relation to and, indeed, refracted through Western colo­nialism. Precisely because the West was an implicit reference point for the Japa­nese empire, writers during the Japanese colonial period frequently claimed that the Japanese resembled or were closer to the people they colonized than were other imperialists to their colonized. I refer to this second discourse as the "rhetoric of sameness:' studying its function within Japanese imperial discourse.

The position of Japan between its colonies and the West was not stable and in­variable throughout the entire colonial period but evolved in accordance with Japan's changing position in the global order. At times, Western empires were held up as models to emulate, but at others, they were negative examples to be carefully avoided. Regardless of whether it was copied or eschewed, the West was an implicit refer­ence point by which Japan charted its own course as empire and measured its own progress. After Japan acquired its first colonies, it justified its new position as colo­nial master by invoking the civilizing mission that Western powers had earlier used to impose unequal treaties on it, and legitimated its rule over others by appealing to the superiority of the Japanese race. By contrast, in the 1940s, the Japanese gov­ernment legitimated its expansion by arguing that Japan was liberating its fellow Asians from the evils of Western imperialism in the name of racial brotherhood.

22 INTRODUCTION INTRODUCTION 23

While these two historical moments are diametrically opposed to each other in con­tent, they are structurally quite similar.

The three-sided structure of Japanese imperial discourse can also be seen in Japa­nese appropriations of the notions of race and civilization in the late nineteenth

century—notions that Japanese imperialists used to articulate a rationale for Japan's rise, as an empire and to legitimate its colonial rule over others. Yet the race that ruled most of the world was the white race and the term civilization referred to Western civilization. How were these terms translated into Japanese culture and how did Japan fit itself into the hierarchies they underwrote? How did Japanese impe­rialists use these terms to legitimate Japan's colonial rule over others? In the Meiji period, Japanese intellectuals did not construct their own hierarchies of empire ex nihilo before setting out to colonize their neighbors. Rather, they revised these Eu-rocentric hierarchies, in which Japan initially occupied an inferior position and was subordinated to the West, in order to make room for a non-Western colonial power. Indeed, Japanese emergence as an empire was premised on a prior form of self-colonization. The Japanese people were the first objects of the government's colo­nizing mission, even before they began colonizing foreign peoples.

In Japanese adoption of Western notions of race and racial hierarchy in the late nineteenth century, early proponents of the science of racial improvement (eugen­ics) used two terms, minzoku and jinshu, to signify the mix of biological and cul­tural characteristics that meant race, although jinshu had a more scientific conno­tation and minzoku was the more popular term .60 From the early Meiji period, these writers played an active role in disseminating to the general public notions of race along with "scientific" methods of classifying different races based on physical ap­pearance, bodily measurements, blood types, and common languages and cultural traditions. Such notions were closely intertwined with ideas of racial hierarchy, the superiority of the white race and the inferiority of colored races.

Fukuzawa Yukichi, an important figure of the Japanese enlightenment, intro­duced his readers to the world's five continents in his 1869 book ShOchü bankoku ichiran (Catalog of the World's Countries in the Palm of One's Hand)' and in Sekai kunizukushi (World Geography; also 1869), which was adopted as a school text­book in the early Meiji period. He also described the five color-coded groupings that inhabited these different continents: Caucasians (Europeans) are white, Asians "slightly yellow:' Africans black, the people of the Pacific Islands brown, and the in­habitants of "the mountains of America" red .6' Besides identifying these racial groups (Fukuzawa uses the term jinshu) and estimating their population, he char­acterizes each race in terms of its cultural and moral level within a clear hierarchy, following the categories of Western racial hierarchy. The whites "have enlightened spirits and have attained the highest stage of civilization"; the yellow race "is strong in the face of adversity, diligent in study, but its abilities are narrow and it is slow to progress"; the red race is "aggressive and loves to fight"; the blacks are "lazy by disposition and know nothing of progress and enlightenment"; and the "brown race is "violent and prone to revenge "62

While Fukuzawa briefly discusses Oceania in his geography book, he does not use the more expansive term nanyO, which later writers employ to refer to the South Seas or to identify the race of its inhabitants. 61 Shiga Shigetaka, who spent ten months touring the South Pacific aboard a Japanese navy training vessel in 1886, 'prided himself on being the first to introduce the Japanese public to the nan)'O, a geocultural region like the East (tOyO) and the West (seiyo), the two civilized worlds known to Japan. In Nan'yö jiji (Conditions of the South Seas; 1887), he provides a :geographical description of South Seas island-nations under threat from Western imperialism as well as a racial classification of their inhabitants as Malays 64

Warn­ing his readers that the battle for the "survival of the fittest" was taking place in the South Seas he contends that Japan must modernize if it hopes to escape coloniza­tion by a more advanced or superior country. Drawing on neo Darwinist theo ries, Shiga argues that "the white race is superior to the yellow, the black and the brown:' and that these inferior races are "threatened with extinction when they be­gin commerce with the whites "65 By introducing the Japanese to recent events in the South Seas and to the notions of social Darwinism, Shiga hoped to sound a warn­ing alarm to the Japanese about the perils they faced: "We must devise a strategy to protect our future as a race and to defend ourselves in this competition with the white race In a world in which strong nations dominated others by force weak ,nations could only survive by cooperating with one another. Shiga goes on to ar­gue that only an alliance between China and Japan could check the irresistible ad­vance of the white race throughout the world Shiga treats the Japanese as an "in­ferior" race when he compares them with the-white race, but he later describes them as a "superior" race in comparison with the Ainu of Hokkaido. Rather than stand­ing at the bottom of a racial hierarchy, the Japanese (and the yellow race) are placed in a middle position.66

During the 188os and 189os, the Japanese weighed the pros and cons of the gov­ernment plan to abolish separate residential districts for foreigners in cities such as Yokohama and Kobe and to permit mixed residential districts. 67 This proposal gen­erated a significant controversy that offers a window onto Japanese views on racial hierarchy and on Japan's racial identity in the late nineteenth century. Inoue Tetsu-jirO, professor of philosophy at Tokyo Imperial University, vigorously opposed mixed residence on the grounds that "the Japanese are inferior to most Westerners intel­lectually, financially, physically, and in every other respect [so] they are inevitably doomed to be defeated in any competition [with the superior Caucasians] ." To prove his point, he cited the precedents of the South Seas islands of Hawaii, Tonga, and Fiji, whose "inferior" populations began to decline when European settlers arrived, with the native population of Hawaii declining by over 8o percent during a century of Western encroachments. 61 According to Inoue, Hawaii could serve as an object

24 INTRODUCTION INTRODUCTION 25

lesson to Japan since Hawaii "is very similar to our country." Since "the ancestors of the Japanese come from the south' it stood to reason that they, too, were an in‑

ferior race that would not prevail in a competition with Westerners. By contrast, Taguchi Ukichi, a proponent of free trade and an early advocate of southern expan­sionism (nanshinron), favored mixed settlement areas and argued that the Japanese were not racially inferior to Westerners. In addition, he denied that they were of South Seas origins. Taguchi later argued, in a 1905 essay, that the Japanese were re­lated to the Hungarians, that both were "beautiful white people and that their lan­guages shared common linguistic roots. 1169 Though Taguchi and Inoue took opposing positions on the issue of mixed settlements, they were in agreement that the south­ern peoples belonged to an inferior race .70 Both thinkers placed the whites at the top of the racial hierarchy and southern races at the bottom; the question that divided them was where the Japanese stood within this hierarchy.

In the late nineteenth century, these diverse thinkers generally envisaged inter­national racial hierarchies through a distinctly social-Darwinist lens. Shiga and In‑

oue treated.Japan as one of the "weaker" or inferior races, that is, losers in the strug‑

gle for survival, and argued that Japan should resist Western pressures and protect its national identity. At a time when Japan occupied a weak position in the inter‑

national order, these writers regarded the struggle for survival from the perspec­tive of the inferior races. Most writers switched rapidly to the perspective of the stronger-races after Japan's victory in the Russo-Japanese War.

Japan's defeat of Russia signified that the nation had attained military parity with the West. The victory by an Asian country over a major European army generated

new attempts in the West to define Japan's position in the racial hierarchy. William

Elliot Griffis, a natural science teacher who had worked in Japan from 1870 to 1874, reclassified Japanese as "Caucasian" rather than "Mongoloid:' based on the popu‑

lar notion that the Ainu, the original inhabitants of Japan, were a Caucasian race.7'

Within Japan, military victory was seen as a vindication of the superiority of Japa­nese blood and of the Japanese racial make-up. Henceforth, many intellectuals mo‑

bilized social-Darwinist postulates to justify Japan's new position as the leading na‑

tion of Asia. In particular, a group of writers associated with southern expansionism began to use racial rhetoric to justify the nation's advance into the South Seas. Much

like Inoue TetsujirO, they believed that the Japanese shared "blood" or racial ties with the people of the region. Unlike him, they argued that the existence of these blood ties justified Japan's further expansion rather than its retreat to the protected shores of Japan's home islands.

In 1915, Inoue Masaji (1876-1947) likened contemporary Japanese to the "su­perior" Anglo-Saxons to justify their new position in the racial hierarchy.

I have long contended that Japanese expansion into the South Seas is a matter of re‑

turning to the ancient past of our race prior to Emperor Jinmu. . . I do not think it

an insult to the Japanese to say that we have a large share of southern blood in our veins. Anthropologists [ jinshugakusha: literally, "race scientists"] have argued that superior and victorious peoples, such as the Anglo-Saxons, formed themselves by ab­sorbing and fermenting several different racial strains. How can the Japanese people, who have received the blood of the southern peoples and purified it, leave the south­ern races to languish in their backwardness? By leading them, developing them, and fostering their happiness, we are not merely laying down the so-called royal road for the barbarians; from our point of view as Japanese, we have the good fortune of be­ing able to return to our original home from the time before Emperor Jinmu. . . . The southern expansion is inevitable, natural, and indispensable. 72

Inoue begins by asserting that the ancient Japanese advanced from the South Seas to their present home. Based on this racial past, he argues that the modern Japanese should return to the place of their birth. Significantly, he does not claim that the South Seas islanders are the only ancestors of the Japanese, but he implies that they are the most ancient ones. Later, as the Japanese "advanced to the north by conquest:' they incorporated the blood of other peoples into their heterogeneous racial make­up. As a racial strain composing the modern Japanese, the southern islanders are distant kin toward whom the Japanese feel both a sense of superiority and a pa­tronizing sense of obligation. Indeed, when Inoue notes that it is not an "insult" to say that the Japanese have southern blood, he implies that he agrees with prevail­ing conceptions about the congenital inferiority of South Seas islanders. Unlike their primitive and .inferior forebears, the Japanese have "fermented" and "purified" their southern blood, transforming themselves into a superior and victorious race, just as the Anglo-Saxons had done. Inoue makes use of pseudoscientific racialist dis­course to identify the Japanese as the Anglo-Saxons of the Pacific, and he also justifies Japan's expansion of the South as a return to its original homeland.

Japanese conceptions of civilization follow a similar trajectory to those of race that I have outlined above. If racial discourse spatializes cultural differences among the various societies in the world, the discourse of civilization temporalizes these differences by plotting all human societies as occupying different stages in a uni­versal scheme of historical development.73 To be sure, long before the opening of Japan, Japanese thinkers possessed well-developed notions of "civilization" and "bar­barism:' which often were signified by visible markers such as style of dress or be-havior.74 Unlike Tokugawa notions of civilization, however, the modern notion of civilization and enlightenment (bunmei kaika) was a dynamic idea that signified that all human societies were progressing along the same path, but that they were moving at variable speeds. In accordance with this idea, differences among social groups were interpreted as different stages of development along a single trajectory: societies were deemed "backward" or "advanced" according to their position along the path of progress.75

In his Bunmeiron nogairyaku (An Outline of a Theory of Civilization), Fukuzawa

26 INTRODUCTION INTRODUCTION 27

Yukichi set forth a theory about the temporal development of civilization: all soci­eties of the world (including Japan) occupy a place in history according to their greater or lesser distance from the most advanced Western countries. He divided all societies into three categories, representing different stages of evolutionary progress: first, the "primitive" stage, in which "neither dwellings nor supplies of food are stable" and in which man "is dependent upon arbitrary human favor or acci­dental blessings"; next, the "half- civilized' stage, where "daily necessities are not lacking since agriculture has been started on a large scale" but where people only "know how to cultivate the old," not "how to improve it"; finally, the' stage of civi­lization, where on the basis of material abundance, humans have invented the "prin­ciple of invention itself" and can "plan great accomplishments for the future and commit themselves to their realization:'76 Civilization is at once the ideal to be at­tained, the end point of a progression, and the summit of a hierarchy.

In addition, Fukuzawa located the different continents of the globe within this temporal schema: "When we are talking about civilization in the world today, the nations of Europe and the United States of America are the most highly civilized, while the Asian countries, such as Turkey, China, and Japan, may be called half-civilized countries, and Africa and Australia are to be counted as still primitive lands"77 Such designations are "common currency," accepted even by those "labeled semicivilized or barbarian:' The distance between their countries and those of the West is "the overriding concern of Asian intellectuals today 1171

Though Fukuzawa borrowed from the previous studies of Buckle and Guizot in his theory of civilization ,79 he freely modified their ideas to fit the case of Japan. Buckle and Guizot had treated climate and race, respectively, as the key factors that accounted for the worldwide dominance of European civilization. Rejecting this determinist approach, Fukuzawa saw civilization as a universal process open to all nations rather than a system confined to Europe. In his view, nothing would pre­vent Japan from assuming its rightful place as a full member of the community of civilized nations, provided that it reformed its institutions and society80

Furthermore, he defined civilization as a "relative" concept inasmuch as it could only be defined in relation to what civilization was not: "civilization is escaping from barbarism and gradually advancing along the path of progress" Nevertheless, the converse of this statement.was also true: barbarism could be defined only from the point of view of civilization. Fukuzawa located the forces opposing civilization in the internal weakness of Asia: Oriental despotism of China and "the old customs of Japan" Most important, he differed from his European sources in the position that he, as enunciator, occupied within the hierarchy of civilization.8' He placed Japan (but also Turkey and China) in the middle stage of half-civilized (hankai), behind the nations of the civilized West but ahead of the primitive societies of Africa, Oceania, and Asia. This notion of the middle ground offered Japan a lever to raise its position in this hierarchy, precisely because, as Fukuzawa also pointed out, civ­ ilization is a relative notion: "If we compare China with countries of South Africa, or to take an example closer to hand, if we compare the Japanese people with the Ezo, then both China and Japan can be called civilized. 1112 At the time he wrote An Outline of a Theory of Civilization, Fukuzawa stressed that Japan needed to develop internally before it could engage in military ventures abroad. Catching up with the West did not yet entail colonizing Asia, but it most definitely included the adop­tion of a Western viewpoint toward Asia.

- The recognition of Japan's position as a middle ground might have served as a

basis for solidarity between the Japanese and other Asians. Instead of joining forces with its neighbors to resist Western encroachments, however, Japan appropriated Western civilization to improve its own position in the international order, thereby weakening this middle ground between civilization and barbarism. After the failed coup d'etat by the pro-Japanese reformists in Seoul (the Kapsin Incident), Fukuzawa published an editorial in the Jiji Shinpo (Jiji Newspaper, Tokyo), titled "Datsuaron" (Disassociation from Asia), in which he urged his countrymen to break with East Asia and to treat its neighbors as Western imperialist nations were doing: "Japan is geographically on the eastern end of Asia, but in terms of our national spirit we have left the old confines of Asia and have joined Western civilization. "13 Yet even as he identified Japan with European-style civilization, he compared the advance of the latter to the spread of measles: "Measles started in Nagasaki and is gradually spread­ing to the east with the spring thaw:' As the spread of Western civilization was in­eluctable, the nations of the East could not resist it any more than the people of Na­gasaki could stop the spread of measles. Like a contagious disease, civilization was an alien product whose introduction had a corrosive effect on Japanese society How­ever, the Japanese should "encourage its spread and make sure that all countrymen infect themselves with its spirit" because the transmission of civilization was ulti­mately beneficial, strengthening the health of the nation. 84 Besides recommending that the Japanese "infect themselves with the spirit (of civilization) Fukuzawa urged them to "quarantine" themselves from Korea and China, because these countries "violate the natural law of the spread of civilization:' During the Meiji period, Japan had broken with Asia in its "national spirit' but "from the perspective of civilized Westerners' Japan would be judged in the same way as were China and Korea "be­cause of the geographical proximity of the three countries:' Fukuzawa feared that Japan would be the object of a Western orientalizing gaze that would be unable to distinguish Japan from its "lawless" and "unscientific" neighbors. 85

Rather than let foreigners impose their civilization upon Japan, Fukuzawa sum­moned the Japanese to preempt them and to step forward as the agents of Japan's cultural transformation: the Japanese must colonize themselves and internalize the orientalizing gaze of the West that views Asia and Japan as uncivilized. In his "GaikOron" (On Diplomacy; 1883), Fukuzawa exhorted his countrymen to "first abandon their old customs and usages and to reform everything from their politi‑

28 INTRODUCTION INTRODUCTION 29

cal system, laws, and education down to the smallest details of their everyday lives in society. . . and to strive to borrow from Western ways, resolving to make their

country in the eastern region of Asia a new Western nation."86 In a later passage of the same essay, he called for the Ainu and RyükyU on the periphery of the nation to practice a similar form of self-colonization, but in this case the model to be copied was Japan itself. Indeed, he argued that the process of civilization will be easier for these peoples because the degree of difference in people and culture is not so great as that between Japan and the West; 87 here he offered a template of colonial discourse that would later be applied to other Asians. In a first stage of self-colonization or auto-Orientalism, the Japanese accepted and internalized Eurocentric standards and value judgments that judged Japan as backward and uncivilized." In a second and later stage, they discovered barbarians on their periphery and set out to colonize them in the name of civilization. The motto of dissociating from Asia was achieved not only by civilizing Japan but also by discovering uncivilized others among Japan's Asian and Pacific neighbors.89

THE RHETORIC OF SAMENESS/SIMILARITY

In an essay titled "Japanese Colonialism:' in his book The Japanese Nation, Nitobe describes the policies of blockade and extermination that the Japanese employed to force aboriginal headhunters of Taiwan to surrender their lands in the early twen­tieth century. "When they are practically caged, we make overtures to them. We say, 'If you come down and don't indulge in headhunting, we will welcome you as brothers" This rhetorical transformation of headhunters into brothers resembles the Western "civilizing mission:' in which the representatives of advanced nations assume the burden of raising backward savages from their benighted state and thereby redeem their humanity. However, Nitobe's claim of "brotherhood" is premised not simply on his notion of "civilization" and shared humanity but also on a belief that the aborigines are of Malay origin and hence are racial cousins of the Japanese. "These Malay tribes resemble the Japanese more than they do the Chi­nese, and they themselves say of the Japanese that we are their kin and that the Chi­nese are their enemies "'0 In midsentence, Nitobe shifts from speaking in the first person to recording the viewpoint of unidentified, plural aboriginal speakers. This curious shift from "they are like us" to "we are like you" is not uncommon in Japa­nese imperial rhetoric. In this statement, we have a perfect instance of what I would call imperial ventriloquism, the attribution of views by the colonizer to the colo­nized in Order to create the false impression of dialogue, whereas in fact the speaker's only real interlocutor is himself. Nitobe puts words into the mouths of the colonized. He was far from the only practitioner of this imperial ventriloquism.9i In calling attention to the solipsistic nature of Japanese colonial discourse, I do not mean to imply that the colonized did not participate in the production of discourses

on the original sameness of the Japanese and the people they colonized. In fact, there were many writers in Korea and Okinawa who also wrote about the original unity of Japan and its colonies.92 Nevertheless, the Japanese were the creators and main practitioners of this discourse.

Often, Japanese speakers based their notion of similarity with the colonized on claims of physical resemblance or common descent. Takekoshi recalled meeting an aboriginal student at medical school in Taipei who "bore no resemblance to ordi­nary Formosan students but reminded me much of Japanese students from Kyushu" Indeed, he recommended that specialists undertake historical study to compare the Taiwan savages with "the Kumaso family in Kyushu. . . or with the ferocious chief­tain Nagasunehiko," the leader of a clan in the ancient past that had resisted Em­peror Jinmu in his subjugation of the eastern territories.93 Like Takekoshi, Fujiwara Toyo, one of the founders of physical education for women, compared the bodies of Taiwanese aborigines to those of Japanese peasants. "The sëiban say one must not kill the Japanese because they are brothers, but they regard the Taiwanese who have emigrated from China as enemies and want to kill them. Since the facial struc­ture of the seiban is exactly the same as that of Japanese peasants, one might sur­mise that the Japanese in the ancient past used to have a physical constitution like that of the seiban."94

If some scholars stressed Japan's similarities of physical constitution and ap­pearance, Others emphasized that the Japanese were racially related to the people they colonized. By far the best known of these theories, the so-called nissen dOsoron, held that Korean and Japanese races were of common ancestry. The linguist Kanazawa ShOzaburO (1872-1967) coined the term "nissen dOsoron" in an epony­mous 1929 book. Noting the similarity between Korean and Japanese languages, he claimed that the former was a branch dialect of the latter, just like the RyUkyU language.95 Anthropologists, who compared present-day inhabitants of both places, held that the Koreans were all but indistinguishable from the Japanese in their cra­nial and bodily measurements 96 Based on early historical interaction between Ko­rea and Japan, Hoshino Hisashi and Kume Kunitake argued that Japan had ruled the peninsula from the fourth to the seventh centuries; this past experience offered a precedent for the Japanese annexation of Korea in the modern period.97 As the empire expanded into Manchuria and Mongolia, Japanese scholars argued that both Japanese and Koreans were part of a single racial and cultural grouping that en­compassed most of Northern Asia. First proposed by nineteenth-century European linguistic scholars, this Ural-Altaic hypothesis was used to justify further conquests on the Asian continent in the name of shared racial heritage. They also used such theories to justify treating Manchuria, Korea, and Mongolia as regions (rather than nations) and to detach China from its borderlands.98

As these examples show, Nitobe's rhetorical move exemplifies a general feature of Japanese colonial discourse: its tendency to advance claims of similarity between

INTRODUCTION INTRODUCTION 31

the colonized and the colonizer. In its emphasis on similarity; this discourse stands opposed to the typical discourse of Western colonial powers.99 In most of their colonies, Europeans drew a clear line between themselves and those they ruled, and they actively discouraged any border crossings and mixing. When theorists use these European models to describe colonial relationships, they characterize them as bi­nary relationships in which the colonizer constitutes himself or herself in opposi­tion to the colonized.'00 If Western colonizers tended to emphasize the "otherness" of the colonized, Japanese colonial administrators made sameness a mode of rule in the Japanese empire.'°'

Why did the Japanese couch their claims of imperial rule in this rhetoric of sim­ilarity and analogy rather than that of binary opposition and difference? One might argue that this Japanese rhetoric had a certain basis in reality, notably in-the geo­graphical and cultural characteristics of the lands Japan colonized. As historians have pointed out, the Japanese empire differed from other empires in that Japan colonized territories in close proximity to the home islands and ruled peoples with whom it shared a common culture and history.'02 'While this assertion certainly con­tains a grain of truth, it does not explain why Japan made neighborliness and same­ness a feature of its rhetoric of rule. To start with, Japan most certainly colonized adjacent lands from necessity rather than from choice. By the time Japan entered the world-wide scramble for overseas colonies, the other world powers had already seized most of the world's "unclaimed" territories in Africa, Asia, or the Pacific. In addition, until the first decades of the twentieth century, Japan lacked a powerful navy, was unable to project its power far from the home islands, and could only col­onize geographically proximate territories.

Another reason why Japanese writers drew on the rhetoric of similarity has to do with the policies of assimilation the Japanese adopted to transform the colonized into future Japanese. As is well known, the dominant Japanese policy toward the colonized was dOka, a word formed by two graphs meaning "to make the same" and translated as assimilation. However, if the Japanese sought to make the colonized the same as themselves, then it followed that they were perfectly aware that they were separated from those they ruled by differences. These differences resulted from centuries of political separation, to say nothing of cultural and linguistic barriers, and had created a huge gulf between Japan and its colonies. In the context of as­similation policies, the rhetoric of sameness was less an acknowledgment of real similarities between colonizer and colonized than a legitimizing discourse that ex­plained away the gap between a similarity posited in the past and a difference ac­knowledged in the present. It also concealed the contradictions of colonial policy discourse and masked the realities of discrimination and the structures of Japanese domination.103

Indeed, the Japanese rhetoric of sameness was no less hypocritical than the West­ern rhetoric of otherness, since it legitimated policies that were discriminatory and racist toward the colonized. "To make the same" tended to foreclose the option of granting the colonized political rights and equality with the Japanese. Particularly late in the colonial period, "making the same" entailed coercive policies toward Ko­reans, Taiwanese, and Micronesians such as forced name changing, prohibitions of indigenous languages, compulsory worship in Shinto shrines, and induction of qual­ified males into the Japanese military. Such policies, implemented with great bru­tality in all Japanese colonies from the late 1930s, smothered the identities of the colonized and repressed any form of cultural difference. When they imposed these policies colonial officials claimed that they were granting "benefits" on the colo­nized by "raising them to the level of the Japanese" In a vicious and perverse cir­cle, the greater the violence that colonial authorities visited on the colonized, the greater the benefit they claimed to be conferring upon them.

Even though Japanese. imperialists often couched their arguments in this rhet­oric of sameness, it was by no means the only idiom available in their lexicon to justify colonial rule. Indeed, some writers downplayed racial and cultural similar­ities between themselves and other Asians and emphasized their affinities with West­ern powers. Fukuzawa Yukichi described the Sino-Japanese War (1894-95) as a war of civilization against barbarism, thereby conflating Japan with the civilized West and distancing it from China. During the Russo-Japanese War (1904-5), the Japa­nese government dispatched Suematsu KenchO and Kaneko KentarO to Western na­tions to convince the Western public that Japan was waging a war on behalf of West­ern values and to refute "yellow peril" propaganda.104 In a speech given in London, Suematsu ridiculed the idea—promoted by Yellow Peril ideologues—that Asians shared either practical interests or common cultural traditions.

Can anyone imagine that Japan would like to organize Pan-Asian agitation of her own seeking, in which she must take so many different peoples of Asia into her own confi­dence and company—people with whom she has no joint interests or any commu­nity of thought and feeling?. . . How then could it be expected for one moment that the various peoples of the East, with their varying degrees of intelligence, their con­flicting interests, and their long-standing feuds and jealousies, could ever have cohe­sion enough to range themselves under one banner against the powers of the Occi­dent? And if they could do so, is it to be imagined that Japan would enter upon so quixotic an enterprise as to place herself at the head of so unmanageable a mob?... The peoples of the East are, some of them, politically independent; others are under the sway of one or another European power. To combine them in a single undertak­ing would be a task utterly impracticable and unpromising. Japan has already cast her lot with the Occident, and in the eyes of many Asiatics it is to be remembered that the Japanese are no less "Yang-Kwai" (foreign devils) than the Occidentals1105

If some writers sought to play up Japan's resemblances to the people they colo­nized, others categorically asserted that the Japanese were racially superior to all other Asians or advanced claims that they were racially unique. Like Natsume Söseki

32 INTRODUCTION INTRODUCTION 33

in Manchuria, travelers to the colonies associated other Asians with images of "filth, squalor, and indolence" and identified themselves with values of civilization, dili­gence, and hygiene. 106 In wartime propaganda, the Japanese people were described as belonging to a pure race and possessing a unique culture to distinguish them from their rivals and enemies in the West. Indeed, Japanese "Orientalism" differs from its Western variety; since it fuses assertions of the superiority of the Japanese in relation to other Asians with claims about the cultural uniqueness of Japan in relation to the West. Depending upon time and place, Japanese played up affinities between themselves and other peoples of Asia or stressed differences, claimed to represent Western civilization or asserted their absolute uniqueness. Accordingly, the Japanese rhetoric of sameness was not invoked consistently and did not pos­sess any conceptual coherence. On the contrary, duality, not to say duplicity, char­acterizes Japanese writings about the colonized. 107

At least initially, the rhetoric of sameness reflected the weakness of the Japanese empire and its status as a latecomer. The Japanese themselves had imported West­ern civilization and technologies only a few decades before they claimed to intro­duce these novelties to their colonies. Unlike the Europeans, they could not pre­tend that they were spreading their own universal civilization to the colonized. It is noteworthy that Western legal advisors on Japanese policy toward Taiwan, Japan's first colony, recommended that the government adopt a system of indirect rule in order to reduce the cost to the Japanese treasury. In spite of this advice, the Japa­nese colonial regime favored the more costly alternative of assimilation because they feared that the subject people, left to their own devices, would lack loyalty to the Japanese state. The government laid special stress on the inculcation of the Japa­nese language and emperor worship to foster the loyalty of the subject peoples. A. Japanese educator wrote in 19o1:

Japan is weaker than and inferior to the West in the military, economic, and cultural spheres and is slightly superior to China and Korea. Aside from this, we must hold onto one thin thread of hope. Namely, Japanese ethics, part of our education cur­riculum and a very important subject, is something that we can spread throughout the world. Our spirit is superior to that of the Western peoples. I concede that we must yield to them in matters relating to material conditions. But we do not need to yield to them in ethics. It is a matter of policy that we should introduce material civiliza­tion to the Chinese. But it is our duty to infuse them with Japanese ethics. 108

As a discourse turned toward the colonized, the rhetoric of sameness sought to "in-fuse

frfuse [the colonized] with Japanese ethics:'

As a discourse directed toward the West, this rhetoric showed that the Japanese had a unique relationship with the people they colonized, a relationship that set them apart from colonization on the Western model.109 In 1910, Count Okuma Shigenobu, founder of Waseda University, wrote that "many European countries expanded into countries with different races, different nations and different reli­gion unlike Japan, which expanded into areas occupied by the same race and the same nation.... As for the Taiwanese and the Koreans, they hardly differ from the Japanese in their thinking, their feelings, their customs, manners, and religion ""° Whereas Western imperial powers face "innumerable problems in their attempts to rule over different races. . . in our case, the path to assimilation [dOka] will trans‑

form the Koreans into loyal, obedient To be sure, Okuma did not want

to extend to Koreans the same rights granted to Japanese under the Meiji consti­tution when he spoke of transforming them into "loyal, obedient subjects. 11112 He simply meant that the racial similarity of the Japanese and Koreans would smooth the path to Japan's colonization and obviate the need for Japan to employ violence, the means that Western powers had used to subjugate their radically different and refractory subjects.

Yet even after the Japanese empire became strong in the early part of the twen­tieth century, Japanese writers continued to deploy the rhetoric of sameness to jus­tify its rule over the different people of the Asia Pacific region. Proponents of pan-Asianism argued that Japan was destined to lead East Asia both because it was a modern nation and because it shared a common culture and racial identity with its neighbors. While pan-Asianists initially sought to reform decaying Asia along Western lines, they later sought to oust Western imperialism, notably during the Second World War."' In the 1940s, supporters of the Greater East Asian Co-Prosperity Sphere called for the unification of peoples "closely related to each other geographically, racially, culturally, and economically" and pledged to end centuries of Western hegemony.' 4 This anti-imperial rhetoric both distinguished Japanese imperialism from the predatory Western variety and sought to bind the colonized to the Japanese with claims of racial solidarity. In fact, the terms of this rhetoric were again and again reinterpreted, transformed, and redirected to serve new pur­poses. On the one hand, claims of similarity offered an ideological justification for Japanese colonialism, distinguishing it from that of the West. On the other, the no­tion that the Japanese shared a common identity with the colonized provided an important paradigm for early twentieth-century history, ethnography, archaeology, and so on. When they turned their attention to territories under Japanese imperial control, Japanese scholars often argued that there was an ethno-racial continuity between the Japanese and the people they colonized. In this respect, Japanese im­perialism resembles the continental imperialism (pan-Germanism or pan-Slavism) analyzed by Hannah Arendt, which also used racial and cultural ties to justify ter­ritorial unification.

Whatever the precise motivation of individual exponents of this discourse, I will argue that the fact that the Japanese sought to claim a fictive unity and imaginary kinship with those they colonized is highly significant and calls for close examina­tion. In the course of this book, I will consider the discourses of a variety of writ‑

34 INTRODUCTION INTRODUCTION 35

ers, literary and otherwise, who deployed this rhetoric of similarity. In the works of these writers, similarity was posited at different sites including culture and civ­ilization, race, and physical bodies. This rhetoric assumed two different forms, one strong the other weak, that I will distinguish analytically. In the strong form, the Japanese were said to possess attributes (race, culture) that made them absolutely similar to the colonized, thereby justifying Japanese colonial rule on the basis of common qualities. In the weaker version, the Japanese were, by comparison with Westerners, relatively similar to the people they colonized on these grounds, Japa­nese rule was preferable to that of Western colonizers. Japanese speakers invoked both sets of justifications ever more aggressively in the post-World War I period, when President Wilson's call for self-determination, the Russian Revolution, and the upsurge in nationalist movements throughout the colonized world put all kinds of imperialism, whether Western or Eastern, on the defensive.

Just as Japanese imperial rhetoric often stressed the similarities between colo­nizer and colonized, distinguishing their imperial rule from that of Western pow­ers, it described the Japanese empire in ways that distinguished Japan from the West: anti-Western and "anti-imperialist" terms. A kind of anti-imperial imperialism shaped the ordinary language with which the Japanese spoke of their colonies and referred to themselves and the people they ruled. When Japan acquired Taiwan in 1895, politicians and journalists carefully avoided referring to this "newly acquired territory" as a colony, no doubt because the term "colony" had negative connota­tions.' 5 For example, Hara Kei wrote: "The situation of the [Taiwanese] people is absolutely different from that of the different countries that establish rule over people of different races:' In 1910, speaking of Korea, Prime Minister Hara wrote that it was an extension of Japan rather than a territorial dependency or a colony:"6 In journalism until the First World War, Japanese colonies are often referred to sim­ply as shin Nihon, or the new Japan, and their residents as new Japanese (shin ni-honjin). Not only did the Japanese eschew referring to their "foreign territorial pos­sessions" by the term "colony:' but they claimed to have acquired these lands through neutral, peaceful processes: Japan did not "annex:' it "merged" (heigo suru) with Korea, and the Qing "ceded" (katsujO suru) Taiwan to Japan. In the case of Man­churia, Japanese rulers strove to preserve the fiction that they were dealing with an "independent country:"7 To be sure, these euphemisms concealed the realities of colonial war (Taiwan), military occupation (Korea), and invasion (Manchuria); nev­ertheless this curious nomenclature deserves attention in its own right. Japanese history textbooks still refer to the "cession" of Taiwan and the "merger" with Korea.

People commonly distinguished Japan (naichi) from the colonies (gaichi): these two terms could be rendered in English as inner and outer lands. Indeed, gaichi de­noted not only colonies like Korea and Taiwan, but also Hokkaido and Okinawa, both peripheral spaces between nation and colony that had been incorporated into Japan in the 1870s.1 18 Reflecting its semicolonial status, Okinawa in particular was (and to some extent still is) conceptually aligned with a colonial outside rather than with an inside of the nation. If pre-1868 Japan constituted naichi, then the Japanese who settled in the colonies were called naichijin, rather than simply nihonjin (Japa­nese). (I will note here that the term gaichijin, as a conceptual counterpart to naichi-jin, is rarely encountered in colonial period writings.) "Japanese" clearly included Taiwanese and Koreans, as is shown by 1930 census reports that give Japan's pop­ulation as a hundred million, a figure that included thirty million people in the colonies! 19 In colonial discourse, "Japanese" was an umbrella term that did not dis­tinguish between Japanese and Taiwanese (or Koreans), whereas naichijin specifi­cally excluded the colonized. The distinction between naichijin and nihonjin was a crucial semantic difference that spoke volumes about Japanese colonial policy. In a suggestive passage of his travel memoir Musha, Satö Haruo transcribes a rumor about a massacre of Japanese by rebellious aborigines that he happened to overhear in an inn where he was staying: "All of the nihonjin [Japanese] in the village of Musha had been murdered" He comments sardonically that, "strictly speaking, this should have been 'all the naichijin were massacred: to use the terms that our rulers have taught us" When the speaker blurts out "Japanese" in a moment of excitement, he reveals his assumption that, notwithstanding the euphemisms then in vogue, the "Japanese" did not include the colonized. 120

ORGANIZATION OF THIS BOOK

In Tropics of Savagery, I offer four specific case studies of Japanese imperial litera­ture and of figures of its colonial discourse. Chapters 2 and 4 focus on individual writers: SatO Haruo, who traveled to Taiwan in 1920, and Nakajima Atsushi, who lived in Palau between 1941 and 1942. Chapters i and 3 focus on tropes for "sav­agery:' the figure of the headhunter in Taiwan and the "ogres" of the South Seas in the MomotarO legend. The first two chapters of the book deal with literature related to Taiwan, Japan's first colony, while the last two chapters treat works on Microne­sia and the South Seas. While I have not followed strictly chronological order in this book, my individual chapters trace the stages of Japan's imperial trajectory di­achronically: early conquest, liberal critiques of colonialism, late-colonial integra­tion, and war-time mobilization.

In chapter 1, I recount the history of Japan's brutal military conquest of the Tai­wanese aborigines and examine the legal rationales that colonial officials offered to justify this conquest. As Japanese military force encountered fierce resistance by aboriginal groups, the aborigines came to be viewed in Japanese discourse entirely in terms of "headhunting: Even as Japanese colonial administrators endeavored to stamp out this custom, they carefully preserved the figure of the "headhunter" as a trope, and they circulated images of this "headhunter" to justify the violent subjugation of the savage "Other" and to affirm their civilizing mission. In the

36 INTRODUCTION INTRODUCTION 37

legend of Go HO, a story that was widely disseminated to young Japanese and Taiwanese, a Qing official sacrifices his own life in order to persuade the savages to renounce headhunting. Once the savages were incorporated into the Japanese em­pire, some Japanese writers discovered a savage within themselves. The protagonist of Oshika's "The Savage" goes to the aboriginal lands in search of his inner savage and becomes a "headhunter" to free himself from civilized modernity. Just as an earlier generation of writers drew on global discourses on civilization to write tales of Japan's colonial conquests of savages, late imperial writers appropriated tropes of primitivism prevalent in Western literatures to address a critique of Japanese modernity.

In chapter 2, I examine the early development of Japanese anthropology and its influence on colonial literature. Western scholars introduced the science of anthro­pology to Japan in the 187os and conducted the first scientific investigations into the origins of Japanese people. Within a decade, Japanese scholars "nationalized" this foreign science and brought it to bear on the aboriginal population of Taiwan, which quickly became the first overseas field in which Japanese anthropologists could work. In the next few decades, colonial ethnographers expanded their field­work to embrace all of the new territories that fell under Japan's dominion.

As a genre of writing about aboriginal societies, ethnography provided a model for the writer SatO Haruo, who traveled to Taiwan in 1920 and became acquainted with the ethnographer Mori Ushinosuke. A few years after he returned to Japan, SatO wrote MachO (Demon Bird), a short story based on a passage in Mon's ethnog­raphy. The ethnographer-narrator of "Demon Bird" writes about an episode of per­secution in an unnamed barbarian village. At the same time, the story he tells is an allegory about Japanese persecution of Koreans during the Great Kanto Earthquake. "Demon Bird" is a story that uncovers unexpected links between colony and me­tropolis. The work appeared at a time when criticism of. Japan's colonial policies by liberal and reformist intellectuals was at its peak.

In chapter 3, I consider MomotarO, Japan's most famous folktale, and his trans­formation in the early twentieth century into an allegory for Japanese expansion toward the South Seas. To overcome the dearth of overseas adventurers in Japanese history, advocates of imperial expansion championed the mobilization of this folk­lore hero to spark the interest of Japanese youth in the acquisition of overseas colonies. Nitobe InazO saw MomotarO as a pedagogical tool that could fire the imag­ination of Japan's insular youth and spur them on to colonize the South Seas. Other writers believed that MomotarO was, at best, a flawed model for Japanese colonial­ism, and even blamed the failures of Japanese colonial policy on his harmful in­fluence. In 1925, Akutagawa published a satire of MomotarO in which the peach boy is portrayed as a cruel invader who brutally attacks a group of humanized ogres living peacefully on an island paradise in the South Seas. At the end of this story, young ogres counterattack and fight to win the independence of their homeland.

At the intersection of folklore, propaganda, and parody, MomotarO emerges as a contested site for debating the Japanese imperial project and for defining self and other in the age of empire.

In chapter 4, I look at the South Seas literature of Nakajima Atsushi (1909-42), an important writer who published most of his work during the period of the Greater East Asian Co-Prosperity Sphere. Nakajima, who grew up in Japanese-ruled Korea in the 1920S, went to Palau in 1941 to work as editor of Japanese language text­books in Japanese-ruled Micronesia. Returning to Japan after about nine months, he wrote two collections of stories that offer the reader a compendium of colonial discourses and stereotypes on nan'yO during this late period of the Japanese em­pire. I examine the colonial filter through which the narrator of these works regarded the islands of Micronesia and the people that inhabited them. This late colonial writer offers us a sensitive reflection on Japanese "imperial mimicry" and a self-conscious assessment of the Japanese imperial gaze.

In my conclusion, I examine tropes of savagery in literature of the postwar period. Cannibalism features as a major theme in three prominent pacifist and humanis­tic novels that were published within ten years of Japan's defeat and the loss of its empire. By, focusing on this theme, I examine continuities and discontinuities be­tween tropes of savagery in the colonial era and in the immediate postcolonial period.