PAUL D.

BARCLAY

OUTCASTS
OF
JAPAN’S RULE ON
TAIWAN’S “SAVAGE BORDER,”
1874–1945

EMPIRE
 Luminos is the Open Access monograph publishing program
from UC Press. Luminos provides a framework for preserving and
­reinvigorating monograph publishing for the future and ­increases
the reach and visibility of important scholarly work. Titles
­published in the UC Press Luminos model are published with the
same high standards for selection, peer review, production, and
marketing as those in our traditional program. www.luminosoa.org
 ASIA PACIFIC MODERN
Series Editor: Takashi Fujitani
1. E
 rotic Grotesque Nonsense: The Mass Culture of Japanese Modern Times, by Miriam Silverberg
2. V
 isuality and Identity: Sinophone Articulations across the Pacific, by Shu-mei Shih
3. Th
 e Politics of Gender in Colonial Korea: Education, Labor, and Health, 1910–1945, by
­Theodore Jun Yoo
4. F
 rontier Constitutions: Christianity and Colonial Empire in the Nineteenth-Century
­Philippines, by John D. Blanco
5. T
 ropics of Savagery: The Culture of Japanese Empire in Comparative Frame, by Robert Thomas
Tierney
6. C
 olonial Project, National Game: A History of Baseball in Taiwan, by Andrew D. Morris
7. R
 ace for Empire: Koreans as Japanese and Japanese as Americans during World War II, by
Takashi Fujitani
8. Th
 e Gender of Memory: Rural Women and China’s Collective Past, by Gail Hershatter
9. A
 Passion for Facts: Social Surveys and the Construction of the Chinese Nation-State,
1900–1949, by Tong Lam
11. R
 edacted: The Archives of Censorship in Transwar Japan, by Jonathan E. Abel
12. A
 ssimilating Seoul: Japanese Rule and the Politics of Public Space in Colonial Korea, 1910–1945,
by Todd A. Henry
13. Working Skin: Making Leather, Making a Multicultural Japan, by Joseph D. Hankins
14. I mperial Genus: The Formation and Limits of the Human in Modern Korea and Japan, by
Travis Workman
15. S anitized Sex: Regulating Prostitution, Venereal Disease, and Intimacy in Occupied Japan,
1945–1952, by Robert Kramm
16. O
 utcasts of Empire: Japan’s Rule on Taiwan’s “Savage Border,” 1874–1945, by Paul D. Barclay
Outcasts of Empire
 The publisher and the University of California Press
­ oundation gratefully acknowledge the generous support of the
F
­Philip E. ­Lilienthal Imprint in Asian Studies, established by a major
gift from Sally Lilienthal.
 Outcasts of Empire
Japan’s Rule on Taiwan’s “Savage Border,” 1874–1945

Paul D. Barclay

UNIVERSIT Y OF CALIFORNIA PRESS

University of California Press, one of the most distinguished university
presses in the United States, enriches lives around the world by ­advancing
scholarship in the humanities, social sciences, and natural sciences. Its
activities are supported by the UC Press Foundation and by philanthropic
contributions from individuals and institutions. For more information,
visit www.ucpress.edu.

University of California Press

Oakland, California

© 2018 by Paul D. Barclay

Suggested citation: Barclay, Paul D. Outcasts of Empire: Japan’s Rule on

­Taiwan’s “Savage Border,” 1874–1945. Oakland: ­University of California
Press, 2018. doi: https://doi.org/10.1525/luminos.41

This work is licensed under a Creative Commons CC by NC ND license.

To view a copy of the license, visit http://creativecommons.org/licenses.

Library of Congress Cataloging-in-Publication Data

Names: Barclay, Paul D., author.

Title: Outcasts of empire : Japan’s rule on Taiwan’s “savage border,”
 1874–1945 / Paul D. Barclay.
Other titles: Asia Pacific modern ; 16.
Description: Oakland, California : University of California Press, [2018] |
  Series: Asia Pacific Modern ; 16 | Includes bibliographical references and
 index.
Identifiers: LCCN 2017030926| ISBN 9780520296213 (pbk. : alk. paper) |
  ISBN 9780520968806 (ebook)
Subjects: LCSH: Taiwan aborigines—History—20th century. |
 Japan—Colonies—History. | Taiwan—History—1895–1945.
Classification: LCC DS799.42 .B37 2018 | DDC 951.249/04—dc23
LC record available at https://lccn.loc.gov/2017030926

27 26 25 24 23 22 21 20 19 18
10 9 8 7 6 5 4 3 2 1
For Naoko
 C onte nts

List of Illustrations and Tables ix

Acknowledgments xiii
Note on Transliteration and Translation xvii

Introduction: Empires and Indigenous Peoples, Global Transformation

and the Limits of International Society 1

PART ONE.  THE ANATOMY OF A REBELLION

1. From Wet Diplomacy to Scorched Earth: The Taiwan Expedition,
the Guardline, and the Wushe Rebellion 43
2. The Longue Durée and the Short Circuit: Gender, Language, and
Territory in the Making of Indigenous Taiwan 114

PART TWO.  INDIGENOUS MODERNITY

3. Tangled Up in Red: Textiles, Trading Posts, and Ethnic Bifurcation
in Taiwan 161
4. The Geobodies within a Geobody: The Visual Economy of Race
Making and Indigeneity 190

Notes 251
Glossary 293
Index 301
 L i st of Illustrati ons an d Ta bl es

F IG U R E S

1. Japanese colonial-period ethnonyms for Taiwan Indigenous Peoples   9

2. Inō Kanori’s ethnic map, 1898   10
3. Picture postcard map of Taiwan’s ethnic bifurcation, ca. 1904   34
4. Two Atayal men engaged in “conjoined drinking,” near Wulai, Taiwan,
ca. 1900  45
5. Scorched earth guardline, ca. 1910   47
6. Surrendered Atayal at a stopover in Jiaobanshan, ca. 1910   48
7. Weapons captured from Atayal peoples, ca. 1910   49
8. Saigō Tsugumichi, Japanese soldiers, Paiwan headmen, and the interpreter
Johnson, 1874   72
9.  A certificate attesting submission to the Japanese Expeditionary Forces,
1874  73
10. Relics preserved from the 1874 expedition, ca. 1910   74
11. Bottle and can from the meeting at Dakekan, 1895  84
12. The Linyipu District Administrative Office, 1898   92
13. Gantaban men with severed heads, 1903   97
14. Guards warn of the approach of hostile forces, ca. 1910  107
15. Cutting trees to build the scorched-earth barricades, ca. 1910  108
16. The large Taiwanese labor force, during construction of the aiyūsen,
ca. 1910  108
17. Porters hauling food and water for the guardline expeditionary troops,
ca. 1910  109
18. Watan Yūra, ca. 1900   123
ix
x     List of Illustrations and Tables

19. Watan Yūra, Kōan, Aki, and Pazzeh Watan, 1903   124
20. Kondō “the Barbarian” Katsusaburō, ca. 1910   130
21. Quchi-area residents posing with flag, ca. 1897   132
22. Kondō Gisaburō with Truku peoples, January 1915   137
23. Tata Rara with Japanese interpreter Nakamura Yūsuke, 1896   143
24. Tata Rara with her Puyuma militia, 1896   144
25. Pan Bunkiet, ca. 1900   145
26. The Wulai School for Indigenous Children, ca. 1910   150
27. Jiaobanshan model school for indigenous children, ca. 1930   153
28. Atayal textiles from Japanese ethnological survey, ca. 1915   170
29. A diorama from the 1913 Osaka Colonial Exhibition with Atayal red-striped
capes  171
30. Trading post at Jiaobanshan, ca. 1913  177
31. Atayal women wearing imported clothing and weaving traditional
clothing, 1936  186
32. Map of Taiwan, 1895   193
33. Japanese census map, 1905   193
34. Ethnic map of Taiwan, ca. 1912   193
35. An anthropology journal sketch of Watan Nawi, 1895   200
36. Jiaobanshan emissaries and Governor-General Kabayama as depicted in
Fūzoku gahō, 1895   201
37. Photograph of the Jiaobanshan emissaries and Japanese officials in Taipei,
1895  202
38. Photograph of Habairon, Motonaiban, and Washiiga, 1895   203
39. Ethnographic drawing of Habairon, Motonaiban, and Washiiga, 1896   204
40. Photograph of Habairon, Motonaiban, Ira Watan, Marai, Pu Chin, and
Washiiga, 1895   206
41. Textbook etching of Jiaobanshan emissaries, 1897   207
42. Jiaobanshan emissaries in fanciful setting, ca. 1900   208
43. Photo of Jiaobanshan emissaries in Western press, 1902   209
44. Etching of Sediq woman and Paalan headman in Ministry of Education
textbook, 1904  210
45. Photograph of Sediq woman and Paalan headman, ca. 1897   210
46. Etching of Sediq woman and Paalan headman in commercial textbook,
1908  210
47. Photograph of Mori Ushinosuke, Japanese officers, and Truku headmen,
1910  220
48. Official commemorative postcard depicting indigenous customs, 1911  224
49. Men and women along the Quchi guardline, ca. 1904   226
50. Japan’s Atayal allies along the Quchi guardline, ca. 1904   227
51. Dynamic and static maps of Taiwan’s ethnic diversity, 1912   231
 List of Illustrations and Tables    xi

52. Jiaobanshan as staging area for Gaogan offensives, ca. 1910   234
53. Jiaobanshan woman with basket and pipe, ca. 1930  237
54. Postcard sleeve, “Jiaobanshan’s hidden savage border,” ca. 1930  238
55. Couple in Jiaobanshan, ca. 1930  238
56. Mountains of Jiaobanshan, ca. 1930  239
57. Contrasting photos of Jiaobanshan and Paalan, 1935   240
58. Second-order geobody of Atayal   241
59. Marai and Yūgai of Rimogan, ca. 1903   242
60. Wulai dwelling and granary, ca. 1903   243
61. Yūgai and Marai in textbook illustration, 1919   243

MAPS

1. The major indigenous ethnic groups of northern Taiwan   7

2. Overview of the major indigenous ethnic groups of Taiwan as portrayed in
Japanese-period maps, ca. 1935   8
3. Taiwan, the Ryūkyū Islands, the Langqiao Peninsula, Mudan, and Satsuma,
in East Asia   51
4. The Langqiao Peninsula, ca. 1874   53
5. Administrative centers, contact zones, and political boundaries in Taiwan,
1874–1945  58
6. Dakekan, Jiaobanshan, Quchi, Wulai, and Rimogan, ca. 1910   80

TA B L E S

1. Military encounters between the colonial government and Taiwan

­aborigines, 1896–1909  100
2. Sources of revenue for the colonial administration, 1897–1907  101
 Ac knowle d gme n ts

The research for this book began in an undergraduate seminar room, decades ago.
Since then, I have racked up a record of personal and scholarly debts dispropor-
tionate to the modest results achieved. These begin with my history professors at
the University of Wisconsin, Alfred McCoy, Kathryn Green, Jean Boydston, John
Sharpless, Kenneth Sacks, and Jürgen Herbst. Thanks, Al, for encouraging me to
think big.
Several mentors who became friends at the University of Minnesota shaped
this project and have earned my eternal gratitude. Advisors Byron K. Marshall
and David W. Noble steered a comparative dissertation project to completion
and spent countless hours counseling me on matters profound and trivial. Ann
Waltner, Ted Farmer, Jeani O’Brien, Steven Ruggles, David Lipset, Russ Menard,
Jennifer Downs, Chris Isett, and Wang Liping were all generous with their time,
energy, and ideas. My fellow graduate students Sean Condon, Yonglin Jiang, Joe
Dennis, Yuichirō Onishi, David Hacker, David Ryden, Matthew Mulcahy, Rachel
Martin, Jennifer Spear, Jennifer Turnham, Martin Winchester, and Jon Davidann
were the best classmates and extended family a graduate student could hope for.
Jeff Sommers is a permanent friend and colleague from before and after graduate
school; his outlook and insight have shaped this book profoundly.
Indulgent hosts, true friends, and brilliant associates have promoted my
research in Japan. First and foremost, Vicky Muehleisen, Yamamoto Masashi, and
Jerome Young put me up in Tokyo more times than I can recall. Fumu Susumu at
Kyoto University and Sasaki Takashi at Doshisha University sponsored my early
research and opened their doors and offices to a neophyte. Arisue Ken at Keio

xiii
xiv     Acknowledgments

University hosted me for a year of sabbatical research at Keio University and took
the time to introduce me to everyone who was anyone in my area of research.
I emphatically thank Professor Kishi Toshihiko at the Center for Integrated
Area Studies at Kyoto University for a residency, his friendship and guidance,
myriad introductions, field trips, and workshops. Without Kishi-sensei’s enthusi-
astic support, this book could not have been completed. Thanks to Professor Hara
Shōichirō, director of the center, for making CIAS like a home away from home.
My research in Taiwan has been utterly dependent upon the friendship and assis-
tance of several scholars and friends. University of Minnesota classmate Peter Kang
(Kang Pei-te) has hosted me, shown me around, and provided the foundation for
my investigations. Chen Wei-chi (Tan Uiti) and Chang Lung-chih have been superb
teachers, comrades, and loyal supporters from the start. John Shufelt is a true friend,
intellectual compatriot, host, and fellow explorer of Hengchun Peninsula. Douglas
Fix is an indispensable mentor and model friend; Doug has forgotten more than
I’ll ever know about Taiwan. Caroline Hui-yu Ts’ai has written the most incisive
and thoroughly researched institutional, legal, and cultural history of Japanese rule
in Taiwan; more than that, she took great pains to host me at Academia Sinica’s
Institute for Taiwan History to finish research for this book. Paul Katz is a master
of Taiwanese social history and religious studies and an endless supplier of shipped
documents, connectivity, and nomunication. Professor Clare Huang (Huang Chih-
huei) has taken me to field sites, introduced me to graduate students, patiently
explained the nuances of the difficult postcolonial situation in Taiwan, and shared
rare historical materials; she has changed the course of this research for the bet-
ter. I also thank anthropologists of Taiwan Hu Chia-yu, Kuan Da-wei, Fred Chiu,
Kerim Friedman, Scott Simon, Wang Peng-hui, Aho Batu, and Geoffrey Voorhees
for camaraderie, mentorship, and vast repositories of knowledge. Wu Micha, Wang
Ying-fen, Sandra Jiang, Lin Maleveleve, and Yayut Chen have extended many cour-
tesies and made this project fun. Chen Yi-fang of the Puli Municipal Library opened
new doors and contributed wisdom, energy, and enthusiasm.
From the Shung Ye Japanese Research Group on Indigenous Peoples of Taiwan,
anthropologist, historian of colonial ethnography, and über-senpai Kasahara
Masaharu has inspired, mentored, and supported my work for two decades. Thank
you so much, Kasahara-sensei! I also thank Nobayashi Atsushi for hosting me at
the National Ethnological Museum in Osaka and trying to keep me in the loop.
Professors Shimizu Jun, Miyaoka Maoko, Ōhama Ikuko, Yamamoto Yoshimi, and
Tsuchida Shigeru have shared their networks, knowledge, and research in the true
spirit of collegiality. Nagasako Minako at the Gakushūin Daigaku Archives pro-
vided access to collections and bibliographic support.
Thank you very much to Julia Adeney Thomas, Kirsten Ziomek, Janice
Matsumura, Prasenjit Duara, Dennis Washburn, Chris Hanscom, Murray
Rubinstein, Andrew Morris, Kenneth Ruoff, Hyung Il Pai, Rob Tierney, Alexis
 Acknowledgments    xv

Dudden, Barak Kushner, Ann Heylen, David Ambaras, Kate McDonald, John
Shepherd, Robert Eskildsen, Joseph Allen, Tony Tavares, Emma Teng, Seiji
Shirane, Matthew Fraleigh, Adam Clulow, Sabine Frühstück, and Austin Parks
for the invitations, provocations, encouragement, panels, letters, edits, and shared
documents.
My colleagues in the History Department at Lafayette College have been there
for me in numerous ways over many years. Special thanks to Tammy Yeakel,
Deborah Rosen, Josh Sanborn, Rebekah Pite, DC Jackson, Bob Weiner, Don
Miller, Andrew Fix, Rachel Goshgarian, Jeremy Zallen, and Christopher Lee
for intellectual community and a place to call home. I thank my Asian Studies
Program compatriots: Seo-Hyun Park for brainstorms and crucial bibliography,
and Li Yang, Robin Rinehart, Ingrid Furniss, Il Hyun Cho, and David Stifel for
their intellectual companionship and ongoing commitment to my professional
and personal development.
I am fortunate to have accomplished and unselfish colleagues in anthropol-
ogy at Lafayette College. Thanks, Andrea Smith, Bill Bissell, Wendy Wilson-Fall,
and Rob Blunt, for not allowing me to caricature your discipline. EXCEL Scholars
Wu Haotian, Linda Yu, Li Guo, Sun Xiaofei, Sharon Chen, and Ning Jing have
compiled tables, abstracted articles, and translated Chinese-language and French
documents into English for me over the years.
Digital Scholarship Services colleagues Eric Luhrs, Paul Miller, Charlotte
Nunes, James Griffin III, John Clark, and Michaela Kelly have built databases, con-
structed maps, captured images, hosted workshops, and provided more support
than could be reasonably expected. John Clark created the six beautiful maps for
this book. Neil McElroy, Diane Shaw, Elaine Stomber, Terese Heidenwolf, Lijuan
Xu, Pam Murray, and Karen Haduck at Skillman Library unstintingly supported
this project with access to funds, images, texts, databases, and interlibrary loan
materials, not to mention accessioning and processing all manner of ephemera
and curios. They have spoiled me rotten.
Matsuda Kyōko’s pioneering research in the history of Japan’s colonial anthro-
pology in Taiwan has been an inspiration. I am also heavily indebted to Kitamura
Kae, Kondō Masami, Kojima Rei’itsu, Matsuoka Tadasu, Yamaji Katsuhiko, and
Matsuda Yoshirō for conceptualizing and documenting the history of indigenous-
Japanese relations with admirable depth, nuance, and creativity. These scholars
have set a high standard for this field. In addition, Chou Wan-yao, Wu Rwei-ren,
Ka Chih-ming, Wang Tay-sheng, and Yao Jen-to have produced masterworks in
the historical sociology of Taiwan; even where I’ve neglected to cite them, their
ideas permeate this book.
Thanks to Donald and Michiko Rupnow for the picture postcard that adorns
the cover of this book, and for supporting this research with other rare and won-
derful images. Michael Lewis, Elizabeth and Anne Warner, Richard Mammana,
xvi     Acknowledgments

Lin Shuchin, and David Woodsworth have kindly donated, lent, or provided access
to their private collections. Without their generosity and public-spiritedness, this
book would not have been possible.
Sections of “ ‘Gaining Trust and Friendship’ in Aborigine Country: Diplomacy,
Drinking, and Debauchery on Japan’s Southern Frontier,” Social Science Japan
Journal 6, no. 1 (April 2003): 77–96, appear in chapter 1; “Cultural Brokerage and
Interethnic Marriage in Colonial Taiwan: Japanese Subalterns and Their Aborigine
Wives, 1895–1930,” Journal of Asian Studies 64, no. 2 (May 2005): 323–60, in ­chapter 2;
“Tangled Up in Red: Textiles, Trading Posts, and the Emergence of Indigenous
Modernity in Japanese Taiwan,” in Andrew Morris, ed., Japanese Taiwan: Colonial
Rule and Its Contested Legacy (London and New York: Bloomsbury Press, 2015),
49–74, in chapter 3; and “Playing the Race Card in Japanese-Governed Taiwan, or:
Anthropometric Photographs as ‘Shape-Shifting Jokers,’ ” in Christopher Hanscom
and Dennis Washburn, eds., The Affect of Difference: Representations of Race in East
Asian Empire (Honolulu: University of Hawai‘i Press, 2016), 38–80, in chapter 4.
The editors and readers of these pieces offered valuable advice and suggestions.
Research for this book was funded by fellowships from the Japan Society for the
Promotion of Science, the National Endowment for the Humanities, the Taiwan
Ministry of Foreign Affairs, the Lafayette College Provost’s Office, and the Andrew
Mellon Foundation. The Friends of Skillman Library and the Lafayette College
Academic Research Committee generously supported the publication of this book.
Thank you, Asia Pacific Modern series editor Tak Fujitani and Professor Jordan
Sand of Georgetown University, for encouraging me to write a book. Tak offered
counsel, advice, and support and deserves many thanks for reading this manu-
script in its entirety, and parts of it repeatedly. Jordan has been an unselfish mentor
and colleague. The book was improved thanks to their kind attention. Any errors
of fact and interpretation that remain, despite all of this help, are wholly my own.
At University of California Press, senior editor Reed Malcolm and production
coordinator Zuha Khan are amazing. Thanks for your timely responses to queries
large and small! Jody Hanson designed the beautiful cover for this book, providing
a much-needed lift as I came down the homestretch. Also, my heartfelt thanks to
copy editor Erica Soon Olsen and production editor Francisco Reinking for their
heroic labors in seeing this project through to completion.
Finally, my parents, David and Mary Barclay; Keiko Ikegami; and dearly
departed Papa Ikegami cannot know the depth of my gratitude. Uncle Bill, Uncle
Akita, brother John, and sister Barbara, thank you so much for a lifetime of inspi-
ration and always being there.
I dedicate this book to my wife, Naoko. She and our daughter, Megumi, have
lived this project without complaint and have supported it in more ways than can
be expressed in writing.
 N ot e on Translite ration an d Tra n sl ation

Japanese-language words in the text are transliterated in the modified Hepburn sys-
tem, except for the place-names Tokyo and Osaka. The default system for Chinese-
language words is Hanyu Pinyin. However, Taiwanese personal and place-names
that are commonly transliterated in Wade-Giles or other non-Pinyin systems have
been left as I have found them. There is no standard system for transliterating
Austronesian personal names. Where possible, I have followed usage from Chou
Wan-yao’s New Illustrated History of Taiwan. Korean words are romanized in the
reformed system. The McCune-Reischauer system is used for names of authors
and publications that are cataloged under this system.
All translations from Japanese sources are by the author unless otherwise indi-
cated. Chinese translations are the author’s adaptations of translations by research
assistants Wu Haotian, Linda Yu, Li Guo, Sun Xiaofei, and Ning Jing.

xvii
 Introduction
Empires and Indigenous Peoples, Global Transformation
and the Limits of International Society

P R O L O G U E : T H E W U SH E R E B E L L IO N A N D
I N D IG E N OU S R E NA I S S A N C E I N TA I WA N

On October 27, 1930, terror visited the small community of Japanese settler-­
expatriates in the picturesque resort town of Wushe, an administrative center
nestled on a plateau in the central mountains of Taiwan.1 On that day, some 300
indigenes led by Mona Ludao raided government arsenals, ambushed isolated
police units, and turned a school assembly into a bloodbath. All told, Mona’s men
killed 134 Japanese nationals by day’s end, many of them butchered with long dag-
gers and beheaded. Alerted by a distressed phone call from an escapee, the Japanese
police apparatus, with backing from military units stationed in Taiwan, responded
with genocidal fury. Aerial bombardment, infantry sweeps, and local mercenaries
killed roughly 1,000 men, women, and children in the ensuing months. A cor-
nered Mona Ludao removed to the countryside and then killed his family and
hanged himself to avoid capture. Subsequently, the Japanese government relocated
the remaining residents of Mona’s village, Mehebu, forever wiping it off the map.2
Over the course of Japanese rule (1895–1945), the Taiwan Government-­General
forcibly relocated hundreds of other hamlets like Mehebu. The invasive and
exploitative policies that provoked Mona and his confederates also eroded pre-
colonial forms of social organization, authority, and ritual life among Taiwan’s
indigenes. As it severed bonds between indigenes and their lands, in addition to
prohibiting or reforming folkways it deemed injurious to its civilizing mission,
the government-general nonetheless laid the groundwork for the emergence of
Taiwan Indigenous Peoples as a conscious and agentive historical formation. By
arresting the diffusion of Chinese language and customs into Taiwan’s interior,
1
2     Introduction

restricting geographic mobility across the so-called “Savage Border,” dividing the
colony into normally and specially administered zones, and sanctioning a bat-
tery of projects in top-down ethnogenesis, the government-general inscribed a
nearly indelible “Indigenous Territory” on the political map of Taiwan over the
five decades of its existence.
This book will argue that successive, overlapping instantiations of state power’s
negative and positive modalities precipitated the formation of modern indigenous
political identity in colonial Taiwan. This process paralleled other nationalist
awakenings forged in the crucible of foreign occupation. As state functionaries
smashed idols, compelled assimilation, and asserted the authority of a central
government, their fellow nationals reified, commodified, and preserved the mate-
rial, cultural, and territorial expressions of native distinction. These Janus-faced
vectors of state building can be found wherever governments targeted citizenries,
imperial subjects, or marginalized out-groups for inclusion into a new kind of
national political space. Applying these axioms to the case of Taiwan Indigenous
Peoples under Japanese colonial rule, Outcasts of Empire argues that the process
Ronald Niezen dubs “indigenization” is a historical concomitant of competitive
nation building in the age of high imperialism (1870s–1910s).
Rightly emphasizing the importance of transnational activist circuits, global
NGOs, and the increased salience of international rights conventions, Niezen and
others consider the decades following 1960 the incubation period for “interna-
tional indigenism.”3 Laura R. Graham and H. Glenn Penny stress how indigeneity
“emerged as a legal and juridical category during the Cold War era” in response
to “growing concerns about environmental degradation during the twentieth cen-
tury together with the emergence of human rights discourses. . . . ”4 Writing about
the Taiwan case study explored in this book, Wang Fu-chang asserts that indig-
enous political consciousness is a decidedly recent arrival, erupting in its current
form in the 1980s.5
While recognizing the importance of the movements of the 1960s and beyond
for indigenous cultural survival in the twenty-first century, this book argues that
the early twentieth century is a better place to look for the systemic wellsprings of
indigenism.6 Rather than viewing indigenism as a postwar development enabled
by a more or less functioning international system, Outcasts of Empire suggests
that nationality, internationalism, and indigenism were mutually constituted for-
mations, rather than sequentially occurring phenomena.
The pages that follow examine the politics, economics, and cultural move-
ments that informed the Japanese colonial state’s partitioning of Taiwan’s indig-
enous homelands into a special zone of administration known as the Aborigine
Territory. The administrative bifurcation of Taiwan began as an expedient measure
in the 1890s, reflecting the dependence of the Taiwan Government-General on
Qing precedents and straitened colonial budgets. By the 1920s and 1930s, how-
ever, the peoples today known as Taiwan Indigenous Peoples7 were cast for good
 Introduction    3

beyond the bounds of the colonial state’s disciplinary apparatus. The so-called
Takasagozoku (Formosan Aborigines) were accorded a special status as imperial
subjects because they were believed to lack the economic competence to thrive
in the colony’s “regularly administered territories.”8 In a more positive sense,
indigenes were invested with a cultural authenticity that marked them as avatars
of prelapsarian Taiwan antedating Chinese immigration, based in part on high
Japanese appraisals of Austronesian cultural production.9 From the 1930s onward,
the distinctiveness of indigenes as non-Han Taiwanese was elaborated and pro-
moted by the state, the tourism industry, and intellectuals, laying the groundwork
for the successor Nationalist Party government of Taiwan (Guomindang or GMD)
to rule the island as an ethnically bifurcated political field.10
The deterritorializing and reterritorializing operations that underwrote the
emergence of Taiwan Indigenous Peoples during the period of Japanese colo-
nial rule had locally distinctive contours.11 But these interrelated processes were
embedded in a global political economy dominated by a capitalist business cycle
and international competition. In East Asia, the Japanese state refracted these
transnational forces throughout its formal and informal empires. A parallel and
instructive set of events in neighboring Korea illustrates this point.
In 1919, the Japanese state brutally suppressed a Korean uprising known as the
March 1 Movement. That year, across the peninsula, around one million Koreans
loudly protested the draconian administrations of governors general Terauchi
Masatake and Hasegawa Yoshimichi (1910–1919) on the occasion of former king
Gojong’s funeral. As was the case with the Wushe uprising of 1930 in Taiwan, the
magnitude and vehemence of the protests were taken as a negative verdict on
Japanese rule. The savagery of Japan’s suppression of the uprising, which may have
taken 7,500 Korean lives, became a source of national embarrassment. The sense
that colonial rule should rest on more than naked force, and the awareness that
the world was watching, impelled the Japanese state to embark on reforms that
emphasized co-optation, the active support of Korean elites, and abolition of the
most violent and hated forms of colonial police tactics, such as summary punish-
ment by flogging.12
During the 1920s, the Korean Government-General launched a series of poli-
cies known as “cultural rule” in response to the March 1 debacle. As part of a larger
program to legitimate itself, Japan’s official stance toward Korean literature, archi-
tecture, music, and other cultural forms took a preservationist turn that tempered
enthusiasm for the fruits of Korean ethnic genius with a wariness of insubordina-
tion and a long-standing belief that Koreans were developmentally laggard. The
softening of the government’s posture and policies entailed neither the implemen-
tation of a culturally relativist agenda nor the abandonment of the core principles
of racial denigration. Nonetheless, Saitō Makoto’s “cultural rule” policy repre-
sented a sea change, and it set into motion a series of reforms that laid bare the
contradictory demands made upon the interwar colonial state.
4     Introduction

On the one hand, state power was ultimately maintained through the threat
of force and justified by a theory of Japanese racial superiority. On the other, the
colonial state sought to attain hegemony through the politics of inclusion, which
brought in its train practices that were conducive to the production of modern
Korean subjects.13 Henry Em summarizes the paradoxical long-term effects of
colonial rule in terms that mirror events in Taiwan:
Thus, contrary to conventional [Korean] nationalist accounts that argue that Japa-
nese colonial authorities pursued a consistent and systematic policy of eradicating
Korean identity, we should see that the Japanese colonial state actually endeavored
to produce Koreans as subjects—subjects in the sense of being under the authority
of the Japanese emperor, and in the sense of having a separate . . . subjectivity. . . .
It was in this sense that Japanese colonialism was ‘constructive’ for both the ­colonizer
and colonized. . . . Coercion, prohibition, and censorship, then were not the only (or
even primary) forms through which colonial power was exercised. . . . there was a steady
proliferation of discourses concerning Korean identity emanating from the J­apanese
­colonial state itself—including studies of Korean history, geography, language, customs,
religion, music, art—in almost immeasurable accumulated detail. . . . For the Japanese
colonial state, the goal of exploiting Korea and using it for its strategic ends went hand
in hand with the work of transforming peasants into Koreans, or ‘Chōsenjin.’14

Two parallels are in evidence here. First of all, well-coordinated attacks on

Japanese state power (March 1 in Korea and Wushe in Taiwan), followed by clumsy
and disproportionate responses (the open firing on civilians in Korea, the aerial
bombardments in Taiwan), actuated regime change. In Taiwan, the heated debates
surrounding the Wushe Rebellion pitted Japan’s opposition Seiyūkai (Friends of
the Constitution) against Governor-General Ishizuka Eizō, who was allied with
the ruling Minseitō (Popular Government Party). The Seiyūkai capitalized on
the Taiwan Government-General’s incompetence to call for the resignation of
Ishizuka, who actually stepped down along with his inner circle, a reshuffling rem-
iniscent of Saitō Makoto’s ascension to the governor-generalship of Korea in 1919.15
Again echoing events in Korea, the cowed successor administration in Taiwan
called for a renovation of “Aborigine Administration” and a shift from rule by
naked force and intimidation to government by co-optation and delegation.
Importantly, for our purposes, Taiwanese highlanders were thereafter governed
as members of ethnic groups, whose cultural, political, and economic distinction
from the rest of Taiwan was selectively preserved, with certain elements even cel-
ebrated by colonial administrators, metropolitan voters, and consumers in Taiwan
and Japan alike.16 As was the case for Japanese cultural rule in 1920s Korea, the
new era of aborigine administration proved compatible with the emergence of a
discourse on ethnic integrity, one that overrode localisms. The artifacts and struc-
tures that coalesced during this period would then resurface in the postcolonial
era in the form of indigenous ethnonationalism.17
 Introduction    5

In contrast to Korea, public expressions of indigenous patriotism were

s­uppressed in Taiwan for over four decades after the Japanese Empire crum-
bled in 1945.18 The 1.5 million migrants to Taiwan in the Chinese Nationalist
Party (GMD) exodus represented yet another wave of colonization for Taiwan’s
majority population.19 During a long stretch of one-party rule under martial
law (1949–87), GMD-sanctioned history excluded discussions of indigenes as
autochthons because it regarded Taiwanese history as a regional variant of main-
land China’s.20
After martial law was lifted in 1987, politicians put distance between them-
selves and the GMD by supporting an indigenous cultural renaissance to signal
the island’s distinctiveness from the mainland.21 On a parallel track, the founding
of the Alliance of Taiwan Aborigines in 1984 ushered in a wave of organized indig-
enous activism.22 Thereafter, coordinated action between Han Taiwanese national-
ists and Indigenous activists produced a number of political and cultural reforms
aimed at promoting a measure of autonomy and correcting the most egregious
forms of public denigration. As a result, the notions that non-Han peoples are the
island’s original inhabitants and that Taiwan’s multiracial composition should be
celebrated rather than overcome are mainstream political positions23—much as
they were in the 1930s, when Japan ruled the island.
Within this broader context of renaissance and revival, Puli-based freelance
writer Deng Xiangyang’s energetically researched oral and documentary histories
articulate a local perspective on the Wushe uprising that posits indigenes as anti-
colonial heroes and avatars of an authentic, pre-Chinese Taiwanese past. Deng’s
books draw on biographical and family histories and photographs culled from his
local network of Sediq acquaintances.24 Along with graphic artist Qiu Ruolong,
who also has extensive contacts and family relations in the Sediq community, Deng
coproduced a television drama and children’s book about the Wushe Rebellion.
Qiu and Deng have popularized the heroic suffering of the Sediq people, the harsh
labor conditions and sexual harassment that contributed to the revolt, and the
brutality of the Japanese counterattacks. All of these themes were subsequently
dramatized in the blockbuster John Woo production Seediq Bale (2011), a feature-
length film that recounts the story of the rebellion in romantic hues that recall Last
of the Mohicans and Dances with Wolves.25
As I watched and rewatched this film, I was struck by its fidelity to the Japanese
inquest reports into the causes of the rebellion. The main characters, the key
scenes, and the plot structure are immediately recognizable to anyone who has
sifted through the documentation generated by the rebellion. Insofar as some
Japanese characters are made out to be racist buffoons deserving of grisly deaths,
this film can be considered anticolonial. At the same time, its deep engagement
with a reservoir of colonial-era tropes, documents, and narrative structures high-
lights the entanglement of colonialism and postcolonial nationalism that marks
6     Introduction

the Korean experience, shown here in a new context: the making of an indigenous
people.
While there were similarities, as noted above, the complex process by which
residents of Mehebu and Paaran (see map 1) became Sediq was different from
the trajectory that saw natives of Seoul, Gyeongju, and Pyongyang transformed
into Chōsenjin (people of Joseon, or Koreans). In the case of Taiwan Indigenous
Peoples, the translocal, subimperial, and putatively organic identities fostered
under the government-general’s variant of cultural rule were not Taiwanese, per
se, but by turns Indigenous Formosan (Takasagozoku) or attached to particular
ethnolinguistic groups (Amis, Bunun, Paiwan, Atayal, Tsou, Rukai, Saisiyat, and
Yami) (see map 2).
For example, in 1930, Mona Ludao and his followers appeared in official docu-
ments, journalism, and commercial publications as “savages,” “barbarians,” or
“Formosan Tribes” (banjin, seibanjin, banzoku).26 In many respects, they were
governed as such: policy before 1930 emphasized their backwardness, mainly by
excluding them from the tax base due to purported economic incompetence. The
translocal identifier banjin ascribed little importance to matters of ethnic identity
and was, in fact, symptomatic of a pre–Wushe Rebellion approach to governance
that paid scant attention to subject formation.27
In the press, in government statistics, in police records, and in ethnological
writing, indigenes were also identified as members of particular units called sha
in Japanese (Mehebu and Paaran, for example). The sha were units of governance
pegged to residential patterns, although they did not necessarily reflect local con-
ceptions of territoriality and sovereignty. Rather, the category sha (in Chinese, she)
was imposed by the Qing, long before the Japanese arrived, as a blanket term for
any indigenous settlement or cluster of hamlets.28 Like the banjin designator, affili-
ation with a sha did not confer ethnic or cultural status upon the governed.
Terminology anchored in derivatives of the terms ban and sha suggested con-
tinuity from Qing times and a relative disinterest in indigenous interiority. On
the other hand, as early as 1898, Japanese ethnologists began to classify residents
of Mona’s hometown of Mehebu as Atayals. This neologism originated with Inō
Kanori and signaled a different way of imagining Taiwan’s non-Han population(s).
The term Atayal first appeared in Japanese documents in 1896 to identify an ethnic
group noted for facial tattooing, a common language that spanned several water-
sheds and valleys (and sha), and the production of brilliant red textiles.29 The term
Atayal, which connoted membership in a culture-bearing ethnos, rarely surfaced
in policy-making circles during the first two decades of colonial rule. From early
on, however, the term was inscribed in an academic counterdiscourse, as exempli-
fied by a color-coded map. The map’s novel subethnic components—territories
for the Atayal, Bunun, Tsou, Amis, Paiwan, Puyuma, and Tsarisen peoples—­
overwrote Qing-period cartographic voids. This architectonic prefigured today’s
officially sanctioned view of Taiwanese multiculturalism (see figures 1 and 2).
 ¯ 7DLSHL !
(
!
(
6KHQNHQJ

( ;LQGLDQ
!
!
(
'DNHNDQ 4XFKL
!
( :XODL
!
(
-LDREDQVKDQ
!
(
!
( <LODQ
Saisiyat 5LPRJDQ !
(

;LWRX

1DQ¶DR
Atayal

0DOHSD
0DVLWREDQ ! ( ;LQFKHQJ
( 7UXNX
! !
(
!
( 6HGLT
:XVKH ! ( 7RGD
( !
(!
! ( 0HKHEX
+XDOLDQ
3DDODQ !
(
!
(
3XOL
Bunun Amis
 0LOHV

Map 1. The major indigenous ethnic groups of northern Taiwan, with the Atayal and Sediq
settlements most frequently mentioned in this book. The Sediq territory reflects the demarca-
tions of Japanese official surveys in the second decade of the twentieth century. Today, much of
the territory labeled Sediq by the Japanese is now considered Truku territory.
 ¯
Saisiyat

Atayal

Sediq

Tsou
Amis
Bunun

Rukai
Puyuma

Paiwan

Yami

25 Miles

Map 2. An overview of the major indigenous ethnic groups of Taiwan, as portrayed in

Japanese-period maps, ca. 1935. As this book argues, the shapes of these territories, their names,
and their numbers have been historically contingent.
 Introduction    9

ATAYAL = Ethnic Group (banzoku)

Xitou = Linguistic/Ethnic Group (ban)
Palaan = Indigenous Town (sha)

Southern Tribes Northern Tribes

SAISIYAT

PAIWAN PUYUMA BUNUN TRUKU SEDIQ ATAYAL

Lower Tgdaya Toda Truku Quchi qian- Xitou Nan’ao

Hengchun (Wushe) Dakekan
Wulai
Paalan
Raga Jiaobanshan
Hōgō
Tuilasok Beinan Gantaban Rimogan
Sabaree Mehebu Masitoban
Koalut
Mudan
Kuskus

figure 1. Japanese colonial-period ethnonyms for Taiwan Indigenous Peoples.

During the colonial period, the ethnic designator Sediq that marks Mona Ludao
and his descendants was also slow to catch on, even after the Wushe uprising of
1930.30 Nonetheless, the political and cultural salience of labels such as Atayal,
under which the Sediq were subsumed at that time, increased in the 1930s. TGG-
sponsored indigenous youth corps, a new locus of political power, were organized
around these labels of identity and difference.31 In addition, museum collections,
censuses, and language dictionaries inscribed these categories for reactivation
decades later.32
There are obvious differences in scale between modern Korean nationalism
and Taiwan indigenous renaissance. The Korean Peninsula is populated by well
over sixty-five million people, while the Sediq lands of contemporary Taiwan
claim about ten thousand; Taiwan’s sixteen recognized indigenous ethnic groups
total roughly three hundred thousand souls. But for the purposes of this study, the
more important issue is how these two former areas of the Japanese Empire have
diverged, despite their common experiences as test cases for the efficacy of cul-
tural rule, referred to above. Unlike Koreans who reside on the peninsula, Taiwan’s
indigenous peoples cannot claim a nation-state to express their shared heritage
and territorial distinction in the hard currency of sovereignty. At the same time—
and this separates them from ethnic minorities who cannot imagine themselves
as indigenous—conditions are such for Taiwan Indigenous Peoples that territorial
sovereignty in one form or another is a plausible, if not desirable, way forward.
figure 2. Inō Kanori’s ethnic map of Taiwan’s indigenous territory, 1898. The colored zones
represent indigenous ethnic groups. The white areas represent nonindigenous (Han) Taiwan. A
broken line indicates the borderline between the two. This map was published in 1900 as part
of a government report: Inō Kanori and Awano Dennojō, Taiwan banjin jijō (Taipei: Taiwan
sōtokufu minseibu bunshoka, 1900), n.p.
 Introduction    11

T H E D E C L I N E O F NAT I V E S A N D T H E E M E R G E N C E O F
I N D IG E N OU S P E O P L E S I N WO R L D H I ST O RY
American and Japanese successes in nation building during the 1870s and 1880s
were not isolated events. After 1860, Germany and Italy also consolidated, and they,
along with France, put into place new central governments. . . . A handful of nations
put together the skills and aggressiveness to create concentrated industrial plants
whose products and profits translated into military power. The resulting new impe-
rialism of the post-1860 years redrew the maps of Africa and Asia.—Walter LaFeber33
As there are no unoccupied territories—that is, territories that do not belong to [a]
state . . . [T]he characteristic feature of the period under review is the final partitioning
of the globe—final, not in the sense that repartition is impossible; on the contrary,
repartitions are possible and inevitable—but in the sense that the colonial policy of
the capitalist countries has completed the seizure of the unoccupied territories on our
planet. For the first time the world is completely divided up, so that in the future only
redivision is possible, i.e., territories can only pass from one “owner” to another, instead
of passing as ownerless territory to an owner.—V. I. Lenin34

In their recent intervention into the field of international relations, political sci-
entists Barry Buzan and George Lawson make a strong case for the nineteenth-
century origins of the “full international system.”35 They convincingly argue that a
“global transformation” was occasioned by the confluence of shifts in the histori-
cally concatenated domains of production, state building, and ideology. Following
historian Jürgen Osterhammel, Buzan and Lawson point out that today’s inter-
national system is not a creature of Europe’s landmark Treaties of Westphalia
(1648) or Versailles (1919), but is rather a product of nation-state–sponsored
industrial capitalism’s global impact as an integrative and disintegrative force.36
Osterhammel also views the nineteenth century as a world-historical pivot, while
decentering the story of global modernity with a conscious move away from unre-
flexive Eurocentrism. Osterhammel and Buzan and Lawson take inspiration from
C. A. Bayly in their regard for ideation and social formations as more than the epi-
phenomena of high politics and industrial progress.37 But for all of their breadth,
erudition, and nuance, none of these magisterial syntheses of nineteenth-century
global political economy and history has accounted for a persistent phenomenon
that was part and parcel of the transformations they so ably map. I refer here to the
emergence of indigenous peoples, which is the other side of the coin of the birth of
the modern nation-state system.
As Lenin asserted a century ago (see epigraph to this section), a handful of
nation-states managed to write their sovereignty over the earth’s surface during
Osterhammel’s, Buzan and Lawson’s, and Bayly’s period of global transformation.
This feat was accomplished by harnessing the power of capitalism to a novel form
of intra- and interstate sovereignty. In the emergent international system, state
sovereignty was imagined as distributed evenly within clearly defined and lim-
ited national borders. Unlike their early-modern dynastic-state predecessors, with
12     Introduction

their relatively low state capacities and patchworks of graded, plural, and oscillat-
ing sovereignties, the new order was premised on the axiom that no space could
belong simultaneously to multiple sovereigns. Perhaps even more radical was the
axiom that no space on the globe could be excluded from the system.38
Colorful early twentieth-century globes and Mercator projection maps visually
encapsulated the new system’s logic. Formerly unknown white spaces and great
deserts of the world were filled in with various imperial colors. One color ended
where another began. The shapes given to these territories were serially reproduced
and became instantly recognizable logos. Thongchai Winichakul called these enti-
ties “geo-bodies.”39 Today, we understand that these globes and maps expressed
wishful thinking. Surely, the United States had not extended its authority over
thousands of Philippine Islands and their millions of inhabitants with the stroke of
a pen in Paris in 1898.40 So too with the French in Vietnam in 186241 or the Japanese
in Taiwan in 189542 at the conclusion of similar paper agreements. Superficially,
such treaties, accords, and protocols transferred sovereignty from dynastic states
to nation-states. But the notion of “transfer” obscured a thorny problem: dynas-
tic states conceptualized sovereignty in significantly different terms, so the sover-
eignty that was inherited by nation-states was necessarily incomplete, partial, and
unsatisfactory by the standards of “international society”43 as it was being recon-
figured in the nineteenth century.
In the years and even decades following annexation, each imperial nation-state
aspired to reconcile the contradictions of an international system premised on
universal reach but composed of legally plural empire-states riddled with semi-
autonomous enclaves. Through negotiation, warfare, and experiments in applied
colonial science, empire-states attempted to create a world of contiguous imperial
geobodies in the face of deficits of intelligence, rebelling populations, challeng-
ing terrain, and fiscal constraints. As Jane Burbank and Frederick Cooper put it,
“ . . . when territories were taken over, colonized people did not simply fall into
whatever role striving industrialists could imagine for them. Empires still came up
against the limits of their power at the far end of lines of command, where they
had to mobilize conquered communities and find reliable intermediaries—all at a
cost that did not exceed the benefits.”44
Continuing in this vein, these historians point out that the far-flung European
nineteenth-century empires, products and engines of a voracious industrial capi-
talism, were not sui generis entities. The great protagonists of high imperialism
employed and built upon older imperial forms, albeit with more firepower and
competitive edge. For them—and for Lauren Benton, another influential scholar
of comparative imperialism—the ideal type of an international society exhausted
by unitarily sovereign states was only that, an ideal. Giving it too much credence,
as the telos of imperial history or as a dominant formation, belies an exclusive
focus on the utterances of metropolitan visionaries and a disregard for how
empires actually functioned.
 Introduction    13

This book shares the view of Burbank and Cooper and of Benton that empires
are best understood as successors to other empires and that postcolonialism did
not resolve itself into a world of sovereign nation-states based on former colo-
nial boundaries. At the same time, it will argue that hallmarks of nineteenth-
century imperialism—its symbiosis with industrial capitalism, a geostrategic
setting of multipolar competition, and a susceptibility to liberal and anti-­imperial
­inflections—produced a condition I will call indigenous modernity. Taking
Burbank and Cooper’s emphasis on the cost constraints of imperial integration to
heart, Outcasts of Empire will describe a kind of fiscal exhaustion, one that occurred
during the slice of world-time described above, as the ground for the emergence
of indigenous peoples, autonomous regions, and other forms of quasi sovereignty
that mottle the ideally solid surfaces of the international system’s geobodies.45
A particular type of virtuous circle or positive feedback loop lay at the back
of state efforts to call forth disciplined populations, the forerunners of today’s
national citizenries. In colonial Taiwan’s “normally administered” districts, the
government-general recovered its investments on a “big bang” of outlays for
population censuses, land surveys, and a buyout of the old regime’s elite because
subsequent state revenues exceeded the costs of government. These monies were
then plowed back into state building or reinvested in industry.46 This virtuous cir-
cle of discipline begetting increased state budgets begetting governmentality was
broken, however, in the militarily resistant, sparsely populated, topographically
forbidding, and linguistically checkered highlands of Taiwan. In these “specially
administered” territories, natives became indigenous peoples.
In our time, population recovery, courtroom victories, and cultural renaissance
are integral to the story of indigenous peoples. The old tropes of dispossession,
deracination, and endemic poverty remain relevant, but narratives of the “disap-
pearing native” ring false in the twenty-first century. This book starts with the
understanding that indigenous peoples did not become extinct. At its widest aper-
ture, Outcasts of Empire asks how Bayly’s, Osterhammel’s, and Buzan and Lawson’s
models of global transformation might look different if the emergence of indigen-
ism is viewed as an integral component of the international system.
To accomplish this task, this study chronicles the making of an indigenous
people at a spatial and temporal junction of the aforementioned great transforma-
tion. The case study is the emergence of the Atayal, Bunun, Saisiyat, Tsou, Amis,
Paiwan, Rukai, and Puyuma peoples in Taiwan under late Qing and Japanese rule
from the 1870s through the 1940s. By shifting the optic away from interstate com-
petition and the development of a global division of labor between the First and
Third Worlds,47 and focusing instead on the production of bifurcated sovereignty
within nationalized and colonized spaces, this study attempts to explain disjunc-
tions and discontinuities internal to emergent national political formations. The
intensification and extensification of global capitalism,48 it will argue, instantiated
the political geography of bifurcated sovereignty in Taiwan.
14     Introduction

The narrative begins in the 1860s, with a discussion of shipwrecks and ransoms
on Taiwan’s Langqiao Peninsula. On the southern extremity of the Qing Empire,
Langqiao’s political topography presented numerous hurdles to the extension of
the international system.49 Through an analysis of the diplomatic, military, and
institutional solutions implemented to solve these problems, I show how a vision
of world order became globally normative (although it remained aspirational at
best). Subsequent chapters analyze how the Japanese colonial state attempted to
implement the new models of sovereignty throughout Taiwan from the 1890s
through the 1940s.
At a more granular level, Outcasts of Empire focuses on the lives, actions, and
­aspirations of men such as Kondō Katsusaburō (a Japanese colonist) and women
such as Iwan Robao (daughter of a Sediq headman), individuals who left faint
but discernible traces in the documentary record. Jun Uchida’s term “brokers of
empire”50 describes them well, as does Daniel Richter’s sobriquet “cultural brokers.”51
By either definition, these brokers thrived in the twilight era of the legally pluralis-
tic dynastic empires—in our case, the Tokugawa, Qing, Joseon, and Shō dynasties
in Japan, China, Korea, and the Ryūkyū Islands, respectively. Outcasts of Empire
chronicles the shifting structural positions of the headmen, interpreters, trading-
post operators, and trackers who dominated the political economy of the early-
modern hinterlands as they were displaced or repurposed in the new international
order.
As a narrative history that pays due attention to personalities who have yet to
receive scholarly attention, the empirical heart of this study is pericentric.52 The
reconstruction of the particulars of long-obscured frontier history is necessary
because it indicates that the nineteenth century’s reterritorialization project was
more costly and time-consuming than metrocentric histories would have it. At
the same time, the macrohistorical frame cannot be ignored, because interstate
competition propelled successive Japanese and Qing projects in state building on
the edges of Taiwan’s governed spaces. One cannot explain the lavish spending for
these endeavors without putting them into a context of perceived national peril
and the desire to create a world order that could accommodate much higher vol-
umes and velocities of commerce, diplomatic communications, and migration.
I am not the first scholar to argue that indigeneity and the international system
are related historical phenomena. However, the case that their co-creation dates
back to the early twentieth century and that their entanglement commenced as
part of the global movement to make the surface of the inhabited earth coextensive
with national geobodies, has yet to be made. In making this claim, I argue that the
historical process of indigenization was an integral component of the international
system’s emergence. The mechanisms uncovered by a study of their co-creation, I
believe, provide an answer to one of modernity’s big conundrums: why the dream
(or nightmare) of a world exhausted by nationally governed territories is receding
 Introduction    15

farther and farther toward the horizon, despite the exponential increase in state
capacities and communication technologies.
Behind such seemingly innocuous statements as “Ankara spoke with Moscow,
to Washington’s chagrin” lies the conceit that there are people in Ankara, Moscow,
and Washington who can speak for the populations containerized in Turkey,
Russia, and the United States. These metonyms suggest that leaders have the power
to restrain, encourage, and regulate their populations to the extent that agree-
ments between them bind their populations, as well. Our ability to imagine that
this system is operative, even as we are mindful of frictions and imperfections,
presupposes a long history of state making that will be explored in the following
pages. As we shall see, the project was animated by high ideals and mendacious
ideologies, by altruism and greed. One can say that the UN charter presupposes
such a world order, as did the Kellogg-Briand Pact of 1928 outlawing war. At the
same time, the Berlin Conference of 1885, the Taft-Katsura accord of 1905, and the
League of Nations Covenant of 1920 invoked the same ethos to justify collusion
among powerful states to deny sovereignty to colonized peoples.53
This book contributes to the literature on global transformation by foreground-
ing the immense quantities of labor-power required to operationalize the ideal-
typical political units that constituted the international system. For the women
and men of Meiji Japan (1868–1912) who either went to Taiwan or participated in
its colonization as citizens of the metropole, Taiwan’s former Qing frontier was a
vast, complex, unregulated, and challenging hinterland. Its conquest and pacifi-
cation was chronicled in minute detail, because each negotiation, punitive expe-
dition, or survey appeared as a necessary step to consolidate a Japanese empire
that must either “eat or be eaten.” And yet, for all of its intensity, momentum, and
destructive fury, the project ran aground in the central mountains in the year 1915.
The location and timing of the project’s abandonment, I will argue, tells us much
about the patchy sovereignty that still characterizes the earth’s political geography.

S TAT E S , SE T T L E R S , A N D NAT I V E S

Historian C. A. Bayly has labeled native peoples “the losers” in the global trans-
formation. Bayly’s formulation helpfully avoids the anachronistic and circular
term “primitive” and provides clues for identifying global patterns in the configu-
ration of modern political space. Bayly writes, “What can be said is that small
cultural groups which were not centrally involved in intensive peasant commodity
production came under unprecedented political, cultural, and demographic pres-
sure where previously they had been able to bargain with residents of the settled
domains or early representatives of European and American power. The ‘cultural
terms of trade’ moved decisively in favor . . . of all settled societies producing a
permanent agricultural surplus or industrial artifacts.”54
16     Introduction

In our analysis, we will designate Bayly’s “settled societies” as settlers, defined

as social formations involved in “intensive peasant commodity production.”
Following Eric Wolf, we stipulate that settler populations are distinguished from
“lineages, clans, tribes, and chiefdoms”55 by their relationships to states. Settlers
surrender surplus wealth to a state via some form of routinized taxation, whereas
Bayly’s native peoples (hereafter abbreviated natives) inhabited what James Scott
terms the ungoverned peripheries of a world not yet exhausted by national borders
or the direct control of dynastic states.56 As Bayly intimates, demographically pre-
ponderant settlers overwhelmed smaller, relatively mobile populations of natives
by squatting and organizing self-defense in the course of the nineteenth-century
global transformation. More importantly, however, settlers yielded a portion of
their surplus to states, which converted agricultural productivity into organized
military force, at a remove from direct native-settler encounters.
Natives, in contrast, lived under political systems that fissured, subdivided,
and recombined within the limits imposed by the politics of redistribution and
reciprocity, which could never attain the surplus-extracting capabilities of states.57
States, not settlers, turned geographically indeterminate margins, borderlands,
and frontiers into demarcated and permanently defended political boundary lines.
The dynamic of this tripartite state-settler-native relationship, then, is a central
plotline in the delineation of the spaces of centralized state sovereignty in world
history.
As a third party, the state could either mobilize agricultural surplus to restrain
land-hungry settlers or deploy it to advance settler interests vis-à-vis natives,
depending on a host of variables. One should not understand this process as an
uninterrupted, unidirectional march, as depicted in Frederick Jackson Turner’s
frontier thesis, with the state an assumed ally of the tax- or tribute-paying settler.
As Richard White has noted, based on the North American example, the state’s
intervention on behalf of settlers was often determined by the relative value cen-
tral administrators placed on alliances with native peoples.58 Moreover, when mul-
tiple empires were in competition, natives could leverage imperial backing against
settlers. But as John Shepherd has shown regarding the Qing frontier in Taiwan,
state support for settlers was also determined by a crude cost-benefit analysis. If
the military expenses required for an aggressive settlement policy were justified by
the projected increase in land revenues, the state would back the settlers.59
As White and other scholars have shown regarding the North American natives
under British and U.S. colonial rule, it was when states abandoned native peoples
in favor of taxpayers (actual or potential), in light of fiscal, geostrategic, or even
electoral pressures, that nonstate spaces were likely to enter the realm of centrally
administered territory. Broadly speaking, this dynamic of state expansion was
reversible in the early modern period but became much less so with the advent
of international relations undergirded by capitalist relations of production. As
the increased volume and pace of global trade called forth greater regulation of
 Introduction    17

ungoverned territory by centralized administrations, states sided with settlers to

turn the tide against the natives.
An observer at century’s turn might well have agreed with Lenin that states, by
then, had carved up the earth’s occupied territory. Bayly estimates that “only” four to
eight million people constituted the world population of natives circa 1900, a num-
ber seemingly too small to stop the march of progress. But if the large portion of the
earth’s surface occupied by natives on the eve of World War I is taken into account,
Lenin’s statement appears premature. The remaining spaces of native autonomy
constituted formidable obstacles to state-building visionaries and the settlers who
acted as shock troops for national territorial expansion. To be sure, during the global
transformation, vast tracts of the most productive native land were incorporated
under the authority of nation-states and dynasties through dispossession, slaughter,
or epidemiological catastrophe. Nonetheless, large areas remained administered by
or in the name of native peoples beyond the spaces of national sovereignty.

L E G A L C E N T R A L I SM A N D L E G A L P LU R A L I SM

The Japanese state in Taiwan aimed to effect “the transformation of differenti-

ated and layered political order into an homogenous space”60 from the early 1900s
through about 1915, at the height of a historical epoch when “territory rather than
status and allegiance increasingly defined the jurisdiction of the state.”61 Its pre-
decessor state in Taiwan, the Qing, exemplified a contrasting regime of “weak” or
multicentric “legal pluralism.” Throughout the Qing Empire, heterogeneous com-
munities and ranked status groups stood in differentiated legal relationships to the
apical center of authority in Beijing. Implicit in the notion of multicentric legal
pluralism is the possibility that sovereignty can be graded, and even diminished,
at the margins of a polity. In contrast, nation-states ideally impose equal measures
of sovereignty over the whole surface of their territories to facilitate defense, tax
collection, conscription, economic development, and the mobility of labor, among
other desiderata. At the least, expanding nineteenth-century empires attempted
to shut down competing sources of authority and to bring plural systems under a
hierarchical state-centered umbrella.
While the implementation of legal centralism was a costly affair, it was not a
project abandoned lightly in the arena of competitive imperialism. At that time,
border territories that appeared “ownerless,” because they operated by nonstate
logics, were subject to occupation by other states by the lights of Western interna-
tional law. At the height of social Darwinism, Japanese leaders often felt compelled
by geostrategic considerations to extend or consolidate territorial sovereignty in
order to protect Japan’s flanks, under time pressure and with finite resources. The
end result of this conjuncture in Taiwan was its current bifurcated sovereignty.
Bifurcated sovereignty is a descendant of weak legal pluralism in roughly the
same way that indigenous peoples are descendants of native peoples. In other
18     Introduction

words, the processes described by revisionist historian of American foreign rela-

tions Walter LaFeber and influential theorist of empire V. I. Lenin as the final push
toward state ownership of the earth’s inhabited space was halted where native
peoples could not be brought within imperialism’s centrally administered ter-
ritories. Today these peoples are known as indigenous, and their territories are
administered under different sets of rules from the rest of the nation’s spaces.
That bifurcated sovereignty is here to stay, despite the overwhelming pressure for
administrative and legal integration brought to bear by industrial capitalism and
international competition, requires historical explanation.
Thongchai Winichakul has described modern, Mercator-projection map-logos
as geobodies. The geobody has a fixed shape—its borders are no longer expand-
ing (or oscillating)—and its boundaries are clearly delineated. These two features
distinguish it from most dynastic realms of the prenational era. Thongchai’s case,
Thailand’s geobody, was formed in the context of international competition and
technology transfer that characterized the long nineteenth century. To preempt
encroachment by aggressive Western powers, who considered the ability to carto-
graphically represent the precise limits of sovereignty to be an index of sovereignty,
the Siamese court sent officials to its unmapped and unbounded peripheries to
establish the geographic outer limit of Siam’s sovereignty, defined as the exclu-
sive right to govern the polities and populations within its national territory. The
maps created under this modernizing regime, adopting the cartographic sciences
and conventions of international society, established clear limits to the Thai state’s
spatial reach, while forestalling British attempts to extend the boundaries of Indo-
Burma at Thailand’s expense.
The inscriptions of national and imperial boundary lines over the erstwhile
middle grounds that separated dynastic states were more than defensive projects
to consolidate sovereignty within a container. The construction of geobodies can
also be read as a multistate solution to a chronic problem facing maritime empires
during the second industrial revolution.62 The dynastic state’s relatively laissez-faire
approach to rule in distant borderlands was inimical to the orderly functioning of
a world economy whose industrial plants and working masses were sustained by
the timely shipment of oceangoing bulk commodities. This new epoch, wherein
national populations became dependent upon long-distance trade for daily exis-
tence, I call the age of high-velocity capitalism.
High-velocity capitalism increased the number of disputes arising over ship-
wrecks in far-flung ports, while it swelled the vehemence with which states
weighed in on such affairs. As the Qing and Tokugawa dynasts learned to their
peril, foreign governments, representing the interests of the merchants and cus-
tomers dependent upon oceanic trade, began to demand the right to negotiate
with central authorities about incidents arising in the harbors and shores that sho-
guns and emperors had kept at arm’s length since the mid-seventeenth century.
 Introduction    19

When authorities in Edo attempted to accommodate Washington’s demands to

facilitate the increased volume of oceanic economic activity in the 1850s, for exam-
ple, they quickly ascertained that one central authority in Edo could not make
the coastlines of Japan hospitable to foreign merchant marines by administrative
fiat. The legacy of Edo’s inability to command its borders from a central node of
authority is well known: the Tokugawa government crumbled in a civil war that
was ignited by confusion over how to deal with coastal traffic. Having learned how
the fate of central government was now dependent upon the extension of sover-
eignty throughout the national geobody, the Meiji oligarchs made the eradication
of legal pluralism and the instantiation of legal centralism key goals of the post-
1868 Japanese state.
A mere five years later, the Meiji state joined the global war on legal plural-
ism by pressing the Qing court on its handling of the 1871 Okinawan shipwreck
on Taiwan known as the Mudan village incident. In December 1871, Mudan vil-
lagers on the Langqiao Peninsula allegedly murdered fifty-four shipwrecked sail-
ors from Miyakojima (Ryūkyū Islands). The fate of the castaways became a cause
célèbre for ambitious Japanese politicians, who turned the islanders’ misfortune
into an opportunity for the young Meiji state to assert its territorial claims over
the Ryūkyūs (later called Okinawa Prefecture). In April 1874, Japanese admiral
Saigō Tsugumichi broke a string of diplomatic stalemates by descending on the
Langqiao Peninsula at the rear of three attack groups to avenge the Ryūkyūans,
recover their remains, and clear the coasts of wreckers.63
Within a month, Japanese troops laid waste to the southern Taiwanese vil-
lages it held responsible for the murders. Negotiations among Japanese, Qing,
and British officials in Beijing prevented armed confrontation between East Asia’s
major powers. Nonetheless, over five hundred Japanese troops perished from dis-
ease and exposure awaiting Tokyo’s orders. Ultimately, the Japanese recognized
Qing suzerainty throughout Taiwan and secured a face-saving indemnity, while
the Qing pledged to ensure the safety of distressed sailors on Taiwan’s coasts.
Recast as a chapter in the clash between high-velocity capitalism and legal plu-
ralism, the Mudan incident’s family resemblance to the American shipwrecks off
the coast of Japan in the 1840s and 1850s becomes apparent.64 The Langqiaoans’
plunder of shipwrecks was analogous to the hostility of shogunal officials who
refused succor to stranded whalers or even imprisoned them.65 From the stand-
point of heavily capitalized maritime commercial powers, stronger central author-
ities with shared commitments to safeguarding commerce were a prerequisite to
opening up East Asian markets.
In the myriad “Mudan village incidents” that litter nineteenth-century annals,
one obstacle to resolution was the uncertain, broken, and long chain of command
that connected border areas to dynastic capitals. When consular officials relayed
the protests of distressed travelers or merchants to the shogun, emperor, or king
20     Introduction

about rough treatment on their peripheries, dynasts could hardly get to the root of
the matter because their states functioned without the integrated, centralized, and
ground-level bureaucracy that characterized nation-states.
Unlike the local hong merchants, interpreters, strongmen, and gentry who con-
nected dynastic centers to their geographic peripheries, nation-state functionaries
were dependent upon the center. Indeed, replacing intermediaries and brokers
with centrally appointed officials was a venerable method of statecraft in East Asia
long before the 1871 affair. During the Ming dynasty, the tusi system of inherited
chieftainships was abolished in favor of staffing southwest China’s frontier with
appointed officials.66 In the imperial Chinese case, the rule of avoidance, coupled
with the centralized system of recruitment, appointment, and promotion, estab-
lished an administrative grid united by common language, purpose, and chain of
command. Institutionally speaking, the spatial extension of this grid represented
the limits of the emperor’s direct access to reliable information and the enforce-
ability of his edicts.67
However, under the Ming and Qing, centralized bureaucracy only reached
down to the district level. An unpaid subbureaucracy of gentry, yamen runners,
and local notables took over from there to manage the day-to-day functions of
government. The problem of central command and control was only exacerbated
in Central Asia, where a completely different system of state-to-society mecha-
nisms was instituted under the banner of the Bureau of Border Affairs. As Pars
Cassel points out in his study of Qing legal pluralism, under the imperial sys-
tem, nonofficials were not permitted the “luxury” of claiming to be subjects of the
emperor in the sense that they might have any ritual, legal, or moral claims upon
his attention, or he on theirs.68
From Beijing’s perspective, then, the Mudan incident was “local” and “periph-
eral” because its Paiwanese perpetrators, and the Hakka Chinese who rescued
the survivors, lived below and beyond the Qing’s centralized administrative grid.
One solution to the problem, attempted in the later 1870s and 1880s under the
leadership of Shen Baozhen and Liu Mingchuan, was to put military pressure on
Taiwanese border tribes in order to increase the size of the Qing tax base to pay for
more expeditions to pacify and secure the interior.69
Such punitive expeditions required rudimentary military intelligence or the
help of local allies, neither of which was beyond the reach of the Qing. As William
T. Rowe put it, “when it chose to, the [Qing] state could certainly marshal the
resources to despotically terrorize its subjects.” In the new order, however, such
power was insufficient. Rowe continues: “on a day-to-day basis [the Qing] left
many of the functions we might think of as governmental to private individuals
and groups.”70 It was these private individuals and groups who were considered
threats to a system that sought to make the world safe for the circulation of higher
volumes and velocities of commerce.
 Introduction    21

D I S C I P L I N E , G OV E R N M E N TA L I T Y, A N D B IO P O L I T IC S

In the 1874 negotiations with the Zongli Yamen over the resolution of the Mudan
incident, Ōkubo Toshimichi claimed that the Qing were not sovereign in Langqiao
because the court, the entity that answered to national leaderships in London,
Tokyo, and Washington, lacked the capacity to regulate the daily affairs of
Langqiaoans.71 In Lenin’s parlance, the Langqiao Peninsula was still among the
ownerless territories of the world. For Ōkubo, who was being advised by French
legal scholars and an American consul, establishing “ownership” or sovereignty
over a governed territory meant educating and “improving” natives to prevent
disruptive behavior, or punishing them for transgressions against merchants and
outsiders in a timely manner. This type of sovereignty involved a level of territorial
integration between borderland and capital unthinkable for the expansive, mul-
tiethnic, and legalistically plural Qing dynasty. As Meiji leaders themselves were
to learn in the 1870s and 1880s, territorial integration on this scale, at this level of
intensity, required new software and hardware for governance.
As Ōkubo was chastising Qing officialdom for its lassitude regarding southern
Taiwan, the young Tokyo government that ruled in the name of Meiji was execut-
ing its own ideological offensives, administrative overhauls, and financial strata-
gems to avoid a repeat of the Tokugawa regime collapse (which was attributable to
weak horizontal integration). This project entailed obvious centralizing measures
such as replacing feudatory leaders with centrally appointed governors, raising a
conscript army, funding and populating a public-school system, fielding a police
force, and establishing a court system. These measures required tax increases, but
getting citizens to pay such taxes, voluntarily and without recourse to revenue-
squandering and economy-stultifying structures of oppression, presupposed the
prior existence of these selfsame institutions.
To bootstrap itself out of this dilemma required nothing less than moving Japan
from a society of punishment to one of discipline. This quantum leap entailed the
instantiation of a governmental approach to statecraft. Here, I am borrowing from
geographer Matthew Hannah’s adaptation of Foucault’s terminology to analyze the
problem of state building in nineteenth-century North America. His model is apt
because, as Yao Jen-to has observed, the government-general seems to have taken
a page out of Hannah’s playbook to govern Taiwan.
According to Foucault, the French dynastic state, for all its unchecked and ter-
rifying power, was relatively weak. Beyond the palace precincts, he writes, a pleth-
ora of religious, clan-based, guild, and local leaders maintained the old regime’s
social order with recourse to bribery, corruption, and benign neglect. The problem
with these “innumerable authorities,” from the view of the insurgent bourgeoisie,
was that they “cancelled each other out and were incapable of covering the social
body in its entirety.”72
22     Introduction

It was the Chinese and Japanese social body’s uncovered areas and multiple
sources of authority that vexed the Euro-Americans who asserted a “natural right”
to protect commercial intercourse throughout the world in the mid-nineteenth
century. For Foucault, as was the case for treaty-port powers vis-à-vis the Qing
and the Tokugawa, the unpredictability of royal power from above was of a piece
with its weaknesses when directed outward. Under multicentered legal pluralism,
to borrow Benton’s terminology, “each of the different social strata had its margin
of tolerated illegality: the non-application of the rule, the non-observance of the
innumerable edicts or ordinances, were a condition of the political and economic
functioning of society.”73 Here Foucault speaks of poaching, smuggling, pilfering,
unregulated use of the commons, and other petty offences that did not concern the
crown directly. In Qing-era Taiwan, as we shall see, ransoming, or paying “aborig-
ine rent” to borderland toll-states, or redeeming severed heads from sham battles
for bounties, were functional equivalents of these tolerated illegalities.
The political economy of high-velocity capitalism, in Foucault’s exposition,
demanded a new approach to law and sovereignty. The late eighteenth century’s
increase in “commercial and industrial ownership [and] the development of ports,”
along with “the appearance of great warehouses . . . [and] the organization of huge
workshops,”74 found the sporadic and uneven consistency of royal authority insuf-
ficient for the protection of widely dispersed, ubiquitous, and moveable property.
Privately held and deployed assets were now too numerous and valuable to be ren-
dered secure, everywhere and always, by the king’s punitive brand of justice. For the
treaty-port powers, the same principle held: how could the emperor’s or shogun’s
occasional displays of sovereign/punitive power, or the compromised and negotia-
ble justice of magistrates, village headmen, and yamen runners, be entrusted with
the volumes of commerce that would pass through multinational customs houses?
For Foucault, the bourgeoisie’s solution was to instantiate “a systematic, armed
intolerance of illegality,” which he termed “public power.”75 Public power is the
aggregate effect of discipline, or what may be termed panopticism. This modal-
ity regulates the conduct of millions across vast distances because it conditions
individuals as objects and authors of a dispersed, circulating form of behavior-
modifying power as teachers and students, doctors and patients, jailers and jailed,
parents and children, and so on. According to Foucault, again recapitulating the
demands of treaty-port consuls in East Asia, it
. . . became necessary to get rid of the old economy of the power to punish, based on
the principles of the confused and inadequate multiplicity of authorities, . . . punish-
ments that were spectacular in their manifestations and haphazard in their applica-
tion. It became necessary to define a strategy and techniques of punishment in which an
economy of continuity and permanence would replace that of expenditure and excess.
In short, penal reform was born at the point of junction between the struggle against the
super-power of the sovereign and that against the infra-power of acquired and tolerated
illegalities.76
 Introduction    23

Here, the “struggle . . . against the infra-power of acquired and tolerated ille-
galities” called forth a public power that would educate, diagnose, cure, and
police populations to pay taxes, work hard, serve in the military, and respect pri-
vate property willingly if not energetically. As Caroline Ts’ai puts it, “disciplinary
power works through the construction of routine. This power is ‘constitutive.’ As
such, the Foucauldian approach . . . locates power in the ‘micro-physics’ of social
life.”77 Properly established, public power circulates continuously, everywhere and
always. Its absence, as Ōkubo argued in his diatribes against the Zongli Yamen in
1874, signals a state’s incompatibility with modern political economy made mani-
fest in the international system. For Foucault, discipline (government through
reeducation and behavior modification), rather than punishment (the use of terror,
edict, and brute force to obtain compliance), was a sine qua non for nation-states
with capitalist economies.
Following Foucault, Hannah demonstrates how societies cross the watershed
from punishment to discipline through a burst of state-directed energy, in the
form of censuses, land registration, and bureaucratic staffing. These projects lay
foundations for rule by governmentality. Discipline and governmentality are over-
lapping concepts. As Hannah puts it, discipline constitutes individuals as objects
and authors of knowledge and power, but it does so with a specific goal in mind—
the creation of docile workers and obedient citizens. Takashi Fujitani’s Splendid
Monarchy provides a relevant example of discipline making in operation.
To unify a social body riven by horizontal and vertical cleavages, the Meiji
oligarchs instantiated discipline by cultivating identification with the emperor,
imperial history, and a new national culture. The enterprise aimed to flatten status
distinctions and weld fissured territorial divisions.78 First, the emperor was made
visible across the realm by a series of lavish processions and tours, providing a
common point of reference for a people only vaguely, if at all, acquainted with the
image, person, or backstory of the state’s new symbolic, unifying center. However,
these costly processions were insufficient. Their evanescence, novelty, and circum-
scribed routes could not in and of themselves cover an entire social body to pro-
duce “a modern citizenry with an interiorized sense of themselves as objects of an
unremitting surveillance.”79 But it was a start.
Statues of mythical and actual national heroes were installed across the land-
scape, while state-authored rituals and liturgies were publicized and enforced.
Importantly, the imperial institution—including the imposing architecture of
a national capital—was relocated to Tokyo as a node of panoptical power. Now
the emperor would stay put, and the people would move. Meiji, instead of being
the object of the public’s gaze, was now the author of an omnipresent discipline-
making gaze. Meiji’s mass-produced portraits were made objects of daily ritu-
als in public schools, allowing him to be seen looking down on his subjects,
daily, to the far edges of Japanese sovereignty. The emperor’s paramountcy was
augmented by mass military spectacles that drew tens of thousands near the
24     Introduction

imperial palace, also presided over by the all-seeing emperor as the master of
ceremonies.80
To simplify, the Meiji oligarchs used the imperial person and associated pag-
eantry to both represent a body politic to itself and interiorize loyalty and obedi-
ence to a symbolic center. This massive project consumed not only the costs of
mobile pageantry but also expenses associated with Tokyo’s urban renewal and
the building of a conscript army. In microcosm, Emperor Sunjong and his retinue
traversed the Korean Peninsula in 1909 on a similar errand, to forge identification
with a subordinate national monarch. However, Sunjong reigned at the pleasure of
a Japanese resident-general named Itō Hirobumi. Consequently, Sunjong’s proces-
sion was met by protests and ended up working against the rump Yi dynasty.81 Its
utter failure to replicate the effects of the earlier Meiji processions suggests that
display and symbolism alone were insufficient in and of themselves to foment
a sense of nation among a nascent citizenry. Moreover, Japan’s 1909 exercise in
imperial pageantry reveals that measures effective in the home country might not
be exportable to the colonies.
While there would be no imperial processions in Taiwan,82 other more por-
table methods of bootstrapping social formations into the era of discipline were
imported from Japan. The Meiji land tax reform (LTR) was implemented from July
1873 until the end of 1876, at about the same time that the imperial processions
were setting out from Kyoto and Tokyo. Like the grand imperial tours, the LTR
involved meticulous planning by Japan’s top officials. At a cost of about forty mil-
lion yen,83 the central government assessed “85.44 million parcels of rice paddies
and all other types of land, and issu[ed] 109.33 million certificates of land owner-
ship” to Japanese citizens.84
The purpose of the LTR was to simplify, standardize, and make more equitable
(across prefectures) the patchwork system of Tokugawa land tenure and taxation.
Under the old system, different domains were taxed at various rates, and taxes
were paid in kind, based on a percentage of annual yield. For central planners
whose annual budgets required forecasts of income and expenses, the unpredict-
ability of the old system presented headaches. In addition, the differential tax rates
caused discontent. Therefore, the two-and-half-year project sought to convince
citizens that the new government was impartial. But more importantly for our
purposes, the reformed system levied taxes based on assessed land value to make
government receipts more predictable, to commoditize land (by issuing certifi-
cates), and to yoke peasants to market discipline by demanding cash payments.
In the short run, central-government receipts remained constant, and per cap-
ita tax rates declined in most years. Over the longer haul, Kōzō Yamamura argues,
the reform strengthened the government while it enriched the populace, because
efficiencies arising from a land market drove up aggregate productivity.85 Less san-
guine accounts do not deny the efficiency of the new system for administrators or
its ability to generate national wealth. However, because tax rates were fixed and
 Introduction    25

redeemable only in cash, bad harvests, isolation from market centers, or low prices
often sent smallholders without reserves into tenancy.86
The LTR in Japan and the one implemented later in Taiwan exhibit a governmen-
tal approach to statecraft, rather than a disciplinary modality. Governmentality,
rather than targeting individuals as objects of regulation, per se, targets the national
political economy. As Hannah puts it, “Governmentality, like discipline, constructs
(not merely “manipulates”) its objects, but unlike discipline, it constructs them as
objects that should not be unduly manipulated. . . . As such, governmentality, too,
is fundamentally structured around cycles of social control linking observation,
normalizing judgment and regulation. But unlike its genealogical predecessors,
governmentality . . . helps the governing authorities decide whether there are lim-
its to its ability to enforce or achieve the norm . . . In short, governmentality at a
national scale involves more respect for the integrity and autonomous dynamics
of the social body.”87
Seen in this light, the 1873–76 LTR was the quintessential governmental project.
A nationwide cadastral survey of farmholdings and the validation of holdings with
land certificates were undertaken to lay the foundations for a national political
economy that could be managed, in some sense, by accountants and actuaries in
Tokyo. Government yields were increased not by the principle of “squeezing the
peasant like a sesame seed to get additional oil” but rather by monetizing agri-
culture, establishing a fixed tax rate, and creating the conditions for a national
land market, thereby letting “the invisible hand” (and the enforcement of eviction
notices) raise overall yields.88
The Meiji state’s LTR was repeated a quarter century later in Taiwan. Minister
of Civil Affairs Gotō Shinpei (1898–1906) oversaw Taiwan’s first national land
survey and imperial Japan’s first population census. As with the home island, the
land registration, assessment, and deeding project was a gargantuan enterprise.
The cadastral survey itself produced 37,869 maps of villages. The project cost over
5,225,000 yen. Like its 1873 Japanese predecessor, Gotō’s LTR in Taiwan simpli-
fied and streamlined an early-modern system of land tenure and taxation. The old
Qing system, with its multiple layers of usufruct and subsoil rights and various
forms of payment, not only hid land from the tax collector but was difficult to
administer from Taipei.89
The related population census, another measure implemented to regularize
government receipts, took three years to prepare. Enumeration day was October 1,
1905. It required “842 supervisory staff, 1,339 assistants, and 5,224 census tak-
ers.”90 In terms of scale, level of detail, and cost, these two enterprises not only set
Taiwan apart from less capital-intensive modern colonies—they also preceded the
first national home-island census by fifteen years. From Yao Jen-to’s perspective,
the TGG’s exorbitant spending and attention to detail distinguished the Taiwan
Government-General from the British government in India, which staffed its
­census projects more lightly. Yao concludes that the Taiwan Government-General
26     Introduction

did not take its military dominance as an invitation to remain ignorant about sub-
ject populations.
Rather, Japan entered the imperial arena in the 1890s, after statistics had finally
become a basis for statecraft in Europe. Therefore, the advantages of followership
allowed Gotō to put the Taiwan Government-General on a scientific footing from
its inception. In Yao’s account, however, the censuses and LTR were punitive, and
not governmental in the sense Hannah uses the term. Yao rightly points out that
accurate statistical data render an alien population visible, knowable, and to some
extent manageable for outsiders by representing it in “combinable, mobile and
stable” units (borrowing from Bruno Latour).91 On the other hand, Yao asserts
that these expensive and scientific instruments were effective because they aided
the government in hunting down rebels.92 He also equates “capitalism” with orga-
nized theft by noting that land registration made Taiwan’s real estate available for
confiscation by Japanese corporations after the hidden acreage was exposed in the
cadastres.93 In these two instances, systematic and comprehensive knowledge was
weaponized for naked exploitation.
In contrast, Ka Chih-ming has analyzed this same land reform program as a
governmental undertaking. For Ka, the purpose of Gotō’s land reform was only
in part to increase the government’s tax yield by registering more land. It also
served to cement a TGG class alliance with Taiwan’s smallholders against absen-
tee landlords who collected “small rents” and then forwarded “big rents” to the
capital (after skimming). According to Ka—and this echoes the circumspection
with which U.S. and Japanese land surveys proceeded in noncolonized spaces—­
cultivators cooperated with the cadastres in exchange for fee-simple titles to their
land. Moreover, the Taiwan Government-General could reallocate these titles
thanks to a large buy-off from absentee landlords—paid for by funds raised on the
bond market in Japan.
From a fiscal perspective, Gotō’s gambit worked. From 1905 onward, Taiwan
no longer required subsidies from Tokyo and became exceptional in the annals of
Japanese colonialism by becoming a boon, rather than a drain, on Japan’s national
economy.94 Ka writes, “With order restored and revenues ensured through deficit
financing, the administration implemented a number of measures to increase pro-
duction and expand the market: a thorough land cadastral not only revealed tax
evasion but greatly facilitated the subsequent reform that created a modern private
land ownership system. The land survey and reform guaranteed private property
rights, facilitated the effective use of land, and provided incentives to increase pro-
duction for the market.”95
Anticipating Ka’s analysis by many decades, the 1906 Japan Yearbook exclaimed
that the “success of [Japanese] colonial policy in Formosa is conclusively dem-
onstrated in the Revenue Column.  .  .  . The item of ‘Subsidies from Central
Government’ that was steadily diminishing finally disappeared . . . ”96 W. G. Beasley,
also noting the parallel between Japan’s 1870s and 1880s reforms and Gotō’s early
 Introduction    27

twentieth-century expenditures in Taiwan, concluded that the bundle of projects

reviewed above had secured “financial stability which . . . made it possible to meet
budget deficits by floating bonds, placed with a newly formed Bank of Taiwan.”97
This quantitative yardstick for success—achieving black-ink balances on reg-
ularly scheduled budgets—represents the type of governmental rationality well
articulated by Max Weber in his landmark study The Protestant Ethic and the Spirit
of Capitalism. Weber’s linked concepts of quantification and continuous renewal
resonate strongly with contemporary governmental projects in colonial Taiwan.
Weber writes:

. . . capitalism is identical with the pursuit of profit, and forever renewed profit, by
means of continuous, rational, capitalistic enterprise. For it must be so: in a wholly
capitalistic order of society, an individual capitalistic enterprise which did not take
advantage of its opportunities for profit-making would be doomed to extinction.
. . . The important fact is always that a calculation of capital in terms of money is
made, whether by modern book-keeping methods or in any other way. . . . Everything
is done in terms of balances: at the beginning of the enterprise an initial balance,
before every individual decision a calculation to ascertain its probable profitableness,
and at the end of a final balance to ascertain how much profit has been made.98

This passage, on the one hand, reflects the “eat or be eaten” worldview of a number
of Meiji intellectuals and statesmen in the 1890s. For example, influential author,
public intellectual, and publisher Tokutomi Sōhō, who came to see Japanese impe-
rialism as a means of national survival at century’s turn,99 would have assented
to substituting “nation” for “enterprise” in Weber’s quotation: nations that do not
compete successfully are doomed to extinction.
Weber’s formulation also encapsulates a core principle of governmentality in
Taiwan under Japanese rule. Beginning with the LTR and the census, the Taiwan
Government-General quantified human, environmental, and built assets in the
colony, and more importantly, it set and recalibrated policy based on income
and expenditure projections. As Hannah points out, a governmental state relies
upon statistical record keeping and computation to manage its heterogeneous
resources and budget for multiyear projects. In the competitive imperial setting,
failure to balance the colonial budget portended scrutiny from the Diet, which
held the purse strings of the military and supplementary budgets for the Taiwan
Government-General. Therefore, Weber’s foregrounding of capitalism’s continu-
ously renewed commitment to positive balances captures the spirit of capitalism
as an ethos but also isolates a distinctive feature of statecraft in the postdynas-
tic era. In this dispensation, national expressions of public power underwrote
territorially aggrandizing war machines. National militaries, as income-starved
institutional behemoths, not only competed against each other but also competed
with domestic spending projects that sustained the public power presupposed by
militarized states.
28     Introduction

Weber’s contemporary Takekoshi Yosaburō issued a book-length encomium to

Gotō Shinpei and Japanese rule in Taiwan the same year that The Protestant Ethic
first appeared. Takekoshi articulated the governmental ethos in his stout defense
of the embattled government-general in Taiwan. In 1904 he wrote, “If the trade
[between Taiwan and Japan] continues to grow as it has done during the last six of
seven years, Japan will by about the year 1910 have received back an equivalent of
all the subsidies [to the Taiwan Government-General], together with the interest
upon them. From that time Japan will have reached the goal of colonial enterprise,
and be able to look to her colony for substantial support.”100
Based on his own projections, the economic historian Takekoshi inserted a
table that projected TGG annual income, expenditures, public loans (princi-
pal plus interest), and a positive overall balance for the years 1903 through 1922.
Unsurprisingly, each year cost more than the previous one.101
In another example of Weberian rationality at work in colonial Taiwan, TGG
councilor Mochiji Rokusaburō issued his famous “opinion paper on the Aborigine
problem” (Bansei mondai ni kansuru ikensho) in December 1902. With an eye to
the bottom line, Mochiji declared that the “savage-territory problem” (banjin ban-
chi mondai) was, strictly speaking, a land management problem. As peoples who
did not submit to the Qing, the “savages” (banjin/seibanjin/genjin) were not cov-
ered in the articles of the Treaty of Shimonoseki that formally ceded Taiwan to
Japan. Moreover, Taiwan’s seiban (savage) population lacked recognizable organs
of government and therefore stood outside the rules of civilized warfare and diplo-
macy. As Mochiji put it, “sociologically speaking, they are indeed human beings
(jinrui), but looked at from the viewpoint of international law, they resemble ani-
mals” (dōbutsu no gotoki mono).
Mochiji reasoned that, insofar as “savage territory” could be made to pay for
the expenses of punitive expeditions and policing, they would be launched. But if
expeditions were sent simply for the purpose of revenge or a display of dominance,
Mochiji argued that they should not be undertaken.102 Accordingly, the military
expeditions targeted the camphor-rich north and left the south mostly untouched.
Mochiji here displayed a clear aversion to punishment (in Foucault’s sense) and
showed himself ever the rational bureaucrat.
Takekoshi added urgency to Mochiji’s plan a year later with the plea that “time
was running out” for a solution to the “problem of the savages.” He wrote that
the militarization of Japanese-Atayal relations “does not mean that we have no
sympathy at all for the savages. It simply means that we have to think more about
our 45,000,000 sons and daughters than about the 104,000 savages.”103 Takekoshi
thereby embraced home-island Japanese as “sons and daughters” while banishing
indigenes from the national family.
Thereafter, the Taiwan Government-General and the Tokyo government
dribbled out funds to cobble together remnant Qing private armies and expand
the police force in order to secure lucrative areas of camphor production. It took
 Introduction    29

the political clout of fourth governor-general Sakuma Samata (1906–15), a hero

of the 1874 invasion of Taiwan, to pry the lid off the imperial treasury. Drawing
upon connections to Prime Minister Katsura Tarō and father of the conscript
army Yamagata Aritomo, and with the Meiji emperor’s blessing, Sakuma secured a
whopping fifteen million yen, in 1909, for a five-year “big bang” spending program
to solve the “aborigine question” once and for all.
While the scale, ambition, and cost of Sakuma’s “five-year plan to control the
aborigines” was commensurate with the 1873–76 LTR in Japan and the 1898–1905
Gotō spending package in Taiwan, its rationale was not quite the same. In the vein
of these early programs, the idea was to create conditions for future wealth accu-
mulation with a large initial investment. But in this case, the human beings in the
targeted area were not considered sources of labor power or entrepreneurial skill
but were seen instead as obstacles (unless they could be mobilized to subdue other
indigenes). There was no attempt here to instantiate the virtuous circle of gov-
ernmentality. Therefore, Sakuma’s five-year plan, and the TGG policies regarding
indigenes more generally, adopted a punitive and biopolitically inflected rational-
ity of statecraft, to the exclusion of disciplinary concerns.
This was the crux of the matter: typically, in a nation-building context, disci-
pline precedes or coincides with governmentality. However, in a colonial situation,
wherein the “natives” (in Bayly’s sense of the term) are not reckoned as a potential
source of labor but rather as impediments to resource extraction, discipline is—by
the yardstick of governmental calculations—wasteful. Foucault’s notion of “bio-
politics”104 is operative here.
Sabine Frühstück’s account of the “modern health regime” in Meiji Japan illus-
trates the positivities of biopower. Her study analyzes state initiatives to “force a
population to be healthy.” Several biopolitical projects, involving vital statistics,
the regulation of sex, and government attention to nutrition, exercise, and inocu-
lation, improved Japan’s racial stock so it could stand up to the West and compete
in the international arena.105 This ensemble of state interventions, many cotermi-
nous with the LTR of 1873–76, was aimed at building a healthy conscript army and
industrial labor force.
However, like the LTR, the regime of hygiene aimed to improve national power,
not individual health and wealth. As many farmers were dispossessed by land
reform, the health of state-regulated military prostitutes was sacrificed for the gen-
eral health of the body politic. Turned outward or against internal Others, then,
a biopolitical regime marks out subpopulations as expendable in the name of the
greater good. Fujitani and Achille Mbembe follow Foucault by insisting that arro-
gating the power to sustain life entails the assertion of a right to extinguish it. This
duality has been referred to as “necropolitics.”106
Patrick Wolfe’s analysis of the race politics in settler societies emphasizes this
point. In North America, he writes, constituting the black race as a population and
then nurturing its demographic resilience (partly through the legal fiction of the
30     Introduction

“one-drop rule”) had a biopolitical logic: labor-pool management. On the other

hand, the connected project of Indian removal (through both physical removal
and the legal fiction that “mixed blood” Indians were, in fact, not indigenes) had a
different economic logic: the land inhabited by indigenes was worth more, calcu-
lated in terms of cash value, if Indians could be removed.107
Importantly, in this model, racism creates races—not the other way around.
Extenuating circumstances may even impel biopolitical regimes to reverse race-
based policies in the name of self-preservation. In Takashi Fujitani’s analysis of
ethnic management during the Pacific War, the demands of mobilization pushed
the Japanese and American regimes of the 1940s to transition from vulgarly to
politely racist outlooks in the name of national security. As Japanese-Americans
in the United States and Koreans in the Japanese Empire were identified as sources
of labor and fighting power, each regime sought to inculcate that hallmark of disci-
pline, the spirit of voluntarism, among formerly excluded populations. The shift to
polite racism and the new emphasis on disciplinary measures such as compulsory
assimilation programs aimed at the “conduct of conduct” announced a redefini-
tion of marginalized communities in terms of cultural capacities or peculiarities,
rather than as biologically defined Mendelian populations.108
Leo Ching has made a related argument regarding Taiwan’s indigenes in the
1930s and 1940s. In Ching’s analysis, a combination of alarm over the Wushe
Rebellion and wartime mobilization requirements pushed the Taiwan Government-
General to imagine the former “savages” (seibanjin) as the Formosan Aborigines
(Takasagozoku). This new politely racist rhetoric did not entail equal treatment
or the end of exploitation, but it portended a shift in administrative rationality.
Henceforth, the Takasagozoku would not be formally excluded with the vulgarly
racist epithet banjin. They would instead be targets of disciplinary ministrations
aimed at a particular kind of subject formation, one defined by a willingness to
die for the emperor.109 Huang Chih-huei’s study of the Takasagozoku volunteers
who claimed to have exhibited even more “Japanese spirit” than their home-island
officers in the Philippines campaigns of 1943 and 1944 stands as a testament to the
efficacy of “imperial subjectification” among some indigenes in Taiwan.110
This book argues that “imperial subjectification” among indigenes can also be
understood as an indigenization movement. While Ching’s chronology and analy-
sis are persuasive as a two-stage model to chart the mechanisms and meanings of
the shift from vulgar to polite racism in TGG policy, its emphasis on the ideational
aspects of citizenship to the exclusion of its legal-economic bases leaves room for
some debate.
The Taiwan Government-General’s assimilation campaigns were preceded
by the multistage trading-post and guardline projects. During the Kodama and
Sakuma eras (1898–1915), northern tribes were the expendables in the government-
general’s biopolitical calculus. These punitive and often bloody campaigns encir-
cled erstwhile borderland “natives” (in Bayly’s sense) with administered territories,
 Introduction    31

reconfiguring their homelands as an ethnic enclave.111 While they were regarded as

“savage,” and governed largely through the negative techniques of sovereign power,
as objects of biopolitical management they were sequestered, “censused, mapped,
and museumed.”112
As “specially administered” subjects, the Taiwan Government-General posi-
tioned indigenes as exterior to the more intensively disciplined and governmental-
ized territories governed by baojia (mutual-surveillance and tithing organizations)
in “regularly administered” Taiwan.113 From 1898 through 1902, the government-
general revived the old Qing-period system of mutual responsibility and tithing
as an efficient mechanism for registering, policing, and mobilizing a Han labor
pool. By 1920, the surface of “normally administered Taiwan” was covered with
baojia districts.114 These had all been subject to land markets and taxation and were
highly commoditized as integral parts of the imperial political economy. However,
negatively defined indigenous territory, by 1920, was already “public land” and
excluded from the baojia system. By the time the imperial subjectification cam-
paigns of the 1930s were launched, access to the indigenous territory by outsiders
had been suppressed by a permit system for three decades.
Against this backdrop, the imperial subjectification policies of the 1930s were
implemented through a new generation of indigenous youth corps (seinendan).
These corps leaders had attained adulthood under Japanese rule in an ethnic
enclave. As the government-general mobilized these corps through drill, parade,
song, schools, panindigenous gatherings, and finally military recruitment, it solid-
ified the standing of the indigenous territory as a separate entity, or what I have
termed a “second-order geobody.” Therefore, the “imperial subjectification” poli-
cies, even though they were conducted in the Japanese language, could only further
isolate the indigenous territory from the social, political, and economic currents in
Han Taiwan that were preparing the ground for postcolonial nationality.
During kōminka (imperial subjectification), indigenes became proficient in
spoken Japanese. They hoisted Hinomaru (Japanese) flags and belted out impe-
rial anthems with apparent gusto. Moreover, some abandoned “evil customs”
such as in-home burials and face tattoos. Nonetheless, the notion that they were
“becoming Japanese” overlooks an important aspect of modern citizenship: its
material basis.
Unlike their neighbors, the specially administered peoples of Taiwan had
never been citizens or subjects of agro-bureaucratic empires-cum-dynastic
states. As “natives” practiced in the art of not being governed, most of Taiwan’s
Austronesians were not accustomed to the routines of unremitting, coordinated,
and surplus-generating agricultural and manufacturing toil that characterized the
lot of their lowlander neighbors under Tokugawa, Qing, and Joseon rule. Looking
backward, it would seem that the success of disciplinary projects such as imperial
panopticism, the Taiwan Government-General’s graft of a Japanese police appara-
tus upon a Han Taiwanese baojia system, or the cultivation of national Korean and
32     Introduction

Taiwanese bourgeoisies were all premised on longue durée processes begun before
the great age of transformation.
Colonial rule in Han-dominated Taiwan took on a coloration of nation building
in Japan, under a regime of governmentality/discipline,115 whereas the indigenous
territory of Taiwan was ruled under a hybrid regime of punishment and biopol-
itics, at least until the 1930s. However, these regimes did not operate indepen-
dently but were built together, one upon the back of the other. Initially, Japanese
officials thought that highlanders could be mobilized on Japan’s side against low-
landers in the colonial state’s war against armed Han resistance. Looking back at
these early years, commentators considered 1895–98 to have been a “honeymoon
period” of Japanese-Indigenous relations. During this interval, brokers like Kondō
the Barbarian flourished as trading-post operators, land speculators, and contract
employees of the state. But the debt incurred by the Taiwan Government-General
to set off the “big bang” of governmentality in the plains put pressure on the inland
sea of Taiwan’s camphor forests.
As Japanese officials were well aware, frontier violence, which began to spike in
1898, could have been reduced by simply backing the rights of indigenes vis-à-vis
Japanese lumber companies and their employees. Indeed, such a policy had much
to recommend itself while Japan sought allies in its war against Han guerillas. In
fact, the biggest camphor merchants in Xinzhu Prefecture in the late Qing- and
early Meiji-period camphor industry were Saisiyat men, most notably Ri Aguai.
Some indigenes were more than willing to sell camphor—the resource so ardently
coveted by Japan’s state builders and merchants—to outsiders for monetary gain.116
But the state could not exercise regulatory functions to control the volume or qual-
ity of camphor under “middle ground” or “contact zone” conditions. State plan-
ners wanted to standardize quality, regulate output, and project estimated profits
from its monopoly bureau for annual budgetary purposes.117 However, unlike in
the plains, where smallholders and some merchants, as well as scholars, could
be co-opted to provide the means for igniting the motor of disciplinary rule, the
Japanese found few property holders to make a parallel class alliance in the high-
lands. With this crucial element missing, the Japanese state pressed forward in its
efforts to make the Aborigine Territory into a state space with a series of punitive
expeditions and economic embargoes.
Sediq, Atayal, and Truku settlements surrendered their arms en masse and
were at times relocated to valleys within reach of the police apparatus. After estab-
lishing this beachhead, the Japanese state embarked on a series of cartographic
and infrastructural projects to make the indigenous territory legible to the state.
But as Matsuoka Tadasu has demonstrated,118 the post-1915 Japanese state balked,
in a sense, by not following up its military victories with policies of reterritorial-
ization that would have rendered the non-Han spaces of upland Taiwan indistinct
from the rest of the island colony. Legible, yes, but subject to modern discipline—
not quite.
 Introduction    33

To restate the main argument: national and indigenous political formations are
homologous, historically linked, and symbiotic. Modern state building in the age
of high-velocity capitalism entailed heavy governmental outlays to create com-
mensuritized sociopolitical formations for sustaining the timely circulation of
information, goods, and people, all under the pressure of international competi-
tion. In the emergent international system, at least ideally, one national geobody’s
sovereignty ended where another began. The indigenous geobody was distinctive.
As an administered territory defined by its exteriority to the full array of citizen-
making projects associated with governmental and disciplinary tactics, the indig-
enous geobody was a “second-order geobody.” It was discrete and bounded, and it
took on the formal properties of a geobody. Instead of achieving national sover-
eignty, however, it remains a subunit of a first-order geobody, the arena of disci-
pline- and citizen-making projects in the age of global transformation.
In Taiwan under Japanese rule, the virtuous circle of governmentality that is
presupposed by legal centralism reached its geographic limit at the boundary of
the Aborigine Territory. Therefore, Taiwan’s second-order geobody can be consid-
ered the ethnogeographic expression of the Japanese regime’s aspirations to master
a territory just beyond its grasp. A first-order geobody describes a spatially config-
ured unit of administration within which a mixture of Foucauldian discipline and
Gramscian hegemony produces more or less self-actuating populations who sur-
render revenues to central authorities over and above the cost of funding myriad
projects in governmentality. Such policies, programs, and installations underwrite
the accumulation of capital, and its diversion to state coffers, at rates sufficient to
balance accounts as calculated in annual government reports. Second-order geo-
bodies are found at the extremities of empire, where various combinations of local
resistance, rugged terrain, sparse population, and other factors rule out the cre-
ation of revenue-neutral regimes of governmentality. Although not brought within
the fold of governmentality, the second-order geobodies were nonetheless consti-
tuted as objects of accounting, surveillance, and biopolitical ministration because
the modern international system—unlike the interdynastic order it ­displaced—did
not allow for ungoverned territories to exist “beyond the pale.”

I N D IG E N OU S M O D E R N I T Y

In its 1905 census, the Taiwan Government-General attributed Malayan and

Mongolian racial characteristics to populations that also fell respectively under
“special” (savage) and “regular” (civilized) administration119 (see figure 3).
This innovation confirms Charles Hirschman’s hypothesis that modern racial
epistemology had a hand in hardening the once porous occupationally and con-
fessionally defined sociopolitical boundaries that characterized the early-modern
world. As was the case in colonial Malaysia, the racializing process in colonized
Taiwan employed circular logic. In each colony, so-called Malays (indigenes) were
Figure 3. This ca. 1904 postcard divides Taiwan into two races, territorially defined as the
“Han race” and the “savage race.” Courtesy of the Rupnow Collection.
 Introduction    35

distinguished by their putative incapacity to participate in a modern, capitalist

economic system, with all of its risks and benefits. Thus, Malays were placed under
“special administration” and deprived of any chance to participate in a modern,
capitalist economic system.120 TGG officials asked: Did villagers have written deeds
to their lands? Did they grow irrigated (surplus) rice? Could they honor contracts?
If the answer was “no,” then they were Malays.
On the one hand, Japanese men on the spot defined race by recourse to ste-
reotypes and institutions disconnected from the craniometry and anthropometry
that stood at the foundations of scientific racism. Nonetheless, at the back of the
Malay label was a large corpus of Japanese race science. This bundle of scientis-
tic ethnic maps, tables, photographs, and assessments of character mostly went
ignored by policy makers in the early decades of Japanese rule. Yet it reemerged
in the 1920s as the substrate for further elaboration, and it has been energetically
revived in post–martial law Taiwan since the 1990s.
The classificatory work of three men in particular has drawn sustained attention
from scholars of Japanese rule in Taiwan and of the history of anthropology in Japan’s
empire. They will play important roles in this study, as well. These three pioneering
government ethnologists—Torii Ryūzō (1870–1953), Inō Kanori (1867–1925), and
Mori Ushinosuke (1877–1926)—completed their most celebrated work during the
first two decades of colonial rule, between 1895 and 1915. Methodologically, they
arrayed various word lists, artifacts of material life, and elements of intangible cul-
ture into the tabular form that characterized fin-de-siècle anthropology’s compara-
tive method. Based on this data, they sorted Taiwan’s diverse upland settlements
into a nine-tribe schema. These tribes were represented as commensurate units
juxtaposed like nation-states on a Mercator-projection map.
Reflecting their engagement with transnational currents in the young science
of ethnology, these men were dissenters from the crudest forms of polygenetist
racism that informed evolutionary anthropology. Rather, like Franz Boas and
kindred spirits in England and Germany, these men advocated the worth and
importance of nonliterate, small-scale societies as objects of study and subjects of
history. To be sure, like Boas himself, they were self-assured in their superiority as
members of industrial, urbanized societies. They neither aimed for nor achieved
the Malinowskian ideal of empathy or access to the interior lives of members of
the societies who were the objects of their taxonomic labors. As survey ethnolo-
gists, however, they were much more concerned about specificity, particularity,
and the well-being of the peoples they studied than armchair ethnologists who did
not concern themselves with face-to-face encounters. Reflecting their complicated
relationship to the politics of conquest and ethnogenesis, their work was too fine-
grained in some regards, and too abstract in others, to appear immediately useful
to administrators in the early decades of colonial rule.121
However, on the heels of Governor-General Sakuma’s campaign against the
northern tribes, which ended in 1915, several trends conspired to put the latently
36     Introduction

pluralist conception of Taiwan devised by Torii, Inō, and Mori on a permanent

footing, one that lasts to this day.122 First, reflecting the post–World War I ethos of
cultural appraisal (which was eclipsing racist denigration) that accompanied the
Wilsonian moment, the new apostles of folklore and ethnic authenticity portrayed
themselves as the stewards of territory inhabited by the non-Han Malayans.123 This
operation figuratively swept the landscape of Chinese claims to prior occupation.
Moreover, an ethnic tourism trade emerged in Taiwan’s highlands in the 1920s.124
Its myriad graphic and material products were organized around the ethnic-­
pluralist model and institutionalized a view of indigenous Taiwan that rendered it
a culture garden (see chapter 4).
To put these developments into global context, we recall that Bayly’s narrative of
displacement, depopulation, and dispossession ends in 1914 on a cautious note: he
did not write off the small-scale societies that Eric Wolf once termed the “­peoples
without history.” Bayly noted that indigenous resilience and creativity, the new
human science of anthropology, a market for primitivist art in the metropole, and
a growing revulsion to settler avarice in the mother countries had prevented the
natives from becoming extinct. Turned on their heads, the forces Bayly regards as
brakes on the extermination of natives can be reconceptualized as a seedbed for
the creation of indigenism.
Anthropologist Ronald Niezen defines indigenism as the movement of First
Nations and indigenous peoples to secure political rights, or even national self-
determination, based on claims to primordial and prior ties to particular territories.
Niezen has observed that indigenism has flourished with the strengthening of inter-
national society after World War II. He argues that the First Nations or indigenous
peoples of the world, who number some “three hundred million people from four
thousand distinct societies,”125 are now a self-conscious, transnational force, whose
current methods of rights articulation, preservation, and reclamation would have
been scarcely imaginable in the prewar era. One of this book’s tasks is to reconcile
the corrosive effects of nation building and capitalism on the natives during the
global transformation that left the world with “four to eight million” people existing
on the margins of expanding states circa 1900 and the rise of indigenous peoples as
an integral component of international society in the current dispensation.
In this case study of Taiwan under Japanese colonial rule, the period that sepa-
rates Bayly’s narrative of decline and Niezen’s assertion of interstate system / First
Nations symbiosis—1915 to 1945—brackets the years when central administrators
in Taipei institutionalized and culturally invested the second-order geobodies that
comprise Taiwan’s indigenous territory. Prasenjit Duara writes that during the
interwar period, Wilsonian notions of ethnic self-determination, combined with
strains of historical pessimism in the West, blunted the impulse to conquer new
colonies outright or to expand state borders at the expense of natives. Duara argues
that a new, globally circulating “symbolic regime of authenticity” based political
legitimacy on primordial ties to territory instead of on the rights of the conqueror
 Introduction    37

to rule the conquered. Thus, “cultural rule” in Korea after 1919 and Japan’s insis-
tence that Manchukuo be ruled as a de jure nation-state rather than as a colony
were consonant with the Wilsonian outlook. In both cases, Japanese ideologues
anchored claims to legitimacy in the language of a brand of stewardship that was
exercised in the name of the authentic nation/nationals of each dependency.126
In Taiwan, the discourse on sovereignty and authenticity dovetailed with an
ethnic tourism trade that ennobled former “savage headhunters” as ethnic groups
worthy of study, preservation, and even emulation. Indigenous artists, tourism
industry workers, native informants, and politicians responded to those oppor-
tunities by producing the artifacts, icons, and performances that put flesh on
the administrative-academic skeleton of territorially defined ethnic difference.127
Under conditions of economic and physical isolation, coupled with cultural
investment, the ungoverned peripheries of yore became the indigenous territo-
ries of today—now imagined as timeless, authentic, and buffered from the flow of
world history.
Acknowledging that historical forces created indigenous peoples is not the same
thing as claiming that indigenous claims for rights recovery in land, resources, or
human rights rest on invented traditions and are thus hollow. As Arif Dirlik put
it, “[a] critique of “cultural essentialism” that offers no articulated means to distin-
guish between the essentialism of indigenous ideology, and the essentialism of a
Confucian revival or [political scientist Samuel] Huntington’s vision of war among
civilizations, may be methodologically justifiable; but it is, to say the least, morally
irresponsible and politically obscene. Indigenous claims to identity are very much
tied in with a desperate concern for survival; not in a “metaphorical” but in a very
material sense.”128
The example of the Saisiyat group in northern Taiwan provides a counterweight
to breezy notions that indigenous peoples themselves, as self-defined collectivi-
ties, were invented by colonialist discourse or urban metropolitan longings. While
the notion that their homeland is circumscribed by a fixed boundary line pop-
ulated by bearers of a unique cultural ensemble is indeed a colonial invention,
Qing-period records indicate that Saisiyat peoples identified themselves as Saisiyat
before Japanese colonization. Moreover, they have retained and defended a sense
of corporate identity without recourse to projected or reified cultural markers as
core elements of identity. At the same time, colonial-period photographs, arti-
facts, and documents, as material objects, have—to use Hu Chia-yu’s apt phrase—
“enlivened” the past and served as touchstones for perseverance in the face of
settler encroachment and institutionalized racism. Hu’s invocation of Marshall
Sahlins reinforces Dirlik’s point and bears repeating: “The awareness of indigenous
culture as a value to be lived and defended is not just a simple and nostalgic desire
for the fetishized repositories of a pristine identity, but signifies the demand of
them for their own space within the world cultural order, and recognizes their
existence in the context of national or international threats.”129
38     Introduction

The goal of this book is to think through the processes and structures in world
history that coproduced nation-states and indigenous peoples. It is less about the
invention of tradition among indigenous peoples than it is about the incomplete
consolidation of legal centralism characteristic of state building in the era of com-
petitive imperialism and high-velocity capitalism. The world that was obliterated
by the twin forces of capitalism and nationalism was not that of the natives but
rather that of the in-between spaces of ambiguous sovereignty, the buffer zones at
the margins or interstices of centrally administered, surplus-extracting dynastic
and nation-states. In this niche, chiefs, interpreters, trading-post operators, track-
ers, headmen, and professional border crossers brokered the coexistence of politi-
cal autonomy and economic interdependence during the global transformation.
As it turns out, the success or failure of projects to instantiate governmentality
in Taiwan largely hinged upon whether or not a district or area had fallen under
Qing administration. If this case study has broader applicability, the disciplined
social fields considered normative in our models of international relations may, in
fact, be products of longue durée processes that spanned successive empires. This
finding suggests that today’s international order is premised on sociological mod-
els that are relevant to a minority of the earth’s population. Moreover, it suggests
that bringing historically undisciplined populations within the ken of interna-
tional society (“nation building”) is probably beyond the means or will of any state
that possesses the surplus wealth to pull it off. This is because such projects exhaust
the patience of citizenries who must foot the bill, and because governments that
ignore the centripetal pull of popular sentiment by definition lack the stability and
resources to complete long-term projects for exporting discipline beyond their
own borders.130

O R G A N I Z AT IO N O F T H I S B O O K

This introduction has argued that the process Ronald Niezen dubs “indigeniza-
tion” and the phenomenon James Clifford terms “Indigènitude” are historical
concomitants of nation building during the era of high imperialism (1870–1914).
Anthropologists and sociologists consider the decades following 1960 to be the incu-
bation period for the global indigenism movement. Indigenism, in their view, rests
upon an infrastructure of interstate coordination anchored in UN human-rights
conventions, transregional NGO activity, and democratically sanctioned political
protest. By looking for the origins of indigenism in the late nineteenth century, I
have argued that seemingly disparate formations—a supra-state, nation-states, and
stateless peoples—emerged as contemporaneous formations. The case study of
Taiwan under Japanese rule suggests that nationality, internationalism, and indi-
genism were mutually constituted, rather than sequentially occurring phenomena.
Chapter 1 of this study begins with the murders of fifty-four shipwrecked
Okinawans on Taiwan’s Hengchun Peninsula. This cause célèbre put the opposing
 Introduction    39

logics of the dynastic and international systems on a collision course. Hengchun’s

mountain ranges were resistant to rule from a distance under the Qing, hindering
the recovery of remains and the punishment of the alleged killers. The first part of
chapter 1 explains how the lack of horizontal integration between the Hengchun
Peninsula and the Qing Empire necessitated “wet diplomacy.” This mode of inter-
action stressed particularistic, emotionally charged attachments requiring period-
ical renewal in the absence of administrators, courts, and policemen. To eradicate
this obstacle to efficient resource management, which survived the transfer of
sovereignty from the Qing to Japan in 1895, the colonial government built armed
bunkers, relay stations, and guard posts along a scorched-earth trail known as the
aiyūsen to enclose the Atayal settlements of northern Taiwan. From 1903 through
1915, government forces extended the barrier and then marched it inward. The
Taiwan Indigenous Peoples, as a modern political-cultural formation, emerged
from the ruins of these scorched-earth campaigns as a disarmed, geographically
isolated, and dispossessed minority population.
Chapter 2 advances the argument that linguistic diversity occasioned institu-
tional creativity and strategic frustration for colonialists in Taiwan. It examines
state-led efforts to yoke an unruly Qing periphery to a centralized administration
through its language and education policies. In Taiwan, administrators required
methods to govern a patchwork of Chinese- and Austronesian-language-speaking
populations. Commercial interdependency between the island’s varied social for-
mations spawned more organic solutions to the problem of cross-border com-
munications, as well. Top-down, centrally installed measures, such as language
training for colonizers or the colonized, were most amenable to administrative
cohesion. However, formal language instruction was not scalable, due to cost con-
straints. Bottom-up arrangements, such as intersocietal adoption and marriage-
centered alliance, were cost-effective but less answerable to central command. The
Japanese state ultimately opted for intermarriages as a strategy—a holdover from
Qing times. This policy came up short because it fostered identification with par-
ticular local power structures. It also allowed colonists to establish kin-ordered
bases of authority independent of their bureaucratic chains of command. The
political marriages, therefore, contradicted the more general colony-wide impulse
to homogenize bureaucratic, legal, and economic space, and they led to intelli-
gence failures and discontent among the ruled and even to the Wushe Rebellion
of 1930.
The next chapter demonstrates how gifts performed multiple functions in the
borderland economy of Qing-period and Japanese-ruled Taiwan. Travelers brought
offerings to the mountainous hinterland to pay for services rendered or expecta-
tions of future assistance. In addition, the presentations of gifts were occasions
for recording the emotional states of recipients and for gauging the dispositions
of little-known peoples as either “greedy,” “honest,” or “uncorrupted.” Gifts also
fostered trade dependency among indigenes. As the Taiwan Government-General
40     Introduction

intensified resource extraction from camphor forests in the 1900s, it regulated gift
giving as an arm of policy, to reform indigenous mores or to reward and punish
highlanders. One class of these gifts, those involving dyed red thread or cloth, took
on additional meaning, as these materials were disassembled and reassembled into
“traditional” Atayal textiles. These were either consumed locally or reexported to
trading posts, anthropologists, curators, and tourists. In this last guise, Atayal red-
dyed textiles became exhibits in discourses on authenticity, progress, and primitiv-
ism. By tracking the uses of textiles across these many domains, c­ hapter 3 illustrates
the productive, though problematic, interdependence of indigenes, settlers, and
metropolitans in the production of colonial modernity and global indigenism.
Chapter 4 examines a battery of twentieth-century Japanese official, academic,
and commercial publications about Taiwan indigenes. These artifacts deployed
texts, pictures, and maps to manufacture a “large reservoir of cultural imaginar-
ies”131 that have reemerged as reference points and resources for the indigenous
renaissance in the twenty-first century. In what began as a quest to wrest resources,
impose administrative order, and promote immigration to an erstwhile Qing
borderland in the late 1890s and 1900s, Japan’s colonial administration, tourism
industry, and scholarly apparatus, by the 1930s, had completed numerous projects
in ethnic typification, geobody construction, and racialization to institutionalize
indigeneity in Taiwan. This concluding chapter analyzes these three interrelated
processes to demonstrate how the increased intensity of resource extraction in
colonial Taiwan intersected with historical trends in reprographic technology and
new forms of state making to ethnically pluralize the island’s non-Han populations
as second-order geobodies.
 PART ONE

The Anatomy of a Rebellion

 1

From Wet Diplomacy to Scorched Earth

The Taiwan Expedition, the Guardline,
and the Wushe Rebellion

I N T R O DU C T IO N

On October 7, 1930, Japanese patrolman Yoshimura Katsumi passed through the

hamlet of Mehebu (near Wushe) during a wedding celebration. There, a Sediq
man, one Tadao Mona, beckoned him for a drink. Yoshimura brusquely refused,
thereby slighting Tadao publicly. Tadao’s hands were bloody from butchering
meat for the festival repast. Nonetheless, he touched Yoshimura’s freshly laun-
dered uniform and left a red stain. Yoshimura angrily struck Tadao twice on the
hand with a cane. Along with his younger brother Bassao Mona, Tadao wrestled
Yoshimura to the ground and returned the blows. The next day, their father, head-
man Mona Ludao, went to the nearby Japanese police box with bottles of millet
wine in hand to formally apologize. Much to Mona’s chagrin, the branch chief,
Sugiura Kōichi, refused the gift and apologies. Instead, Mona and his sons were
reported to Sugiura’s superiors. Mona was told to expect severe punishment; the
possibilities ranged from detention to a punitive expedition against his entire vil-
lage. Yoshimura’s report and Sugiura’s handling of the matter drove Mona into the
arms of other rebellious Sediq men. Mona Ludao himself led the bloody Wushe
Rebellion three weeks later. In the October 27, 1930, assaults, Tadao Mona’s squad
killed Yoshimura Katsumi in the first wave.1
These events of early October, known as the “Yoshimura beating incident”
(Yoshimura ōda jiken) have been recounted (and sometimes embellished) in
numerous government inquests, Japanese memoirs, document collections, and
overviews of the Wushe uprising.2 A popular 2001 comic book reconstructed the
dispute for a general audience of post–martial law Taiwanese, while the block-
buster 2011 movie Seediq Bale dramatized it in a multilingual feature film.3 There
43
44     The Anatomy of a Rebellion

are conflicting facts in these accounts. We are left to wonder: What type of meat
bloodied Tadao’s hands? How many Japanese patrolmen visited the wedding? How
many bottles of wine did Mona offer to Sugiura? Did Mona jump in and help
assault Yoshimura, or did he break up the fight? But regarding the main sequence
of events, all accounts agree: a wedding took place on October 7; a fight between
Yoshimura and Tadao erupted; and Mona Ludao’s attempt at informal mediation
was rejected in favor of an administratively determined punishment.
From Mona’s perspective, Sugiura’s recourse to superiors constituted a rejection
of Mona’s chiefly authority to resolve disputes on the basis of his status as a revered
Sediq headman. At the same time, the administrative hand-off subordinated
Sugiura; it was an admission that he was not Mona’s equivalent, a “local Japanese
chief ” with wide discretionary powers. This indeed was a momentous change, and
it signaled the termination of rule by the outcasts of empire on both sides of the
“savage border.” The Yoshimura beating incident exemplified the regime’s efforts to
dispense with an intermediary layer of quasi officials at its geographic extremities.
In the new order, the authority of Japanese officials and indigenous leaders
would be derived solely through competition and recruitment through the public-
school system and the police bureaucracy. At least theoretically, this new hierar-
chy, on both sides of the ethnic divide, reached all the way up to the emperor in
Tokyo, thereby unifying the island administratively—almost.

W E T D I P L OM AC Y A N D E A R LY JA PA N E SE C O L O N IA L
RU L E

Six decades prior to Yoshimura’s pointed refusal to drink with Tadao Mona, out-
siders with business on the edges of Qing territory in Taiwan accepted alcoholic
beverages from indigenes and stuck around to imbibe them at close quarters. We
can think of such transactions as instances of “wet diplomacy.” Throughout the
island, alcohol was consumed as a liquid, but in northern Taiwan, the preferred
method was called “conjoined drinking” (gōin), which entailed contact with an
interlocutor’s saliva (see figure 4). The term “wet diplomacy” is therefore descrip-
tive of a particular type of activity, but it also denotes an ideal-typical form of
human interaction. The phrase is not used in this study to distill a particularly
“Japanese” form of sociability but rather is adopted to account for the frequent and
repetitive descriptions of diplomatic drinking in the source material, and for the
insistence of historical actors themselves that the relationships thus consecrated
were personal and not abstract or contractual. This ideal type contrasts with durai
(dry) interaction, which is fungible. Wet bonds must be renewed with periodic
expressions of fealty, while dry relationships are rule governed and attach to social
roles or official ranks rather than individuals.4
In the late 1890s and early 1900s, the early years of formal Japanese rule in
Taiwan, wet diplomacy was the coin of the realm along Taiwan’s savage border. It
Figure 4. Two Atayal men engaged in “conjoined drinking,” near Wulai,
Taiwan, ca. 1900. The man on the right is Bato Watan of Rahao. Narita Takeshi,
Taiwan seiban shuzoku shashinchō (Taipei: Narita shashin seihanjo, 1912), 39.
46     The Anatomy of a Rebellion

evinced continuity with the precedent-setting practices of the Taiwan Expedition

in the 1870s. In both instances, Japanese officials, merchants, and scholars entered
into personal relations with indigenes to secure forbearance, goods, labor, or guid-
ance. In the course of these transactions, indigenous merchants, chiefs, and inter-
preters established bonds of trust with particular outsiders. These relationships
frequently involved intermarriage (see chapter 2). In the wet diplomatic context,
favoritism and loyalty were expected—in fact, they were the whole point.
Indigenous leaders sought access to Japanese goods and political patronage to
gain advantage over local rivals or to shore up authority within their segmented
polities. Therefore, from their perspective, the notion that all Taiwanese would
be treated equally as children of the emperor was anathema. This Japanese policy
was formally known as “nondiscrimination” (isshi dōjin). Under it, all loyalty and
obligation were directed toward a single apex, the emperor—thus making isshi
dōjin structurally analogous to modern nationalism, which also channels loyalty
upward as it postulates horizontal fraternities of juridically equal and interchange-
able subject-citizens. Indigenous Taiwan’s small-scale, nonstate polities were inim-
ical to the instantiation of isshi dōjin, however, because they were held together
by the politics of redistribution. Here, chiefly authority was augmented by pub-
lic displays of favoritism vis-à-vis other indigenes and successful warfare against
neighboring enemies.
Japanese isshi dōjin proclamations promised impartiality for all Taiwanese but
were nonetheless premised on Japanese preeminence. That is to say, even the most
fair-minded Japanese officials believed that indigenes were savages squatting on
the emperor’s land. However, as long as chiefs were receiving gifts, publicly engag-
ing in joined-mouth drinking with Japanese colonists, or themselves apportioning
banquet victuals, the vagaries of language allowed indigenous leaders to convert
wet-diplomatic encounters into political capital vis-à-vis followers and rivals. But
in the long run, the logics of centralized administration and bureaucratic rational-
ity were incompatible with the protocols of wet diplomacy.
To overcome the paradoxes of wet diplomacy and establish a less particularistic
and labor-intensive form of rule, the TGG stationed guard posts along a scorched-
earth trail known as the aiyūsen (see figure 5) to enclose the Atayal, Sediq, and
Truku settlements of northern Taiwan. This denuded landscape provided clear
lines of fire, while it physically divided northern Taiwan into two distinct zones:
one for imperial subjects and another for outcast rebels. From 1903 through 1915,
government forces extended the scorched-earth barrier and then marched it
inward.5 The Indigenous Peoples of Taiwan, as a modern political-­cultural for-
mation, emerged from the ruins of these scorched-earth campaigns. The sur-
render ceremonies themselves, and the nakedly asymmetrical power relations
they enforced, epitomized “dry” interpersonal relations—the lists of demands
Japanese policemen read from their elevated platforms were couched not in the
language of reciprocity and benevolence but as orders for compliance upon pain
 From Wet Diplomacy to Scorched Earth    47

Figure 5. Scorched earth guardline, ca. 1910. Narita Takeshi, Taiwan seiban shuzoku
shashinchō (Taipei: Narita shashin seihanjo, 1912), 163.

of punishment (see figure 6). The trade-starved, bombed-out, and battered indig-
enous delegations were summoned to these spaces as supplicants. To resume trad-
ing relations and stop the shelling, they were forced to surrender their weapons
and listen to lists of conditions enforced by carrot-and-stick methods, rather than
renewed displays of friendship, as called for in wet diplomacy (see figure 7).
Yet after the highlands were largely disarmed, wet diplomacy did not dry up.
The guardline campaigns had, in fact, exhausted the Taiwan Government-General.
In the late 1910s through early 1930s, state-society relations were left in the hands
of rural district police officers, their indigenous wives, and local chiefs such as
Mona Ludao. These districts remained outside the tax base and were labeled “spe-
cial,” “savage,” or “aborigine.” I will refer to this phase of Japanese rule as the era of
“native authority,” following Mahmood Mamdani.6
Wet diplomacy solved several problems for the Japanese, as it did for the Qing
in earlier times: with a meager budget, the government-general established itself
in the highlands despite linguistic and cartographic ignorance. From the strategic
choke points thus established, information could be collected and relationships
forged. However, wet diplomacy hindered the efficient and reliable extraction
48     The Anatomy of a Rebellion

Figure 6. Surrendered Atayal at a stopover in Jiaobanshan, ca. 1910. Endō Hiroya, ed., Banhi
tōbatsu kinen shashinchō (Taipei: Endō shashinkan, 1911), n.p.

of resources from the upland forests, and it blurred the boundaries of state sov-
ereignty along the so-called “savage border.” This chapter plots the history of
indigenous-Japanese relations around the forced march from wet diplomacy to
scorched earth, and then from scorched earth to native authority. It further posits
an isomorphism between wet diplomacy and the low-velocity milieu of tributary
interstate relations. By the same token, it will argue for a congruence between dry
human relations and the horizontal integration of space that is presupposed by the
regime of rule from a distance. This latter mode of governance, in turn, sought to
facilitate the generation, accumulation, and funneling of surplus wealth to state
coffers.

T R E AT Y- P O RT TA I WA N A N D T H E M U DA N
V I L L AG E I N C I D E N T

As with the 1930 Yoshimura beating incident, the 1871 shipwreck of the Ryūkyūans
in Taiwan could have been resolved locally, on the spot. In each case, recourse
to central authorities occurred in the context of increased volumes and veloci-
ties of resource extraction in these imperial peripheries. Such pressures catapulted
 From Wet Diplomacy to Scorched Earth    49

Figure 7. Weapons captured from Atayal peoples during Sakuma Samata’s five-year plan,
photographed at the Taiwan Government-General Museum in Taipei, ca. 1910. Endō Hiroya,
ed., Banhi tōbatsu kinen shashinchō (Taipei: Endō shashinkan, 1911), n.p.

otherwise unexceptional events to historical turning points. The 1930 Wushe

uprising was set off by the Japanese government’s heightened demand for timber
in central Taiwan. Yoshimura himself was brought to Wushe to supervise a Sediq
labor force in felling and hauling gigantic cedars, at fixed wages, to complete a
series of public works projects. Yoshimura was not locally embedded, as some
of his superiors were, but was a specialist hired to accelerate the process of forest
exploitation. When his tone-deaf rejection of Tadao’s entreaty caused a ruckus, his
supervisor Sugiura construed the wedding brawl as an assault on imperial author-
ity and sent his report up the chain of command, escalating matters beyond the
point of no return. Six decades earlier, a longer chain of events, with a larger cast
of characters and an even higher body count, presaged a similar set of problems at
the outer margins of state-administered territory in Taiwan.
The increase in Asia-bound commerce that followed the Opium Wars (1839–42)
turned the Taiwan Strait into a major thoroughfare in the China trade. After
two treaty ports were opened there in 1860, Taiwan’s “poorly charted and still-
unlighted coasts” and rocky shallow waters became a graveyard for boats and their
crews. One tally estimates that about “150 foreign vessels  .  .  . foundered in the
50     The Anatomy of a Rebellion

vicinity of the Taiwan coast between 1850 and 1869 of which over thirty were plun-
dered or burned.”7 Another calculates that at “least forty-six Western merchant
ships” disappeared or were destroyed from 1861 through 1874.8 Most infamously, in
December 1871 near Bayao Bay, Taiwanese murdered 54 shipwrecked sailors from
Miyakojima (Ryūkyū Islands / Okinawa), an act that occasioned the invasion of
over 3,000 Japanese troops beginning in May 1874 (see map 3).
Most historical accounts of the 1871 shipwreck are based on the recollections of
a Japanese official and two Miyakojima survivors. Oral testimonies of the protago-
nists, their rescuers, and local officials were jotted down in 1872 and subsequently
combed over by generations of scholars in Taiwan, Japan, the Ryūkyū Islands, and
beyond. Chou Wan-yao summarized this scholarship in a recent review article,9
while Miyaguni Fumio has republished the primary-source materials.10 Ōhama
Ikuko has mined archives in Taiwan and Japan (including Okinawa) and recon-
structed the events with great care.11 In addition, local historian and native of
Kuskus village Valjeluk Mavaliu has collated written evidence with his geographi-
cal and genealogical knowledge,12 while Mudan resident and oral historian Gao
Jiaxin (Lianes Punanang) has interviewed a dozen or so Paiwan elders (including
Valjeluk Mavaliu),13 to provide a local perspective.
The drama was set in motion by the November 30, 1871, departure of four trib-
ute ships from the castle complex of Shuri (near Naha), on the island of Okinawa.
The ships were returning to Miyakojima and the Yaeyama Islands in the Ryūkyū
Kingdom. Within sight of Miyakojima, the ships were blown off course and suc-
cumbed to a typhoon on December 12, 1871. One of the ships from the Yaeyama
Islands was lost forever, but another landed on Taiwan’s western coast. The survivors
of the latter wreck made it back to Naha through the good offices of Qing officials. Of
the two Miyakojima ships, one made it safely back to its port of origin, but the other—
the famous one—capsized off the east coast of southern Taiwan, near Bayao Bay.14
There were sixty-nine passengers on the ill-fated Miyakojima ship. Three of
them perished trying to get ashore. The other sixty-six made landfall five days after
the horrible storm. Soon after, they encountered two Chinese men who reportedly
warned them away from traveling inland among dangerous Paiwan peoples. It is
difficult to know what transpired at this juncture. The survivors’ deposition indi-
cates that the sixty-six Ryūkyūans were robbed by the Chinese and decided to part
ways with their hosts. They spent the night in makeshift outdoor lodgings and set
out the morning of December 18. As they wandered westward, they encountered
men with large earrings and distended earlobes, presumably Paiwanese.
The Ryūkyūans followed the Paiwanese to a settlement of fifteen or sixteen
thatched homes, Kuskus, where the lost mariners were given water and food and
were put up for the evening. The provision of water by Kuskus residents, for them
at least, symbolized an offer of protection and friendship, according to Valjeluk
Mavaliu.15 The deposition claims that during the night, they were robbed again
by their Kuskus hosts. In the morning, a departing group of hunters ordered the
Map 3. Taiwan, the Ryūkyū Islands, the Langqiao Peninsula, Mudan, and Satsuma, in
East Asia.
52     The Anatomy of a Rebellion

Ryūkyūans to stay put and offered to return with enough game to provide a feast.
However, the presence of so many armed men, coupled with the rumors of head
hunting that had greeted them on shore two days earlier, impelled them to make a
break for it while the hunting party was absent.
Dozens of the fleeing Ryūkyūans found shelter in the home of seventy-three-
year-old Hakka trading-post operator Deng Tianbao (named “Old Weng” in
the deposition16) and his thirty-year-old son in a nearby settlement of five or
six houses. Most Taiwanese Hakka (Kejia) people immigrated from Guangdong
Province. Throughout Taiwan, the Hakka established settlements in foothills and
acted as commercial conduits between Austronesians in the mountains and Han
people of Fujianese extraction (Hok-los) in the plains.17 Valjeluk Mavaliu writes
that Deng was married to a Paiwanese woman named Utjau,18 a common practice
for cross-border merchants in this area.
That same day, Paiwan men caught up with the Ryūkyūans. They forcibly entered
Deng’s compound. Dozens of Ryūkyūans were dragged out and slaughtered in
a courtyard; others were caught in flight. Nine Ryūkyūans avoided detection in
Deng’s home, and another three escaped (and were captured by other Paiwanese
subsequently). The other fifty-four were killed in the melee. The next day, the nine
survivors were removed to the much larger Hakka settlement of Poliac (Baoli) and
put under the care of the village head, Yang Youwang—Deng’s son-in-law. Poliac
was an ammunition depot, trade junction, and place of arms manufacture. Hakka
merchants transshipped powder, shot, and guns to Paiwan buyers inland from this
location (see map 4).
In addition to securing the safety of these nine, Yang also arranged for the ran-
som of the three separated men. Yang Youwang then sheltered these twelve survi-
vors for about forty days, before sending them to the administrative seat of Taiwan
Prefecture (Taiwan-fu, today’s Tainan). The Ryūkyūans were returned to Naha in
July 1872, over seven months after the four ships were hit by the typhoon.19
Although it became a truism among Japanese officials and subsequent chroni-
clers that Paiwanese Mudan villagers murdered the seafarers, residents of Kuskus,
today known as Gaoshifo, were the assailants.20 It has been proposed that the vio-
lence resulted from the Ryūkyūan visitors’ ignorance of guest etiquette (they ate
and ran), or that the captors could not find buyers to pay ransom and thus killed
them.21 Gao Jiaxin’s oral histories provide another plausible explanation. Sixty-six
adults, unable to communicate in the local languages, walked into Kuskus farm-
land and began taking food and trampling over hotly contested village bound-
aries. On top of these nuisances, efforts to feed and shelter dozens of strangers,
who brought no provisions and could not make themselves understood, strained
Kuskus resources and hospitality. Finally, the intruders were killed as punishment
for their multiple misdeeds.22 The best estimates put the population of Gaoshifo at
250 at the time.23 Therefore, one can imagine that caring for sixty-six guests would
have been a major, to say nothing of disruptive, undertaking in this sparsely popu-
lated area of rural Taiwan.
 From Wet Diplomacy to Scorched Earth    53

¯
( Fenggang
!

!
(
Mudan

Bayao
( Kuskus
! Bay
!
(
Stone Gate

Checheng
Langqiao (
!!
(
Bay Langqiao
!
( ( Poliac
!
Siaoliao

Tuilasok
Sabaree !
(
Hengchun !
(
!
(

( Koalut
!

Roads
5 Miles

Map 4. The Langqiao Peninsula, ca. 1874.

Whatever the motivations of the people who killed the Ryūkyūans, the slaugh-
ter did not immediately raise eyebrows in Japan. While a few officials who hap-
pened to be in Beijing or Naha in mid-1872 got wind of the massacre, it did not
become a full-blown international incident until April 1874. The 1872 repatriation
of the twelve survivors from Poliac to Fenggang to Taiwan-fu to Fuzhou appears
54     The Anatomy of a Rebellion

to have been by the book. In fact, the Qing system for repatriating Ryūkyūan ship-
wreck victims on the shores of China (including Taiwan) settled 401 incidents over
the course of the seventeenth to the nineteenth centuries. Qing policy throughout
its realm was to provide food, clothing, and ship repairs for stranded mariners.
Reciprocal agreements with neighboring powers returned Chinese shipwreck sur-
vivors to the Qing realm from Japan, Korea, Vietnam, and the Ryūkyūs in over 700
incidents during the same period.24
Edwin Pak-wah Leung has argued that this regional repatriation system worked
without visible complaint until the arrival of the Western-centered treaty-port
regime in the late nineteenth century. He notes that the Ryūkyūan court itself did
not request Japanese officials to intercede on its behalf regarding the fifty-four vic-
tims of the 1871 shipwreck. It was Japanese officials visiting Okinawa who lobbied
for the invasion. The Ryūkyū monarch Shō Tai (Shang-tai), instead of requesting
Japanese aid, sent a reward to Chinese officials in Fuzhou after the return of the
twelve survivors from Taiwan. In fact, Ryūkyūans had been stranded on Taiwanese
shores on fifty-three recorded occasions between 1701 and 1876 and were repatri-
ated by the Qing in each instance.25 Because such incidents were unexceptional
and because Japan’s expedition to avenge the deaths of the fifty-four Ryūkyūans
occurred over two years after the event, the question of the timing of the 1874 inva-
sion has been of great interest to historians. Quite rightly, scholars have looked to
Japan’s national politics and diplomatic history for answers, decoupling the May
1874 invasion from the December 1871 shipwreck.
At the time of the Meiji Restoration (1868), state-to-state relations within East
Asia had yet to become international and are more properly termed interdynastic,
or the “tributary system.” To assure that diplomacy demonstrated the sovereign’s
centrality and paramountcy, tributary courts received official delegations from
other states at specified times and for periods of limited duration. The missions
followed carefully prescribed routes, performed minutely orchestrated guest rit-
uals, and returned home. Much has been made of the cultural chauvinism and
hierarchical ordering that informed diplomatic practice under this system. In
this analysis, however, the important hallmarks of tributary forms are not their
validation of a generalized, transhistorical sense of Chinese superiority but rather
more portable features that can be extended to other premodern regimes. Namely,
tributary relations were tailored to the specificity of each diplomatic transaction.
Guest rituals in this system were adjusted for differences between supplicant states
or even for the size of missions. Tributary relations also eschewed the notion that
interstate relations be put on a permanent, contractual, and routinized basis. In
their emphases on the necessity for periodic renewal and forthrightly particular-
istic forms,26 Qing, Tokugawa, and Joseon guest rituals bore a family resemblance
to the “wet diplomacy” described above.27
According to Takeshi Hamashita, “modernity” did not overpower encrusted
East Asian “tradition” with the coming of the Western treaty ports at the conclusion
 From Wet Diplomacy to Scorched Earth    55

of the Opium Wars. Rather, the Atlantic powers overlaid a new set of institutions,
practices, and personnel onto existing Chinese, Japanese, and Korean maritime
trade circuits. Independent of the new trade, intra-Asian commerce experienced
its own growth spurt in the nineteenth century. East Asian politicians and mer-
chants themselves could be eager proponents of trade-volume expansion, tariff
standardization, harbor improvements, and efficient customs agencies. In a word,
Hamashita demonstrated that the routinization of cross-border commercial prac-
tices within Asia, rather than threatening dynasties, could, in fact, strengthen cen-
tral authority.28 The ideological divide in dynastic circles was not so much between
moderns and traditionalists but about means to increase, regulate, or reject higher
volumes of foreign trade in order to protect sovereignty.29
These debates about how to meet the challenges and opportunities presented by
the arrival of industrial capitalism in East Asia raged in Japan during the interval
between the December 1871 killing of the Ryūkyūans in Deng Tianbao’s court-
yard and the May 1874 Japanese invasion of Taiwan. Simultaneously, the Japanese
government struggled to assert authority throughout its three major islands of
Honshū, Kyūshū, and Shikoku. Rural rebellions against increased taxes, compul-
sory education, and military conscription, as well as the “chastise Korea debate”
(seikanron) at the higher levels of government, formed the backdrop for the 1874
invasion. Saigō Tsugumichi’s occupation of southern Taiwan, therefore, not only
aimed to clear the Langqiao Peninsula of parasites who leeched wealth from the
global system but also advanced the Meiji project to stamp out regionalism and
resistance to central authority within Japan.30 As Norihito Mizuno put it, the var-
ied projects subsumed under the Taiwan Expedition—to colonize the Langqiao
Peninsula, to siphon off samurai discontent with a foreign adventure, and to assert
exclusive sovereignty over the Ryūkyūs—all conformed to a single logic: the young
Meiji state’s heightened threat perception vis-à-vis its industrialized competitors
in the arena of nation-state imperialism.31
It will be recalled that the lack of succor for shipwrecked American whalers
at the hands of Japanese local officials energized the U.S. government to pursue a
“water and wood” treaty with the Tokugawa Shogunate (1603–1867) in 1853. The
1854 Treaty of Peace and Amity, signed between the United States and Japan on the
heels of two U.S. naval missions led by Commodore Matthew Perry, was aimed at
a Japanese government actively hostile to shipwreck victims, and by extension,
to untrammeled access to the ocean’s natural resources.32 America’s whalers were
extending their range deeper into the Pacific, making the problem of shipwrecks
on Japan’s shores a persistent one if Americans were to keep increasing oil con-
sumption. Under threat of naval bombardment, the Tokugawa Shogunate signed
the 1854 treaty with the United States and a subsequent 1858 treaty to open trade
and treaty ports.
The government in Edo managed to dodge the bullet of foreign invasion,
but its actions met with dire consequences. After the shogunate signed these
56     The Anatomy of a Rebellion

unpopular instruments, civil wars erupted throughout the realm, in addition to a

wave of political assassinations. When rebellious Satsuma (a southwestern domain
in Kyūshū) shore batteries fired on British ships in 1863, also to protest Japan’s new
policies of openness, British naval bombardment nearly decimated the castle town
of Kagoshima.33 Outgrowths of this civil strife, brought on by a central govern-
ment’s overreach to satisfy the demands of mercantile powers, fomented the col-
lapse of the Tokugawa Shogunate in 1868.
As the Tokugawa regime was teetering on its last legs, U.S. consul to Amoy
(Xiamen) Charles LeGendre reprised, on a smaller scale, the 1853–54 U.S. missions
to Tokyo in Langqiao, Taiwan. In the fall of 1867, LeGendre concluded a verbal
agreement with the Langqiao headman Toketok to protect shipwrecked sailors
from assault, robbery, and ransoming on Taiwan’s south coast.34 The agreement,
like the 1854 Japan-U.S. treaty, reflected the efforts of maritime powers to make the
seas safer for greater volumes and velocities of commerce. The LeGendre-Toketok
negotiations, like the U.S.-Japan talks of the late 1850s, averted armed confronta-
tion in the short run, while they placed unbearable strains upon Toketok’s ability
to hold together a loose confederation of subordinate polities. Toketok’s problems,
in fact, mirrored those of the Tokugawa government vis-à-vis Western gunboats.

L E G E N D R E E X T E N D S T H E T R E AT Y- P O RT SYST E M T O
L A N G Q IAO

In April 1867, LeGendre made his first journey to Taiwan to investigate an

American shipwreck that became known as the Rover incident. The consul’s visit
ignited a seven-year fuse that exploded with the landing of four Japanese warships
in Langqiao in May 1874. The barque Rover bottomed out on Taiwan’s south cape
in March 1867. After it capsized, the Rover’s captain, his wife, and his crew were
murdered on shore by Koaluts (a Paiwan tribe) while trying to secure help. The
one survivor of the attack, a Chinese man, reported the incident to authorities
in Gaoxiong after making his escape. Thereafter, a British steamer left the same
port to investigate but was turned back by the eruption of musket fire from the
camouflaged and elevated positions of the Koaluts. LeGendre headed to the south
cape for a ten-day tour.35 He failed to find survivors or remains since he could not
secure local guides.
While LeGendre plotted his next move in Xiamen, U.S. vice admiral H. H.
Bell launched a frontal attack from the location of the Rover’s wreck. Ignoring the
advice of LeGendre, Bell skipped a stop at the port town of Siaoliao. LeGendre had
suggested that Bell activate the “mediating networks” of Hakka, Hok-lo, bicul-
tural, and Paiwan villagers that connected the port of Siaoliao to the protected
and isolationist Koalut polity.36 Going it alone instead, Bell successfully landed,
but soon after, one Lieutenant Captain Mackenzie was shot through the heart by
a Koalut marksman during an uphill infantry charge against unseen defenders.
 From Wet Diplomacy to Scorched Earth    57

The American warship pulled anchor and shelled the Koaluts from a safe distance,
apparently inflicting great damage. But without access to kin, trade, and alliance
networks, Bell could not locate survivors or discover who specifically had killed
the crew of the Rover.37
As a treaty-port consul in Xiamen, which subsumed Taiwan on the American
diplomatic organizational chart, LeGendre took the Langqiao Peninsula’s opacity
as a professional and personal affront. A letter to LeGendre from the Taiwan cir-
cuit attendant regarding the Rover referred to Articles 11 and 13 of the 1858 Treaty
of Tianjin to state “that whenever within the jurisdiction of the Emperor of China,
anyone shall molest Americans, the military and civil authorities must, on hear-
ing of the same, try to punish the authors.” However, the note added—and this
became the sticking point for years to come—“the Savage country does not come
within the limits of our jurisdiction, and our military force is not able to operate
in it.”38 Finding this response unsatisfactory, LeGendre took the matter all the way
to the top (short of an imperial audience). In person, he proposed to the governor-
general of Fujian and Zhejiang, Wu Tang, that the Langqiao Peninsula be occupied
by Chinese inhabitants and garrisoned permanently by Qing forces.39
Langqiao’s meager population—about fifteen thousand people, of whom three
thousand were indigenous, seven thousand or so bicultural, and another four or five
thousand Hakka and Hok-lo40—and its variety of unknown (to Chinese officials)
languages and inaccessible terrain41 did not commend the extension of civil and
military administration to this remote corner of empire. Understandably, Wu did
not formally consent to LeGendre’s proposals, but neither could he ignore them. It
is testament to the force of British and French arms and the demonstration effects
of the 1860 sacking of Beijing42 that LeGendre could obtain a meeting with Wu,
who sat at the apex of an administrative unit that governed perhaps thirty million
Qing subjects.43 It was surely in this larger context that the American’s numerous
requests, if not honored, were generally entertained.
Wu Tang’s successor, Ying Gui, therefore allowed LeGendre to accompany
a military tour of Langqiao led by General Liu Mingdeng in September 1867.44
LeGendre recounted travel to Langqiao from the prefectural capital, Taiwan-fu,
as an arduous affair. Although ranking Qing officials extended LeGendre diplo-
matic courtesies, transportation was halting. LeGendre rode on a sedan chair in
proximity to General Liu, but his journey was delayed by bad weather and slow
communications. As one moved south, the last Chinese city on the coast, Fangliao,
was literally the end of the road. To go beyond, Liu’s troops (who numbered about
five hundred), built a forty-mile path—supposedly at LeGendre’s urging—before
the expedition could proceed (see map 5).45
LeGendre’s contemporary accounts of his progress through Taiwan, which he
appears to have kept scrupulously accurate within the limits of his knowledge, do
not suggest that Taiwan was divided into two separate zones, Chinese and indig-
enous (see figure 3). He would adopt this divisive view later, after his trips to the
 ¯ TAOYUAN !
( Taipei
TAIPEI

!
( ( Wulai
!
XINZHU Dakekan
( Yilan
!

MIAOLI
YILAN

TAIZHONG

NANTOU Wushe
!
(
!
(
Puli

JIAYI HUALIAN

TAINAN

!
(
Taiwan Fu
TAIDONG

Gaoxiong !
(

AHOU !
(
Fangliao
Mudan
!
(

25 Miles

Map 5. Principal administrative centers, contact zones, and political boundaries mentioned in
this book. The Taiwan Government-General’s 1909–20 prefectures are in all capitals. The lighter
area of the map is TGG-demarcated Aborigine Territory (ca. 1905–45).
 From Wet Diplomacy to Scorched Earth    59

island ceased and his knowledge was, in a sense, weaponized for use in the dip-
lomatic arena. As Bruce Greenfield wrote of the Lewis and Clark journals, expe-
dition accounts subsume numerous genres of writing, each with its own source
of authority. For LeGendre the amateur scientist and aspiring author, recording
surfeits of local detail established his credibility and retained reader interest.46 At
the same time, LeGendre wrote in a combative, legalistic style to prosecute his
official duties. In this register, later complicated by his advocacy of a Japanese
invasion, his descriptions of Taiwan lost much of their veracity and took on a
hortatory aspect.
During this 1867 procession and follow-up tours in 1869 and 1872, LeGendre
entered a region of marked ethnic diversity and hybridity; its indigenes—predom-
inantly Paiwan—presented outsiders with a confusing array of settlements, con-
federations, and diplomatic forms.47 The principle Paiwan people that LeGendre
dealt with are known as the Eighteen-Tribe Confederation. The eighteen tribes
were considered to have been under Toketok’s leadership in Western sources,
though the bases of that leadership and what it entailed were murky to contempo-
raries and remain an object of research and speculation.
These eighteen tribes (including the Kuskus who killed the Miyakojimans and
the Koalut who killed Mackenzie) were of Paiwan ethnicity. The primary political
and residential unit was a village, with populations ranging from seventy-five to
four hundred people; these villages were mostly located in the mountains beyond
the western coastal plain. Paiwan villages were militarily predominant in Langqiao
before 1875. They collected taxes and tribute from Han settlers. Among the Paiwan
lived a smaller population of Amis people, migrated from Taidong, and people of
plains indigene descent known as Makatao. Four lineages—of Puyuma or Beinan
ancestry—produced the locally paramount “big stride chiefs” (ōmata tōmoku)
such as Toketok and Isa. The four noble clans of lower Hengchun viewed them-
selves, and were treated by Paiwans, as distinctly “foreign.” Later Japanese eth-
nologists referred to the lineages as Skaro peoples, or Paiwanized Puyuma. The
four Skaro clans were revered as descendants of militarily powerful invaders; the
southern clans that Toketok and Isa headed were also feared as efficacious and
dangerous magicians.48
Villages, such as Kuskus, Sabaree, and Tuilasok, were located less than forty
meters above sea level and sat astride well-traveled river corridors. Others, such as
the feared Koaluts and Mudans, were more geographically isolated. Hakka peoples
in Langqiao inhabited trade junctions between the plains and the mountains and
frequently intermarried with Paiwan peoples. Today Hakka people are normally
grouped with Han Chinese, but outsiders such as LeGendre and his interpreter,
William Pickering, considered them a distinct people. Most remarked upon was
the village of Baoli (Poliac), proximate to the two largest coastal Han settlements,
Checheng and Siaoliao. This Hakka village also guarded the entrance to the valley
that led to Kuskus (see map 4).
60     The Anatomy of a Rebellion

Less well-defined groups, variously called “plains savages” or “half-castes,” are

less easily described by ethnic labels. I will refer to them as bicultural, although
many of them were tricultural. The man known to history as Miya belongs to this
category. He was familiar with Paiwanic and southern Chinese languages and
was related through kinship, commercial, and professional networks to a Paiwan
confederation chief, a Qing general, and the headman of Siaoliao town. One of
Langqiao’s few ports, Siaoliao was a Han-dominated coastal village that LeGendre
referred to as “half-caste.” Miya would become LeGendre’s factotum in Langqiao
and then go on to help the Japanese establish military operations there in May 1874.
Han peoples known as Hok-los, mainly from Fujian, were the majority pop-
ulation of nonindigenous people in Taiwan. They also resided in Langqiao, and
they were known to be in competition with Hakka settlements, if not engaged in
actual armed conflict. As LeGendre and Pickering, his interpreter, were quick to
point out, however, common cause, such as the arrival of Qing soldiers or defense
against Paiwans, could bring these two groups together as temporary allies.49
The Paiwan peoples of the extreme south, who most concerned General Liu
Mingdeng’s Qing scouts in September 1867, maintained their military preponder-
ance with supplies of shot, powder, and metal, purchased from Hakka traders in
exchange for mountain products (furs, meat, medicinal ingredients, loot from
shipwrecks). Further north near Fenggang, where Japanese troops would camp
in mid-1874, Han settlements paid taxes or tribute to Paiwan villages.50 So-called
“aborigine rents” were relatively light; they did not fund infrastructure but were
paid out as protection money. Villagers also paid taxes to Paiwans farther south,
with the important exceptions of larger settlements such as Checheng and Siaoliao.
To summarize, for people who hoped to conduct business in Langqiao circa
1870, it was imperative to know how particular settlements were connected
together and to secure the aid of people who could mediate the appropriate con-
stellations of polities suited to a particular type of errand. Administrative ranks
such as LeGendre’s were of little help unless they were attached to armies on the
move, generous caches of gifts, or personal connections to embedded locals.
Returning to LeGendre’s mission: A little over two weeks after the U.S. con-
sul’s September 6, 1867, arrival, he and General Liu entered the town of Langqiao
behind a color guard of fifty flag bearers. The townspeople prepared shrines and
memorial tablets for the arriving mandarin and foreign dignitary. The locals pros-
trated themselves accordingly. Eight sedan bearers carried LeGendre. The consul
was not bothered by the fact that he could neither read the ceremonial calligraphy
produced for his benefit nor understand conversations between Liu and the recep-
tion committee. Separating himself from the seemingly unwholesome conditions
of the Chinese town, LeGendre camped outside Langqiao with an interpreter and
a few attendants. From there, he ran a shadow operation in parallel to the more
expansive Qing effort, sending emissaries to find Toketok, the man whom General
Liu had planned to negotiate with, as well.51
 From Wet Diplomacy to Scorched Earth    61

LeGendre believed that Liu’s large armed retinue would impede negotiations,
so he ended up meeting Toketok on neutral ground with five interpreters and a
guide. The October 10, 1867, tête-à-tête lasted about forty-five minutes and ended
abruptly. Toketok brought about two hundred men under arms; he had mustered
some six hundred for a previous meeting in Poliac, even though his home village
of Tuilasok had a population of less than one hundred.52
From LeGendre’s perspective, Toketok was willing to negotiate because
LeGendre himself and the war hero Lieutenant Captain Mackenzie (shot by
a Koalut man) were brave men whom Toketok could respect. Time and again,
LeGendre emphasized the personal nature of his relationship with Toketok and
praised the frankness and simplicity of the Paiwan leader. He also claimed that
Toketok would not deal with Han people or Qing emissaries because of their lack
of courage.53 During the 1874 expedition itself, Toketok’s successor as paramount
chief of the eighteen tribes reportedly expressed a reluctance to deal with anyone
but LeGendre,54 so important was the personal connection forged during the 1867,
1869, and 1872 visits to Langqiao.
It is impossible to weigh the extent to which personal ties such as the ones
articulated by LeGendre mattered to the Langqiaoans. LeGendre’s linguistic
transactions occurred through interpreters, usually more than one. Moreover, he
could not read or understand Chinese dialects. Thus the speech he imputed to
Toketok, or that which LeGendre’s successor, Commander Douglas Cassel, heard
from Toketok’s successor, Isa (see below), must be taken as approximations. While
Toketok may have resented Qing officials and found common cause with the frus-
trated treaty-port official LeGendre, and while it was undoubtedly a bold stroke to
meet Toketok with such a small escort, LeGendre’s diplomatic offensive had at its
back a massive Qing army and at its front a seasoned commercial interpreter who
laid much of the groundwork for the meeting. Moreover, as shall be explored later
in this book, Toketok was able to ritually incorporate LeGendre into a sphere of
exchange and tribute by orchestrating the receptions of gifts and dispositions of
banquets in accord with local forms (see chapter 3).
William Pickering recalled several visits to Toketok as an emissary from
Langqiao residents and asserted that LeGendre’s meeting was but an afterthought
and a ceremonial ratification of negotiations largely completed by Pickering him-
self. More importantly, while LeGendre was discussing the finer points of shore
safety with Toketok on October 10, over five hundred Qing troops, spies, and
interpreters were fanning out across the peninsula threatening to exterminate the
Paiwan and advertising the fact that many more Qing soldiers lay in reserve on the
mainland.55 LeGendre’s introduction to the region, it should be recalled, was his
grand entrance to Langqiao with General Liu at the back of a color guard and in
the front of an army.
Despite this general atmosphere of intimidation, the actual terms of the “treaty”
(which was never acknowledged by LeGendre’s superiors in Beijing, despite his
62     The Anatomy of a Rebellion

repeated efforts) suggest that Toketok was not overawed by the Qing troops. Some
of the stipulations resemble the 1854 U.S. treaty with Japan: shipwrecked mariners
were to be provided with basic provisions and shelter and would be brought ashore
at only specified locations. However, there were major differences. For example,
the 1854 treaty with Japan stipulates that U.S. sailors could not be confined at close
quarters like the Dutch in their compound in Nagasaki but must be allowed to
move about freely. It also leaves open the possibility of putting U.S. consuls at one
or both “wood and water” stations (Shimoda and Hakodate). In contrast, Toketok’s
agreement restricted foreigners to coming ashore in a very specific location and
did not allow them to “visit the hills and villages” of Paiwans. When Toketok revis-
ited the 1867 discussions, he also asked for the right to kill visitors without red flags
(unless they were victims of nautical accidents).56
It appears that some sailors were actually saved thanks to the agreement with
Toketok. However, LeGendre’s personalized brand of treaty making was insuf-
ficient to effect his stated aims. Upon returning to Langqiao in 1869 and 1872,
LeGendre had to reactivate local networks anchored by Miya—who was not a sal-
aried Qing official—just to locate Toketok. Then he was informed that the treaty
would have more force if it could be renewed annually with personal meetings
between the U.S. consul and the Paiwan headman. In addition, as the celebrated
massacre of the Ryūkyūans revealed, Langqiaoans understood the agreement as
valid only for sailors who could be linked to LeGendre himself (non-Asian mari-
ners). And just as problematic, from a treaty-power viewpoint, was the fact that
Toketok’s promises did not carry much weight beyond Tuilasok, one of the small-
est settlements of the eighteen tribes.57
After LeGendre left Taiwan in late 1867, Governor-General Ying Gui wrote a
memorandum to the Qing central administration in Beijing with recommenda-
tions to keep foreigners out of Langqiao completely. Instead of backing LeGendre’s
proposal to integrate Langqiao more tightly into the treaty-port economy, he
tried to prevent the area from becoming a flash point for diplomatic disputes with
Westerners. Ying Gui also proposed stationing troops and officers and appoint-
ing a few local headmen (Hok-lo, Hakka, and bicultural) to rescue shipwreck vic-
tims after the fact. These representatives and troops would be stationed in coastal
Fenggang and Fangliao, not in Paiwan territory as LeGendre had envisioned.
The Qing plan was premised on an indirect form of supervision.58 In short, these
proposed Qing measures were consonant with the older, low-velocity tributary
mechanisms described above for ameliorating, rather than preventing, incidents
like the Rover.
The indefatigable LeGendre countered again. In an official letter to his Qing
counterpart, he urged that better roads, a large garrison, and monthly displays of
Qing naval force off the south coast—leavened with threats of trade embargoes on
powder, shot, and other necessities—would bring the Paiwan to heel and reduce
the menace they posed to commerce once and for all. Writing in 1869, LeGendre
showed awareness that his propositions would dramatically increase Qing budget
 From Wet Diplomacy to Scorched Earth    63

outlays. And indigenous villages would be forced to deal routinely with soldiers
and settlers under his plan—a sure recipe for bloodshed.
According to the Qing records of the October 1867 proceedings, Taiwan offi-
cials agreed with General Liu that fort construction on the south cape would pro-
voke the Paiwanese (whom they termed “savages”).59 Based on the events of 1875,
we know that the Qing ministers had cause for concern. In the years immediately
following the Taiwan expedition of 1874, Koalut, Saparek, and Mudan raids into
lowland territory killed many, and one follow-up Qing expedition was massa-
cred.60 Even in 1867, without hindsight, it seemed to Qing officials that anything
less than extermination or permanent occupation would add to the disorder that
LeGendre sought to alleviate. As LeGendre himself pointed out, nonindigenous
Langqiaoans who paid taxes to Paiwanese were getting off lightly. Were the Qing
to extend its administration south of Fangliao, the financial burden would fall on
them. Like his American counterparts who negotiated treaties with the Tokugawa
Shogunate, then, LeGendre was urging the Qing to become more activist in ways
that would surely create chaos, violence, and tax increases. The goal, in each case,
was to expand the arena of global commerce in the name of a “general good” that,
at least for the foreseeable future, did not include very many East Asians.61
It was not the case that all residents of Taiwan suffered at the hands of the
activist American consul. For example, the residents of LeGendre’s base of opera-
tions in Siaoliao, where Japanese troops would also set up shop in 1874, eagerly
joined efforts to extend the treaty-port economy’s reach. Wages for guide service,
in addition to transshipment opportunities for the spoils of wreckage and profits
from more legitimate forms of coastal trade, enriched Siaoliao residents in the
1860s and 1870s. Even before the arrival of LeGendre in 1867, the post–Opium War
increase in intra-Asian junk traffic redounded to Siaoliao’s economic benefit.62
When LeGendre received word in early 1872 that Toketok wished to see him
again, it was through Miya (the son of a Siaoliao headman mentioned above) and
his network that arrangements were made. Again the occasion was shipwreck
trouble, this time the London Castle. To assist in this case, LeGendre took the ini-
tiative without an official invitation and started for Taiwan in his capacity as a
“piracy-suppressing consular official.” LeGendre briefly docked at the walled city
of Taiwan-fu to apprise “civil and military authorities of the island” of his mis-
sion. Without personally visiting local officials in the prefectural capital, LeGendre
steamed south and docked at Fangliao on March 1, 1872, with his picked men.
LeGendre expected to find a garrison at Fangliao, in addition to personnel
entrusted with securing castaways from Paiwanese hosts or captors, per his nego-
tiations with Qing officials in 1867 through 1870. LeGendre found instead that the
Qing had attached a single “military messenger” to the local headman. LeGendre
turned to unnamed civilians for aid, but they proposed a service fee that LeGendre
declined to pay, perhaps on principle. The next morning, LeGendre’s team dis-
embarked farther south, near Siaoliao. When the party arrived on March  2, a
U.S. consulate interpreter named James Johnson—himself a native of Xiamen, a
64     The Anatomy of a Rebellion

veteran of the American Civil War, and a resident of New Jersey—sought out a
Siaoliao man named Miya.63 Miya had been LeGendre’s guide to the area in 1867
and 1869, and for his efforts as a member of Liu Mingdeng’s militia in 1867, he “was
awarded the fifth military rank by the Taiwan” circuit intendant.64
Johnson finally located Miya. He and LeGendre then hatched a plan to visit
Sabaree, a Paiwan settlement about an “hour’s walk from” the coast. LeGendre and
Miya hired another guide and “twenty seven Chinese coolies” to haul the required
gifts for diplomacy with the Paiwans. LeGendre described Miya and his other
Siaoliao guide as “half-castes,” by which he meant people conversant in spoken
Chinese (probably the Minnan variant) and a Paiwan dialect.65
A Paiwan man named Isa (also known as Esok or Yeesuk) presented him-
self to Miya and Johnson as the responsible leader in Sabaree, the via point for
LeGendre’s meeting with Toketok. Like Miya and Johnson, Isa had met LeGendre
on a previous occasion. After spending one evening in Sabaree, the U.S. con-
sul, Miya, the Siaoliao guide, James Johnson, and twenty-seven porters walked
another two hours inland to Tuilasok. Today, Highway 200 in Pingdong County
runs through this same corridor; it is about two and a quarter miles from Sabaree
to Tuilasok. In 1872, Tuilasok was the abode of Toketok. In Tuilasok, LeGendre
and his large entourage were made to wait because Toketok had been detained
on matters in a different village. Upon Toketok’s return, animals were slaughtered
for a feast.
LeGendre inquired about the disposition of shipwrecked sailors in the area
since the 1867 agreement. It turned out that Toketok had aided a few groups but
had never heard from Qing authorities about their ultimate safety. LeGendre took
this lack of communication or recompense as a sign of Qing bad faith. LeGendre’s
other reports emphasized that Paiwan settlements were jealous about the entry
of outsiders and that Toketok was contemptuous of Qing officials. James Horn’s
reports, quoted approvingly by William Pickering, painted Toketok as a shrewd
manipulator and ransom taker. It is not clear why LeGendre expected the relation-
ship between Qing officials and a toll-state broker such as Toketok to have been
more cordial, routinized, and intimate. In fact, despite having a Siaoliao network
in place and having met Toketok on two previous occasions, it was no small matter
for the U.S. consul himself to locate Toketok and conduct business.
It was during this visit to Siaoliao that LeGendre first heard about the plight of
the fifty-four Ryūkyūans. Based on his interview with Toketok, LeGendre darkly
hinted that the Chinese themselves were responsible for their deaths because the
Qing government had refused to staff frontier outposts with agents who could
expedite the ransom payments that would have saved the Ryūkyūans.66
The various eighteen tribes, as they were called, had populations of a couple
hundred people—some more, some less. Their economies were of modest scale,
compared to the nations that sent ships halfway around the world. Housing large
groups of strangers who required shelter, food, water, and “supervision” strained
 From Wet Diplomacy to Scorched Earth    65

local resources—to judge from the constant refrain that payment to the hosts was
a sticking point for release. Sensibly, LeGendre urged that regular gift distribu-
tion and timely payments were key to securing the victims of shipwrecks. Oddly,
LeGendre appears not to have considered the probability that large and routine
payments to confederation leaders such as Toketok would have only increased
their independence from the Qing, as well as the Hakka and bicultural settlements
who provided LeGendre himself access to Paiwan representatives.
But if strong and independent chiefs posed threats to the smooth functioning
of a treaty-port maritime economy, so did weak ones. As Toketok explained it him-
self at the 1872 meeting, his authority among the confederated eighteen tribes was
contested and not wholly dependable. At the banquet, LeGendre maneuvered Isa
of Sabaree and Toketok of Tuilasok into explaining their exact political relation-
ship, which turned out to be one of rivalry. As the public face of diplomacy with
outsiders, Toketok made plenty of local enemies by collaborating with ­outsiders.
Accordingly, Qing officials did reward Toketok with gifts and payments to help
him maintain that authority, noted LeGendre. (We will find Mona Ludao facing
the same dilemma in 1930, when Taiwan was under Japanese rule.)67
Having said his farewells to Isa and the assembled Sabaree men at his second
feast of the day, LeGendre headed back to Siaoliao on the coast, never to engage
in palavers, gift exchanges, and toasts with Paiwan political leaders again. The date
was March 5, 1872.68

T H E M E I J I R E S T O R AT IO N A N D T H E 1 8 7 4 I N VA SIO N
O F TA I WA N

Between 1867 and 1872, Charles LeGendre made a total of eight trips to Taiwan.69
As LeGendre was shuttling back and forth from Xiamen to Taiwan-fu and mak-
ing circuits of the island, the Tokugawa Shogunate in Japan was overthrown. The
successor state was promulgated in 1868 with a sixteen-year-old emperor named
Mutsuhito as the figurehead and symbolic apex/center. Mutsuhito is known to his-
tory as the emperor Meiji. A group of samurai, mostly from the old feudal domains
of Satsuma and Chōshū, strove to build a state that could resist the impositions
of industrialized Euro-American powers. Thus, the so-called Meiji oligarchs set
out to collapse the status-group distinctions and regional inequalities sanctioned
by the Tokugawa dynastic state. At the same time, they positioned the nascent
national geobody within a complex regional system at the intersection of Western
international society and the East Asian tributary system.
Almost from the start, Meiji diplomats tried to do away with tributary proto-
col because it diminished the dignity of the Japanese emperor by referring to him
as a mere “king,” a subordinate to the Middle Kingdom (Qing) emperor. Facing
their neighbors, the Meiji oligarchs promoted the new form of state-to-state rela-
tions (see introduction). Japan’s central authority, the Council of State, launched a
66     The Anatomy of a Rebellion

number of ill-fated diplomatic initiatives to the Korean court in the 1870s to abol-
ish older practices, finally imposing a trade treaty in 1876 through the use of force.70
Regarding Japan’s more powerful neighbor, the Qing dynasty, negotiation was
the only way forward. An early treaty was signed with the Qing in 1871. Many in
Japan, however, found it objectionable for retaining the irksome extraterritoriality
clauses of Western treaties with Japan and China. To revise the 1871 Qing treaty,
diplomat Yanaihara Sakimitsu went to Beijing in 1872. There, he read about the
murder of the fifty-four Ryūkyūans in Langqiao from the official gazette, Jingbao.
Yanaihara sent the clipping to Minister of Foreign Affairs Soejima Taneomi in
Tokyo in May 1872. This letter put the incident on the central government’s radar
even before the survivors returned to Naha. After the shipwreck victims were
repatriated on July 12, a visiting Kagoshima man, Ichiji Sadaka, caught wind of
their plights while in Naha and relayed the news to Kagoshima Prefecture gover-
nor Ōyama Tsunayoshi.
Since 1609, the Ryūkyūs had been tributary to both Satsuma domain (which
subsumed Kagoshima castle town) in Japan and the Qing empire. On August 29,
1871, the domains in Japan were abolished, and the Ryūkyūs were, on paper at
least, folded into the new entity of Kagoshima Prefecture.71 However, there was still
a court in Shuri castle of Naha, and the Ryūkyūans considered themselves tribu-
tary to the Qing. It was precisely during the Miyakojima sailors’ ordeal in Taiwan
(November 30, 1871, through July 12, 1872) that central administrators in Tokyo
were devising institutional and pragmatic stratagems for integrating Ryūkyū into
the Japanese state, destroying the legitimacy of the Shō dynasty, and separating
Ryūkyū from China. Because of its timing, then, the shipwreck was a godsend for
Japanese who sought a firmer international claim to exclusive ­sovereignty over
the Ryūkyūs.
Kagoshima governor Ōyama’s August 31, 1872, letter urged Meiji to punish the
Taiwanese for harming his subjects (the ill-fated Miyakojima tribute mission) in
order to “spread the imperial prestige abroad and succor the angry spirits of Your
islanders below.”72 In October 1872, these bottom-up rumblings from Kagoshima
gained traction in Tokyo with the aid of Charles De Long, the American minister
to Japan. Seeking to drive a wedge between the Qing and Japan, De Long stirred
the pot by drawing Minister of Foreign Affairs Soejima Taneomi’s attention to the
Taiwan shipwreck and suggesting that it be made the occasion of an armed expedi-
tion. To further his aims, De Long introduced Soejima to LeGendre, who was just
six months past his last visit to the Langqiao headmen Isa and Toketok. The two
statesmen discussed Taiwan with the aid of LeGendre’s extensive map collection.
By December 1872, Soejima convinced the Council of State to hire LeGendre as a
high-ranking advisor to the Japanese court.73
The Meiji emperor himself received Soejima and LeGendre on March 9, 1873,
and issued a rescript granting Soejima full authority to avenge the deaths of the
Ryūkyūans.74 They both set off for Beijing in mid-March 1873 to ratify a revised
 From Wet Diplomacy to Scorched Earth    67

version of the Sino-Japanese Friendship Treaty of 1871.75 But there was also an
unannounced goal for this mission: Soejima and Yanaihara Sakimitsu would put
the Ryūkyū shipwreck case before Qing officials to notify them of an invasion of
the areas of Taiwan “beyond Qing government.” LeGendre thought the Qing might
actually cede a portion of the island and allow its permanent occupation by Japan
without a fight. On June 21, back-channel negotiations conducted by Yanaihara,
his interpreter, Zheng Yan’ning (Tei Entei), and three Qing officials named Mao
Changxi, Dong Xun, and Sun Shida addressed the 1871 shipwreck.
The Japanese side, according to later testimony (the correspondence was not
preserved), claimed to be the rightful defenders of the beheaded Ryūkyūans.
They informed Qing negotiators that Japan would send a punitive expedition to
Taiwan to find the murderers and punish them. The Japanese delegation argued
that Taiwan was so starkly divided between its governed, Chinese territory and its
ungoverned, indigenous territory that a foreign army could militarily occupy the
latter without appreciably affecting the former.
Qing statesmen, led by Li Hongzhang himself, did not acknowledge these side
talks as part of the formal diplomatic mission. The Qing view that emerged from
the discussions nonetheless defined the Ryūkyū Kingdom as tributary to the Qing;
the 1871 shipwreck was an internal matter. According to Beijing, the Qing mag-
istrate’s repatriation of twelve survivors to Naha demonstrated sovereignty. They
considered the case closed. Qing officials stated that the savages (banmin, seiban)
were but one of many border peoples whom the empire allowed to maintain sepa-
rate customs and usages. Qing claims that the assailants were beyond the pale of
civilization (kegai) and yet within the imperial realm were seized upon by Japanese
negotiators, who made much of the term kegai in subsequent diplomatic squabbles
regarding the righteousness of the 1874 invasion.76
By failing to restrain wreckers on the south coast, the Japanese side reasoned,
the Qing had forfeited its claims to jurisdiction in Langqiao. The Japanese delegates
argued that it was their right and duty to “chastise” the “barbarians,” either before
a European power filled the vacuum by annexing Taiwan as a colony, or before
enraged Japanese vigilantes stormed the island in defiance of all governments.77
The diplomats returned to Japan with the Mudan village incident left unresolved.78
Foreign Minister Soejima resigned in the aftermath of the October 1873 debates
among Council of State oligarchs over invading Korea (the seikanron affair).
Therefore, plans for the invasion of Taiwan remained dormant that winter. In the
spring of 1874, however, they were revived to mollify Satsuma men such as Saigō
Tsugumichi. In the wake of the seikanron hollowing out of Japan’s central adminis-
tration, the Taiwan invasion held together a fractured oligarchy by throwing a bone
to expansionists who remained in Tokyo. On May 4, Ōkuma Shigenobu was put in
charge of the “Taiwan Savage Territory Office” to occupy Langqiao. As the plans to
invade Taiwan became public knowledge, Western diplomats declared their neu-
trality in deference to the Qing, though sub rosa support for Japan was also in
68     The Anatomy of a Rebellion

evidence.79 Recalibrating in light of foreign pressure, Ōkubo Toshimichi, a leading

member of the council, cabled the expedition’s commander, Saigō Tsugumichi, to
delay the mission. In contravention of Ōkubo’s order, on May 2, 1874, Saigō sent
the first transport with 1,200 soldiers to Taiwan, with three more to follow.80
The Japanese expeditionary force deployed 3,658 men between May and
December 1874. LeGendre himself did not participate, but his network was put to
good use by General Akamatsu Noriyoshi and Douglas Cassel, a lieutenant com-
mander in the U.S. Navy charged with operations planning. The first Japanese ship
arrived in Langqiao Bay on May 7. It docked at Siaoliao, where LeGendre had
disembarked for the 1872 mission recounted above.
The lengthy correspondence between Ōkuma Shigenobu and Charles
LeGendre reveals that cooperation with local agents in Siaoliao was anticipated,
as well as some sort of alliance with Toketok and the Tuilasok Paiwans. LeGendre
recommended that Siaoliao headman Miya be sought out before troops disem-
barked. According to LeGendre’s instructions, the Siaoliao and Tuilasok men were
to be threatened with an invasion force of twenty-five thousand men and utter
extermination. They had narrowly escaped suspicion for collusion in the deaths
of the fifty-four Ryūkyūans, LeGendre warned through Cassel, because they had
extended courtesies to LeGendre in 1872.81
Commander Cassel arrived in Langqiao with Fukushima Kunari, the Japanese
consul at Xiamen, on May 6, 1874, aboard the vessel Yūkōmaru. Cassel had also
taken a commission from the Japanese government to execute the punitive expe-
dition to Taiwan. The Chinese-American interpreter James Johnson was sent
ashore to locate Miya and his relatives. A payment of cash was offered to retain the
services of the leading family of Siaoliao, while LeGendre’s threats and promises
were issued on shipboard. The following week, through Miya’s orchestration, the
Japanese hired around five hundred Chinese laborers, organized by headmen, at
wages of thirty cents per day. The local economy received another boon from the
sale of provisions and land rentals. Journalist Edward House remarked sardoni-
cally about the entrepreneurial spirit of local Langqiaoans, who leveraged their
bargaining power without hesitation.82
On May 10, Cassel sent Johnson, Miya, and attached Siaoliao interpreters to
arrange a meeting with Tuilasok leaders. Miya had reported Toketok’s death ear-
lier, so plans to work through his successor, Isa of Sabaree, were put into place.
The first diplomatic encounter between Japanese officers and the Langqiao con-
federation was held on May 16. The encounter took place at the village of Wangsha,
a small settlement called “semi-savage” in Japanese accounts and “half-caste” in
English accounts—signifying its location in the foothills at the entrance to the
areas of Paiwan domination.
Three Paiwan headmen agreed to meet with the Japanese on neutral ground.
They refused to enter the Japanese camp near Siaoliao, and they resisted requests to
enter their own villages beyond Wangsha in the mountains. The Sabaree chief Isa,
 From Wet Diplomacy to Scorched Earth    69

who took charge of the Paiwan delegation, brought a large retinue of armed retain-
ers. Cassel acted as General Akamatsu’s translator, it appears, while he also issued
his own declarations as LeGendre’s surrogate. Johnson translated for Cassel (from
English into southern Chinese), while Miya translated for Johnson (from southern
Chinese into a Paiwan dialect). Journalist Edward House reported: “The interview
was not very long, and although the colloquy was necessarily slow, requiring triple
translation each way—from Japanese into English, thence into Chinese and again
into the savage dialect, with the same process in reverse in replies—there was little
occasion for extensive discussion. The mere forms of meeting and recognition, and
the interchange of a few reassuring words were about all that was really required.”
Three Snider rifles were given to Isa and the other Paiwan headmen, and the meet-
ing was concluded with drinks and a feast.83
Aspects of the scene described above were repeated late into the summer of
1874. In published writings, journalist Edward House and LeGendre emphasized
the warm relations that permeated indigenous-outsider relations in Langqiao.
Saigō Tsugumichi’s 1902 obituary claimed that he wore the silver bracelet
received from Paiwan headmen at one of these feasts into old age. This bracelet
was immortalized in a postcard set issued on the twentieth anniversary of his
death and in a photo attached to a 1936 tribute to his Taiwan days, financed by
Saigō’s family and admirers.
But Gatling guns, military parades, and intimations of larger troop deploy-
ments were subsequently brandished to intimidate the Paiwan emissaries who
received Japanese gifts. And as Japanese troops began to use this firepower to
destroy Paiwan bodies, villages, and landscapes, it became easier to arrange meet-
ings. Thereafter, the conferences lost the character of diplomatic encounters and
became occasions for the Japanese to enumerate demands and specify the param-
eters of acceptable conduct.
This paradigm shift is illustrated by the second major meeting between Japanese
forces and Paiwan headmen, held on May 25, 1874. This meeting occurred three
days after a battle that broadcast the Japanese Army’s capacity for carnage loudly
and clearly. The May 22, 1874, Stone Gate Battle broke out during the search for
perpetrators of an earlier Paiwan ambush. According to Cassel, the Japanese troops
blundered into Stone Gate for lack of discipline—merely to avenge the death of a
single soldier. The two or three hundred Paiwanese defenders at Stone Gate were
led by Alok and his son. Both were killed in the melee, along with another thirty or
so Mudan warriors. Many more were wounded. On the Japanese side, between four
and seven were killed and over a dozen wounded (accounts vary). News of Alok’s
death and the ferocity of Japanese troops—who severed a dozen Paiwan heads and
brought them back to camp with other trophies of war—prompted Isa and other
leaders to seek out Commander Saigō in order to avoid becoming the next victims.84
The May 25, 1874, conference was attended by Commander Saigō himself and
was held at Miya’s personal residence. The main forces of Paiwanese guards and
70     The Anatomy of a Rebellion

Japanese soldiers were kept away, since there was mutual suspicion. The five Paiwan
delegates themselves arrived with a cow, ten chickens, hides, and millet wine. Saigō
“ordered a very large package of very handsome presents to be brought,”85 which
“consisted of two superb Japanese swords, packages of silk, woolen and cotton
cloths, and a variety of . . . ‘fancy goods.’ ”86
The Japanese side, through the same three layers of translation that obtained
during the first meeting on May 16, admonished Isa and the others that

. . . there were two tribes of people with whom we [the Japanese forces] have a deadly
quarrel, and that not one single man of these should escape the death which they
deserved at our hands. These were the [Mudan], and the [Kuskus] who had helped
them, and that, as sure as the sun rose in the East and set in the West every man
of these wicked people should surely die by our hands. But that we had heard that
some of the people from [Sabaree] and [Tuilasok], and the other Southern tribes
had availed themselves of the mountain roads to go round and join our enemies,
these treacherous and murderous [Mudan], and I wished to solemnly warn him that
if this were really done, the most dreadful punishment would fall upon him and all
his people.87

Isa, who led this delegation, did not require Miya’s third layer of interpretation;
like Toketok’s brother and his adopted son, Bunkiet, Isa could understand spoken
Chinese. He assented to the demands that Japanese troops have free access to the
peninsula and that Mudan and Kuskus men not be harbored. Cassel then prom-
ised to distribute flags to sixteen of the eighteen tribes of Langqiao—Isa asked to
receive them personally. The absence of the flags would then identify the Mudan
and Kuskus foes as fair game for Japanese soldiers to fire upon at will. Imperial
Japanese presumption of indigenous collective guilt, and the delineation of free-
fire zones marked out for generalized punishment, would reemerge in the early
twentieth century in the camphor wars against the Atayal, Saisiyat, Truku, and
Sediq peoples of northern Taiwan.
Since the Japanese did not have actual flags on hand for villages to hoist for
protection, numbered certificates with the names of the villages and headmen
were issued and then distributed as promissory notes on the flags (see figure 9).
It is hard to imagine how such placards might have shielded civilians from col-
lateral damage or cases of mistaken identity. In any event, Japanese official his-
tories reproduced facsimiles of them in their accounts of the occupation, as if to
document the army’s precision and discrimination, in the midst of a rough-and-
ready display of brute force that sent a message not dissimilar to that of the French
crown’s execution of the regicide Damian, as famously reconstructed by Michel
Foucault.88 Against this backdrop of imminent threat, Cassel wrangled an agree-
ment out of Isa for permission to build another fort in Langqiao, this time on the
east coast. The sufficiently rattled Isa accepted. After these terse communications
were relayed back and forth, Saigō’s gifts were distributed while cups of “Chinese
 From Wet Diplomacy to Scorched Earth    71

samshu” were passed around so the gathered parties could “drink Friendship”
(Cassel’s words) to conclude the proceedings.89
The expedition’s only set-piece, scripted military operations—drawn up to pun-
ish the alleged killers of the Ryūkyūans—were raids on Mudan and Kuskus settle-
ments from June 1 through June 3. These search-and-destroy missions were carried
out by three detachments. They hacked through tough terrain and numerous
improvised Paiwan defense works to cross the narrow peninsula. The sweltering
heat, swollen rivers, Saigō’s nocturnal bivouac, and army food shortages became
part of the mission’s lore. Although it was celebrated in press reportage and color-
ful wood-block prints as a glorious affair, the climax was less so: the burning of a
few Paiwanese settlements in Mudan and neighboring Ernai. These villages were
abandoned before the Japanese arrived, because they had been warned away by the
skirmish at Stone Gate.90
To formalize the victory over Mudan, a triumphal procession was held at the
Kameyama headquarters just south of Siaoliao on June 5. Three days later, del-
egates from numerous tribes came to meet Saigō. In American accounts, nine dif-
ferent certificates were redeemed for proper flags. In a Japanese account, based
presumably on more detailed reports, sixteen flags were distributed that day.91 At
this meeting, the Japanese troops reinforced the lessons taught at Stone Gate and
in the village-burning exercises of early June by displaying their Gatling guns and
massed troops in formation. Resistance was futile. Again, gifts—though of a less
generous nature—were distributed. Here, Isa consented to a Japanese fort at a spe-
cific location on the east coast, while refusing the offer of rent for the land.
The June 8 meeting can be interpreted as the culmination of a series of meet-
ings with lower-Hengchun headmen that commenced in 1867 with LeGendre’s
first confab with Toketok. This sequence of complex encounters can be viewed
as a progression of sorts. That is to say, over the course of seven years, wet diplo-
matic exchanges between Paiwan leaders and outsiders—effected by offers of gifts
and pledges of mutual obligation—gave way to more scripted encounters wherein
formalistic, rule-laden edicts were issued simultaneously with threats of violence.
Initially, for LeGendre to meet Toketok on September 10, 1867, with a few inter-
preters, it required weeks of advance planning and mediation. Toketok made the
U.S. consul wait and finally granted him a brief audience in his own compound.
The first Japanese meeting with Isa, on May 16, 1874, was held on neutral ground,
while the second one, on May 25, occurred in Miya’s courtyard within shouting
distance of Camp Kameyama. For the June 8, 1874, meeting, Saigō and his men
didn’t budge but instead waited for the Langqiao headmen to visit Saigō’s head-
quarters (see figure 8). As Edward House reported, by then the Japanese had reck-
oned the total Paiwanese population of Langqiao at three thousand people—fewer
than the number of Japanese forces brought to bear on the peninsula. The outcome
of the military contest was longer in doubt by June 8. Therefore, the Paiwan emis-
saries were treated as supplicants.
72     The Anatomy of a Rebellion

Figure 8. Saigō Tsugumichi, Japanese soldiers, Paiwan headmen, and the interpreter James
Johnson. [Saigō Tsugumichi and Mizuno Jun in Hengchun, 1874], lw0242, East Asia Image
­Collection, Lafayette College, Easton, PA, accessed May 13, 2016, http://digital.lafayette.edu/
collections/eastasia/lewis-postcards/lw0242.

On July 1, a fourth meeting included the recently vanquished Mudan repre-

sentatives. Sakuma Samata (who would become the fourth governor-general of
Taiwan in 1906) acted as Saigō’s delegate at this session, which was held in Poliac,
the Hakka settlement that functioned as the gateway to the interior. Much had
happened since the June 8 conference at Camp Kameyama. On June 22, the Qing
plenipotentiary Shen Baozhen visited Langqiao and demanded that Saigō cease
operations. Saigō was not empowered to negotiate with Shen, but the writing was
on the wall. Japanese troops would be leaving in the near future and would not
colonize Taiwan. Nonetheless, Sakuma read a battery of prohibitions and warn-
ings to the assembled headmen and issued flags. By November, at least fifty-three
flags had been doled out to Paiwan emissaries (see figures 9 and 10).92
But wet diplomacy was not yet dead in Hengchun, despite the complete reversal
in power relations. In August, the Tuilasok headman invited Commander Saigō
for a banquet with representatives from the interior of southern Taiwan. Saigō
traveled by sedan chair across the peninsula. They stopped at Isa’s compound in
Sabaree and then arrived for a large banquet in Tuilasok. His translator Mizuno
Jun’s description documents indigenous-Japanese interpersonal relations at their
dénouement. Mizuno recalled:
 From Wet Diplomacy to Scorched Earth    73

Figure 9. A certificate attesting submission to the Japanese

­ xpeditionary Forces, 1874, and the schematic drawing of the
E
­surrender/alliance flag, from Inō Kanori’s general history of Taiwan.
Inō Kanori, Taiwanshi (Tokyo: Fuzanbō, 1902), facing p. 76.

A chief from a neighboring village with about ten men showed up, and a twilight
feast ensued. A Chinese style wooden table was set up out of doors, and on top of
it was pork gravy, deer, and pheasant with sweet potatoes, all cooked in a big pot
and spread out on the table. The local sweet potato shōchū was ladled out of a big
rice-cooking pot, and we were strongly urged to drink. While many of that day’s
74     The Anatomy of a Rebellion

Figure 10. Photograph of families in Lower Hengchun, with relics preserved from the 1874
expedition, ca. 1910. Customs of Savage Tribe, ip2026, East Asia Image Collection, Lafayette
College, Easton, PA, accessed July 24, 2017, http://digital.lafayette.edu/collections/eastasia/
imperial-postcards/ip2026.

honored [Japanese] guests were renowned for their drinking abilities, they were no
match for the amount of alcohol served at this banquet. This huge banquet appeared
to consume a lifetime of wealth for the Indigenes to prepare, and this should have
produced the best feelings of affection. However, the strong insistence that people
continue to imbibe produced some uneasiness and perplexity . . .
Men then joined in a circle and began to dance and sing. Before leaving, Saigō
attempted to do a little dancing himself, to the great delight of all present. The danc-
ing was not unlike the obon dancing done in the Tokyo area. When the savages saw
Saigō do their dance that evening, they were pleased beyond limit.93

Unlike the exchanges of toasts, drinks, and presents at the May 16 feast attended
by Isa in Wangsha, Saigō’s interactions with Tuilasok hosts in August were of no
diplomatic import. By late July, all combatants had declared submission to Japan,
and fighting had ceased. Monotony had set in among the soldiers.94 Morale sank
to new lows, and homesickness, boredom, and fatigue became widespread. Danny
Orbach has unearthed a Japanese veteran’s memoir stating that young recruits,
egged on by their officers, were shouting from the treetops to Saigō himself, “Let’s
go home!,” when the commander made his rounds.95
In addition to flagging morale, malaria visited Japanese military camps. From
late August through November, the outbreak—abetted by insufficient shelter, heavy
 From Wet Diplomacy to Scorched Earth    75

rains, lack of medicine, food scarcity, and poor sanitation—took over 550 Japanese
lives. Of the army’s 500-strong contingent of laborers, over 120 perished.96 Reports
in the Japanese press about illness at camps began to surface in late July; by October
they had become routine. Journalistic reports were based on rumors and provided
wildly fluctuating statistics. But all reports agreed that Taiwan’s climate was “mias-
mic” and unhealthy for Japanese people, an image of the island that persisted into
the twentieth century. As Chen Xuan has pointed out, however, many Paiwan peo-
ple also died from diseases spread by Japanese troops and workers, while Fujisaki
Seinosuke, a colonial official, himself blamed lack of resources, planning, and med-
ical knowledge for the high death toll. In other words, the ravages of illness, which
were fiftyfold more lethal than combat in Langqiao, cannot be attributed to “tropi-
cal conditions” but rather to imperial overreach.97
Up until the time of the troops’ departure in December 1874, small groups of
Japanese scouts made forays around the peninsula to learn about its social condi-
tions and ethnic complexion. Numerous visits to headquarters near Siaoliao or the
northern encampment in Fenggang for banquets or displays of military hardware
were frequent. Importantly, no attempts were made to extract resources for com-
mercial export from Langqiao or to police the day-to-day activities of the penin-
sula’s inhabitants. There were reports of insubordinate troops committing sexual
harassment and petty crimes,98 but Langqiao did not become the site of the ethnic
cleansing threatened by Cassel aboard the Yūkōmaru on May 7, 1874.

L A N G Q IAO A N D T H E SP E C T E R O F P U B L IC P OW E R

After the December 1871 killing of the fifty-four Ryūkyūans, Qing district and pro-
vincial officials treated survivors with utmost care; at the next rung below, local
literati and traders in southern Taiwan paid ransoms to save some lives and took
it upon themselves to properly inter the remains of the dead. The slain Ryūkyūans
were enshrined in mass graves that were subsequently feted at biannual rites to
calm their spirits. In short, the murders of Ryūkyūans may have occurred beyond
the formal limits of Qing jurisdiction, but they did not occur beyond the pale of
civilization, as diplomats contended. Edwin Pak-wah Leung’s and Chou Wan-yao’s
contentions that the dynastic system for handling shipwrecks worked without
complaint are important, for they speak eloquently to the Qing court’s centrip-
etal orientation, its narrow base of legitimacy, and its incompatibility with high-
velocity capitalism.
As Chou and Leung indicate, the complaint against the Qing for the deaths of
the Miyakojimans was launched by Kagoshima activists, not the Ryūkyūan court.
The activists who remonstrated for revenge, it should be noted, were not interested
in extending the parameters of the treaty-port system. Their project was nonethe-
less intimately connected to the horizontal integration of the Meiji state. As power
was centralized in Tokyo, and the “general good” redefined as the bottom line
76     The Anatomy of a Rebellion

on a national balance sheet instead of in terms of the exemplary moral conduct

of officials,99 the Satsuma activists—all samurai—advocated an attack on Taiwan
to retain political relevance. While it is tempting to view them as members of an
emergent national citizenry, aggrieved by the loss of newly imagined conationals
from Miyakojima, the Satsuma men Kabayama, Saigō, and Ichiji were “domain
nationalists,” in all likelihood.100 Nonetheless, the fact that low-ranking men in a
geographically remote prefecture were able to exert political pressure on the Tokyo
government is significant. The post-Tokugawa state could not ignore bottom-up
expansionist impulses, for it had to shore up its multiregional coalition via appeals
to patriotism. If the Meiji oligarchs were to build a state strong enough to remain
sovereign in an unforgiving social-Darwinist world order, they would require
more than tacit acceptance of the central government’s primacy. Satsuma’s active
commitment to the Meiji modernization project was crucial, so Ōkubo and the
central leadership launched the expedition even against foreign objections and
those of fiscal hawk and pragmatist Kido Takayoshi.101
Qing support for limited forms of restorative justice and its tolerance of wreck-
ing and unsanctioned violence on the periphery of its realm were relics of dynastic
rule. The Qing did not rest its legitimacy on claims to embody, represent, nurture,
and defend every one of its citizens, no matter where they were located. Its rep-
resentatives even spoke of Paiwanese subjects as “beasts” and people of such low
status that honorable governments would avoid any association with them. The
Qing’s inward orientation, as so many have noted, was not conducive to building
national strength, as was made painfully clear when Li Hongzhang was left to fight
Japan’s imperial forces with his regional navy in 1894. Its anachronistic and low-
velocity approach to diplomacy was best articulated in a formal response drafted
by four members of the Zongli Yamen to pointed questions of Ōkubo Toshimichi
at the September 14, 1874, negotiations for Japan’s withdrawal from Taiwan. The
Qing ministers wrote:

In the aboriginal territory the Chinese Government let the indigenes keep their cus-
toms. Those who are able to pay tax pay it. Those who are talented enter nearby
schools. We enforce generous and lenient policy to bring them up to a high level
of edification. They are subject to the officials of the nearby districts. China stresses
a gradual process of government, and has no intention to forcefully or too rapidly
subjugate them. The natives of Hainan Island are treated the same. China has many
regions of similar condition and every province has its own practices.102

In other words, the Qing realm included many areas deserving of special treatment,
depending on local conditions and levels of economic development. In some areas,
it was appropriate to govern very little, if at all. Consequently, at the margins of
Qing rule in Langqiao, LeGendre and then Saigō were forced to rely upon a collec-
tion of headmen, brokers, and guides for hire to locate allies in their search for the
remains of fellow countrymen or in pursuit of malefactors who had killed them.
 From Wet Diplomacy to Scorched Earth    77

Jürgen Osterhammel has argued that such “cross-border lobbies” of “pirates

and partisans, semiprivate military operators and warlords” stitched the world of
prenational empires together. It could not have been otherwise, given the limited
state capacities of these far-flung multiethnic states, separated as they were by vast
stretches of unadministered territory. As the international system consolidated
itself around industrialized nation-states, however, these interstitial groups were
displaced, ideally, with functionaries who answered to center-generated directives.
Osterhammel terms this process “horizontal integration.”103
Ōkubo rebutted the Zongli Yamen’s position paper, quoted above, with point-
by-point objections that called for the horizontal integration of Langqiao with the
Qing empire. Ōkubo asserted that “a nation is responsible for the control of crimes,
and when crimes are not punished, then there is no law, therefore no national
jurisdiction.” In addition, he stated that “tax is a contract agreed to between the
ruler and the ruled, and must be applied uniformly upon every citizen and subject
within the domain. Making any exception would reduce the payee into tributary
status and as not owing allegiance to the ruler.” To his list of criteria for demon-
strating sovereignty, Ōkubo added compulsory education. He charged that “the
education of only a very small minority of a tribe could not be regarded as ‘civiliz-
ing’ the whole aboriginal population.”104 Thus, Ōkubo advocated universal educa-
tion for rural Taiwan at a time when the Japanese system itself only operated in fits
and starts. Kido Takayoshi, Japan’s minister of education in 1874, argued against
Ōkubo’s plan for outfitting an expedition to Taiwan for this very reason: Japan’s
own public-school system was woefully underfunded, and the nation could not
afford foreign adventures until it put its own house in order.105
Ōkubo railed against Qing officials who “preside remotely and abandon the
indigenes to their criminal undertakings.” Such leniency, he urged, was “tanta-
mount to not having any jurisdiction.” Lastly, Ōkubo attacked Qing policies of
attraction and voluntary submission, stating that “two hundred years of gradu-
alism was a bit too slow.”106 Some thirty years later, Japanese parliamentarian
Takekoshi Yosaburō would make an identical argument to his own central gov-
ernment: TGG policies of voluntary submission through benevolence were too
gradual, leaving the empire vulnerable at its extremities, he argued.107
The endless rounds of wet diplomacy that characterized political transactions
in Langqiao before 1874 were symptomatic of “the infra-power of acquired and
tolerated illegalities” that provoked LeGendre and formed the basis of Ōkubo’s
case against the Qing. At the same time, as Carol Gluck, Takashi Fujitani, Robert
Eskildsen, Marlene Mayo, Danny Orbach, and Hyman Kublin have demonstrated
amply, the Japan of 1874 had barely made a beginning on stamping out tolerated
illegalities, regionalism, dynastic myopia, and threats to central authority within
its own borders. It is thus unsurprising that many of Ōkubo’s charges against the
Qing were based on international law as interpreted by Gustave Boissonade, who
accompanied Ōkubo to Beijing on September 10, 1874, as his advisor.
78     The Anatomy of a Rebellion

The French legal scholar also coached Itō Hirobumi and Inoue Kaoru on the
fine points of international law as they extended the treaty-port system to Korea by
military force and diplomatic cunning in the mid-1870s. Therefore, when Ōkubo
alluded to constructions of sovereignty in Beijing that resembled the instantiation
of public power, he referenced the abstract writings of jurists such as Emer de
Vattel, Friedrich Martens, and Johann Kaspar Bluntschli, and not conditions in
Japan. Ōkubo was not suggesting that the Qing behave more like the Meiji gov-
ernment, but he was rearticulating his own fledgling government’s centralizing
aspirations in a different setting.108 The Qing negotiators simply refused to play by
these new rules (the Koreans would have less success rebuffing such arguments
in 1876), and the Japanese ended up withdrawing from Taiwan that December.109
LeGendre himself did not envision the creation of a disciplinary society or
the establishment of public power to end Langqiao’s threats to shipping. While
LeGendre was still getting his feet wet in treaty-port diplomacy in the late 1860s,
his vision for southern Taiwan was indigenous autonomy under Qing suzerainty.
Like Native Americans or New Zealand’s Maori peoples, Taiwan’s Paiwans could
be left to manage their own affairs but would be denied the ability to enter into
agreements with foreign powers. The Qing would be responsible for “restraining
them” and indemnifying injured foreigners. If the Qing could not accomplish these
tasks, it would forfeit the territory.110 A few years later, in mid-1874, after LeGendre
had been removed from Taiwan for a couple of years, his view hardened. His new
plan B for treaty-port era reterritorialization was extermination. If they could
not be pacified, foot draggers who impeded access to sea-lanes were expendable.
As an alternative to extermination, LeGendre proposed making Langqiao into a
Japanese penal colony, developing it along Australian lines.111 In all of these sce-
narios, there was no provision for dispersing power throughout a population to
make it self-actuating, as was the goal of the Meiji oligarchs vis-à-vis Japan’s own
citizens. At the same time, the Qing policy of gradualism, or letting border peoples
maintain separate customs until they “came around,” was also ruled out.112
This newly emerging liminal space between the old Qing posture of letting dis-
tant tribes in Hainan or Langqiao maintain their own customs until they came
into the fold of “civilization” and the concurrent Meiji policy of horizontal inte-
gration at all costs for Honshū, Kyūshū, and Shikoku created the ground for the
transformation of southern Taiwan’s natives into indigenous peoples.

T H E Q I N G - JA PA N E SE I N T E R R E G N UM A N D U E N O
SE N ’ IC H I

In 1890, Foreign Minister Aoki Shūzō ordered Ueno Sen’ichi to Taiwan to investi-
gate its commercial potential. In late December, Ueno set sail for Danshui (a port
near Taipei).113 In early 1891, Ueno departed Danshui “to meet a chief (doshū)”
as part of his mission. Ueno described his destination as “the Tokoham region,
 From Wet Diplomacy to Scorched Earth    79

which occupies a majority of the savage [territory] (seiban chū mottomo daibubun
o shimuru Tokohamu chihō ni shite).”114 Ueno used the term “Tokoham” (Dakekan)
loosely; the route he described brought him not there but rather to Atayal peoples
then known as the eight villages of Marai (near today’s Wulai). Ueno departed
Taipei on a sedan chair, crossed a broad plain to Xindian, and from there traversed
several creeks to the villages known as the Quchi tribes (Wulai, Raho, Raga, and
Rimogan—see map 6).
By 1891 the Quchi area was a well-established trading junction for Atayal
peoples to barter furs, bones, organs, and medicinal plants for cloth, iron, gun-
powder, and ornaments with Chinese traders. It can be considered the functional
equivalent of Poliac in Hengchun. Quchi analogously connected Xindian, and by
extension Taipei/Danshui, to the mountain settlements and forests south along
the Nanshi River (then known as the Xindian River). As early as 1871, European
and American writers knew to travel there to pick up guides and information
on their way to the interior. On the map Charles LeGendre published that year
(which made its way into Japanese military archives along with Ueno’s handwrit-
ten report), the river junction abutting “Koo-cheu” (Quchi) is tagged “Chiankoey
the name of all the places where exchanges between the Chinese & Aborigines take
place.” According to Consul LeGendre, the Border Tower, just downriver, was a
no-man’s-land between a southward-moving Han frontier and the indigenous vil-
lages tucked in the mountains.115 As was the case when LeGendre advanced south
past Fangliao in Langqiao, traders moving south past Quchi were dependent upon
such intermediaries as they could find and went past the “savage border” at their
own peril.
In 1882 assistant Imperial Maritime Customs officer William Hancock also
stopped in “Kochu” (Quchi) en route from Xindian. The year Hancock wrote,
oolong tea exports from Taiwan were increasing annually, a trend that may have
contributed to some of the frontier chaos and violence he witnessed. Hancock
remarked on the presence of trading posts and Atayal-Chinese interpreters.
According to Hancock, Chinese settlers in Quchi were still in danger of head-
hunting raids, despite their inroads as settlers and tea cultivators. When he finally
obtained passage to an Atayal village, he noted the presence of many severed
Chinese heads. Since these borderland “illegalities,” to use Foucault’s terms, did
not directly affect Westerners, Hancock did not remonstrate with the Qing to fix
the problem, in contrast to the activist consul LeGendre just a decade earlier.
Hancock hired the services of a bicultural interpreter/tracker to continue
inland. He stopped at a trading post to purchase samshu liquor for distribution.
His guide, who quickly donned Atayal garb, weapons, and adornment before
alighting, brought a small group from the interior, who were then duly plied with
liquor. Convinced that Hancock was generous, they consented to a visit to an area
that appears to have been close to Rimogan, upstream from Wulai. There, in a
scene reminiscent of Saigō’s big August 1874 feast in Tuilasok, Hancock danced in
 Railroad Trunk Line
Pushcart Railroad
Government Road
Taipei
!
(
¯
Rivers

Shenkeng
Taoyuan !
(
!
(

Xindian
!
(

!
( Quchi

!
( Dakekan

ATAYAL !
(
Wulai

Jiaobanshan
!
(

Rimogan
!
(
Yilan
!
(
5 Miles

Map 6. Dakekan, Jiaobanshan, Quchi, Wulai, and Rimogan, ca. 1910. The lighter area of the map represents the Aborigine
Territory.
 From Wet Diplomacy to Scorched Earth    81

a circle with Atayal women, ate roast pork, and poured down cups of alcohol with
his hosts. He was led back to Quchi by a different path than the one by which he
had arrived, and he noted in his consular report that the village was well hidden
from outsiders.116
Following in Hancock’s tracks a decade later, Ueno finally found a guide “who
could speak the Aborigine language a little.”117 Ueno ordered him to arrange a par-
ley with the paramount chief (shūchō no gotoki tōmoku). Taking his leave, Ueno’s
interpreter removed his “Chinese-style” clothing, pinned up his queue, donned a
tunic, waist dagger, and Atayal-style cap, and set off to find the paramount chief.118
After a long wait, the interpreter returned with six Atayal people. The guests were
a “paramount chief ” named Watan Yū, his wife and daughter, and three retainers.
After Ueno distributed the expected presents, Watan Yū produced sweet potatoes
and a bundle of rice stalks from a “head carrying bag.” After this exchange of gifts,
the chief, along with the other five, stood up at once and took turns striking Ueno
in the chest with their right arms. Ueno asked the Chinese-language interpreter
(tsūben) what this meant, and the interpreter replied that it was a sign of affection
and happiness. It literally meant that the indigenes considered Ueno as kin, as a
member of the “tribe/race” (dōjinshu). Ueno then reciprocated by striking Watan
Yū’s chest with his right hand.
As the festivities continued, Ueno showed Watan a bottle of sweet potato
shōchū purchased at a trading post along the way. As Ueno was about to open it,
Watan asked if he could take the bottle back to his own village for later consump-
tion. Ueno convinced him that both sides should consume it together, on the spot,
in the spirit of fraternity. Quite joyfully, according to Ueno, the imported liquor
was consumed. Ueno’s attention to detail reveals his belief that every gesture and
gift was freighted with meaning. Ueno noted that, as in a Japanese tea ceremony,
after taking a drink, each person would wipe the rim of the bowl to show purity of
intention before passing the bowl to the next person.119
Upon his return to Tokyo, Ueno published a detailed record of his observations,
along with information gathered from Western missionaries and Qing sources he
was privy to as a Japanese consul. These appeared in the February 1892 issue of
the Tokyo Geographical Society Journal. Ueno also hit the lecture circuit to alert
Japanese entrepreneurs about business opportunities in Taiwan and on the con-
tinent.120 Ueno then addressed the Tokyo Geographical Society’s March 22, 1895,
meeting. Brandishing the earrings and cape he had received from Watan Yū and
his wife, Ueno reported that the unconquered “savages” regarded the Chinese as
enemies. Like Hancock before him, he portrayed Atayal people as innocents, who
stood little chance against their cunning civilized neighbors, the Chinese.121
During his tenure as consul at Wonsan, Korea, Ueno recopied and edited
his 1891 travel notes and submitted a handwritten draft to the Imperial Army’s
General Staff, signing the updated report October 25, 1894.122 Three months later,
the General Staff published Ueno’s report under the title “A Gazetteer of Taiwan.”123
82     The Anatomy of a Rebellion

All of these iterations institutionalized the notion that Taiwan’s population was
divided between victimized indigenes and Han aggressors and that the former
constituted a population of natural allies who could be won over with gifts of cloth,
jars of liquor, and a willingness to participate in wet diplomatic transactions.124
Because Saigō and Ueno were short-timers, their interactions with indigenes
were perforce of a different order than those of Qing officials or Han settlers. Thus
their codified and serially reproduced assurances that relations of hostility or inti-
macy were attributable to the respective characters of Chinese, Taiwan Indigenous
Peoples, and Japanese were based on false premises. These Japanese visitors,
backed by the armed might of an occupational force or the implicit threat of Qing
armies, were not charged with extracting raw materials from Taiwan’s interior, nor
did they take responsibility for maintaining the peace between rival indigenous
groups or between lowlanders and highlanders. Their faith in the efficacy of wet
diplomacy would be severely tested when the Japanese government took over from
the Qing and found itself torn between the notions of public power as expressed
by Ōkubo Toshimichi in Beijing and the practical necessity of ruling within the
constraints of finite resources. They would begin, as it turned out, by letting border
peoples keep their own customs and by exercising leniency.

F E A ST I N G , G I F T I N G , A N D D R I N K I N G : ATAYA L -
JA PA N E SE D I P L OM AC Y I N T H E C O L O N IA L E R A

On August 2, 1895, Admiral Kabayama Sukenori signed papers transferring the

sovereignty of Taiwan to Japan aboard a ship in Jilong Harbor. Within a month,
forty-six men under a garrison commander named Watanabe trekked southeast
toward the mountains to initiate official relations with Atayal representatives.
He chose the junction town of Dakekan, a gateway to the mountain trade. After
marching for a couple of miles, Watanabe waited for a delegation of Atayals.
Emissaries numbering seven, two of them women, showed up. The soldiers dis-
tributed gifts of liquor, tobacco, silver coins, and tinned mackerel to the seven
red-caped representatives.
The expected drinking soon ensued. As two o’clock rolled around, the chief
reminded Watanabe that a five-mile journey back into the mountains awaited. He
promised to meet the Japanese later at the same spot.125 On September 4, a mission
headed by section chief of industrial development (shokusan-buchō) Hashiguchi
Bunzō (future Taipei governor) set out from Taipei to follow up. After negotiat-
ing the actual spot and timing over days of weather delays and messages shuttled
through an interpreter, the summit commenced on September 8.
The Japanese lined up the twenty-two Dakekan-area residents on a field. Then
Taipei governor Tanaka Tsunatoku read a statement: “You have come from afar,
down from the mountains, bringing even your women and children. We are over-
joyed to meet with you here. We set out for Dakekan yesterday to meet with you
 From Wet Diplomacy to Scorched Earth    83

because we wanted to establish good relations and to apprise you of matters of

official import. To see you here in good health bearing a compliant (onjun) attitude
makes all of us very happy.”126
These introductory remarks, made to “test the attitude of the indigenes,” were
followed by a toast, and the liquor was produced. Like Watanabe and Ueno before
them, Hashiguchi and Tanaka found themselves at close quarters with their Atayal
counterparts. In Hashiguchi’s case, the intimacy produced mild discomfort tinged
with amusement. Hashiguchi explained at a lecture before the Tokyo Geographical
Society: “As everyone knows, when Germans make agreements, they face each
other, put their hands like this, (Hashiguchi motions in imitation) and drink beer
together. The indigenes are a little different. Touching their mouths together, they
both drain the same cup. I don’t mean to boast, but when I said ‘how about some
sake,’ immediately one came at me with his mouth and I joined him . . . If one
­cannot bear to do such things, one cannot associate or communicate with the
­indigenes at all.”127
This was not the first time a high-ranking Japanese official joined mouths
with an Atayal man. In September 1873, Mizuno Jun—then a government lan-
guage scholar and interpreter—accompanied future governor-general of Taiwan
Kabayama Sukenori to Suao Bay (near Yilan) on reconnaissance. Upon meeting
a delegation in the Nan’ao area, Kabayama’s group distributed spirits manufac-
tured in Fujian in exchange for servings of the local alcoholic beverage, probably
a fermented millet wine. Farther along, at a subsequent feast, shōchū was ladled
out to the thirsty and tired Japanese travelers. The headman grabbed Kabayama’s
colleague Naritomi Seifū by the shoulder and forced their mouths together so they
could down the liquor out of the same cup.128
Returning to the September 1895 mission: after the initial toasts, Tanaka
announced to the gathered men and women that the relationship between Japan
and the Dakekan peoples would conform to the principles of nondiscrimination:

Here we must communicate to you a matter of utmost importance. You should listen
silently and with full attention. This island of Taiwan, the parts inhabited by you
[Indigenes] and by the Chinese (shinajin), has wholly become a possession of Great
Japan, under the rule of our Great Emperor. Henceforth, the garrison troops and
officials posted here will administer both Taipei and Dakekan, and govern everyone
uniformly (mina ichiritsu ni waga nihonkan kore o kanri suru). The previous Chinese
administrators have fled and left, so from this day forward, you are all, just like us,
the children of the Emperor, you are our brothers. Having listened well to our of-
fice’s order, you will exert yourself to be loyal subjects of the Emperor and the Great
Empire of Japan.129

Tanaka’s speech was communicated through Atayal interpreters and received in

the context of joined-mouth drinking rituals and presentations of gifts. At the
same time, it was filled with “dry” sobriquets written in lofty and abstract phrases
84     The Anatomy of a Rebellion

Figure 11. Drawing of Watan Nawi’s wife,

Lemoi Maton, present at the meeting during
Watanabe’s expedition to Dakekan. In the head-
hunting bag, note the alcohol bottle and can,
which were gifts brought back from the meeting
with representatives of the colonial government
near or in Dakekan (Daxi). Tokyo jinruigakkai
zasshi 11, no. 115 (1895): back cover.

that disavowed special treatment or a budding personal relationship. Japanese

officials would propagate these mixed messages throughout the interior in subse-
quent “first-contact” events.
Despite the layers of linguistic confusion, Hashiguchi’s and Tanaka’s adherence
to Ueno’s notes on hospitality and their willingness to participate in wet diplomacy
achieved their short-term goal. Twelve of the original twenty-two Atayal were per-
suaded to visit the Dakekan garrison. Seven of these, including the Jiaobanshan
headman Watan Nui (who had dealt with Watanabe on August 29), returned to
Jiaobanshan soon thereafter (see figure 11). They were issued a beef cow and a written
note signed by Hashiguchi and Tanaka as a receipt to commemorate Japan’s expres-
sion of sincerity and its announcement that Taiwan was now “imperial territory.” It is
not clear from the documentation whether Watan Nui, like Toketok and his brother
in 1867, requested this written record or if the Japanese team felt it served a purpose.
 From Wet Diplomacy to Scorched Earth    85

The remaining five Atayal emissaries were lured to Taipei with the promise of
additional wool blankets. Accompanied by their Atayal interpreter, Pu Chin, they
went to the capital for meetings with Governor-General Kabayama, presentations
of gifts, and photo sessions. The Dakekan contingent of six (including Pu Chin)
was welcomed by a brass band and feted at an elaborate banquet with Kabayama
himself. They also visited a military police station, where they taunted a Han
Taiwanese prisoner with threats to cut off his head and bring it home as a souvenir.
These events generated sufficient excitement to be chronicled in several magazines
and newspapers.130 The Shanghai illustrated Dianshizhai Pictorial even published
an engraving of the Japanese band serenading the tattooed Atayal men, women,
and children as they dined in an ornate Chinese banquet hall.131
On September 12, the five emissaries were sent back to Dakekan on a boat,
where they continued to drink, sing, and nap.132 Thus the government-general
established its initial method of rule in the uplands, whereby Japanese rural
officials “governed” their indigenous subjects by personally distributing gifts to
headmen and retainers at events that involved alcohol consumption and intimate
physical contact, far removed from the habitations of the indigenes themselves.
As did Ueno Sen’ichi in 1891, Hashiguchi returned to Tokyo and publicized his
observations and opinions regarding Taiwan’s non-Han population. On October
13, 1895, he lectured at the Oriental Club (Tōhō Kyōkai). There, Hashiguchi reit-
erated Ueno’s assertion that indigenes and Han were mortal enemies, thanks to
the Chinese practice of breaking their promises. Hashiguchi characterized the
“brigands” (dohi) fighting the Japanese government as Hakka from Guangdong
Province. The seiban (raw barbarians), by contrast, were well disposed to the
Japanese.133 Hashiguchi showed up at the Tokyo Geographical Society on October
22, 1895, to repeat his performance.134 Colorful speeches like Hashiguchi’s, larded
with asides, anecdotes, and speculation, remained popular in prewar Japan. At
the same time, there was also an outpouring of attentive journalism and academic
writing, including notes and observations about the indigenous alcoholic bever-
ages mentioned in passing by LeGendre, Mizuno, Ueno, and Hashiguchi.

I N D IG E N OU S A L C O HO L P R O D U C T IO N A N D
C O N SUM P T IO N A S C U LT U R E

As a low-ranking official in the Bureau of Education, ethnologist Inō Kanori con-

ducted Japan’s first extensive ethnological survey of the island’s non-Han popu-
lation. Inō’s influential descriptions confirm Hashiguchi Bunzō’s report on the
etiquette observed in the north and indicate that there were several types of liquor
found beyond the pale of the former Qing administration. Here is Inō’s description
of Atayal spirit production:
To make liquor, millet or rice is steamed in a wooden basket steamer. It is mixed in
a pot with water and yeast purchased from the Chinese. This is covered with leaves
86     The Anatomy of a Rebellion

and brews for three or four days. Then the unfermented yeast is filtered out and the
liquor is ready.
We should say a word about the related matter of the actual drinking. There is
the special custom of two people touching the sides of their mouths together and
drinking out of the same vessel. This should be called a drinking ritual (inrei) which
expresses good will or courtesy.135

Based on more extensive fieldwork, fellow government ethnologist Mori Ushinosuke

emphasized that Atayal people were not habitual drinkers; they manufactured and
consumed spirits only on special occasions. Mori’s account echoes Inō’s but employs
native terminology:

Rice or millet is placed in a steamer basket (called a koro) and then put on top of
a kettle and steamed. It is transferred to a winnowing basket and cooled, and then
the appropriate amount of powdered yeast is added and mixed in. It is put in a pot
and a little water is mixed in, and the whole thing is covered with leaves. Then heat
is applied to certain areas of the pot and it brews for three or four days. Now more
water is added and the liquor is filtered in a device made from wisteria, and it is
ready to drink.
In some villages, after brewing, the refined portion of the liquor is consumed first
and then the rest is expressed. To distribute the liquid portion, a bamboo cup called
a zō is dipped into the middle of the kettle.136

Inō’s earlier account also provided recipes and commentary on manufacturing

processes and consumption practices among Bunun, Tsou, and Paiwan peoples.
Inō noted that Atayal and Bunun both engaged in joined-mouth drinking, while
southern groups used double-chambered drinking vessels for the same purpose.
Inō Kanori’s diary documents his own direct participation, revealing that even
anthropologists were expected to get wet if they would ply their trade in northern
Taiwan. On May 26, 1897, in Wulai (just south of Taipei), Inō was welcomed by a
young bantei (brave) named Watosinai, whom Inō had previously met. Upon see-
ing his acquaintance, Watosinai struck Inō in the chest and then opened up some
liquor. Subsequently, Inō and his men joined in a banquet with the village head-
man, Wataniurrak (Watan Yūra), and Watosinai, during which the younger man
broke out into exuberant singing before the group turned in for the evening.137 It is
probable that this was the same Watan Yū who drank with Ueno in 1891.
Through firsthand observation and his study of Qing records, Inō also learned
that in addition to welcoming strangers, indigenous Taiwanese engaged in com-
munal drinking at fixed times on the ritual-agricultural calendar. On September
2, 1897, in Tōsha (near Sun Moon Lake), Inō’s party was feasted on native liquor as
part of a thirty-day harvest celebration. Inō and his party drank the festival liquor
with young and old, finally joining in a ring, dancing and singing as participants in
a scene reminiscent of Saigō Tsugumichi’s 1874 adventure. Similar festivities were
held when Inō’s party arrived in Beinan on November 9, 1897.138
 From Wet Diplomacy to Scorched Earth    87

Okada Shinkō, prefect of Jiayi in 1905, recorded investigations of Tsou cus-

toms near Alishan. Okada emphasized that, though liquor was the staple bever-
age (excepting water) among the Tsou, it was not produced regularly and stored
for everyday use but was only for special occasions.139 This observation would be
repeated in many later reports on various groups of indigenes in Taiwan. In sum,
it is evident from contemporary firsthand observations that alcohol consumption,
before 1905, was variable and often restricted to particular seasons, occasions, or
festivals. For Japanese administrators in the early years, social drinking was an
effective method of bringing indigenes down from the hills to government stations
and trading posts in order to gather information.

E STA B L I SH I N G C O N N E C T IO N S T H R O U G H W E T
D I P L OM AC Y

Just as the five emissaries from Jiaobanshan were leaving Taipei in September 1895,
Hashiguchi Bunzō ordered Captain Kawano Shuichirō to lead Japan’s first official
embassy to the Yilan Plain on September 14, 1895. Kawano reported that under the
Qing, the border trade with the Nan’ao and Xitou Atayal was conducted privately
by interpreters. In 1889, however, after a military campaign to establish a Qing pres-
ence in upland Yilan (at a cost of a thousand Chinese lives), a Pacification Office
(fukenju) was built at the mountain pass Dingpobuwu to supervise trade with the
Xitou tribes. Another fukenju outpost was built at Alishi to manage trade with the
Nan’ao tribes. Thus, like Watanabe and Hashiguchi before him, Kawano headed for
a well-established junction settlement to initiate diplomacy with Atayal peoples.
Kawano’s expedition left Yilan on November 16, 1895, for Dingpobuwu and
wound through narrow passes beset by bamboo thickets. The troop’s ears were
assaulted by howling dogs and shrieking waterfowl. The Japanese ferried across a
large ravine and forded the same stream over twenty times. To avoid scaring the
Atayal away, “fourteen or fifteen” men were ordered to take a rest under cover of
a thicket. The government translator and two others were sent ahead. Kawano’s
men hiked along a valley from which they saw smoke rising from the many active
camphor stoves. At a clearing, the Japanese presented themselves to ten emissar-
ies from two Xitou villages, Xiyanlaowa and Mosu. Among them were eight men,
including the Xiyanlaowa headman Yawa Ui, and two women—Kawano recorded
all of their names and ages. Kawano’s speech made no mention of nondiscrimina-
tion but emphasized Japan’s military superiority, while it demanded compliance.
Here is the text Kawano entered in his report:

The Great Empire of Japan has recently defeated the Qing in battle. Now Taiwan and
the Pescadores islands belong to us. Henceforth, you shall follow the directives is-
sued by the Great Japanese Empire. Since the local villagers (jinmin) shall henceforth
absolutely cease their violent acts, you shall in turn cease committing violent acts
88     The Anatomy of a Rebellion

against the residents of this district. Because they were enemies of Japan, we killed
ten thousand Qing subjects in this region; we sent the other ten thousand back to
China. You should not disobey the commands of the Japanese empire and bring
disaster upon yourselves. [We shall pacify your hearts, everywhere cultivating virtue
and encouraging industry. Eventually, we shall open a path of intercourse between
yourselves and Japan.]140

Echoing the June 8, 1874, meeting between Paiwan emissaries and Commander
Saigō in Hengchun, Kawano and his men distributed a Japanese flag, gifts, drink,
and food to the assembled Xitou men and women. The meeting featured trappings
of a negotiation, with its attendant toasting, gifting, and oathing, but the visitors
were backed with overwhelming military force just beyond the scene of the meet-
ing. In this regard, parleys on Taiwan’s “savage border” in the opening decade of
Japanese colonial rule were not so different from treaty-port negotiations between
Qing, Tokugawa, and Joseon officials and gunboat diplomatists.
In some cases, as with LeGendre’s early meetings with Toketok or Akamatsu’s
first meeting with Isa, the threat was imperfectly understood or ignored. Such was
the case with Kawano’s mission. Despite the bluster, the Xitou men firmly refused
Kawano’s entreaties to visit the walled city of Yilan for further parleys, just as Watan
Nui of Jiaobanshan refused a similar request from Hashiguchi to visit Taipei. They
also rebuffed Japanese requests to travel farther inland. The Xitou emissaries did
offer to distribute the remaining gifts to other tribes and act as intermediaries
between the Japanese and the inland villages.141 The mission succeeded in estab-
lishing rudimentary communication, but the new subjects of empire successfully
parried Japanese attempts to visit their habitations.
The Watanabe, Hashiguchi, and Kawano missions all took place near the Taipei
Basin or Yilan Plain, at active nodes of indigenous-Chinese-Qing communication,
trade, and warfare. The junction city of Puli, farther south, shared these attributes
but was farther inland and more remote from the coastal ports and cities. While
the highest mountains in the Langqiao visited by Saigō in 1874 were about 200
meters above sea level, the Puli basin sat at about 500 meters, and Japan’s first
Sediq contact settlement of Paalan was at an altitude of about 1200 meters. At these
higher altitudes, the Hiyama mission broke new ground among villages not found
on treaty-port maps of Taiwan.
Puli prefect Hiyama Tetsusaburō made his first overtures to the Tgdaya (Sediq)
tribes of Paalan in December 1895. Following precedent, Hiyama dispatched a
“northern tribe” wife of a jukuban (cooked barbarian) to carry gifts and a message
to the most powerful Tgdaya chief, the headman of Paalan.142 After her second
trip and much delay, four Sediq scouts came to test the waters in Puli. They were
promptly treated to drinking, dancing, and photographs with a Captain Ishihara
of the Puli garrison. On December 31, the Paalan headman brought five “subordi-
nate” village headmen and a contingent of three hundred retainers, plus women
 From Wet Diplomacy to Scorched Earth    89

and children, down to Puli. The Japanese put them up on the grounds of the old
Qing Pacification Office post, sending them provisions of sake, rice, and salt. The
next day Hiyama Tetsusaburō, a government interpreter (tsūyakukan), and “two or
three low-ranking officials” delivered an inaugural Japanese speech:
We are overjoyed that you have come here from afar. From the beginning, we Japa-
nese, as your brothers, have desired to meet and parley with you. Fortunately, the
whole of the island has been restored as the emperor’s territory. Now we have come
to establish our administration on this part of the island. We are genuinely pleased to
have this opportunity to meet with you.
We have heard that you have always killed and injured Chinese people. This is a
very evil practice. If you are really our brothers, as we have already mentioned, you
absolutely must stop murdering. You may come visit our offices at any time. We, in
turn, would like to go to your abodes. Today, since you have visited our offices, we
have prepared a meager offering of drinks and side dishes. So let us drink with glad-
ness to your arrival!

After hearing this initial Japanese declaration, the Tgdaya contingent replied with
a counterspeech: “Murder and carnage do not come naturally to us. In the past, the
Chinese have gathered strength in numbers and committed murder when catch-
ing one or two of us alone. Therefore, when we find a small group of Chinese, we
murder them. But now that the Japanese are our brothers, we will certainly not act
this way anymore.”
With this exchange of solemn promises and admonishments, a stone-burying
ceremony ensued. The anonymous correspondent who chronicled this exchange
explained, “even should this stone and things thus buried disintegrate, our prom-
ises to each other shall not be altered.”143

I N S T I T U T IO NA L I Z I N G W E T D I P L OM AC Y

In January 1896, district officials and Taipei authorities issued decrees and created
organs of government for the purpose of routinizing and regulating contacts with
indigenes. As a former Qing center of administration and hub of the border trade,
Dakekan was the logical starting point. The old walled city of Dakekan housed
the government-general’s first dispatch station (shutchōjo), the forerunner of the
Pacification Offices.144 Its chief, Miyanohara Tōhachi, applied to the government
general for a modest 143 yen to fund “Barbarian Feasts” (seiban kyōhi) as a first
step to addressing the area’s endemic violence. Miyanohara acknowledged that
a recent rash of murders and beheadings by Atayal people could be attributed to
Japanese disarmament policies that left Chinese dwelling near the “savage border”
defenseless. Nonetheless, Miyanohara trotted out a list of good reasons for Atayal
to murder Han sight unseen: Qing policy was one-sided and relied exclusively on
force; Han merchants cheated indigenes in trading; Han cultivators encroached
90     The Anatomy of a Rebellion

on Atayal lands; and finally, Han were just as guilty of murdering uplanders as the
mountaineers were of raiding the townspeople.
Miyanohara insisted that the Atayal were stubborn, obstinate, and unreason-
able. However, they were very well disposed toward the Japanese and could be
counted on to keep an agreement. Thus, to prevent further violence, Miyanohara
invited over ten village headmen and admonished them to stop killing Han. To
govern through suasion, Miyanohara argued, regular feasts for village leaders were
the ticket.145 The government organ that would forever be associated with feasting
and drinking, the Pacification Office, was chartered to begin work on April 1, 1896,
soon after Miyanohara made his recommendations.
The early years of Japanese “Aborigine administration” (riban)—the
“Pacification Office period”—were later criticized for their leniency, along the
same lines for which Ōkubo criticized lax Qing rule in Langqiao in 1874. As a
policy that left border areas opaque and hindered Taiwan’s economic integration,
the gift policy became a lightning rod for invective. These criticisms perhaps over-
stated the government-general’s largesse, however. The initial proposed budget of
¥1,293 for gifts, or about ¥215 per prefecture, was less than half the monthly cost
of subsidizing a local Chinese militia to stand guard against the very same Atayal
men and women targeted to receive these gifts. In short, the feasts and gift distri-
butions held at the dispatch stations wrapped the savage-border militia’s iron fist
in a velvet glove. Just as LeGendre and Saigō professed the warm relations between
non-Chinese foreigners and indigenes as a product of each side’s sincerity, the
actual meetings were held with overwhelming firepower in reserve on the side of
the visitors.
As a veteran of the 1874 expedition, Satsuma man Kabayama Sukenori took a
special interest in indigenous affairs. Yet his programs had to meet strict budget
constraints, due to the cost of the war against Han insurgents and the fact that
Taiwan was a drain on Tokyo’s finances. Therefore, the twenty policemen origi-
nally requested by the governor-general for each of eleven proposed Pacification
Offices, at a total cost of ¥111,181, were cut out of the budget. Instead, each station
received one interpreter (tsūyaku of hannin rank), two assistants to the chief (shu-
jiho of hannin rank), and two “operatives” (gishu of hannin rank); only eight out
of the eleven stations would have a chief (shuji of sōnin rank), for a total of sixty-
three officials, including chiefs.146 In May 1896, Bureau of Industrial Development
Chief Oshikawa Noriyoshi, under whose purview the Pacification Office fell,
requested forty-two additional “staffers and special commissioners (koin and jimu
shokutaku)” for the rural outposts.147
In our first concrete example of how Japan governed the indigenes “on the
cheap,” these 105 pacification officers were charged with intelligence gathering,
diplomatic relations, commerce, and public works among a population commonly
believed to cover 60 percent of Taiwan’s territory and comprising at least eight
major language groups (and many dialects). Although the indigenous population
 From Wet Diplomacy to Scorched Earth    91

was estimated at around only 100,000 souls, they inhabited forbidding terrain and
guarded their autonomy with widely available firearms.
Dispatch station chief Miyanohara Tōhachi was appointed first Dakekan
Pacification Office head on May 25, 1896. The station itself commenced activities
on June 30.148 Within the month, local headmen accepted Miyanohara’s invitations
to Dakekan city for parleys. He was not sure exactly where these villages were
located, however. Things got off to a rocky start when his Han tsūji (interpreter)
refused to go inland for fear of a bandit leader known as Jian Ai. This so-called ban-
dit was ensconced in the mountains beyond Dakekan in Shinajii, with his Atayal
wife and a number of followers. Shinajii was home to many of the emissaries who
had met with Hashiguchi the previous September.149 Miyanohara heard that Jian
Ai was turning the Atayal against the Japanese, so he wanted to apprehend him.
Miyanohara hired two local interpreters in Dakekan as well as two banpu
(female Atayal) to accompany them to the interior. Their immediate target was
the paramount headman Daima Weixian of Yiheng. The four interpreters reached
him, and then Daima sent six scouts to Dakekan. They were duly feasted, toasted,
and lodged on the evening of July 21, 1896. The Japanese sent them back to Yiheng
with fancy cigars to entice Daima to broker a parley.
On July 28, a contingent of chiefs, men, women, and children representing six
villages arrived at Dakekan. Miyanohara recorded, “When the indigenes arrived,
the two banpu explained to us what sort of feast to prepare. They like nothing more
than pork, salted fish, and native liquor.” As it turned out, Jian Ai answered the
call, while Daima did not. Miyanohara offered Jian Ai amnesty for his past crimes
against Japan in return for a pledge to become an interpreter and informant on
conditions near Shinajii, conditions to which Jian Ai apparently assented. Chief
Daima’s emissaries attributed his absence to an injured leg.150
Saitō Otosaku, the chief of the Linyipu Pacification Office (see figure 12),
counted fifty-two villages under his charge. He reported that in August 1897,
twenty of his twenty-seven men were in sick bay.151 The turnover rate was high.
In its second year, the total number of salaried officials in the Pacification Office
increased to sixty-five (from fifty-seven the year before), but forty-one of these
officials were new. In some posts, like Puli, the turnover rate was 100 percent for
titled officials.152 And here was a potentially fatal flaw: from the indigenous side,
particularistic bonds were being forged via wet-diplomatic protocols, while from
the Japanese side, duty on the savage border was a sort of exile to be endured and
then left behind as quickly as possible.
The Pacification Office installations themselves were built over a period of two
months, from June 2 to August 3, 1896. Oshikawa Noriyoshi, Hashiguchi Bunzō’s
successor in the Industrial Promotion Section, issued a thirteen-point circular
defining the Pacification Office’s mission. Oshikawa underscored the importance
of explaining to chiefs and men of influence that the Japanese were impartial
rulers. He wrote that the simple, savage indigenes could quickly be won over by
92     The Anatomy of a Rebellion

Figure 12. The Linyipu District Administrative Office, successor to the Pacification Office,
on eve of expedition to the mountain tribes, December 15, 1898. Twenty indigenous porters are
loaded down with gifts to bring to the interior. Karl Theodor Stöpel, Eine Reise in das Innere
der Insel Formosa und die erste Besteigung des Niitakayama (Mount Morrison) (Buenos Aires:
Companñía sud-americana de billetes de banco, 1905), 42.

gift giving, and he instructed officers to meet with headmen on appointed days,
distribute presents, and explain Japan’s imperial mission. Loyal villages would
be rewarded with further gifts, while disobedient ones would be punished by
withholding presents.153 This last caveat presupposed that indigenous hunger for
imported goods was strong enough to leverage cooperation in such operations as
mapping, labor dragoons, and even turning over local peoples to Japanese police
officers. For many areas of Taiwan, however, such conditions would not obtain
until the 1920s or even later.

T H E O R IG I N S O F SP E C IA L A DM I N I S T R AT IO N

During the early years of Japanese rule, members of parliament and other public
intellectuals advocated for the extension of the Japanese constitution to Taiwan.
The debate was similar to ones raging in the United States about the Philippine
Islands. At issue was the capacity of colonized populations to “play by the rules,” or
bear the rights and responsibilities of citizens. While the “extend the constitution”
 From Wet Diplomacy to Scorched Earth    93

faction lost the early rounds of the debate, over the course of Japanese colonial
rule in Taiwan, a great deal of legal integration between the mother country and
the colony actually occurred. However, the geographic reach of this integration
stopped at the areas beyond the old Qing savage border, in the areas placed under
“special administration.”
Early debates about the legal bifurcation of Taiwan, into normally and specially
administered zones, occurred in the spring of 1897. Then, the Pacification Office
heads met under the aegis of Oshikawa to chart a course for the future based on their
first eight or nine months of field experience. One of the burning issues regarded
punishments for head taking, robbery, and other violent crimes—the myriad “ille-
galities” that fester in ungoverned areas beyond the reach of “public power,” to use
Foucault’s phraseology. Hiyama Tetsusaburō, the first Japanese prefect posted to
Puli, argued that Japanese law could not be applied to the Sediq peoples under his
charge. Echoing the Zongli Yamen members who contested Ōkubo two decades
prior, Hiyama argued that, if penalized severely for murder, as called for in Japanese
law, Sediq people would rebel en masse. Besides, Hiyama added, if Japan enacted a
strict, consistent law code for the Sediq and Atayal peoples, it could only be applied
to those who had shown themselves at the Pacification Offices. In the end, he rec-
ommended an ad hoc form of punishment, appropriate to the means at the disposal
of the overstretched colonial state: “for those who take heads, cut off their gifts com-
pletely, stop distributing liquor, and stop giving them their needed salt.”154
On the last day of the aforementioned conference, the Civil Affairs Bureau of
the Taiwan Government General issued regulations to the prefects, subprefects,
and gendarmes (kenpeitai) in regions with significant non-Han populations. The
regulations are prefaced by a recognition that hunting was crucial to indigenous
livelihoods, explaining that a complete ban on guns was inadvisable. Nonetheless,
the following restrictions were put into effect to check what was perceived as ram-
pant disorder:
1. Merchants must have a certificate/warrant (shōmei) from the Pacification Office
head to supply indigenes with gunpowder—no exceptions (kyokashō) or special
permits (tokkyo shō) are allowed.
2. Those who are not merchants must have a permit/permission (kyoka) from
the Pacification Office to supply indigenes with gunpowder; for hunting rifles,
the police and gendarmes may supply up to 3 hyakume (about 375 grams) of
gunpowder or 500 percussion caps/fuses; to supply over this limit requires
permission of the prefect.
3. The distribution of gunpowder to indigenes falls within the purview of the
Pacification Office and therefore should be witnessed by officials.
4. The Pacification Office must issue a monthly report to the police and g­ endarmes
stating the amount and kinds of gunpowder distributed to indigenes.155

Some Pacification Office heads judged the 1897 regulations on trade as too lenient.
In the fall of 1897, Nagano Yoshitora, who replaced Hiyama Tetsusaburō as Puli
94     The Anatomy of a Rebellion

Pacification Office chief in July, argued that Atayal tribes were more bloodthirsty
than the Bunun tribes who resided south of Puli. Nagano agreed that the new reg-
ulations on gunpowder and shot should be observed in the south, but that a com-
plete ban should be put into effect for Atayal. In keeping with the early Japanese
view that the indigenes were, at heart, good subjects, Nagano suggested that the
northern “savages” be reformed by taking them on sightseeing tours of Taipei.156
After sufficient transformation, the trade in gunpowder and shot, so vital to peo-
ples who relied on hunting for protein, could be resumed.157 In fact, Nagano’s rec-
ommendation was taken to heart, and a delegation of indigenous representatives
toured Japan that year; they were received by throngs of curious onlookers. In the
short term, the tours did not have the intended effect of reducing violence related
to camphor forests, however.
In response to a report that the Wuzhishan Pacification Office had punished a
camphor worker’s murder by levying a “traditional fine” of shell-money currency,
Division of Interior head Sugimura Shun took exception. “Traditional” punish-
ments, even if supplemented by embargoes of trade or the cessation of gifts, he
argued, would embolden indigenes and cause them to despise the weak govern-
ment. Outraged, Sugimura suggested that Pacification Office chiefs use gifts as
weapons, and that trade embargoes, as means of collective punishment, could
force the hands of indigenous leaders. Before resuming trade, the Pacification
Office should
1. Order that the indigenous perpetrator be delivered.
2. After receiving the delivered perpetrator, deal with them by applying the law.
3. During the time that the perpetrator has not been delivered, the whole village
should be held responsible: the distribution of gifts will be suspended; it goes
without saying that the sale of arms and ammunition will be strictly banned.
4. During the period referred to in item 3, other villages will be ordered to cut off
commerce/communication with the [offending] village.158

In July 1897, Kawakami Chikakata, Pacification Office head in southern Taiwan’s

Fanshuliao Prefecture, also railed against the patchy sovereignty that character-
ized Japan’s government along the savage border. Kawakami noted that many
camphor harvesters were forced to present gifts to indigenous headmen for
access to the forests; some demanded rattan, others tea leaves. These audacious
headmen even collected tribute outside the so-called Aborigine Territory. The
indigenes were acting under the mistaken impression that they “owned the
land” and could force outsiders to pay tribute, which in turn required the ser-
vices of the hated “go-betweens,” who were completely untrustworthy, according
to Kawakami. In the end, commerce and “Aborigine education” were impeded
because of blurred distinctions between public and private. Kawakami suggested
that all lands abutting the Aborigine Territory be put under the jurisdiction of the
Pacification Office.159
 From Wet Diplomacy to Scorched Earth    95

T H E F U KA HO R I M I S SIO N A N D T H E E N D O F W E T
D I P L OM AC Y

While the Pacification Office was still operative, in February 1897, an expedition
led by one Captain Fukahori Yasuichirō disappeared in the Neng’gao mountain
peaks just northeast of Puli while searching for a potential railway route to Hualian
Harbor. Captain Fukahori left Puli in late January 1897, and by March, all fourteen
of his men were given up for dead. In April the local garrison chief initiated a pro-
longed, frustrating series of negotiations with area headmen to find the remains
and the Fukahori expedition’s murderers. Reminiscent of LeGendre’s search for
the remains of shipwreck victims in 1860s Langqiao, the usual methods of gift
presentations backed by threats were used, but the area headmen stonewalled the
investigation.160 To pour salt in the wounds, a camphor worker was slain by Tgdaya
locals at the same time. In retaliation for Fukahori’s disappearance, the stymied
investigation, and the fresh killing, the government imposed an embargo on salt,
ammunition, and guns for eight whole years.161
As we have seen, the Pacification Office officials operated through concentric
rings of intermediaries at this time. The whole structure was bound together by the
distribution of material goods and the forging of personal bonds through drink-
ing rituals. The first intermediaries were the “Aborigine interpreters,” male and
female, who were sent to mountain redoubts to bring headmen to the Pacification
Office stations with promises of gifts, food, and drink. The next ring of inter-
mediaries were the “conciliated/submitted” indigenes who carried invitations to
other villages, bringing more headmen, interpreters, warriors, and children to the
Pacification Office’s banquets, gift distributions, and trading fairs. Although this
system performed well enough for an understaffed, poorly informed, outnumbered
set of rural outposts, it could not protect camphor harvesters from indigenes who
felt cheated of customary tolls or were defending what they considered ancestral
territory. Moreover, indigenes desperate for salt, guns, ammunition, or machine-
made cloth traded through other indigenous groups to subvert the embargo.162
Rightly sensing that the understaffed, lightly armed, and linguistically disadvan-
taged Pacification Office could not extend sovereignty by wet diplomacy alone, the
government-general decided to abolish it altogether and place indigenous affairs
within the purview of the Third Section of the lowest rung of the territorial admin-
istrative hierarchy, the District Administrative Office (Benmusho, hereafter DAO).
After the DAO replaced the Pacification Office in the summer of 1898, Japanese
rural administrators repeatedly called for tougher measures against the indigenes
who fought back against aggressive timber harvesting. Time and again, admin-
istrators called for trade blockades to be used as incentives for good behavior. In
other words, now that the Japanese, like the Qing before them, were taking sides
in an economic war for rights to harvest timber along the savage border, mentions
of Japan’s “special relationship” to indigenes disappeared.
96     The Anatomy of a Rebellion

In February 1900, Governor-General Kodama Gentarō lambasted the policies

of his predecessors Kabayama, Mizuno, and Hashiguchi. With insurrection ebb-
ing in the plains areas, Kodama declared, “we must shift our military forces to the
savage territory. Those who live there are stubborn, and live like wild beasts; if we
continue to feast them with liquor and food, staying with a policy of attraction,
it will take many months and years for them to reach even a limited degree of
evolutionary development.”163 And here was the crux of the matter: wet diplomacy
would not make indigenes into loyal, governable, imperial subjects in a timely
manner. For Kodama, the clock was indeed ticking, as Japan was a relatively poor
country playing in the high-stakes international game of competitive colonialism.
In late 1900, Japan was undergoing a massive armament expansion to prepare for
war against Russia and could not afford to manage a colony in the red.
The years 1896 through 1900 paved the way for the aggressive measures that
followed. During this interval, gift distribution centers became trading posts, with
fixed exchange rates. Having established the posts as primary nodes of exchange,
the government-general then began to impose selective embargoes against mis-
chief-makers.164 By this method, devised from expedience but also based on field
reports about wet diplomacy from the 1870s and 1880s, Japanese officials in north-
ern Taiwan’s mountain districts had cultivated an effective network of collabora-
tors, trading partners, informants, and mercenaries to go along with a rudimentary
knowledge of local political conditions.
The effectiveness of wet diplomacy as a stepping-stone to the eradication of
indigenous sovereignty is best illustrated by a notorious 1903 incident on the
Sediq/Bunun border just south of Wushe. The grislier details and some of the
specifics come down to us through memoirs many decades after the events, so
there are factual discrepancies in these accounts. The cataclysm had its origins
in the salt embargo placed on the Tgdaya tribes after the disappearance of the
Fukahori mission in early 1898. On December 16, 1902, the Puli subprefect dis-
patched a “savage woman [banpu] named Iwan” to sound out Tgdaya tribes about
an upcoming Japanese punitive expedition to the north. Iwan learned that they
sought a truce with Japan in order to resume trade in salt. Besides, they were
engaged in a cycle of revenge feuding with neighboring Toda and needed arms
and ammunition.165 Iwan was a common name in the Tgdaya region, so it is hard
to say if Iwan Robao (see chapter 2) is the person mentioned in the above report.
Nonetheless, Deng Xiangyang, Aui Heppa, and Pixo Walis have all attributed the
planning of the “Bukai incident” that followed Iwan’s reconnaissance to Kondō
Katsusaburō and his Paalan wife, Iwan Robao.166 This interpretation is consistent
with other documented facts, but I have yet to find direct documentary evidence
of Kondō’s involvement.
On October 5, 1903, the continued economic blockade drove Tgdaya men into
a bloody trap. Short of salt and firearms, men from Paalan consented to a meeting
with Gantaban men of the Bunun group south of Puli, at Shimaigahara (Bukai).
 From Wet Diplomacy to Scorched Earth    97

Figure 13. Gantaban men at the Puli branch office with severed heads of Paalan and
Hōgō men, October 1903. Jun’eki Taiwan Genjūmin kenkyūkai, ed., Inō Kanori shozō Taiwan
Genjūmin shashinshū (Taipei: Jun’eki Taiwan Genjūmin hakubutsukan, 1999), 115. Photograph
courtesy of the publisher: The Shung Ye Museum of Formosan Aborigines, Taipei.

The meeting’s pretext was Gantaban’s promise to funnel embargoed goods to

Paalan behind the government’s back. On the appointed day, the Gantaban trad-
ing delegation served the Paalan men large quantities of alcohol to initiate talks.
Thereafter, concealed men rushed in, killing 95 of the 100 Paalan men and taking
their heads.167 The Japanese press reported 104 dead along with a yield of 57 rifles,
130 spears, and 127 daggers. The Bunun warriors brought 27 of the heads to the
Puli subprefect, who appears to have expected this outcome (see figure 13).168 The
TGG file gives similar casualty figures while noting that Wushe’s once reluctant
posture toward the Japanese was much improved.169

T H E C A M P HO R I N DU S T RY A N D T H E G UA R D L I N E

The causes of armed conflict along the indigenous-Han border were multiple.
Japanese officials reported a surge in revenge killings and ritually sanctioned head
taking in the wake of Japanese arms confiscation sweeps that were part of the anti-
bandit campaigns of 1896 through 1902. But most of the killings represented efforts
by indigenes to defend hunting grounds, fallow fields, or forests previously subject
98     The Anatomy of a Rebellion

to fees levied by headmen on camphor harvesters. These camphor-related killings

dated from the treaty-port period (1860–1895), reached a fever pitch by 1900, and
climaxed in a major battle between the Japanese state and Ri Aguai in late 1902.
The camphor tree (Cinnamonium camphora, or Camphora officinarum, or
kusunoki in Japanese) “is a rather small-leafed evergreen, attaining . . . [a] height
of from sixty to one hundred feet, with wide-spreading branches and a trunk two
to four feet in diameter.” By century’s turn, the western plains had been denuded.
Camphor was still abundant, however, in the “debatable land between the Savage
territory” and the foothills.170 The aggressive collection of the product in the latter
half of the nineteenth century incited warfare between highlander and lowlander,
creating a frontier that resembled a permeable membrane; commerce flowed
through, but armies and lowlanders were kept at bay. During the last decade of
Qing rule, government losses from these battles were horrendous, totaling hun-
dreds of casualties per encounter.171
Camphor, traditionally used as an insect repellent and medicine, had been har-
vested in Taiwan since the seventeenth century, but developments in technology
and politics increased the product’s value sharply in the late nineteenth century. In
1858, Englishman Alexander Parkes invented a way to use camphor in the manu-
facture of celluloid, which meant that “combs, tobacco pouches . . . and indeed
everything which had before been made of ivory, coral or tortoise-shell now came
to be made of celluloid.”172 Camphor also became a crucial ingredient in the pro-
duction of smokeless gunpowder during the latter half of the century.173 In the
same year that Parkes put camphor on the industrial map, the Treaty of Tianjin
forced Taiwan’s two major ports to open to international trade as part of the settle-
ment of the so-called Arrow War. With demand for camphor up, increased num-
bers of foreign traders called on Taiwan’s ports, and an influential comprador class
emerged. By 1878, the export economy of camphor, tea, and sugar “absorbed about
twenty percent of [Taiwan’s] population.”174 The Lin family of Wufeng, who later
supplied the Japanese colonial government with its first hired troops to guard the
camphor districts, were part of this class.175
The old Lin guardline stretched from Taizhong city to Puli.176 Iriye Takeshi, an
infantryman who surveyed the area in 1896 for the army and wrote extensively
for the trade press about Taiwan’s indigenes, reported that these private guards
(aiyong) could be bought off and had conspired to allow the murder of Japanese
telegraph-wire stringers in February 1896.177 By 1898 camphor manufacturers were
hiring their own guards in addition to the ones employed by the Lins and subsi-
dized by the Japanese government. One company paid out ¥1,200 per month for its
own security.178 That same year, the Taiwan Government-General began recruiting
guards for its rural installations in the foothills. With the emergence of camphor
as a pillar of state revenue collection, the government-general abdicated its role as
referee between indigene and Han and jumped onto the playing field as a partisan
 From Wet Diplomacy to Scorched Earth    99

of the lumber business. The total number of government and privately employed
camphor guards totaled over 1,100 in 1898.
The majority were Taiwanese resident in the border areas, working under
Japanese police officers. With these forces, sporadic Japanese-led military offen-
sives were launched against villages accused of attacking government installations,
experimental camphor plots, or Taiwanese plainsmen. These episodic “punitive
expeditions” (tōbatsu) had mixed results—at times whole villages were burned,
at other times companies of Japanese officers with their squads were routed and
killed. Many battles were inconclusive.179
Recorded instances of armed conflict totaled 79 in 1897; they more than trebled
to 271 in 1898. In each of the years between 1898 and 1901 (inclusive), over 500
deaths and 100 injuries were dealt out to government forces by indigenes near the
guardlines. The great majority of the casualties were Han and jukuban Taiwanese
(see table 1).
A look at the financial stakes for the central government in Japan explains why
the Taiwan Government-General was willing to absorb so many losses. In 1896,
expenses related to war and civil government in Taiwan cost the Japanese Treasury
11 percent of its annual budget. From 1895 to 1902, subsidies to Taiwan were about
7 percent of the Japanese national budget. By 1898 there were calls in Japan for
the sale of the new colony back to China or to a European power.180 In this atmo-
sphere of crisis, Governor-General Kodama Gentarō and his civil minister Gotō
Shinpei took charge of the colony in February 1898.181 Gotō’s short-term solution to
the colony’s financial problems was a government monopoly on opium, camphor,
and salt. By instituting a government monopoly on camphor in June 1899, the
government-general not only received the profits from a lucrative export product
but also forced up the world price, since the only major competitor was Japanese
camphor. As table 2 illustrates, with the exception of 1903, which showed unchar-
acteristically low yields for the monopoly, camphor consistently supplied 13 to 30
percent of the Taiwan-generated revenue for the colonial state during the critical
1900 through 1907 period of Japan’s rule.
To finance projects in governmentality that would ultimately make the Han-
dominated lowlands into a disciplinary society (see introduction), Gotō secured
a Japanese government bond issue of ¥35 million in 1898. The proceeds were ear-
marked for railroads, land surveys, harbors, and other improvements. Between
1899 and 1905, the Taiwan Government-General received ¥31.2 million in install-
ments from this bond issue, which gradually overtook and then replaced Tokyo’s
annual subsidies. (For this period, Taiwan’s annual revenues, subsidies included,
were a little over ¥10 million.) The home government released this money expect-
ing the government-general to repay it out of the colonial budget.182
To increase the output of camphor and reduce injuries to the labor force, all
privately employed guards were brought under the command and employ of the
 Table 1.   Military encounters between the colonial government and Taiwan aborigines, 1896–1909

Killed Wounded Total

Year Cases Total casualties Casualties per case
Jpn Twn Jpn Twn Killed Wounded

1896 41 63 16 79 1.93
1897 79 151 15 166 2.10
1898 271 21 536 8 126 557 134 691 2.55
1899 293 21 510 6 144 531 150 681 2.32
1900 314 95 430 34 81 525 115 640 2.04
100     The Anatomy of a Rebellion

1901 342 7 503 2 121 510 123 633 1.85

1902 273 8 303 23 98 311 121 432 1.58
1903 140 4 225 7 53 229 60 289 2.06
1904 185 27 254 13 100 281 113 394 2.13
1905 196 43 284 8 67 327 75 402 2.05
1906 169 71 173 15 103 244 118 362 2.14
1907 172 61 269 53 247 330 300 630 3.66
1908 89 24 68 14 38 92 52 144 1.62
1909 203 35 155 41 143 190 184 374 1.84
Total 2,767 417 3,710 224 1,321 4,341 1,576 5,917 2.14

Sources: Data from Formosa Bureau of Aboriginal Affairs, Report on the Control of the Aborigines in Formosa (Taihoku: Government of Formosa, 1911), adapted from table iv, 45;
Mochiji Rokusaburō, Taiwan shokumin seisaku (Tokyo: Fuzanbō, 1912), 393; Taiwan jijō (Taihoku: Taiwan Government-General, 1923), 93; Riban gaikyō (Taihoku: Taiwan sōtokufu
keimukyoku [Taiwan Government-General Bureau of Police Affairs], 1935), 70–71.
 Table 2. Sources of revenue for the colonial administration, 1897–1907

Tax revenue Net revenue from monopolies Revenue generated in Camphor profit Camphor as % of Camphor as % of
Year
(¥ millions) (¥ millions) Taiwan (¥ millions)a (¥ millions) monopoly revenue local revenueb

1897 2.63 0.42 3.05 0.00

1898 2.91 1.47 4.38 0.00
1899 3.51 0.75 4.26 –0.24
1900 3.27 2.63 5.90 1.75 66.5% 29.7%
1901 3.52 1.83 5.35 1.09 59.4% 20.4%
1902 3.81 2.64 6.45 1.10 41.8% 17.1%
1903 3.82 1.85 5.67 0.27 14.7% 4.8%
1904 5.67 3.00 8.67 1.29 42.9% 14.9%
1905 7.65 3.54 11.19 1.99 56.1% 17.8%
1906 8.34 5.16 13.50 1.77 34.3% 13.1%
1907 8.54 7.00 15.54 3.58 51.1% 23.0%

a
For this calculation, I am only considering revenue generated in Taiwan (the total of tax revenue plus net revenue from monopolies); I am excluding subsidies from the Japanese
treasury and bond sales.
b
For this calculation, camphor profit is shown as a percentage of revenue generated in Taiwan.
Source: Data from Ka Chih-ming, Japanese Colonialism in Taiwan: Land Tenure, Development, and Dependency, 1895–1945 (Boulder, CO: Westview Press, 1995), 55–63.
From Wet Diplomacy to Scorched Earth    101
102     The Anatomy of a Rebellion

Japanese government in 1902. Then, on January 27, 1903, all affairs pertaining to
the indigenous territories were moved from various provincial and subprovincial
offices to the jurisdiction of the central police headquarters in Taipei. In March
1903, Kodama and Gotō convened a conference of all high-ranking officials on
the island to discuss a plan for ending “the Aborigine Problem.” The committee
was chaired by Mochiji Rokusaburō, a councillor in the Ministry of Civil Affairs.183
During his ten years in Taiwan, Mochiji worked in provincial administration,
engineering, communications, and education.184 In December 1902, he was sent
on an inspection tour of the aborigine border. The resulting “Mochiji line” was
approved and codified as a working document at the “Provisional Section for the
Investigation of Affairs in the Aborigine Territory,” convened in March 1903.185
As a proponent of the governmental modality discussed in the introduction,
Mochiji urged calculated restraint after an initial “big bang” of brutal landscaping
to create the playing field for modern economic growth. Instead of eradicating
the indigenes, which would have been an expensive, drawn out, and ultimately
pointless enterprise (from the standpoint of efficiency), Mochiji argued that
Japanese offensives be limited to areas that would bring in enough revenue to jus-
tify the expenditures of life and treasure.186 With these priorities in mind, Mochiji
proposed a three-pronged approach to savage border policy: (1) to unify com-
mand under one head; (2) to divide the indigenous territory into the north and
south—treating the northerners (Atayal, Saisiyat, Truku, and Sediq) as hostiles
and the southerners (Bunun and presumably Paiwan) as allies; and (3) to revive
the Chinese guardline system to surround the hostiles in the northern district.187
Mochiji’s academic credentials made him a true exemplar of “Meiji youth,” a gen-
eration for whom social Darwinism formed the backbone of a modern, scientific
worldview.188
According to Mochiji, in cases where inferior races came into contact with
superior ones, “history” taught that the inferior race was either eradicated in the
struggle for survival or assimilated to the superior race.189 His analysis suggested
that indigenous defense of local forests was an affront to imperial prestige: “Until
we solve this problem with the indigenes, we will not have sufficient cause to boast
to the outside world of our nation’s will and ability to expand and be enterprising.
The Aborigine territory occupies 56% of the island’s surface, and is a storehouse of
mineral, forest, and agricultural wealth. Unfortunately, the savage and cruel indi-
genes have thrown up a barrier to this storehouse of natural resources.”190
Echoes of treaty-port rhetoric are audible in Mochiji’s position paper. In the
1860s and 1870s, it was Paiwan treachery that endangered free trade and use of
the sea-lanes, thus thwarting “progress.” Now it was the “aborigine problem” on
Taiwan’s forested inland sea that frustrated imperial Japan’s drive to become a first-
rank nation-state.
From 1903 until 1909, the Taiwan Government-General extended guardlines
around the perimeter of northern indigenous territories. The extension process
 From Wet Diplomacy to Scorched Earth    103

involved building trails, stringing wire, and erecting guardhouses in cooperation

with hired, allied, or coerced Taiwanese. The case study of the Quchi alliance in
chapter 4 indicates that the networks, knowledge base, and control over trade
established during the Pacification Office and DAO periods made it possible to
close this perimeter before marching it inward. Large-scale coordinated military
offensives were the exception during the 1903–09 extension period; more com-
mon were skirmishes that came from attacks on advance posts, camphor work-
ers, and police installations. To build the first strongholds, Japanese forces put
mortars and other forms of mountain artillery at commanding heights to provide
cover for timbering, wire-hanging, and construction crews. All of these men were
exposed to sniper fire, boulder cascades, and sabotage from above. The Bureau of
Aboriginal Affairs recorded a total of 1,154 such confrontations for the 1903–09
period, for a total of 1,693 killed and 902 wounded on the government side. The
great majority of the casualties were Taiwanese (see table 1). One estimate of the
number of indigenous casualties put the figure at 30 percent of the number of
Japanese casualties.191
The next phase brought overwhelming force to bear on mobile, motivated, and
fugitive populations in the deeper mountain recesses of central northern Taiwan.
The price tag for this endeavor and its timing support this book’s contention that
the attempted integration of former Qing peripheries into the Japanese polity was
a continuation of the treaty-port project to accommodate higher volumes and
velocities of global commerce by expanding the radius of public power. While
Governor-General Sakuma lobbied to fund a military “surge” in Taiwan’s northern
and central mountain forests, the government-general also battled to secure funds
to expand and deepen Jilong Harbor.
In sync with guardline construction to pipe flows of camphor from interior to
port, Jilong Harbor had been subject to dredging and incremental improvements
since 1898 to push bulk commodities, including camphor, onto the world mar-
ket. Initially, the government in Tokyo dribbled out only small fractions of TGG
funding requests, often reducing or suspending allocations. In 1906, the twenty-
second Diet finally released a large sum—over ¥6 million—for a seven-year har-
bor improvement plan. But increased demand for Taiwanese rice and camphor
exceeded the harbor’s capacity to service sufficiently large cargo ships to clear the
docks. Therefore, the Diet earmarked another ¥5 million in 1912, to be disbursed
over a fourteen-year period.
By 1917, a total of ¥11,720,000 had been allocated, out of ¥15,700,000 million
requested over an eighteen-year period, to build Jilong Harbor.192 For the dura-
tion of colonial rule, Jilong dominated Taiwan’s foreign trade in ship traffic and
tonnage. Unlike the budget for Sakuma’s punitive expeditions, however, the Jilong
Harbor budgets were paid completely out of the Tokyo treasury.193 Nonetheless,
their roughly equal cost suggests the urgency and magnitude of the “Aborigine
problem” for advocates of an aggressive policy circa 1909.
104     The Anatomy of a Rebellion

Sakuma’s famous five-year plan (1910–14) was actually his second five-year plan.
The first five-year offensive was launched in 1907. Sakuma sought mass surrenders
and, apparently, glorious victories crowned with parades and triumphal arches.
These failing, Sakuma’s staff submitted a budget to the twenty-fifth Diet with
¥15,399,000 slotted for the construction, staffing, and movement of the guardlines,
to be disbursed and spent over a five-year period.194 This enormous sum was the
lion’s share of the estimated ¥18 million that Japan poured into the camphor wars
from 1904 through 1916.195 Even the abandoned first five-year plan was ambitious,
financially. In 1907 and 1908, a total of ¥3.58 million was allocated for the extension
and movement of the guardlines.196
The expense of this seemingly inconclusive war had its critics. At the February
1909 session of the twenty-fifth Diet, the firebrand representative Sasaki Yasugorō
grilled Prime Minister Katsura, Communications Minister Gotō Shinpei, and
Home Minister Harada Tōsuke for requesting large outlays of blood and trea-
sure to subdue a “dying race.” Sasaki was known as the “King of Mongolia” and
as a “continental adventurer” (tairiku rōnin) for his exploits in Central Asia dur-
ing the Russo-Japanese War.197 While Mochiji invoked world history to espouse
social Darwinism, Sasaki held up Zheng Chenggong and other imperial powers
as models of competence in aborigine administration, compared to the Taiwan
Government-General, which Sasaki considered inept.
Sasaki accused the cabinet members of refusing to answer questions, criti-
cized them for their inattentiveness to the science of colonial statecraft, and even
charged that measures to annihilate the indigenes were unconstitutional. At the
heart of Sasaki’s long harangue, which was interrupted by applause, was the charge
that the cost of indigenous suppression had no end in sight and would increase ad
infinitum until it broke the national budget. He doubted that suppression of the
indigenes was an urgent national matter.198
Despite Sasaki’s sharp questioning and the thirty signatures his inquiry gar-
nered, the budget passed. Fujii Shizue attributes the approval to Sakuma’s connec-
tions with Yamagata Aritomo. A hero of the Meiji Restoration, Yamagata was also
the father of Japan’s modern army, a prime minister, and a mentor to sitting prime
minister Katsura Tarō. In addition, Fujii argues, the Meiji Emperor himself took a
personal interest in the undertaking and lent his prestige to Sakuma’s proposal.199
Out of an estimated 1909 aborigine population of 122,000, there were 28,242
Atayals, the principal targets of the guardline movements. These guardlines were
manned and advanced by forces that varied between 6,000 and 7,000 inspectors,
engineers, laborers, guards, and police officers (mostly Taiwanese guards).200 A
description of the guardline’s physical aspect explains how these thousands were
deployed:
The guard-line has two components. The first is a road opened up in the mountain
areas. . . . The other is the guard station. This is a kind of a dugout, a sentry box with
 From Wet Diplomacy to Scorched Earth    105

gun holes which is protected from bullets and shells by impenetrable sandbags, earth,
stone, etc. Where necessary, electric wire fences have been hung and land mines have
been buried. An electric current was passed through the wires, and many upland-
ers and their domesticated animals have died from electric shock. At this time, field
and mountain artillery, as well as mortars were set up . . . As for the land mines, they
were installed by the police inspectors and their assistants; the construction was left
to police officers, who were all Japanese, and the assistant patrolmen, who were of
Chinese descent, were not allowed to install them for the maintenance of secrecy. It
appears that the electric wire fences began to be hung beginning in 1905. The elec-
tric switch was controlled by a Guardline Headquarters police inspector; since there
were only one of these stations per guardline, even the guards were electrocuted
from time to time.201

By 1912 the lines contained 756 guard stations, 427 branch stations, and 196 super-
intendent stations, extending to a total length of 226 miles.202 The organization of
these facilities was quite methodical, at least in its schematic form:

The width of the guardline’s path was 1.8 meters, and the foliage for 100 meters on
either side of the path was cut down so for the sake of visibility. The Chinese guards
protected the guard stations, and were equipped with rifles, which allowed them to
be mobile. Two to four guards manned each guard post. A branch superintendent
station was placed every four to five guard stations; officers and assistant policemen,
as well as guards were stationed here. For every four to five branch stations a superin-
tendent station built; besides the police and guards, inspectors and assistant inspec-
tors were stationed here, as well as the intermittent medical personnel or reserve
unit . . . amazingly, one of these posts was erected every 220 meters, which testifies to
the strength of the resistance to Japanese rule.203

The guardline’s organizational chart mirrored an ideal of centralized administra-

tion: priorities were set, and knowledge was concentrated, at the bureaucratic apex,
while sacrifice was demanded from those at the bottom. At the pinnacle stood the
police headquarters in Taipei, executing policy that was drafted with an eye to the
“greater good” of imperial Japan, as defined by the imperatives to balance annual
colonial budgets. At the nerve center of the guardline itself were the control sta-
tions. Here, Japanese inspectors were attended by physicians and heavily guarded.
The inspectors were entrusted with the electrical switches, as well. Below them
were Japanese nationals at branch stations and substations who were apprised of
mine placement, and then at the bottom rung were the exposed Chinese and juku-
ban units. They absorbed the preponderance of casualties and, for fear of security
breaches, were kept in the dark about operational secrets (the placement of mines
or the ignition of the electrical fences), to their own physical peril.
Schematic maps from the extension and movement phases of these pacifica-
tion campaigns represent the scorched-earth installations as circles and triangles
arrayed on a “line.” However, written accounts described a “hundred-meter-wide
106     The Anatomy of a Rebellion

clear-cut” path, while photographs attest to three-dimensional, undulating eco-

logical boot-prints. In a word, the term “guardline” (aiyūsen) soft-pedaled the
physical magnitude of the scorched-earth installations. The photographs below
document Japanese, Taiwanese, and indigenous guards, tree fellers, and porters
who walked this tightrope during the period 1903–15 (see figures 14–17).
As guards, porters, and soldiers, Taiwanese recruits were exposed to enemy
fire. In 1911, for example, 151 Taiwanese, compared to only 4 Japanese, were killed
advancing the guardline.204 Overall, over 2,000 government forces perished
(including Japanese), with another 2,200 wounded, in the camphor wars of 1904 to
1915.205 Taking a longer view, the death toll attributed to the war against indigenous
people—including the spike in resistance from 1897 to 1901 and the dead from
the Wushe uprising—amounted to 7,080 Taiwanese and Japanese, with another
4,116 wounded (1895–1935).206 Caroline Hui-yu Ts’ai reports that some “55,600
[Taiwanese] men” were commandeered to serve in the subjugation campaigns as
corvée labor.207 At its peak, expeditionary forces proper numbered 12,000 men
(including regular army and navy units).
Acknowledging the danger these exposed patrolmen, sentries, porters, and
infantrymen faced, day and night, the Taiwan Government-General promulgated
criteria to award special cash bonuses for meritorious service on the guardline in
July 1905. The rationale stated that this type of duty was no different from fighting a
war.208 In April 1907, another TGG internal order announced that medals would be
awarded to the bravest scouts and soldiers on the guardline; the governor-general
himself would bestow top honors; battalion leaders would dole out lesser awards.
The medals and prizes gave these punitive expeditions and related activities the
trappings of a glorious undertaking.209
There were three tiers within which the pay levels for the one-time bonuses
were set. Awards for inspectors and assistant inspectors ranged from the most
meritorious award of ¥60 (grade A), down to the grade D award of ¥30. Policemen
started at ¥35 and could earn as little as ¥12. Lastly, assistant policemen earned
from ¥20 to ¥8 (descending from A- to D-class awards). In other words, the most
meritorious assistant policemen would earn ¥10 less than the least meritorious
inspector and ¥15 less than the most meritorious policeman.
A list of sixty-four awards submitted to Governor-General Sakuma by Nantou
Prefect Koyanagi Shigemichi in May 1906 included commendations for twenty-
eight assistant patrolmen, all with Taiwanese names. The remaining inspectors
and policemen had Japanese names, meaning that Japanese guardline combat-
ants outearned their Chinese compatriots handily, while casualties and the risk of
death were much higher for Taiwanese participants.210
At the height of the five-year campaign, Ōtsu Rinpei, who directed operations as
Sakuma’s police chief, wrote the preface for a collection of bidan (glorious tales) about
the brave men of the guardline. Some of the men lionized in this volume were killed
in action; others were severely wounded. Ōtsu introduced these fifty-four sketches
Figure 14. Guards with bamboo and carved wood signaling devices to warn of the approach
of hostile forces, ca. 1910. Narita Takeshi, Taiwan seiban shuzoku shashinchō (Taipei: Narita
shashin seihanjo, 1912), 157.
Figure 15. Cutting trees to build the scorched-earth barricades, ca. 1910. “[T]here are many
old trees with huge trunks, and it can require many days for a single tree. . . . ” Narita Takeshi,
Taiwan seiban shuzoku shashinchō (Taipei: Narita shashin seihanjo, 1912), 161.

Figure 16. The large Taiwanese labor force, mostly sedan porters, at rest during construction
of the aiyūsen, ca. 1910. At its peak strength, the Japanese guardline forces employed over fifty
thousand Taiwanese laborers. Tōbatsu junsatsu kinen shashinchō (unpublished photograph
album). Courtesy of the Chiyoda City Hibiya Library and Museum, Tokyo.
 From Wet Diplomacy to Scorched Earth    109

Figure 17. Indigenous Taiwanese hauling food and water for the guardline ­expeditionary
troops, ca. 1910. “Savage of Hormosa [sic],” ip1219, East Asia Image Collection, Lafayette
­College, Easton, PA, accessed July 26, 2017, http://digital.lafayette.edu/collections/eastasia/
imperial-postcards/ip1219.

by magnifying the importance of the war. He wrote that Taiwan presented three
challenges to Japan: bandits, pestilence, and savages. The former two were eradicated
with policing and medicine in the early years of Japanese rule, while the last obstacle
was still being overcome. These tales describe the actions of men running into nests
of headhunters to secure strongholds, fighting gamely against ambushes, or collect-
ing intelligence in the face of grave danger. In total, the collection lauds the actions
of twenty-six Taiwanese guards and twenty-eight Japanese policemen. In contrast to
the cash bonus system’s racially determined hierarchy of danger and compensation,
the glorious tales were presented in chronological order with no preference for rank,
age, or nationality.211 The roughly equal representation and equalitarian format of
Ōtsu’s pamphlet stood in stark contrast to the patently discriminatory nature of day-
to-day operations, be it measured in terms of risk or reward.
There were four major offensives in Sakuma’s campaign, beginning with the
attack on the Gaogan Atayal near Jiaobanshan in 1910. The picture that emerges is
not so much that of a master plan coming to fruition as that of a centralized state
with vastly superior technology and resources exhausting itself in a border region
whose political economies, topographies, and traditions of armed resistance con-
firmed the wisdom of Mochiji’s formula—some areas of Taiwan would not be
worth the investment of blood and treasure to administer. In fact, the original
110     The Anatomy of a Rebellion

¥15.4 million budgeted for the campaign was insufficient. Supplementary requests
drove the costs up to ¥16.24 million. Although official summary reports buried
this figure, a contemporary economic analysis of the government-general’s devel-
opmental policies discovered that over half of this burden—about ¥8.3 million—
was paid out of the national treasury in Tokyo as supplemental funding.212
Although the five-year budget was drawn up with great precision, planners
could not predict the course of the war. Early movements were repulsed, soldiers
ran away, and the plan fell behind schedule. The 1914 Japan Year Book, which
assiduously updated annual entries regarding the war’s progress, reported omi-
nously: “The [first] three campaigns cost about [¥9 million], but only about one-
tenth of the program was effected.” In other words, well over half the budget had
been spent, and four of the five budget years had passed, but only 10 percent of the
plan had been completed.213 Time and money ran out that summer. The vaunted
plan abruptly ended with Sakuma’s declaration of victory over the Truku peoples
in August 1914.214 Although southern and eastern Taiwan remained untouched,
and the proposed network of roads and fences to secure the perimeter of the
indigenous territory remained a dream, Sakuma and his officers erected trium-
phal arches in various parts of Taiwan and paraded to celebrate.215 The Emperor
Taishō granted Sakuma an audience on September 19, 1914. The seventy-one-year-
old general reported to the throne, “the indigenous suppression undertaking has
been concluded.”216
Japan declared war on Germany on August 23, 1914, a week before the Truku
campaign ended. The six-week siege of the German port at Qingdao cost 1,400
Japanese casualties (400 deaths), while the Japanese Navy steamed southward
to occupy German holdings in the South Pacific.217 The opening of the so-called
Great War marked the end of imperial Japan’s attempts to incorporate indigenous
Taiwan into the empire. Sakuma’s successor, Governor-General Andō Sadami,
closed down organs and installations associated with the five-year plan between
July and September 1915.218 In 1917, the guard posts themselves were dismantled.
Their personnel rotated into police administration as mobile units in more scat-
tered but densely staffed police boxes.219
Tōgō Minoru and Satō Shirō, who finished their tome on Taiwanese colonial
development in February 1916, wrote hopefully that the resources expended dur-
ing the five-year campaign would be recovered as more indigenous territory in
Taiwan was added to the tax base through incorporation into “normal administra-
tion.”220 The official digest of Taiwan’s administration for 1915, however, was less
optimistic. It issued a skeptical and prophetic warning: “The work of subjugation
has thus been completed . . . , but this may not mean that the savages have been
reduced to submission. Some of them may still remain incorrigible at heart.”221
The subsequent 1916 edition was more pointed. The “work of subjugation,” it
reported, “has thus been completed over all the aborigines’ region, at least for the
present. Strictly speaking, out of the total tribes of 672 with 129,715 inhabitants, 551
 From Wet Diplomacy to Scorched Earth    111

tribes representing 116,744 have vowed allegiance and the balance of the 121 tribes
with 13,000 are still to be dealt with.”222 As testament to the TGG’s lack of appetite
for resuming Sakuma’s aggressive policies after World War I, subsequent Japan
Year Books repeated the 1916 summary verbatim, as boilerplate, for ten years run-
ning through 1927.

F R OM M I D D L E G R OU N D T O NAT I V E AU T HO R I T Y

The historical contours of wet diplomacy in Taiwan’s uplands recall the fate of
North America’s “middle ground.” Historian Richard White’s study of eighteenth-
century frontier diplomacy argues that as long as Algonquians and Ojibwas had
the leverage to force British and French agents onto the “middle ground” between
state and nonstate spaces, gifts, feasts, and oaths were required to secure indigenes’
provisional allegiance. By the early 1800s, writes White, the United States had neu-
tralized the French and British threats to its sovereignty. Without rival powers to
divide its energies, Washington trained its guns on Indians and enforced directives
in Algonquian and Ojibwa territory with ultimatums and firearms instead of with
banquets and gifts.223
Echoing this pattern, wet diplomacy in Taiwan prospered when indigenes
occupied the middle position between contending powers: the Japanese state and
Han guerilla fighters. The government-general’s fear of a two-front war recom-
mended a policy of negotiation, feasting, and gifting. Fiscal and political pres-
sure to produce income from the camphor monopoly made wet diplomacy less
tenable, however. After Gotō Shinpei’s mass annihilation of Han rebels in 1902,
Taiwan’s middle ground became harder to maintain. Even so, while the Taiwan
Government-General from 1903 onward trained its forces on indigenous peoples,
tōmoku and seiryokusha (headmen and men of influence) remained important
intermediaries, resembling the functional group “native authority” as described
by Mahmood Mamdani.
The members of the native authority, who were recruited and coerced by the
Taiwan Government-General in the 1910s and 1920s during the peak of the guard-
line movements, had achieved local leadership status through prowess in media-
tion or battle, as leaders of ritual groups, or as members of chiefly lineages. The
government-general supported the authority of these men with field trips to Tokyo
and Taipei, stipends, badges, medals, and cash awards for meritorious service. This
generation of mediators, despite these emoluments, were not utterly dependent
upon the state for authority nor were they disciplined imperial subjects. Some
remained incorrigible at heart.
By 1933, there were 186 Atayal, 92 Bunun, and 128 Paiwan officially recognized
tōmoku. Of 431 headmen for all of Taiwan, only 22 could understand Japanese; a
mere 20 had attended a Japanese educational facility (see chapter 2). Therefore,
while members of the native authority might have been amenable to bribes,
112     The Anatomy of a Rebellion

threats, or expedient alliances, they could not be counted on to prosecute Japan’s

interests for love of country, the emperor, or a belief that the new order would
bring progress—the hallmarks of discipline (or interiority).224 In short, Japan’s ver-
sion of native authority in Taiwan was a poor instrument for establishing public
power in the interior, and it can be viewed as a stopgap measure for a government
that could no longer expand its territory through the use of massed force.
Mona Ludao, who would lead the Wushe Rebellion in 1930, himself played the
role of tōmoku. He traveled to Nantou City and Japan as a government-sponsored
“tourist” in 1911, along with dozens of other tōmoku and seiryokusha. At the height
of the guardline movements, four troops of headmen were brought to Japan to
overawe them with views of Japan’s economic, demographic, and military power.
According to TGG police records, Mona himself was a revered warrior and a
physically imposing man, recognized by other Tgdaya as a worthy successor to his
father, Ludao Bai, a local chief (domoku).225
Ludao Bai’s adoption of Japanese son-in-law Kondō Gisaburō, in fact, had one
foot in the world of wet diplomacy and another in the scorched-earth policy. To
consecrate the marriage of Kondō and Ludao Bai’s daughter Chiwas and activate
the new Japan-Tgdaya alliance, the Taiwan Government-General contributed six
oxen for slaughter and twenty oil cans of sake, most of it distributed at a wedding
feast at Kasumigaseki (near Wushe) in 1909. Ludao Bai’s son Mona Ludao medi-
ated between Sediq men and their new Japanese in-laws during the drunken and
tense proceedings. The debauch carried on for about a week. Afterward, about 650
of Ludao Bai’s followers, thanks to this wedding feast and promises (unfulfilled) of
enemy heads, supported guardline extensions at the expense of Truku and Toda
settlements just north of Wushe (see chapter 2).226
A decade later, the government-general called upon Mona Ludao. Much has
been written about the Salamao incident of 1920. Official sources consider Mona
a conspirator against Japan during this affair, whose bad aims were thwarted by
leaks. Oral histories and documentary evidence, on the other hand, suggest that
Mona helped lead, or at least mobilize, about 560 Tgdaya troops to punish Salamao
rebels. The incident began with an influenza epidemic attributed to the presence
of outsiders. The most infamous in a series of Salamao raids and ambushes was an
attack on a police station that took nineteen Japanese lives. As a condition of their
earlier surrenders, and to avoid suspicion of conspiracy, warriors from Malepa,
Xalut, and several Tgdaya villages signed on to fight with Japan. Fujisaki Seinosuke
judged Japan’s crushing victory a turning point, noting that the wrath of imperial
forces frightened indigenes throughout northern Taiwan into submission.
Reminiscent of the Stone Gate Battle of 1874 and the Gantaban slaughter of
1903 (see figure 13), Salamao heads were displayed publicly to heighten the dem-
onstration effect. Twenty-seven heads were photographed in front of the Wushe
police station as trophies of the punitive expedition. Five heads were brought back
to Malepa, where they were feted at a large head-taking festival—a long-banned
 From Wet Diplomacy to Scorched Earth    113

practice that the police allowed on this one occasion, since the killings were gov-
ernment sanctioned.227
Mona Ludao’s involvement in the 1920 maneuvers is sporadically documented;
the contested nature of the records reflects his structural position as a member of
native authority, which inclined him toward double-dealing. His clan expected
material benefits for his family and followers by taking Japanese in-laws such as
Gisaburō or contributing fighting men and laborers to Japanese expeditions. At
the same time, chiefly cooperation was forced, on pain of the resumption of bom-
bardment from mountain guns or of trade embargoes, leaving men like Mona
with little choice. On the other hand, alliances with Japan, as the above campaigns
showed, provided young Atayal, Sediq, Bunun, and Truku men opportunities to
take heads and exhibit bravery in battle—rare occasions, since head taking had
technically been outlawed.
But to favor Mona’s men, thereby propping up his stature as a tōmoku, the
Taiwan Government-General perforce eroded the prestige of Mona’s Truku and
Toda counterparts or the Salamao chiefs, just as the government-general had
taken sides against Paalan (Tgdaya) in 1903, when they sanctioned the Gantaban
head-taking binge. In other words, the favoritism entailed by wet diplomacy was
anathema to the rationalization of state machinery, while it was the very antithesis
of isshi dōjin (impartiality).
Mona himself, according to one memoir, was known to burst into anger at the
mention of his role in Salamao. Official records describe him as a reluctant impe-
rial soldier—the government-general marked him as a malcontent. Kondō himself
complained that his Wushe in-laws had donated twenty thousand man-hours of
labor to building the guardline in early 1909, while their fields lay unattended.228
Kondō argued that long-term resentment at the government’s ingratitude, which
could have been allayed with a few kegs of sake and an official thank-you or a few
more guns and bullets for Mona and his allies, ultimately led many locals to sym-
pathize with and participate in the Wushe Rebellion of 1930.229
Therefore, Sugiura’s rebuff of Mona’s offer of mediation after the Yoshimura
beating incident on October 8, 1930, should be seen as the last straw, one that
added humiliation to long-standing torment. Of course, the relationship was also
unsatisfactory from the TGG side. Accordingly, Japanese officials took steps to
break their reliance on Mona’s generation of headmen in the 1930s and to replace
them with more dependent and reliable intermediaries, whose loyalty would not
be purchased with canisters of alcohol, head-hunting licenses, or the promise of
preferential treatment at government gun depots.
 2

The Longue Durée and the Short Circuit

Gender, Language, and Territory in the Making of
Indigenous Taiwan

Linguistic variation was the order of the day in nineteenth-century East Asia, as
for most of the world. Before the advent of universal education and vernacular
dictionaries, humanity’s thousands of mother tongues were geographically cir-
cumscribed.1 As we saw in chapter 1, communication by relay through chains
of interpreters severely limited the scope and subtlety of linguistic transac-
tions between Taiwan’s natives and newcomers.2 However, as long as states did
not require routinized and rapid communication between capitals, and capitals
between peripheries, the slow and halting process of cross-border translation did
not become an overriding concern of central administrators.
Higher state capacities, longer imperial reaches, and administrative integration
became imperative for dynastic states with the arrival of international society in
East Asia on the helms of steam-powered gunships. This was the lesson future
Meiji Restorationists learned from Perry’s gunboats in 1854 and the Qing from the
1874 Taiwan Expedition that brought 3,658 Japanese occupiers to Langqiao. Now, a
ship’s crew running afoul of coastal wreckers on distant shores could be construed
as injurious to national dignity or, perhaps worse, the steady flow of population-
sustaining commerce. Left unresolved, such occurrences often spiraled into inter-
national incidents that brought indemnities and unwanted treaties in their wake.3
Accordingly, the Qing launched numerous attempts to extend its administra-
tive reach deeper into Taiwan’s interior after the shock of ’74, with varying degrees
of success. Nonetheless, when the Japanese regime replaced the Qing in the sum-
mer of 1895, the riddle of central command and control across Taiwan’s complex
linguistic landscape still awaited a solution.
When Japanese troops and officials arrived in northern Taiwan during May
1895, only a small minority of Taiwanese could speak the dialect of Chinese, Beijing
114
 The Longue Durée and the Short Circuit    115

“official speech,” that Japanese interpreters could understand. The language bar-
rier was made more insurmountable by the Meiji-period drift away from Chinese-
language training in favor of Western languages. For example, of the forty students
enrolled at the Army Staff College in 1885, “twenty-five studied French, fifteen
German, and none Chinese.” As for study abroad, the numbers reveal the same
lack of institutional interest in China. Of the seven hundred ninety-two graduates
of the Army Staff College in 1883–1914, eighty-one studied in Germany, thirty-three
in France, twenty-nine in Russia, twenty-four in Britain, and thirteen in China.4
A few unconventional military men and commercial innovators kept their
sights on China despite the “civilization and enlightenment” period’s fascination
with the West. Most famously, the “continental adventurers” (tairiku rōnin) advo-
cated a China-first orientation for Japan’s foreign relations. Foremost among them
was Arao Kiyoshi, who transferred to the China section of the General Staff from
the Kumamoto infantry in 1885. Aside from presumed cultural ties and a collec-
tive sense of grievance at Western gunboat diplomacy, rōnin of Arao’s disposition
believed that Japanese could not hope to compete successfully with Western mer-
chants in Europe and the Americas. Therefore, Arao and like-minded rōnin saw
Sino-Japanese trade as a potential engine for Japanese national wealth and large
military budgets in the era of “survival of the fittest.”
Arao was sent to Shanghai in 1886 to gather military intelligence. He con-
ducted operations under cover of pharmaceutical giant Kishida Ginkō, founder
of the Ginza-headquartered pharmaceutical company Rakuzendō. Kishida opened
Rakuzendō branches in Shanghai, Fuzhou, and Hankou in the 1880s. Kishida was
also Japan’s first journalistic war correspondent, whose eyewitness coverage of the
Taiwan Expedition of 1874 made him a national figure. On the mainland, Kishida
provided lodging, food, and cover for Japanese students of Chinese language and
for government spies. Under Kishida’s and Arao’s tutelage, Rakuzendō “employees,”
Japanese men wearing queues and dressed in tunics, fanned out across the conti-
nent to learn the languages, geography, commercial habits, and folkways of China.
After his return from China in 1889, Arao vigorously promoted plans for a
large, government-assisted Japan-China Trading Company to coordinate strat-
egy and resources for Japanese commercial expansion onto the continent. Arao
also proposed an attached training institute for Chinese language, geography, and
business practices. From cabinet members to public audiences of potential stu-
dents, employers, and entrepreneurs, thousands of Japanese heard Arao’s message.
Promised public funding for the trading company and research institute, however,
evaporated with the fall of the Kuroda cabinet and the seating of the fiscally con-
servative Constitution Diet in 1890.
Nonetheless, the Nisshin bōeki kenkyūjo (Sino-Japanese Commercial Research
Institute) opened with a class of 150 students, selected by exam from a pool of
300 applicants that same year. Arao’s Nisshin bōeki kenkyūjo established a rigor-
ous four-year curriculum that taxed many students beyond endurance. The school
116     The Anatomy of a Rebellion

was always on the brink of bankruptcy. Had it been a completely private concern,
it surely would have collapsed. Arao’s dense network of military connections in
Tokyo made the difference. Vice-Chief of Staff Kawakami Sōroku, in charge of
intelligence gathering in China, garnered emergency funding from the Diet and
kept the research institute afloat. Though the institute cannot be called a govern-
ment school, the Japanese Army lent it facilities for recruitment. In June 1893, its
first graduating class of 89 finished just in time for the start of the Sino-Japanese
War. Over 70 of these 89 graduates served as army interpreters or spies during the
1894–95 conflict.5
Arao’s research institute could not, however, meet the demands of a protracted
continental war. Apparently, the first call for interpreter applications met its yield
target in September 1894. The theater of war quickly expanded, however, and by
October the army began inducting less-qualified translators out of desperation. For
example, forty students at the Kumamoto Kyūshū Gakuin were given accelerated
training and pressed into service to make up for the shortage.6 Mori Ushinosuke,
who studied Chinese at a Nagasaki commercial school before the war’s outbreak,
recalled:
At the time, anyone who could communicate in Chinese was recruited by the Army
and sent to the Liaodong battle area; soon a shortage of manpower developed, and
they began to use men who had barely a whiff of Chinese language training. If I think
about it now, it is a kind of joke, because even those of us who could not understand
Chinese were added to the ranks of these men. Since I did not continue with this
work, I have long forgotten the words of Chinese that I did learn. But though I have
no knowledge of Chinese at the present time, my associates and I, the ones unskilled
in Chinese, were soon studying and being drilled, and we became very good Chinese
translators.7

Those who answered the call to serve in the theater of war were classified as army
auxiliaries and paid at the bottom of the officer’s scale, between twenty-five and
fifty yen per month. When the mainland phase of the war concluded in April 1895,
many wartime interpreters went on to work for Japanese firms in mainland China,
clustering around Shanghai and Hankou. Others went to Taiwan.8
Of the roughly 150,000 Japanese men activated for the Sino-Japanese War from
June 1894 through the dissolution of imperial headquarters on March 31, 1896, over
a third served in Taiwan at one time or another.9 Eleven translators accompanied
Colonel Fukushima Yasumasa’s expedition to Danshui on August 9, 1895. Their
mission was to establish a Japanese capital in Taipei. The interpreters came directly
from Lüshun, Liaodong Peninsula, still wearing winter jackets. Highlighting the
scarcity of interpreters, Fukushima refused an urgent request to spare a few to the
Imperial Guard. Indeed, in the report he filed, interpreters accompanied every
landing party and participated in missions great and small, dangerous and routine.
For their contributions to the mission, completed on August 25, 1895, interpreters
 The Longue Durée and the Short Circuit    117

were awarded thirty-yen bonuses for meritorious service, on top of monthly sala-
ries that ranged from thirty to sixty-five yen.10
Nonetheless, the bonuses did not provoke a flood of reenlistments. The Daily
Yomiuri announced on September 12, 1895, that the military police (kenpeitai) in
Taiwan “needed sixty translators.” Of the three regiments sent from Hiroshima to
Taiwan on September 2–3, one left with two translators; the rest embarked mute,
waiting for help to catch up later, if it could be found. Men with abilities in spoken
and written Chinese, who could “transmit orders, warnings, and information to the
natives” were scarce indeed, pushing salaries to between thirty and seventy yen.11
The veterans and graduates of the commercial schools who could be persuaded
to cross the strait did not speak Minnan dialects (in Japanese, called Taiwango or
Forumosago, or Taiwanese) or indigenous languages. By the same token, most
Taiwanese did not speak the Beijing dialect known to Japanese interpreters as
Southern Administrative Chinese (to distinguish it from Manchu, or Northern
Administrative Chinese). Therefore, additional talent had to be recruited in
Taiwan itself. In the early years, the military hired Taiwanese fluent in Mandarin as
“auxiliary translators” (fukutsūyaku) to work with the Japanese translators. Brush
talking was heavily utilized, leading to frustration and miscommunication.12 A
Yomiuri correspondent provided a brief sketch of the early history of communica-
tion problems in Taiwan in this August 19, 1895, dispatch from Houlong, Miaoli:
At first, when our military took Jilong and then occupied Taipei [May–August 1895],
natives [dojin] who could speak Mandarin [kango] were rare. This caused difficulties
for our official translators [tsūyakukan], needless to say. It got to the point where cer-
tain lieutenants asserted they would not use official translators. Although the num-
ber of natives who spoke Mandarin was insufficient as far as the official translators
were concerned, happily everyday communication could be conducted without too
many tie-ups. Moreover, as the days passed, many of the official translators learned
to understand Taiwanese [dogo].
As this division began its advance south, from Taipei to Xinzhu, and then from
Xinzhu to Houlong, the language changed radically along with the topography. . . .
The natives in the Houlong area do not speak Mandarin, so many discussions were
conducted by brush talking [hitsudan], which was bothersome. To be sure, Hou-
long’s numerous so-called jukuban speak their own local version of Chinese, so some
can speak a little Chinese. Nonetheless, unless some way can be found to translate
this [local dialect] into Mandarin, the [Japanese] official translator cannot under-
stand. This means that if another translator is brought in, three layers of interpreta-
tion have accumulated . . .
Again, brush talking is no simple matter, there are many idiosyncratic characters
and colloquial phrases [in use among the locals]; often standard Chinese writing is
not understood.13

The small corps of Japanese and Taiwanese military police with smatterings of
language training could not meet the demand for translators that came with the
118     The Anatomy of a Rebellion

transition from regimes of punishment to regimes of discipline. By November

1896, the Civil Affairs Bureau and provincial administrations employed seventy-
nine translators, all of them Japanese.14 Since the colony was in acute fiscal crisis
in its early years, augmenting this expensive corps was impossible. An economical
compromise was proposed on July 25, 1897. Taiwan prefectural officials proposed
ten-yen monthly pay supplements as incentives to keep policemen proficient in
Taiwanese from escaping to the private sector and to encourage further efforts at
language acquisition, but the proposal was rejected.
The administration of Kodama Gentarō and Gotō Shinpei (1898–1906) is often
credited with bringing discipline to Taiwan. As part of a flurry of new measures,
in April 1898, Imperial Order 68 was issued. It stipulated that hannin rank (fourth
class) civil officials, police officers, and jail guards were eligible for stipends of up
to seven yen per month as auxiliary translators. Qualified applicants would be paid
bonuses to act as translators in addition to performing their prescribed duties.15
Evaluations of the first round of applications for auxiliary-translator pay supple-
ments (kensho) in May 1898 indicate that about 18 percent of Japanese policemen
had at least some familiarity with “Taiwanese,” while 6 percent of patrolmen and
9 percent of officers could “communicate effectively” or were considered “fluent.”16
Some 82 percent of Japanese policemen had no familiarity with a Taiwanese lan-
guage in mid-1898. Even in the mother country, this state of affairs appeared alarm-
ing. In June 1899, the Taiwan Society, a group of powerful lobbyists that included
ex-governors-general and some of Japan’s top politicians,17 lauded the kensho pol-
icy, explaining that the language barrier had fomented much “Taiwanese resent-
ment and Japanese anger.”18 In October 1899, the government-general responded
again. Taiwanese language institutes were attached to subprefectures (benmusho)
and police stations to increase the number of qualified auxiliary translators. In
September 1900, the Land Survey, Railway, and Camphor Monopoly Bureaus
added language institutes to their local stations, as well.19
The Taiwan Government-General instituted the camphor monopoly in 1899
as one of several programs to address the fiscal crisis. After a rocky start, it estab-
lished itself as a cash cow, thus impelling the state to abandon its role as peace-
maker between indigenes and camphor workers. The government-general allowed
existing private armies to guard camphor fields from the start. But with the cam-
phor monopoly in place, it deployed its own policemen to protect the workers. It
also devised legal machinery to dispossess indigenes of forest lands. These related
projects touched off a rash of violent reprisals, which elevated the communication
gap with rebelling Atayal, Sediq, and Truku speakers in Taiwan’s north to a matter
of national security.
The government-general thereafter launched programs in Austronesian-
language education for Japanese policemen and Japanese-language training for
Taiwanese. It also sanctioned and bankrolled intercultural marriages to over-
come linguistic divides that were formerly bridged by unreliable, free-booting
 The Longue Durée and the Short Circuit    119

interpreters. In the short term, these policies appeared successful. By 1915, the
government-general had achieved a military victory by disarming most “north-
ern tribes,” as we saw in chapter 1. But while the state more or less established a
monopoly on the legitimate use of force along the savage border by 1923, it did
not horizontally integrate the fractured political and linguistic terrain of the so-
called Aborigine Territory, nor could it instantiate a disciplined society beyond
the edge of the baojia units (regularly administered territory). The costs associated
with training its own officials in myriad indigenous languages or building a school
system that could shift the burden of translingualism to the indigenes proved too
high for the always cash-strapped government-general.
Therefore, the government-general abandoned its mission to enforce the stan-
dard of sovereignty outlined in Ōkubo Toshimichi’s 1874 admonitions regarding
the Qing’s uneven and dissipated sovereignty on its borderlands. What had been
an expedient policy of benign neglect in the late 1890s hardened into a conscious
policy of legally pluralistic, ethnically bifurcated administration in Taiwan.
During the 1920s, budgetary constraints and the appearance of calm reduced
further the state’s allocations for manpower and infrastructure in the highlands.20
Yet the Taiwan Government-General stepped up its resource-extracting activi-
ties. Postmortems of the Wushe Rebellion blamed the overbearing demands of
the state, in the disciplinary vacuum left by government on the cheap, for the
conflagration. As we shall see, the stunted communications network encour-
aged by recourse to political marriages and reliance on tōmoku (headmen) cre-
ated numerous blind spots for the state. As the 1874 invasion shocked the Qing,
the 1930 Wushe uprising prompted the government-general to launch sweeping
administrative reforms.
The first order of business was to reform the police service and rid aborig-
ine administration of underachievers and shady characters.21 The tōmoku who
remained “incorrigible at heart” were the next targets. A concerted effort to reform
aborigine customs utilized discipline-making techniques, with an emphasis on
“imperial subjectification” (inculcating “Japanese spirit”). However, although these
policies were superficially assimilationist, they did not portend the horizontal inte-
gration of the colony but rather reinforced the boundaries of an ethnic enclave.
To intensify imperial subjectification (kōminka), the government-general
demoted “chiefs” (shūchō, tōmoku, shuryō, domoku) or “village elites” (­seiryokusha)22
and promoted a generation of young “pioneers” (senkusha) and youth associations
(seinendan)23 to replace them as intermediaries. In the bargain, the daughters of the
chiefs and village elites ceased to function as prominent political actors. Women
such as Tata Rara, Iwan Robao, Yayutz Beriya, Yawai Taimu, Chiwas Ludao, and
others, and the Japanese men who married into their families, were no longer
needed to maintain imperial rule in the uplands. This shift appears to have smooth-
ened the linguistic interface in the indigenous territories, preparing the ground for
its horizontal integration into the rest of the colony.
120     The Anatomy of a Rebellion

At the same time, decades of legal discrimination, economic protectionism,

and third-rate education had locked the majority of indigenes, no matter how
proficient in Japanese, into the status of perennial subalterns. By 1930, aborigine
administration had become its own interest group, with publications, associations,
and an esprit de corps. Moreover, the ethnic tourism industry, academic anthro-
pology, and Japanese literary production discursively constituted the indigenous
territory as “the real Taiwan.”
To be sure, indigene enrollments in night schools, police-station annexes
(small installations dedicated to education), and elementary schools increased
in the 1930s, as did the number of years of schooling for a minority of students.
On the other hand, the variety of subjects taught, the quality of instructors, and
opportunities for higher education remained limited by comparison with the lofty
rhetoric of imperial subjectification. By the 1930s, although it went under the ban-
ner of assimilation and was dedicated to the diffusion of Japanese language and
elements of home-country comportment, indigenous education came to mean
“education to remain indigenous.”24 At the time of war’s end in 1945, the second-
order geobody of Taiwan’s indigenous territory had achieved unprecedented levels
of linguistic and infrastructural integration, via Japanese language-training pro-
grams, state-organized pan-indigenous institutions, and the Taiwan Government-
General’s road building program. In some respects, the indigenous territory bore
many features of a community that could be imagined as a nation on the eve of
Japan’s departure with the empire’s defeat. On the other hand, the indigenous terri-
tory’s denizens lacked rights in property, were cut off from the economic life of the
rest of Taiwan, and were relegated to the role of toiling, honest, and loyal children
of the emperor. In a word, the 1930s imperialization policies finished the job of
turning erstwhile Qing border peoples into modern ethnic minorities.

T H E G E N D E R D I V I SIO N O F L A B O R I N C R O S S - B O R D E R
C OM M U N IC AT IO N S

Before the Qing government could even sign the Treaty of Tianjin of 1858, foreign
adventurers in Taiwan were testing the waters for trade and missionizing possibili-
ties. In 1857, future British consul Robert Swinhoe wrote: “I had the pleasure of see-
ing a few [Indigenous] women, who were married to Chinese at . . . [Langqiao] . . .
a Chinaman named Bancheang, of large landed property, traded with the Kalees
[Rukai] of the hills . . . He was constantly at variance with the Chinese authorities
who had outlawed him, but could not touch him, as he was so well defended by his
numerous Chinese dependants, and the large body of Aborigines at his beck. This
man was wedded to a Kalee.”25
As we learned in chapter 1, there was no shortage of “outlaws” like Bancheang
in late nineteenth-century Langqiao. Swinhoe’s vignette, however, is notable for
exposing an element of Qing borderland society obscured by the outsized impact
 The Longue Durée and the Short Circuit    121

of the Mudan village incident on historiography. The preponderance of docu-

mented transactions between natives and newcomers in 1860s–1870s Langqiao
ran through the Siaoliao-Sabaree-Tuilasok corridor, or the Checheng-Poliac-
Kuskus route along the Sichong River corridor. Male mediators such as Siaoliao
native Miya and Poliac subofficial Yang Youwang (see chapter 1) are prominent in
these records, because of their pivotal roles in the dramatic events of the epoch.
Douglas Fix’s research into Langqiao social history suggests, however, that female
mediators also operated in this milieu: “Another group, unnamed older women—
nearly all labeled ‘aborigine’ wives of headmen—seem to appear just as frequently
as the [Miyas] in my sources. Their existence and recorded activities indicated
that cross-community mediators were ‘born’ at the intersection of kinship connec-
tions, munitions markets, and castaway exchanges. Furthermore, the regularity of
their mediation suggests that a gendered analysis of the historical constitution and
reproduction of frontier communities is long overdue.”26
Xu Shirong’s quantitative study of late Qing Hengchun social structure also
notes the high incidence of intermarriage between Hakka residents of Langqiao
and indigenous peoples. According to Xu, these intermarriages were woven into
relations of mutual economic interdependence.27 It may well be that male inter-
mediaries such as Miya or Yang, who would qualify as “local gentry” under some
definitions, typically dealt with conspicuously marked and well-armed official ret-
inues of outsiders, while female interpreters such as those mentioned by Swinhoe,
Fix, and Xu interfaced with smaller parties of traders, explorers, and missionaries.
For example, while traveling in a small party, William Pickering—whom we
met earlier as LeGendre’s interpreter—encountered a chief whom he referred to
as a “T’ong-su” (tongshi), the “headman of the tribe, responsible to the Chinese
government.” Pickering spoke to him through an “old woman, named Pu-li-sang,
[who] was no novice to the ways of civilization, as she had, years ago, been mar-
ried to a Chinese, and also had lived for some time with the [Tsou] Bangas . . . ”28
On the eve of the Japanese invasion in 1874, American naturalist Joseph Steere
also noted the role of indigenous women as mediators, writing that the “Kale-
whan [Rukai], in times of scarcity, frequently sell their daughters to the Chinese
and Pepo-whans [plains indigenes], who take them as supplementary wives and
make them useful as interpreters in thus bartering with the savages. While we were
among the Kale-whan the chief offered to sell us three girls of the tribe at twenty
dollars each.”29
In world history, this was not an unusual arrangement; there are numerous
examples from other times and places. In many cases, intercultural marriages were
symbiotic, because outsiders provided access to far-flung trade networks or mili-
tary alliances that could shore up a family’s ascendancy in a fragmented milieu
of stateless polities. In Taiwan, examples of such unions can be found in records
going all the way back to the early seventeenth century. In fact, the term tong-
shi, employed by Pickering above, evokes a long history of semiofficial, bicultural
122     The Anatomy of a Rebellion

intermediaries who functioned on the fringes of Qing territory beginning in the

Dutch era (1622–61).30
Over the centuries of Qing rule (1683–1895) over Taiwan, tongshi were the pri-
mary conduits of trade and diplomacy during the era of tributary interstate rela-
tions in East Asia. The English word “interpreter” is a good approximation, but
tongshi were more than linguistic adepts. They organized commerce, mediation,
and taxation in the most sparsely populated areas of Taiwan. As would be the case
during the period of Japanese colonial rule, tongshi often appeared as suspicious,
disruptive, and not altogether trustworthy characters in Qing records.
Many of these intermediaries provided continuity between the successive
Qing and Japanese imperial regimes in Taiwan. It is evident that ample room for
cross-border maneuvering existed in Taiwan even after two centuries of Qing rule.
On the eve of the handover to Japan in 1895, for example, Qing critics judged
Shen Baozhen’s and Liu Mingchuan’s kaishan fufan policies to “open the moun-
tains and pacify the indigenes” a failure. After twenty years and “several hundred
million taels spent on road construction, military deployment, suppression and
pacification of aborigines,” Han settlement had not increased appreciably outside
southernmost Taiwan, while outlays remained exorbitant.31 Nonetheless, several
individual campaigns and measures bore fruit, which supplied a small coterie of
seasoned intermediaries who in turn formed the basis for the successor Japanese
state’s attempt to integrate the indigenous territories into its empire. In the south
(to be discussed in more detail below), a walled city, Hengchun, was built in 1875 as
a base for pacification campaigns into the mountains. There, veteran Paiwan and
Puyuma participants in the treaty-port diplomacy of the 1860s and 1870s signed
on as Qing subofficials; many would later work for the Japanese. In the north,
administrative beachheads were established near Wulai, Yilan, and Dakekan,
which also became early base areas for Japanese rule in the indigenous territory.
The latter three northern examples were gateways into Atayal-dominated
mountain interiors. There, various types of Atayal-Chinese interpreters mediated
mountain-plain or state-society relations. Males, whether as subofficials or com-
mercial operators, were called tongshi (interpreter). This title could be attached
to a chief who collected information and mobilized labor for the Qing or to an
interpreter for hire who had acquired language proficiency and political connec-
tions through intermarriage. In 1885, Governor Liu Mingchuan dispatched his son
Chaohu to Quchi (near Wulai) at the head of three thousand soldiers. The eight
villages of the area were named for the paramount chieftain Marai, who submitted
to Liu that year. Marai was then deputized to collect information, coordinate gift
distribution, and report to Liu for six Qing liang (Japanese ryō) per month.
Marai passed away around 1888 without a successor. A Wulai chief named
Watan Yūra, the man we met in chapter 1 as Ueno Sen’ichi’s drinking partner and
source of Atayal clothing and jewelry, filled the vacuum as the paramount head-
man of the Quchi tribes32 (see figure 18). The interpreters known to us as Aki and
 The Longue Durée and the Short Circuit    123

Figure 18. Watan Yūra, ca. 1900. Goto Shimpei, “Formosa Under
Japanese Administration,” The Independent 54, no. 2,796 (July 3, 1902):
1,582.

Kōan have been photographed with Watan Yūra (see figure 19). Aki and Kōan were
Chinese by birth but became “Atayalized” over the decades. They moved into the
mountains sometime in the 1870s.
Kōan was married to an Atayal woman and worked with the Japanese regi-
ments stationed in Xindian in February 1896. Taiwan was still under military rule;
the Xindian garrison was fighting Han rebels in Yilan. As the troops set out to
survey a direct overland route from Taipei to Yilan, Kōan presented himself as an
interpreter with local connections to the Quchi tribes. He arranged for Japanese
forces to meet with the paramount chief of the eight villages, Watan Yūra, near
Quchi on February 10. This led to a long series of parleys. The Japanese provided
rice, liquor, gifts of various kinds, and assurances of protection from Chinese set-
tlers in return for Quchi acquiescence in Japanese roadbuilding. These negotia-
tions were carried out through an official Japanese translator, Kōan, and an Atayal
man named Shiron, who had lived in Taipei and learned Chinese in the 1880s at
Governor Liu Mingchuan’s Academy for Aborigine Boys.
If Kōan’s wife participated in these negotiations, her name was not mentioned.
Female interpreters were referred to not as tongshi but as banpu (indigenous
124     The Anatomy of a Rebellion

Figure 19. Watan Yūra, Kōan, Aki, and Pazzeh Watan in 1903. Mori Ushinosuke, Taiwan
banzoku zufu, vol. 1 (Taipei: Rinji Taiwan kyūkan chōsakai, 1915), plate 23.

females), though they often performed the same tasks as tongshi. The Atayal
woman Awai and the Sediq woman Iwan Roba0, for example, took the mar-
riage route to becoming interpreters, as had Kōan. The most celebrated Atayal
female interpreter, Yayutz Beriya, married a Japanese pharmacist named Nakano.
Beyond taking the marriage route, Yayutz subjected herself to years of rigorous
formal training in the Japanese school system to achieve prominence as a language
teacher for Japanese policemen and as a famous interpreter.33
In addition to the tongshi and banpu who learned languages through immersion,
another type of Qing-period interpreter is exemplified by Pu Chin, a Dakekan-area
native who learned to speak and write Chinese in a Qing school for indigenous
children. In response to the booming export economy in camphor (see chapter 1),
Governor Liu Mingchuan established the Fukenju (Bureau of Pacification) in 1886
to regulate relations between indigenes and Taiwan’s Han population.34 Its charter
specified that “barbarian women (banba) should . . . greet and feast the barbarians
(banjin) who come down from the mountains.” The Japanese gloss to this pas-
sage added, “these are taken mainly from banpu (indigenous women) married to
Chinese men (Shinajin).35
To mitigate the influence of the banpu, whose loyalties were necessarily divided,
Governor Liu built a school to create a loyal corps of male interpreters. His academy
 The Longue Durée and the Short Circuit    125

opened in 1890 and accepted twenty students; it added ten more in 1891—all sons
of chiefs or men of influence, aged sixteen or seventeen sui.36 Graduate Pu Chin
was twenty-three years old and working in industry when he was hired by the
Dakekan garrison to interpret for the Japanese in August or September 1895.37 As
recounted in chapter 1, the talks orchestrated by Pu Chin were initiated by the
Dakekan garrison commander, one Captain Watanabe. On August 29, Watanabe
sent a seiban tongshi (“savage interpreter”) to arrange for Jiaobanshan-area Atayal
peoples to come down the mountain to speak with the Japanese. The next day,
two women and five men showed up on neutral ground—down from the moun-
tains, but beyond the rice fields abutting Dakekan walled city. Watanabe’s retinue
brought over forty people, including soldiers, porters, and military police. It is
likely that the younger of two Atayal women was an interpreter as well. She was
about eighteen, wore manufactured Chinese clothing, and “could pass for Japanese
but for her braided hair.” She appeared to be Atayal village’s delegate in dealings
with Chinese.38
A week later, current Taipei governor Tanaka Tsunatoku and future Taipei
governor Hashiguchi Bunzō, each with a couple of staffers, left Taipei by train to
meet the Dakekan garrison. This time, the Japanese arrived with an official trans-
lator (tsūyaku-kan), who translated Japanese into English and vice versa, and an
unnamed Portuguese “temp” (koin) who translated between Taiwanese (Taiwan
dogo) and English. In addition, they utilized two Atayal-Chinese interpreters, Pu
Chin and Washiiga, to translate between Taiwanese and the Dakekan dialect of
Atayal. At this parley, the relays were as cumbersome as those of Captain Cassel
and Isa near Sabaree back in 1874 (see chapter 1).
The female Atayal interpreter named Washiiga was about nineteen years old.
She had divorced a Chinese man at age sixteen, wore some “old Chinese cloth-
ing,” and understood the “Formosan-language somewhat.” Pu Chin, probably with
Washiiga’s assistance, provided Hashiguchi and Tanaka with the names, domiciles,
family relations, titles, and ages of the Atayal contingent in his all-important regis-
ter.39 Tanaka’s report, Hashiguchi’s report, and the Tokyo Asahi newspaper account,
all based on eyewitness testimony, provide different names for the Atayal people,
suggesting the uncertainty of the linguistic environment. Tanaka and Hashiguchi
did not even agree on the number of Atayal people present at the lineup at their
initial meeting, so opaque was the language situation.
While Governor Tanaka assiduously recorded the script of the speeches he
made to the assembled Dakekan men (see chapter 1), he noted in his manuscript
report that only the “gist” was communicated. His spoken words, after all, were
converted from Japanese to English, English to Taiwanese, and then Taiwanese
to Dakekan Atayal, so anything other than a few cardinal points would have been
impossible to communicate.
To convince those who had traveled to Dakekan from Jiaobanshan to continue
on to Taipei, the troops offered to surrender the four blankets they had lent to
126     The Anatomy of a Rebellion

the Atayal contingent, one blanket per person. Hashiguchi reported that the sol-
diers tried to explain to the Atayal people that the blankets were not the soldiers’
private property but belonged to the platoon. The Atayal, for their part, believed
that they already owned the blankets, even before the offer was made. Hashiguchi
concluded that the Atayal people could not distinguish the concepts of lend and
give, partly due to the small number of words in their language.
Rather, according to Hashiguchi, they conflated the concepts of give and lend
into an Atayal concept somewhere in between that meant “to turn over, relin-
quish.”40 While Hashiguchi admitted that he was speculating, he had hit upon an
important sticking point. The concepts of private property, ownership of goods,
or keywords like lend and borrow, as we shall see in chapter 3, went beyond lan-
guage difficulties and suggested fundamental differences in political economic
cosmologies.41
After duly filing his reports and recommendations based on this encounter (see
introduction and chapter 4 for details), Hashiguchi ordered subprefect Kawano
Shuichirō to Yilan’s Atayal territory. In September, “quite fortunately and by chance,”
they located four tsūji (interpreters) of former Qing employ and their banpu (indig-
enous women) wives. Kawano described the women as “pure indigenes with facial
tattoos.” The male interpreters themselves were described as kabanjin (transformed
barbarians), a term applied to indigenes who had assimilated politically or cul-
turally to Han folkways, though to a lesser extent than the jukuban (cooked bar-
barians). Kawano deputized the women as tsūben (go-betweens/interpreters) and
ordered them to bring the chief of Xiyanlaowa village (Xitou) to the plains for a par-
ley. The women then went inland and convinced a chief to meet Japanese officials
on neutral ground.42 Awai, one of the tsūben hired by Kawano, was thirty years old
and had been married to the jukuban Chen Xilai for eight years. Although she wore
some Chinese clothes, she retained distinctive tattoos and spoke the local Atayal
dialect.43 Kawano also recorded the text of his lengthy speech. This long-winded
admonition was presumably communicated through the government interpreter
and the tsūben to the gathered Atayal men and women.
As was the case with the Hashiguchi/Tanaka mission and feasts during the 1874
invasion, Kawano’s scribes were careful to transcribe and translate into Japanese
a song relayed to the gathered listeners in Atayalic.44 It is impossible to know if
Japanese officials were putting words into the mouths of the gathered singers.
In any event, Kawano, like Hashiguchi, was denied access to the abodes of these
emissaries; he returned to Taipei satisfied that he had soothed the feelings of the
Atayal people and allayed their suspicions of the new regime.

T H E T R OU B L E W I T H T O N G SH I

While male tongshi such as Pu Chin and female tsūben such as Awai, located in
and around the Qing Pacification Offices, were readily available to lead Japanese
 The Longue Durée and the Short Circuit    127

Army contingents to particular villages to read proclamations and distribute gifts,

it would take many hundreds of interpreters to satisfy the needs of Japanese cam-
phor industrialists, people in the tea business, and officials who were laying the
groundwork for an island-wide infrastructure. As indicated by an early report in
the magazine Taiyō, which described the frustratingly complex process of media-
tion in Taiwan’s interior, locating and securing a sufficiently large and adept corps
of translators would be a daunting task:
An interpreter is a Chinese who is good at indigenous languages and is conversant
with conditions among indigenes. Each hamlet or group of hamlets has an interpret-
er (a hamlet is similar to our buraku [village]). The “local man” (shatei) works under
the interpreter, procuring whatever is needed for each side to facilitate communica-
tion, diplomacy, and so on. Each hamlet without fail has a local man. These local
men and interpreters are not always to be found in the hamlets, though. They have
their own households with affairs to attend to, and they hire themselves out to people.
Moreover, these interpreters can apply for government service. Officials screen
them, and . . . they become responsible for relaying directives and taking care of all
manner of things related to the indigenes . . . One imagines that their income is not
small.
.  .  . The interpreters and local men assume an arrogant and insulting manner
toward the indigenes, who in turn treat the interpreters very humbly.45

Saitō Kenji, a camphor merchant and engineer, reinforced the unflattering por-
trait of interpreters in Taiyō from a commercial perspective, for an audience of
Japanese who were in Taiwan or likely to travel there. He wrote that most inter-
preters were jukuban, whom he described as “evolved savages” (shinka no seiban).
According to Saitō, Chinese interpreters were rare. As liminal figures of uncertain
affiliation, wrote Saitō, interpreters swindled and cheated Japanese merchants at
every turn, holding them hostage with threats of head-hunting reprisals should
they be dismissed by their employers. Saitō noted that interpreters could be hired
for a fixed monthly salary but that it would be better to spend a couple of months
learning “the indigenous language” oneself, instead.46
According to a Japanese infantryman named Iriye Takeshi, who recorded his
observations just after the cession in 1895, the availability of Atayal interpreters in
the north was directly related to increases in trade volume in the treaty-port era.
Iriye described the codependence of commerce and intermarriage:
In [jukuban] territory, indigenes give their daughters away and force them to marry.
This is a recent occurrence; it did not happen in former times. . . . All of these types
of marriages are related directly to [cross-border] trade . . . A meeting of the minds
is reached between man and woman, and the marriage is arranged. Before any im-
morality is committed, the chief ’s permission is sought and the marriage carried
through. When a jukuban takes a seiban concubine, the new bride is given Chinese
clothing and so on. . . . For the return trip home, they are given a water buffalo, two
jars of liquor, and black and red cloth.47
128     The Anatomy of a Rebellion

Anthropologist, interpreter, and police informant Mori Ushinosuke recorded that

bride price—in the form of red skirts (up to one hundred in some cases), cattle,
pigs, guns, daggers, farming implements, and even farmland—was customary at
Atayal weddings.48 Therefore, the above marriages chronicled by Iriye, while finan-
cially motivated, were not casual unions. In fact, they were publicly sanctioned.
Indigenous parties to such unions took care to assure that marriage alliances did not
result in too much familiarity between lowlanders and ­highlanders—­permission
to visit natal villages was often limited to outmarried daughters and not granted
to their husbands. Since chiefs enriched themselves and accrued political power
by interposing themselves between villages farther inland and delegates of for-
eign powers, thereby monopolizing outflows of mountain products and inflows of
imported prestige goods, such strictures make sense. They suggest that outmar-
rying was part of a complex political strategy to maintain power in a milieu that
required the distribution of gifts to maintain leadership status.49
To contextualize these practices, it should be noted that the selling, purchase,
and exchange of daughters to further the goals of a village, family, or economic
unit was common in rural northern Taiwan among the Han, as were arranged
marriages in Japan for the same purpose. Han Taiwanese families in the Taipei
Basin out-adopted birth daughters in exchange for adopted female daughters, to
raise the latter as future daughters-in-law. This practice prevented expensive wed-
ding costs and forestalled problems related to bringing outsiders into the family
compound.50 In Japan, mukoyōshi (son-in-law) adoption was a way to preserve
lineage property in a system of inheritance based on primogeniture for families
without male offspring. That is to say, the apparent ease with which Atayal chiefs
outmarried daughters to secure future commercial intermediaries would not have
appeared callous or outlandish to Han or Japanese people at the time.
From these and many other reports, a general pattern emerges. A minimum
of two interpreters (usually three) brokered Japanese-indigenous communication
in the early years of colonial rule. Japanese men known as “official translators”
(tsūyakukan or tsūyakusei) interpreted for government officials to the Chinese
or “evolved barbarian” interpreters (tongshi). They interpreted between the “offi-
cial Chinese” spoken in Beijing and Taiwanese Chinese (Minnan dialect), while
the intermediaries known simply as “local men” (shatei) or “aborigine women”
(banpu) interpreted from Taiwanese Chinese to an indigenous language. Despite
their low official status, the local men and banpu were indispensable for sojourns in
the indigenous territories. When Pacification Office chief Saitō Otosaku’s Japanese
official translator, Shōjiro Kunitei, was incapacitated with malaria during an early
circuit of Alishan’s environs, Saitō’s mission continued to function. Saitō would not
proceed, however, without his local man.51 An official from the interpreter’s bureau
in Miaoli underscored the point: four different languages were spoken behind the
mountains abutting Miaoli city. It was simply impossible to conduct business there
without local men.52
 The Longue Durée and the Short Circuit    129

T H E N EW C L A S S O F JA PA N E SE I N T E R P R E T E R S

As recounted in the previous chapter, Hiyama Tetsusaburō’s first mission to Puli

took on the same trappings as other “first embassies” to plains-mountain inter-
faces. In one way, however, the Hiyama mission was unique. From their base in the
newly established Puli Pacification Office, Hiyama himself, his associate Kondō
Katsusaburō, and at least two other Japanese men made alliances with Tgdaya
tribes by becoming sons-in-law. According to Irie Takeshi, Puli prefect Hiyama
Tetsusaburō’s wedding, held around January 1, 1896, was a major event. Hiyama
doled out the expected blankets and jars of liquor at his wedding to Chiwas, the
daughter of the Paalan headman Pixo Sappo.53 By April 5, 1896, Hiyama’s marriage
was publicized in the Japanese press as a human interest story.54 Hiyama was a col-
orful politician from Tokyo who was lampooned in the Tokyo press for abscond-
ing to Taiwan to make an easy salary by exploiting his connections to Mizuno
Jun.55 His tenure in Puli did not last; he was expelled from office on May 25, 1897,
for collusion in a burglary that ended in a murder.56
Hiyama’s most notorious associate, Kondō Katsusaburō, would spend the rest
of his life in Taiwan, achieving fluency in Taiwanese Chinese and at least one dia-
lect of a northern indigenous language, probably Sediq. He was considered by
some to be an expert on indigenous affairs on a par with Mori Ushinosuke him-
self.57 Contemporary references to Kondō label him as a tongshi (interpreter), sho-
kutaku (commissioner), or buppin kōkanjin (aborigine trader).58 Kondō was also
a decorated translator/explorer, who is credited for brokering complicated deals
with Bunun, Atayal, Sediq, and Truku headmen and emissaries (see below). His
long analysis of the causes of the Wushe Rebellion were picked up in various offi-
cial, journalistic, and popular postmortems, thus spreading his name, for a time,
widely (see figure 20).
Kondō Katsusaburō was born on December 10, 1873, the eldest son of Kondō
Mankichi and Chiyo (née Kawahara) of Myōzai County, Urashō Town, Tokushima
Prefecture. His family was in the indigo business. Kondō himself was a restless
youth and purportedly blew 130 yen of his family’s money in Ōsaka’s warehouse
and entertainment district instead of procuring indigo. The penniless Kondō
made his way to Kobe, built up capital as a jobber and peddler, and finally opened
a profitable store. His earnings were sufficient to repay his family. Kondō next
joined the Ōkura Group to work in the commissariat during the Sino-Japanese
War. Kondō sailed from Shimonoseki to Inchon and returned to Japan with the
Nagoya army for the triumphal return. Kondō could not raise enough money for
passage to America to learn the milling business, so he ended up in Hong Kong
via Shanghai. His interest in Taiwan, piqued by stories of headhunters in the newly
acquired land, sent him there on the next ship.59
A fragmentary memoir asserts that Kondō actually swam into Jilong Harbor
shirtless, after having been thrown overboard from an American ship for scuffling
130     The Anatomy of a Rebellion

Figure 20. Back row center, Kondō “the Barbarian” Katsusaburō, ca. 1910. Courtesy of the
Rupnow Collection.

with crew members.60 Kondō’s first known destination in Taiwan, Puli, was a
crossroads of savage-border commerce and a melting pot. Conditions there rap-
idly deteriorated after the Japanese declared themselves masters of the island.
Puli’s walled fortifications, coupled with its location on a plateau high up in the
mountains, made it an ideal rebel (“bandit”) base.61 Puli was captured by opposi-
tion forces on July 11, 1896, as the last in a series of towns and cities under the con-
trol of one Kien-I. It was retaken by the Japanese on July 22.62 The Puli Pacification
Office (bukonsho) officially opened on July 23, 1896, the day after Kien-I’s defeat,
with Hiyama Tetsusaburō as its first chief.63
Under Hiyama’s patronage, Kondō was granted a permit to “trade in indig-
enous products” in Puli on May 19, 1896.64 As a first step to becoming a celebrated
“aborigine hand” (bantsū), Kondō wed Iwan Roba0, the daughter of a Paalan vil-
lage headman named Chitsukku,65 sometime in 1896.66 By all accounts, Kondō
was a talented linguist; he quickly became a much sought-after interpreter. In
January 1897 Kondō and Iwan joined the ill-fated Fukahori Yasuichirō expedition.
Before the expedition entered the high mountains beyond Puli, however, Kondō
fell ill with malaria. Iwan accompanied the mission as an auxiliary interpreter,
working with Kondō’s replacement, a plains indigene named Pan Laolong, also a
merchant specializing in the “aborigine trade.”67 Pan and Iwan survived because
they abandoned the mission before it came to grief, rightfully fearing danger as
the troop traveled beyond areas previously secured through parleys with local
 The Longue Durée and the Short Circuit    131

chiefs. The frigid weather warned off other potential interpreters on the mission.
Fukahori was able to hire about ten Sadō (near Truku village) men to lead him,
but their assistance was useless in Xalut, where a different language was spoken.
Because Fukahori’s team was engaged in the visible project of mapmaking and had
famously scolded and beat curious Tgdaya men for inquiring about these activi-
ties, the Xalut people feared that Fukahori was an advance team for a punitive
expedition.
The wanderings of this troop were not unlike those of the Ryūkyū castaways
in Langqiao in 1871. Without a common spoken language, miscommunication
was the rule. The local peoples, rightly fearing large numbers of mobile outsiders,
finally reacted with violence and slaughtered all fourteen of Fukahori’s men. An
indigenous-district policeman who reconstructed these events in 1936 opined that
Fukahori’s mission would have succeeded had the locally embedded and multilin-
gual Kondō Katsusaburō not fallen ill.68
Ostensibly to locate Fukahori’s killers, on August 20, 1897, Kondō left his wife,
Iwan, and headed north to the Truku-Toda village complexes toward Neng’gao
Mountain. He brought two of his trading-post employees, Nagakura Yoshitsugu
and Itō Shūkichi, in addition to securing the services of Itō’s wife, a Toda native
named Tappas Kuras.69 After a grueling, rain-soaked march, Kondō secured the
patronage of a trade-minded Toda headman named Bassau Bōran. The headman
was a former customer at Kondō’s Puli establishment. As Bassau’s houseguest and
adopted son, Kondō literally “went native” in the course of his twenty-month stay,
earning the sobriquet Seiban Kondō (Kondō the Barbarian).70
In the meantime, the Second Combined Squadron and a search party from the
Puli Pacification Office conducted independent investigations in March 1897. The
latter was ineffectual, for the same reasons that LeGendre’s and Bell’s investigations
of Koalut in 1867 failed: they had no means for incentivizing cooperation. The
more heavily armed and staffed squadron, however, determined that Fukahori’s
party had died in battle on a forlorn riverbed deep in the mountains. They recov-
ered a few tattered articles of clothing and a pair of eyeglasses.71 In March 1900,
two years after Fukahori’s disappearance, Kondō himself accompanied Puli sub-
prefect Ōkuma Hirotake to recover five of the actual skulls and more artifacts in
Sado, the place Kondō had ingratiated himself with Bassau.72
While Kondō and Itō bartered their way around Toda-Truku territory with
the assistance of their new extended families, other noteworthy unions between
Japanese men and indigenous women began to sprout up, including that of Jiku Shō
Min and his Atayal wife, Bariya Nōkan. In 1895 or 1896, Japanese army interpreter
Jiku (aka Watan Karaho) married into a Wulai family near Quchi. He claimed to
be the heir apparent to his father-in-law’s chieftaincy.73 Inō Kanori’s diary entry
for May 23, 1897, referred to Jiku as the point man for directing Japanese officials
on journeys from Xindian to Quchi (Ueno Sen’ichi’s route in 1891).74 Jiku signed
several of his reports with the moniker Savage Righteous Army. His associations
132     The Anatomy of a Rebellion

Figure 21. Quchi-area residents posing with flag, ca. 1897. East Asia Image
Collection, Lafayette College, Easton, PA.

with camphor merchant and hydroelectric power entrepreneur Dogura Ryūjirō

suggest that he used his relations as a labor force and perhaps as a private army.75
As Inō Kanori’s and Mori Ushinosuke’s guide to Wulai, Jiku was also a key player
in the birth of Japan’s ethnology of Taiwan.76 As Inō was making his famous tour,
Jiku dispatched a Rimogan resident named Taimu to survey the location of dozens
of inland villages. Jiku submitted a color map, several dozen sketches of cultural
artifacts, and a taxonomy of Atayal language groups in an official report.
 The Longue Durée and the Short Circuit    133

Jiku included a description of a Raga (south of Wulai) couple with his attached
photograph—perhaps the first photo in Taiwan to accompany an ethnological sur-
vey. He described the sitters as a male named Yūmin and a female named Akashi.
Jiku referred to them as pioneers of “aborigine civilization” because they requested
flags and Japanese clothing (see figure 21).77
Kondō Katsusaburō also translated for ethnologists Inō Kanori and Torii
Ryūzō,78 while Kondō’s wife, Iwan Robao, was subject to Torii’s interviews and
anthropometric surveys.79 Therefore, the bicultural marriages described above
were indispensable for gathering practical as well as more abstract forms of
intelligence regarding indigenous peoples in the early years of colonial rule.
Despite their benefits, in 1899 Sanjiaoyong (near Jiaobanshan) district offi-
cer Satomi Yoshimasa complained to Taipei governor Murakami Yoshio that
Japanese-Atayal unions were causing friction with local men. Satomi suggested
that Japanese civilians be forced to apply for permits before marrying indigenous
women. Murakami then proposed a system of punishments for Japanese men who
abandoned their local wives. In addition, Murakami recommended state support
for divorcées, whose local marriage prospects had been ruined by public asso-
ciation with foreigners. Murakami’s questionnaire, circulated to magistrates and
police posted in aborigine territory, illuminates official concerns:
1. To what extent are there contracts or gifts of cash or merchandise on the occa-
sion of a wedding (kashu no sai)?
2. [Provide] the name and occupation of the person who married a banpu (banpu
o metorishi mono).
3. Is it possible for one who has been a secondary wife/concubine of a Japanese to
remarry an aborigine (naichijin no shō / mekake to naritaru mono wa futatabi
banjin ni kasuru)?
4. In the case of abandonment, how is the banpu managing? How do the aborigi-
nes feel about the situation?
5. Was the union with a Japanese (naichijin ni kasuru) entered into as a lifelong
commitment, or was it rushed into avariciously for the short term?80

Despite these problems, there appeared to be no other way to establish paths of

communication with Austronesian language speakers. Even if official interpreters
could learn indigenous languages, they were not stationed in places where they
might have been useful. According to the personnel rosters for the first year of the
Pacification Office’s operation (1896), there were eight translators posted among
its eleven stations.81 The following year, the translator positions were all gone,
although four of the 1896 translators remained as administrative assistants (shu-
jiho), now performing “routine duties and tasks to be determined by the station
chief, including interpreting.”82 In the late summer of 1897, one Pacification Office
head reported in the Daily Yomiuri that even when translators were employed,
communication was still difficult because they did not understand the local dia-
lects very well.83 The Pacification Office was dissolved the next year, and aborigine
134     The Anatomy of a Rebellion

administration was moved to Section 3 of the new District Administrative Offices

(Benmusho). As a rule, the Benmusho stations lacked translators altogether.
But without Japanese staffers trained in indigenous languages, wrote one
pacification officer in 1898, Japan was doomed to rely upon the shady local men
who cheated indigenes by plying them with booze. The local men also frustrated
Japanese policy by trafficking contraband rifles.84 One obvious measure to dispense
with Japan’s reliance on interpreters, without relying on the fractious intercultural
marriages, was formal language training for pacification officers and policemen.
In March 1897, as a first step, the head of the Industrial Section in Taipei ordered
Pacification Office heads to compile word lists and phrase books based on trad-
ing-post conversations, official meet and greets, and negotiations of camphor con-
tracts with indigenes. The memorandum’s author wrote that indigenous languages
had no historical affinities to Chinese or Japanese and that mastering them was
difficult. The circular proposed systematic word collection and phonetic analysis
with the cooperation of the Pacification Office. Materials started trickling in by
October 1897.85
However, as anthropologist Mio Yūko’s careful analysis of the manuscript cop-
ies of the responses reveals, the stacks of word lists and conversational phrases
were made by linguistic amateurs, written down in katakana (a Japanese sylla-
bary), and they ended up collecting dust in government offices. The ambitious
project did not produce training manuals or textbooks, leaving the Japanese to
learn indigenous languages through immersion.86
To incentivize immersion learning, the so-called kensho (stipend) system was
implemented. In this program, police officers and assistant policemen were given
monthly pay supplements if they could pass language exams in Hok-lo, Hakka,
or an Austronesian language.87 The pay incentives for Han Taiwanese assistant
patrolmen to learn Japanese succeeded. The Taiwan Government-General thus
cultivated a cohort of Japanese-speaking Han Taiwanese to handle interpreting
chores in the plains and ports. However, the stipend system was premised on the
assumption that a comprehensive and fair exam could be administered in the tar-
get language. It is clear from the foregoing, however, that Japanese officials were
not in possession of instructional materials or a language pedagogy that could
produce examinations, examiners, and a curriculum for Austronesian languages.
Therefore, an exception was made. The indigenous-language tests would con-
sist of abbreviated interviews with Japanese district officials, who determined if
a candidate understood an indigenous language sufficiently to “communicate
and understand complex matters” and “relay orders without doing damage.”
Examinees would be sorted into three categories of proficiency.88 A set of 1907
regulations stated that exams for indigenous languages be limited to oral inter-
views. That same year, bango (indigenous language) was officially defined as nine
separate tongues: Atayal, Saisiyat, Bunun, Tsou, Tsarisen, Amis, Puyuma, Paiwan,
and Yami.89
 The Longue Durée and the Short Circuit    135

The practical application of government ethnologist Inō Kanori’s taxonomy

(see figure 2) had important long-term consequences. There are many more
Austronesian languages in Taiwan than the nine listed above. As Inō’s colleague
and rival ethnologist Mori Ushinosuke observed at the time, Inō’s taxonomy
lumped dissimilar languages together, while it split similar languages into sepa-
rate categories. Mori himself advocated a sixfold classification that fused Saisiyat
with Atayal and grouped Puyuma, Tsarisen (Rukai), and Paiwan together, based
on his own research as the Meiji era’s most dogged survey ethnologist of Taiwan.
Despite Mori’s reservations, however, these nine groups became census catego-
ries, museum labels, and place-names on ethnic maps. As inscribed ethnonyms,
they were carried forward with considerable institutional momentum into the
late twentieth century (with the exception of Tsarisen, which essentially became
Rukai). Until indigenous activists themselves met with some success in revising
the taxonomy in the twenty-first century, these nine language groups constituted
the building blocks of indigenous Taiwan’s second-order geobody.90
On July 27, 1909, police headquarters circulated another memorandum urging
colonial police stationed in the northern districts to learn Taiwanese and indig-
enous languages. The Yilan district station employed three instructors in the office
and another at a garrison. At other locations, however, no qualified instructors
could be found, so the men were reduced to studying word lists on their own.
One district head noted that lack of contact with Atayal people for actual practice
hindered progress. The Xinzhu station chief claimed that most of his men could
say a few words in an indigenous language but could not conduct everyday con-
versations. The best he could do was provide nonsystematic practice for the police-
men under his command with indigenous mercenaries or the indigenous wives of
Japanese policemen.91
It is not surprising that conversation partners for uniformed Japanese and
Taiwanese officials were scarce in 1909. We saw in chapter 1 how Kondō Katsusaburō
and Iwan Robao (or possibly other informants to the Japanese government)
orchestrated a trap for Hōgō- and Paalan-village Tgdaya (Wushe) men, who lost
most of their weapons and warriors on a single day, on October 5, 1903. After their
defeat, emissaries from Wushe notified a local police office that the severity of the
trade embargo had inflicted suffering, and they offered to supply troops and por-
ters to help build a guardline. The Taiwan Government-General kept Hōgō wait-
ing. They finally sent negotiators in late 1905. Kondō Katsusaburō then acted as the
go-between to orchestrate Wushe’s surrender in 1906. In order to resume trade, the
Paalan and Hōgō headmen allowed the government-general to build a guardline
path to the summit of Shoucheng Dashan, just below its 2,420-meter peak. This
commanding height overlooked dozens of settlements of strategic interest in and
around Wushe.92 To avoid being shot on sight, the Wushe-area men were told to
bring Japanese flags to the Kasumigaseki station and check in before hunting in
the area.93 From this new section of guardline, which cut the Tgdaya villages off
136     The Anatomy of a Rebellion

from the trade hub of Puli, the government-general was able to encircle the Atayal,
Truku, and Sediq tribes north of Wushe in subsequent campaigns.94 For brokering
the surrender ceremony at the Kasumigaseki station, Kondō was awarded seventy
yen for meritorious service. This was an extraordinary bonus. At this time, the top
cash award for ranking inspectors was only sixty yen.95
Perhaps hoping to reproduce the 1906 breakthrough attributed to Kondō
Katsusaburō, who was married to Iwan Robao of Paalan, Police Bureau head Ōtsu
Rinpei urged the government to actively encourage, instead of merely tolerate,
mixed marriages between low-ranking Japanese officers and indigenous women
from chiefly lineages.

T H E P O L I T IC A L M A R R IAG E S U N D E R S A K UM A
S A M ATA’ S A DM I N I S T R AT IO N

Following the Russo-Japanese War (1904–05), Governor-General Sakuma Samata’s

administration (1906–15) accelerated the policy of military conquest over the
Atayal villages that had refused to surrender their rifles and sovereignty during
the previous decade of colonial rule. Although Wushe had formally submitted in
1906, many of the tribes farther upland and inland still remained defiant. Sakuma
pledged to break the resistance of the northern tribes once and for all by extend-
ing the heavily armed guardline to the heart of Atayal country to encompass Xalut
(which subsumed Masitoban), Truku, and Toda (see map 1).96
Sakuma’s right-hand man on indigenous policy, Police Bureau chief Ōtsu Rinpei,
inspected the forward posts of the guardline in spring 1907. His report proposed to
remedy the alarming interpreter situation with a “political-marriage” policy:
I have sensed an extreme paucity of indigenous-language translators at every instal-
lation on my tour. This being the case, it is not difficult to imagine that many, many
more translators will be necessary as we bring the indigenes within the fold of gov-
ernment. I need not mention that success in managing the indigenes hinges upon the
ability of our translators . . . . The quickest route to cultivating translators would be to
give occasional financial assistance to the appropriate men and have them o ­ fficially
marry banpu .
[We] recognize that such marriages can cause trouble because indigenes . . . will
say that such marriages are unfair and benefit only the villages or lineages of the ban-
pu’s new husband. By the same token, if the Japanese husband is rotated to another
post and abandons his wife, it will offend indigenous sentiment. Nonetheless, our
present lack of translators is troubling. Because marriage to an indigenous woman
will raise living expenses, occasional moneys should be provided for the low-ranking
men who cannot afford it.97

In August 1908 Ōtsu commissioned Kondō to secure Wushe allies in order to add
another spur to the system of guardlines already in place. Kondō himself initially
demurred but then reconsidered when Ōtsu offered him a land grant of thirty
hectares near the Kasumigaseki police station outside Puli. In October Kondō
 The Longue Durée and the Short Circuit    137

Figure 22. Left, in Japanese clothing, Kondō Gisaburō, with defeated Truku peoples, January
1915. The photographer, Mori Ushinosuke, attested to Kondō’s linguistic prowess in Truku and
credited him with helping to defeat the Truku pictured here. Note the oilcan in the foreground:
a bamboo indigenous drinking vessel is resting on top. Mori Ushinosuke, ed., Taiwan banzoku
zufu, vol. 1 (Taipei: Rinji Taiwan kyūkan chōsakai, 1915), plate 31.

found allies in the settlements of Hōgō (headman Aui Nukan) and Mehebu
(headman Ludao Bai), two Tgdaya groups who considered Truku and Toda ene-
mies.98 According to Kondō, Ludao Bai and Aui Nukan demanded that he and his
younger brother Gisaburō (see figure 22) take wives in Wushe to seal the bargain.
Accordingly, Kondō paid the price of one pig to Paalan’s chief to ratify his “divorce”
from Iwan Robao to pave the way for his second marriage.99
In January 1909 the brothers Kondō were married to Obin Nukan and Chiwas
Ludao. In late February, soon after the formalization of the alliance by marriage,
Kondō Katsusaburō led some 654 Wushe warriors in the general attack on Toda
and Truku, which was successfully concluded by March 1909.100 The official report
states that Wushe drew the Taiwan Government-General into a local feud by blam-
ing Toda for recent attacks on the Japanese guardline. The report’s authors were
skeptical of the charges against Toda but went along because they thought it expe-
dient to use Wushe’s manpower to gain a strategic foothold in difficult terrain. In
138     The Anatomy of a Rebellion

addition to Kondō’s 654 Wushe (Tgdaya) allies, who apparently acted as laborers,
the government fielded 580 fighting men for the climactic attack on Toda.101 These
maneuvers secured more commanding heights for Japanese cannon and mortars
as well as strategic passes for communication, which allowed the government-
general to effectively hem in the Sediq territory from all sides.102
As an eldest son and a future koshu (household head), Kondō Katsusaburō
expected to secure his Japanese family’s future by obtaining a large, government-
registered farm near Puli for the care of his aged father, per Ōtsu Rinpei’s promise
in August 1908. Indeed, Kondō’s father, who had immigrated to Taiwan by this
time, passed away in March 1909, leaving Kondō legally responsible for the care of
his mother, younger brothers and sisters, and a number of nieces. Kondō spent the
years 1909 through 1916 pressing claims for the promised acreage while continuing
to accept government commissions as a guide and interpreter.
As the years wore on, his patron Ōtsu Rinpei returned to Japan; other officials
familiar with Kondō’s case left Nantou Prefecture. Finally, Wushe district police
inspector Kondō Shōsaburō (no relation) deeded the land to his own relative, pre-
empting Katsusaburō’s claim. This perceived slight angered Katsusaburō’s younger
brother Gisaburō to the boiling point, according to Katsusaburō’s testimony. In
1916 Gisaburō tendered his resignation in Taipei but was instead transferred out of
Wushe to a new post in Hualian District at Pushige.103
Sixteen-year-old Kondō Gisaburō had come to Taiwan in 1901. Gisaburō
worked his way up through the ranks to become a sergeant in the indigenous-
territory police by the time of his infamous transfer, thanks to his facility with the
Sediq dialects spoken near Wushe. Igarashi Ishimatsu, a Japanese official of long
experience in Wushe, recalled Gisaburō as a companionate husband to Chiwas
Ludao and a figure of local respect. Igarashi claimed that Chiwas’s devotion to
Gisaburō allowed the Japanese to squelch an uprising by Xalut and Salamao in 1913.
Overhearing the plan from her brother Mona Ludao, Chiwas alerted Gisaburō,
who in turn sounded the alarm and averted disaster for the Japanese.104 Such
timely intelligence is all that Ōtsu Rinpei could have hoped for when he called for
political marriages back in 1907. However, this success story would be short-lived.
After Gisaburō’s 1916 transfer to Pushige, he never returned, leaving Chiwas
bereft. She eventually remarried a local man but was never compensated with
death benefits, nor was she given a sinecure, as was done with other abandoned
wives.105 Kondō Katsusaburō recalled that Chiwas had not been entered into
Gisaburō’s koseki (household registry), making it impossible for the elder Kondō
to file a claim on Chiwas’s behalf.106

T H E TA I SHŌ - P E R IO D P O L I T IC A L M A R R IAG E S

From 1909 to 1914, Sakuma Samata’s “five-year plan to control the aborigines”
(gokanen keikaku riban jigyō) brought most of Atayal country under direct
 The Longue Durée and the Short Circuit    139

colonial rule. When the temporary organ established to administer Sakuma’s five-
year plan (the Bureau of Aboriginal Affairs) was disbanded in July 1915, the Taiwan
Government-General declared victory in the north.107 Thereafter, the indigenous
territory was governed by Japanese police officers who “possessed the power of
life and death” over their colonial subjects. The indigenous-district police were
charged with peacekeeping, educational, and medical functions. They became the
eyes, ears, hands, and feet of the Taiwan Government-General in the lands under
“special administration.” Unlike their predecessors, who lived near trading posts
or in border towns, Japan’s new “aborigine hands” would reside among the tribes
and villages that they governed. As salaried officials (in contrast to tongshi), they
would take direct orders from their superiors and transmit these orders to the
“submitted tribes.”
Two particularly well-known Japanese-indigene marriages, both initiated to
advance the guardline in Atayal country, survived into an era that saw increased
surveillance and conscripted labor. One of these marriages ended in 1925 with the
sacking of Shimoyama Jihei after a drunken altercation with his superior officer,
the other in 1930 with the killing of Sazuka Aisuke in the Wushe uprising.
Shizuoka native Shimoyama Jihei arrived in Taiwan in 1907 as a young army
veteran, two years before Sazuka’s arrival in 1909. During Sakuma’s five-year cam-
paign, the Malepa dispatch station was overrun and its staff wiped out. To reestab-
lish control, the Nantou District head of indigenous affairs sent Shimoyama to run
things, after ordering him to marry Pixo Doleh, the daughter of Malepa’s headman.
According to Pixo’s niece, Malepa assented to the marriage out of economic des-
peration to obtain needed patronage and rifles.108 According to their son Hajime,
Pixo became proficient at Japanese conversation. Shimoyama, on the other hand,
could not speak Atayal. Pixo bore him two sons and two daughters. Shimoyama’s
other betrothed, Shizuoka native Katsumata Nakako, arrived in Puli after he and
Pixo had started their Malepa family. Nakako also began to bear Shimoyama’s chil-
dren. For a time, Shimoyama shuttled between Puli and Malepa. No longer able
to afford two households, he finally moved his Japanese family to Malepa. Pixo
Doleh, the daughter and sister of local headmen in a fiercely monogamous society,
was humiliated.109
Shimoyama’s open display of contempt for Atayal sensibilities came to the
attention of Neng’gao district commander (gunshu) Akinaga Nagakichi in 1925.
One oral history claims that Akinaga was tipped off by a disgruntled Atayal youth.
In any event, Akinaga broached the delicate issue of bigamy in front of assem-
bled Malepa men. Shimoyama retorted that Pixo’s relatives had consented to this
arrangement. Besides, an angered Shimoyama replied, he had been assured that
he could break off the political marriage to Pixo after three years’ time, even if
there were children. For his part, Akinaga took umbrage at being contradicted
publicly by a subaltern. Tempers flared. Shimoyama challenged Akinaga by grab-
bing his lapels. Akinaga then threw Shimoyama down on a table. Shimoyama
140     The Anatomy of a Rebellion

was dismissed for insubordination the next day, a fact that demonstrated clearly
that the government-general believed it could handle affairs in Malepa without
him. Although his abrupt departure left Pixo bitter, Shimoyama secured a job
for her at a Wushe infirmary at a salary of forty yen per month before he left.
Nonetheless, Shimoyama’s son Hajime recalled that his mother was never entered
into Shimoyama’s family register (koseki) as a legally recognized wife. Nonetheless,
due to Pixo’s high status in local society and official concerns about the ramifica-
tions of mistreating a headman’s daughter, Shimoyama was not able to treat her as
a disposable concubine.110
Nagano Prefecture native Sazuka Aisuke (1886–1930), unable to abide impover-
ished country life in Japan’s “snow country,” fled to Taiwan via Tokyo one January
morning in 1909 with money stolen from his father.111 Five years later, as combined
police, army, and navy forces bombarded resistant villages near Hualian Harbor in
the last spasms of Sakuma’s “campaign to end all campaigns,” Sazuka married Yawai
Taimu, the daughter of an Atayal headman in Masitoban, just north of Wushe.
Sazuka’s wife, Yawai, was remembered as one who spoke proper Japanese and
worked harder than other Atayal women to assimilate to Japanese customs.112 As
one postmortem account of the Wushe Rebellion put it, Yawai married outside her
local society only to be mistrusted by her fellow Atayal. She was also looked down
upon by Japanese colonists for being a “savage.”113 In a word, Sazuka and Yawai
were representative outcasts of empire. One left the bleak prospects of Nagano
behind to seek fortune in a colonial periphery, while the other was cast out by her
father to shore up his position as a local headman.
For his willingness to persevere in aborigine country, Sazuka was finally made
chief of the Wushe branch station on March 31, 1930.114 Wushe was an adminis-
trative center and showpiece for Japanese policies vis-à-vis the northern Taiwan
Indigenous Peoples, a real step up from isolated Masitoban. There were eleven
settlements under his new watch, and Wushe boasted a post office, schools, hot-
springs inns, trading posts, and dormitories for students and government workers.
Although it was a promotion, the reassignment to Wushe turned out to be Sazuka’s
death sentence. Because of their roots in Masitoban, Yawai and Sazuka were viewed
suspiciously in Wushe. Residents in Wushe’s villages were Sediq Tgdaya, who
spoke a different language than Masitoban Atayals. Shimoyama Hajime recalled
that Sediq visitors to Atayalic Malepa spoke Japanese with their hosts, since Sediq
and Atayal were mutually unintelligible.115 In the Xalut settlements of Malepa and
Masitoban, Sazuka relied upon his wife’s Masitoban networks for local knowl-
edge and, to some extent, for his legitimacy as a leader for some two decades.
His linguistic facility and political experience in these former posts, however, pro-
vided Sazuka with little useful intelligence in the new setting. On the contrary,
Sazuka’s affiliation with a non-Sediq Atayal—his wife, Yawai Taimu—and her
father inclined Tgdaya residents within his jurisdiction to accuse Sazuka of favor-
itism toward Xalut peoples. They expected Sazuka and Yawai to exact vengeance
 The Longue Durée and the Short Circuit    141

against Tgdaya peoples in relation to recent intra-indigenous (though state-aided)

violence.116 And although Ōtsu Rinpei thought low-ranking rural policemen such
as Sazuka could maintain bicultural households on their modestly supplemented
salaries, Sazuka was an infamous embezzler, who was known to withhold pay and
salt it away for his personal use. After his death in the rebellion, it was discovered
that he had amassed ¥2,000 in a savings account and ¥20,000 in cash.117
Shimoyama Jihei’s ignominious firing and Sazuka Aisuke’s violent demise can
be considered the final blows to a system of rule predicated on a gender division
of labor for cross-cultural communication that dated back to early Qing times. In
this system, locally embedded women, as the daughters of headmen or local indig-
enous notables, were either adopted or married into outsider families to act as buf-
fers between politically separated but economically linked social formations. The
men in this equation—be they Han merchants, settlers, or officials, or Japanese
policemen or trading-post operators—capitalized on their wives’ bilingualism and
local kin networks to make a go of it on the so-called savage border. In hind-
sight, it is clear that this ad hoc arrangement was perhaps adequate for maintain-
ing a certain volume of cross-border trade or for gathering military intelligence
for short-term expeditions. But intercultural marriages fell far short of provid-
ing a mechanism to consolidate a stable system of outsider rule among the Sediq,
Atayal, and Truku peoples of northern Taiwan.
According to statistics compiled by Kitamura Kae, twenty-nine Japanese police-
men stationed in the special administrative district were proficient in an Atayalic
language circa 1930, to rule over hundreds of settlements—many of them still hos-
tile. According to Shimoyama Hajime, the son of Pixo and Jihei, the language bar-
rier forced Japanese policemen to use blows instead of admonishments. Hajime
extrapolated from his father’s beatings to hypothesize that generalized physical
abuse was a long-term cause of the Wushe uprising. In an interview years after the
fact, Hajime even wondered if better language skills among the Japanese police
could have averted the slaughter that took place on October 27, 1930.118

T E AC H I N G JA PA N E SE T O I N D IG E N O U S P E O P L E S

As Atayalic language competence among Japanese policemen declined in the

1920s, the Japanese language skill of indigenes was on the rise. The formal instruc-
tion of indigenous peoples in Japanese language was consonant with the general-
ized imperial impulse to diffuse national culture outward. It also aimed to cultivate
a corps of interpreters whose loyalty was directed toward the state rather than their
Taiwanese in-laws or private trading houses. While this project did not bear fruit
in time to prevent the Wushe uprising, it represented the centerpiece of Japan’s
assimilation policies toward indigenes.
In 1896, soon after establishing itself in Taipei, the Taiwan Government-General
established fourteen Japanese-language training institutes (kokugo denshūjo)
142     The Anatomy of a Rebellion

throughout the island. Older students took crash courses in language and basic
arithmetic to clerk and translate for the government. Younger students, as objects
of the regime’s “attraction policy,” studied “written Chinese, geography, history,
singing and gymnastics.”119 The Hengchun walled city institute in southern Taiwan
served a mixed population of “acculturated indigenes” (jukuban) and Han stu-
dents. Nearby, an annex school (bunkyōba) was built in Tuilasok; it was the first
government school for indigenes.
The Paiwan settlement of Tuilasok was the home of Toketok, the man with
whom LeGendre had struck agreements in 1867 and 1869. Toketok died in 1873,
thus leaving his nephew Tsului to represent Tuilasok in meetings with Saigō
Tsugumichi and other foreigners. By Tsului’s side sat Jagarushi Guri Bunkiet—the
son of a Tuilasok woman and a Guangdong immigrant surnamed Lin. Bunkiet
was born in Checheng or Poliac in 1854 and was purchased by the sonless Toketok,
who groomed the bilingual Bunkiet to be a headman.120 Bunkiet came to the notice
of Qing officials when Hengchun walled city was built in 1875 as a base for Shen
Baozhen’s “open the mountains, pacify the indigenes” campaigns.
For his success as a labor recruiter, the Qing awarded Bunkiet the surname Pan.
He thereafter worked, presumably as an interpreter, in the 1890s campaigns to pacify
the Hengchun tribes, earning a Qing meritorious rank of the fifth order (wu-pin).121
In November 1895, Bunkiet, bearing a flag his family had received from
Saigō twenty-one years earlier, attended Hengchun dispatch-station chief
Sagara Nagatsuna’s first official audience.122 That same month, a Japanese garri-
son stationed in Sabaree mistakenly killed several men and burned down even
more dwellings. According to one version of the story, the garrison heard shots
fired from the direction of a Koalut village. In another version, they mistook
an armed contingent of Paiwan men for hostiles. In any event, skittish Japanese
troops responded by tying eight villagers to trees. While Koaluts were fleeing, the
Japanese proceeded to torch twelve houses. They also shot and killed six residents
in the confusion. News of the depredations had spread to neighboring settlements
when Bunkiet stepped in. With Bunkiet’s intercession, the Japanese government
assuaged local anger by disbursing twenty yen each to the families of the dead
and five yen each to the victims of the massacre. For this service, Bunkiet was
appointed commissioner attached to the Hengchun dispatch station.123
As the government-general’s new point man for the lower Hengchun, Bunkiet,
in concert with other local leaders, brokered the formal submission of forty-four
tribes in Taidong that same February.124 Bunkiet and another bicultural mediator,
a Puyuma woman known to us as Tata Rara, most famously assisted the govern-
ment-general on May 18, 1896. Instead of merely brokering an agreement, Tata
mobilized Paiwan and Puyuma troops to militarily defeat a remnant Qing officer
named Liu Deshao (Liu Tek-chok) and his followers.
The Qing formally surrendered Taiwan on April 17, 1895, but (because of numer-
ous rebellions) it took until April 1, 1896, for Japan to officially declare an end to
 The Longue Durée and the Short Circuit    143

Figure 23. The Puyuma “Joan of Arc” Tata Rara with Japanese interpreter
Nakamura Yūsuke, 1896. Adolf Fischer, Streifzüge durch Formosa (Berlin: B.
Behr Verlag, 1900), 323.

the state of war and institute a civil government. Liu Deshao remained in southern
Taiwan, operating as a local potentate beyond the reach of the government-general
past the cessation of military rule on the island.125 In accord with Shen Baozhen’s
“open the mountains, pacify the indigenes” program, the Qing had begun to sta-
tion officials and soldiers in Taidong (near Beinan) in 1875. Around two thousand
men were posted throughout the Hengchun Peninsula. They were not especially
brutal, wrote the Japanese commentator Hara Segai, but they did make inspec-
tions, conscript labor, and take millet and rice without paying. The Qing govern-
ment accordingly became loathsome to the people, inciting head-hunting attacks
on plainsmen and provoking armed standoffs. According to Hara’s report, when
the Japanese took Taiwan in 1895, the Qing abandoned these forces without alert-
ing them of the cession. The remnants resorted to looting and stealing to feed
themselves, and the area descended into chaos.
During his first diplomatic mission to Taidong in February 1896, Sagara
Nagatsuna consulted with Zhang Yichun, a prominent merchant and husband of the
multilingual interpreter Tata Rara. Zhang’s wife, Tata, then lobbied Sagara’s transla-
tor, Nakamura Yūsuke, to intervene against Liu’s remnant troops (see figure 23).126
Tata Rara was the daughter of Chen Ansheng, a chief who dealt with LeGendre
during his perambulations in the 1870s. One source even referred to her as “Chen
Tata.”127 Tata Rara spoke Southern Min Chinese, as well as the Puyuma, Paiwanese,
144     The Anatomy of a Rebellion

Figure 24. Tata Rara with her Puyuma militia, 1896. Adolf Fischer, Streifzüge durch Formosa
(Berlin: B. Behr Verlag, 1900), 324.

and Amis languages. Like Bunkiet, she earned a salary of six yuan per month as
a Qing subofficial,128 presumably for services rendered during the kaishan fufan
campaigns.
In May 1896, Tata and Bunkiet organized a militia of three hundred indigenes
to expel the Qing remnants. Tata Rara herself personally led the men through a
hail of musket fire to secure the surrender of a thousand Qing troops. She appears
to have been a formidable commander. One German publication referred to her
as the “Puyuma Joan of Arc,” while a Japanese report referred to her as a “heroine”
(see figure 24).
Bunkiet and Tata, as liaisons and military leaders, earned cash rewards and
plaudits for the defeat of Liu Deshao. Bunkiet received a commendation called
Order of the Sacred Treasure, Sixth Rank, on December 15, 1897 (see figure 25).129
Tata Rara belatedly received her reward of forty yen on February 9, 1900.130
After the suppression of Liu Deshao, Bunkiet and Sagara Nagatsuna applied to
have the colony’s first Japanese-language institute built for indigenous students.
On September 2, 1896, a plan to build this modest school was approved. Bunkiet
then mobilized his followers to supply land, building materials, and labor in time
for a September 10 opening ceremony. One encomium to Bunkiet reports that
he registered his own nephew to lead by example and that over thirty youths in
Tuilasok followed suit.131 A less celebratory account by a Japanese teacher reported
Figure 25. Tuilasok interpreter, headman, and Qing/Japanese official Pan Bunkiet, ca. 1900.
The medal on his suit is a sixth-grade Order of the Sacred Treasure, conferred in 1897 by the
Japanese government. Jun’eki Taiwan Genjūmin kenkyūkai, ed., Inō Kanori shozō Taiwan
Genjūmin shashinshū (Taipei: Jun’eki Taiwan Genjūmin hakubutsukan, 1999), 177. Photograph
courtesy of the publisher: The Shung Ye Museum of Formosan Aborigines, Taipei.
146     The Anatomy of a Rebellion

that the first group of twenty-seven students did not know why they had matricu-
lated, except that the headman Bunkiet had sent them.
Three neighboring Hengchun Paiwan settlements, Koro, Tekiri, and Kauwan,
responded to Sagara’s announcement of the “urgent matter” of Japanese-language
education with much less enthusiasm. They temporized by convening councils to
deliberate the matter. Tekiri and Kauwan did not apply for schools, and by 1900,
only two institutes had been built in all of Hengchun. Koro village finally applied
on the condition that the school be combined with a gun and ammunition depot.
As Kitamura points out, the language institutes were not gifts from the govern-
ment but were built with locally provided materials and labor. “Buy-in” could not
be assumed. Several Paiwan schools were shut down temporarily or closed because
of armed attacks from neighboring settlements.132 Reminiscent of the Meiji state’s
early efforts to spread compulsory schooling throughout Japan, central ambitions
outstripped budgets, parents could not always spare the foregone labor of the chil-
dren in the fields, and resentment by locals who paid for these unfunded mandates
led to hundreds of cases of vandalism.133
Despite these hiccups, the lower Hengchun villages were far ahead of indig-
enous settlements in the north in terms of formal education.134 We recall here that
Ōkubo Toshimichi’s parting shots to Qing officials in Beijing at the conclusion
of his negotiations in 1874 included criticism of the lack of schools in the Paiwan
settlements in Langqiao. While the Qing did not exactly dot the landscape with
schoolhouses, it did make a beginning. After Japanese forces left in 1874, Hengchun
city drew population by decree but also through attraction, as the Qing filled in an
administrative grid with baojia (mutual responsibility) units and zhuang (admin-
istrative villages). Educational institutions to transform indigenous peoples into
Sinophone Qing subjects were also piloted. This flurry of activity directed at hori-
zontal integration tipped the balance of power in favor of the Han, vis-à-vis the
formerly dominant Paiwan, during the twenty-year interval between the Taiwan
Expedition and the assumption of Japanese rule.135
When Inō Kanori arrived in Hengchun at the behest of the TGG education
bureau, the Tuilasok annex school was already in operation. In late October 1898,
Inō stopped in the walled city, visited the local Pacification Office, and finally
reached Tuilasok. During his tour, Bunkiet invited Inō into a ceramic-roofed
home, well appointed with Chinese furniture. Bunkiet put on an elaborate feast for
his Japanese guests and showed Inō a treasured copy of the admonition that Saigō
Tsugumichi had issued in 1874, warning the Paiwan to be obedient to the Qing.136
In his synthetic report on conditions throughout Taiwan’s interior, Inō wrote
that the lower Hengchun towns such as Tuilasok were the most “advanced” of all
indigenous peoples, due to frequent intermarriage with Chinese in their prox-
imity. Inō singled out Tuilasok and Sabaree especially as places where tiled-roof
homes and the use of Chinese language were common—these villages were, in
fact, hardly distinguishable from Han settlements, according to Inō. The “lower
 The Longue Durée and the Short Circuit    147

Hengchun tribes,” south of the line stretching from Mudan to Fenggang (see
map 5), were therefore geographically advantaged compared to the more isolated
and less “advanced” upper Hengchun villages, according to Inō’s evolutionary
taxonomy.137 After the Qing began to fund reclamation and Han migration, the
southernmost Paiwan tribes, such as Sabaree, Tuilasok, and Koalut, were forced to
adopt the Qing hairstyle, the queue. In the northern part of the peninsula, how-
ever, around Mudan and Kuskus, Paiwan men retained their Paiwan hairstyles.138
In 1902 the Japanese Hengchun district officer applied to abolish the status of
Mudan, Kuskus, Sabaree, and Tuilasok as “savage villages” (Japanese sha, Chinese
she) and promote them to “townships” (Japanese shō, Chinese zhuang). The appli-
cation, like Inō’s survey ethnography, noted in-migration of Han and Paiwanese
acculturation and intermarriage with Han as factors in favor of township status.
Another reason the lower Hengchun tribes qualified for “normal administra-
tion” was their cultural competence regarding private property and the conduct
of commerce. Most notably, unlike indigenous peoples more generally, the lower
Hengchun Sinicized indigenes understood how to calculate the worth of trade-
goods in abstract monetary units—rather than in terms of customary equiva-
lences in bartered goods (see chapter 3). In a similar vein, the application noted
that lower Hengchun Paiwanese adjudicated disputes through mediation. They
did not follow formal court proceedings, but at least disputes were not solved by
feuds, vendettas, or divination.
As a surplus-producing, stockbreeding, and agricultural area, lower Hengchun,
it was argued, should contribute to the cost of government by paying taxes—which
meant extending all rights and obligations of colonial subjects (hontōjin) to Koalut,
Sabaree, Tuilasok, and selected other villages. Thus, on April 28, 1904, the for-
mer abodes of Isa, Toketok, and the wreckers of the Rover were, administratively
speaking, no longer “savage” in the eyes of the Japanese government.139 Bunkiet
himself is said to have led a group of Paiwan residents out of the mixed-residence
Tuilasok to resettle in Mudan Bay in 1901.140
The evidence offered above suggests that Japan’s “savage border,” reckoned as
a cultural boundary that separated economic moderns from indigenous peoples,
reflected the extent of the Qing government’s reach by 1895: wherever Paiwanese
had been forced to wear the queue, normal administration followed. In fact, the
lower Hengchun settlements of Mudan and Kuskus (the villages whose men
reportedly did not wear the Qing queue), located just north of Sabaree, Tuilasok,
and Koalut, remained classified as sha (savage settlements) well into the 1930s.
As map 4 and figure 51 indicate, Japan’s ethnic classification excluded Tuilasok,
Koalut, and Sabaree from the Paiwan zone by 1912, while most of lower Hengchun
remained firmly within it. The colorful maps attached to 1930s statistical com-
pendia followed suit. On the Paiwan side, as late as 1934, Mudan and Kuskus were
still sending children to the “aborigine schools,”141 despite the region’s status as
a relatively wealthy and economically developed location. Thanks to its suitable
148     The Anatomy of a Rebellion

riverbanks, Mudan became an early site for the cultivation of paddy rice in 1909.
Within a couple of decades, Mudan was a model rice-production center,142 though
it still remained within the indigenous zone.
This seemingly arbitrary line—which divided fellow Paiwanese in a surplus-
producing and fertile area of Taiwan into normally and specially administered
peoples—had one important consequence. Like the qualifications for indigenous-
district policemen (see chapter 1), which were lower than those for policemen
stationed in Han districts, the indigenous school curriculum was adapted to spe-
cial conditions—it was watered down. From the beginning, indigenous students
were not expected to perform as well as their Han counterparts in Hengchun city
or the other common schools and language institutes. Even under auspicious
­conditions—with an eager and powerful collaborator in Bunkiet and a surplus-
producing Paiwan population versed in the Southern Min dialect of Chinese—
Sagara Nagatsuna and the higher-ups in Taipei designed the Tuilasok annex school
to keep Hengchun residents of Paiwan ancestry “down on the farm.” Although
it has been celebrated as the first indigenous education facility in the history of
Japanese colonial rule, the Tuilasok annex was, in fact, a precursor to the third-tier
educational system that would define Japanese rule in the highlands.143
By March 1898 the Taiwan Government-General had built sixteen Japanese-
language training institutes and an additional thirty-six annex schools; four of the
latter were for indigenous children (all in Taidong or Hengchun, in the south).
That same year, the institutes were discontinued and replaced by a more ambitious
common-school (kōgakkō) system. Its six-year curriculum “consisted of ethics,
Japanese language, classical Chinese (composition, reading, calligraphy), arithme-
tic, music and gymnastics.”144 However, in two locations, Hengchun and Taidong,
the vocationally oriented institutes were left in place until 1905. As one official
report put it, the “reason is not far to seek. The existence of so many aborigines in
these districts accounts for this exception.”145 In 1904, thirteen language institutes
and annexes served indigenous children, with an enrollment of 803 (759 males and
44 females).146
Even after the remnant language institutes for indigenous children were discon-
tinued in 1905, indigenous schools remained separate and inferior, but under the
new name aborigine common schools (banjin kōgakkō).147 The government-gener-
al’s own English-language propaganda was forthright: “According to the new [1905]
arrangement, the school course covers only four years instead of six . . . while the
number of subjects was made less than those of the [common schools]. The aborig-
ine pupils have in the regular course only three subjects: Morals, Japanese, and
Arithmetic. They have no need of Chinese classics, Science, and Commerce. Only
Agriculture, Manual Training, and Singing (one or all three) may be added to the
regular course of study according to the intellectual development of the tribe.”148
Because the residents of Tuilasok, Sabaree, and Koalut bore the burden of
taxation after their incorporation into the regularly administered territory, their
 The Longue Durée and the Short Circuit    149

children were theoretically eligible to attend six-year common schools with a full
curriculum.149 Aborigine common schools were built among Amis and Paiwan
peoples who lived under normal administration and remained concentrated in
the “advanced” south and the eastern rift valley into the 1930s. By 1935, there were
exceptions: five aborigine common schools had been built in the specially admin-
istered territories. Of these five, only one was located in the north among the
Atayal, Sediq, and Truku populations.150
Eika Tai has noted that the common schools for Han children were aimed at
creating “citizens” (kokumin), whereas that term was absent in planning docu-
ments and curriculum for the indigenous common schools. This nomenclature
reflected the government-general’s pre-1930s approach to indigenous education: it
would create not fellow nationals but rather docile and useful subjects.151

T H E E X PA N SIO N O F I N D IG E N OU S E D U C AT IO N T O
T H E N O RT H

Due to the violence surrounding the camphor wars, which surely fed into Japanese
officialdom’s low estimation of Atayal cultural capacity, the first “northern tribes”
school was established in Quchi in 1908. The Quchi school was called an aborigine
youth school (bandō kyōikujo), a new institution that complemented the aborigine
common schools. The youth schools were creatures of the guardline movements
discussed in chapter 1. In contrast to schools for Han, making “citizens” was not
part of their writ.152
At the Quchi school and others like it, indigenous children were taught com-
portment (bowing, sitting respectfully, interacting with superiors), lifestyle
“improvement” (wearing geta [clog-sandals], using chopsticks, wearing pants),
ethics (loyalty), farming (intensive, fixed plot), and handicraft production (see
figure 26). The youth schools were staffed by policemen and district-office staffers.
The guidelines were drawn up in 1908, in the midst of the war against northern
indigenes. They were built in part to win over the indigenous peoples and also to
train interpreters.153 The pioneering Quchi school was moved upriver to Wulai in
1911. The bulk of instruction was in field cultivation, tree felling, and other “reg-
ular employments. Japanese-language training was the second most important
pursuit.”154
In 1915, the same year that Japan’s major military offensives against indigenous
peoples were terminated with the retirement of Sakuma Samata, the Taiwan
Government-General issued first-year and second-year readers for indigenous
children titled Banjin dokuhon (A reader for indigenes).155 A user’s manual for
Banjin dokuhon (published in 1916) forecasted an ethnically bifurcated admin-
istration and the birth of indigeneity in modern Taiwan. Taiwan’s educational
administrators in effect hardened the line that LeGendre drew between indigenes
and Han for political purposes in 1872 (see chapter 1), that census-bureau desk
150     The Anatomy of a Rebellion

Figure 26. The Wulai School for Indigenous Children, which opened ca. 1910. The standing
adults wear the trademark embroidered upper garments with decorative buttons, as immortalized
by Mori Ushinosuke’s Quchi portraits. The male children sport Japanese-style haircuts and
imported clothing, while the girls wear Atayal leggings with Chinese-style upper garments.
THE SCHOOL OF URAI SAVAGE TRIBE, lw0410, East Asia Image Collection, Lafayette
College, Easton, PA, accessed July 26, 2017, http://digital.lafayette.edu/collections/eastasia/lewis-
postcards/lw0410.

jockeys inscribed between “Mongolians” and “Malays” in 1905, and that guardline
architects scorched into the earth in the 1910s. The manual explained:
In April 1914 . . . aborigine common school regulations were promulgated, begin-
ning a new chapter in indigenous education. [Initially . . . ], the aborigine common
schools used the Ministry of Education elementary-school reader designed for
naichijin (Japanese), or the Taiwan Government-General common-school reader for
hontōjin (Han Taiwanese). . . . However, it goes without saying that these two readers
were inappropriate for aborigines, who are of a different race, who differ in language,
whose environment and climate is not the same, and whose lifestyles and cultures
are varied. Therefore, the government-general ordered the formation of an editorial
board for a new set of Aborigine Readers. . . .156

While this manifesto suggests that indigenes had varied cultures and lifestyles,
it was premised on the notion that they were all subunits of a Malay race that
was also inscribed in the 1905 census categories. Education bureaucrats thus based
the different levels of education, for Han and indigenous Taiwanese, on perduring
attributes such as environment, race, and culture. The phrasing of the guidelines
 The Longue Durée and the Short Circuit    151

amplifies Matsuoka Tadasu’s view that 1915 was the turning point when Japanese
colonists began to accept the specially administered territory as a permanent,
rather than expedient, division.
In 1902, by contrast, the district officer of Hengchun applied to move Tuilasok,
Sabaree, and Koalut under normal administration ostensibly because these com-
munities had been Sinicized. However, Sinicized referred to those who spoke
Minnan Chinese (a Fujianese tongue) and wore a Central Asian hairstyle (the
Qing queue). Therefore, the term Han in this case did not have an ethnic referent
but served as a proxy for the ability to produce surplus wealth and participate in a
commodified economy. We can read this circa 1904 usage of Han as equivalent to
liangmin (good subject), an old Qing term that separated most imperial subjects
from pariahs, bannermen, nobles, and outlanders.
But under the 1916 educational guidelines for indigenes, reflecting wider cur-
rents in the post–World War I Wilsonian dispensation, indigenous status was
being moved out of the plastic category of “uncivilized” and into the static cat-
egory of “possessing unique attributes” as a racial/cultural marker, while the term
Han moved in the same direction as it also became particularized as an ethnonym.
As Sakano Tōru has argued, the first decades of Japanese colonial rule saw a con-
flation of “civilization” with “Sinicization” in the discourse on “aborigine improve-
ment.” However, by the 1930s, “imperialization/Japanification” had become firmly
entrenched as the telos of progress in official discourse.157
The pointed suppression of Chinese-character (Japanese kanji, Chinese hanzi)
use in the 1915 readers furthered the ethnic bifurcation of Taiwan into Han and
indigenous moieties. While common schools for Han students158 and late Qing
indigene schools in Taiwan focused on Chinese reading and writing,159 Japanese
authorities in post-Sakuma Taiwan regarded Chinese characters as optional for
indigenes, almost a luxury. The first two years of indigenous instruction, accord-
ing to reformed rules, were conducted in katakana (a syllabary). Teachers stressed
oral communication and diction; hiragana (another syllabary) was added in the
third year, with some very elementary Chinese characters. In the fourth year, indi-
genes learned a few dozen kanji related to their customs, surroundings, and voca-
tions but not nearly enough to read a Japanese newspaper or a Chinese poem.160
Of course, Japanese itself, like Chinese, was hardly a monolith, and what con-
stituted Japanese in 1915 required spelling out. Because many indigenous-district
policemen hailed from northeastern Japan and Kyūshū, the 1916 teacher’s guide,
echoing those for “national-language” teachers in the home islands, mandated
the use of Tokyo-standard pronunciations. It inveighed against the rural Japanese
policemen’s “ragged Japanese” and “vulgar colloquialisms” and “mispronun-
ciations,” and it urged proper diction.161 Shimoyama Jihei’s son recalled that the
Kyūshūans and Okinawans who staffed the aborigine police passed their rough
Japanese on to their students; female Atayal women commonly used locutions like
ore and o mae (very casual forms of male speech).162
152     The Anatomy of a Rebellion

Again promoting ethnic bifurcation, Japanese educators also pushed the notion
that, as members of the Malay race, the indigenes were ethnic affines of the Japanese.
Teachers could use the fourth-year reader’s account of the emperor Jimmu’s east-
ward conquests, it was suggested, to incline Taiwan’s “Malays” (indigenes) to dis-
identify with the Han race and imagine themselves as long-lost cousins of the
Yamato (Japanese) ethnos.163
So while the emphasis of indigenous education was practical and focused on
everyday conversation, it also held within it the seeds of more complex forms of
indoctrination. The fourth-grade reader taught students that a typical ban-sha
(aborigine settlement) contained about 30 families and 150 residents. Indigenous
residents generally farmed their plots but occasionally hunted in the mountains.
Within each ban-sha, two buildings were most prominent: the schoolhouse and
the police station. According to the reader, before the police station arrived, farm-
ing was difficult because indigenous peoples were in a perpetual state of war. Now
they could farm peacefully. And, if they studied hard enough in school, they might
be sent to the capital to study Japanese. One sample lesson reproduced postcards
from a fictitious older brother in Taipei. The postcards reproduced pictures of the
governor-general’s residence and a large, ornate school building, as well. In addi-
tion to the large buildings of the capital, the reader also illustrated the countless
thousands of Japanese troops that paraded in Tokyo and stood ready to put down
rebellious imperial subjects.164 In short, the reader reproduced in miniature the
cardinal lessons of the indigenous sightseeing tours to Taipei and Tokyo, wherein
headmen were made to understand imperial might and wealth and, hopefully,
were dissuaded from further rebellious activity.
On the one hand, after 1928, some institutes adopted more challenging curricula.
Thereafter, making kokumin (citizens) was a stated goal of indigenous schooling.
Chinese characters were now part of the curriculum, as well.165 However, propos-
als to send professional teachers to the indigenous territory, either to sharpen
and reinforce the education outlined above or to improve the lives of indigenes,
were continually thwarted. Indigenous schools were staffed by policemen for the
duration of colonial rule. Throughout the 1920s, enrollments in aborigine youth
schools actually increased. Nonetheless, allocations for instructional materials and
physical plant were routinely diverted to fund shortfalls elsewhere in the colo-
nial budget. In the end, writes Kitamura Kae, most Japanese-language schools for
indigenous peoples, with the exception of the demonstration school at Jiaobanshan
(see figure 27), became thinly disguised work camps.
Crops were raised by students to pay for their upkeep and the costs of infrastruc-
ture. Language lessons were dispensed in rickety buildings with chronic shortages
of desks, chairs, and school supplies.166 While common schools for Han students
faced financial constraints because they were paid for with local taxes levied on col-
onized Taiwanese, the indigenous schools faced even tougher constraints because
they were funded by a central government that was perpetually cash-starved.
 The Longue Durée and the Short Circuit    153

Figure 27. Jiaobanshan model school for indigenous children, ca. 1930. [The Kappanzan
School and Classroom], ip1532. East Asia Image Collection, Lafayette College, Easton, PA,
­accessed July 26, 2017, http://digital.lafayette.edu/collections/eastasia/imperial-postcards/ip1532.

With the arrival of Taiwan’s first civilian governor-general, Den Kenjirō, in 1921,
a few highly qualified and ambitious Han and indigenous students were permit-
ted to enter elementary schools with Japanese children. In 1921, eight indigenous
students throughout Taiwan met the tough standards. The denigrating word “sav-
age” was dropped from the name of the aborigine common schools in 1922, but
the course of study remained restricted to four years of practical education. The
exceptions who proved the rule were children of Japanese policemen and indig-
enous women who entered the elementary school in Wushe alongside Japanese
schoolchildren. Most indigene school attendees—themselves a slight percentage
of the total population—went to the Wushe aborigine common school. This was
the only common school for indigenous children for all of northern and central
Taiwan; all other Truku, Sediq, and Atayal children, if they attended at all, were
taught at Japanese police stations.
Two Wushe common-school graduates, Hōgō natives Hanaoka Ichirō (Dakis
Nobing) and Hanaoka Jirō (Dakis Nawi), entered the Puli elementary school as
teenagers in 1921. Puli was a walled city and gateway to the mountain settlements
in central Taiwan. The two subsequently took jobs as a school instructor and a
police assistant, respectively, near Wushe, and became famous for the suicide note
they left on the morning of the October 27, 1930, Wushe Rebellion.167 While they
were not directly implicated in the attacks, they blamed “excessive forced labor”
154     The Anatomy of a Rebellion

for the rebellion and reported that they had “been detained against [their] will by
the barbarians.” The passionately scrawled note criticized both sides, reflecting the
painfully torn identity of men who had ascended the meritocratic ladder but also
identified with the people they had left behind.
The Hanaokas (not related) had married local Hōgō women at a Japanese-style
wedding ceremony near a government office in the 1920s. Their wives, Kawano
Hanako (Obing Nawi) and Takayama Hatsuko (Obing Tado) were the daugh-
ter and niece of a leading Sediq political figure, the Hōgō paramount chieftain
(sōtōmoku) Tadao Nōkan. Hanako and Hatsuko were frequently photographed in
Japanese clothing and attended the Puli elementary school, like their husbands.
According to Deng Xiangyang, these marriages were arranged by the government
to provide examples of successful assimilation by indigenous peoples.168
The rumors of the two Hanaokas’ complicity in the October 27, 1930, slaugh-
ter of Japanese civilians and police in the model indigenous administrative hub
of Wushe suggested that too much education for indigenous peoples would only
produce malcontents. Matsuda Yoshirō has written that after the 1930 uprising,
indigenous peoples were no longer permitted to enroll in advanced educational
institutes.169 Kitamura Kae’s statistics show that in the 1920s, fewer than ten stu-
dents from the indigenous territory per year were selected to attend the elemen-
tary schools; by 1930 there were only twenty-eight total indigenous graduates of
elementary, middle, normal, nursing, medical, and women’s schools combined.170
But the more important trend was the content, context, and purpose of indigenous
education in the post-Wushe years, which shaded into the full-blown mobilization
period after 1936. If the Taishō period can be characterized by the vocational bent
of indigenous education, the Shōwa years featured an emphasis on spiritual trans-
formation and imperial subjectification.
Inspector Yokoo Hirosuke took a leading role in framing post-Wushe policies
for the indigenous areas. Yokoo obliquely referred to the “Hanaoka” syndrome
with a reference to French colonial education in Africa. Yokoo believed that the
French were moving too quickly on their assimilation policies. Therefore, the
African educated elite returned to home villages resenting their own countries.
Some suffered the anomie that bedeviled proletarian intellectuals in other colo-
nies. Yokoo was not against education per se. He insisted that its quality and extent
be improved and that Japanese-language training be more thorough. Yokoo stated
his belief that indigenes should internalize and themselves propagate loyalty to the
emperor and a sense of imperial mission. Somewhat paradoxically, Yokoo empha-
sized respect for the indigenous character (although he called for the reform, if not
eradication, of many indigenous practices). While he was an advocate for imperial
subjectification, Yokoo also believed that indigenous education should not exceed
the cultural capacities of its targets.171
TGG policies indeed reflected the twin priorities of imperialization and respect
for indigenous particularity in the 1930s. The period saw decreasing expenditures
 The Longue Durée and the Short Circuit    155

per indigenous student for education172 but also saw renewed efforts to propagate
Japanese language through other forms of schooling, including ubiquitous night
schools attached to police boxes. One official estimate in 1937 put the overall “can
understand basic Japanese for everyday tasks” rate for indigenes at 34 percent for
males and 24 percent for females. Among Atayal peoples (Atayal, Truku, Sediq),
these numbers were 40 percent and 32 percent, respectively. By 1942, roughly
50 percent of all indigenous peoples were proficient in conversational Japanese,
with 58 percent of Atayal men reaching that level of proficiency.173
As indigenous proficiency in Japanese increased during the 1920s through the
1940s, the Taiwan Government-General hollowed out the incentive structure for
Japanese policemen to learn indigenous languages. Before 1922 all policemen who
passed a language test received stipends, but after 1931 that number decreased to
90 percent for a minority of “fluent” policemen and to only 20–30 percent of the
majority “proficient” Austronesian language speakers. Diminishing incentives
were reflected in low proficiency rates: in the 1930s, only 2.5 percent of indigenous
district policemen were fluent in a local language, while 26 percent were merely
proficient. That meant that about one in twelve dispatch stations had an officer
fluent in a local language. But since the percentage of lower-ranking Japanese
patrolmen in the indigenous area also decreased, while the number of indigenous
policemen increased to about 50 percent of the police force,174 the number of bilin-
gual indigenous policemen actually increased; they just weren’t Japanese.
During the period that I have characterized as falling under “native author-
ity,” circa 1910 to 1930, cross-cultural communication was anchored by female
indigenous relatives of influential headmen who married Japanese policemen. The
problems encountered by Kondō Gisaburō, Sazuka Aisuke, and Shimoyama Jihei
revealed that system’s limitations and contradictions. The reliance on chiefs not
only circumscribed the influence and knowledge of policemen attached to them,
but as we saw in chapter 1, the older generation of headmen themselves were unre-
liable vectors of public power. Therefore, after Wushe, they were replaced by new
leadership: the indigenous youth corps (Takasagozoku seinendan).175
These new youth leaders were called “pioneers” (senkusha). They replaced the
tōmoku and seiryokusha (headmen and men of influence) as the intermediar-
ies between Japanese officialdom and the Taiwan Indigenous Peoples. With the
advent of a new indigenous leadership recruited from aborigine youth schools,
a less locally embedded intermediary cohort emerged. This new leadership was
groomed to lead a society of farmers, stockbreeders, and fishermen who were
destined, if Japanese colonial rulers had any say in the matter, to live their lives
out in the indigenous territory. There was no question of integrating the highland
economy with the rest of Taiwan. After the Wushe Rebellion, it was determined at
the highest levels of government in Taiwan that Taiwan Indigenous Peoples could
only thrive under the paternalistic care of a functionally specialized aborigine
administration.176
156     The Anatomy of a Rebellion

As Japanese became the language of governance within the indigenous ter-

ritory, Japanese policemen were rotated more freely over the surface of the
territory. But their new ethos and sense of professional identity tied them to
the second-order geobody of indigenous Taiwan. The new role of indigenous
women in the post-Wushe era is perhaps best symbolized by the famous Atayal
heroine Sayon. In September 1938, Sayon and ten other Atayal women were help-
ing a local Japanese policeman named Takita cross a river to get to his military
post and join the war in China. Weighed down with three of his suitcases, Sayon
herself fell in and drowned, becoming a martyr and the subject of song, film, a
memorial bell, and more.177 The seventeen-year-old Sayon was neither an inter-
preter cohabiting with Takita nor a leader in war on the order of Iwan Robao or
Tata Rara.
Instead, she was carrying luggage to help a mobile, uniformed, male agent
of Japanese empire exit her village peacefully. While many treatments of Sayon
emphasize the hoopla generated by her supposed spirit of sacrifice, she was essen-
tially a porter. In fact, as the indigenous territories became sites of ethnic tourism
in the 1930s, the luggage-carrying indigenous woman became a common sight.178
To be sure, the “normal administration” that indigenes were being excluded
from meant heavy taxes, floggings by policemen without a trial (until 1921),
and discrimination in public schools, administrative jobs, and the police force.
Moreover, public education in the normally administered territories was far from
universal or equal to the quality of education for children of Japanese colonists.
Only 2.19 percent of school-age Taiwanese children attended schools in 1900. The
rate reached 4.66 percent in 1905 and 5.67 percent in 1910.179 By 1930, however, the
rate was 30 percent, and by 1940, 60 percent.180
Even after Gotō’s administration put an end to large-scale armed Han resis-
tance in 1902, rebellions continued through 1915. In 1907 rebels led by Cai Qingrin
killed twenty-four TGG policemen and frontier sentries in the “Beipu incident.”
The next day, insurgents killed all Japanese residents of Beipu town, for a total of
fifty-seven fatalities in two days. The government-general killed about ninety reb-
els in its suppression sweeps. Over one hundred suspected rebels were arrested;
nine were sentenced to death, and ninety-seven were either confined or put on
corvée labor. Other planned rebellions were discovered before they came to
fruition in 1912 and 1914. The hundreds of conspirators apprehended, however,
suggests that Han-dominated lowland Taiwan could not be classified as a disci-
plinary society before World War I, insofar as brute force was required to main-
tain order.181
The Ta-pa-ni or Xilai’an incident of 1915 marked the last spasm of Han armed
resistance.182 Former colonial policeman Yu Qingfang (1879–1915) and a­ ssociates
led this millenarian uprising. According to Paul Katz, author of a definitive social
history,183 ”the number of Taiwanese and Japanese killed during the fighting
 The Longue Durée and the Short Circuit    157

[was] estimated to have exceeded 1,000 people. A further 1,957 Taiwanese were
arrested . . . of whom 1,482 were put on trial and 915 sentenced to death. A total of
135 Taiwanese were executed before the Taishō Emperor issued a decree of clem-
ency, while scores more died in prison.”184
Han armed resistance never again reached the scale and intensity of the Beipu
and Ta-pa-ni revolts. According to Caroline Hui-yu Ts’ai, by “1920, when the colo-
nial administration of Taiwan delegated part of its power to local governments . . .
structural integration by and large had been completed.” In other words, over a
two-decade period of experimentation, the Taiwan Government-General had
effectively co-opted the baojia system for taxation and vital statistics while mobi-
lizing it to make even Taiwan’s rural population visible and amenable to policing
and labor mobilization. Despite the coextension of territorially defined Japanese
policing units (gun) with local self-responsibility units (baojia) throughout Taiwan,
Ts’ai concludes that structural “integration . . . did not necessarily result in social
integration. It took the two decades of the interwar period  .  .  . for the colonial
administration to appropriate Taiwanese society for its use.”185
Nonetheless, in comparison to the indigenous territories, lowland Taiwan can
be said to have become a disciplinary society in many respects. After the Ta-pa-ni
rebellion was put down, normal administration began to diverge more sharply
from special administration. By 1915, the year the reader for indigenes was pub-
lished, a middle school had been established for Han Taiwanese. A few vocational
schools were added after 1919 and a university (with only rare seats for Taiwanese)
in 1928. Despite its restrictive and discriminatory nature, the Japanese education
system for Han Taiwanese was robust enough to foment “major changes in the
structure of Taiwanese elites. In the 1910s, the landlords and gentry . . . had begun
to give way to the rising groups of professionals trained in colonial institutions.”186
Over the next two decades, the court system, especially for commercial law,
became more equitable; corporal punishment was abolished; and a great number
of rights guaranteed by the Meiji constitution for citizens on the main four islands
were extended to Taiwanese under regular administration.187
Although social mobility through education for Taiwanese was restricted to
medicine and education, it was structurally homologous to Japan’s home-island
system. Most students would obtain basic, functional literacy and morals educa-
tion, and a small percentage of the population would be trained to lead the rest of
society. Under colonial rule, these leaders were supposed to be models for collabo-
ration and assimilation, but many led movements that challenged TGG preroga-
tives. The formation of a Taiwanese political class capable of mobilizing a petition
movement for home rule and equal treatment in Japan had developed by 1914.
This large-scale civil-society activism attests to the power of precolonial educa-
tional traditions and the strength of island-wide networks among Han Taiwanese.
But the nonviolent political resistance movements of the 1920s and 1930s, as
158     The Anatomy of a Rebellion

Ming-cheng M. Lo has demonstrated, evolved in the colonial school system. In

contrast, residents in the specially administered districts did not mount petition
movements, establish opposition newspapers, or lobby the government in Tokyo.
Instead, they attacked, burned, and looted exposed police boxes or Japanese settle-
ments on occasion, most infamously on October 27, 1930.
 PART T WO

Indigenous Modernity
 3

Tangled Up in Red
Textiles, Trading Posts, and Ethnic Bifurcation in Taiwan

For a major exhibition titled Rainbow and Dragonfly, held from September 2014
through March 2015 at the National Taiwan Museum in Taipei, a large billboard
greeted visitors and passersby with four distinct examples of Taiwanese indige-
nous textile manufacture, of varied pattern and age, all dominated by the color red.
Inside, displays of Atayal cloth occupied fully one-half of the palatial museum’s
first floor. It was a complex installation: one-hundred-year-old, disembodied fab-
rics were juxtaposed with updated designs on sleek black mannequins; colonial-
era ethnological surveys shared space with large color photos of contemporary
weavers consulting them; and numerous imperial-era postcards illustrated the
diversity, antiquity, and daily uses of the displayed fabrics.
The Taiwan Government-General established the National Taiwan Museum in
1908, at the height of Japan’s camphor wars against the Atayal, Sediq, and Truku
peoples who produced much of the cloth in this exhibition. The museum’s current
structure was completed in 1915, the year Sakuma Samata left office and declared
victory in the war against the northern tribes. As the scorched earth, free-fire
zones, and land-dispossession policies of that era displayed the negative effects
of sovereign power, the coterminous planning, construction, and stocking of the
museum exemplified the positivities generated under TGG auspices in the colo-
nial period.
The Rainbow and Dragonfly exhibition, housed in a neoclassical, neocolonial
edifice, was in good measure a dusting off, repackaging, and repurposing of the
artifacts collected among indigenes during the Japanese occupation. Therefore,
it revealed how colonial ethnography, museum practices, and photography were
(and remain) active partners in the sustenance and revival of Atayal culture, which

161
162     Indigenous Modernity

was also packaged in this exhibition as an element of Taiwan’s national heritage.

The silhouette outline maps of Taiwan’s geobody that decorated the exhibition
mark out Atayal and Paiwan homelands as the familiar second-order geobodies
found on Inō’s 1900 map (see figure 2). Rainbow and Dragonfly illustrated the dif-
ficulty of imagining modern Taiwan without its indigenous complement or indig-
enous Taiwan without its modern complement.
This chapter recounts a 150-year history of indigenous-outsider transactions
centered on red-dyed cloth. It explains how the oldest artifacts in the Rainbow and
Dragonfly exhibition conjoined global trade circuits, cross-cultural technological
adaptations, colonial collecting, and Atayal social reproduction. In a word, this
thread of colonial and postcolonial history illustrates how indigeneity and moder-
nity coproduced each other.

G L O BA L E N TA N G L E M E N T A N D L O C A L R E SP O N SE S

U.S. consul to Taiwan James W. Davidson dated the “beginning of the commer-
cial career of [Taiwan]  .  .  . from 1858, when the two Hong Kong firms, Jardine
Matheson & Co. and Dent & Co. first engaged the Formosan trade.” Between that
year and 1865, British, Russian, French, and American diplomats signed treaties
with the Qing to open Danshui, Tainan, Jilong, and Gaoxiong to foreign commerce.1
The next decade saw a lucrative tea, camphor, and sugar export boom. Taiwanese
merchants and laborers organized and directed most of the trade, while a few scat-
tered Western missionaries, exporters, and a lone British consul constituted the
meager foreign presence.2 However, the frequent disappearance of mariners on and
off the coast (see chapter 1) prompted a few foreign agents to leave the security of
ports to ransom survivors or locate their remains. Others traversed the island out
of curiosity, to find commercial opportunities, or to win souls for Christ.
The guarantees of safety to foreigners spelled out in the Treaty of Tianjin of 1858,
however, were moot in interior destinations beyond the limits of Qing adminis-
trative control. For such journeys, local trading-post operators and subofficials
known as tongshi (interpreters) outfitted visitors with provisions, advice, and
guides. Among the most noted items of purchase were gifts for indigenous hosts.
U.S. consul to Xiamen Charles LeGendre’s 1869 meeting with the Tuilasok chief
Toketok (nudged to fruition with a baggage train of gifts) and Captain Douglas
Cassel’s 1874 meeting with Sabaree chief Isa (also concluded with lavish displays
of generosity) are two prominent examples of the importance of gifts for border-
land diplomacy (see chapter 1). So too with the aforementioned treks by Ueno
Sen’ichi, Hashiguchi Bunzō, Kawano Shuichirō, and other Japanese emissaries to
the “savage territory” in the 1890s. Gifts performed multiple functions in these set-
tings. They were initially a means of paying for services rendered or expectations
of future assistance, to be sure. But they were much more.
 Tangled Up in Red    163

Presentations of gifts were also occasions for recording the emotional states of
recipients and gauging the dispositions of little-known peoples as either “greedy,”
“honest,” or “uncorrupted.” Gifts also fostered trade dependency among indigenes.
The regulation of gift giving was also implemented to reform mores, punish mis-
deeds, and incentivize compliance. Finally, gifts were recycled and repurposed.
Manufactured containers were brought back to villages for reuse, while red-dyed
textiles were disassembled and reassembled into traditional cloth and clothing,
to be reexported to anthropologists, curators, and tourists as cultural items. As
commodities in the Japanese curio, art, and souvenir markets, or as objects of
aesthetic or scholarly contemplation, the reexported cloth and clothing became
touchstones for metropolitan discourses on progress, primitivism, and cultural
relativism. As such, these objects took their place next to the Gwanghwamun Gate
in Seoul, Korean celadon wares, or bamboo and wickerwork from Taiwan.3
In a famous consular report that recounted his 1869 encounter with Toketok,
Charles LeGendre wrote:
I gave the chief one hundred and eighty yards of red camlet, a small pistol, a single-
barrel shot gun (unserviceable), and a spear . . . an ivory spy-glass and case . . . some
beads, and a quantity of rings, bracelets, and a case of gin . . . Toketok had not ex-
pected this attention, and he was evidently much touched by it. “If you have brought all
this to buy me,” said he, “you have taken a useless care, for you had my word; but if you
hand me these presents as a token of friendship, I receive them with pleasure.”4

Toketok’s reported speech, to the effect that the 180 yards of red camlet and other
gifts were mere “tokens of friendship” and that LeGendre had “taken a useless
care,” is significant for reprising assertions that indigenes were disinterested in
material gain or profit. Toketok’s soliloquy, recorded after the fact and at the end of
a translation relay from Paiwan to Minnan Chinese to English, does not jibe with
less florid descriptions of similar transactions. Other travelers to Taiwan recorded
Toketok’s businesslike collection of cash fees for the upkeep of stranded sailors.5
LeGendre himself remarked in unpublished writings that Toketok’s leadership
of the eighteen tribes was financially draining and that the leader was known to
drive hard bargains for ransoms of foreigners.6 Douglas Fix has argued that when
LeGendre arrived in Langqiao in 1867, Toketok’s influence was at an apogee. By
1869, his hold on power had become precarious, in part for lack of funds.7 But if
LeGendre’s imputation of native simplicity to a master negotiator rings false, it was
not exceptional.
Edward House, a journalist who accompanied the Japanese invasion in 1874
and championed all of LeGendre’s political positions, made similar observations.
House had occasion to witness coastal Han Taiwanese bargain with troops for
land, provisions, and wages as the army built its first camp, while House also
reported on Japanese conferences with the Tuilasok and Sabaree leaders to nego-
tiate alliances, passage, and land for an east coast garrison. At a June 8 meeting,
164     Indigenous Modernity

House wrote of Paiwan emissaries that “hints of the presents that were awaiting
them at head-quarters did not affect their resolution, and it seemed impossible to
move them, when suddenly Isa, stirred by what impulse I cannot imagine, unless
it may have been the recollection of having made a promise at the time of his last
visit, announced that he would go.”8
For House, as for LeGendre, indigenes were indifferent to gifts but anxious
to honor commitments. When Isa and a group of his confederate Paiwan head-
men ceded a plot of land on the east coast for Japanese occupation, House wrote
that offers “of payment were made, but the chiefs declined compensation, with
the carelessness to gain which I have spoken of as characteristic of them.”9 House
drove home his larger point with a comparison: “The savages [House’s term for the
Paiwanese] have nothing whatever of the Chinaman in their exterior aspect, and
their ways of life are totally separate. The divergence of their disposition is most
strikingly shown in the contrast between the insatiate greed of the West Coast
[Han] and the indifference to gain of the mountaineers.”10
LeGendre himself had several occasions to express vitriolic attitudes toward
assorted Chinese officials and interpreters for their bad faith and greed, presag-
ing Japanese scorn for Han Taiwanese who purportedly cheated indigenes in the
so-called “border trade.” In detailed reports of individual encounters, however, we
find that the stark line House and others drew between the “Chinaman” and the
“mountaineer” did not hold, calling into question their judgments that Han and
indigenous attitudes toward profit or material gain were antithetical. For example,
Toketok—the quintessential noble savage in the passage quoted above—himself
slept in a Chinese bed, while one of his brothers could read and write Chinese.
Isa, his successor as the dominant political figure in lower Hengchun, understood
spoken Chinese. The “half-caste” allies and hirelings whom LeGendre and House
distanced from China for having lived beyond the reach of Qing territory were
referred to as “Chinese” by Douglas Cassel in unpublished correspondence—a
contradiction that the American consul adjusted with a heavy editorial hand.11
In the long run, as this chapter shall argue, the triad of reinforcing stereo-
types—impartial Japanese, avaricious Chinese, and innocent indigenes—became
a self-fulfilling prophecy, as Japanese policies grounded in this discourse would
eventually isolate Atayal, Sediq, Truku, Bunun, and Paiwanese from the market
economy.12 To be sure, the economies of upland rural Taiwan were not nearly as
monetized or commoditized as the Han-dominated lowlands when Japanese colo-
nists arrived in 1895. Therefore, a different sensibility and approach to commerce,
gain, and valuation plausibly obtained in and beyond the savage border of Taiwan.
At the same time, since this stereotype was resistant to empirical disconfirmation
and durable in the face of historical change, it also functioned as a racial attribute,
fixing certain populations with particular immutable characteristics in the dis-
courses and legal apparatus of the colonizing power.
 Tangled Up in Red    165

Like other travelers during the treaty-port period, LeGendre had an ax to grind
with the local Qing officials, who seemed to thwart his every initiative, and Hakka
intermediaries who profited from their monopoly on access to interior settlements.
The everyday forms of resistance LeGendre faced are unsurprising, since his abil-
ity to travel to the interior of Taiwan was a direct result of the success of British
arms in mainland China. Traders and diplomats in post-1842 Chinese treaty ports,
LeGendre prominent among them, discovered that the stroke of a pen in Tianjin
or Beijing did not magically transform Taiwan into a welcoming site for foreign
residence or an arena of untrammeled access.
There is probably more than a kernel of truth to the perception that indigenes
were more open to negotiation with foreigners than Han Taiwanese, since they
sought allies in struggles against settlers or Qing officials over resources and terri-
tory. That is to say, the “greedy Chinese, innocent indigene” trope may have fairly
reflected the experiences of individual diarists. There is also the possibility that
Toketok’s attitude toward gifts was context specific, dynamic, and grounded in a
cosmology more intricate and meaningful than the sort of crass considerations
that suggest themselves in LeGendre’s and House’s accounts. Anthropologist
Kamimura Tōru has suggested as much by reconstructing a network of Parijarijao
(“Paiwanized” Puyuma) chiefs and big men that was connected by patterned
exchanges resembling a kula ring.
In Kamimura’s view, the Langqiao Peninsula’s indigenous Skaro lineages, with
Tuilasok’s Toketok first among them as the big stride chief (ōmata tōmoku), ritu-
ally subordinated common Paiwan village chiefs (such as the Mudan headman,
for example) through gifting relations. These transactions resembled tributary
relations in some regards—as they featured hierarchical yet reciprocal ritual
exchange—while they partook in some aspects of a kula ring, insofar as hier-
archies were ratified only when equilibrium was restored at the termination of
a cycle of gifting and countergifting. In Kamimura’s reconstruction, Toketok,
as the big stride chief, and Isa and the secondary great chief (futamata tōmoku)
traveled widely to seasonal festivals to present finished goods such cloth, ritual
daggers, alcohol, and millet cakes “downward.” In return, subordinate chiefs,
themselves members of aristocratic Paiwan lineages, countergifted “upward” with
domesticated and hunted meat and presentations of large quantities of alcohol
at banquets. When LeGendre and his men showed up with a large cargo of fin-
ished goods, which he lavished upon Toketok in exchange for alcoholic bever-
ages and roast meat, Toketok situated LeGendre, quite publicly, in the position
of one to whom fealty is rendered in return for protection, thereby inverting the
ōmata tōmoku’s place in the gifting cycle—he became a receiver of finished goods
instead of a supplier. If such were the case, then Toketok’s hesitation to receive
gifts can be interpreted as appropriate behavior for an ōmata tōmoku in a ritual
gifting context. Kamimura regards the instances of Toketok’s more avaricious
166     Indigenous Modernity

bargaining as “enclave” transactions conducted outside the sphere of ritual rela-

tions and networks.13
Paiwan and Puyuma chiefly genealogies, foundation myths, and the specifics
of prestations at festivals were not recorded until after the Japanese arrived as a
colonizing power in 1895. Therefore, this more nuanced view of indigenous dispo-
sitions to gifts was not available to LeGendre, House, Mizuno, Kabayama, and the
other men who became Toketok’s and Isa’s unwitting amanuenses. In the absence of
countervailing explanatory frameworks, their treaty-port era jottings were elevated
from commercial intelligence reports to sociological verities. For one, the mus-
ings of LeGendre and House, a U.S. consul and a journalist for the New York Post,
respectively, were widely disseminated and archived for posterity in libraries and
repositories. In addition, and more importantly, hundreds of LeGendre’s reports,
missives, and memoranda were translated into Japanese and read by Japanese offi-
cials. Ueno Sen’ichi, an assiduous consumer and translator of treaty-port docu-
ments, elaborated upon LeGendre’s discourse in his seminal official writings in the
1890s, while Mizuno Jun himself echoed similar sentiments. Mizuno’s early pro-
nouncements as Taiwan’s highest ranking civil official in 1895 cemented the “greedy
Chinese, innocent indigene” binary in TGG circulars and archived reports.
As a permutation of the above-mentioned treaty-port era discourse, the trope
of the “innocent” indigene took on new valences in colonial Taiwan. The Qing
dynasty and its corrupt mandarinate were no longer relevant. The new foils were
the untrustworthy bicultural border denizens discussed in chapter 2, the “inter-
preters” (tongshi). This substitution of tongshi for mandarins mirrored the con-
figuration of northern Taiwan’s camphor forests as another “closed country” to be
opened as an arena for high-velocity capitalism. Partly because they contained for-
ests that sustained populations of animals, sources of clean water, fish, vegetables,
and other resources, indigenes defended these lands with alacrity.
Nonetheless, the contest here was not between proto-environmental subsis-
tence producers and rapacious capitalist invaders. Rather, the savage territory was
an active zone of ongoing economic exploitation, like Chinese ports before the
arrival of Western gunboat diplomacy. As Takeshi Hamashita has written of this
maritime economy, the new entrants—this time the Japanese—did not displace
an isolated traditional economy with profit-driven, translocal commercial activ-
ity. Instead, they overlaid a new system of rules and procedures and connected an
existing system to wider currents of global trade. In this milieu, aptly described by
Antonio Tavares as a late-imperial exchange economy, indigenes, Han Taiwanese,
and operators who straddled both groups exploited these forests for commodities,
not just birds, deer, firewood, and mushrooms.
As the Taiwan Government-General turned its attention to camphor as a fiscal
boon for its ailing balance sheet, Japanese attributions of greed and faithlessness
to tongshi (interpreters) reflected frustration with “toll states” that earned money
by taxing and monopolizing access to camphor producers. For actuarially minded
 Tangled Up in Red    167

administrators in Taipei and Tokyo, this intermediary layer of brokers presented

an obstacle to the efficient extraction of high volumes of product.14 The trope of
Chinese greed and indigenous honesty also justified colonial rule. As Japanese
displaced interpreters, private militias, and Qing armies as agents of dispossession,
the insistence upon indigenous innocence and victimhood validated Japanese pol-
icies to annex land, engage in forced relocation, and restrict Han immigration.
Because indigenes did not understand the concept of private property, ran the
argument, they could not make legal claims to title deeds for land. Consequently,
their legalistic transformation into economic wards of the state allowed the gov-
ernment-general to sell off “excess land” to logging companies, in a repeat of the
gambit used in Hokkaidō vis-à-vis Ainu peoples in 1899.15
Like any other colonial project, however, Japanese rule contained its own inter-
nal tensions, while it underwent historical vicissitudes. A competing discourse
generated by these same cross-border exchanges identified Atayal red-striped
capes as an element of a distinctive cultural ensemble. From the viewpoint of colo-
nial ethnology, any cloth, implement, word list, or social convention that attached
to “Atayalness” possessed a measure of intrinsic worth. As Scott Simon has noted,
the ethnonym Atayal is a Japanese-era creation. Today, this imposed category has
become an object of heated political contestation in the postcolonial period.16 On
the other hand, its origins were far from nefarious. The Japanese man who put the
word Atayal into play recorded it as an autonym, a term provided to him in col-
laboration with agentive Atayal peoples.
In fact, the term Atayal had its origins in a dissenting discourse. Yamaji
Katsuhiko’s study of the centrality of Atayal textiles and Paiwan woodcarving to
Japanese-indigenous relations over the past century is important in this regard.
Echoing the findings of Hu Chia-yu, Yamaji notes that early Japanese descriptions
of Atayal cloth juxtaposed them with other elements of ethnic identity—face tat-
toos, housing architecture, origin myths, and so on—to construct Atayal people
as an ethnos. Hu’s translation from Inō’s famous 1900 survey is instructive: “The
less abstract the costume design becomes, the higher its intellectual level tends to
be. The more concrete [a] costume decoration becomes, the higher its intellectual
level. . . . Thus, the weaving patterns of the Atayal costumes consist of straight lines
and angles and elements arranged without clear order . . . , which reflect a lower
intellectual level. The embroidery designs on the Paiwan costumes consist of pic-
torial animal figures . . . which demonstrates a higher intellectual level.”17
By the 1920s, however, such antiseptic (and at times derogatory) descriptions
gave way to aesthetic engagement and emotional investment. As was the case
with Paiwan wood carvings, Bunun song, and Amis dance, Atayal weaving pro-
vided Japanese academics, tourists, and impresarios with a means to rejuvenate a
Japanese self ravaged by the atomism and dislocations of urbanized modernity.18
Haruyama Meitetsu has argued that the colonial history of Taiwan cannot be
written in isolation from political developments in Tokyo. For Haruyama, the
168     Indigenous Modernity

twists and turns in the history of Taiwan’s administration were intimately linked
to regime turnover in Japan’s central government. Haruyama rightly insists that
Japan’s ruling elites were a dynamic and internally divided group and that one
must specify “who, what, and where” when imputing motives and causality to “the
metropole” in the study of colonial Taiwan.19 The same can be said for Japanese
consumer, artistic, and literary tastes. They too were various and protean, and they
constituted a dynamic aspect of Taiwan’s history, as well.
In his 1991 book Entangled Objects, Nicholas Thomas devised a useful frame-
work for elucidating the long-term imbrication of indigenous renaissance with
colonialism and its legacies, and for bringing translocal networks and economies
to bear on the siloed studies of nationalist history or community-based ethnogra-
phy. In his study of asymmetrical historical interactions between British imperial
agents and Fiji and Samoan Islanders, Thomas proposes the “entanglement” meta-
phor as an alternative to “incorporation” (a triumphalist or extinction narrative of
global capitalism’s rise to dominance) or “comparison” (the critique or lionization
of capitalism through comparison with putatively alternative economic logics).
The “entangled objects” model, which I employ in this analysis, steers a course
between ascertaining the rate and extent of the periphery’s transformation by the
core (“incorporation”) and the meticulous reconstruction of internally coherent
ideal-typical systems of meaning (“comparison”).20 The former method is that
criticized by Haruyama as too deterministic and reliant on Marxist stage theories,
while the latter approach comes from the efforts to write “internal” histories of
indigenous peoples that filter out “external” influences.
To illustrate, from an “incorporation” perspective, LeGendre’s presentation of
the red cloth to Toketok is salient because it facilitated an agreement between a
U.S. emissary and a powerful chieftain. The arrangement was but one in a series
of ad hoc accords whose breakdown fomented the Japanese invasion of Taiwan in
May 1874. As a result of this invasion, the Qing initiated more aggressive policies
against indigenes from 1875 through 1895, further eroding their autonomy. Shortly
thereafter, mechanized Japanese military might brought indigenous populations
to heel beginning in 1903. By 1915 or so, Japan delivered the coup de grâce to indig-
enous sovereignty, ushering in an era of subordinate existence in the global divi-
sion of labor and the colonial racial pecking order. In this rendering, 180 yards of
red camlet is fungible—any gift might have served the same function. This con-
clusion is at odds with the inordinate attention paid to the specific contours of
materiality by contemporary observers. Moreover, as an extinction narrative, the
“incorporation model” does little to help us understand current developments in
indigenous peoples’ rights recovery and renaissance or the continued attractions
of primitivist consumerism.21
Using the comparative method would lead us in a different direction, and
it also raises problems. Based on the voluminous travel reports and diplomatic
correspondence available for southern Taiwan and anthropological fieldwork
 Tangled Up in Red    169

studies of Paiwanese kinship, political structure, and ideology, one could ascer-
tain the role of red camlet in the redistributive political economy of the Eighteen-
Tribe Confederation. Using models constructed by political anthropologists of
Polynesia, one could then hypothesize to what extent Toketok was a chief or a big
man, or try to figure out how his brushes with the global economy transformed his
leadership style from one type to the other.
Having established the internal logic of Langqiao’s political economy and the
meaning of gifts within it, one could then contrast it with the logic of monopoly
capitalism and national sovereignty to identify what is distinctive, autochthonous,
and original about Paiwanese social organization.22 Like incorporation analyses,
comparativist studies yield important insights. But as Thomas notes, the method-
ological insistence upon “difference” in comparative work has a tendency to con-
sign actors in these systems to parallel, containerized temporalities, an analytical
fiction that is belied by the phenomenon of meaningful cross-cultural exchange.
An “entangled objects” analysis instead lingers a bit longer on the materiality
of LeGendre’s gift to take notice of the fact that indigenes refashioned imported
red cloth to create what has come to be known as Atayal traditional clothing (see
figure 28). Illustrations and Japanese ethnological displays (see figure 29) helped
stabilize this marker of Atayal tradition.
These garments in turn took on a variety of local meanings, values, and usages
that had little relevance to the story of international relations but were, nonetheless,
crucial for social reproduction in indigenous societies.23 Following LeGendre’s red
camlet through another iteration, we observe that the articles that have come to be
known as traditional Atayal cloth were in turn reappropriated by Japanese of vari-
ous stripes for a multitude of purposes during the colonial period (and beyond).
As museum pieces, objects of study, and popular items at souvenir stands, Atayal
textiles came to symbolize either a particular ethnic group or the artistic genius of
an ancient, vanished race of Austronesians (see below).
Fast-forward to contemporary times: Taiwan Indigenous Peoples have refash-
ioned elements of Japan’s colonial-period repository of material culture. The
revival of Atayal weaving practices, partly based on consultation of textiles col-
lected during colonial times and preserved in Japan, now illustrates claims of
Atayal distinctiveness and autochthony.24 The same can be said for Paiwan, Rukai,
and Saisiyat cultural revivals in the post-1980s milieu.25
As we have seen in the introduction and the first two chapters of this book,
indigenes received, in the form of gifts (with strings attached), tons of red cloth—
as remnants, garments, bolts, or even flags. We have also seen how the Taiwan
Government-General clamped down on gifting by 1900. In the post-Pacification
Office dispensation, what LeGendre called “tokens of friendship” were deployed as
the candy backed by the whip of trade embargoes and punitive expeditions. During
the period of Japan’s “primitive accumulation” of an indigenous storehouse of cul-
tural artifacts, anthropologists entered into these transactions. They wrote about,
170     Indigenous Modernity

Figure 28. Illustration of Atayal textiles from Japanese ethno-

logical survey, ca. 1915. Kojima Yoshimichi and Kōno Kiroku, eds.,
Banzoku kanshū chōsa hōkokusho (Taipei: Rinji Taiwan Kyūkan
Chōsakai, dai 1-bu, 1915), n.p.

sketched, photographed, and collected these items, then supplied museums and
expositions with evidence of indigenous ethnic integrity, local genius, and archaic
vibrancy. Concurrently, trading-post operators accepted Atayal cloth as specie and
retailed it to tourists and collectors, diffusing weavings widely to Japanese and
international collectors.
 Tangled Up in Red    171

Figure 29. A diorama from the 1913 Osaka Colonial Exhibition with Atayal red-striped
capes prominently displayed as “The household of the Taiwanese natives and its customs and
manners.” “Grand Colonial Exhibition at Tennoji Park,” ip1472, East Asia Image Collection,
Lafayette College, Easton, PA, accessed July 26, 2017, http://digital.lafayette.edu/collections/
eastasia/imperial-postcards/ip1472.

Much of the indigenous cultural material that was photographed or preserved

for posterity was collected during the “lost decade” of Japanese rule (1895–1905),
which is usually considered a period of passivity, at worst, or simplistically as one
dominated by a policy of “nurture.” However, during this decade, the peoples
beyond the savage border who had kept the Qing state at arm’s length dazzled
Japanese officials in search of allies and informants with their song, dance, festi-
vals, folktales, and handicrafts. Indigenous emissaries to feasts and parleys posed
for photographs and paintings; indigenous artifacts were collected by rural offi-
cials in exchange for tobacco and red thread that was subsequently distributed
to shore up headmen’s claims to authority; and indigenous textiles purchased in
cross-border trade were boxed up and shipped to Japanese museums and indus-
trial expositions to line the pockets and enhance the profiles of Japanese anthro-
pologists (see figure 29). These myriad transactions had the cumulative effect of
institutionalizing a view of indigenes as racially and culturally distinct from each
other and from Han Taiwanese.
The contemporary world is politically and economically dominated by set-
tler states and their majority populations.26 Therefore, indigenous peoples, for
both commercial and political reasons, often find themselves forced to perform
172     Indigenous Modernity

identities that meet the expectations of outsiders, whether in court or in the mar-
ketplace. Cultural reification was certainly a prominent legacy of Japanese colo-
nial rule in Taiwan. The objects discussed in this chapter, then, remain in play
as articles of ethnic pride and as exhibits in the high-stakes game of indigenous
renaissance and recovery.27

T O K E N S O F F R I E N D SH I P A N D T H E I N N O C E N T
A B O R IG I N E

The themes of victimization, trade dependence, and exploitation punctuate

descriptions of indigenous-outsider relations from the very first written records
about Taiwan. Chen Di’s 1603 Account of the Eastern Barbarians (Dongfan Ji)
explains that Fujianese traders brought “agates, porcelain, cloth, salt, and brass” to
Taiwan to trade for deer horns, hides, and meat. Chen laments that indigenes had
“developed some desires” leading “rascals [to] cheat them with junk.”28 Two decades
later, Commander Cornelis Reyersen observed indigenes exporting diverse ani-
mal products for “coarse porcelain and some unbleached linen.” These exports
were brokered by the “Chinese living there, who . . . married local women.”29 A
February 1624 Dutch East India Company record suggests that trade dependency
gave Chinese immigrants near the future Taiwan-fu leverage disproportionate to
their numbers: “In almost every house . . . one, two, three, nay sometimes even five
or six Chinese are lodged, whom [indigenes] keep very much under control. . . .
Likewise they themselves are bullied by the Chinese for not giving them food or
not working hard enough. The Chinese immediately threaten to deprive them of salt,
which means they are dependent on them.”30
As we saw in chapter 2, the history of interpreters throughout the Qing period
reveals a long record of economic ties that exceeded the boundaries of imperial
administration. Jumping ahead to the treaty-port period, we return to the disap-
pearance of fifty-four shipwrecked Ryūkyūans on Taiwan’s southern peninsula in
1871. The calamity brought Meiji Japan’s first official visitors to Taiwan. In par-
ticular, the observations of Mizuno Jun merit scrutiny. Mizuno would return to
Taiwan as the top civilian official in the Taiwan Government-General at its incep-
tion in 1895.
In 1873, Ambassador Soejima Taneomi dispatched a twenty-two-year-
old Mizuno, who was studying Chinese on the continent, from Hong Kong to
Danshui to investigate conditions surrounding the deaths of the Ryūkyūans.31
Upon arriving at Dakekan, Mizuno’s party trekked eastward on a steep wood-
cutter’s path toward the so-called savage border. Mizuno was told that areas of
Chinese habitation were marked by the russet color of denuded forests, while the
indigenous areas were lushly green. Mizuno’s informants told him that this divide
ran the length of Taiwan. It was patrolled by armed Han and indigenous guards
on either side.
 Tangled Up in Red    173

In a clearing used to initiate cross-border parleys, on May 23, 1873, Mizuno

hailed a group of passing Atayal people. The men fled to the hills at the sight of
Mizuno’s armed Chinese guides. Two Atayal women, however, stayed behind.
They explained that their village had been the victim of a ruse. Chinese traders had
promised the delivery of Western goods to lure unsuspecting Atayal people into
the clearing. The Chinese subsequently kidnapped the Atayal men and ransomed
them later in exchange for titles to land. To overcome their reticence, Mizuno
offered to distribute large quantities of red cloth, à la LeGendre, if the two women
could bring an Atayal headman down from the mountains. Mizuno noted that
indigenes coveted red cloth most of all. The next morning, he presented the two
women with gifts of red cloth, matches, small daggers, and pearls. That afternoon,
the chief sent a different women’s contingent down the mountain. Mizuno sup-
plied each with a “foot or two” of red cloth. Finally, Mizuno presented the chief
with a live pig and two large jars of shōchū liquor. With this presentation of “tokens
of friendship,” Mizuno had accomplished the “principal goal of his mission, to
look into savage strengths and weaknesses, degrees of intelligence and ignorance,
and manners and customs.”32
Although Mizuno did not record population figures, estimates of military
strength, or routes to interior villages (as later travelers would), the intelligence he
collected on his mission to Dakekan was put to use a quarter century later. Every
transaction in Mizuno’s account was premised on the Dakekan peoples’ desire to
obtain imported goods—twenty-two years before the onset of Japanese colonial
rule. At the same time, as first governor-general Kabayama Sukenori’s civil admin-
istrator, Mizuno issued proclamations and oversaw policies based on his estima-
tion that the seiban (raw savages, i.e., unassimilated) were generally victims of the
cunning and duplicitous Han Taiwanese.
George Taylor, an imperial maritime customs agent for the British crown,
attested to the popularity of red woolens in southern Taiwan during the same era.
Taylor observed that the red color alone made the cloth desirable, since the bright-
ness achieved by Western dyeing techniques could not be achieved by local meth-
ods. These “serges,” as they were called, were quickly pulled apart and combined
with sturdier local ramie, hemp, and china-grass fiber to make durable clothing.
Taylor complained that for everyday use, Taiwanese weaves were still dominating
the market, much to the embarrassment of the British official, who had hoped that
his machine-made cloth would flood the market.33
Ueno Sen’ichi’s 1891 report (see chapter 1) stated that purchasing presents for
the indigenes was most necessary for entering the savage territory. Accordingly,
Ueno brought along “liquor, tobacco, glass beads, Western red-dyed thread, brass
buttons, white ceramic buttons, and so on.” Like so many Japanese officials who
followed him, Ueno insisted that presents to indigenes be distributed equally, from
the youngest child to the paramount chief. If one indigene were treated too kindly
and another too carelessly, hard feelings would result, wrote Ueno. In return for
174     Indigenous Modernity

these “tokens of friendship,” chief Watan Yūra produced sweet potatoes and a bun-
dle of rice stalks from a “head-carrying bag” and presented them to Ueno. After
this exchange of gifts, Ueno’s mission ended. He concluded that Atayal people were
simple and trusting, but they were also quick to anger and never forgot a slight.34
Like Ueno, Captain Watanabe, during the previously discussed August 29, 1895,
mission to Dakekan, distributed gifts equally among Atayal emissaries to a diplo-
matic meeting. According to Watanabe’s report, Atayal valued manufactured goods
highly, as they swaddled the empty liquor vessels, great and small, into bundles to
carry back to their villages (see figure 11). Yet the chief showed no interest in the
silver coins he was given.35 Hashiguchi Bunzō’s follow-up mission also commenced
with the distribution of red cloth, tinned meat, handkerchiefs, ornamental hair-
pins, short daggers, tobacco, and alcohol.36 Hashiguchi wrote that the Jiaobanshan
embassy men were adorned with trademark Atayal red-striped capes. Hashiguchi
emphasized the importance of red serge (a rough woolen), which he distributed
in equal shares. To produce the distinctive red garments, Hashiguchi reported, the
women took the serge apart and wove the dyed thread together with locally pro-
duced ramie fiber.37
Based on his experience with the Jiaobanshan emissaries, Hashiguchi, as
director of the Office of Industrial Promotion (Shokusan-bu), proposed that each
Japanese garrison near the savage border stock gifts for distribution to neighboring
tribes. The Civil Affairs Bureau accordingly sent memoranda to the subprefects of
Tainan, Miaoli, Yunlin, Yilan, Hengchun, and Puli explaining their importance. It
specified that scarlet cotton fabric, red beads, flower hairpins, cigars, daggers, red
blankets, red serge, and hand towels were all to be stocked. Every item on the pro-
posed inventory matched one that was distributed by Hashiguchi to the Dakekan
emissaries three days earlier, on September 8, 1895.38
As recounted in chapter 1, after the Pacification Office was opened in mid-1896,
station chiefs were ordered to meet with headmen on appointed days and distrib-
ute presents. The belief that indigenes treasured their gifts and would do just about
anything to receive them informed the office’s optimistic charge. With bolts of red
cloth, bags of salt, and bottles of sake as inducements, these hundred men would
survey the political, demographic, mineral, vegetative, and military strength of
some seven-hundred-odd settlements in uncharted territory.39
In addition to having power as political and diplomatic implements, red-dyed
textiles were also known as inducements to commence trade relationships. On
September 30, 1895, an anonymous Civil Affairs Bureau translator (tsūyakukan)
told a Japanese metropolitan readership that adroit gifting could open Taiwan’s
interior to Japanese camphor merchants. The cost of gifts to chiefs was about
one yen per camphor tree in Miaoli, he reported. These gifts included red, black,
brown, or purple cloth scraps for women and guns, swords, sake, and tobacco for
men. To obtain the camphor trees, he wrote, local Chinese also traded Nanjing
coins or rings and bracelets made of pearl and lead. The selling price of a camphor
 Tangled Up in Red    175

tree was sixty or seventy yen—a more than sixtyfold return on investment. This
particular correspondent surmised that indigenes did not understand the value of
currency or the difference between silver and gold. He wrote that they accepted
one or the other based on color preferences for yellow or white and that one could
even use shards of glass or chunks of metal for currency in some areas.40
In October 1895 Ueno Sen’ichi’s 1891 report resurfaced in a commercial guide
published by the Japanese Ministry of Agriculture and Commerce in Tokyo. The
guide reiterated Ueno’s advocacy of red cloth as a rain-making gift and as the price
of entry to the “savage territory.” Moreover, it duly reproduced Ueno’s capsule
history of Han aggression, which partly attributed land dispossession to the sup-
posed indigenous character traits of simplicity, illiteracy, and lack of foresight. The
Chinese, according to this iteration of Ueno’s now recycled report (itself a distilla-
tion of translated, paraphrased, and plagiarized passages from Western travelers),
were greedy, cunning, and unscrupulous.41
A November 1896 report issued by the Industrial Promotion Section began its
discussion of camphor with a predictable admonition that travelers stock “liquor-
meat-cloth,” especially Western imported red-dyed cloth. It then excoriated the
residents of Beipu as dishonest indigenes to justify stationing armed guards in the
area to protect camphor workers. In addition, countering the economic innocence
argument, the report announced that indigenous headmen in Qing times required
frequent cash payments, in addition to feasts and gifts, to keep the peace.42
A series of internal memoranda from the Luodong (Yilan) section chief for
aborigine affairs also engaged in the rhetoric of indigenous victimization in com-
merce. He urged honest, fair-dealing Japanese immigrants to insert themselves
into the “aborigine trade.” If they didn’t, he warned, the profits would remain in
the hands of unscrupulous Chinese who set off violent cycles of revenge feuds
by cheating Xitou and Nan’ao peoples.43 The Luodong office claimed to have bro-
kered marriages between the commercially minded jukuban (acculturated sav-
ages) of Alishi village and daughters of the Atayal Nan’ao villages to strengthen
government ties with the mountain peoples. The district office held a large wed-
ding banquet at which over a hundred Nan’ao Atayal guests were feted in May
1899. The Japanese official also encouraged the adoption of indigenous males
into jukuban households in order to recruit these bicultural couples to act as
interpreters. Qualifying the stereotype of Atayal disinterest in material gain, this
scheme was premised on the lure of trade goods and wealth to inspire the Nan’ao
villagers to place their daughters and sons among the trading villages at the foot
of the mountains.44
As we have seen, Japanese accounts of diplomatic gifting repeatedly insisted
that indigenes were acutely sensitive to “equal distribution,” whether it was equal
measures of cloth or portions of boiled meat. It is hard to discount such reports as
projections or stereotypes, since they exist in so many forms. But did such behavior
mean that indigenes were egalitarian, unselfish, and innocent of the profit motive?
176     Indigenous Modernity

At distribution events, a Japanese official could proclaim that his emperor would
not play favorites. In kind, indigenous emissaries equally distributed goods among
their own followers, at least in the presence of Japanese officials. But these happy
structural isomorphisms had their limits. According to infantryman Irie Takeshi,
Puli prefect Hiyama Tetsusaburō doled out the expected blankets and jars of liquor
at his wedding to the daughter of the Paalan chief near Wushe. Complications
arose, however, when Hiyama, unable to distinguish Toda and Truku men from
the Paalan men, distributed gifts to everyone. Angered by the fact that their rivals
from Toda and Truku received gifts, the Wushe men ambushed them after the ban-
quet and took their gifts at gunpoint. Irie reported that the Wushe men asserted
their right to receive the gifts first and then to redistribute them to other locals as
they saw fit. After all, their chief had conducted a marriage alliance with Hiyama.45
Hiyama’s successor in Puli, Nagano Yoshitora, triggered the same response by dis-
tributing presents to Toda emissaries after a feast in 1898. On their way back home,
the Toda men were ambushed by Wushe men. In the ensuing battle, Toda and
Wushe suffered fifteen and two casualties, respectively.46
Moreover, detailed lists of gift items in the manuscript records of traveling
district officials reveal that more expensive gifts were earmarked for headmen.
These unequal distributions were, in fact, routine.47 In summary, some individu-
als and groups of indigenes did not view material goods as fungible commodities
that could be reduced to a value expressed in monetary terms. Nonetheless, these
goods were still highly valued, were sites of contestation and competition, and
were deployed by indigenes as either commodities or political currency.

T R A D I N G P O S T S , B E HAV IO R M O D I F IC AT IO N , A N D
P U N I T I V E E M BA R G O E S

Saitō Otosaku, Pacification Office chief in Linyipu, articulated the growing chorus
of criticisms directed at Hashiguchi’s gifting policy in an 1898 white paper. In his
preface, Saitō wrote: “One must take care in distributing gifts to the indigenes; if it is
done carelessly, it can lead to feelings of injustice and foment anger, or it can cause
lethargy and shiftlessness . . . We must not distribute gifts without a reason; we must
certainly not distribute luxury goods; we must not give in to demands for goods;
when gifts are requested, we should give no more than is absolutely necessary.”48
In language that reflects a newfound confidence in the Japanese government’s
ability to command rather than placate, Saitō composed a well-calibrated scale of
gift categories. He reserved the Hashiguchi-style “tokens of friendship” for first-
time visitors to government offices. Return “guests” would have to earn their gifts.
For example, to receive goods classified in Saitō’s top-shelf categories, indigenes
would have to perform “labor on roadwork, afforestation projects, or stock-rais-
ing/farming enterprises . . . or service as savage auxiliaries (banhei).” Such efforts
would be rewarded with “thick cotton shirts; black cloth, light cloth, table salt,
 Tangled Up in Red    177

Figure 30. Trading post at Jiaobanshan. Riban Gaiyō (Taipei: Taiwan sōtokufu minseibu
banmu honsho, 1913), n.p.

matches, and so on”; “buttons, all colors of wool thread, combs, tobacco, Nanjing
pearls, and so on”; and, for especially meritorious service, guns and ammunition.
Saitō recommended that farm tools, seeds, and stock be freely given because
they would wean the indigenes from their hunting economy. He argued that fire-
arms and ammunition, though necessary for the time being, should be phased out,
because hunting in and of itself was a vestige of savagery. Saitō also envisioned
the government-managed trade as a source of profit. Lesser categories in Saitō’s
typology, which ranged from hoes and hatchets to hairpins and fans painted with
nishiki-e scenes, were to be stocked as trade items.
A plan that may have been influenced by Saitō’s report was implemented in
Yilan in mid-1901. It restricted trade to government-licensed agents who would
operate with set exchange rates. For example, one deer pelt was listed as equivalent
to two feet of red woolen cloth or nine catties of table salt; one bear bladder equaled
two iron pots or one suckling pig; three catties of wood ears equaled five catties of
salt or a skein of thread; and so on. The government would profit from this trade
and use the proceeds for “indigenous betterment,” which meant building schools
and transporting and lodging chiefs’ families who came down the mountain to
enroll in the pilot education programs.49 The conversion of gift-­distribution cen-
ters into trading posts was central to the trade-for-education program attempted
in Yilan in 1901 (see figure 30). While these projects can be understood as attempts
178     Indigenous Modernity

to give indigenes a “fair shake” vis-à-vis their wealthier Han neighbors, these trad-
ing posts quickly turned into instruments of punishment, conquest, and income
generation for the Japanese state.
In 1899 Gotō Shinpei launched the Taiwan Government-General’s camphor
monopoly, along with other measures, to increase revenue to support his vision
of an efficiently run surplus-extracting colony. As Antonio Tavares has shown, the
official Japanese plan for camphor—to export large quantities of uniform quality
with profits accruing to Japanese capitalists—posed a direct threat to the Atayal,
Saisiyat, and other northern indigenes, whose leaders were accustomed to leas-
ing forest land, collecting tolls, or organizing production themselves. Accordingly,
cases of Atayal and Saisiyat violence against Japanese officials accelerated.50 In
response, the Pacification Offices were abolished in June 1898, and rural instal-
lations for aborigine management began to emphasize the importance of embar-
goes, smuggling, illicit trade, and contraband. As northern tribes put up stout
armed resistance to logging-company encroachments, government officials began
to worry less about the injury done to indigenes by crafty Chinese and to express
outrage that Han traders would subvert Japanese bans on weapons, ammunition,
or even salt to blacklisted tribes.
Taipei prefectural governor Murakami Yoshio urged that the tribes responsible
for a June 1900 armed uprising near Dakekan, which cost over a thousand Chinese
and Japanese lives and destroyed much property, be completely cut off from trade
and from receiving gifts. Friendly villages would have only a partial ban on trade,
in Murakami’s plan. Murakami sent out strict regulations requiring merchants to
be registered and calling for a complete ban on commerce in guns and salt to trou-
bled areas. Murakami believed that villages could be crushed and brought to heel
after a few months of deprivation of life’s necessities.51 In September of the same
year, the Balisha district chief in Yilan sent out a similar memorandum, calling for
selective trade embargoes against villages who defied the government’s authority.
He stipulated that feasting and trade would be permitted for tribes who had made
amends for their crimes or who were above suspicion.
Such a policy might have seemed wildly optimistic in the era when Ueno
and Hashiguchi were being led by their guides into terra incognita to purchase
interviews with “demanding” headmen and chiefs. Yet, by 1900, by following the
Pacification Office’s directive to regularly supply gifts to headmen as an incen-
tive to “heed invitations to arrive and be transformed,” the Yilan district offices
rerouted enough traffic or created enough new demand to be in a position to open
and close the spigot.52
In 1902 two major developments conspired to minimize the centrality of the
diplomatic and pedagogical functions of “entangled objects” and accentuate the
punitive power of their regulation. First, the Ri Aguai Rebellion, which pitted
an indigenous–Hakka–Han coalition of camphor producers against the govern-
ment-general, taught the Japanese that force would be required to make northern
 Tangled Up in Red    179

Taiwan’s interior safe for capitalism. Under Ri Aguai’s domination, camphor

production was too decentralized, too beholden to toll-state politics that ran on
bribes, and too unproductive to meet the camphor monopoly’s requirements for
black ink on the annual balance sheet of colonial management.53 Second, Gotō
Shinpei, through a combination of adroit manipulation and cold-blooded assas-
sination, brought the Japanese campaign against armed Han resistance to comple-
tion that same year. In late 1902 an Indigenous Affairs Section was placed under
the Police Bureau.54
In July 1905 Government Order 56 strictly regulated all “aborigine trade” mer-
chants. Although private traders were still allowed to operate, they required gov-
ernment permits. All trade items were to be registered and declared, along with
the names of all employees and coworkers.55 The long list of surrender ceremo-
nies that punctuate the annals of aborigine administration after 1906 shows that
resumption of borderland trade was important to indigenous leaders. In January
1906 the four villages of Fanshuliao (Ahou Prefecture) promised not to seek con-
traband trade goods from camphor workers on the savage border as a condition
of resuming trade. In May 1906 the Wushe (Tgdaya) tribes agreed to leave their
weapons at home to conduct business at trading posts and to stay at specially des-
ignated lodgings during sojourns for commerce.56
In the summer of 1909, the government-general put strict embargoes and rations
at the center of its much publicized “Five-Year Plan to Subdue the Indigenes.” On
October 9, 1909, after applying for terms of surrender, certain Dakekan tribes were
permitted “one rice-bowl of salt per month per person” as “gifts,” though not as
“trade items.” The tribes agreed to cease taking heads, to surrender their guns,
and to submit to biannual inspections for weapons. Toda’s terms of surrender on
October 17, 1909, also stipulated that indigenes could not negotiate the price of
goods at the reopened trading posts. Moreover, matches, salt, and daggers would
be excluded from the list of trade items and supplied as gifts to Toda residents in
amounts determined by the government-general. The following day, the Malepa
tribes submitted to similar terms, accepting a ban on trade in salt, matches, and
daggers in exchange for subsistence-level handouts. On November 11, 1909, the
Xalut tribes also surrendered, again foreswearing the right to trade in salt, matches,
and daggers.57
On April 1, 1910, the government-general began to operate its own trading posts,
instead of merely supervising trade. The management of this trade was entrusted
to the Taiwan branch of the Patriotic Women’s League (Aikoku Fujinkai),58 while
the former system of privately run licensed trading posts remained as a paral-
lel system operated by Han Taiwanese. Echoing Saitō Otosaku’s memorandum
of 1897, this system announced its intention to reform indigenous character by
suppressing the instinct to hunt.59 Special commissioner Marui Keijirō inspected
these trading posts in 1913 and urged that the Women’s League be stripped of the
contracts (and that private trading posts be abolished, as well). In a painstakingly
180     Indigenous Modernity

detailed accounting of exchange rates for adzuki beans and salt rations, among
other commodities, Marui pointed out that the Women’s League was fleecing the
indigenes. He believed, consistent with earlier reports about Han avarice, that such
unfair dealing would provoke anger in the long run. Marui argued against the
participation of Chinese merchants on the grounds that their bad moral character
was corrupting by its very nature. He even linked the presence of venereal dis-
ease among indigenes to Chinese traders. As a substitute, he recommended the
Saitō plan as the only way forward: use indigenous consumer appetites and trade
dependency as a lever to reform indigenous character through the promotion of
weaving, planting, and stock breeding.
A heavy hand was required, Marui argued, because indigenes were still men-
tally deficient in terms of their capacity to function as economic moderns. For
example, Marui suggested that they be paid only in tools or other durable goods
instead of cash for labor on roads and other public works. Why? They might fool-
ishly spend their wages. Marui expressed frustration that indigenes hiked kilo-
meters of mountain trails to save six thousandths of a yen on a catty of salt. He
attributed this stubbornness to the well-known desire of indigenes to be treated
fairly—if salt cost 5 sen per catty in Wushe, he wrote, then Toda men would not
pay 5.6 sen for it at the nearest trading post but would instead walk all the way to
Wushe to get the “fair price.”60
Soon after Marui’s report was released in October 1914, Government Order
85 called for the establishment of official trading posts to replace the Women’s
League institutions. These posts would be operated by police captains (keibu) and
assistant captains (keibu-ho), who would report to district heads. All posts would
work with fixed barter schedules, to be set at the prefectural level. Following
Marui’s plan nearly to the letter, the circular that accompanied the new trading-
post regulations stated:

The goal of trade in the indigenous territory is completely for education. To have suc-
cess, we will pay high prices for cereals, legumes, ramie, rattan, and wicker goods; we
will sell farming implements and pig and cattle stock at low prices. We want to instill
an agricultural ethos among them. We will pay low prices for deer antlers, deer penis,
animal pelts, and bones, to discourage hunting.
Moreover, villages that are not submissive will have their rations of salt severely
limited. If we interrupt the flow of salt, that will give them some time to reflect upon
their situation. This is a way to exercise coercion without resorting to brute force. It is
a “soft policy.” We are willing to sacrifice profit for the government to attain our goal
of making the indigenes into farmers.61

In the short term, the new policy failed to deter hunting. In the February 1917 issue
of the government organ Taiwan Gazette, ethnologist Mori Ushinosuke explained
that 70 percent of the value of indigenous “exports” exchanged at trading posts
consisted of animal products obtained in the hunt. Mori wrote that indigenes
 Tangled Up in Red    181

could not afford metal pots, salt, or fabrics without hunting. In addition, hunting
put meat on the table. Mori highlighted the economic irrationality of indigenous
agriculture when prices for crops were so low: at the time of his writing, a deer
penis still brought five yen, bear bladders ten yen, and a good set of deer antlers
thirty to forty yen in some markets.62 In short, a couple of well-aimed shots or
smartly set traps was equivalent to months of toil in the fields or on road crews.
As we have seen in the foregoing, indigenes were, at least in some contexts,
demonstrably motivated by the prospect of cash earnings, material benefit, and
personal advantage, according to the records of the Taiwan Government-General
itself. But these anecdotal examples, no matter how legion, did little to disrupt the
view that indigenes were ipso facto irrational, childlike creatures who lived on the
“fruits of the chase” and required tutelage. A 1935 article in a Japanese policemen’s
magazine titled “The Economic Sensibilities of Savages and [Savage] Customs”
provides but one of many examples of this resilient trope.
It maintained that indigenes did not customarily buy or sell goods based on
considerations of market price but instead valued them in accord with tradi-
tional value. This estimation was the same one that ruled out the upper Hengchun
Paiwanese from being included in the regularly administered territories with the
lower Hengchun Paiwanese in 1904 (see chapter 1). With so much home-brewed
liquor to be had for free, the article continued, there was little incentive for drink-
loving indigenes to worry about having cash on hand. Therefore, they spent their
ready cash until they were flat broke. These sweeping generalizations were illus-
trated with a large photograph of several Tsou men downing bamboo cups of
wine with gusto. The author conceded that some indigenes maximized profit and
adopted a market mentality, thanks to colonial policies of tutelage. He congratu-
lated the police and hopefully noted that the indigenes were losing their backward
habits and becoming more like Japanese.63
But if indeed some Atayal and Saisiyat farmers were making profits in busi-
ness, it is hard to see how Japanese policies were to thank for this result. Imposed
trade dependency, embargoes on necessities, fixed prices at the trading posts, and
an onerous licensing system were not aimed at producing profit-seeking, utility-
maximizing individuals but rather hard-working, surplus-producing, and pacified
imperial subjects.

T E X T I L E S I N E X P O SI T IO N S , M U SE UM S , A N D
P HO T O G R A P H S

Colonial-period writers, especially officials, considered the early years of the

occupation to have been ones of passivity regarding aborigine policy. Outposts
were lightly staffed and underfunded, to be sure. Moreover, most of the govern-
ment’s resources were poured into the war against the so-called “bandits” (dohi)
from 1896 through 1903—who were mostly nonindigenous, as far as the Japanese
182     Indigenous Modernity

could tell. Therefore, there is some truth to the view that the era of “red-cloth
­diplomacy” was a historical cul-de-sac. But for the ethnologists whose taxonomic
work, photography, collecting, and display efforts are still bearing fruit in Atayal
ethnic revival movements, the years 1895–1903 were remembered as a golden age.
To bring our story full circle, then, we return to the National Taiwan Museum,
site of the Rainbow and Dragonfly exhibition. The revered aborigine expert Mori
Ushinosuke was the first curator of the indigenous materials at the museum, and
he left a long shadow. Mori was an interpreter for Tokyo University anthropolo-
gist Torii Ryūzō, who made four anthropological surveys of Taiwan between 1896
and 1900. Mori was also a junior contemporary and sometime rival of the famous
historian, folklorist, ethnologist, and taxonomist Inō Kanori (1867–1925), a for-
midable collector in his own right. To this day, Mori Ushinosuke’s ethnographic
photographs are prominently displayed in the National Taiwan Museum’s perma-
nent exhibition and in other Taiwanese museums. Along with Inō’s and Torii’s
collected materials, photographs, and biographical information, these collectors
and the artifacts that contributed to ethnology are often celebrated and, to the best
of my knowledge, the collectors are rarely considered as plunderers or exploiters
of indigenous heritage.64
During the first decade of colonial rule, at the pacification outposts, govern-
ment halls, and army garrisons, these Meiji-period anthropologists photographed,
measured, interviewed, and collected artifacts from the indigenous representa-
tives who showed up to receive gifts or have a social drink. Inō Kanori’s field notes,
Torii Ryūzō’s published travelogues, and Mori’s serialized memoirs all describe
the period as a time when demobilized Japanese soldiers, Han-indigenous inter-
preters, ethnologists, government officials, and indigenous headmen assembled to
conduct business, exchange information, and size each other up.
In the era before participant observation, anthropologists like Inō worked
quickly and gathered evidence opportunistically. Discussions about gift items, as
we saw with Hashiguchi’s confused conversations about the loan/gift of blankets
to the Jiaobanshan emissaries, acted as prompts to initiate discussions between
parties that had very little to discuss, given the language problems that plagued
these encounters. The precious materials also functioned as the currency of access.
Inō wrote that in the “course of distributing various items colored red, which
they generally like, such as scraps of red cloth, red yarn, red Japanese flags, and
ornate hairpins,” he had divined key aspects of the Atayal guests’ mental life
(shisō). There are echoes of Ueno Sen’ichi’s 1891 report in Inō’s 1896 update, but
there are important differences. Like Ueno, Inō observed a reverence for the gifts,
polite manners in receiving them, and a lack of selfishness among the recipients:
they insisted that everyone receive the same gifts.
But Inō’s report did not compare the Atayal people (as he would later call them)
favorably to the Han Taiwanese, nor did he dwell on their simplicity or inno-
cence.65 Instead, Inō learned the local names for the numerous types of clothing
 Tangled Up in Red    183

and adornment that were fashioned from these gifts and illustrated his account
with several carefully labeled sketches. Having established that the female visi-
tors from the Wulai area were similar in appearance to the women brought from
Dakekan by Hashiguchi in 1895 (see above), Inō classified them as coethnics. To
the question, “By what name do you refer to yourselves?,” Inō heard the reply,
“Taiyal,” from members of each contingent. Inō recorded this term in Roman
script, announcing a new scientific outsider’s perspective on non-Han peoples in
Taiwan. As he noted, the Qing terms shengfan and shufan (literally “raw savages”
and “cooked savages”) were externally imposed political categories. Atayal, in con-
trast, was a self-designated ethnonym, according to Inō.66
In 1898, Inō launched a bulletin to publish research on Taiwan Indigenous
Peoples. For the inaugural issue, he published a photomontage with representa-
tives of eight tribes. The Atayal man in the montage was from the Wulai area,
probably a member of the troop that visited Taipei in 1896. Due to technological
constraints and the infrequent access Inō had to sitters, Inō’s montage was cobbled
together out of black-and-white studio portraits and field photographs of uneven
quality. This illustration could not capture the brilliant reds that were distributed
to Atayal woman at Japanese outposts as the raw materials for the textiles that
would in turn mesmerize Japanese souvenir hunters, ethnographers, and art fanci-
ers. To remedy this problem, Inō commissioned a color painting as a substitute for
display at the 1900 Paris Universal Exposition. While the photomontage displayed
the Atayal man and woman in a simple vest and Chinese blouse, respectively, the
painting took considerable artistic license to adorn them in ornate, bright-red tra-
ditional Atayal clothing.67
It appears that Inō rarely left ethnological-survey encounters empty-handed.
Inō Kanori’s large collection of cloths, carvings, and other implements formed the
basis for the Taiwan National University Museum of Anthropology collection. The
lion’s share of Inō’s over 430 specimens were obtained from his family in Iwate
Prefecture by Utsurikawa Nenozo in the late 1920s, and they were the material
foundation for the academic study of indigenous material culture in Taiwan at the
PhD level.68
Mori Ushinosuke returned to Inō’s site in late 1902 and early 1903 to photo-
graph Watan Yūra and his family. Mori took multiple portraits of individuals
and groups from Wulai and Rimogan (just upriver), with a focus on garments
and cloth production especially. Five of these portraits were exhibited in nearly
life-size reproductions for the five million visitors who attended the 1903 Osaka
Industrial Exhibition. The Osaka posters were transported to the 1904 St. Louis
World’s Fair and then picked up by American news services for further reproduc-
tion (see chapter 4). Besides the large impact the Osaka Expo had on the propaga-
tion of Inō’s map and taxonomy and Mori’s photographs, it also had a direct link
to contemporary Taiwan. According to the National Taiwan Museum’s hundredth-
anniversary guide, Mori
184     Indigenous Modernity

first came into contact with the exhibit of aboriginal culture at the [Osaka Industrial
Exhibition] in 1903. Five years later, the Japanese government established the Af-
filiated Memorial Museum of the Business Property Bureau to commemorate the
completion of Taiwan’s railroad network. Ranging from collecting objects for display,
assortment of exhibit facilities, to the allotment of proper space, Mori had a hand in
every aspect of exhibits. In 1915, the Taiwan Viceroy’s Office Museum . . . was finally
completed and the specimens . . . were transferred to the new location. Mori trans-
ferred to the new museum and worked there until his retirement in 1924.69

To the extent that the TGG museum was the first stop for visitors and a school of
colonialism for Japanese officials, Mori’s foundational work as the supplier, cura-
tor, and analyst of large collection of Atayal fabric in Taipei perhaps did more
to associate “Atayals” with culture bearers than any of his myriad activities. On
the photographic front, at the height of their popularity, Mori’s textile-rich Wulai
photographs were reproduced in Japanese geography textbooks, commercial pub-
lications, and government reports, while Inō’s ethnic map found its way into the
Japanese school curriculum.70
By 1915, the year Sakuma’s scorched-earth campaign terminated, any primary
or high-school teacher in Japan had at hand the materials to demonstrate that
Taiwan was inhabited by a number of ethnic groups, each in possession of its own
customs, languages, and territories. With the installation of Mori’s collection at
the Taipei museum, the same could be said of any important guest or ambitious
official in Taiwan. This was a true accomplishment for the ethnologists. At this
time, the overwhelming image of Taiwan Indigenous Peoples, even in textbooks
but especially in newspapers, photo albums, and postcards, was of armed savage
enemies of the state who would either soon go extinct or assimilate to Japanese
culture.

E T H N IC T OU R I SM A N D I N D IG E N O U S A RT F O R M S

As the frontier wars over camphor wound down in the 1910s, Atayal villages near
Jiaobanshan and Wulai became regular stops for Japanese tourists and visiting
dignitaries. The Atayal textiles that incorporated the red threads introduced in
the treaty-port period could now be obtained at tourist-friendly trading posts as
authentic indigenous cloth. During Japanese colonial rule, these garments made
the transition from items of local consumption and everyday use to exported,
high-quality handicrafts and art objects. In 1920, the protagonist of Satō Haruo’s
novella Wushe reported that the trading posts were stocking inferior knockoffs
of the “genuine indigenous textiles” he sought,71 while visiting Crown Prince
Hirohito himself viewed an Atayal weaving demonstration in a Taiwan exhibition
hall in 1923. The prince reportedly expressed admiration for their purity, color,
and boldness of expression.72 The indigenous trading posts and weaving demon-
strations were also on the itineraries of Prince Chichibu in 1925 and Prince Asaka
 Tangled Up in Red    185

in 1927. A photo of the sword-bearing Asaka and his police escorts towering over
three female Atayal weavers was splashed on the cover of the December 1927 issue
of the Taiwan Gazette.73
In 1933 the eminent scholar and critic Ozaki Hotsuma urged colonial officials
to enforce Japan’s Important Arts Preservation Law in Taiwan so that traditional
Atayal textiles, along with Paiwan woodcarvings, could be preserved as “national
treasures.” Ozaki rued the extent to which indigenous culture had been degraded
in Taiwan since its golden age. He believed that Atayalic artistic abilities had
peaked in the distant past, when the world’s most archaic form of linear patterned
cloth had emerged in the mountains of Taiwan. He argued that these artifacts, if
preserved in a repository for scholarly and artistic appreciation, would reflect well
on the empire itself. For Ozaki, the “normally administered areas” of Han resi-
dence possessed nothing of interest, except for derivative pieces imported from
the continent.74
Like the Japanese aesthetes who praised the genius of Goryeo-era pottery while
Japanese merchants undercut its production by flooding Korea with cheaper man-
ufactured wares,75 Ozaki did not connect the current “degraded state” of Atayal
weaving to Japanese imports or other policies that eroded traditional forms of pro-
duction. As early as 1900, the Government-General began to facilitate the con-
struction of textile factories in Xindian. The local indigenous affairs field office
recorded with satisfaction that Atayal women were being trained in Xindian and
in Wulai to run the machines.76 By 1938 journalist Harrison Forman, who shot
numerous photographs at the same indigenous village visited by Princes Hirohito,
Chichibu, and Asaka in the 1920s, observed that Atayal textiles had become luxury
goods for local people. Traditional clothing required two weeks of labor to pro-
duce a single garment, wrote Forman, while secondhand Japanese clothing sold
for about the price of two day’s labor on a road gang.77 In a letter to the editor of
Natural History, Forman lamented that “the women too are dressed by the Japanese
in cotton kimonos, which are symbolic of a movement that will rob another one of
the few remaining native groups of the world of their own traditions and culture.”78
Forman’s photograph, like a similar one by journalist Adachi Gen’ichirō taken in
1936, depicts Atayal women in Japanese clothing at work producing traditional
garments for export, all while consciously posing for ethnic-tourism photographs
(see figure 31)

L E G AC I E S A N D D I L E M M A S

Since the late 1980s, NGOs, the central government, and county offices have dis-
persed funding for indigenous language school curricula, the revival of dormant
public rituals, and the manufacture of indigenous textiles, sculptures, and other
items of material culture. As a result, the post-1990 affirmations of face-tattooing
and head-taking, and the rediscovery of Atayal textiles and traditional music and
Figure 31. Atayal women wearing imported clothing and weaving traditional clothing, 1936.
Tanaka Kaoru and Adachi Gen’ichirō, Taiwan no yama to banjin (Tokyo: Kokon Shoin, 1937),
facing p. 92.
 Tangled Up in Red    187

dance, have erased much of the public and private stigma formerly attached to
Atayal culture by Japanese and Han neighbors over the past century.79
The consultation of colonial-era ethnological writings, illustrations, and pho-
tographs has been a crucial component in many, if not all, of these revivalist
projects.80 The Japanese-period documentary record has been important for the
authentication of native traditions in Taiwan. This recourse to materials preserved
and stored by outsiders is testament to the transformative effects of the govern-
ment-general’s relocation and assimilation policies, followed by decades of cultural
oppression under the GMD. In a word, both the Japanese and the Taiwanese states
have had a hand in deracinating Taiwan’s Austronesian populations. Perhaps more
significantly, under GMD rule, 80–90 percent of Taiwan’s indigenes converted to
Christianity, while roughly 50 percent migrated to urban environments antitheti-
cal to the maintenance of early twentieth-century markers of ethnic distinction.81
Recent ethnographic research suggests that top-down cultural politics are inte-
gral to the twenty-first-century indigenous renaissance in Taiwan. Anthropologist
Michael Rudolph has noted that, as Christians, many indigenes did not initially
identify with the revitalized symbolism, languages, and ceremonies that were
being promoted by “elite traditionalists.” A period of time was required for adap-
tation and reappropriation. Scott Simon has written extensively about the emer-
gence of an indigenous elite in the 1950s in response to GMD changes in property
law and rural administration. These political operators often use identity politics
to fulfill their own political ambitions and are sometimes viewed with derision by
their rank-and-file constituents. Mitsuda Yayoi corroborates the general pattern:
educated elites formulated and promoted particular versions of indigenous ethnic
identity and subsequently mobilized followers to achieve state recognition for a
given interpretation of tradition.82 Hu Chia-yu has identified the same dynamic
but tempers this view with the judgment that after a period of elite domination,
the revitalization movement found a home in everyday life and rural indigenous
villages. In the twenty-first century, she argues, “the enhancement of local cultures
and indigenous identities are intertwined with the promotion of Taiwanese con-
sciousness and identity . . . ”83
Can indigenous renaissance in Taiwan, therefore, be likened to top-down
cultural revitalization projects in other postcolonial situations? Critical scholar-
ship of the Indian case has suggested that unifying symbols of Indian, Hindu, or
Maharashtra continuity, cohesion, and distinctiveness have been invoked by cul-
tural elites to quash internal dissent in the name of national survival.84 In Taiwan,
on a much smaller scale, energized groups of activists have won official recognition
for their ethnic groups to become eligible for office holding, public funding, and
political patronage that accrue with the state’s imprimatur. Since 2001 the number
of recognized indigenous groups in Taiwan has climbed from nine to sixteen.85
While the motives of such leaders and their followers are mixed and complicated,
these battles have occurred in an institutional framework that incentivizes the
188     Indigenous Modernity

homogenization of particular ethnic identities and the accentuation of differences

among them.86 And it was precisely these two processes that Inō Kanori and Mori
Ushinosuke sought to consolidate as collectors, exhibitors, editors, and writers in
the early twentieth century.
Some would consider it irresponsible to put the Taiwan indigenous renaissance
on a par with postcolonial Indian nationalism. The move to historicize putatively
timeless entities such as the Atayal can undermine indigenous claims to an auton-
omous political identity, according to this line of thought.87 James Clifford’s classic
study of the courtroom travails of the Mashpee Indians, whose legal claims to
rights and resources hinged on their ability to document the autochthony and
continuity of their community by recourse to visible markers of culture, is a case
in point.88 To suggest that any indigenous identity has been staged, as I have done
in this chapter, can be considered an attack upon the claims to collective redress
that are part and parcel of indigenous rights recovery movements.
However, not all proponents of preservationism are indigenes laboring against
staggering odds to regain stolen rights or establish a modicum of dignity. In post–
martial law Taiwan, Han intellectuals who have invoked indigenous “otherness”
as a tool for revitalizing Taiwanese national culture or for pulling Taiwan out of
China’s cultural orbit are primarily interested in Taiwan Indigenous Peoples as
symbols. The symbolism of authentic, timeless, and non-Han indigenous peoples
secures Taiwan’s Austronesian heritage in this discourse.89 The problem here is
that the preservationist ethos encoded in essentialist definitions of ethnic belong-
ing can backfire by creating unreasonable expectations that have grave real-world
consequences. The notion that indigenous peoples are inauthentic or not truly
indigenous if they do not wear traditional clothing or bear other markers of ethnic
difference easily recognized by outsiders is, in fact, a common one. This fixation
on authenticity shades into the political view that visibly assimilated indigenous
peoples should ipso facto lose rights or privileges (such as access to waterways and
hunting grounds or preferential treatment on university-entrance or civil-service
exams).90 The specter of the inauthentic (and undeserving) indigenous person is
never far from the surface.
As was the case in the period of Japanese colonial rule, there are scholars today
who find certain elements of indigenous cultures intrinsically valuable and of high
aesthetic worth. Han anthropologists have worked in recent years to reinstate recon-
structed forms of indigenous dance and song into the fabric of everyday life by pro-
moting public performance as education. One critic of this movement has asked:
Shall those young [indigenes] who have been in contact with Han society for a long
time identify with an image  .  .  . that has stagnated for several decades or centu-
ries? Or shall they identify with a culture that has—as a result of inevitable historical
development—interacted with other ethnic groups? And what kind of “Aborigines”
shall non-Aborigines identify with? Is it possible that [Han preservationists]—in
 Tangled Up in Red    189

­ rder to redress the Han’s former hegemony or for reasons of political correctness—
o
unconsciously bring all possibility for the Aborigines’ pluralist cultural development
to an end?91

In this view, the interests of curators, ethnologists, neoprimitivists, and progres-

sive Han activists are pitted against the majority of indigenes who have lived
among the Han for many decades and who do not wish to turn back the clock.
This proposition assumes that visible markers of identity, which perhaps define
indigenes to outsiders, somehow exhaust self-conceptions of belonging from an
emic perspective. Hu Chia-yu’s study of Saisiyat memory, identity, and ritual life
controverts such a view. She indicates that indigenous ethnic identities in Taiwan
are also maintained through oral transmissions that interact in complex ways with
material artifacts. Hu writes:
.  .  . non-verbal expressions are heavily emphasized [in] Saisiyat ritual practices.
However, the persuasiveness of ritual materials is combined with elaborated multi-
sensory operations. The process of practicing rituals, touching sacred objects, tasting
ancestral foods and drinks, listening to ritual speeches or songs, and making body
movements are all perceived as major sources of sensory multiplicity. The senses as
embodied powers are mediated through the material properties of sacred objects,
ritual foods, or stylized bodily actions. Thus . . . the ancestral past is continuously
sensed, recognized and articulated in the present to build and secure permanence in
the Saisiyat community.92

In other words, the meanings of these objects for those who treasure them as mne-
monic sites cannot be divined by mere visual inspection or abstract contempla-
tion. They derive their identity-making efficacy through use in specific contexts
that are largely unavailable to outsiders.
The three Meiji-era collectors who did so much to preserve, categorize, and
enliven these objects for outsiders, Inō Kanori, Mori Ushinosuke, and Torii Ryūzō,
could scarcely have imagined, I think, that Saisiyat, Atayal, Bunun, and other
indigenous groups would still be around in the twenty-first century to utilize their
collections. If their taxonomic labors and collecting proclivities have furthered
indigenous renewal and persistence, as it seems they have to some extent, this
long-term consequence was unintended, although not contrary to the spirit of
their work in its own context.
 4

The Geobodies within a Geobody

The Visual Economy of Race Making and Indigeneity

In the early twentieth century, Japanese official, academic, and commercial pub-
lications deployed texts, pictures, and maps to manufacture the “large reservoir
of cultural imaginaries”1 that have reemerged as reference points and resources
for Taiwan’s indigenous renaissance in the twenty-first century. In what began as
a quest to wrest resources, impose administrative order, and promote immigra-
tion to an erstwhile Qing borderland in the late 1890s and 1900s, Japan’s colonial
administration, tourism industry, and scholarly apparatus, by the 1930s, had com-
pleted numerous projects in ethnic typification, geobody construction, and racial-
ization to institutionalize indigeneity in Taiwan. This concluding chapter analyzes
these three interrelated processes to demonstrate how the increased intensity of
resource extraction in colonial Taiwan intersected with historical trends in repro-
graphic technology and new forms of state making to ethnically pluralize the
island’s populations under the umbrella of an Aborigine Territory.

T H E G L O BA L T R A N SF O R M AT IO N A S T H E AG E O F
A N T H R O P O L O G IC A L T Y P I F IC AT IO N
. . . [I]n premodern European discourses, non-Western peoples tend to be charac-
terized not in any anthropologically specific terms, but as a lack or poorer form of
the values of the centre. . . . My analytical fiction . . . [postulates] . . . a shift from an
absence of “the Other” (as a being accorded any singular character) to a worldview
that imagines a plurality of different races or peoples. The distinctively modern and
anthropological [discourse] projects natural differences among people that may be
rendered at one time as different “nations,” at another as distinct “races” or “cultures.”
The underlying epistemic operation—of partitioning the human species—makes

190
 The Geobodies within a Geobody    191

possible a variety of political and ethnographic projects: particular populations may

be visible as objects of government; they may serve as ethnological illustrations or
subversive counter-examples in comparative social argument; and these reified char-
acters may be available for appropriation in anti-colonialist, nationalist narratives.2

In his definition of pluralism, or anthropological typification, Nicholas Thomas

argues that modern colonial epistemology “partitions” subject populations into
internally cohesive, distinctive, ontologically stable groups whose unique traits are
embodied in representative figures, or anthropological types. To put these modern
anthropological paradigms into relief, Thomas defines premodern ethnology as its
foil, in terms strikingly similar to the assessment of Qing ethnology by Japanese
ethnographer Inō Kanori. For Thomas, anthropology’s interest in the languages,
technologies, and belief systems of remote and small populations reflects, in part,
modern cultural pluralism’s axiom that all peoples are inherently worthy as mem-
bers of the human race. Pluralism represents an epistemic rupture from Christian
and Enlightenment paradigms of incorporation/assimilation, argues Thomas. This
is so because difference configured plurally has something to offer multiple con-
stituencies, as it gains traction as the object of scientific inquiry, exotic interest,
ethnic pride, or racialist pessimism. In the new dispensation, pluralism is invested
with an emotional intensity that forestalls the withering away of diverse cultural
formations under the impact of the homogenizing forces consolidated during the
long nineteenth century’s global transformation (see introduction).
For Meiji ethnologists, who defined themselves as pluralists in Thomas’s sense
of the word, the first Chinese travelers to Taiwan were the very picture of pre-
modern traveler/anthropologists. They postulated an intellectual chasm between
themselves and their Qing predecessors in terms not unlike the ones Thomas uses
to separate premodern and modern anthropological imaginaries. For example, in
1905, a decade into Japanese colonial rule in Taiwan, Inō wrote:
When the Chinese first learned of Taiwan’s location, they acknowledged the e­ xistence
of the island’s own people, or “the natives.” . . . But at the time, they only ­recognized
the natives as a different people, with different language and customs, but did not
give them a particular name.  .  .  . In Ming times, the name “Eastern Barbarians”
(dongfan) was used . . . .After the Qing occupied Taiwan, there were two major divi-
sions, based on the presence or absence of political compliance, the seiban and the
jukuban . . . They did not, [however,] make observations about race.3

As the Qing period wore on, Inō wrote, Chinese observers began to add details
to their reports and make distinctions among the various non-Han peoples of
Taiwan. They even propounded theories for the origins of the Taiwanese natives.
But in the final analysis, the “Chinese observations about the Taiwanese natives’
essential nature, in relation to race, were extremely crude. There are no theories
propounded here worthy of reference in scholarly discourse. The reason for this
can be attributed to one cause: they based their theories on a general view that
192     Indigenous Modernity

recognized the natives as beyond the pale of civilization, almost as a different spe-
cies, one largely excluded from humanity.”4
It is clear from this passage that by “race,” Inō meant something more than
a bundle of genetic traits that defined a Mendelian population, biologically or
somatically. While Inō’s project, as we shall see, essentialized and homogenized
each subgroup of indigenous people through the process of typification, it also
vindicated their humanity.
Thomas rightly insists that modern colonizers were far from unified in their
conceptions of populations that underwent “anthropological typification.” For
Thomas, typification could shade into a version of relativism. Preservationist eth-
nologists or missionaries, in some cases, attempted to militate against the necrop-
olitical Darwinian outlook of their fellow colonialists. In our case, Gotō Shinpei’s
point man on aborigine policy, Mochiji Rokusaburō, wears the black hat. In the
face of their exculpatory discourse, Mochiji excoriated anthropologists for their
relatively positive assessments of indigenes. In Mochiji’s view, there were only two
kinds of “savages”: compliant and rebellious. The rebels existed beyond the bounds
of sovereignty. Mochiji had little patience with pluralists and implicitly blamed
them for Japan’s lack of industrial progress in the uplands.

M E I J I NAT IO NA L I SM A N D G E O B O DY C O N S T RU C T IO N
I N C O L O N IA L TA I WA N

For Thomas, “colonialism’s culture” locates positivities (“cultures”) among the

“ungoverned,” “undergoverned,” or “savage” peoples who are the objects of com-
mercial, missionary, or military forces in the expanding international system. In
contrast, early modern ecclesiastical/dynastic agents in such areas posited a nega-
tivity, or an absence of attributes thought to be constitutive of the imperial center,
in their accounts of “barbarians/savages.”5 In this analysis, Taiwan’s solidly colored
secondary geobodies became placeholders for the positivities, or cultures, that dis-
placed the blank, absent spaces of “savagery” in the Qing imaginary. The maps in
figures 32–34 illustrate a sequence of contrasting logics.
The map in figure 32 is taken from an 1895 Japanese commercial publication;
it is based on Western maps of Taiwan from the treaty-port era. In it, Taiwan is
bifurcated into an eastern “savage” half and a western “Chinese” zone. The line
separating the two zones is two-dimensional, stark and clear. This cartographic
convention, which does not demarcate administrative boundaries or conform to
natural land forms or emic senses of place but is rather a purely ideological con-
struction, replicates the discourse on “separate Taiwans” advocated by Charles
LeGendre after 1872 (see chapter 1).6 In LeGendre’s view, over half of Taiwan lacked
government and civility. These maps, in fact, made the argument that over half of
Taiwan was terra nullius, and up for grabs in terms of international law.
Figure 32. Map of Taiwan, 1895 (published in October). Figure 33. Japanese census map, 1905. The Special Figure 34. Ethnic map of Taiwan, ca. 1912. Takeuchi
Akishika Kenkitsu, Taiwan shiyō (Tokyo: Seibidō, 1895), Population Census of Formosa 1905: Report of the Sadayoshi, Taiwan (Taipei: Taiwan nichinichi shinpō,
n.p. ­Committee of the Formosan Special Census Investigation 1914), 620.
(Tokyo: Imperial Printing Bureau, 1909), n.p.
194     Indigenous Modernity

The 1905 Japanese census map (figure 33) designates the white space in its
middle as the “Aboriginal Territory.” This map reveals the limits of Japanese offi-
cialdom’s ability to inspect, enumerate, and travel to Taiwan’s interior during the
camphor wars discussed in chapter 1. The census map indicates that Japanese offi-
cialdom regarded the Aborigine Territory as a limit, beyond which it could not
properly conduct the basic functions of a modern state. The third map (figure 34)
filled in the blank white space of the 1905 census map with culture areas. Each
shaded polygon represents a shuzoku (race/tribe) in possession of a unique lan-
guage, material culture, and corpus of myths and legends. This map was paired
with a photomontage of male and female ethnic types. The polygons in figure 34
are second-order geobodies in this analysis. The pictured map was published in
1914 in a commercial publication; it was based on Inō Kanori’s map in figure 2,
composed between 1898 and 1900.7
The extension of Thongchai’s geobody concept to second-order geobodies is
justifiable, I think, because indigenous-territory geobodies within geobodies, like
national geobodies, were strategic responses to the era’s competitive imperialism.
In each case, cartography was fundamentally about the mastery of space from a
remote, central location at a bureaucratic apex. Whereas late nineteenth century
Thai monarchs adapted an alien form of cartography to preserve autonomy in the
face of an encroaching international system, the Japanese anthropologists in our
story sketched their ethnic maps to extrapolate knowledge gained in limited and
specific encounters to assert mastery over territories whose state of lawlessness
and opacity was perceived as detrimental to the survival of the Japanese Empire.8

T H E G L O BA L T R A N SF O R M AT IO N
A S T H E AG E O F R AC E

As self-aware non-Westerners, Japanese imperialists are said to have practiced a

sort of mimesis as they internalized and reapplied race as a discursive weapon vis-
à-vis other Asians, or to have refracted Western idioms in the process of becom-
ing a full-blown colonial power. This narrative of Japanese race making as one of
mimesis and/or refraction focuses our attention on the question of how race was
“domesticated by the Japanese” and then redeployed in the colonies.9 As we shall
see, the mimetic framework is not without value for the study of race making in
colonial Taiwan. Indeed, formally trained and westward-facing ethnologists like Inō
Kanori, Torii Ryūzō, and Mori Ushinosuke had a role to play in the popularization,
operationalization, and conceptualization of indigenes as members of a distinct
race or races. They truly exhibited the processes of mimesis and refraction, so to
speak. But their formulations were mediated, modified, and contested in the mass
media by metropolitan visionaries, postcard publishers, editorialist-constables,
colonized subjects themselves, and myriad others who were at best loosely affili-
ated with explicit programs of civilization and enlightenment or Westernization.
 The Geobodies within a Geobody    195

Moreover, the Qing precedents in aborigine administration, image production,

and category construction, all of which deeply informed Japanese projects in race
making, fell largely outside the history of Euro-American cultural history. In light
of race’s multiple and independently accreted genealogies, then, the single-origin,
diffusionist model of race as subfield of European intellectual history is of limited
utility for a long-term analysis of indigenous Taiwan’s encounter with the global
transformation.10
In light of these considerations, this chapter will adopt a looser construction
of “race” as a category and social force. Following W. J. T. Mitchell, I regard race
as a “medium,” akin to a language game, rather than as a Euro-American pseu-
doscientific category that creates its own referent. Mitchell has argued that the
medium of race provides the iconography and language for representing any
theory, ideal, or argument that asserts the internal homogeneity and “groupness”
of particular “kinds” of human beings and their distinctiveness from similarly
configured groups (races). Mitchell’s conceptualization recalls Thomas’s definition
of typification/pluralism, insofar as it minimizes the difference between “­culture”
and “species” as bases of invidious comparison. But it goes further. Mitchell
concedes that the racial medium is often inflected with the familiar metaphor
of species difference. But “race talk and race thinking” can also adopt the imag-
ery and discourse of culture, class, status, or gender to configure difference as
natural.11 Mitchell’s capacious definition is salient here because, as we shall see,
photographer-­anthropologists, publishers, propagandists, and merchants rarely
portrayed Taiwanese racial difference in strict, Linnaean terms or even in logi-
cally consistent language or iconography. Rather, race was the thread or mediating
category that conjoined a mix of ethnonyms, slurs, stereotypes, and ideals into a
language of difference and hierarchy.
Mitchell, along with a number of other analysts, including Patrick Wolfe, con-
siders modern formulations of race to be the aftereffects and not the cause of
racism. In other words, racism is “what hurts.” Racism is the practice of discrimi-
nation, genocide, exclusion, isolation, or segregation implemented against a race.12
Qing-period (1700s–1895) resource wars along the frontier of Han settlement on
the island of Taiwan produced a racism whose institutionalization provided the
terminology and raw materials for Japanese race scientists to contend with in the
Meiji period and beyond.13 Therefore, the use of imported Western ideas involving
somatological or linguistic criteria for racial classification did not produce racism
in Taiwan but did provide a new language game, or medium, for framing policy
options that could exacerbate or mitigate preexisting forms of racism.
To understand the genesis, meaning, and staying power of ethnically pluralist,
biologically invidious, and territorially bounded indigenous formations in Taiwan,
we must not only ask what men like Gotō Shinpei hoped to achieve by institu-
tionalizing the scientifically accredited notions of hierarchy and order imbibed
during his study of medicine in Germany. We must also identify and locate the
196     Indigenous Modernity

dispersed interests, mechanisms, and micropolitical arrangements that allowed

the medium of race to proliferate and sustain itself across vast social fields.14 The
power of anthropology and its related discourses is not to be found in connections
between individual ethnologists and particular policies but rather in the rupture it
introduced into discourse.
As the following sections will illustrate, even if few Japanese or Taiwanese read
ethnological reports emanating from the highlands in the early twentieth cen-
tury, ethnology reflected and inspired a plethora of lithographs, line drawings,
paintings, ethnic maps, and photographs of indigenes. These artifacts circulated
via newspapers, picture postcards, photo magazines, museum exhibitions, and
exposition guides from the onset of colonial rule in 1895 right down to the end
of the Pacific War in 1945. By the 1930s at the latest, it would have been impos-
sible for tourists, colonial administrators, or politically conscious Japanese citi-
zens to imagine a Taiwan shorn of its indigenous presence. Despite their relatively
small numbers, about 2 percent of the island’s population, the Indigenous Peoples
produced by Japanese graphic artists, novelists, administrators, scholars, and the
indigenes they interviewed, conducted trade with, photographed, and married
covered half of the island’s surface. The Atayal, Sediq, Truku, Bunun, Paiwan, Ami,
Yami, Saisiyat, Rukai, and Tsou peoples were officially, academically, and popularly
understood to be Taiwan’s original inhabitants and were composed of several con-
tiguous but distinct culture zones—in an interwar milieu that saw the conqueror’s
right to rule displaced by the discourse on sovereignty and authenticity.
The affirmation of indigenous autochthony and ethnic integrity was much
more than the discovery of truths obscured by Sinocentric Qing discourse. As
a broad-based repudiation of the notion that Taiwan’s interior and eastern rift
valley were inhabited by peoples “beyond the pale” (Thomas’s “poorer version of
the values at the center”), the creation of indigenous geobodies in Taiwan was,
dialectically speaking, the negation of a negation. That is to say, the indigenous
geobodies that coalesced in 1920s Taiwan did not restore Indigenous Peoples to
their rightful place as original owners of the island, but rather witnessed the birth
of two-way symbolic traffic between a nationalized citizenry and an excluded
minority that operated, for the first time, without a mediating layer of brokers
such as Kondō the Barbarian, Pan Bunkiet, or Watan Yūra. This new configura-
tion, which I have referred to as “indigenous modernity” in chapter 3, describes
an international system of nation-states riddled with pockets of quasi sovereignty.
This order is sustained by the persistent and even heroic efforts of indigenes to
reclaim ancestral lands and maintain corporate identities in the face of detrib-
alization and assimilation movements. As currently configured, it also requires
indigenes to perform identities in ways consistent with the aspirations, fantasies,
and expectations of settler/majority populations who still control most levers of
state power.
 The Geobodies within a Geobody    197

U E N O SE N ’ IC H I A N D T H E P R E H I S T O RY O F G E O B O DY
C O N S T RU C T IO N I N TA I WA N

The Taiwan Government-General set up its Taipei administration in June 1895.

Even before officials could safely report for duty, publications about the curious
folkways of the empire’s new subjects began to circulate in Japan. Especially promi-
nent were travel accounts of hill peoples, collectively known by terms that are
variations on the theme “savage,” such as banjin, seibanjin, yabanjin, or banzoku.15
Takigawa Miyotarō’s Shinryōchi Taiwantō (Our new territory: The island of Taiwan,
published in June 1895) was typical in one respect: it recycled the report of Ministry
of Foreign Affairs staffer Ueno Sen’ichi, who visited Taiwan in 1881 and 1891.16
Ueno’s report, which was introduced briefly in chapter 1, is an amalgam of Ueno’s
firsthand observations, smatterings of cribbed material, and Ueno’s translations of
British Imperial Maritime Customs agent George Taylor’s notes on ethnography.17
In addition to cropping up as source material for commercial publications, Ueno’s
notes on Taiwan were widely disseminated within Japan’s military apparatus. Thus,
we can assume they had been combed over by any Japanese official, soldier, or civil-
ian with a position of responsibility in Taiwan in the post-1895 period.
Ueno began with a discussion of Qing data and classification, continued with
a digest of George Taylor’s observations, and concluded with his 1891 travelogue
of a visit to the environs of Quchi, home of several prominent Atayal players in
the construction of indigenous modernity. To orient Japanese readers, Ueno refer-
enced the term “eighteen tribes of Langqiao.” He reminded readers that this famil-
iar sobriquet applied only to southern Taiwanese and that there were many more
than eighteen tribes, even in Hengchun. According to an 1879 Qing census of the
area, wrote Ueno, there were some fifty-eight villages under the jurisdiction of
Hengchun, plus another forty-six villages of the “Puyuma race” (shuzoku), who
also lived in Hengchun district.18
Ueno’s updates supplemented a voluminous body of Japanese records from
the 1874 invasion. Much had changed in Hengchun as a result of Shen Baozhen’s
“open the mountains, pacify the barbarians” policies of the post-1874 period. This
updated baseline data was presumably useful to Japanese constables and soldiers
who fanned out across Taiwan in 1895 and 1896 to set up the rudiments of rural
administration. In a less practical vein, Ueno’s report also provided a somewhat
confused digest of existing ethnographic information. Ueno wrote that all of
Taiwan’s so-called shengfan (raw aborigines) could be subdivided into four ethnic
groups/races (shuzoku): the Paiwan, Zhiben (Depon), Amis, and Pingpu (Pepo).
This taxonomy, inherited from Taylor, left out much of the island’s indigenous
population and presented some logical inconsistencies. For one, Ueno subsumed
the Plains-dwelling tribes (pingpu-zu) under the rubric shengfan. In contrast, later
systems considered pingpu-zu to be synonymous with jukuban, a category defined
antithetically to seiban. Like his crude taxonomic scheme, Ueno’s descriptions of
198     Indigenous Modernity

origin myths followed Taylor verbatim. Taylor himself relied upon oral testimony
and legends to fix ancient indigenous migration routes to Taiwan, instead of hav-
ing recourse to comparative linguistic data or the maps of the global distribution
of cultural traits preferred by academic ethnologists of the time. For all of its prob-
lems, this shaky overview formed the baseline for Japanese ethnologists. The “four
to five different tribes of shengfan discovered by European writers” mentioned in
Inō’s 1895 manifesto refer to Taylor’s classification, while Torii Ryūzō’s early clas-
sificatory work also cites Taylor as a pioneer.19
Ueno’s report described the costume, ornaments, and physical appearance of
his Atayal interlocutors but did not hazard to classify indigenes into subgroups—
he instead used the Chinese term fanren (savage). As we saw in chapter 2, Ueno’s
report focused instead on the cultural practices that facilitated communication
between Atayal people and outsiders, to serve as a practical guide for the con-
duct of business. Thus, the material items it assiduously described—tobacco, glass
beads, red-dyed cloth, bottled spirits—were not evidence of autochthony or cul-
tural integrity but rather were elements of a hybrid material repertoire that had
sprung up between the economically interdependent but politically independent
peoples on each side of Taiwan’s “savage border.”

T H E TA I WA N G OV E R N M E N T- G E N E R A L’ S F I R ST
E N C OU N T E R W I T H I N D IG E N OU S P E O P L E S

The newly ensconced Taiwan Government-General’s first official mission to

Dakekan-area Atayal peoples followed Ueno’s gifting and drinking playbook,
some two decades later. Mori Ushinosuke classified the Atayal members of this
mission as the cismontane Dakekan tribes (mae-Taikokan-ban). In the early 1900s,
this group consisted of nine settlements.20 Jiaobanshan was the largest of these.
The Japanese press and colonial bureaucracy generated at least seven written
reports of the events bounded by Captain Watanabe’s August 29 exploratory meet-
ing near the Dakekan garrison and the return home of the Jiaobanshan delegation
on September 12. Contemporary imagery derived from the embassy suggests that
Japanese officials, writers, and publishers did not bring ethnologically informed,
pluralistic categories with them to Taiwan in 1895. These would be formulated
and disseminated over the course of colonial rule, on site. For example, articu-
late officials like Hashiguchi Bunzō and Captain Hirano Akio employed Qing
terminology for “savagery” (seiban) and mentioned specific settlements and the
names of their leaders. They did not, however, use the term Atayal. Nor did they
attach culture zones to ethnic maps of the island. As we shall see, the first wave
of Japanese-produced icons of indigeneity in Taiwan conform to Thomas’s ideal-
typical premodern imaginary, wherein other peoples “tend to be characterized not
in any anthropologically specific terms, but as a lack or poorer form of the values
of the centre.”21
 The Geobodies within a Geobody    199

The first published Japanese illustration of an Atayal person, based on an actual

face-to-face encounter, appeared in an article in the September 29, 1895, issue of
the mass-circulation Tokyo Asahi shinbun. The article included a crude sketch
of the woman known as Papau Iron, later known as Habairon. This line draw-
ing made no attempt to capture individual features but was rather a schematic
drawing of facial tattoos. According to this article, Habairon was about thirty-five
years old in 1895; she is described as a woman bearing forehead and cheek tattoos,
which signified her betrothal (probably an incorrect interpretation). The other
female in the party of six Atayal who trekked beyond Dakekan and traveled all the
way to Taipei with Hashiguchi Bunzō and Governor Tanaka (see chapter 1) was
described as a nineteen-year-old woman named Washa Buta (Washiiga in subse-
quent sources). Washiiga lacked tattoos because she had married a Chinese man
at the age of sixteen. In contrast to the “savages” (seiban), Chinese did not require
tattooing as part of the marriage contract, according to the report. The article also
specified that female cheek tattoos consisted of four bands, indicated in the small
line drawing referred to above. Despite its crudity, the newspaper drawing helped
to fix the female Atayal face tattoo as one of the most easily recognizable icons of
indigeneity in Taiwan.
A second, more detailed sketch portrayed two participants in the Jiaobanshan
encounter—an Atayal couple who did not travel all the way to Taipei—and
appeared as the back-cover illustration for the 115th issue of the Journal of the
Tokyo Anthropological Society (see figure 35). This sketch illustrates a short narra-
tive description of Japanese-Atayal diplomatic exchanges in September 1895. The
narrative was penned by Hirano Akio, who accompanied the September 4, 1895,
Hashiguchi mission to Caoling Ridge at the foot of Jiaobanshan. Like the Tokyo
Asahi illustration, there is no attempt at ethnic or racial classification in the article.
The personal names “Watan Nawi, Jiaobanshan village head, and Lemoi Maton,
wife of Jiaobanshan village head” are attached, under the banner of seiban no fuku
(savage attire). The illustration of Hirano’s note, like the cruder drawing in the
Tokyo Asahi, focuses on facial tattoos as the marker of indigenous appearance, dis-
tinguishing between male and female patterns: men wore a vertical set of bars on
their foreheads, while women wore the ear-to-mouth tattoos.22 The clothing, orna-
ments, coiffure, and headdress of the Jiaobanshan residents are labeled with great
care. The detached coiffures and ornaments abstracted isolated cultural elements
from individual historical actors to put Taiwanese seiban within comparative eth-
nography’s global data set of cranial indexes, hair-strand widths, and ear shapes.
At the same time, the faithfully rendered can and bottle in Lemoi Maton’s backsack
(see figure 11) situates the image in the shared time of frontier diplomacy, while
the choice of headman Watan Nawi, who is a protagonist in the narrative account,
elevates the illustration above the anonymity usually associated with “types.”
These two illustrations—one mass market, the other scholarly—made gestures
toward ethnographic accuracy by specifying the gender differences in tattooing
200     Indigenous Modernity

Figure 35. An anthropology journal sketch of Watan Nawi, 1895. Tokyo jinruigakkai zasshi 11,
no. 115 (1895): back cover.

practices among so-called seiban. The next illustration, from the November 1895
issue of the general-audience magazine Fūzoku gahō, overlooked such niceties
(see figure 36). This drawing harked back to a venerable Japanese representational
practice from the 1874 expedition.
According to the Fūzoku gahō article’s text (a faithful reproduction of excerpts
from the official report filed by Hashiguchi Bunzō), the female indigenous inter-
preter Washiiga wears Chinese clothing and speaks an indigenous language,
while the older woman, Habairon, was an unacculturated tattooed female in
local garments. In the Fūzoku gahō illustration, however, the indigenous women
lack tattoos altogether, and all of them wear a combination of Atayal capes and
Chinese-style blouses.23 In contrast to the ethnologically fastidious drawings in
the anthropology bulletin described above, the men in the Fūzoku gahō wear facial
tattoos that run from mouth to ear—a patently female pattern. More than any
other detail, the tattoo mix-up confirms that the Fūzoku gahō artist was not work-
ing from photographs or eyewitness accounts. It is doubtful that the artist even
read the text being illustrated.
Nonetheless, one element of this otherwise fanciful image did have a referent.
Kabayama Sukenori, Taiwan’s first governor-general, famously hosted the five
Jiaobanshan emissaries and their interpreter Pu Chin in Taipei in September 1895
(see chapters 2 and 3). The inset portrait of Kabayama in figure 36 appears to be
based on the commemorative photograph of the embassy.
 The Geobodies within a Geobody    201

Figure 36. Early mass-media illustration of the Jiaobanshan emissaries

and Governor-General Kabayama. “Seiban no junfuku,” Fūzoku gahō 103
(November 28, 1895): 17–18.

In this version of the photograph (figure 37), one Jiaobanshan emissary stares at
the ground, while another is in a crouching position—not unlike the postures rep-
resented in the Fūzoku gahō illustration. The other Jiaobanshan emissaries adopt
postures alien to the Fūzoku gahō drawing, however. These deviations from the
photograph seem purposeful. The creative repositioning of Kabayama as a smartly
attired military hero, who towers over the distorted figures of supplicant indigenes,
in fact echoes 1874 wood-block prints of the Japanese expedition to Langqiao. As
Robert Eskildsen has pointed out, colorful wood-block illustrations of the 1874
invasion in early Meiji-period newspapers were rife with exaggerated contrasts
202     Indigenous Modernity

Figure 37. Photograph of the Jiaobanshan emissaries and Japanese officials in Taipei, 1895.
Far left, an unnamed foreigner, probably the Portuguese interpreter who translated from
English to Taiwanese for the Hashiguchi/Tanaka mission; 4, acting ­governor Tanaka Tsunatoku,
who read a declaration of Japan’s sovereignty to the assembled ­villagers near Caoling Ridge; to
Tanaka’s left, Hashiguchi Bunzō, who led the mission and ­became a key figure in early Japanese
policy as governor of Taipei prefecture; next, a Jiaobanshan man named Marai (about eighteen
years old); squatting below him, Ira Watan (about twenty-one years old), also from Jiaobanshan;
3, seated, Civil Affairs Minister Mizuno Jun; 2, seated, ­Governor-General Kabayama; standing,
between Kabayama and Mizuno, Habairon (about thirty-five years old), from Shinajii; seated,
to Kabayama’s left, Washiiga (about nineteen years old), the Jiaobanshan woman who acted
as interpreter at Caoling Ridge; standing, behind Washiiga, Motonaiban, an approximately
forty-three-year-old Shinajii man; to Motonaiban’s right, Pu Chin, an Atayal Chinese-language
interpreter. Akiyoshi Zentarō, ed., Nihon rekishi shashinchō kinko no kan, zohō shihan hakkō
(Tokyo: Tōkōen, 1914), 106.

that posed heroic and upright Japanese soldiers lording it over gaudily attired and
obsequious Paiwanese people.24 By 1895, however, the smartly dressed Japanese
military man (figures 36 and 37), with medals and uniform, formed a contrast to
both the crouching indigenes and the pike-wielding samurai of the 1870s.
From 1895 through 1900, embassy-related imagery was predominant in
Japanese iconography of Taiwan’s non-Han population. For example, one photo-
graph of three Jiaobanshan emissaries (see figure 38) was converted into an etch-
ing for the Journal of the Tokyo Anthropological Society’s January 1896 number (see
figure 39). This image illustrated a text by ethnologist Inō Kanori.
Figure 38. Photograph of Habairon, Motonaiban, and Washiiga, 1895. Nihon Jun’eki Taiwan
Genjūmin kenkyūkai, ed., Inō Kanori shozō Taiwan Genjūmin shashinshū (Taipei: Jun’eki
­Taiwan Genjūmin hakubutsukan, 1999), 115. Photograph courtesy of the publisher: The Shung
Ye Museum of Formosan Aborigines, Taipei.
204     Indigenous Modernity

Figure 39. Ethnographic drawing of Habairon, Motonaiban, and Washiiga that appeared
in Tokyo jinruigakkai zasshi 11, no. 119 (1896), n.p. This drawing was titled simply “Taiwanese
­Savages” (Taiwan seiban).

These portraits of the three emissaries Habairon, Motonaiban, and Washiiga

were taken in Taipei or Dakekan walled city, to judge from the large masonry
structure in the background. Whereas the Fūzoku gahō illustration attempted
to establish a “savage territory” setting by replacing the built environment of
 The Geobodies within a Geobody    205

the courtyard with a bamboo grove, Inō’s commissioned etching effaced the set-
ting of the photograph completely by whiting out everything but the sitters. This
form of decontextualization was a favorite tool of visual anthropologists in the
Victorian era, because it facilitated global comparisons along certain narrowly
defined somatic criteria. But unlike the decontextualized picture that illustrated
Hirano Akio’s note (figure 35), the illustration in figure 39 does not supply labels
for items of material culture. Moreover, it is puzzling (in light of subsequent fab-
rications) that the facial tattoos described in the Tokyo Asahi article—the ethno-
graphic marker of Atayal cultural distinction in this period—have been obscured,
instead of accentuated, in this etching. Inō did, however, provide readers with a
sharp illustration of Atayal textile wizardry, an element that was muted in Hirano’s
otherwise more detailed sketch.
The photograph in figure 40 was also staged during the Jiaobanshan embassy. It
became a template for several mass-circulation images, including a photograph in
a middle-school geography textbook. This portrait of Habairon, Washiiga, Marai,
Ira Watan, and Motonaiban, with their interpreter Pu Chin, was shot in the gover-
nor-general’s reception hall in Taipei. As a comparison with figure 37 indicates, the
Atayal guests sat in the governor-general’s and admiral’s wicker chairs for a smaller
group portrait, in much less stiff poses. Ira Watan (seated) and Marai were broth-
ers (aged twenty-one and eighteen, respectively), and Motonaiban was a forty-
three-year-old widower (also seated). Washiiga and Pu Chin are standing on each
side of Marai. Habairon is seated in a wicker chair. She appears to be playing a
mouth harp, which also became, like facial tattoos, a marker for Atayal femininity
in Japanese iconography.
As we learned in chapter 1, newspaper and magazine accounts of the Jiaobanshan
embassy attached personal names to all of the emissaries pictured in figure 40,
as well as the twenty-two people who greeted Hashiguchi outside Dakekan on
September 8, 1895. This textbook illustration, however, is captioned Taiwan doban,
with no other supporting information. The phrase translates to “native savages
of Taiwan” and utilizes idiosyncratic characters. Japanese accounts published in
Taiwan always use the character ban, a carryover from the Qing period, to describe
indigenes. In contrast, home-island publications used a mixed and inconsistent
batch of signifiers for Taiwanese, even after standardized terminology had been
fixed by the government. This 1898 textbook featured only this one photograph of
Taiwanese people, making the Jiaobanshan emissaries representatives of the whole
island for young readers.25
The photograph in figure 40 reappeared in an early 1900s postcard, indicating
its commercial appeal. The publishers used the more conventional term Taiwan
seiban to label the emissaries. Like other pre-geobody-era publications, this one
lumps the Atayal men and women from the Jiaobanshan embassy together with a
Puyuma man and a group of Bunun or Tsou men (sitters in the other two photos),
under the blanket category “savage.”26
206     Indigenous Modernity

Figure 40. Photograph of Habairon, Motonaiban, Ira Watan, Marai, Pu Chin, and Washiiga,
1895. The photograph, captioned “Taiwan doban,” was circulated in Satō Denzō, Chirigaku
kyōkasho: chūtō kyōiku, Nihon chizu furoku (Tokyo: Hakubunkan, 1898), n.p. This image is
reproduced from Sugiura Sosaku, Taiwan meishō fūzoku shashinchō (Osaka: Sugiura shōkō
baiten, 1903), courtesy of the National Taiwan Library. Thanks to Joseph Allen for locating
this image.

The 1897, 1898, and 1900 editions of Sanseidō’s Teikoku chirigaku kyōkasho
(Imperial geography textbook) present a radically altered version of the same pho-
tograph in a woodcut (see figure 41). Here, the artist removed Habairon’s facial
tattoos and mouth harp. The artist also depicted the seated Ira Watan as a stand-
ing model. The built environment, Pu Chin, Marai, Washiiga, and Motonaiban
have been removed from the scene, probably to allow the reader a clearer view of
the clothing and wicker chair that were thought to typify Taiwanese customs and
manners (fūzoku) at the time. The caption, “a picture of Taiwanese” (Taiwanjin
no zu), obscures the indigenous or Atayal identity of the sitters. The Qing terms
seiban (savage) and jukuban (acculturated savage) for indigenes do appear in the
text but not in connection with this illustration. The Sanseidō text equated eastern
Taiwan with the “savage territory,” identifying it as home to the Mudan villagers
who massacred Japanese victims back in 1874.27 Because the tattoos on Habairon’s
face are not apparent in the photograph that this etching has abstracted, it is not
 The Geobodies within a Geobody    207

Figure 41. Textbook etching of Jiaobanshan emissaries, 1897.

Kamei Tada’ichi, Teikoku chirigaku kyōkasho (Tokyo: Sanseidō,
1900), 152.

surprising that they were omitted. Compared to the Fūzoku gahō illustration in
figure 36, this textbook etching is a faithful representation of the photograph on
which it is based.
Yet another iteration of the Jiaobanshan embassy photograph was published
concurrently, in 1900 (see figure 42). In this etching, the editors have retained
Habairon’s original hand and arm positions but have removed her mouth harp.
The male Ira Watan now wears female cheek-to-chin tattoos, while the male Marai
has been endowed with ponytails. Sheep, banana trees, and buildings on stilts have
208     Indigenous Modernity

Figure 42. Jiaobanshan emissaries in fanciful setting, ca. 1900. Anonymous, Shōgaku chiri
kansan (Tokyo: Shūeidō, 1900), 29.

replaced the courtyard of Kabayama’s reception area. Presumably, the point was to
place the “savages” in their native habitat, symbolized here by elevated granaries,
hilly topography, tropical foliage, and domesticated livestock. Although this was
an educational text, it resembles the Fūzoku gahō illustration in its propensity for
fabrication, enhancement, and misrepresentation.28
Two years later, an American current-events magazine called The Independent
published a cropped version of the base photograph for these etchings (see
­figure  43). Its caption also erased the indigenous or non-Han identities of the
­sitters. Following the lead of their Japanese counterparts, they referred to Ira
Watan, Marai, Motonaiban, and Pu Chin as generalized “Natives of Formosa.”29
The only other photograph in the Independent article is of three Atayal people (see
figure 18). It suggested to American readers that typical Taiwanese were adorned
with little else than striped capes or beaded vests, that women wore face tattoos,
and that everyone walked barefoot (see figure 18).
The preponderance of indigenes and the paucity of Han Taiwanese in Japanese
textbook portraiture—as exemplified in the five editions discussed above—was
given much broader extension and a state imprimatur in the first geography text-
book published, in 1903, by the Japanese Ministry of Education (see figure 44).30
In its textbook, it adapted a photograph presumably taken on August 26, 1897 (see
figure 45), when Inō Kanori visited Paalan—one of the major settlements near
 The Geobodies within a Geobody    209

Figure 43. Photo of Jiaobanshan emissaries in Western press,

1902. Goto Shimpei, “Formosa Under Japanese Administration,” The
­Independent 54, no. 2,796 (July 3, 1902): 1,582.

Wushe, and the place where Hiyama Tetsusaburō had married into a local family
two years earlier. Hiyama’s father-in-law, headman Pixo Sappo, is probably the
man in the photograph. According to Inō’s journal, Pixo gave Inō a tour, with the
help of a Sediq woman named Iwan.31 This Iwan was probably Iwan Robao, Kondō
Katsusaburō’s wife.
As with the proliferation of Jiaobanshan embassy images, the people most acces-
sible to Japanese official-scholars also became icons for an ethnic group (although
their names have become lost to history). This photograph also appeared in Karl
Theodor Stöpel’s Eine Reise in das Innere der Insel Formosa und die erste Besteigung
des Niitakayama (Mount Morrison)32 and as an etching in Kamei Tadaichi’s Teikoku
shin chiri (The empire’s new geography) (see figure 46).33 A photograph of Pixo
and an Atayal female with three other armed men, from the same photo shoot,
was also published in the widely circulated and republished reference work by U.S.
Consul James Davidson.34
Figure 44. Etching of Sediq woman and Paalan Figure 45. Photograph of Sediq woman and Paalan head- Figure 46. Etching of Sediq woman and Paalan
headman in ­Ministry of Education textbook, 1904. man, ca. 1897. Jun’eki Taiwan Genjūmin kenkyūkai, ed., Inō headman in commercial textbook, 1908. Kamei
Monbushō, ed., Shōgaku chirisho kan ni (Tokyo: Kanori shozō Taiwan Genjūmin shashinshū (Taipei: Jun’eki Tadaichi, Teikoku shin chiri (Tokyo: Sanseidō,
Monbushō, 1904), 69. Taiwan Genjūmin hakubutsukan, 1999), 142. Photograph 1908), 138.
courtesy of the publisher: The Shung Ye Museum of Formo-
san Aborigines, Taipei.
 The Geobodies within a Geobody    211

The official 1903 textbook depicts a man leaning on a rifle; he wears the trade-
mark Atayal “savage machete” (bantō); he and the woman also smoke pipes. The
female wears the striped Atayal cape so prominent in the other illustrations, but
the stripes are muted in the illustration—only visible on the arm bands (­figure 44).
The photograph reveals the bolder stripes shown in the other illustrations
(­figure 45). This alteration can be attributed to the graphic simplification required
to turn grayscale photographs into line drawings, perhaps. But there are notable
fabrications here. First, the woman in the photograph is not smoking a pipe. And
she is not bare-legged and barefoot, as in the drawing. Finally, as we saw in the
etching of the Jiaobanshan photograph, her arms have been moved and her basket
removed, presumably to accentuate the features of clothing and accoutrement that
illustrators thought most salient. That exposed, well-defined, and fabricated bare
feet were emphasized here suggests an imputation of savagery to the sitters—or
perhaps the eroticization of the photo.35
The 1908 commercial textbook by Kamei (figure 46) presents readers with a
more muddled etching: it is not as well defined as a line drawing, and its attempt
to represent multiple shades of gray cannot be considered wholly successful.
Nonetheless, it reproduces faithfully all of the elements shown in the photograph
itself, with the exception of a small dog in the background.36
The portrait of Pixo and the Atayal female remained in print in official Japanese
textbooks until the 1903 edition was replaced with the 1910 edition. On the one
hand, the textbooks, official and commercial, deethnicized their Atayal subjects
(ten years after Inō’s ethnic labels had gained traction in official and scholarly
circles) by invoking the Qing categories of seiban and banjin, respectively. On
the other hand, the widely circulated textbook images highlighted ethnic mark-
ers of Atayalness such as striped clothing, pipes, machetes, and earrings. Even if
commercial artists, education bureaucrats in Tokyo, and publishers did not know
the difference between a Han Taiwanese, an Atayal person, or someone from
Hengchun, the ethnographic lens that produced their base photographs put the
identifying icons of Atayal ethnicity into play for a national audience.
Inō Kanori, who was at least indirectly involved in the production of the
imagery described above, lodged an editorial complaint about the Ministry of
Education textbook’s illustration, but not because it was ethnographically suspect.
Instead, Inō criticized the overrepresentation of indigenous peoples in textbooks.
Not only was the majority population of Han Taiwanese given short shrift but also
the textbook was “less than ideal” because it did not make distinctions among
Hok-lo and Hakka Taiwanese. Moreover, he added, all seven indigenous ethnic
groups should have been illustrated in the textbook. The 1903 Monbushō geog-
raphy textbook also claimed that eastern Taiwan was largely wild and untamed,
ignoring the large rift valley inhabited by cattle herders and farmers; Inō also took
exception to this misrepresentation.37
As was the case with the Philippine Islands, where American image makers
favored photographs of scantily clad Ifugao and Igorot hill peoples over suited
212     Indigenous Modernity

and hoopskirted Filipino and Filipina urbanites, the visual rustification of Taiwan
sketched above created the impression that Japan was colonizing an ungoverned
territory. In the late nineteenth century, sovereignty was associated with “civiliza-
tion” in the emergent international system. The painting of colonial acquisitions
such as Borneo, Taiwan, or the Philippines as spaces of savagery, at this time, justi-
fied colonial occupation under the doctrine of terra nullius. Therefore, the pre-
dominance of Jiaobanshan embassy and Atayal imagery in early colonial-period
Japanese publications does not necessarily attest to the emergence of a pluralist
ethos in Japan, although in some respects it laid the groundwork for a plural-
ist sensibility by familiarizing large audiences with a visual vocabulary that could
portray erstwhile savages as culture bearers.

T H E S C I E N C E O F R AC E C OM E S T O TA I WA N

On November 3, 1895, the celebrated ethnologist and pioneer of Taiwan history

Inō Kanori set out for Taiwan from Ujina Harbor aboard the Aikoku Maru.38 It was
around this time that Inō drafted a manifesto that linked his projects in typifica-
tion, geobody construction, and race science to the sound management of Japan’s
new colonial possession. He wrote:
The people of Taiwan are known by three types: Chinese (shinajin), cooked barbar-
ians (jukuban), and raw barbarians (seiban). As for the Chinese, of course their de-
scendants will become obedient citizens—it should not present much difficulty to
govern them. However, the raw and cooked barbarians need to be investigated from
the perspectives of natural as well as conjectural science. Thereafter, an administra-
tion and an educational policy can be structured. As for “cooked” and “raw,” these
are general terms formerly used to reflect degrees of submission to government. If
we look at it from a scientific point of view, however, there are at least four or five dif-
ferent tribes/races (shuzoku) [of aborigines], as we know from looking at the articles
written by foreigners who have investigated this area. But what about the intrinsic,
distinctive (koyū) physiologies, psychology, and local customs of the various tribes?
What about their connections to the Philippine Islands and neighboring islanders?
To this day, these are unsettled issues. Today, the hands of our countrymen, the clari-
fication of these questions, will, it goes without saying, contribute to our political
goals . . . And we shall also obtain results in regard to our scholarly aspirations.39

From November 1895 to January 1896, with the help of Tashiro Antei, a high-
ranking bureaucrat, botanist, and seasoned observer of Ryūkyūan customs and
manners, Inō established an institutional link between the small community of
Japanese anthropologists in Taiwan and Tsuboi Shogorō’s Anthropological Society
at Tokyo University.40 With ambitious intellectual plans but no budget, Tashiro and
Inō’s society would publish its correspondence and research findings in Tsuboi’s
Journal of the Tokyo Anthropological Society. Its anthropological specimens would
be stored in Inō’s Taipei dormitory.41
 The Geobodies within a Geobody    213

The first goal of the new organization was to establish a scholarly taxonomy of
names for the different non-Han tribes of Taiwan. Inō called for scientific inves-
tigations along the following lines: biological, psychological, ethnographical,
linguistic, geographic, and religious.42 In chapter 1, Kawano Shuichirō’s first dip-
lomatic mission to Yilan in November 1896 was described in some detail. One of
the bicultural intermediaries, the Xitou woman named Awai, appears in Inō’s first
ethnological report from Taiwan. Awai acted as informant on language, village
names, and local songs in the Xitou region of Yilan. Inō reproduced her testimony
with his own commentary in his December 1895 digest of Tashiro’s report.43 By
the time Inō published his famous 1900 synthetic ethnography, Taiwan banjin jijō,
complete with an ethnic map of Taiwan, he had aggregated numerous interviews
with bicultural interpreters such as Awai as tabular results for use by bureaucratic
higher-ups who lacked direct knowledge about local populations. At this level of
abstraction, Atayal interpreters like Washiiga and Awai fell out of the picture as
informants, as did the “aborigine hands” Jiku Shō Min and Kondō Katsusaburō.
Like Habairon, Awai’s visage was turned into a crude sketch to display her facial
tattoos in Inō’s article, while her role as conduit of information, trade goods, and
topographical intelligence was highlighted in Kawano Shuichirō’s report. We can
surmise that Awai’s, Iwan’s, and Washiiga’s prolonged association with Japanese
ethnologists and policemen played a role in the ethnicization of Atayal peoples.
As Vicente Rafael has observed regarding American census categories for
Philippines Islands residents, the physical process of information gathering itself—
rather than the reception of written digested reports—was an exercise in identity
formation for both the field-level staffers who conducted surveys and the people
who provided them with information.44 Thus, it was not only anthropologists,
their readers, and government officials who were buying into the new, anthro-
pologically informed pluralist ethos. Hundreds of Taiwanese indigenes, many of
them community leaders, received monetary rewards, payment in kind, or access
to patronage for participating in interviews, linguistic inquiries, photo shoots, and
bodily measurement exercises with Japanese survey ethnologists and census tak-
ers. They were learning, in other words, that they were Atayal, Bunun, and Paiwan
peoples, instead of members of particular lineages, settlements, or ritual groups.
For Inō, the anthropologist’s value-added labor was not participant observation
but his global viewpoint—and ability to collate information scientifically. Referring
to Lieutenant Hirano Akio’s published notes,45 Inō warned readers that amateur
field reports could contribute to science only after the application of his own edito-
rial hand. Hirano described Watan Nawi’s cap as made of tanuki hide. Inō pointed
out that the tanuki (a Japanese raccoon dog) did not exist in Taiwan and that other
reliable eyewitnesses reported that such caps were made of deerskin. Inō insisted
that such details were important for ethnologists,46 who were set apart from men
of practical affairs, such as Hirano, for their ability to construct evidence-based
taxonomies, which would, in turn, provide a scientific foundation for colonial rule.
214     Indigenous Modernity

In addition to interviews with indigenes or conversations with rural policemen

who themselves had daily contact with mountain dwellers, Inō relied upon Qing
documents, as his January 1896 report to the Journal of the Tokyo Anthropological
Society reveals. In this issue, Inō referenced an unnamed Chinese record to give
population estimates for Dakekan. Inō also relayed a report on methods of alco-
hol brewing and noted that the process might be further clarified by looking at a
corroborating passage in the Danshui Prefecture “Local Customs” section of the
old Taiwan gazetteers.47 In later reports, he would cite specific Qing records as
evidence for a number of local practices.48
Timothy Tsu has argued that men like Inō represented a new force in Japanese
Sinology. During the Tokugawa period, Japanese scholars read Chinese texts to
understand the principles of statecraft, history, and family relations in order to
become better human beings and governors. It was not uncommon for Chinese
scholars to visit Japanese literary societies. They were lavishly wined and dined as
they shared their expertise with provincial aficionados in Japan. Modern social-
scientific approaches to data collection changed all of this. To govern Taiwan more
efficiently, Japanese officials mined Chinese court records, land deeds, population
registers, and treatises for evidence, not for theoretical insight or moral guidance. As
Peter Pels put it in his discussion about a similar transition in colonized India, the
colonial social scientists entered into a direct relationship with texts, individuals, and
populations in order to correct the superstitious, backward, and parochial outlooks
of local mandarins. Thus, Inō would be an energetic compiler, editor, and translator
of Qing documents, but he remained committed to the idea that these documents
supplied only the raw materials for scientific analysis, subject to his corrections.49
In the March 1896 issue of the Journal, Inō reproduced documentary evidence
from the Qing archives to expose its shufan/shengfan nomenclature as anthropologi-
cally bankrupt. He documented how certain villages were classified in older Qing
records as shengfan but reclassified as shufan in later gazetteers. Sometimes the two
terms were used interchangeably with terms such as yefan, another word for “savage”
with no particular ethnic referent. Inō also turned a critical eye to fellow country-
man Ueno Sen’ichi, whose 1895 report for the Japanese army was summarized above.
Inō chided Ueno for mistakenly using the toponym peipoban (plains tribe) as an
ethnonym. If we recall Inō’s use of the word koyū (particular) in his early declaration
of “needs and opportunities” for Taiwan research, we can understand his objections.
In the context of race studies as construed by Inō, koyū referred to the distinguish-
ing characteristics of a tribe that fixed its essential nature. Terms like “unaccultur-
ated” (shengfan) and “plains tribe” described accidental characteristics shared by any
number of tribes, villages, and confederations throughout the Qing empire.
To go beyond poking holes in other accounts, Inō had no choice but to follow the
paths of Ueno, Hashiguchi, Kawano, and Hirano to gather information from and
about peoples whose languages had yet to be recorded in dictionaries. One of his
early informants, the woman who would become known to Japanese simply as Ai,
 The Geobodies within a Geobody    215

was born around 1880 to chief (shuchō) Pira Omin of Kimunajii (near Dakekan).
Meiji anthropologists prefaced much of their work in the language of hardship,
solitary travel, danger, and exploration. Undoubtedly, such posturing reflected
aspects of their experience. On the other hand—and quite sensibly, in a new colony
among unknown tribes—these ethnologists, like the merchants and rural officials
discussed in the previous chapters of this book, initially turned to the most willing
and available informants. And these people tended to congregate around old Qing
Pacification Offices and trading entrepôts, Dakekan being the most prominent.
Inō’s informant Ai was seven years old when Governor Liu Mingchuan’s
Pacification Office was established in Dakekan in 1886. A Christian missionary
named Chen Cunxin arrived circa 1890. Chen soon adopted Ai, which explains
her facility in the Minnan dialect of Chinese by the time Inō showed up. As we
have seen, such adoptions were common enough along the savage border, with
the aim of each side to gain access to important suppliers, protectors, and patrons
from across the border. Inō Kanori paid Chen and Ai a visit in January 1896. He
took Ai on as a Japanese-language student while in turn interviewing her about
her childhood in Kimunajii and taking rudimentary lessons in the Dakekan vari-
ant of Atayalic. This relationship continued for over a year, until Ai’s tragic death
on March 19, 1897, from a fever. According to Inō, Ai was making rapid progress
in Japanese before she passed away. In his memorial article in the Journal of the
Tokyo Anthropological Society, Inō characterized Ai as a shining example of Atayal
people’s “potential for civilization.” He also acknowledged his debt to her as a native
informant on language, history, and customs.50 So, before Inō barely had time to
situate himself in the new colony, he had already relied upon the services of several
bicultural Atayal women, including Iwan, Awai, and Ai, to make a beginning on
filling in the blank ethnological spaces on the old Qing map.
Although Inō considered himself to be a race scientist, his narrowly anthro­
pometric investigations were peripheral to his taxonomic project. In contrast to
the grim, serious, and consequential anthropometry used by United States anthro-
pologists to lend academic legitimacy to the “one-drop rule” or to “blood-quantum”
politics,51 Inō’s quantitative forays are incidental to his larger body of work or even
to individual reports. For example, to determine Ai’s racial status, Inō drew up a list
of body parts and compared Ai’s features to those of the “typical Japanese girl.” Inō’s
method here was a far cry from Mori Ushinosuke’s subsequent use of standardized
forms, based on the famous British Ethnological Society Notes and Queries forms.52
The major headings on Inō’s list were even inconsistent across studies. In Ai’s case, he
listed skin color, bodily hair and coiffure, eye shape, and “other parts of the face” as
important constituents of racial identification. For each feature, Inō used a Japanese
woman as the proxy for Asian and marked “similar” or “not” (if “not,” he explained
how a feature was different). Under the column bodily hair, Inō marked “similar,”
but under mouth, he wrote that Ai’s was “comparatively bigger.” For all of his efforts,
however, Inō believed this method of racial classification was inconclusive. For one,
216     Indigenous Modernity

Inō’s sample was much too small. But even if this problem were ignored, Inō wrote,
some of Ai’s physical features were Asian, while others were Malayan, so there was
no way to decide her racial affiliation based on his anthropometric investigations.
In the end, Inō believed linguistic evidence was more definitive, because Ai’s spo-
ken bango (Atayal) had the flat tone (compared to the rising and falling tones of
Chinese) and harsh consonants characteristic of Malay languages.53
Inō freely mixed cultural and somatic criteria in this case to arrive at a fore-
gone conclusion: Ai was not Chinese or Japanese but was instead a member of the
Malay race. This determination linked her ethnographically to the peoples of the
Philippine Islands and the Malay Peninsula, while extracting her from the fron-
tier political economy of Dakekan. In the years to come, Inō would labor to rep-
resent the “pure Malay” indigene graphically—an immense project to which we
will return below. In a subsequent report based on a larger sample, Inō conducted
anthropometric studies of the inhabitants of Wulai. Almost as if to invite the criti-
cism of later scholars, Inō adopted different categories for measurement for his
Wulai subjects than the ones he used for Ai, rendering a scientific comparison
impossible. Inō had twenty-seven Wulai residents to work with, which led him to
state that his statistical sample was too small to be meaningful.54
On the other hand, this March 1896 series of interviews, which took place in
Taipei, changed the course of Japanese representational practices vis-à-vis indi-
genes forever and has cast a long shadow on current practices in self-representa-
tion. It was in Taipei that Inō determined that his Wulai informants were similar
enough in dress, ornament, language, and comportment to his Dakekan infor-
mants to be classified as coethnics under the rubric Atayal.

W U L A I , QU C H I , A N D T H E ATAYA L I Z AT IO N O F T H E
N O RT H E R N T R I B E S

Inō’s March 1896 interviews with Wulai residents not only provided him with impor-
tant clues for the classification of Taiwanese ethnic groups but also set the stage for
the production of photographs by activating networks of ethnologists, military men,
camphor capitalists, and headmen from the settlements along the Xindian River.
Inō Kanori’s travel notes from 1897 and his 1900 ethnological report list Jiku Shō
Min as his local informant for Quchi. Jiku came to Taiwan with the Imperial Guard
in 1895. In 1896, under orders from first governor-general Kabayama Sukenori,
Jiku made his way to Quchi to “pacify the locals.”55 Subsequently, he married into
a leading Quchi family to establish a beachhead for Japanese logging concerns in
conjunction with camphor entrepreneur Dogura Ryūjirō. As an adopted outsider,
Jiku took the title Watan Karaho to reckon himself a true local chief (tōmoku).56
Because Inō referred to Jiku in his famous 1900 compendium as his guide to Quchi,
we can surmise that Jiku himself facilitated the March 1896 exchanges between Inō
Kanori and the Wulai contingent in Taipei recounted above.
 The Geobodies within a Geobody    217

A little over a year after the interviews with the twenty-seven Wulai residents
in Taipei, Inō and Awano Dennojō formed an expedition crew to begin a 192-day
ethnographic survey tour of Taiwan by order of the Bureau of Education. They
began by recruiting two Han Taiwanese aborigine-language interpreters, named
Kōan and Aki, both inhabitants of the interior for over twenty years. Both were
married to indigenous women. Inō could hardly distinguish Aki from an Atayal
man; he lacked a queue, wore Atayal clothing, wielded a machete, and traveled
barefoot. Inō’s party also hired Awai, presumably for interpreting, along with her
two children, in the village of Rahao.57 This Awai may have been the same inter-
preter who acted as Tashiro Antei’s informant in the fall of 1895.
Jiku Shō Min organized the purchase of gifts and supplies, as well as the hiring
of interpreters. When the party reached Wulai, just south of Taipei, Inō and Awano
were welcomed by a young bantei (brave) named Watosinai, whom Inō had met
previously (probably in Taipei).58 Watan Yūra, whose name shows up in the manu-
script records of Japanese rural administration as a representative of Wulai and
in Inō’s list of informants from the March 1896 interviews, was also present. Inō
did not return to Wulai/Quchi after the May 1897 meeting, but Mori Ushinosuke,
again under the aegis of Jiku Shō Min, rekindled these ethnologist-informant rela-
tionships in 1902..
Inō’s island-wide survey concluded in December 1897. At this time, we should
recall, etchings based on photographs of the Jiaobanshan embassy and a few
actual photographs had been circulating in Taiwan and Japan as the most widely
distributed illustrations of the island’s peoples. On April 23, 1898, Inō Kanori
made his first major intervention to controvert these ethnologically vague and, in
his view, irresponsible portrayals. Shifting the ground toward a pluralist model,
Inō unveiled his newly devised taxonomy of Taiwan Indigenous Peoples with a
photographic montage composed of representative “types” at an exposition in
Danshui.59 On August 16, 1898, Inō’s montage appeared on the cover of the first
issue of Banjō kenkyūkaishi (Journal of the society for research on aborigine con-
ditions).60 The same montage was reproduced for a general readership on August
5, 1899, in Taiwan meisho shashinchō (A photo album of Taiwan’s famous sites).61
Inō’s montage, prefiguring the map he would publish in 1900, depicted Taiwan’s
ethnic groups as commensurate units, suggesting ontological parity for each. Unlike
the map, however, the montage could not provide visual consistency across ethnic
groups. Some of the photos are studio portraits, others were shot in situ. Some dis-
played noted cultural markers, others did little to distinguish their subjects from
other Taiwanese. For example, the “tattooed face savage/Atayal” exemplar wears no
jewelry, has no visible tattoos, and is carrying an imported weapon. This man, who
hailed from Quchi, was probably familiar to Inō. In the montage, he is referred to as
the “tattooed face savage” (keimenban),62 which is also glossed in furigana as Atayal.
The new ethnonym Atayal redefined the “savages” in terms of unique attributes
manifested in language and material culture, neither connected to nor dependent
218     Indigenous Modernity

upon their relationship to a Japanese or Qing imperial center. In other words, the
ethnonym Atayal asserted a presence. But in asserting an ethnic presence, Inō also
effaced the individual’s identity. In contrast with the Jiaobanshan embassy sitters,
whose names are easy to discover, clues regarding the identity of the man known
simply as “tattooed face savage” remain fugitive.
A glass plate of Inō’s “tattooed face savage” is held at the U.S. Library of Congress
as part of the Bain News Service photographs,63 along with a portrait of two well-
armed “chiefs,” two photographs of headhunters with fresh heads, and a picture of
a skull shelf.64 The Bain negative is captioned “typical fighting man of the head-
hunters, Formosa.”
The Japanese word for “savage” is also burned into the plate. The other chiefs
in the Bain News Service collection are also referred to as savages, while the other
three Bain negatives provide graphic evidence of seiban barbarity by displaying
severed heads or skulls.65 The term Atayal is not attached to any of these photos.
Other pictures of the armed man from Inō’s 1898 montage have been reproduced
in the 1999 book Kanori shozō Taiwan Genjūmin shashinshū (Images of Taiwan
indigenous peoples from the Inō Kanori archive). One shows the man seated with
a rifle and a Japanese flag; the other shows him directing soldiers with weapons and
again carrying a Japanese flag. The editors of this volume argue, convincingly, that
he probably fought on the side of the Japanese state against Han rebels in late 1895
and 1896, as an irregular in the employ of Japanese station chiefs stationed near
Quchi.66 His image was also reproduced on cabinet cards, a popular photographic
medium from the 1860s through the 1890s.67
In contrast to the ethnically ambiguous Atayal photographs in the 1898 mon-
tage, Inō’s canonical 1900 textual description of the Atayal describes them as face-
tattooed peoples, richly adorned in geometrically carved earrings and accessories
that included buttons, brass, shiny metal, and colored threads.68 To reconcile the
difference between Inō’s textual/cartographic representations of bounded, discrete
ethnic blocks among Taiwan’s non-Han population and the motley assemblage
of hybrid-themed photographs in the 1898 montage, Inō commissioned a color
painting for the 1900 Paris Universal Exposition.69
The 1900 painting adds head-feathers, necklaces, hats, and earrings to each rep-
resentative “savage type” to accentuate cultural differences among them.70 The pro-
cess of fabrication is even more evident in the similarly augmented illustration for
the 1907 Tokyo Industrial Exhibition, also curated by Inō.71 In the 1900 and 1907
exposition montages, the gun has been airbrushed out of the Atayal man’s hand,
large bamboo earrings have been painted in, and forehead and chin tattoos have
been darkened or added. In the 1898 montage, the Bain News Service photograph,
and the cabinet card, the facial tattoo on the man’s forehead is invisible or barely
visible, and the chin tattoo is obscured by shadow. An examination under high
magnification of the high-resolution negative stored in the Library of Congress
reveals a very faint forehead tattoo and rules out a chest tattoo. The photos of
alternative poses from the same session show no evidence of visible facial or chest
 The Geobodies within a Geobody    219

tattoos. Yet the reproduction in James Davidson’s 1903 book (for which Inō Kanori
was the ethnological advisor) exhibits pronounced forehead, chin, and chest tat-
toos, the result of doctoring.72
In the “ethnicized” 1900 and 1907 paintings, the female Atayal “type” shed her
Chinese-style upper garment for an Atayal sleeveless vest, decorative chest embroi-
dery, and a striped handwoven cape. The reornamented Atayal woman wears eth-
nologically correct earrings and necklaces, as well. Whereas the ethnicized male
Atayal model was transformed with a paintbrush or pencil, his female counterpart
was removed from the 1898 photomontage and replaced by a different sitter in the
1900 and 1907 composite portraits.73
As curator for the 1900, 1903, 1904, and 1907 expositions, Inō Kanori was
charged with educating the public about the non-Chinese population of Taiwan.
In the course of preparing the visual component of his exhibits, Inō turned “sav-
ages” into Atayal and other nondescript males into Yami and Paiwan by embel-
lishing visual documents. It would be accurate to view these fabrications, in the
context of colonial census taking, mapmaking, and museum curation,74 as exer-
cises in top-down ethnogenesis.

M A K I N G P HO T O G R A P H S I N Q U C H I : V I SUA L C U LT U R E
A N D T O P- D OW N E T H N O G E N E SI S

James C. Scott’s characterization of a parallel project in British Burma suggests that

Inō’s difficulties were not unique. As was the case with British officials, Japanese
agents in upland Taiwan found even the act of naming tribes to be fraught with
problems because of the hybridity they regularly encountered. Of these turn-of-
the-century classificatory schemes, Scott writes that “a major reason why trait-
based designations of ethnic or tribal identity fail utterly to make sense of actual
affiliations is precisely that hill groups themselves, as manpower systems, absorbed
whomever they could. This absorptive capacity led to great cultural diversity
within hill societies.”75 These multiform, ethnically diverse “manpower-absorbing”
societies were precisely the kind of formations Japanese officials encountered on
the savage border circa 1900. In his study of the “Nanzhuang incident” of July 1902,
Antonio C. Tavares observes that Saisiyat tribes under chieftains in the late Qing
period were conglomerates of Han, non-Han, and hybrid actors, with hierarchies
of power and wealth that ran counter to entrenched East Asian taxonomic spec-
tra composed of descending markers of value from central and paramount hua
(civilized) to peripheral and lowly yi (barbaric).76 Mori’s 1902 description of Watan
Yūra’s extended family (whom Inō met in May 1897, pictured in figures 18 and 19)
illustrates how, like the Nanzhuang operators, Wulai headmen adopted outsiders
to create hybrid formations, to the chagrin of ethnic cartographers like Inō:
Wulai settlement’s paramount chief Watan Yūra’s eldest daughter Pazzeh and his ad-
opted son [yōshi] are both about twenty years of age. This couple is the poken, mean-
ing that if the line of succession is maintained, they will become chiefs . . .77
220     Indigenous Modernity

Figure 47. Photograph of Mori Ushinosuke, Japanese officers, and Truku headmen, 1910. Mori
is facing away from the viewer and wears the dimpled hat. Courtesy of the Rupnow C
­ ollection.

To the right of Watan Yūra is standing a plains aborigine [peipozoku] (cooked

barbarian [jukuban]) male without tattoos. About twenty years ago, it is said that he
entered this settlement/village and became a savage brave [bantei] . . . To the left are
four wives of savages [banjin]. They are wearing Chinese-style blouses and gaiters.78

The networks that emerged through intermarriage, business dealings, armed con-
flict, and research among Japanese officials, ethnologists, merchants, and leading
families from Quchi—we saw them in action during Inō’s 1897 and 1898 infor-
mation-gathering exercises—provided the setting for Mori Ushinosuke to make
qualitatively superior photographs in the ethnographic genre. In contrast to our
earliest examples of ideal-typical “head-hunting chiefs” photographed in Taipei
studios (figure 21), the photographs made in Quchi during the establishment of
the guardline were shot at or near the residences of the subjects (figure 19). The
increased number of poses and more intricate stagecraft evident in the Quchi pho-
tographs suggest a heightened degree of familiarity between photographers and
subjects.79 Mori himself was rarely photographed. One photograph was taken circa
1914 at the end of the Truku campaigns; Mori is wearing the hat, with his back
toward the viewer (see figure 47).
In one recorded instance of Mori Ushinosuke’s networking prowess, on January
28, 1903, Mori brought Rimogan resident Marai Watan and his wife Yūgai Watan
to Taipei to view a local theater production of Ishiyama Gunki, meet the staff of the
Taiwan Daily News, and see the sights. According to the short newspaper write-up,
 The Geobodies within a Geobody    221

the Rimogan couple “came down from the mountains to visit Quchi town once or
twice a year, but had never been to Taipei.”80
Since Marai’s portrait was displayed at the Osaka Exposition, which opened on
March 1, 1903, it is reasonable to assume that Mori photographed this couple just
before or after their visit to the capital city in January. Their most widely published
portrait was shot in April 1903. It was included in a number of official publica-
tions and formed the basis for at least two picture postcard designs (see figure 59).
Postcards were also generated from different poses of the same couple, indicating the
ideological and commercial appeal of their likenesses as rendered by Mori’s camera.81
Instead of using painting or drawing to “Atayalize” Yūgai, Mori was able to
accomplish her racialization by requesting different poses. While Yūgai’s anthropo-
metric portrait shows her Chinese-style upper garments clearly, her more commer-
cially reproduced pose was staged to present a more “authentic” Yūgai.82 In the latter
photograph, she dons a cape of local design, which conceals her Chinese clothing.
This de-Sinicized version of Yūgai’s portrait showed up in various picture postcards,
in a 1932 book by ethnographer Koizumi Tetsu, in a collection of government sta-
tistics, and on the sleeve for a set of Taiwan Indigenous Peoples picture postcards.83
During the over 150 days that it remained open, the Osaka National Industrial
Exposition of 1903 drew over five million visitors.84 The Taiwan Government-
General lobbied intensively for exposition space under the energetic leadership of
Gotō Shinpei. For Gotō, the Osaka Exposition presented an opportunity to educate
Japanese on the home islands about the strategic importance of Taiwan, its potential
economic benefits, and the cultural and culinary attractions of the island. A prevail-
ing home-islander image of Taiwan, which Gotō hoped to dispel, was that of a sav-
age place where demonic tribes practiced cannibalism.85 Against this backdrop, the
“tattooed face savage” photos of men with large guns glaring at cameras in studio
settings were hardly appropriate. Gotō’s opportunistic aversion to savagery dove-
tailed with Inō Kanori’s preference for cultural themes in evidence in the 1900 Paris
Exposition painting. Thanks to Mori’s new photographs from the environs upriver
from Quchi, Inō was able to accommodate both desiderata without recourse to
another commissioned painting. He simply exhibited Mori’s photographs.86
Mori shot portraits of Pazzeh Watan, daughter of the headman Watan Yūra,
near Wulai in February 1903, just after his January trip to Taipei with Marai and
Yūgai. Pazzeh is seated to the viewer’s far left in figure 19. This group portrait shows
her wearing Atayal leggings with a Chinese blouse and headdress—a common
sight along the old Qing savage border. This photograph appeared in a Taiwan
Government-General publication of 1911, in the general-circulation magazine
Taiyō in 1917, and in Mori’s 1915 and 1918 ethnological picture albums, as well as
on Shōwa-period picture postcards.87 It was also picked up for syndication in the
United States.88 For a different pose, Pazzeh put on bamboo earrings, locally woven
fabrics, necklaces, and a diamond-shaped breast cover festooned with white but-
tons for her anthropometric portrait. Her bust and profile from this shot appeared
222     Indigenous Modernity

in dozens of formats in Japanese and foreign publications and continues to be

reproduced to this day. The differences in Pazzeh’s dress and ornamentation in the
two portraits from February 1903 recall Inō’s alterations of the 1898 montage dis-
cussed above: features considered to be signs of autochthony and distinction were
accentuated, while any other visible signs of hybridity were obscured.

T H E V I EW F R OM T H E M E T R O P O L E : C O N SUM E R
D E M A N D M E E T S C O L O N IA L P O L IC Y

Photographs, as slices of time that freeze events, cannot narrate themselves with-
out the help of external markers, such as captioning or juxtaposition with other
photographs.89 In the case of Mori Ushinosuke’s Quchi-area photographs, cap-
tions told various stories. The text for some portraits supported Mori’s ethnicizing,
pluralizing ethos. An early 1900s postcard of Pazzeh applied the caption “Shinkō
[administrative district], Urai-sha [place name] Woman, Taiyal Tribe,” sans termi-
nology for “savage,” for example.90
Mori’s and Inō’s original intentions were more often violated once they lost edi-
torial control, because policy was moving in a different direction than representa-
tional practices favored by Japan’s ethnologists.91 In December 1902, only a couple
of months before Mori took Marai’s and Pazzeh’s photos for the Osaka Exposition,
councillor Mochiji Rokusaburō, the brains behind Gotō Shinpei’s aborigine pol-
icy, issued his famous “opinion paper on the aborigine problem,” which averred
that “sociologically speaking, they are indeed human beings (jinrui), but looked at
from the viewpoint of international law, they resemble animals (dōbutsu no gotoki
mono).”92
As the scorched earth campaigns were commencing, in June 1904, parlia-
mentarian Takekoshi Yosaburō arrived in Taiwan to write a progress report.93 In
September 1905, Takekoshi published Taiwan tōchi shi, his encomium to Japanese
rule based on his 1904 visit; he included profile pictures of Watan Yūra’s daughter
Pazzeh Watan, who figured prominently in Mori’s Quchi photographs, as a name-
less Atayaru-zoku jo (Atayal tribe female) next to exemplars from each officially
recognized tribe. Most of these photographs were also exhibited at the 1903 Osaka
Exposition. These ethnically labeled profiles and frontal shots, with blank back-
grounds in the anthropometric genre, were interleaved with photos of Japanese
military maneuvers and Gotō’s inspection tour of camphor forests. In this con-
text, the ethnicized portraits of Marai and Pazzeh appeared as emblems of recal-
citrant Atayals, marked as an expendable race. In contrast to the photographed
Jiaobanshan emissaries who arrived in Taipei in September 1895 to initiate rela-
tions between the government-general and the cismontane Dakekan tribes, the
names of Watan, Marai, Pazzeh, and Yūgai were not mentioned in newspaper
accounts of guardline movements in and around Quchi. And by the time their
photos emerged in the flood of picture albums of Taiwan indigenes published in
 The Geobodies within a Geobody    223

the 1910s, Japanese publicists and anthropologists had ceased attaching personal
names to photographs of indigenes altogether.
During the era of the guardline, 1903–1915 (see chapter 1), the rhetoric on Atayal
savagery was amplified and reinforced by the captions for various contemporary
postcards that identify Pazzeh Watan as a seiban, bellicose seiban, or (incorrectly)
as a member of the Taroko seiban, the most militarily resistant subgroup of Atayal
peoples.94 This view received the imprimatur of the Taiwan Government-General,
which issued its second set of commemorative postcards on October 15, 1905. One
featured a studio portrait of armed Atayal men captioned as “head hunters” in
English and labeled “Taiwan savages” (Taiwan banjin) in Japanese.95
The government-general would issue three more commemorative postcards
featuring indigenes during the guardline period. Martial themes were prominent.
The 1908 set, for the thirteenth anniversary of colonial rule, commemorated the
Stone Gate Battle of 1874 with a colorful painting of the battlefield and a tomb-
stone to the fallen warriors.96 A paired card reprised an 1874 photograph of Saigō
Tsugumichi surrounded by his own soldiers and several emissaries from Paiwan
villages in Hengchun (figure 8).
These cards were issued two years into the administration of Governor-General
Sakuma, who himself was the hero of this 1874 battle. The fourteenth-anniversary
card (1909), issued on the eve of Sakuma’s “five-year plan to conquer the aborigines”
(see chapter 2), features the governor-general himself. Whereas his fellow veteran of
the 1874 campaign Saigō sits among his Paiwan emissaries in the 1908 commemora-
tive series, Sakuma is elevated high above the more than forty Atayal and Saisiyat
men and women gathered at his residence in the photo from the 1909 series. Sakuma
is placed even higher than Prince Kan’in, who was conducting a royal tour in 1908.97
To look at the body language and positioning of Paiwan and Jiaobanshan sitters
(figure 37) in their respective commemorative photographs, one might surmise
that they were guests, or at least subordinate allies, at functions presided over by
Saigō or Kabayama (in 1874 and 1895, respectively). But the fourteenth-anniver-
sary photograph of Governor-General Sakuma allows for no such reading. Here,
even the most highly placed indigenous person’s head is even with Sakuma’s feet,
while most of them are huddled on the ground in positions of abjection. In fact,
the 1909 postcard can be read as the wish fulfillment of the Fūzoku gahō artist who
imagined Kabayama in a similarly commanding position over the Jiaobanshan
emissaries in 1895 (figure 36).
The last of the guardline-period TGG postcards with indigenous themes was
transitional. It struck a balance between a pluralist sensibility and the discourse on
savagery (see figure 48). Issued in 1911, at the height of the war against Atayal peo-
ples in and around the tramontane Dakekan tribes, it included a Mori Ushinosuke
photograph of an Atayal dwelling and attached granary.98 This scene was common
in anthropological reports of the time. The card also featured a well-circulated
photograph of Koalut dancers from Hengchun.99 The photographs were innocuous
224     Indigenous Modernity

Figure 48. Official commemorative postcard depicting indigenous customs, 1911 [“Taiwan
Aborigines Dancing” and “Taiwan Aborigine Dwelling with Granary”], ip1441, East Asia Image
Collection, Lafayette College, Easton, PA, accessed May 13, 2016, http://digital.lafayette.edu/
collections/eastasia/imperial-postcards/ip1441.

enough, but they were framed by a heavily embossed decorative element composed
of a machete, a head-hunting backsack, a gun, some spears, and two arrows, next
to a trademark Atayal wisteria-woven cap. The composition therefore connoted
menace, if not danger. At the same time, it presupposed its audience’s familiarity
with particular cultural repertoires of the Atayal and Paiwan indigenous ethnic
groups, so that viewers could read the embossed icons as symbols of indigenous
bellicosity. Assuming that these mass-produced, high-quality commemorative
sets were designed with great deliberation, one can conclude that the pluralistic
terminology and iconography so laboriously engineered by Inō, Mori, and a cast
of hundreds was gaining traction at the apex of administration in Taiwan by the
1910s, if only incrementally. The photo captions in figure 48 themselves are mute
regarding ethnic designation and retailed generic words for “savage” instead.
The predominance of head-hunting and savage tropes in the government-­
produced postcard sets and in the state-produced and commercial punitive expe-
dition albums reflects the power of metropolitan policy directives and consumer
tastes to shape the meanings attached to widely circulated photographs of Taiwan’s
indigenous people. With adroit captioning and the creative deployment of design
elements, Tokyo and Taipei publishers and editors yoked print culture to TGG
policy and Japanese consumer tastes. At the same time, ethnologists on exposi-
tion-planning committees, editorial boards of academic journals, and lower-level
 The Geobodies within a Geobody    225

corners of the colonial bureaucracy launched counterhegemonic arguments with

artifacts, photographs, drawings, and texts. To be sure, advocates of a pluralist
agenda such as Inō Kanori and Mori Ushinosuke were largely motivated by con-
cerns of professional advancement in the emergent field of anthropology. They
also had an eye for the survival and growth of the discipline itself. The ethnologists
pursued these ambitions as they extolled Japan’s modernity in the global arena of
scientific publication. Whether they supported the discourse on savagery or eth-
nicity, the poses, set designs, captioning, and layout decisions that shaped visual
culture surrounding indigenous peoples responded to dictates largely divorced
from the specific locations where sitters, photographers, and stage managers came
together to expose wet- or dry-plate slides to light in the presence of living beings
and the environment.
To ground these images in their historically specific contexts of creation, we
must return to the Atayal-dominated regions of Quchi and Jiaobanshan—each a
fecund hearth of pluralist and savage-trope imagery in the early twentieth century.
If indigenes themselves had anything to say about how they were being repre-
sented in Japanese photography, they exercised agency at these sites.

T H E S AVAG E - B O R D E R M I SE - E N - S C È N E O F C O L O N IA L
P HO T O G R A P H Y I N TA I WA N

In May 1903, two months after the opening of the Osaka Exposition and the
first public display of Watan Yūra’s family portrait, Japanese authorities impris-
oned nine Quchi men. They were used as hostages to secure the cooperation of
other Quchi leaders to build a cordon sanitaire to separate camphor fields and
tax-paying settlements from Atayal who were attacking camphor harvesters. This
stratagem met with some success. During the years 1904 and 1905, the local knowl-
edge, military prowess, and political agility of Quchi residents made it possible
for the Taiwan Government-General to establish the toehold it needed to build
its guardline across the most contested areas just south of Taipei. Thereafter, from
its Quchi base, the government-general built several hundred miles of guardline
across northern Taiwan.100
The role of Quchi auxiliaries in the advancement of the guardline is attested
by two photographs produced there circa 1905. One depicts an extended family,
consisting of eleven men, four women, and two children (see figure 49); the other
depicts seven armed men (see figure 50), who also posed for the family portrait
in figure 49. The brush-written caption to the extended family portrait places the
scene at the “Guardline Control Station,” meaning that it was taken in 1904 at
the earliest. It was reproduced in the September 2, 1905, Taiwan nichinichi shinpō
(Taiwan daily news). Therefore, the men, women, and children captioned as
“Wulai Savages” in the newspaper were photographed about two years after Mori’s
famous Quchi photographs were shot in early 1903.
226     Indigenous Modernity

Figure 49. Men and women along the Quchi guardline, ca. 1904. Courtesy of the National
Taiwan Library (Taipei). Unpublished photograph album, Taiwan shashin (National Taiwan
Library Document 0748–71, plate 82).

The “family snapshot” (figure 49) appeared on an early 1900s postcard, in a

few Japanese-language books, in several foreign publications, and in geography
textbooks.101 A comparison of the careers of the 1903 (figure 19) and 1905 Wulai
(figure 49) group portraits suggests that a threshold was crossed in 1905 with the
militarization of relations between the Japanese government and the Quchi-area
tribes. In each group portrait, four women crouch to the viewer’s left, dressed in
a combination of Chinese and Atayal clothing. Standing and squatting men with
pipes and diamond-shaped breast ornaments are the center of attention. Children
wrapped in trademark striped Atayal capes sit in the foreground of each portrait,
rounding out the cast of characters. However, the structure in the background
of the 1905 photograph (figure 49) is new. This distinctive feature is a guardline
control station. This military installation is not identified in the published ver-
sions of the photo, although it is indicated with a hand-brushed caption in an
unpublished album housed in the former TGG library in Taipei. Therefore, editors
and consumers likely viewed figure 49 as a family portrait rather than as the war
photograph that it was.
In the seven Japanese-language publications that include a photo from the
1905 shoot (figures 49 and 50), only the mixed gender and age “family portrait”
 The Geobodies within a Geobody    227

Figure 50. Japan’s Atayal allies along the Quchi guardline, ca. 1904. Courtesy of the National
Taiwan Library (Taipei). Unpublished photograph album, Taiwan shashin (National Taiwan
Library Document 0748–71, plate 102).

appears. In the 1905 and 1906 versions, the sitters are referred to as “Urai [Chinese:
Wulai] tribe” people, but in the 1907, 1912, 1925, 1928, and 1933 versions, they are
captioned as members of the Atayal tribe. Thus, in Japanese-language publica-
tions, the 1905 photograph of a large group in front of an unacknowledged guard
station (figure 49) was captioned to ethnicize the Atayals. This framing operation
was simultaneously culture-affirming and detemporalizing vis-à-vis the photo-
graph’s sitters.
In contrast, of the four English-language publications (two by Japanese pub-
lishers for foreign consumption, and two by foreigners) that picked up the 1905
shots, three adopted more explicitly martial photos of the armed men in the road
(figure 50). These publications also used the term Atayal to describe the sitters—in
1910, 1916, and 1923. Based on this analysis, we can say that somewhere between
1906 and 1907, the term Atayal began to catch on with publishers, even for pho-
tographs that did not obviously illustrate the unique folkways of the sitters but
instead lumped them together with the world’s garden-variety savages with guns.
In sharp contrast to the shot of seven armed men blocking the road in the
1905 series, the spin-off photographs from the 1903 portrait (figure 19) eschew
martial themes in favor of domestic scenes of cloth production and residential
228     Indigenous Modernity

architecture. Mori took several photographs, besides the family portrait of Watan
Yūra, during the same session. These scenes from daily life circulated widely as
postcards and illustrations for a variety of publications.102
Mori Ushinosuke’s Quchi-area photos, comprising subjects from Wulai and
Rimogan, were reproduced in at least forty different postcard designs and in doz-
ens of newspapers, magazines, journals, and other media in subsequent years.103
While Mori the ethnologist may have reached dozens of anthropologists through
his reports in the Journal of the Tokyo Anthropological Society or thousands of
readers through his serialized travelogues in Taiwan newspapers, the audience for
his photographs of representative Atayal culture bearers was global.
There are many reasons why the period 1903–05 was propitious for mass-circu-
lation Atayal photographs. First, by the time Mori arrived in late 1902, a network of
acquaintances connected him to several sitters and photographic sites. Therefore,
as the multiple poses and variations in the Rimogan shots indicate, Mori was able
to work in a better climate for photographic production than his predecessors Torii
and Inō, who both operated on tight schedules among unfamiliar peoples during
their forced-march surveys of the late 1890s. The state of photographic technol-
ogy at the time demanded long exposure times and the participation of willing
sitters to produce high-resolution, interesting, and well-composed photographs.
During the period 1902–05, many Quchi people were on the government’s payroll
or employed by Japanese industry. Quchi residents may have charged money to
be photographed, as well (a common practice a little later on), or they may have
been favorably disposed toward their Japanese trade partners, employers, acquain-
tances, and in-laws (such as men like Jiku Shō Min).
Second, just after Mori’s prints from Quchi began to arrive in Osaka for the
exposition and in the Tokyo University anthropology lab for analysis, a picture
postcard boom was ignited by the Russo-Japanese War. The clearest and best
extant copies of the Quchi photos, the Wulai guardline photo, and even one copy
of the Jiaobanshan emissary photo are in the form of collotype prints on sturdy
postcard stock. In the 1903–07 period, several postcard producers availed them-
selves of Mori’s photographs and other official sources to propagate these high-
quality reproductions. Textbook publishers raided this storehouse to represent
indigenous Taiwan to every grade-school student in Japan.
However, not long after these iconic individual and group portraits were
shot between late 1902 and early 1903, local Quchi men overran Taiwan’s first
hydroelectric power station and killed fifteen workers and guards in 1905.104 The
razed power station was built on the Xindian River near Quchi village under
the direction of Dogura Ryūjirō.105 Dogura was an intimate of Jiku Shō Min and
Mori Ushinosuke. He had also obtained the rights to develop hundreds of hect-
ares of Atayal land around Quchi for camphor production in 1899. The termi-
nation of official photographic activity in Quchi coincides with these events,
unsurprisingly.
 The Geobodies within a Geobody    229

The reprisals for the 1905 incident must have been harsh, though details are hard
to come by. We know that within a very short time, Wulai was transformed into a
hub of Japanese assimilation policies and hot-springs tourism. In December 1908,
travel writer and explorer Mary T. S. Schaffer and three Canadian companions vis-
ited Wulai to get a glimpse of the world-renowned “head-hunters of Formosa.”
Like many a visitor after her, Schaffer wrote excitedly about the prospects of meet-
ing a savage head taker and even purchased a postcard of a freshly taken head
held by its exultant killers. But when she actually met Atayal people on the trail,
she found them to be friendly and happy to pose for photographs in exchange for
silver coins. Included in Schaffer’s collection of magic-lantern slides was a color-
tinted profile of Pazzeh Watan, the daughter of Watan Yūra (the woman in figure
19, to the viewer’s far left).106 U.S. Consul to Taiwan James Davidson also used col-
orized glass-lantern slide images of Mori’s sitters for his public presentations.107 In
fact, all of Schaffer’s and Davidson’s photographs were created before 1905, when
Wulai was still ethnologically available, so to speak.
The period 1903–05, then, was a window in time. Mori arrived in Wulai late
enough to benefit from the networks established by Inō, Jiku, Dogura, and Watan
Yūra, and from recent advances in camera technology, but early enough to ride
the wave of the Russo-Japanese War postcard boom. The luck of good timing also
allowed Mori to make photographs before the 1905 attacks on the power station
flipped over the chessboard of frontier diplomacy along the Xindian River south
of Taipei.

P U N I T I V E E X P E D I T IO N P HO T O A L BUM S

The period between the 1905 Quchi incident and the establishment of an ethnic
tourism industry in Jiaobanshan in the early 1920s marks a transitional period in
the politics of geobody construction in Taiwan. During this epoch, a plethora of
“punitive expedition” photo albums and picture postcards of war imagery from
the guardline dominated the photography of Atayal peoples.108 Superficially, the
preponderance of combat, logistics, and scenes of encampment would suggest that
the pluralistic imaginary developed by Inō, Torii, and Mori was losing ground
during the run-up to World War I.109 However, second-order geobody construc-
tion in fact consolidated itself, oddly enough, in part through the circulation of
war albums. While some of these guardline-focused albums eschewed the use
of ethnonyms and culturally pluralist maps, others enshrined indigenous eth-
nic categories and normalized them by prefacing the narrative portions of their
accounts with thumbnail sketches of Taiwan’s ethnographic complexion. This
representational strategy—which situated dynamic, eventful, and individualized
photographic portraiture in a framework that presented such images to readers
as illustrations of timeless ethnic traits—would continue in Taiwan to the end of
colonial times.
230     Indigenous Modernity

In all such productions, which could run to over six hundred pages, the author
begins the account with an atemporal, spatially organized discussion of Atayal,
Saisiyat, Paiwan, Tsou, Bunun, Amis, Tsarisen/Rukai, Puyuma, and Yami social
organization, material culture, ritual practices, religion, and language. These
sketches are prefaced by a map that locates each second-order geobody and arrays
them like nation-states on a Mercator projection. Photographs from the Mori and
Inō storehouse and others in the same genre then illustrate the features of each
group. The second part of such publications, beginning with war albums in the
1910s, proceeds to narrate the movement of the guardline and Japanese troops into
the interior of Taiwan. Major battles are recounted, statistics of “territory mastered”
are inserted. In this process, the hard-and-fast line between indigenous Taiwan
and Han Taiwan was naturalized, along with the ethnic topography proposed by
Inō in his 1900 Taiwan banjin jijō (Conditions among Taiwan’s savages). This plu-
ralizing operation worked in tandem with images and texts that paradoxically cele-
brated a war that portended to wipe out the cultures so lovingly portrayed between
the same covers. Narita Takeshi’s 1912 album, Taiwan seiban shuzoku shashinchō
(Savage tribes of Taiwan photo album), set the tone (see figure 51).
Narita’s map depicts two competing logics. Its slightly smaller inset map clas-
sifies space (Thongchai’s definition of territorialization) according to its degree
of submission to the colonial government. The legend associates different colors
with varied military and civil conditions. The endpoint of the trajectory implied
by the dynamic categories “half-submitted,” “reluctantly submitted,” “completely
submitted,” and “completely unsubmitted” is a fully “submitted” Taiwan. Narita’s
larger map partakes in a different temporal logic. It depicts Taiwan’s indigenous
ethnic groups as a system of second-order geobodies. These static categories and
polygons, which were retained in dozens of official and commercial maps into the
1940s, depicted Taiwan as an eternally and essentially multiethnic geobody. The
dynamic and static portraits of indigenous spaces, however, had one thing in com-
mon: the indigenous, non-Han portion of Taiwan was separated from Han Taiwan
by a one-dimensional boundary line, the one depicted in figures 39 through 41.
By 1912 Narita’s map could truthfully depict Quchi and Wulai as “completely
submitted.” However, there was an even more compliant classification, which was
reserved for the Amis in the eastern rift valley, the Puyuma, and the tribes of lower
Hengchun: taxable regions under normal administration. In other words, by 1912,
it was possible to be classified as an indigenous tribe even if every single member
of an ethnicity’s second-order geobody was already a tax-paying Taiwanese sub-
ject, living in the normally administered area. Such was the case with the Amis,
whose designator does not appear in 1930s statistical studies of indigenous politi-
cal economy and demography. Nonetheless, the Amis’ putative living space was
coded as a second-order geobody and solidly colored apposite other ethnic groups
on prefatory maps to these compendia.110 Thanks to their traditions of surplus
wealth production and historical familiarity with a commoditized economy, it was
Figure 51. Dynamic and static maps of Taiwan’s ethnic diversity, ca. 1912. Narita Takeshi,
Taiwan seiban shuzoku shashinchō (Taipei: Narita shashin seihanjo, 1912), n.p.
232     Indigenous Modernity

possible for the Amis peoples, as well as the Tuilasok, Koalut, and Sabaree Paiwan,
to move out of the “special administration” category during the first decade of
the twentieth century under Japanese rule. However, the colorful polygons that
marked out their separation from Han-dominated Taiwan on mass-produced,
scholarly, and official ethnic maps never budged.
Over the course of Japanese rule, static ethnic maps resembling figures 2
through 4 and figure 41 were reproduced millions of times in photo albums, on
picture postcards, in school textbooks, and in newspapers for public consumption
in Taiwan, in Japan, and around the world. In short, while Japanese military paci-
fication and police occupation among the Atayal, Sediq, Truku, Tsou, and Paiwan
succeeded to a level unprecedented in Taiwanese history, the movement of the
Qing-period savage border in concert with settler encroachment, assimilation, and
state aggrandizement was finally arrested under the Japanese regime in the 1910s.
Although more punitive expeditions and forced relocation campaigns were yet to
come, the notion that “special administration” in indigenous Taiwan was a tempo-
rary expedient went extinct. Administrators, scholars, and imperial subjects came
to view all indigenes, no matter their outward signs of assimilation, as members of
aboriginal ethnic groups that lived beyond the pale of dynamic capitalist political
economy. This vision of indigeneity turned out to be a self-fulfilling prophecy.

AT TACHING PHOTO GR APHS TO MAPS: THE CULTUR AL

INVESTMENT OF SEC OND-ORDER GEOB ODIES

In 1905, the year of the Atayal raids on the Quchi power station, the government-
general cut off and surrounded Jiaobanshan, the other site of serially reproduced
Atayal imagery, with fortified guard posts, clear-cut fields to aid in the detec-
tion and shooting of indigenes, and electrical wire. In 1906 Governor-General
Sakuma Samata took the reins and turned up the pressure on Atayal peoples who
resisted the commoditization of forest lands and subordination to a bureaucratic
surplus-extracting state. Government forces embargoed recalcitrant Jiaobanshan-
area tribes into surrendering their weapons by extending the guardlines into the
interior. The last push culminated in a 107-day Chintōzan campaign that cost
hundreds of dead on either side.111 In local memory, the 1907 battle is consid-
ered the beginning of Japanese rule in Jiaobanshan (today’s Fuxing Municipality
in Taoyuan Prefecture). By 1925 the former hot spot in the camphor wars had
become a show village for indigenous culture, despite its minority indigenous
population. Jiaobanshan’s transition from battleground to ethnic theme park, it
could be argued, presents a microcosm of the island-wide processes that secured
the indigenous habitations of Taiwan as second-order geobodies.
In 1908 the main trunk railway line from Jilong to Gaoxiong opened. This feat
of civil engineering was the first step in the creation of a robust ethnic-tourism
industry in Taiwan, though it would take years to mature. The numerous colonial
 The Geobodies within a Geobody    233

railway and commercial tourist guides that followed the completion of the trunk
line provide us with a window into the transformation of Jiaobanshan from ungov-
erned space to accessible ethnic resort town. The only Atayal person to appear in
an early 1908 guide occupies the corner of a montage. The text surrounding the
illustration describes the loss of Japanese life in a massacre near Beipu, the site of a
major uprising in 1907. In jarring contrast to later guides, this pamphlet only men-
tions indigenes in connection with the establishment of rural garrisons, guard-
lines, and reclamation projects.112
In 1910 a light-gauge pushcart railway connected the walled city of Dakekan to
formerly inaccessible Jiaobanshan; it was by far the most expensive and lengthy
rail extension of the year’s “savage territory” improvements.113 This railway fed sup-
plies and troops to Jiaobanshan for Japanese military actions against the tramon-
tane Dakekan settlements known as the Gaogan tribes (an Atayal people). Endō
Hiroya’s Banhi tōbatsu kinen shashinchō (Aborigine punitive expedition commem-
orative photo album), dated 1911, features photographs of Jiaobanshan town occu-
pied by Japanese soldiers and policemen. The main road is littered with military
infirmaries. A few “Jiaobanshan banjin” are pictured standing next to policemen
(see figure 52).114
Late 1910 marked the end of large military maneuvers against Gaogan. By 1912
Jiaobanshan was sufficiently pacified to be considered the site for a visit from the
Meiji emperor himself. For the emperor’s expected visit, a special reception hall,
later depicted in numerous picture postcards, was erected in the main settlement
area.115 But, as a privately produced 1915 travel guide reminded its readers, even
though the area was undergoing reclamation and the population was increasing,
the next stop, Gaogan, was still garrisoned for defense against indigenes.116
Just after this 1915 guide appeared, Jiaobanshan became an obligatory stop on
imperial inspection tours of Taiwan. The Imperial Household Ministry archive in
Tokyo contains several lavishly illustrated albums of these tours, which included
those of Prince Kan’in in 1916, Crown Prince Hirohito in 1923, and Prince Chichibu
in 1925. During this liminal decade, the Japanese royals affixed commercial pho-
tographs and their own snapshots to keepsake albums. The overall impression is
of Jiaobanshan as a borderland that was fast becoming an indigenous theme park.
By 1920, the light-gauge pushcart railway that connected the town of Jiaobanshan
to the railway stop at Daxi (the new name for Dakekan) had become a conduit
for tourists, freight, and officials. Travelers ascended the mountain pushed by two
Taiwanese laborers, with rest stations and inns all along the way.117 A 1921 travel
guide emphasized the scenery and the pushcart ride. It also created the boilerplate
for subsequent Jiaobanshan descriptions in the official travel guides:
Nine ri and seven cho southeast of Daxi [Dakekan] walled city is the savage border. If
you clamber your way up, you’ll be able to intimately inspect conditions among the
aborigines. This area is zigzagged and steep. Climbing up the mountain path by tram-
way is extremely slow, but descending is extremely fast, for a speedy return. At the
Figure 52. Jiaobanshan as staging area for Gaogan offensives, ca. 1910. Endō Hiroya, ed.,
Banhi tōbatsu kinen shashinchō (Taipei: Endō shashinkan, 1911), n.p.
 The Geobodies within a Geobody    235

top of the mountain, things suddenly open up into orderly quadrilateral residences;
this is Jiaobanshan. There is the balmy fragrance of camphor and the reception hall
built from building materials from Arisan. This area is a tableland that commands a
view upstream of the Dakekan ravine and is ringed by verdant linked ridgelines. . . .
It is thus known not only for its steep terrain but also for its spectacular scenery.118

Sōyama Takeshi writes that “indigenous territory” tourism had become a main-
stay by 1924, when the “injuries from indigene violence” had finally subsided and
railways extended to the interior.119 The 1924 version of the railway guide boasted
that Jiaobanshan was not only a “strategic hold in aborigine administration” but
also the site of two Japanese inns, a police station dispatch office, a post office, a
school for aborigine children (figure 27), an aborigine trading post (figure 30), a
Mitsui Corporation outpost, and a camphor company collection depot. The guide
also mentioned that visitors could pick up valuable indigenous manufactures at
good prices. Jiaobanshan, concluded this entry, could not be overlooked as a place
to “get to know the aborigines.”120
Jiaobanshan’s reputation even extended to the rare foreign visitor. In 1925, H. L.
Bagallay of the British Foreign Office described Jiaobanshan as “a ‘show village’
about twenty-five miles from the capital, where ‘tame’ savages are exhibited to offi-
cial visitors and tourists . . . ”121 The 1927, 1930, and 1934 Taiwan Railway guides
repeated the 1924 description above, suggesting that ethnic tourism had become
routinized in Jiaobanshan by the early 1930s.122
The 1934 guide adds that Jiaobanshan, Wulai (see above), and Alishan, all sites
for indigenous tourism, were now in “normally administered territory.” However,
the official 1935 Indigenous Household Register statistics indicate that all Atayal
and Tsou peoples—the indigenous peoples of these three sites—resided beyond
administered territory. Therefore, tourists were actually chatting with and pho-
tographing indigenes who were either visiting the tourist sites as employees or
who had removed to the normally administered territory.123 In these three popular
sites, tourists could have the best of both worlds. They could enter Jiaobanshan,
Wulai, or Alishan without filling out burdensome paperwork and paying extra
fees required to travel beyond the line between special and normal administration
and “experience savage life” without unduly upsetting their tour itineraries.124 Ease
of access and proximity to civilization were indeed the major themes in touristic
imagery of Jiaobanshan in the 1930s. Photos of Japanese schools (run by police
officers) filled with indigenous children and Hinomaru flags decorated maga-
zines, atlases, and picture postcards of the 1920s and 1930s (figure 31). Much of the
textual rhetoric accompanying these images extolled the conversion of a former
head-hunting people into rice-growing, peace-loving subjects of the emperor.
Recall that in 1895, culturally hybrid figures—the Atayal woman Washiiga in
a Chinese blouse or the Atayal man Pu Chin wearing a queue—were prominent
in Japanese imagery because interpreters were the point of access to Jiaobanshan.
Anthropologist Inō Kanori, in 1896, abstracted a photograph of Jiaobanshan
236     Indigenous Modernity

residents taken in built environments by whiting out its background to produce

more suitably anthropological types. Into the 1910s and even the 1920s, Chinese
blouses like Washiiga’s were prominent in photographs of Atayal women in border-
land junctions such as Jiaobanshan or Wulai, as exemplified by Mori Ushinosuke’s
photo of a “Dakekan tribe woman” in July 1906, at the height of the camphor wars
near Jiaobanshan, and by dozens of picture postcards.125 By the 1930s, however,
Chinese blouses were scarce in Japanese photography, while Japanese garments
became commonplace. For ethnic-tourism operators, the question became how to
balance the messages of progress and exoticism by packaging accessible villagers
as avatars of a more primitive and even timeless past.
One answer was found in the skillful use of graphic arts to construct Jiaobanshan
as a gateway into the Atayal geobody. This effect was achieved by manipulating
particular icons of Atayal ethnicity such as facial tattoos, tobacco pipes, back-strap
looms, and wicker backpacks. These icons of the unique Atayal cultural repertoire
were detached from lived environments, bundled together, and marketed as por-
table keepsakes from an unspoiled corner of the savage territory. On a circa 1930
postcard, a woman’s facial tattoos, wicker backpack, and pipe were all logoized in
such a fashion (see figure 53).
The sitter for this postcard image is also the mascot for the sleeve that houses
a set of eight Jiaobanshan postcards (see figure 54). Like Washiiga, Habairon, and
Marai in the 1896 anthropology journal (figure 39), this woman was rendered into
line art and then imaginatively placed in a new setting. But instead of the back-
ground that corresponded to blank space on the 1905 census map (figures 33 and
39), the Wakayama Prefecture–based Taishō Company placed her in front of a
mountain range and Jiaobanshan’s trademark suspension bridge. Tourists walked
single file along this popular attraction, which connected the souvenir stands
and the model school for indigenes to the mountains. The sitter in figure 53 was
photographed in front of a fence and a neatly edged footpath on level ground,
presumably in Jiaobanshan’s model ethnic village. The graphic artist thereafter
transformed her into a highlander guarding a secret passageway for the postcard
sleeve titled Bankai no hikyō (Undiscovered savage border).
Advances in reprographic technology allowed for further abstractions and
mash-ups. This sitter’s tattooed face, as well as her backpack, were logoized into
monochrome stamps that adorned other cards in the set, tying otherwise placid
scenes that connoted settled life and domesticity to the tattooed mascot who
greeted visitors to the mountainous and supposedly mysterious savage territory
(figure 55). The photographs on the cards themselves depict an enclosed, flat settle-
ment composed of Japanese-constructed dwellings and public buildings. But the
last card in the set (figure 56) depicts the setting, much like the sleeve, as an exten-
sion of Taiwan’s central cordillera, rather than as the island’s most convenient place
to sample indigenous culture in the course of a weekend tour.
Figure 53. Jiaobanshan woman with basket and pipe, ca. 1930. KAPPAN-ZAN TAIWAN,
wa306, East Asia Image Collection, Lafayette College, Easton, PA, accessed July 26, 2017, http://
digital.lafayette.edu/collections/eastasia/warner-postcards/wa0306.
Figure 54. Postcard sleeve, “Jiaobanshan’s hidden savage border,” ca. 1930 (published by
Taisho of Wakayama, Japan). [“The Savage Border’s Unexplored Boundary: Mt. Kappanzan’s
Variegated Coloration.”] This image is reproduced from an Internet auction.

Figure 55. Couple in Jiaobanshan, ca. 1930. KAPPAN-ZAN TAIWAN, wa0288, East Asia
I­ mage Collection, Lafayette College, Easton, PA, accessed July 26, 2017, http://digital.lafayette.
edu/collections/eastasia/warner-postcards/wa0288.
 The Geobodies within a Geobody    239

Figure 56. Mountains of Jiaobanshan, ca. 1930. KAPPAN-ZAN TAIWAN, wa0291, East Asia
Image Collection, Lafayette College, Easton, PA, accessed July 26, 2017, http://digital.lafayette.
edu/collections/eastasia/warner-postcards/wa0291.

Jiaobanshan worked as a show village, then, because it could provide the expe-
rience of “savage Taiwan” in such a way as to suggest that this “tame” settlement
was indeed an exemplar of the rest of aborigine country. In this example, the deft
use of imaginative geography, cropping, extrapolation, and logoization created a
snapshot of Atayalness that was quickly apprehended and could be brought back
or sent to Taipei, Gaoxiong, Tokyo, or Osaka in the form of a souvenir postcard set.
In 1935 the government-general hosted the Taiwan Hakurankai (Taiwan Expo)
to celebrate its fortieth anniversary. One Taiwan Expo album held dozens of pho-
tographs that also appeared in the monthly bulletin Riban no tomo (Aborigine
policemen’s companion), published by the Keimukyoku Ribanka (Police Bureau,
Aborigine Affairs Section) and in numerous postcard designs. These photographs
were mostly shot by Segawa Kōkichi, an intimate of the Aborigine Affairs section
chief Suzuki Hideo, and successor to Mori Ushinosuke as Japan’s official portrait
photographer of anthropological types.126 In the Atayal section of the album, the
Jiaobanshan photo shows the “Japanizing” of Atayal models in “various kinds of
antiquated and new-style clothing.” Right below it is a photo of Atayal from the
Wushe settlement of Paalan, the home of Iwan Roba0, Kondō the Barbarian’s first
wife. The Paalan men and women show no outward signs of participation in the
market for imported cloth or building materials (figure 57).
Figure 57. Contrasting photos of Jiaobanshan and Paalan, 1935. Top, “New and ancient
various kinds of clothing: Xinzhu-shū Kappanzan-sha”; bottom, “Playing the mouth-harp: two
maidens: Taizhong-shū Paalan-sha,” from Suzuki Hideo, ed., Taiwan bankai tenbō (Taipei:
Riban no tomo hakkōjo, 1935), 13.
 The Geobodies within a Geobody    241

Figure 58. Second-order geobody of Atayal [“Taiyal tribe”]. Suzuki Hideo, ed., Taiwan
bankai tenbō (Taipei: Riban no tomo hakkōjo, 1935), 6.

To suture these two photos together ideologically—the one displaying

Japanization and the other what is being Japanized—the first page of the Atayal
section of the souvenir book subsumes both images under the second-order geo-
body Atayal (figure 58). Because it was located in Atayal country and was con-
figured horizontally within a common two-dimensional geobody, along with
the visibly unassimilated inhabitants of Paalan, Jiaobanshan could represent the
whole of untouched northern Taiwan, while exhibiting imperial progress toward
Japanizing a wilderness that, a mere thirty years earlier, could only be represented
as a borderland on the extremity of the Qing Empire. Here, in the dioramas,
photographs, and ethnic maps of the 1935 exhibition, we find the fruition of the
tandem operations of cartographic and racialist geobody construction. Whether
strolling the fairgrounds, making a day trip to Wulai or Jiaobanshan to buy sou-
venirs, or perusing scholarly, commercial, or official photography of Taiwan
Indigenous Peoples, Japanese people could quickly apprehend 1930s Atayal culture
242     Indigenous Modernity

Figure 59. Marai and Yūgai of Rimogan, ca. 1903. Customs of Savage Tribe [sic], ip1471,
East Asia Image Collection, Lafayette College, Easton, PA, accessed July 26, 2017, http://digital.
lafayette.edu/collections/eastasia/imperial-postcards/ip1471.

and extrapolate it across the surface of Taiwan’s interior. The ensemble of artifacts,
maps, performances, images, and texts associated with Jiaobanshan represented
Taiwan’s aborigines at once as a population frozen in time and as a race that had
been rescued from the neglect of the Qing dynasty by energetic Japanese officials.
For those who could not make the trip to an exhibition or a souvenir stand,
government-published textbooks—compulsory reading for nearly the whole of
Japan—employed the same stratagems. When the Ministry of Education finally
revised its elementary-school geography texts in 1910 and 1919, it partially heeded
Inō Kanori’s criticisms by including portraits of Han Taiwanese. Each edition also
updated its indigenous imagery, both relying on Mori’s photographs from Wulai
and Rimogan. The 1910 edition displayed profile mugshot etchings of Pazzeh of
Wulai and Marai of Rimogan and captioned each as generic banjin (savages). The
1919 edition substituted Yūgai of Rimogan for Pazzeh of Wulai to present readers
with a male and a female icon from the same settlement. Mori’s June 1903 photo-
graph (see figure 59) was the source of this illustration.
A couple of months before Mori photographed the Rimogan couple, he also took
a scenic photograph of Atayal dwellings near Wulai (figure 60). This composition
was reproduced in several government and commercial publications throughout
the colonial period. The updated 1919 textbook rendered these photographs from
Figure 60. Wulai dwelling and granary, ca. 1903. Source: Mori Ushinosuke, ed., Taiwan
­banzoku zufu, vol. 1 (Taipei: Rinji Taiwan kyūkan chōsakai, 1915), plate 11.

Figure 61. Yūgai and Marai, in textbook illustration, 1919. Monbushō, ed., Jinjō
shōgaku chirisho kan ni (Tokyo: Monbushō, 1919), 16.
244     Indigenous Modernity

distinct Atayal settlements into emblems of the same savage territory by insert-
ing the Rimogan portrait of Yūgai and Marai into a Wulai setting (see figure 61).
For the editors of the 1903, 1910, and 1919 geography textbooks, then, all indig-
enous people constituted a subpopulation. The 1919 version that merged Wulai
and Rimogan imagery added a physical setting to cement the view that indigenes,
no matter where they were photographed, were essentially hill people.127

I N D IG E N OU S M O D E R N I T Y: A T WO - PA RT I N V E N T IO N

A careful student of Taiwanese history and Japanese colonial rhetoric, upon

attending the 1935 Taiwan Expo, might have asked why, after forty years of colonial
rule, the indigenous people of Paalan (figure 58) were still dressing in handwoven
fabrics, living in drafty slate houses, and working in pursuits far removed from the
booming capitalist economy of the lowlands. After all, the Japanese state had man-
aged to mobilize resources sufficient to win the Russo-Japanese War (1904–05),
dislodge the Germans from Shandong in 1915, and overrun the three northeastern
provinces of China in 1931. Why had it made such a poor showing in the uplands
of Taiwan after four decades of opportunity?
The answer to this question brings us back to Foucault’s formulation of the
“disciplinary society.” A number of scholars of Japanese colonial rule in Taiwan
have observed that, in the lowlands where some 97 percent of Taiwanese lived, the
Japanese colonial state was able to scale back its raw displays of punitive power
by 1915 or so, with the crushing of the Ta-pa-ni rebellion.128 At the same time,
the extension of the baojia system and the development of commercial law, pub-
lic health and education, commoditization labor, and a system of safeguards for
the protection of private property made great inroads into the warp and woof of
daily life, in cities and in the countryside. This interlocking set of institutions and
practices created a Taiwanese political economy that paid more to Tokyo in tax
revenue than it cost to maintain in terms of expenditures taken from the national
operating budget in Tokyo. A critical mass of lowland Han Taiwanese were,
however grudgingly, bought into a disciplinary order. The incentive system that
garnered this buy-in was itself expensive. To provide title deeds to land, access
to commercial courts, irrigation projects, harbors, schools, hospitals, and other
improvements, in such a way that they are in fact paid for by the colonized people
themselves, required a “big bang” on the order of Gotō Shinpei’s large spending
spree of the early twentieth century. By the mid-1910s, the state—in varied degrees
of alliance with Taiwanese capitalists, smallholders, professionals, policemen, offi-
cials, and baojia heads—had created a population that could not only produce
surplus wealth but would also surrender enough of it to the government to pay for
its own regulation. This form of society, wherein population increases redound
to the power of the state (instead of to its detriment), is the desideratum of the
disciplinary state.
 The Geobodies within a Geobody    245

Yao Jen-to has concurred, asserting that the Japanese created a disciplinary
society and biopolitical regime in Taiwan based on his interpretation of the state’s
profligate use of statistics and its vaunted interest in “forcing Taiwanese to be
healthy.” Curiously, Yao overlooks the fact that the first Japanese census left the
indigenous territories completely blank and that much of Taiwan still fell outside
the Government-General’s tax base when Japan surrendered in 1945.129
The lands beyond the pale of Yao’s analysis were at first defined as a zone of spe-
cial administration, or the indigenous territories, because they were little known
to Japanese administrators and did not appear on the ready-made grid inher-
ited from the Qing. To rectify this situation, Sakuma Samata initiated his own
“big bang” in the form of coordinated military offensives that culminated in the
gokanen keikaku riban jigyō (five-year plan to control the aborigines). As we have
seen, this scorched-earth approach succeeded to the extent that rebelling subjects
were mostly disarmed. The mountain territories were, in fact, made safe for land
and forest surveys. Abutting forests were cleared of toll states, and the obfuscating
layer of tongshi (bicultural interpreters) that complicated relationships between
primary producers and camphor capitalists was expelled from the savage border.
This is all to say that the post-Sakuma “aborigine district” was not a product of
state neglect. After 1910 especially, and into the 1930s, the state proceeded to inven-
tory, catalog, regulate, regiment, spatially array, and even nurture indigenous pop-
ulations in order to extract wealth from the highlands. While this panoptical and
surveillance-like activity bore superficial marks of discipline, it excluded the most
important element: the state did not produce individuals in Taiwan’s indigenous
territories. Rather, it produced tribes, settlements, ethnic groups, and aborigines.
Collective punishment, ad hoc justice, forced relocation, fixed prices at trading
posts, poorly compensated corvée labor, and nonrecognition of ownership rights
in forest lands were the lot of those who lived under special administration in
colonial Taiwan. In these lands beyond the pale of the Han-dominated lowlands,
the state did not invest in banks, courts, high schools, train stations, and other
discipline-inducing infrastructure. In the specially administered areas, Taiwanese
could not secure title deeds, compete for status in a public school system, open
businesses in bustling port towns, or sell agricultural produce, hunted goods, or
handicrafts at market prices.
Given the bounty of natural resources to be found in the highlands and the
alacrity with which indigenes learned the Japanese language and new trades, it
might seem puzzling that the state did not press its apparent advantage in 1915 to
bring the highlands within the profitable zone of disciplinary society. The lower
Hengchun Paiwan peoples, led by Pan Bunkiet and then his son, provide a tell-
ing example. These villages were of mixed Han, jukuban, and Paiwan residence.
They were included in the regularly administered territory of Taiwan during the
period Matsuoka Tadasu has characterized as one of temporary expediency. By
1906 almost all of the lower Hengchun Paiwan settlements were recognized as
246     Indigenous Modernity

being sufficiently settled to pay taxes. Therefore, their territories would no longer
be closed to immigrants, and they would fall under a structure of surveillance
similar to neighboring baojia units. Thus, along with the Amis and the Puyuma
of the eastern rift valley, the lower Hengchun Paiwan were, in many respects, inte-
grated into the empire. Nonetheless, as culture bearers and denizens of ethnically
defined subgeobodies, they occupied a liminal space: they were legally, economi-
cally, and administratively considered “civilized,” but they were still accounted
for in population counts as members of non-Han ethnic groups—as Paiwans.130
Moreover, they were still represented at tourist sites, in exhibitions, and in media
as colorful examples of Taiwan’s ethnic diversity. In a word, there was more to
the permanent bifurcation of Taiwan into ethnically demarcated zones of differ-
ence than uneven capitalist development or relative levels of state neglect and
oppression.
There are two major reasons for the arrested development of disciplinary soci-
ety in colonial Taiwan. The first is negative. When TGG visionaries first mapped
out schemes for ruling the island, Taiwan was Japan’s only colony. As Japan added
colonies (Karafuto and the Kwantung leased territory in 1905, Korea in 1910), the
relative size and importance of Taiwan’s uplands in the grand scheme of things
shrank. Consequently, the cost of building a capitalist infrastructure that could
produce a self-regulating disciplinary society in upland Taiwan was beyond the
reach of the government-general by 1915. As we saw in chapters 1 and 2, even when
the empire was relatively flush after the Sino-Japanese War, savage-border policy
was run on the cheap. Although Sakuma’s five-year plan to control the aborigines
was announced with great fanfare and a ¥15 million budget, the funds dried up
before the work was finished, and Japan declared victory in 1914 out of exhaustion.
Nonetheless, from 1915 through 1925, a series of land and population surveys
was conducted throughout the highlands. These projects set the stage for not only
large forest giveaways but also forced migration movements and programs to tran-
sition indigenes into intensive agriculture. As Matsuoka Tadasu has argued, up
until 1915, special administration was a temporary expedient—the original plan
was to economically integrate the island economy.131 But despite the intrusions of
irrigation projects, village relocation, and police government, the TGG policies
kept the indigenous peoples intact as a separate population. Labor migration to
the plains was discouraged; the fruits of intensive agriculture were not exported
beyond the highlands but were consumed in the highlands themselves; and a
majority of indigenes never paid taxes. Forced migration only occurred within the
aborigine territory: peoples were not moved to areas under regular administra-
tion. Matsuoka documents a number of positions, including that of Governor-
General Den Kenjirō, that argued for the dissolution of the indigenous territory
and Taiwan’s administrative integration. Nonetheless, the government-general
halted in its tracks at the old savage border, partly out of bureaucratic iner-
tia but also because the Wushe uprising showed that indigenes were in need of
 The Geobodies within a Geobody    247

different treatment. Moreover, argues Matsuoka, Japanese scholar-officials in the

­post-Wushe era tipped the balance in favor of preservationist policies.132
Matsuda Kyōko’s analysis of Japanese policies, practices, and representational
strategies in 1930s Taiwan confirms Matsuoka’s hypothesis from a different van-
tage. Most importantly, Matsuda demonstrates that preservationism reached
beyond the small circle of intellectuals discussed above and was translated into
actual policy. This brand of preservationism, Matsuda argues, forced policemen,
exhibition organizers, tourism operators, and administrators to walk a fine line.
On the one hand, they needed to demonstrate the efficacy of Japanization pro-
grams and policies aimed at turning supposedly rootless, violent, wandering
hunters into settled and peaceful agriculturalists. On the other hand, these same
social engineers recoiled in horror at the prospect of indigenes becoming dan-
dies who might participate in the consumer culture and urban cosmopolitanism
that characterized Taipei in the 1930s. The reformed yet not modern indigenous
Taiwanese would always be toilers on the soil, possessed of beautiful customs.
They were an ethnically separate population requiring the ministrations of a
trained corps of Japanese aborigine policemen-administrators to both ease them
away from the worst of their old customs and shield them from the worst excesses
of modernity.
The evidence for Matsuda’s general claim is abundant. For one, whereas indig-
enous sightseeing programs in the late 1890s through the 1910s brought small
numbers of headmen and traditional leaders to urban Japan to scare them into
submission by displaying Japanese military equipment and the advanced infra-
structure of the home islands, tours in the 1930s involved large groups drawn
from aborigine youth corps to demonstration farms within indigenous territory
or to other flourishing agricultural sites in Taiwan or Japan. The goal was no
longer to implant visions of Japan’s overwhelming modernity and power but to
educate indigenous people practically for crop production and stock breeding.133
Here we see an echo of the representational complex exhibited in the ethnic maps
and photographs from the 1935 Taiwan Expo: progress was being made among
the indigenes, but they remained and would always remain a rural folk rooted
to the land and engaged in useful pursuits, as members of cohesive, tradition-
bound collectivities. Insofar as such a construction was a projection of Japanese
conservative fantasies about the timeless virtues embodied in the imperiled and
deracinated rural population of the home islands, it makes sense to view the indi-
genization of Taiwan’s interior as a concomitant of Japanese nation building.134
More importantly, Taiwan-based academic, commercial, and official insti-
tutions had financial, emotional, and careerist stakes in the maintenance of the
second-order geobody of the indigenous territory and its subdivision into eth-
nically defined language groups. Perhaps most prominent among them were the
indigenous territory police and the aborigine youth corps that rose to prominence
in the wake of the Wushe uprising. Both institutions were invested in the kind
248     Indigenous Modernity

of modified preservationism described by Matsuda and had grown elaborate and

cross-cutting organizational structures by the 1930s. The new youth leaders were
fluent in Japanese, not Chinese. The lingua franca that linked them to other indig-
enous youth leaders separated them from Han Taiwanese, as it promoted sociabil-
ity and identification with their indigenous police counterparts.
The interests of the indigenous police bureaucracy dovetailed with local boost-
ers who lobbied for the selection of Taroko and Alishan National Parks above
other sites in Taiwan. This group included primitivists like Lan Yinding and Ōzaki
Hotsuma, Japanese alpinists, ethnologists at Taihoku Imperial University, musi-
cologist Kurosawa Takatomo, and the many Japanese tourists, consumers, and
collectors who were emotionally invested in the continued existence of Taiwan
Indigenous Peoples as an identifiably separate and autochthonous population.135
To summarize, this book has considered the emergence of indigenous moder-
nity as a two-part invention. The first part is a combination of military force suf-
ficient to rob peoples in dynastic-state peripheries of their autonomy, coupled
with an insufficiency of state resources to create disciplinary societies. In areas of
Taiwan not already softened up by Chinese immigration and Qing governance,
punishment remained ascendant over discipline into the 1930s. Here, it is impor-
tant to consider the fact that the GMD itself, with many more resources, main-
tained the ethnically bifurcated form of rule in Taiwan. Its historical resilience
indicates that “specially administered territory” is more than a residual category.
Rather, these homelands embody positivities that gained considerable historical
inertia during a moment in world-historical time marked by the ascendance of
cultural pluralism.
These positivities sank roots in Taiwan coterminously with the so-called “cul-
tural policy” in Korea, in tandem with a more diffuse, Wilsonian world order
that moved sovereignty from an achieved status (civilized nations only) toward
an ascribed status (all nations are ipso facto sovereign). The second part of this
invention is then the construction of the second-order geobodies that formed the
bulwarks of cultural pluralism in colonized Taiwan. As we have seen, the artis-
tic, commercial, and scholarly impulses to collect, preserve, and idealize elements
of indigenous cultures produced reified versions of them in museum collections,
postcards, folklore anthologies, and ethnographic portraits. These preserved ver-
sions of indigenous culture have been sources of contestation, while they have also
been utilized as resources for indigenous survival. The extent to which certain
taxonomies, ethnonyms, or other articulations of indigenous culture are Japanese
inventions is debatable. However, the shape and fundamental nature of the sec-
ond-order geobody is not. It was under Japanese colonial rule that polygons such
as the one illustrated in figure 58 were consolidated.
Over the course of its rule, the Taiwan Government-General zoned most of
Atayal country as public land, save for a small percentage reserved for villages
and their residents. This process amounted to land confiscation. Ownership of
 The Geobodies within a Geobody    249

public lands was transferred to the GMD after retrocession in 1945. While the
lands that were taken away from indigenes during the camphor wars and subse-
quent dispossession operations from 1898 through the 1930s were inhabited by
myriad settlements and distinct collectivities then known as Gaogan, Dakekan,
Nan’ao, or Tgdaya peoples, litigation with Taiwan’s central government now aims
to recover an Atayal homeland. In their analysis of the struggles surrounding the
establishment of Maqaw National Park in north-central Taiwan, activist-scholars
Yih-ren Lin, Lahuy Icyeh, and Da-Wei Kuan have argued that its proposed bound-
aries fall squarely within “Atayal people’s traditional territory.” The map they use
to make this argument depicts Maqaw National Park enveloped by a much larger
Atayal homeland. The space demarcated as territory to be restored to Atayal man-
agement, if not sovereignty, is not the home of a particular dialect community,
ritual group, or voting district but rather is none other than the one drawn up by
Japanese ethnologists, as shown in figures 2, 3, 34, 51, and 58.136
There are many good practical, historical, and legal reasons to consider this
land as Atayal and for the central government to comanage public lands with
Atayal people. There is no other effective way to configure autonomy, sovereignty,
and rights in land nowadays, except territorially. The architects of the Japanese
colonial project in ethnic bifurcation, who were many and not exactly in cahoots,
did not predetermine how these thorny problems will be resolved. But this study
suggests that they set the terms of engagement.
 Note s

I N T R O DU C T IO N

1.  Wushe is in today’s Ren’ai Township in Nantou Prefecture, Taiwan.

2.  Taiwan sōtokufu keimukyoku, ed., Takasagozoku chōsasho dai go hen: Bansha gaikyō,
meishin (Taipei: Taiwan sōtokufu keimukyoku, 1938), 132–33.
3.  Ronald Niezen, The Origins of Indigenism: Human Rights and the Politics of Identity
(Berkeley: University of California Press, 2003); James Clifford, Returns: Becoming Indig-
enous in the Twenty-First Century (Harvard University Press, 2013).
4.  Laura R. Graham and H. Glenn Penny, “Performing Indigeneity: Emergent Identity,
Self-Determination, and Sovereignty,” in Performing Indigeneity: Global Histories and Con-
temporary Experiences, ed. Laura R. Graham and H. Glenn Penny (Lincoln: University of
Nebraska Press, 2014), 1–5.
5.  Wang Fu-chang, Zokugun: Gendai Taiwan no esunikku imajineeshon, trans. Matsuba
Jun and Hung Yuru (Tokyo: Tōhō shoten, 2014), 86–87.
6.  Anthropologist Scott Simon writes that “ . . . today’s political debates about ­Indigenous
rights are rooted in an unfolding political dynamic that predates both the global ­indigenous
rights movement and even the arrival of the ROC on Taiwan. What we know today as
Indigenous Formosa is a co-creation of the resulting relationship between the ­Japanese
state and diverse political constellations among many Austronesian peoples across the is-
land.” From “Making Natives: Japan and the Creation of Indigenous Formosa,” in Japanese
Taiwan: Colonial Rule and Its Contested Legacy, ed. Andrew Morris (London: Bloomsbury
Press, 2015), 75.
7.  “Taiwan Indigenous Peoples” is an English translation of Taiwan Yuanzhuminzu,
the officially adopted name for indigenous peoples in Taiwan. This book will use the term
­indigenes to avoid awkward constructions and wordiness.
8.  Japan’s bifurcation of Taiwan fits under the category of “state-centered legal plural-
ism” or “strong legal pluralism,” as formulated by Lauren Benton. She writes that from “the
251
252     Notes

mid-nineteenth century forward, colonial states devised formal typologies, and drew firmer
boundaries, between social formations within colonial spaces that were perceived to be
at different ‘levels of development,’ and often attributed differences to temporal lag—de-
limiting one zone as ‘primitive’ and the other ‘modern’.” Law and Colonial Cultures: Legal
Regimes in World History, 1400–1900 (Cambridge: Cambridge University Press, 2002), 6.
9.  The term Austronesia refers to a “geographic region of the language family spread-
ing from the Western Pacific (e.g., Taiwan) to the Indian Ocean (e.g., Madagascar).” From
“Editor’s Note,” in Austronesian Taiwan: Linguistics, History, Ethnology, Prehistory, ed. David
Blundell (Berkeley and Taipei: Phoebe A. Hearst Museum of Anthropology and Shung Ye
Museum of Formosan Aborigines, 2000), xxi.
10.  The conceptual, administrative, and political vectors of continuity across the TGG
and GMD periods of rule in Taiwan are complicated. Direct continuities are document-
ed most clearly and forcefully in Matsuoka Tadasu, Taiwan Genjūmin shakai no chihōka:
Mainoriti no nijūsseiki (Tokyo: Kenbun shuppan, 2012), 167–85. See also Nobayashi Atsushi
and Miyaoka Maoko, “Araraweru senjūmin shuchō no shosō,” in Senjūmin to wa dare ka,
ed. Kubota Sachiko and Nobayashi Atsushi (Tokyo: Sekai shisōsha, 2009), 296, 306–8, 312;
Chen Yuan-yang, Taiwan Genjūmin to kokka kōen (Fukuoka: Kyūshū daigaku shuppankai,
1999), 21–24; and Simon, “Making Natives,” 82–83.
11.  I am adapting James Hevia’s conceptualization of re- and deterritorialization. See
English Lessons: The Pedagogy of Imperialism in Nineteenth-Century China (Durham: Duke
University Press, 2003), 20–21.
12.  Michael E. Robinson, Korea’s Twentieth-Century Odyssey: A Short History (Hono-
lulu: University of Hawai‘i Press, 2007), 47–52.
13.  Takashi Fujitani, Race for Empire: Koreans as Japanese and Japanese as Americans
during World War II (Berkeley: University of California Press, 2011), 22–25; Kyung Moon
Hwang, Rationalizing Korea: The Rise of the Modern State 1894–1945 (Berkeley: University
of California Press, 2016), 9–10, 15–16.
14.  Henry H. Em, “Minjok as a Modern Construct: Sin Ch’aeho’s Historiography,” in Co-
lonial Modernity in Korea, ed. Shin Gi-Wook and Michael Robinson (Cambridge: Harvard
University Asia Center, 1999), 353 (emphasis added).
15.  Haruyama Meitetsu, “Shōwa seijishi ni okeru Musha hōki jiken,” in Taiwan Musha
hōki jiken kenkyū to shiryō, ed. Tai Kuo-hui (Tokyo: Shakai shisōsha, 1981), 131–36; Robin-
son, Korea’s Odyssey, 49.
16.  Kondō Masami, Sōryokusen to Taiwan: Nihon shokuminchi hōkai no kenkyū (Tokyo:
Tōsui shobō, 1996), 264–66; Matsuda Kyōko, Teikoku no shisō: Nihon “Teikoku” to Taiwan
Genjūmin (Tokyo: Yūshisha, 2014), 191–211; Sakano, Teikoku, 248.
17.  E. Taylor Atkins, Primitive Selves: Koreana in the Japanese Colonial Gaze, 1910–1945
(Berkeley: University of California Press, 2010); Hyung Il Pai, Constructing “Korean”
­Origins: A Critical Review of Archaeology, Historiography, and Racial Myth in Korean
­State-Formation Theories (Cambridge: Harvard University Asia Center, 2000); Hyung
Il Pai, Heritage Management in Korea and Japan: The Politics of Antiquity and Identity
(­Seattle: University of Washington Press, 2013). For Taiwan, see Yamaji Katsuhiko, Taiwan
­Taiyaru-zoku no hyaku-nen: Hyōryū suru dentō, dakō suru kindai, datsu shokuminchika e no
michinori (­Tokyo: Fūkyosha, 2011), 402–5.
 Notes    253

18.  Hu Chia-yu, “Embodied Memories and Enacted Ritual Materials: Possessing the
Past in Making and Remaking Saisiyat Identity in Taiwan” (PhD diss., University College
London, 2006), 48.
19.  Steven E. Phillips, The Taiwanese Encounter Nationalist China, 1945–1950 (Stanford:
Stanford University Press, 2003), 90–114.
20. Huang Chih Huei, “Ethnic Diversity, Two-Layered Colonization, and Complex
Modern Taiwanese Attitudes toward Japan,” in Japanese Taiwan: Colonial Rule and Its Con-
tested Legacy, ed. Andrew Morris (London: Bloomsbury Academic, 2015), 133–53; Asano
Toyomi, “Historical Perceptions of Taiwan’s Japan Era,” in Toward a History Beyond Bor-
ders: Contentious Issues in Sino-Japanese Relations, ed. Daqing Yang, Jie Liu, Hiroshi Mitani,
and Andrew Gordon (Cambridge: Harvard University Asia Center, 2012), 303. An advance
party for the Nationalist Party’s Seventieth Army, which became the replacement authority
of the defeated Japanese empire in Taiwan, landed in Jilong on October 17, 1945, while the
GMD was still nominally the central government of China. In April 1949, the first plane-
loads of fleeing Nationalist officials arrived in Taiwan from Shanghai in the wake of Chinese
Communist Party offensives. On May 19 of that year, governor of Taiwan Province Chen
Ch’eng proclaimed martial law. Chou Wan-yao, A New Illustrated History of Taiwan, trans.
Carole Plackitt and Tim Casey (Taipei: SMC Publishing, 2015): 303, 322–33.
21.  Michael Rudolph, Ritual Performances as Authenticating Practices: Cultural Repre-
sentations of Taiwan’s Aborigines in Times of Political Change (New Brunswick, NJ: Trans-
action Publishers, 2008), 1–4, 36–38; Mitsuda Yayoi, “First Case of the New Recognition
System: The Survival Strategies of the Thao,” in Taiwan Since Martial Law: Society, Culture,
Politics, Economy, ed. David Blundell (Taipei: Shung Ye Museum of Formosan Aborigines,
2012), 166–67.
22.  Ku Kun-hui, “Rights to Recognition: Minorities and Indigenous Politics in Emerg-
ing Taiwan Nationalism,” in Blundell, Taiwan Since Martial Law, 95, 98–99.
23. Rudolph, Ritual Performances, 55; Simon, “Making Natives,” 86–87.
24.  Simon, “Making Natives,” 87.
25.  Michael Berry, A History of Pain: Trauma in Modern Chinese Literature and Film
(New York: Columbia University Press, 2008), 53–107.
26.  Right after the Wushe uprising, highbrow general readership magazines ran feature
articles on the people who had revolted against the empire. In each of these publications,
the terms banjin (barbarian), banzoku (barbarian ethnic groups/tribes), and kyōban (bel-
licose barbarians) describe indigenous peoples. Authors also identified the Wushe villages
as members of the larger Atayal ethnic group. Namikawa Ryō, “Taiwan no Banjin to sono
minzokusei,” Kaizō (December 1930): 42–67; Nagamatsu Asazō, “Hagyakusuru Banzoku,”
Chūō kōron, 173–86. A popular science magazine added one refinement by noting that the
Wushe villages were part of the Sediq subgroup of the Atayal ethnic group. Matsumoto
Akira, “Taiwan banzoku: Riban jigyō to jinruigaku,” Kagaku chishiki 10, no. 12 (Decem-
ber 1930): 8–11. For representative newspaper articles, see “Musha bōdō jiken no gen’in
to shinsō,” Tokyo Asahi Shinbun, November 20, 1930, 2; “Musha no banjin hōki: keisatsu
bunshitsu sono ta o shūgeki,” Taiwan nichinichi shinpō, October 29, 1930, 1. Newspapers also
used a few references to the ethnic marker Atayal amid a blizzard of pejorative terms such
as banjin and banzoku.
254     Notes

27.  Matsuda Kyōko, Teikoku, 189–90; Leo T. S. Ching, “Savage Construction and Civility
Making: Japanese Colonialism and Taiwanese Aboriginal Representation,” Positions: East
Asia Cultures Critique 8, no. 3 (2000): 795–97, 815. Fujitani makes a related point regarding
Japanese rule in Korea in Race for Empire, 25–26, 37–39.
28.  See Uno Toshiharu, “Taiwan ni okeru ‘banjin’ kyōiku,’ Tenbō 4, no. 196 (April 1975):
37, for difference between sha/she and shō/zhuang; see Matsuoka, Taiwan Genjūmin shakai
no chihōka, 41–44, for a detailed description of various Paiwan terms for units of settlement
and affiliation; see the hundreds of tables in the official publication Takasagozoku chōsasho
dai san-hen: shinka (1937), ed. Taiwan sōtokufu keimukyoku, for lists of every indigenous
settlement in Taiwan categorized as a sha/she.
29.  Inō Kanori, “Shinten chihō ni okeru Seiban no jissa (zoku),” Tokyo jinrui gakkai
zasshi 122 (March 1896): 312–13.
30. Japanese ethnological surveys eschewed the term Sediq until 1915, subsuming
Wushe-area residents under the label Atayal. For a thorough review, see Hara Eiko, “Tai-
yaru, Sedekku, Taroko o meguru kizoku to meisho ni kan suru undō no tenkai 1: Taroko
ni okeru undō o chūshin ni,” Taiwan Genjūmin kenkyū 7 (2003): 210. The six-volume, early
1930s compendium Takasagozoku chōsasho classified all Sediq peoples as Atayal, indicating
that Sediq as a category remained submerged even after the rebellion. Jun’eki Genjūmin
kenkyūkai, ed., Taiwan Genjūmin kenkyū gairan: Nihon kara no shiten (Tokyo: Fūkyōsha,
2001), 47. Popular geographies of the empire followed suit by subsuming Sediq and Truku
peoples under the rubric Taiyal, a synonym of Atayal. See Nakama Terahisa, ed., Nihon chiri
fūzoku taikei dai jūgo kan (Tokyo: Shinkōsha, 1931), 180; Yamamoto Mitsuo, ed., Nihon chiri
taikei: Taiwan hen (Tokyo: Kaizōsha, 1930), 320–39. The latter title included mention of the
Sediq as part of the Atayal group on its map but dropped the distinction in captions and
descriptions of particular practices or places. General interest books on the topic included
ethnic maps as a matter of course, with associated thumbnail sketches of each indigenous
group, but did not adopt the term Sediq. For examples, see Tagami Tadayuki, Banjin no
kishū to densetsu (Taipei: Taiwan banzoku kenkyūjo, 1935), 161, and Ōgata Tarō, Takasago-
zoku (Tokyo: Kyōseisha kōdōkyaku, 1942), 18–21. There were notable exceptions, however.
For example, the TGG railway travel guides supplemented descriptions of Wushe’s beauti-
ful scenery and hot springs with detailed discussions of the 1930 uprising, while describing
the rebels as Sediq members of the Atayal ethnic group. See Taiwan Sōtokufu kōtsūkyoku
tetsudō-bu, ed., Taiwan tetsudō ryokō annai (Taipei: Taiwan Sōtokufu Kōtsūkyoku tetsudō-
bu, 1935),142, Japan National Congressional Library Digital Collection, accessed 12/18/2016,
http://dl.ndl.go.jp/info:ndljp/pid/1875798. In another exception, the Oriental Tour Com-
pany’s Taiwan branch also separated the Sediq from the Atayal, going so far as to mark
them separately on the ethnic map. Ueda Hachirō, Takasagozoku no hanashi (Taipei: Tōa
ryokōsha Taiwan shibu, 1941), 11–13.
31.  Matsuda Kyōko, Teikoku, 189–94.
32.  Andreas Wimmer, Ethnic Boundary Making: Institutions, Power, Networks (Oxford:
Oxford University Press, 2013), 52–55. For particulars regarding the Taiwan case, see Yamaji
Katsuhiko, Taiwan Taiyaru zoku no hyaku nen (Tokyo: Fūkyōsha, 2011), 403–4.
33.  Walter LaFeber, The Clash: U.S.-Japanese Relations throughout History (New York:
Norton, 1998), 40–41.
 Notes    255

34.  V. I. Lenin, Imperialism: The Highest Stage of Capitalism (1917; repr., London and
Chicago: Pluto Press, 1996), 77–78 (emphasis added).
35.  Barry Buzan and George Lawson, The Global Transformation: History, Modernity
and the Making of International Relations (Cambridge: Cambridge University Press, 2015), 2.
36.  Buzan and Lawson, Global Transformation, 35, 63.
37.  Jürgen Osterhammel, The Transformation of the World: A Global History of the Nine-
teenth Century (Princeton: Princeton University Press, 2014), xv–xx; C. A. Bayly, The Birth
of the Modern World 1780–1914 (Malden, MA: Blackwell Publishing, 2004).
38.  Gabriel Popescu, Bordering and Ordering the Twenty-first Century: Understanding
Borders (Lanham, MD: Rowman & Littlefield, 2012), 34–38.
39.  Thongchai Winichakul, Siam Mapped: A History of the Geo-Body of a Nation (Ho-
nolulu: University of Hawai‘i Press, 1994), 16–18, 52–56. (This book will use the spelling
geobody.)
40.  The Treaty of Paris was signed between Spain and the United States on Decem-
ber 10, 1898, to cede the Philippine Islands to the latter; it was ratified by the U.S. Senate
on February 6, 1899. A large, prolonged war broke out between Philippine independence
forces and the U.S. Army on February 4, 1899. On the main island of Luzon alone, the war
of conquest continued well into 1902. Glenn Anthony May, Battle for Batangas: A Philippine
Province at War (New Haven: Yale University Press, 1991), 76–77, 211–69.
41.  Nicola Cooper, France in Indochina: Colonial Encounters (New York: Berg, 2001),
13–16; Mary Evelyn Townsend, European Expansion Since 1871 (Chicago: J. B. Lippincott,
1941), 474, 530–31.
42. The Treaty of Shimonoseki was signed on April 17, 1895, formally ending the
­Sino-Japanese War and ceding Taiwan, including the Pescadores Islands, to the Japanese
­Empire. Wu Micha, ed., Taiwan-shi kojiten, trans. Yokosawa Yasuo (Tokyo: Chūgoku sho-
ten, 2007), 137.
43.  Suzuki Shogo, “Japan’s Socialization into Janus-Faced European International Soci-
ety, European Journal of International Relations 11 (2005): 137–64.
44.  Jane Burbank and Frederick Cooper, Empires in World History: Power and the Poli-
tics of Difference (Princeton: Princeton University Press, 2011), 450.
45.  Lauren Benton uses the term “quasi-sovereignty” to discuss a range of categories
and legal frameworks nineteenth-century empires adopted for “enclave populations” sur-
rounded by majority, or constitutionally governed, populations—often mountain or hill
peoples. A Search for Sovereignty: Law and Geography in European Empires, 1400–1900
(Cambridge: Cambridge University Press, 2010), 226–78. I use the term here for contem-
porary populations who are recognized as distinctive from national populations insofar as
members of these groups have special rights, or are absolved of certain obligations, based
on recognized membership in a quasi-sovereign entity.
46.  W. G. Beasley, Japanese Imperialism, 1894–1945 (Oxford: Oxford University Press,
1987), 146–47.
47.  Buzan and Lawson, Global Transformation, 43, 171–96.
48.  David Held and Anthony McGrew, David Goldblatt and Jonathan Perraton, Glob-
al Transformations: Politics, Economics, and Culture (Stanford: Stanford University Press,
1999), 7–27.
256     Notes

49.  Known as Hengchun since 1875, after a series of Qing reforms sought to extend
imperial rule to this former hinterland. In this book, the name Langqiao is used when refer-
ring to events that occurred before 1875, and Hengchun when referring to events after 1875.
50.  Uchida Jun, Brokers of Empire: Japanese Settler Colonialism in Korea, 1876–1945
(Cambridge: Harvard University Asia Center, 2011), 6: “The term ‘brokers’ also captures
the intermediary position of settlers, who operated simultaneously as agents and pawns of
colonial power.”
51.  Daniel K. Richter, “Cultural Brokers and Intercultural Politics: New York-Iroquois
Relations, 1664–1701,” Journal of American History 75, no. 1 (1988): 41: “simultaneous mem-
bers of two or more interacting networks (kin groups, political factions, communities or
other formal or informal coalitions), [who] provide nodes of communication . . . with re-
spect to a community’s relations with the outside world.”
52.  Michael W. Doyle, Empires (Ithaca: Cornell University Press, 1986), 25–28.
53.  Lippincott, 11, 59; Sydney Giffard, Japan Among the Powers, 1890–1990 (New Haven:
Yale University Press, 1994), 33, 79.
54. Bayly, Birth of the Modern World, 433.
55.  Eric R. Wolf, Europe and the People without History (Berkeley: University of Califor-
nia Press, 1982), 88.
56.  James C. Scott, The Art of Not Being Governed: An Anarchist History of Upland
Southeast Asia (New Haven: Yale University Press, 2009), 6.
57.  Wolf refers to their structure as the “kin-ordered mode of production.” Wolf,
Europe, 88.
58.  Richard White, The Middle Ground: Indians, Empires, and Republics in the Great
Lakes Region, 1650–1815 (Cambridge: Cambridge University Press, 1991).
59. John Robert Shepherd, Statecraft and Political Economy on the Taiwan Frontier,
1600–1800 (Stanford: Stanford University Press, 1993).
60.  Caglar Keyder, “Law and Legitimation in Empire,” in Lessons of Empire: Imperial
Histories and American Power, ed. Craig Calhoun, Frederick Cooper, and Kevin W. Moore
(New York: Free Press, 2006), 119.
61.  Adam McKeown, Melancholy Order: Asian Migration and the Globalization of Bor-
ders (New York: Columbia University Press, 2008), 38.
62.  Peter N. Stearns, The Industrial Revolution in World History, 4th ed. (Boulder, CO:
Westview Press, 2013), 109–20.
63.  Robert Eskildsen, “Of Civilization and Savages: The Mimetic Imperialism of Japan’s
1874 Expedition to Taiwan,” American Historical Review 107, no. 2 (2002): 388–418.
64.  The Mudan village incident followed a string of incidents that involved extended
negotiations to recover passengers and property from distressed vessels in southern Taiwan.
65.  Ian Jared Miller, “Writing Japan at Nature’s Edge: The Promises and Perils of Envi-
ronmental History,” in Japan at Nature’s Edge: The Environmental Context of a Global Power,
ed. Ian Jared Miller, Julia Adeney Thomas, and Brett L. Walker (Honolulu: University of
Hawai‘i Press, 2013), 1–3; Jakobina Arch, “From Meat to Machine Oil: The Nineteenth-­
Century Development of Whaling in Wakayama,” in Japan at Nature’s Edge, Miller, Thomas,
and Walker, 41; Peter Duus, The Japanese Discovery of America: A Brief History with Docu-
ments (Boston: Bedford Books, 1997), 7–11; George Feifer, Breaking Open Japan: Commodore
 Notes    257

Perry, Lord Abe, and American Imperialism in 1853 (New York: Smithsonian Books and
HarperCollins), 70–75.
66.  John E. Herman, “The Cant of Conquest: Tusi Offices and China’s Political Incor-
poration of the Southwest Frontier,” in Empire at the Margins: Culture, Ethnicity, and Fron-
tier in Early Modern China, ed. Pamela Kyle Crossley, Helen F. Siu, and Donald S. Sutton
(Berkeley: University of California Press, 2006), 135–61.
67.  Susan Naquin and Evelyn S. Rawski, Chinese Society in the Eighteenth Century (New
Haven: Yale University Press, 1987), 225–27; William T. Rowe, China’s Last Empire: The Great
Qing (Cambridge: Harvard University Press, 2009), 35–39.
68. Pars Cassel, Grounds of Judgment: Extraterritoriality and Imperial Power in
­Nineteenth-Century China and Japan (Oxford: Oxford University Press, 2012).
69.  Chang, Lung-chih, “From Island Frontier to Imperial Colony: Qing and Japanese
Sovereignty Debates and Territorial Projects in Taiwan, 1874–1906” (PhD diss., Harvard
University, 2003).
70. Rowe, China’s Last Empire, 54.
71.  Sophia Su-fei Yen, Taiwan in China’s Foreign Relations, 1836–1874 (Hamden, CT:
Shoe String Press, 1965), 252–54.
72.  Michel Foucault, Discipline and Punish: The Birth of the Prison, trans. Alan Sheridan
(New York: Vintage Books, 1977), 79.
73. Foucault, Discipline and Punish, 82.
74. Foucault, Discipline and Punish, 85.
75. Foucault, Discipline and Punish, 85.
76. Foucault, Discipline and Punish, 87.
77. Hui-yu Caroline Ts’ai, Japan’s Empire-Building: An Institutional Approach to Colonial
Engineering (London and New York: Routledge, 2008), 93–94.
78. Fujitani, Splendid Monarchy: Power and Pageantry in Modern Japan (Berkeley: Uni-
versity of California, 1996), 5–11.
79. Fujitani, Splendid, 25.
80. Fujitani, Splendid.
81.  Christine Kim, “Politics and Pageantry in Protectorate Korea (1905–10): The Impe-
rial Progresses of Sunjong,” Journal of Asian Studies 68, no. 3 (2009): 835–59.
82.  Crown Prince Hirohito would visit in 1923 to much fanfare, thirty years into Japa-
nese rule. See chapter 3.
83.  E. H. Norman, Origins of the Modern Japanese State: Selected Writings of E. H. Nor-
man, ed. John Dower (New York: Pantheon Books, 1975), 181, 247.
84.  Kozo Yamamura, “The Meiji Land Tax Reform and Its Effects,” in Japan in Transi-
tion: From Tokugawa to Meiji, ed. Marius B. Jansen and Gilbert Rozman (Princeton: Princ-
eton University Press, 1986), 382.
85.  Yamamura, “Meiji Land Tax Reform,” 382–99.
86. Norman, Origins, 245–50.
87.  Matthew G. Hannah, Governmentality and the Mastery of Territory in Nineteenth-
Century America (New York: Cambridge University Press, 2000), 23, 25.
88. James I. Nakamura, Agricultural Production and the Economic Development of
­Japan, 1873–1922 (1966; repr., Princeton: Princeton University Press, 2015), 7–9.
258     Notes

89.  Ronald G. Knapp, “Settlement and Frontier Land Tenure,” in China’s Island Frontier:
Studies in the Historical Geography of Taiwan, ed. Ronald Knapp (Honolulu: University of
Hawai‘i Press, 1980), 55–68.
90.  Yao Jen-to, “The Japanese Colonial State and Its Form of Knowledge,” in Taiwan
under Japanese Colonial Rule, 1895–1945: History, Culture, Memory, ed. Liao Ping-hui and
David Der-wei Wang (New York: Columbia University Press, 2006), 47, 54.
91.  Yao, “Japanese Colonial State,” 43.
92.  Yao, “Japanese Colonial State,” 48.
93.  Yao, “Japanese Colonial State,” 52.
94.  Samuel Pao-San Ho, “Colonialism and Development: Korea, Taiwan, and Kwan-
tung,” in The Japanese Colonial Empire, 1895–1945, ed. Ramon H. Myers and Mark R. Peattie
(Princeton: Princeton University Press, 1984), 358.
95. Ka Chih-ming, Japanese Colonialism in Taiwan: Land Tenure, Development and
­Dependency, 1895–1945 (Boulder, CO: Westview Press, 1995), 55, 57.
96.  Yoshitarō Takenobu, Japan Yearbook (Tokyo: Japan Yearbook Office, 1906), 526.
97. Beasley, Japanese Imperialism, 146–47.
98.  Max Weber, The Protestant Ethic and the Spirit of Capitalism, trans. Talcott Parsons
(New York: Charles Scribner’s Sons, 1958), 17–18.
99.  Marius Jansen, “Japanese Imperialism: Late Meiji Perspectives,” in The Japanese Co-
lonial Empire, 1895–1945, ed. Ramon H. Myers and Mark R. Peattie (Princeton: Princeton
University Press, 1984), 65–66.
100.  Takekoshi Yosaburō, Japanese Rule in Formosa, trans. George Braithwaite (Lon-
don, New York, Bombay, and Calcutta: Longmans, Green, 1907), 138.
101. Takekoshi, Japanese Rule, 139.
102.  Mochiji Rokusaburō, “Bansei mondai ni kansuru iken,” in Riban shikō dai ikkan,
ed. Inō Kanori (1918; Taipei: Southern Materials Center, 1995), 180–228.
103. Takekoshi, Japanese Rule, 230 (emphasis added).
104.  Kyung Moon Hwang defines biopower as the “targeting of . . . [a] . . . population for
bureaucratic outcomes regarding the people’s lives, deaths, and welfare.” Hwang, Rational-
izing Korea, 9–10, 15–16.
105.  Sabine Frühstück, Colonizing Sex: Sexology and Social Control in Modern Japan
(Berkeley: University of California Press, 2003), 17–35.
106.  Achille Mbembe, “Necropolitics,” trans. Libby Meintjes, Public Culture 15, no. 1
(2003): 25.
107.  Traces of History: Elementary Structures of Race (London and New York: Verso,
2016), 7, 18.
108. Fujitani, Race for Empire, 40–66.
109.  Leo T. S. Ching, “Savage Construction and Civility Making: Japanese Colonialism
and Taiwanese Aboriginal Representation,” Positions: East Asia Cultures Critique 8, no. 3
(2000): 810–15.
110.  Huang Chih-huei, “The Yamatodamashi of the Takasagozoku Volunteers of Tai-
wan: A Reading of the Postcolonial Situation,” in Globalizing Japan: Ethnography of the Jap-
anese Presence in Asia, Europe, and America, ed. Harumi Befu and Sylvie Guichard-Anguis
(London: Routledge, 2001), 222–50.
111. Benton, Search for Sovereignty, 222–78.
 Notes    259

112.  Benedict Anderson, Imagined Communities (London: Verso, 1983), 163–85.

113.  Endō Masataka, Kindai Nihon no shokuminchi tōchi ni okeru kokuseki to koseki
(­Tokyo: Akashi shoten, 2010), 139–43.
114.  Ts’ai, “Shaping Administration in Colonial Taiwan,” in Liao and Wang, Taiwan
­under Japanese Colonial Rule, 1895–1945, 97–98.
115. Ka, Japanese Colonialism; Wang Taysheng, Legal Reform in Taiwan under Japanese
Colonial Rule, 1895–1945: The Reception of Western Law (Seattle and London: University of
Washington Press, 2000); Ming-cheng M. Lo, Doctors within Borders: Profession, Ethnic-
ity, and Modernity in Colonial Taiwan (Berkeley: University of California Press, 2002); E.
Patricia Tsurumi, Japanese Colonial Education in Taiwan, 1895–1945 (Cambridge: Harvard
University Press, 1977).
116.  Hu Chia-yu, “Embodied Memories,” 31–32.
117.  Antonio C. Tavares, “The Japanese Colonial State and the Dissolution of the Late
Imperial Frontier Economy in Taiwan, 1886–1909,” Journal of Asian Studies 64, no. 2 (2005):
361–85.
118.  Matsuda Yoshirō, Taiwan Genjūmin, 98–117, 149.
119.  Rinji Taiwan kokō chōsa-bu, Rinji Taiwan kokō chōsa kijutsu hōbun (Taipei: Taiwan
sōtokufu, 1908), 56–57.
120.  Charles Hirschman, “The Making of Race in Colonial Malaya: Political Economy
and Racial Ideology,” Sociological Forum 1, no. 2 (1986): 330–61.
121.  For intellectual and institutional foundations, see Pai, Heritage Management, 95–
113; for analyses of Taiwan surveys, see Miyaoka Maoko, “Mori Ushinosuke no chosaku
mokuroku oyobi jakkan no kaisetsu,” Taiwan Genjūmin kenkyū 2 (1997), 189–99; Kasahara
Masaharu, “Inō Kanori no jidai: Taiwan Genjūmin shoki kenkyū shi e no sokuen,” Taiwan
Genjūmin kenkyū 3 (1998), 54–78; Matsuda Kyōko, Teikoku, 82–165; Sakano, Teikoku, 227–
39; Paul D. Barclay, “Contending Centers of Calculation in Colonial Taiwan: The Rhetorics
of Vindicationism and Privation in Japan’s ‘Aborigine Policy,’ ” Humanities Research 14, no.
1 (2007): 67–84; Yang Nanjun, Maboroshi no jinruigakusha: Mori Ushinosuke, trans. and
ed. Kasahara Masaharu, Miyaoka Maoko, and Miyazaki Seiko (Tokyo: Fūkyōsha, 2005);
and Chen Wei-chi, Yineng Jiaju: Taiwan lishi minzuzhi de zhankai (Taipei: Taida chuban
zhongxin, 2014).
122.  The GMD picked up the nine-tribe classification in the postwar period; the clas-
sification was not amended until 2001, when the GMD was defeated by the Democratic
Progressive Party, led by non-Mainlanders. Hu Chia-yu, “Embodied Memories,” 45.
123.  Marui Keijirō, Buban ni kansuru ikensho, Bando Kyōiku Ikensho (Taipei: Banmu
honsho, 1914), 21–24.
124. Sōyama Takeshi, Shokuminchi Taiwan to kindai tsūrizumu (Tokyo: Seikyōsha,
2003).
125. Niezen, Origins of Indigenism, 4.
126.  Prasenjit Duara, Sovereignty and Authenticity: Manchukuo and the East Asian Mod-
ern (Lanham: Rowman & Littlefield, 2003), 9–14, 77–78.
127.  Chang Wei-chi, Ueda Akira, and Miyazaki Kiyoshi, “Nihon tōchi jidai ni okeru
Taiwan no Genjūmin kankō no keisei,” Sōgō kankō kenkyū 2 (2003): 47–55.
128.  Arif Dirlik, The Postcolonial Aura: Third World Criticism in the Age of Global Capi-
talism (Boulder, CO: Westview Press, 1997), 227.
260     Notes

129.  Hu Chia-yu, “Embodied Memories,” 12. Material quoted from Marshall Sahlins,
“What Is Anthropological Enlightenment? Some Lessons of the Twentieth Century,” An-
nual Reviews of Anthropology 28 (1999): x.
130.  See Frederick Cooper, “Modernizing Colonialism and the Limits of Empire,” in
Lessons of Empire, ed. Calhoun, Cooper, and Moore, 63–72.
131.  Achille Mbembe, “Necropolitics,” trans. Libby Meintjes, Public Culture 15, no. 1
(2003): 26.

C HA P T E R 1 .   F R OM W E T D I P L OM AC Y T O S C O R C H E D E A RT H

1.  Ikoma Takatsune, “Musha-ban sōjō jiken chōsa fukumeisho,” Taiwan Musha hōki jik-
en: kenkyū to shiryō, ed. Tai Kuo-hui (Tokyo: Shasōshi kaisha, 1981), 290–304; Igarashi Ishi-
matsu, Musha jiken jikki (Puli: Taiwan keisei shinpōsha Puli shikyoku, 1931), 6–8; Mikami
Tamotsu, Taiwan, Musha jiken no konjaku: hisan na jiken to sono uramen shi (Chigasaki,
Kanazawa: self-published, 1984), 53–55.
2.  Taiwan sōtokufu, Musha jiken no tenmatsu in Gendaishi shiryō 22: Taiwan (II), ed.
Yamabe Kentarō (Tokyo: Misuzu Shobō, 1971), 587; Mikami, Taiwan, 51–55; Mukōyama
Hiroo, Taiwan Takasagozoku no kōnichi hōki: Musha jiken (Tokyo: Chūō keizai kenkyūjo,
1999), 43–44; Deng Xiangyang, Kōnichi Musha jiken no rekishi, ed. Shimomura Sakujirō,
trans. Uozumi Etsuko (Osaka: Nihon Kikanshi Shuppan Sentaa, 2000), 46–47; 95–99; Chou
Wan-yao, A New Illustrated History of Taiwan, trans. Carole Plackitt and Tim Casey (Taipei:
SMC Publishing Co., 2015), 177–78.
3. Qiu Ruolong, Wushe shijian: Taiwan yuanzhumin lixi manhua (Taipei: Tianyuan
chengshi chuban, 2001), 112–21; Wei Te-sheng, director, and John Woo, producer, Warriors
of the Rainbow: Seediq Bale (Taipei: Central Motion Picture Corporation and ARS Film
Production, 2011).
4.  Robert J. Lifton, History and Human Survival: Essays on the Young and Old, Survivors
and the Dead, Peace and War, and on Contemporary Psychohistory (New York: Random
House, 1970), 101–11.
5.  For details, see Zheng Anxi, Rizhi shiqi fandi aiyong xiande tuijin yubian qian (1895–
1920) (PhD diss., National Cheng-chi University [Taipei], 2011).
6.  Mahmood Mamdani, Citizen and Subject: Contemporary Africa and the Legacy of
Late Colonialism (Princeton: Princeton University Press, 1996).
7.  Leonard H. D. Gordon, Confrontation over Taiwan: Nineteenth-Century China and
the Powers (Lanham, MD: Lexington Books, 2007), 55.
8.  Sophia Su-fei Yen, Taiwan in China’s Foreign Relations, 1836–1874 (Hamden, CT: Shoe
String Press, 1965), 125; Robert Gardella, “From Treaty Ports to Provincial Status, 1860–
1894,” in Taiwan: A New History, ed. Murray A. Rubinstein (Armonk, NY: M. E. Sharpe,
1999), 167. Shao-hua Liu and Shu-min Huang, “The Damming of Mudan Creek,” in Envi-
ronment, Modernization and Development in East Asia: Perspectives from Environmental
History, ed. Ts’ui-jung Liu and James Beattie (London: Palgrave Macmillan, 2016), 113. A
list of specific wrecks can be found in James Davidson, The Island of Formosa Past and
Present (Yokohama: Japan Times Newspaper Company, 1903), 180–82. For another list of
shipwrecks off the southern coast, see Douglas Fix, “The Changing Contours of Lived Com-
munities on the Hengchun Peninsula,1850–1874,” in Guojia yu yuanzhumin: Ya-Tai diqu
 Notes    261

zuqun lishi yanjiu, ed. Hong Liwan (Taipei: Institute of Taiwan History, Academia Sinica,
2009), 278–80.
9.  Chou Wan-yao, “Cong Liuqiuren chuannan shouhai dao mudanshi shijian: ‘Xin’ cail-
iao ye duoyuan quanshi de keneng,” Taiwan fengwu 65, no. 2 (2015): 23–81.
10.  Miyaguni Fumio, Taiwan sōnan jiken (Naha: Naha Shuppansha, 1998).
11. Ōhama Ikuko, “ ‘Botansha jiken’ saikō: Naze Paiwan-zoku wa Ryūkyūtō-min o
satsugaishita no ka,” Taiwan Genjūmin kenkyū 11 (2007): 203–23.
12.  Valjeluk Mavaliu, “ ‘Botan-sha jiken’ ni tuite no shiken,” trans. Miyazaki Seiko, Tai-
wan Genjūmin kenkyū 10 (2006): 38–52; “Bazuroku (Valjeluk) kara no kotoba: shinrai to
kibō,” trans. Ishigaki Naoki, Taiwan Genjūmin kenkyū 11 (2007): 224–28.
13. Gao Jiaxin, “Sinvaudjan kara mita Botan jiken ue,” trans. Satoi Yōichi, Ryūkyū
Daigaku kyōiku gakubu kiyō 72 (March 2008): 41–63; “Sinvaudjan kara mita Botan jiken
shita,” trans. Satoi Yōichi, Ryūkyū Daigaku kyōiku gakubu kiyō 73 (August 2008): 27–50.
14. Miyaguni, Taiwan sōnan jiken, 287; Chou, “Cong Liuqiuren,” 27; Yen, Taiwan, 157.
15.  Mavaliu, “Bazuroku,” 225–26.
16.  Gao, “Sinvaudjan kara mita Botan jiken ue,” 47; Sugiyama Yasunori, ed., Taiwan
meisho kyūseki-shi (Tokyo: Taiwan Government General, 1916), 251.
17.  Xu Shirong, “Qingmo dao rizhi chuqi Hengchun difang de zuqun fenbu (1870–
1900),” Yuyan wenhua fenbu yu zuqun qianxi gongzuo fang (Taipei: Taiwan Normal Uni-
versity, 2012), 27.
18.  Mavaliu, “ ‘Botan-sha jiken,’ ” 45.
19.  Chou, “Cong Liuqiuren,” 28–30; Sugiyama, Taiwan meisho kyūseki-shi, 249–52.
20.  Mavaliu, “Bazuroku,” 38–52.
21.  Ōhama, “ ‘Botansha jiken’ saikō,” 203–29.
22.  Gao, “Sinvaudjan kara mita Botan jiken shita,” 28.
23.  Xu, “Qingmo,” 38.
24.  Chou, “Cong Liuqiuren,” 25–26.
25.  Edwin Pak-wah Leung, “The Quasi-War in East Asia: Japan’s Expedition to Taiwan
and the Ryūkyū Controversy,” Modern Asian Studies 17, no. 2 (1983): 263–64.
26.  James L. Hevia, Cherishing Men from Afar: Qing Guest Ritual and the Macartney
Embassy of 1793 (Durham: Duke University Press, 1995), 15–20, 116–33.
27.  Mary Elizabeth Berry’s discussion about the circulation of bodies in marriage, adop-
tion, and hostage politics and about political attachments solidified by elaborate gifting
practices in sixteenth-century Japan speaks eloquently to this point; see “Public Peace and
Private Attachment: The Goals and Conduct of Power in Early Modern Japan,” Journal of
Japanese Studies 12, no. 2 (1986): 259–68.
28.  Takeshi Hamashita, “Tribute and Treaties: Maritime Asia and Treaty Port Net-
works in the Era of Negotiation, 1800–1900,” in The Resurgence of East Asia: 500, 150 and
50 Year Perspectives, ed. Giovanni Arrighi, Takeshi Hamashita, and Mark Selden (London:
Routledge, 2003), 26–47. Also see Kirk W. Larsen, Tradition, Treaties, and Trade Qing Im-
perialism and Chosŏn Korea, 1850–1910 (Cambridge: Harvard East Asian Monographs,
2011), 13.
29.  Michael R. Auslin, Negotiating with Imperialism: The Unequal Treaties and the Cul-
ture of Japanese Diplomacy (Cambridge: Harvard University Press, 2006), 36–38; Gordon,
Confrontation over Taiwan, 56–57, 73.
262     Notes

30.  Edward J. Drea, Japan’s Imperial Army: Its Rise and Fall, 1853–1945 (Kansas City:
University of Kansas Press, 2009), 36.
31.  Norihito Mizuno, “An Aspect of Modern Japan’s Overseas Expansionism: The Tai-
wanese Aboriginal Territories in the Early Meiji Japanese Perspective,” Oriental Archive 78
(2010): 177–79.
32.  W. G. Beasley, “The Foreign Threat and the Opening of the Ports,” in The Cambridge
History of Japan Volume 5: The Nineteenth Century, ed. Marius B. Jansen (Cambridge: Cam-
bridge University Press, 1989), 267–73; Peter Booth Wiley, Yankee in the Land of the Gods:
Commodore Perry and the Opening of Japan (New York: Penguin Books, 1990), 501–3.
33.  Beasley, “Foreign Threat,” 274–93.
34.  Charles W. LeGendre, Notes of Travel in Formosa, ed. D. L. Fix and J. Shufelt (Tainan:
National Museum of Taiwan History, 2012), 281, 292.
35. Davidson, Island of Formosa, 115–16; LeGendre, Notes, 251–53; Gordon, Confronta-
tion over Taiwan, 56–58.
36.  Fix, “Changing Contours,” 235–37.
37. LeGendre, Notes, 254–55; Gordon, Confrontation over Taiwan, 59.
38.  Liu Mingdeng and Wu Dating, “Taotai and General of Formosa to LeGendre, June
3, 1867,” in Foreign Adventurers and the Aborigines of Southern Taiwan, 1867–1874, ed. Robert
Eskildsen (Taipei: Institute of Taiwan History, Academia Sinica, 2005), 276–77. The section
of Article Thirteen to which the Taiwan officials referred is the following, it would seem: “If
the merchant vessels of the United States, while within the waters over which the Chinese
Government exercises jurisdiction, be plundered by robbers or pirates, then the Chinese
­local authorities, civil and military, on receiving information thereof, shall arrest the said
robbers or pirates, and punish them according to law.” Treaty of Tianjin (Tien-tsin), 1858,
University of Southern California US-China Institute, accessed January 13, 2017, http://china.
usc.edu/treaty-tianjin-tien-tsin-1858.
39. Yen, Taiwan, 132; LeGendre, Notes, 257–61.
40.  Xu, “Qingmo,” 3–4.
41. LeGendre, Notes, 289.
42. Hevia, Cherishing, 74–118.
43.  Susan Naquin and Evelyn S. Rawski, Chinese Society in the Eighteenth Century (New
Haven: Yale University Press, 1987), 213.
44. Yen, Taiwan, 133.
45. LeGendre, Notes, 257–60.
46.  Bruce Greenfield, “The Problem of the Discoverer’s Authority in Lewis and Clark’s
History,” in Macropolitics of Nineteenth-Century Literature: Nationalism, Exoticism, Im-
perialism, ed. Jonathan Arac and Harriet Ritvo (Philadelphia: University of Pennsylvania
Press), 12–36.
47.  Kamimura Tōru, “Parijarijao shuchō-koku’ dai shuchō no zōyo kōkan keitai no ten-
kei to sono henkei to kussetsu tentō (zenpen): 1867-nen kara 1872-nen made no Taiwan
nanbu kōshun chihō no rekishi jinruigakuteki kōsatsu,” Taiwan Genjūmin kenkyū 18 (2014):
38–74.
48.  Kamimura Tōru, “ ‘Kōshun,’ ” 101–22; Xu, “Qingmo,” 15–16.
49.  W. A. Pickering, Pioneering in Formosa (London: Hurst and Blackett, 1898), 191.
50.  Fix, “Changing Contours,” 255.
 Notes    263

51. LeGendre, Notes, 277–78.

52. LeGendre, Notes, 280–81.
53. LeGendre, Notes, 283.
54. Douglas Cassel, “Letter from Douglas Cassel to Charles LeGendre,” in Foreign
­Adventurers, ed. Eskildsen, 207.
55. Pickering, Pioneering, 195. As Kamimura Tōru writes, the Qing host threw Toketok
and his followers into “panic mode”; see Kamimura, “Parijarijao,” 69–70.
56.  Beasley, “Foreign Threat,” 270–71; Peter Duus, The Japanese Discovery of America: A
Brief History with Documents (Boston: Bedford Books, 1997), 92–96; LeGendre, Notes, 291–92.
57. LeGendre, Notes, 289–93, 309–24.
58. Yen, Taiwan, 141–42; for diplomatic correspondence on competing views between
the Qing officials, see Eskildsen, ed., Foreign Adventurers; regarding Langqiao’s Paiwan resi-
dents, see LeGendre, Notes, 261–77.
59. Yen, Taiwan, 146.
60.  Edward H. House, The Japanese Expedition to Formosa (Tokyo: 1875), 222; Inō Ka-
nori, Taiwan banseishi (Taipei: Taiwan sōtokufu minseibu shokusankyoku, 1904), 615–17.
61. Davidson, Island of Formosa, 169.
62.  Fix, “Changing Contours,” 240.
63. LeGendre, Notes, 309–11.
64.  Shufelt and Fix, “Persons and Corporations, 1850–1875 Appearing in Notes of Travel
in Formosa,” in LeGendre, Notes, 415.
65. LeGendre, Notes, 311–12.
66. LeGendre, Notes, 310–16.
67. LeGendre, Notes, 316.
68. LeGendre, Notes, 310–14.
69.  Douglas Fix, “Charles LeGendre’s Travels in Formosa: A Listing of Itineraries,” in
Charles LeGendre, Notes, 442–48.
70.  Marlene Mayo, “The Korean Crisis of 1873 and Early Meiji Foreign Policy,” Journal of
Asian Studies 31, no. 4 (1975): 798–801; Hamashita, “Tribute and Treaties,” 46–47.
71.  Namahiri Tsuneo, Kindai Higashi Ajia-shi no naka no Ryūkyū heigō (Tokyo: Iwa-
nami shoten, 2014), 121–23.
72.  Matsunaga Masayoshi, “Taiwan ryoyūron no keizu:1874 (Meiji 7) nen no Taiwan
shuppei o chūshin ni,” Taiwan kingendai kenkyū 1 (1978): 8; Danny Orbach, “By Not Stop-
ping”: The First Taiwan Expedition (1874) and the Roots of Japanese Military Disobedi-
ence,” Journal of Japanese Studies 42, no. 1 (2016): 33.
73.  Chen Xuan, Meiji Nihon ni okeru Taiwanzō no keisei: Shinbun media ni yoru 1874
‘Taiwan jiken’ no hyōshō (Taipei: National Taiwan University Press, 2013), 13–33; Matsunaga,
“Taiwan ryoyūron no keizu,” 8; Yen, Taiwan, 162–70.
74.  Wayne C. McWilliams, “East Meets West: The Soejima Mission to China, 1873,”
Monumenta Nipponica 30, no. 3 (1975): 241–43; Yen, Taiwan, 180.
75.  McWilliams, “East Meets West,” 244; Yen, Taiwan, 155–56.
76. Namahiri, Kindai Higashi Ajia-shi, 194–95; Matsunaga, “Taiwan ryoyūron no keizu,”
9; Chen Xuan, Meiji Nihon, 31.
77.  Matsunaga, “Taiwan ryoyūron no keizu,” 9; Yen, Taiwan, 188; Chang Lung-chih,
“From Island Frontier.”
264     Notes

78.  McWilliams, “East Meets West,” 265–73; Yen, Taiwan, 187–90; Namahiri, Kindai Hi-
gashi Ajia-shi, 194–95.
79.  Gordon, 198–200.
80.  McWilliams, “East Meets West,” 274–75; Namahiri, Kindai Higashi Ajia-shi, 204–5.
81.  Charles LeGendre, Memo No. 24 to Saigō Tsugumichi, March 31, 1874, in Japan Cen-
ter for Asian Historical Records, National Archives of Japan, A03030001600, accessed on
January 12, 2016, https://www.digital.archives.go.jp/das/image/M0000000000000901592.
82.  Xu, “Qingmo,” 20, 26; Chou, “Cong Liuqiuren,” 75; Fix, “Changing Contours,” 264–
65; House, Japanese Expedition, 30–50; Cassel, “Letter,” 204; LeGendre, Memo No. 24.
83. Quote from House, Japanese Expedition, 65–66 (emphasis added); Mizuno Jun,
“Seiban shiki,” in Tairo Mizuno Jun Sensei, ed. Tairokai (Taipei: Tairokai jimusho, 1930),
224, concurs with this version of events, as does Cassel, “Letter,” 205–7.
84. Fujisaki Seinosuke, Taiwan no Banzoku (Tokyo: Kokushi Kankōkai, 1930), 527;
House, Japanese Expedition, 123; Cassel, “Letter,” 209–10; Taiwan sōtokufu keimukyoku, ed.,
Takasagozoku chōsasho dai go-hen: Bansha gaikyō meishin (Taipei, 1938), 307.
85.  Cassel, “Letter,” 212.
86. House, Japanese Expedition, 102.
87.  Cassel, “Letter,” 213.
88.  Michel Foucault, Discipline and Punish: The Birth of the Prison, trans. Alan Sheridan
(New York: Vintage Books, 1977), 3–5.
89. House, Japanese Expedition, 102–3; Cassel, “Letter,” 215; Shidehara Hiroshi, Nanpō
bunka no kensetsu e (Tokyo: Fuzanbō, 1938), 358–59.
90. House, Japanese Expedition, 133; Cassel, “Letter,” 208–9.
91. Sugiyama, Taiwan meisho kyūseki-shi, 265.
92.  Fujisaki lists the flag numbers and the names of chiefs and villages for over thirty of
the flags distributed between June 19 and September 20; according to his count, the fifty-
third and last flag was distributed on the latter date; see Taiwanshi to Kabayama taishō
(­Tokyo: Kokushi kankōkai, 1926), 490–93; Yamamoto Un’ichi, “Kōgun no banjin buiku ni
tsuite (shita),” Riban no tomo, June 1, 1934, 4. For illustrations, see Paul D. Barclay, “The
­Relics of Modern Japan’s First Foreign War in Colonial and Postcolonial Taiwan, 1874–2015,”
Cross Currents: East Asian History and Culture Review E-Journal 17 (December 2015): 131–38.
93. House, Japanese Expedition, 214–15; Mizuno, “Seiban shiki,” 281–82; Suzuki Sakutarō,
Taiwan no banzoku kenkyū (Taipei: Taiwanshi Shiseki Kankōkai, 1932), 506–8; Shidehara,
Nanpō bunka, 363–64.
94. House, Japanese Expedition, 114–25.
95.  Orbach, “By Not Stopping,” 52.
96.  Mizuno, “Seiban shiki,” 291–95; Fujisaki, Taiwanshi, 494.
97.  Chen Xuan, Meiji Nihon, 125–31; Fujisaki, Taiwanshi, 494.
98.  Orbach, “By Not Stopping,” 48.
99.  Mark Ravina’s intellectual biography of Saigō Takamori documents the difficulty of
this conversion process for a major political figure educated under the dynastic system. The
Last Samurai: The Life and Battles of Saigō Takamori (Hoboken, NJ: John Wiley and Sons,
2004).
100.  Chen Xuan, Meiji Nihon, 13–21.
101. Yen, Taiwan, 203; Orbach, “By Not Stopping,” 46.
 Notes    265

102. Yen, Taiwan, 254.

103.  Jürgen Osterhammel, The Transformation of the World: A Global History of the
Nineteenth Century (Princeton: Princeton University Press, 2014), 394–95.
104. Yen, Taiwan, 256.
105.  Byron K. Marshall, Learning to Be Modern: Japanese Political Discourse on ­Education
(Boulder, CO: Westview Press, 2004), 48.
106.  All quotations from Yen, Taiwan, 256.
107.  Takekoshi Yosaburō, Japanese Rule in Formosa, trans. George Braithwaite (London,
New York, Bombay, and Calcutta: Longmans, Green, 1907), 230.
108. Yen, Taiwan, 252, 261, 268; Alexis Dudden, Japan’s Colonization of Korea: Discourse
and Power (Honolulu: University of Hawai‘i Press, 2005), 46–51.
109. House, Japanese Expedition, 190–215.
110. LeGendre, Notes, 295.
111.  LeGendre, “Has Japan the Right to Assume Suzerainty over Aboriginal Formosa?,”
in Foreign Adventurers, 181–85; LeGendre, Notes, 331.
112. Yen, Taiwan, 257–58.
113.  Hayashi Masako, “Ueno Sen’ichi: Nisshin sensō mae no Taiwan ninshiki no senku-
sha,” Taiwan kin-gendaishi kenkyū 2 (1979): 30–60.
114. Ueno Sen’ichi, “Taiwantō jissen roku,” Tokyo chigaku kyōkai hōkoku 13, no. 11
(­February 1892): 21–48; Sanbō honbu, ed., Taiwan shi (Tokyo: Sanbō honbu, January 1895),
83–85.
115. LeGendre, Notes, 49–52, appended map.
116.  William Hancock, “Tamsui Trade Report for the Year 1881,” 32–40, archived at Doug-
las Fix, ed., Formosa: 19th Century Images (Portland, OR: Reed College), accessed on August
2, 2016, http://cdm.reed.edu/cdm4/document.php?CISOROOT = /formosa&CISOPTR =
2264&REC = 12.
117.  Sanbō honbu, Taiwan shi, 83–85; Ueno, “Taiwantō jissen roku,” 39–42.
118.  In many of the details of his transit and the descriptions of his interpreter, Ueno’s
reports follow Hancock’s so closely that they appear to be plagiarized in parts.
119.  Ueno, “Taiwantō jissen roku,” 42–45; Sanbō honbu, Taiwan shi, 90–91.
120.  “Tōhō kyōkai no ensetu,” Yomiuri shinbun, March 22, 1892, 1.
121.  “Chigaku kyōkai no kōwa,” Yomiuri shinbun, March 25, 1892, 1.
122.  Ueno Sen’ichi, “Taiwantō shisatsu fukumei dai yon gō: Taiwantō seiban fūzoku,”
Bōei kenkyūjo toshokan, Tokyo, Japan. Document 10-Taiwan-M28–1.
123.  Sanbō honbu, Taiwan shi.
124. See chapter 3.
125.  Jiryū Yamabito, “Taiwan zakki nijū-san: Seiban kaidō (dai ikkai),” Tokyo Asahi shin-
bun, September 28, 1895, 2; “Seiban no junfuku,” Fūzoku gahō, November 28, 1895, 18.
126. Tanaka Tsunatoku, “Dakekan seiban kaiken Taipei chiji hōkoku,” in Taiwan
sōtokufu banzoku jijō kōbun ruisan genbun, vol. 1, ed. Huang Liyun (Museum of Ethnology,
Academia Sinica, Nankang, Taipei, Taiwan), n.p.
127.  Hashiguchi Bunzō, “Taiwan jijō,” Tokyō chigaku kyōkai hōkoku 17, no. 3 (October–
December 1895): 309–28.
128. Fujisaki, Taiwan no banzoku, 523; Mizuno, “Seiban shiki,” 199–200.
129.  Tanaka, “Dakekan,” n.p.
266     Notes

130.  Matsubara Iwagorō, “Seiban kōtsū shimatsu (ue),” Kokumin shinbun, September
28, 1895, 1; Matsubara Iwagorō, “Seiban kōtsū shimatsu (shita),” Kokumin shinbun, Septem-
ber 29, 1895, 1; Jiryū, “Taiwan zakki nijū-san,” 2; Hashiguchi, “Taiwan jijō,” 309–28; “Seiban
no junfuku,” Fūzoku gahō, 17–21; Davidson, Island of Formosa, 343.
131.  Wang Peng-hui, “Faces of Ethnic Others in Press: Daishizhai Pictorial’s Southern
Chinese Empire,” Taiwan Historical Research 19, no. 4 (2012): 114–17; Hashiguchi, “Taiwan
jijō,” 319–21.
132.  Jiryū, “Taiwan zakki nijū-san,” 2; Hashiguchi, “Taiwan jijō,” 317.
133.  “Taiwan shoken,” Yomiuri shinbun, October 15, 1895, 2.
134.  Hashiguchi, “Taiwan jijō,” 308.
135.  Inō Kanori and Awano Dennojō, Taiwan banjin jijō (Taipei: Taiwan sōtokufu min-
seibu bunshoka, 1900), 16.
136. Mori Ushinosuke, Taiwan banzokushi dai ikkan (Taipei: Rinji Taiwan kyūkan
chōsakai, 1917; repr., Taipei: Southern Materials Center, 1996), 217–18.
137.  Inō Kanori, “Juntai nichijō,” in Inō Kanori no Taiwan tōsa nikki, ed. Kazunari Mori-
guchi (Taipei: Taiwan fūbutsu zasshisha, 1992), 9.
138.  Inō, “Juntai,” 81, 109, 124.
139.  Okada Shinkō, “Arisanban chōsasho,” Taiwan kanshū kiji 5, no. 5 (1905): 379–80.
140.  Kawano’s manuscript report and Shibayama’s account agree on almost all par-
ticulars of this speech; however, the last line, enclosed in brackets here, is found only in
Shibayama’s Taiyō article and is not contained in the manuscript. Fujisaki summarized this
outing based on his extensive official archive in Taiwan no banzoku, 544–45.
141. Kawano Shuichirō, “Banjin kaiken Yilan shichō hōkoku,” in Huang, Taiwan
sōtokufu banzoku jijō; Shibayama Kakuzō, “Seiban kaikenki,” Taiyō 2, no. 4 (1896): 202–6.
142.  This was probably Pixo Sappo, who was chief of Upper Paalan, the foremost of
three Paalan towns (buraku). Inō, “Juntai,” 71–72.
143.  Wang Jiasheng, “Taiwan tsūshin: seibanjin no kōeki,” Yomiuri shinbun, February
27, 1896, n.p.
144. Fujisaki, Taiwan no banzoku, 544.
145.  Inō Kanori, ed., Riban shikō dai ikkan (Taipei: Taiwan sōtokufu keisatsu honsho,
1918; repr., Taipei: Southern Materials Center, 1995), 8–9.
146. Inō, Riban shikō, 10–11.
147.  Wang Shiqing, Qing dai Taiwan she hui jing ji (Taipei: Lian jing chu ban, 1994), 482.
Translation by Wu Haotian.
148. Inō, Riban shikō, 22, 125; Wang, Qing dai, 483.
149.  “Seiban no junfuku,” Fūzoku gahō, 18.
150. Miyanohara Tōhachi, “Seiban jijō Dakekan bukonsho no hōkoku,” in Taiwan
sōtokufu banzoku jijō, vol. 3; Miyanohara Tōhachi, “Dakekan bukonsho hōkoku,” Taiwan
shinpō, September 19, 1896, n.p. Miyanohara’s forty-four-page manuscript report was repro-
duced in its entirety in the Taiwan shinpō in serial form between September 19 and October
14, 1896.
151.  “Saitō bunkonsho shuji no kiso chihō shucchō,” Yomiuri shinbun, August 20, 1897, 3.
152.  Kyū shokuminchi jinji sōran: Taiwan hen, vol. 1 (Tokyo: Nihon tosho sentā, 1997),
78, 82–97.
153. Inō, Riban shikō, 14–15.
 Notes    267

154.  Ueno Shirō, “Shokuminchi tōchiki ni okeru bukonsho to Taiwan Genjūmin to

no kawari ni tsuite” in Mainoriti no koritsu to kokōsei ed. Chūkyō Daigaku shakai kagaku
kenkyūjo (Nagoya: Chūkyō Daigaku shakai kagaku kenkyūjo, 2002), 63–64.
155. Inō, Riban shikō, 41–42.
156.  Matsuda Kyōko, “ ‘Naichi’ kankō’ to iu tōchi gihō: 1897 nen no Taiwan Genjūmin no
‘naichi’ kankō o megutte,” Akademia [Nanzan daigaku] (2013): 85–103; Jordan Sand, “Impe-
rial Tokyo as a Contact Zone: The Metropolitan Tours of Taiwanese Aborigines, 1897–1941,”
Asia-Pacific Journal 12, issue 10, no. 4 (March 10, 2014), accessed February 26, 2017, http://
apjjf.org/2014/12/10/Jordan-Sand/4089/article.html.
157. Inō, Riban shikō, 42–43.
158. Inō, Riban shikō, 56–57.
159. Inō, Riban shikō, 70–71.
160.  Uchiyama Ban’yū, “Puli no tsuyu (Fukabori Taii gūnan no tenmatsu),” Taiwan
kanshū kiji 11, no. 5 (1905): 60–66.
161.  Kaku Kurata, “Chūō sanmyaku ōdan, ichi,” Taiwan nichinichi shinpō, January 24,
1908, n.p.
162.  Mori Ushinosuke, “Jūgonen no bankai, yon,” Taiwan nichinichi shinpō, May 4,
1912, n.p.
163. Inō, Riban shikō, 156.
164.  Takada Tomizō, ed., Taihoku-shū ribanshi (Taipei: Taihoku-shū keimu-bu, 1924),
304–9.
165.  Taiwan nichinichi shinpō, December 19, 1902, n.p.
166.  Gao Yongqing [Pixo Walis], Musha hizakura no kuruizaki: gyakusatsu jiken iki-
nokori no shogen, trans. Katō Minoru (Tokyo: Kyōbunkan, 1988), 13; Aui Heppaha, Shōgen
Musha jiken: Taiwan sanchijin no kōnichi hōki, ed. Xu Jielin (Tokyo: Sōfūkan, 1985), 20;
Deng, Kōnichi Musha jiken o muguru hitobito, 55.
167. Gao, Musha hizakura no kuruizaki, 13–14.
168.  Taiwan minpō, October 9, 1903, and October 21, 1903.
169. Inō, Riban shikō, 455–56; Banjin no dōyō oyobi tōbatsu no gairyaku, in Gendaishi
shiryō 22: Taiwan (II), ed. Yamabe Kentarō (Tokyo: Misuzu Shobō, 1971), 509.
170.  Charles Archibald Mitchell, Camphor in Japan and in Formosa (London: Chiswick
Press, 1900), 3, 45.
171. Davidson, Island of Formosa, 406.
172. Takekoshi, Japanese, 171.
173.  By 1890 world demand for the product had risen sharply because European gov-
ernments were buying up supplies for the production of smokeless powder; see “Camphor,”
Scientific American 62 (April 19, 1890): 242.
174.  Ka Chih-ming, Japanese Colonialism in Taiwan: Land Tenure, Development and De-
pendency, 1895–1945 (Boulder, CO: Westview Press, 1995), 35.
175.  Johanna Menzel Meskill, A Chinese Pioneer Family: The Lins of Wu-feng, Taiwan,
1729–1895 (Princeton: Princeton University Press, 1979), 239–42.
176. Inō, Riban shikō, 28–29.
177.  Iriye Takeshi, “Taiwan banchi zatsuzoku,” Fūzoku gahō (December 10, 1896): 26–32;
Meskill, Pioneer Family, 241–42.
178. Davidson, Island of Formosa, 430.
268     Notes

179.  Kojima Reiitsu, “Nihon teikokushugi no Taiwan sanchi shihai: Musha hōki jiken,”
in Taiwan Musha hōki jiken kenkyū to shiryō, ed. Tai Kuo-hui (Tokyo: Shakai shisōsha, 1981),
63–64; Formosa Bureau of Aboriginal Affairs, Report on the Control of the Aborigines in
Formosa (Taipei: Government of Formosa, 1911), 34–42.
180.  E. Patricia Tsurumi, “Taiwan under Kodama Gentarō and Goto Shimpei,” in Papers
on Japan, ed. Albert Craig (Cambridge: Harvard East Asian Research Center, 1967), 99–101.
181.  Kodama and Gotō are the “heroes” of Japanese colonization of Taiwan; they are
usually given credit for ending rebellion, making the colony profitable, and reducing the
military’s role on the island. Comparisons to Paul Doumer’s regime in French Indochina
come to mind when reading of their efficiency and ruthlessness; see Edward I. Chen, “Goto
Shimpei, Japan’s Colonial Administrator in Taiwan: A Critical Reexamination,” American
Asian Review 13, no. 1 (1995): 29–59.
182. Ka, Japanese Colonialism, 51–52.
183.  Mochiji’s post was within the inner circle of Taiwan’s central government. Only ten
other civil officials (including Gotō) shared Mochiji’s rank of chokunin. Tsurumi, “Taiwan
under Kodama,” 115; Ogata Taketoshi, ed., Taiwan dainenpyō (Taipei: self-published), 49.
184.  Kaneko Fumio, “Nihon shokuminchi jinbutsu (2): Mochiji Rokusaburō no shōgai
to chosaku,” Taiwan kingendaishi kenkyū 2 (1979): 119–28.
185. Inō, Riban shikō, 179–228.
186. Inō, Riban shikō, 180–90.
187.  Mochiji Rokusaburō, Taiwan shokuminchi seisaku (Tokyo: Fuzanbō, 1912), 380–81.
188.  Julia Adeney Thomas, Reconfiguring Modernity: Concepts of Nature in Japanese
­Political Ideology (Berkeley: University of California Press, 2001), 84–110.
189. Inō, Riban shikō, 181.
190. Inō, Riban shikō, 180.
191.  Ishii Shinji, “The Silent War in Formosa,” Asiatic Quarterly Review (July 1913): 90.
192.  Tōgō Minoru and Satō Shirō, Taiwan shokuminchi hattatsu shi (Taipei: Kōbunkan,
1916; repr., Southern Materials Center, 1996), 332; Tomoyuki Kawata, Formosa To-day
(­Osaka: Taikansha and Co., 1917), 40.
193.  Inoue Toshitaka, “Taiwan sōtokufu no chikkō jigyō,” in Nihon tōchi jidai Taiwan no
keizai to shakai, ed. Matsuda Yoshirō (Tokyō: Kōyō shobō, 2012), 157–59.
194. Fujii Shizue, Rizhi shiqi Taiwan zongdufu lifan zhengce (Taipei: Wenyingtang
­chubanshe, 1997), 228.
195. Kawata, Formosa To-day, 13.
196.  Y. Takenobu and K. Kawakami, The Japan Year Book 1913 (Tokyo: Japan Year Book
Office, n.d.), 669.
197.  “Sasaki Terayama-shi,” Tokyo Asahi shinbun, January 1934, 11.
198. Dai Nihon Teikoku gikaishi kankōkai, ed., Dai Nihon Teikoku gikaishi, vol. 7
(­Tokyo: Dai Nihon Teikoku gikaishi kankōkai, 1928), 905, 908–9.
199. Fujii, Rizhi shiqi, 228.
200.  Formosa Bureau of Aboriginal Affairs, Report, 45; Taiwan sōtokufu minseibu ban-
mu honsho, ed., Riban gaiyō (Taipei: Taiwan sōtokufu minseibu, 1912), appendix 4.
201. Kojima, “Nihon teikokushugi,” 62–3; see also Formosa Bureau of Aboriginal
­Affairs, Report, 16–17.
202.  Taiwan sōtokufu minseibu banmu honsho, appendix 3.
 Notes    269

203.  Kojima, “Nihon teikokushugi,” 62–63.

204. Government-General of Taiwan, The Statistical Summary of Taiwan (Tokyo: Japan
Times Co., 1912), 163.
205. Kawata, Formosa To-day, 13.
206.  Hideo Naito, Taiwan: A Unique Colonial Record 1937–8 Edition (Tokyo: Kokusai
Nippon kyōkai, 1938), 81.
207.  Caroline Hui-yu Ts’ai, “Shaping Administration in Colonial Taiwan,” in Taiwan
under Japanese Colonial Rule, 1895–1945: History, Culture, Memory, ed. Liao Ping-hui and
David Der-wei Wang (New York: Columbia University Press, 2006), 108.
208. Inō, Riban shikō, 397.
209. Inō, Riban shikō, 448–49.
210. “Nantō-chō keibu-ho Itō Eitarō banjin tōbatsu ni kanshi shōyo,” Taiwan Sōtokufu
Manuscript Records, 4938-ce/7-wen, 1906–04–01.
211. Taiwan sōtokufu minseibu banmu honsho, ed., Chiban kikō (Taipei: Taiwan
sōtokufu minseibu banmu honsho, 1911).
212.  Tōgō and Satō, Taiwan shokuminchi, 145–46.
213.  Takenobu Yoshitarō, Japan Year Book 1914 (Tokyo: Japan Year Book Office), 702.
214.  Tōgō and Satō, Taiwan shokuminchi, 148.
215. For a documentary/commemorative postcard of the Ninth Regiment’s buglers,
marchers, and the improvised arch, with troops returning to Puli in August 1914, after vic-
tory, see “[(93) The Expeditionary Force’s 9th Company Entering Puli’s Triumphal Arch],”
tj0095, East Asia Image Collection, Lafayette College, Easton, PA, accessed July 23, 2017,
http://digital.lafayette.edu/collections/eastasia/tjwar-postcards/tj0095. For a photo in the
popular press of Governor-General Sakuma’s triumphal return to Taipei that same year,
complete with arch and parade, see “The Triumph of Governor-General Sakuma at Taihoku
Station,” ts0022, East Asia Image Collection, Lafayette College, Easton, PA, accessed July 23,
2017, http://digital.lafayette.edu/collections/eastasia/cpw-shashinkai/ts0022.
216.  Wu Micha, ed., Taiwan-shi kojiten, trans. Yokosawa Yasuo (Tokyo: Chūgoku sho-
ten, 2007), 166.
217. Drea, Japan’s Imperial Army, 137; Mark Peattie, Nan’yō: The Rise and Fall of the Japa-
nese in Micronesia, 1885–1945 (Honolulu: University of Hawai‘i Press, 1988), 41–42.
218. Fujii, Rizhi shiqi, 274.
219. Zheng, Rizhi shiqi, 144.
220.  Tōgō and Satō, Taiwan shokuminchi, 145.
221.  Takenobu Yoshitarō, The Japan Year Book 1915 (Tokyo: Japan Year Book Office,
n.d.), 727.
222.  Takenobu Yoshitarō, The Japan Year Book 1916 (Tokyo: Japan Year Book Office,
n.d.), 708.
223.  Richard White, The Middle Ground: Indians, Empires, and Republics in the Great
Lakes Region, 1650–1815 (Cambridge: Cambridge University Press, 1991).
224.  Yamaji Katsuhiko, Taiwan Taiyaru-zoku no hyaku-nen: Hyōryū suru dentō, dakō
suru kindai, datsu shokuminchika e no michinori (Tokyo: Fūkyosha, 2011), 309; Matsuoka
Tadasu, Taiwan Genjūmin shakai no chihōka: Mainoriti no nijū seiki (Tokyo: Kenbun shup-
pan, 2012), 87; Matsuda Yoshirō, “Senjūmin to Nihongo kyōiku: Arisan Tsou-zoku no sen-
zen, sengo,” in Seikatsu no naka no shokuminchi shugi, ed. Mizuno Naoki (Kyoto: Jinbun
270     Notes

shoin, 2004), 153; Kondō Masami, Sōryokusen to Taiwan: Nihon shokuminchi hōkai no
kenkyū (Tokyo: tōsui shobō, 1996), 296–97.
225.  Ikoma, “Musha-ban sōjō jiken,” 289.
226.  Watanabe Sei, “Musha sōjō no shinsō o aku hitotsu no kagi! ‘Seiban Kondō’ shi no
hansei o monogataru (nijū-san),” Taiwan nichinichi shinpō, February 5, 1931, 5; Hashimoto
Hakusui, Aa Musha jiken (Taipei: Minamikuni shuppan kyōkai, 1930; repr., Taipei: Ch’eng
Wen, 1999), 147–50; Taiwan sōtokufu, Musha jiken no tenmatsu, 610–11.
227. Nakagawa Kōichi et al., Musha jiken: Taiwan Takasagozoku no hōki (Tokyo:
Sanshōdō, 1980), 59–77; Deng Xianyang, Kōnichi Musha jiken o meguru hitobito, ed. Shi-
momura Sakujirō, trans. Uozumi Etsuko (Ōsaka: Nihon kikanshi shuppan sentaa, 2001),
72–74; Hayashi Eidai, Taiwan Hitsuwa: Musha No Hanran, Minshūgawa No Shōgen (Tokyo:
Shin hyōron, 2002), 52–55; Mukōyama, Taiwan Takasagozoku, 36–37; Fujisaki, Taiwan no
banzoku, 848–49; Ikoma, 289; Mikami, Taiwan, 50–51.
228.  Taiwan sōtokufu, Musha jiken no tenmatsu, 611; Hashimoto, 148–49.
229.  Watanabe Sei, “Musha sōjō no shinsō o aku hitotsu no kagi! ‘Seiban Kondō’ shi no
hansei o monogataru,” Taiwan nichinichi shinpō, February 7, 1931, 3.

C HA P T E R 2 .   T H E LON G U E DU R É E A N D T H E SHO RT C I R C U I T

1.  Jürgen Osterhammel, The Transformation of the World: A Global History of the Nine-
teenth Century (Princeton: Princeton University Press, 2014), 785–90; Benedict Anderson,
Imagined Communities (London: Verso, 1983), 43.
2.  As recently as 2014, Taiwan’s “national Proficiency Test of Aboriginal Languages—
used to certify students seeking indigenous status in order to be eligible for government af-
firmative-action policies—[was] offered in a total of forty-two different language varieties.”
P. Kerim Friedman, “The Hegemony of the Local: Taiwanese Multiculturalism and Indig-
enous Identity Politics,” Boundary 2 (forthcoming, 2018). Taiwan’s national government em-
ployed a team of forty-eight translators to translate President Ts’ai Ing-wen’s August 1, 2016,
apology to the IPT into sixteen different languages. Wu Po-wei and Jake Chung, “Apology to
be Published in 18 Languages,” Taipei Times, December 23, 2016, 4, accessed January 2, 2017,
http://www.taipeitimes.com/News/taiwan/archives/2016/12/23/2003661776.
3.  The Margary affair of February 21, 1875, which occurred within three months of Ja-
pan’s departure from Taiwan, provides yet another example of how deadly violence, in-
flicted by peoples within the Qing empire but beyond its administered territories, could
become an occasion for a costly foreign intervention with long-term consequences for the
Qing. In this case, the death of a single British functionary at the hands of Burmese hill
tribes brought the full weight of the British crown to bear on the Qing court, resulting in a
round of treaty negotiations and public-apology missions. See James Hevia, English Lessons:
The Pedagogy of Imperialism in Nineteenth-Century China (Durham: Duke University Press,
2003), 150–53.
4. Stewart Lone, Army, Empire and Politics in Meiji Japan (New York: Palgrave,
2000), 17.
5.  Douglas R. Reynolds, “Training Young China Hands: Tōa Dōbun Shoin and Its Pre-
cursors, 1886–1945,” in The Japanese Informal Empire in China, 1895–1937, ed. Peter Duus,
Ramon H. Myers, and Mark R. Peattie (Princeton: Princeton University Press), 215–21;
 Notes    271

Joshua A. Fogel, Politics and Sinology: The Case of Naitō Konan (1866–1934) (Cambridge:
Harvard University Press, 1984), 12.
6.  Sasa Hiroo, “Nisshin sensō to tsūyakukan,” in Nisshin sensō to Higashi Ajia sekai no
henyō, ed. Ōhata Tokushirō (Tokyo: Yumani shobō, 1997), 373–90.
7.  Mori Ushinosuke, “Taiwan banzoku ni tuite,” in Taiwan banzokushi dai ikkan (Taipei:
Rinji Taiwan kyūkan chōsakai, 1917; repr., Taipei: Southern Materials Center, 1996), 2.
8.  Sasa, “Nisshin sensō to tsūyakukan,” 373–90.
9.  Ōe Shinobu, “Shokuminchi sensō to sōtokufu no seiritsu,” in Iwanami kōza: Kindai
Nihon to shokuminchi, Teikoku tōji no kōkoku, vol. 2 (Tokyo: Iwanami Shoten, 1992), 3–10.
10.  Furuno Naoya, Taiwangun shireibu: 1895–1945 (Tokyo: Kokusho kankōkai, 1991),
58–77.
11. “Kenpeitai-zuki tsūyakukan no saiyō,” Yomiuri shinbun, September 9, 1895, 2.
12.  Taiwan sōtokufu keimukyoku, ed., Taiwan sōtokufu keisatsu enkakushi (Taipei: Tai-
wan sōtokufu keimukyoku, 1938; repr., Tokyo: Rokuin shobō, 1986), 5: 912. Hereafter ab-
breviated TSK.
13.  He Dongquan, “Taiwan nansei roku: shoken zakki,” Yomiuri shinbun, September 1,
1895, 1.
14.  Figures compiled and analyzed from pages of Kyūshokuminchi jinji sōran: Taiwan
hen, vol. 1 (Tokyo: Nihon tosho sentā, 1997), hereafter abbreviated KJS.
15.  TSK 5: 914–15.
16.  TSK 5: 916.
17. Lone, Army, 52.
18.  “Dogo senkō iin,” Taiwan kyōkai kaihō 9 (1899): 61.
19.  TSK 5: 916–18.
20.  For the best illustration of the budgeting shell game that disguised diminished per
capita spending on education under the banner of “increased participation,” see Kitamura
Kae, Nihon shokuminchika no Taiwan Genjūmin kyōiku shi (Sapporo: Hokkaidō Daigaku
shuppankai, 2008), 238–40. One of many Japanese postmortem investigations into the
causes of the Wushe uprising also noted the neglect of aborigine administration in the 1920s
as a problem. See Ikoma Takatsune, ed., “Musha-ban sōjō jiken chōsa fukumeisho,” in Tai
Kuo-hui, ed., Taiwan Musha hōki jiken: kenkyū to shiryō (Tokyo: Shasōshi kaisha, 1981), 318.
21.  Kondō Masami, Sōryokusen to Taiwan: Nihon shokuminchi hōkai no kenkyū (Tokyo:
tōsui shobō, 1996), 263.
22.  Yamaji Katsuhiko, Taiwan Taiyaru-zoku no hyaku-nen: Hyōryū suru dentō, dakō
suru kindai, datsu shokuminchika e no michinori (Tokyo: Fūkyosha, 2011), 307–11; Matsuoka
Tadasu, Taiwan Genjūmin shakai no chihōka: Mainoriti no nijūsseiki (Tokyo: Kenbun shup-
pan, 2012), 86–88.
23. Kondō, Sōryokusen, 294–99; Matsuda Kyōko, Teikoku no shisō: Nihon “Teikoku” to
Taiwan Genjūmin (Tokyo: Yūshisha, 2014), 189–90; Matsuda Yoshirō, “Senjūmin to Nihon-
go kyōiku: Arisan Tsou-zoku no senzen, sengo,” in Seikatsu no naka no shokuminchi shugi,
ed. Mizuno Naoki (Kyoto: Jinbun shoin, 2004), 153.
24.  Matsuda, “Senjūmin,” 158–59.
25.  Robert Swinhoe, “Notes on the Ethnology of Formosa,” in Natives of Formosa: Brit-
ish Reports of the Taiwan Indigenous People, 1650–1950, ed. Henrietta Harrison (Taipei:
Shung Ye Museum of Formosan Aborigines, 2001), 66.
272     Notes

26.  Douglas Fix, “The Changing Contours of Lived Communities on the Hengchun
Peninsula,1850–1874,” in Guojia yu yuanzhumin: Ya-Tai diqu zuqun lishi yanjiu, ed. Hong
Liwan (Taipei: Institute of Taiwan History, Academia Sinica, 2009), 271.
27.  Xu Shirong, “Qingmo dao rizhi chuqi Hengchun difang de zuqun fenbu (1870–
1900),” Yuyan wenhua fenbu yu zuqun qianxi gongzuo fang (Taipei: Taiwan Normal Uni-
versity, 2012), 20–24.
28.  W. A. Pickering, Pioneering in Formosa (London: Hurst and Blackett, 1898), 143.
29.  Joseph Beal Steere, Formosa and Its Inhabitants, ed. Paul Jen-kuei Li (Taipei: Insti-
tute of History [Preparatory Office], Academia Sinica, 2002), 314.
30.  Paul D. Barclay, “Cultural Brokerage and Interethnic Marriage in Colonial Taiwan:
Japanese Subalterns and Their Aborigine Wives, 1895–1930,” Journal of Asian Studies 64, no.
2 (May 2005): 323–60.
31.  Chang, Lung-chih, “From Island Frontier to Imperial Colony: Qing and Japanese
Sovereignty Debates and Territorial Projects in Taiwan, 1874–1906” (PhD diss., Harvard
University, 2003), 105–7.
32. Kurosaki Michio, “Seibanchi Marai-sha tanken (shūzen),” Tokyo Asahi shinbun,
March 4, 1896, 3; Jiang Guizhen, Zaixian chuangtong de shijian, Wulai Taiyazu de wenhua
tuxiang (Taipei: National Museum of History, 2010), 58.
33.  Kirsten L. Ziomek, “The Possibility of Liminal Colonial Subjecthood: Yayutz Bleyh
and the Search for Subaltern Histories in the Japanese Empire,” Critical Asian Studies 47, no.
1 ( 2015): 123–50.
34.  Chang, “Island Frontier,” 95–101.
35.  Fujisaki Seinosuke, Taiwan no banzoku (Tokyo: Kokushi Kankōkai, 1930), 455.
36.  Inō Kanori, Taiwan banseishi (Taipei: Taiwan sōtokufu minseibu shokusankyoku,
1904), 551–54.
37.  Hashiguchi Bunzō, “Taiwan jijō,” Tokyo chigaku kyōkai hōkoku 17, no. 3 (1895): 311–17.
38.  Jiryū Yamabito, “Taiwan zakki nijū-san: Seiban kaidō (dai ikkai),” Tokyo Asahi shin-
bun, September 28, 1895, 2; “Seiban no junfuku,” Fūzoku gahō 103 (November 28, 1895),
17–18.
39.  Hashiguchi, “Taiwan jijō,” 313–18; Hashiguchi Bunzō, “Dakekan chihō seiban kaiken
no tame shutchō jihōkoku,” in Taiwan sōtokufu banzoku jijō kōbun ruisan genbun, vol. 1,
ed. Huang Liyun (Museum of Ethnology, Academia Sinica, Nankang, Taipei, Taiwan), n.p.;
Tanaka Tsunatoku, “Dakekan seiban kaiken Taipei chiji hōkoku,” in Taiwan sōtokufu ban-
zoku jijō kōbun ruisan genbun, vol. 1; Inō Kanori, “Taiwan tsūshin dai sankai,” Tokyo jin-
ruigakkai zasshi 119 (1896): 179–84; Inō Kanori, ed., Riban shikō dai ikkan (Taipei: Taiwan
sōtokufu keisatsu honsho, 1918; repr., Taipei: Southern Materials Center, 1995), 4; Jiryū Yam-
abito, “Taiwan zakki nijū-yon: Seiban kaidō (dai nikai),” Tokyo Asahi shinbun, September
29, 1895, 2.
40.  Hashiguchi, “Taiwan jijō,” 317–18.
41.  Mori, “Taiwan banzoku ni tuite,” features extended discussions of cross-cultural
misapprehension surrounding terms like surrender and alliance between Japanese and
Atayalic speakers.
42.  Kawano Shuichirō, “Banjin kaiken Yilan shichō hōkoku,” in Taiwan sōtokufu ban-
zoku jijō kōbun ruisan genbun, vol. 1; Shibayama Kakuzō, “Seiban kaikenki,” Taiyō 2, no. 4
 Notes    273

(1896): 203–5; Takada Tomizō, ed., Taihoku-shū ribanshi, vol. 1 (Taipei: Taihoku-shū keimu-
bu, 1924), 1; Fujisaki, Taiwan no Banzoku, 544–45.
43.  Inō Kanori, “Taiwan tsūshin dai ikkai,” Tokyo jinrui gakkai zasshi 117 (1895): 97–98.
44. Kawano Shūichirō, “Banjin kaiken,” n.p.; Shibayama, “Seiban kaikenki,” 202–6;
Takada, Taihoku-shū ribanshi,1.
45.  Nakajima Takemuro, “Seibanchi tankenki jō,” Taiyō 2, no. 21 (1896): 247–48.
46.  Saitō Kenji, “Taiwan no shōnō seizō (setsuzen gō),” Taiwan kyōkai kaihō 4 (1899): 15–16.
47.  Iriye Takeshi, “Taiwan banzoku zue,” Fūzoku gahō 129 (December 1, 1896): 17.
48. Mori, Taiwan banzoku shi, 176–77.
49.  Laura Lee Junker, Raiding, Trading, and Feasting: The Political Economy of Philippine
Chiefdoms (Honolulu: University of Hawai‘i Press, 1999); Matsuoka Tadasu, “Gendai Tai-
wan Genjūmin shakairon chūtàn: ‘Tōmoku’ o tegakari to shite,” Taiwan Genjūmin Kenkyū
8 (2004): 83–84.
50.  Margery Wolf, Women and the Family in Rural Taiwan (Stanford: Stanford Univer-
sity Press, 1972), 171–90.
51.  Saitō Otosaku, “Morrison kikō (zoku),” Taiyō 3, no. 14 (July 5, 1897): 180.
52.  “Taiwan minji shibu: tsūyakukan no Taiwan tsūshin,” Yomiuri shinbun, September
30, 1895, 5.
53.  Iriye Takeshi and Hashimoto Shigeru, “Taiwan banchi zatsuzoku,” Fūzoku gahō 130
(1896): 29–30; Chiwas is identified in Mori Ushinosuke, “Sūsū no kishiki ban monogatari,”
Taiwan nichinichi shinpō, June 21, 1925, 3; Pixo Sappo is identified in Mori Ushinosuke,
“Bankai no konjaku (jū-ni),” Taiwan nichinichi shinpō, July 18, 1915, and Inō Kanori, “Jun-
tai nichijō,” in Inō Kanori no Taiwan tōsa nikki, ed. Kazunari Moriguchi (Taipei: Taiwan
fūbutsu zasshisha, 1992), 73.
54.  Araki Masayasu, ed., Shinbun ga kataru meijishi, vol. 2 (Tokyo: Hara shobō, 1979), 79.
Meiji Japan’s most revered student of northern aborigine languages and culture, Mori Ushi-
nosuke, recalled Hiyama’s marriage to a Wushe banpu as the first official Japanese action to
assert itself among the Atayal of Nantou Prefecture (Taiwan nichinichi shinpō, May 4, 1912).
55.  “Hiyama Tetsusaburō-shi no donchi,” Yomiuri shinbun, May 19,1895, 3.
56.  “Hiyama Tetsusaburō-shi no gōtō kyōsa kengi,” Yomiuri shinbun, June 1, 1897, 3; “Hi-
yama Tetsusaburō kōin ni tsuite,” Yomiuri shinbun, June 11, 1897, 3.
57.  Suimoto Sei’ichi, Taiwan hitsuwa (Taipei: Nihon oyobi shokumin sha, 1928), 259.
58.  Torii Ryūzō, “Taiwan chūō sanmyaku no ōdan: shōzen,” Taiyō 7, 9 (1901): 131; Uchi-
yama Ban’yū, “Puli no tsuyu (Fukahori Taii gūnan no tenmatsu),” Taiwan kanshū kiji 11,
no. 5 (1905): 66; Inoguchi Yasuyoshi, ed., Riban shikō dai nikan (Taipei: Taiwan sōtokufu
keimukyoku, 1921; repr., Taipei: Southern Materials Center, 1995), 232.
59. Suimoto, Taiwan hitsuwa, 259–61. This account has many chronological inconsis-
tencies; it sounds like an oral history based on Kondō’s memory; it was published two years
before the Wushe Rebellion. It is therefore not completely reliable.
60. Yamamoto Ryōichi, “Kaitakusha Kondō Katsusaburō to sono shūi: Miyatake
Shōtarō-shi no genkō o moto ni,” Gendai Taiwan kenkyū 19 (2000): 107–8.
61.  Japanese sources use the word dohi (brigand, bandit) to describe rebels against the
state, since they refused to impute any but the basest motives to the rebels’ acts of sabotage,
robbery, and resistance to Japanese occupation.
274     Notes

62. Takekoshi, Japanese Rule, 93–94.

63. Inō, Riban shikō, 22.
64.  Taiwan nichinichi shinpō, December 21, 1930.
65.  Inō Kanori identified him as Tsitsok, father of Iwan. Inō was introduced to Iwan,
who led him to Tsitsok, by an interpreter (probably Kondō) whom Inō met near Puli; the
interpreter’s name is left blank in Inō’s diary. Inō, “Juntai,” 71.
66.  Taiwan nichinichi shinpō, December 21, 1930; Ide Kiwata, Taiwan chisekishi (Taipei:
Taiwan nichinichi shinpōsha, 1937), 278.
67.  Taiwan nichinichi shinpō, December 21, 1930; Fujisaki, Taiwan no banzoku, 606.
68.  Taiwan nichinichi shinpō, December 25, 1930; Yamabe Kentarō, ed., Gendaishi shiryō
22: Taiwan (II) (Tokyo: Misuzu Shobō, 1971), 509; Toyonaga Tōhei, “Aa, Fukahori Tai’i,”
Riban no tomo (January 1, 1936): 2–3.
69.  Taiwan nichinichi shinpō, December 25, 1930.
70.  Taiwan nichinichi shinpō, January 13–14, 1931; Fujisaki, Taiwan no banzoku,
614–16.
71. Higashiyama Kyoko, “Taiwan ryōyūki ni okeru tai-Genjūmin seisaku: Fukaho-
ri Yasuichirō tankentai to Nagano Yoshitora ikensho kara no ikkō satsu,” Shakai kagaku
kenkyū 32, no. 2 (2012): 283–86.
72. “Puli-sha shishō-chō hokuban junshi heiko Fukahori Tai’i ichigyō sōnan ibutsu
hakken hōkoku,” Taiwan sōtokufu kōbun ruisan, 4627-ce, 2-wen, 1900–05–01.
73.  “Hokuban daishū-chō no shukushi,” Taiwan kyōkai kaihō 1 (1898): 102; “Hokuban
ko’ō kettō no zu,” Taiwan kyōkai kaihō 6 (1899): 76–77.
74.  Inō, “Juntai,” 7.
75.  Xia Shengli, et al., Hyakunen Cangsang: Dogura Ryūjirō to Taihoku Kameyama sui-
ryoku hatsudenjo, trans. and ed. Dogura Masao and Dogura Nobuko (Sakurai-shi: Dogura
Masao, 2006).
76.  “Seiban no raisha,” Taiwan nichinichi shinpō, January 30, 1903.
77. “Jiku Shō Min yori seiban,” Taiwan sōtokufu kōbun ruisan, 4534-ce, 2-wen,
1897–07–01.
78.  Torii Ryūzō, “Taiwan chūō sanmyaku no ōdan,” Taiyō 7, no. 9 (August 5, 1901): 131–32;
Inō, “Juntai,” 70–71.
79.  Torii Ryūzō, “Keimenban joshi no tōkei,” 555–56; “Horisha hōmen nite chōsaseshi
jinruigakuteki jikō,” 525–29; “Taiwan Horisha [Musha] ban [tōbu yūgeimenban] no shin-
wa,” in Torii Ryūzō zenshū 11 (Tokyo: Asahi shinbun, 1976), 557–58.
80. Inō, Riban shikō, 148–49.
81.  KJS 1: 22–23.
82.  KJS 1: 78, 82–97.
83.  “Saitō bunkonsho shuji no kiso chihō shucchō,” Yomiuri shinbun, August 20, 1897, 3.
84. Inō, Riban shikō, 105.
85. Inō, Riban shikō, 45.
86.  Mio Yūko, “‘Bango hensan hōshin’ kara mita Nihon tōchi shoki ni okeru Taiwan
Genjūmingo chōsa,” Nihon Taiwan gakkaihō 11 (2009): 157–69.
87. Inō, Riban shikō, 96–97; TSK 5: 915.
88. Kitamura, Nihon shokuminchika, 176–77.
89. Inō, Riban shikō, 504–55; TSK 5: 923–24.
 Notes    275

90.  At the time of this writing, there are sixteen officially recognized indigenous groups
in Taiwan. P. Kerim Friedman has termed the proliferation of localism the “hegemony of the
local” and refers to these subgeobodies as “fractal.” His larger argument is that redefinition
of “locality” in Taiwan along ethnic lines within the population of non-Mainlanders has dif-
fused anti-GMD nationalism by atomizing locality to the point where it becomes harmless in
terms of national-scale political mobilization. Friedman asserts that such “fractal recursivity”
has its origin in the Japanese colonial period. Friedman, “The Hegemony of the Local.”
91. Inō, Riban shikō, 707–9.
92. Inō, Riban shikō, 455–56.
93. Inō, Riban shikō, 457.
94.  Zheng Anxi, Rizhi shiqi fandi aiyong xiande tuijin yubian qian (1895–1920) (PhD
diss., National Cheng-chi University [Taipei], 2011).
95. “Nantō-chō keibu-ho Itō Eitarō banjin tōbatsu ni kanshi shōyo,” Taiwan sōtokufu
kōbun ruisan, 4938-ce/7-wen, 1906–04–01.
96.  For a description of the guardline, its staffing, and its movements, see chapter 2.
97. Inō, Riban shikō, 533–34.
98.  Watanabe Sei, “Musha sōjō no shinsō o aku hitotsu no kagi! ‘Seiban Kondō’ shi no
hansei o monogataru (nijū-ichi),” Taiwan nichinichi shinpō, February 2, 1931, 5.
99.  Watanabe Sei, “Musha sōjō no shinsō o aku hitotsu no kagi! ‘Seiban Kondō’ shi no
hansei o monogataru (nijū-ni),” Taiwan nichinichi shinpō, February 3, 1931, 5.
100.  Watanabe Sei, “Musha sōjō no shinsō o aku hitotsu no kagi! ‘Seiban Kondō’ shi no
hansei o monogataru (nijū-san),” Taiwan nichinichi shinpō, February 5, 1931, 5.
101. Inō, Riban shikō, 638–39; Deng Xianyang, Kōnichi Musha jiken o meguru hitobito,
ed. Shimomura Sakujirō, trans. Uozumi Etsuko (Ōsaka: Nihon kikanshi shuppan sentaa,
2001), 20.
102.  Aui Heppaha, Shōgen Musha jiken: Taiwan sanchijin no kōnichi hōki, ed. Xu Jielin
(Tokyo: Sōfūkan, 1985), 179–80.
103. Watanabe Sei, “Musha sōjō no shinsō o aku hitotsu no kagi! ‘Seiban Kondō’
shi no hansei o monogataru (nijū-kyū),” Taiwan nichinichi shinpō, February 15, 1931, 5;
“Kanyū genya kashiwatasu kyoka (Kondō Katsusaburō),” Taiwan sōtokufu kōbun ruisan,
6821-ce/6-wen.
104.  Igarashi Ishimatsu, Musha jiken jikki (Puli: Taiwan keisei shinpōsha Hori shikyoku,
1931), 118–21.
105. Yamabe, Gendaishi shiryō, 694.
106.  Watanabe Sei, “Musha sōjō no shinsō o aku hitotsu no kagi! ‘Seiban Kondō’ shi no
hansei o monogataru (nijū-hachi),” Taiwan nichinichi shinpō, February 10, 1931, 5.
107.  Paul D. Barclay, “ ‘They Have for the Coast Dwellers a Traditional Hatred’: Govern-
ing Igorots in Northern Luzon and Central Taiwan, 1895–1915,” in The American Colonial
State in the Philippines: Global Perspectives, ed. Julian Go and Anne Foster (Durham, NC:
Duke University Press, 2003), 232.
108.  Hayashi Eidai, Taiwan hitsuwa: Musha no hanran, minshūgawa no shōgen (Tokyo:
Shinhyōron, 2002), 43–45, 103.
109.  Yanagimoto Michihiko, Taiwan Musha ni ikiru (Tokyo: Gendai shokan, 1996), 39;
Hayashi, Taiwan hitsuwa, 56–57.
110. Yanagimoto, Taiwan Musha, 43–47; Hayashi, Taiwan hitsuwa, 57–58.
276     Notes

111.  Mikami Tamotsu, Taiwan Musha jiken no konjaku: hisan na jiken to sono uramen shi
(Chigasaki, Kanazawa: self-published, 1984), 110–11.
112. Hayashi, Taiwan hitsuwa, 84.
113.  Taiwan sōtokufu, “Musha jikenshi,” in Taiwan Musha hōki jiken: kenkyū to shiryō,
ed. Tai Kuo-hui (Tokyo: Shasōshi kaisha, 1981), 374.
114. Mikami, Taiwan Musha jiken, 108.
115. Hayashi, Taiwan hitsuwa, 64.
116. Taiwan sōtokufu, “Musha jikenshi,” 375; Mikami, Taiwan Musha jiken, 109–10;
Yamabe, Gendaishi shiryō, 614–15.
117.  Chiang Yu-lan, “Japanese Policies Toward the Formosan Aborigines: The Wu-she
Incident of 1930” (MA thesis, University of Oklahoma, 1974), 30.
118. Hayashi, Taiwan hitsuwa, 51.
119.  E. Patricia Tsurumi, Japanese Colonial Education in Taiwan, 1895–1945 (Cambridge:
Harvard University Press, 1977), 16.
120.  Edward H. House, The Japanese Expedition to Formosa (Tokyo, 1875), 64; “Letter
from Douglas Cassel to Charles LeGendre,” in Foreign Adventurers and the Aborigines of
Southern Taiwan, 1867–1874, ed. Robert Eskildsen (Taipei: Institute of Taiwan History, Aca-
demia Sinica, 2005), 206.
121.  Zhuang Yongming, Taiwan fang qing song 5: Taiwan yuan zhu min (Hong Kong:
Yuan-Liou Publishing Co., 2001), 83–90; Abe Akiyoshi, Taiwan chimei kenkyū (Taipei: Ban-
go kenkyūkai, 1937), 283–84; Matsuda Yoshirō, Taiwan Genjūmin no Nihongo kyōiku (Kōyō
shobō, 2004), 8.
122.  Miyaguni Fumio, Taiwan sōnan jiken (Naha: Naha Shuppansha, 1998), 311.
123.  Inō Kanori, “Ryōtaigo no riban (shōzen),” Tōyō jihō (February 1912): 42–43; Suzuki
Sakutarō, Taiwan no banzoku kenkyū (Taipei: Taiwan shiseki kankōkai, 1932; repr., Taipei:
Southern Materials Center, 1988), 512.
124.  Rokkoki Shōgorō, “Taitōshō kannai shisatsu fukumeisho,” Zhongguo fangzhi con-
gshu: Taiwandiqu, no. 311 (Taipei: Chengwen, 1984), n.p.; Inō Kanori, “Langqiao no Daiban-
moku Pan Bunketsu,” Taiwan kyōkai kaihō 88 (1906): 21–22; Sagara Nagatsuna, “Taitō fukin
banjin kijun kōshun shutchōjo hōkoku,” in Taiwan sōtokufu banzoku jijō kōbun ruisan gen-
bun, vol. 1, ed. Huang Liyun (Museum of Ethnology, Academia Sinica, Nankang, Taipei,
Taiwan), n.p.
125.  Takekoshi Yosaburō, Japanese Rule in Formosa, trans. George Braithwaite (London,
New York, Bombay, and Calcutta: Longmans, Green, 1907), 93.
126.  Sagara, “Taitō fukin,” n.p.
127.  Matsuda Yoshirō, Taiwan Genjūmin, 42.
128.  Sagara, “Taitō fukin,” n.p.; Inō, Riban shikō, 157.
129.  “Taiwanjin jokun,” Tokyo Asahi shinbun, December 18, 1897, 2.
130. Inō, Riban shikō, 157–58; Inō, “Langqiao no Daibanmoku,” 21–22.
131. Suzuki, Taiwan no banzoku kenkyū, 512–13.
132. Kitamura, Nihon shokuminchika, 69–72.
133.  Byron K. Marshall, Learning to Be Modern: Japanese Political Discourse on Educa-
tion (Boulder, CO: Westview Press, 1994), 47–49; Leonard H. D. Gordon, Confrontation
over Taiwan: Nineteenth-Century China and the Powers (Lanham, MD: Lexington Books,
2007), 68.
 Notes    277

134. Kitamura, Nihon shokuminchika, 68.

135.  Xu, “Qing modao.”
136.  Inō, “Juntai,” 118–19.
137.  Inō Kanori and Awano Dennojō, Taiwan banjin jijō (Taipei: Taiwan sōtokufu min-
seibu bunshoka, 1900), 222.
138. Jun’eki Taiwan Genjūmin kenkyūkai, ed., Inō Kanori shozō Taiwan Genjūmin
shashinshū (Taipei: Jun’eki Taiwan Genjūmin hakubutsukan, 1999), 183–84.
139. Inō, Riban shikō, 326–32.
140.  Xu, “Qing modao,” 15.
141.  Taiwan sōtokufu keimukyoku, ed., Takasagozoku chōsasho, vol. 3 (Taipei: Taiwan
sōtokufu keimukyoku, 1937), 11.
142. Shao-hua Liu and Shu-min Huang, “The Damming of Mudan Creek,” in Environ-
ment, Modernization and Development in East Asia: Perspectives from Environmental His-
tory, ed. Ts’ui-jung Liu and James Beattie (London: Palgrave Macmillan, 2016), 120–21.
143.  Matsuda Yoshirō, Taiwan Genjūmin, 11.
144. Tsurumi, Japanese Colonial Education, 18.
145. Department of Educational Affairs of the Government-General of Formosa,
­Japan, ed., A Review of Educational Work in Formosa 1916, in Hattori Hitsuji, ed., Nihon
shokuminchi kyōiku seisaku shiryō shūsei (Taiwan-hen) dai ikkan (Tokyo: Ryūkei shosha,
2007), 38.
146. Kitamura, Nihon shokuminchika, 70; Matsuda Yoshirō, Taiwan Genjūmin, 47–49.
147.  Department of Educational Affairs, Review of Educational Work, 39.
148.  In 1905, the Tuilasok institute was incorporated into the nearby Mantsui Com-
mon School; Kuskus, on the other hand, opened an Aborigine Common School that same
month. Kitamura, Nihon shokuminchika, 70.
149.  Taiwan sōtokufu kanbō bunsho-ka, The Progress of Taiwan (Formosa) for Ten Years,
1895–1904 (Taipei: Taiwan sōtokufu kanbō bunsho-ka, 1906), 43.
150. Kitamura, Nihon shokuminchika, 277, 291.
151.  Eika Tai, “The Assimilationist Policy and the Aborigines in Taiwan Under Japanese
Rule,” Current Politics and Economics of Asia 6, no. 4 (1999): 277.
152.  Tai, “Assimilationist Policy,” 279–80.
153. Kitamura, Nihon shokuminchika, 112–14; Kondō, Sōryokusen, 292.
154. Inō, Riban shikō, 849–50.
155. Ōe Shinobu, “Shokuminchi sensō to sōtokufu no seiritsu,” in Iwanami kōza:
­Kindai Nihon to shokuminchi, Teikoku tōji no kōkoku, vol. 2 (Tokyo: Iwanami Shoten,
1992), 3–10.
156. This text is reproduced in the original Japanese in Matsuda Yoshirō, Taiwan
Genjūmin, 53 (translation by author).
157.  Sakano Tōru, Teikoku Nihon to jinruigakusha, 1884–1952 (Tokyo: Keishobo, 2005).
158. Tsurumi, Japanese Colonial Education, 20–21.
159. Inō, Taiwan bansei-shi, 52; Xu, “Qing modao,” 36.
160.  Matsuda Yoshirō, Taiwan Genjūmin, 53–54.
161. Kitamura, Nihon shokuminchika, 114.
162. Hayashi, Taiwan hitsuwa, 64.
163.  Matsuda Yoshirō, Taiwan Genjūmin, 55.
278     Notes

164.  Banjin dokuhon maki san (Taipei: Taiwan sōtokufu, 1916), 45–47; Banjin dokuhon
maki yon (Taipei: Taiwan sōtokufu, 1917), 20, 34. All four volumes are reprinted in Li Rong-
zhi and Chen Shuying, ed., Banjin dokuhon (Fukushima: Kurume Daigaku, 2001).
165.  Tai, “Assimilationist Policy,” 285.
166. Kitamura, Nihon shokuminchika, 238–43.
167.  Uno Toshiharu, “Taiwan ni okeru ‘banjin’ no kyōiku,” in Taiwan Musha hōki jiken
kenkyū to shiryō, ed. Tai Kuo-hui (Tokyo: Shakai shisōsha, 1981), 44–45.
168.  Chou Wan-yao, trans. Carole Plackitt and Tim Casey, A New Illustrated History of
Taiwan (Taipei: SMC Publishing Co., 2015), 185–86; Deng Xiangyang, Kōnichi Musha jiken
no rekishi, trans. Shimomura Sakujirō and Uozumi Etsuko (Osaka: Nihon kikanshi shup-
pan sentaa, 2000), 69–71.
169.  Matsuda Yoshirō, “Senjūmin,” 160.
170. Matsuda Yoshirō, “Taiwan senjūmin,” 160; Kitamura, Nihon shokuminchika,
250–52.
171. Kondō, Sōryokusen, 291–93.
172.  Schools under regular administration were paid for out of local tax moneys and
student fees. Kitamura, Nihon shokuminchika, 238; Tsurumi, Japanese Colonial Education,
240–41.
173. Kondō, Sōryokusen, 295.
174. Kitamura, Nihon shokuminchika, 174–78.
175.  The composition and functions of the youth corps are described in Aoki Setsuzō,
Haruka naru toki Taiwan: Senjūmin shakai ni ikita aru Nihonjin keisatsukan no kiroku (Osa-
ka: Kansai tosho shuppan, 2001), 128–34; Matsuda Yoshirō, Taiwan Genjūmin no shakai
teki kyōiku jigyō (Tokyo: Kōyōsho, 2011), 91–120; Kondō, Sōryokusen, 294–305; Matsuoka,
Taiwan Genjūmin, 82–85.
176. Kondō, Sōryokusen, 262–300; Matsuda Kyōko, Teikoku, 190–92.
177.  Leo T. S. Ching, “Savage Construction and Civility Making: Japanese Colonialism
and Taiwanese Aboriginal Representation,” Positions: East Asia Cultures Critique 8, no. 3
(2000): 810–15.
178.  U.S. Consul Gerald Warner and his wife, Rella, toured the main sites on the indig-
enous ethnic-tourism route in 1939, just after Sayon’s death. Several of Warner’s snapshots
at inns and on trails suggest that Sayon was not exceptional as a female burden-carrier for
hire. See http://digital.lafayette.edu/collections/eastasia/warner-negs-taiwan/gr0277.
179.  Tai, “Assimilationist Policy,” 273.
180.  Tai, “Assimilationist Policy,” 281.
181.  Fujino Yohei, “Nihon tōchika Taiwan ni okeru tainichi kanjō: Kanzoku to Genjūmin
no hikaku kara,” Minzoku bunka kenkyū 5 (2004): 52–54; Wu Micha, ed., Taiwan-shi kojiten,
trans. Yokosawa Yasuo (Tokyo: Chūgoku shoten, 2007), 164.
182.  Wu Micha referred to this event as the watershed that divided Han resistance into
an earlier period of armed uprisings and a later period of nonviolent sociopolitical move-
ments, Taiwan-shi, 173.
183.  Paul R. Katz, When Valleys Turned Blood Red: The Ta-pa-ni Incident in Colonial
Taiwan (Honolulu: University of Hawai‘i Press, 2005).
184.  Paul R. Katz, “Governmentality and Its Consequences in Colonial Taiwan,” Journal
of Asian Studies 64, no. 2 (2005): 389.
 Notes    279

185.  Ts’ai, “Shaping Administration in Colonial Taiwan,” in Taiwan under Japanese Colo-
nial Rule, 1895–1945: History, Culture, Memory, ed. Liao Ping-hui and David Der-wei Wang
(New York: Columbia University Press, 2006), 97–98.
186. Ming-cheng M. Lo, Doctors within Borders: Profession, Ethnicity, and Modernity in
Colonial Taiwan (Berkeley: University of California Press, 2002), 37.
187.  Wang Taysheng, Legal Reform in Taiwan under Japanese Colonial Rule, 1895–1945:
The Reception of Western Law (Seattle and London: University of Washington Press, 2000).

C HA P T E R 3 .   TA N G L E D U P I N R E D

1.  James Davidson, The Island of Formosa Past and Present (Yokohama: Japan Times
Newspaper Company, 1903), 173–74.
2.  James L. Hevia, English Lessons: The Pedagogy of Imperialism in Nineteenth-Century
China (Durham, NC, 2003), 31–48; Robert Gardella, “From Treaty Ports to Provincial Sta-
tus, 1860–1894,” in Taiwan: A New History, ed. Murray A. Rubinstein (Armonk, NY: M. E.
Sharpe, 1999), 167.
3. Yŏnghŏ Ch’ŏe, “Yanagi Muneyoshi and the Kwanghwa Gate in Seoul, Korea,” in
Sources of Japanese Tradition, 2nd ed., vol. 2, abridged pt. 2, ed. Wm. Theodore de Bary,
Carol Gluck, and Arthur E. Tiedemann, ed. (New York: Columbia University Press, 2006),
144–47; Yuko Kikuchi, “Refracted Colonial Modernity: Vernacularism in the Development
of Modern Taiwanese Crafts,” in Refracted Modernity: Visual Culture and Identity in Colo-
nial Taiwan, Yuko Kikuchi, ed. (Honolulu: University of Hawai‘i Press, 2007), 222–29.
4.  C. W. LeGendre, Reports on Amoy and the Island of Formosa (Washington, DC: Gov-
ernment Printing Office, 1871), 33.
5.  Thomas Francis Hughes, “Visit to Tok-e-Tok, Chief of the Eighteen Tribes, Southern
Formosa,” in Aborigines of South Taiwan in the 1880s, ed. Glen Dudbridge (Taipei: Shung
Ye Museum of Formosan Aborigines and Institute of Taiwan History, Academia Sinica), 32.
6.  Charles W. LeGendre, Notes of Travel in Formosa, ed. D. L. Fix and J. Shufelt (Tainan:
National Museum of Taiwan History, 2012), 267–69.
7.  Douglas Fix, “The Changing Contours of Lived Communities on the Hengchun Pen-
insula,1850–1874,” Guojia yu yuanzhumin: Ya-Tai diqu zuqun lishi yanjiu, ed. Hong Liwan
(Taipei: Institute of Taiwan History, Academia Sinica, 2009), 246–47.
8.  Edward H. House, The Japanese Expedition to Formosa (Tokyo, 1875), 143.
9. House, Japanese Expedition, 143.
10. House, Japanese Expedition, 141.
11.  “Letter from Douglas Cassel to Charles LeGendre,” in Foreign Adventurers and the
Aborigines of Southern Taiwan, 1867–1874, ed. Robert Eskildsen (Taipei: Institute of Taiwan
History, Academia Sinica, 2005), 205, 209.
12.  Tania Murray Li, “Indigeneity, Capitalism, and the Management of Dispossession,”
Current Anthropology 51, no. 3 (June 2010): 385–414, draws on several examples from colo-
nized Southeast Asia to demonstrate how outsiders’ estimations of indigenous economic
innocence, incompetence, or corporatism became reified in the form of separate legal re-
gimes that shielded them from market forces.
13.  Kamimura Tōru, “Parijarijao shuchō-koku’ dai shuchō no zōyo kōkan keitai no ten-
kei to sono henkei to kussetsu tentō (zenpen): 1867-nen kara 1872-nen made no Taiwan
280     Notes

nanbu kōshun chihō no rekishi jinruigakuteki kōsatsu,” Taiwan Genjūmin kenkyū 18 (2014):
38–74.
14.  Antonio C. Tavares, “The Japanese Colonial State and the Dissolution of the Late
Imperial Frontier Economy in Taiwan, 1886–1909,” Journal of Asian Studies 64, no. 2 (2005):
361–85.
15.  Kondō Masami, “Taiwan sōtokufu no riban taisei to Musha Jiken,” Iwanami kōza:
Kindai Nihon to shokuminchi, Teikoku tōji no kōkoku, vol. 2 (Tokyo: Iwanami Shoten, 1992),
43–44; Melissa L. Meyer, “ ‘We Can Not Get a Living as We Used To’: Dispossession and the
White Earth Anishinaabeg, 1889–1920,” American Historical Review 96, no. 2 (1991): 384.
16.  Scott Simon, “Making Natives: Japan and the Creation of Indigenous Formosa,” in
Japanese Taiwan: Colonial Rule and Its Contested Legacy, ed. Andrew D. Morris (London:
Bloomsbury Press, 2015), 75–92.
17.  Hu Chia-yu, “Taiwanese Aboriginal Art and Artifacts: Entangled Images of Coloni-
zation and Modernization,” in Refracted Modernity: Visual Culture and Identity in Colonial
Taiwan, ed. Kikuchi, 201.
18.  Yamaji Katsuhiko, Taiwan Taiyaru-zoku no hyaku-nen: Hyōryū suru dentō, dakō suru
kindai, datsu shokuminchika e no michinori (Tokyo: Fūkyosha, 2011), 402–5; Hu, “Taiwan-
ese,” 201–2; Kikuchi, “Refracted Colonial Modernity,” 218–22; Tierney, Tropics of Savagery:
The Culture of Japanese Empire in Comparative Frame (Berkeley: University of California
Press, 2010), 64–65.
19.  Wakabayashi Masahiro, “A Perspective on Studies of Taiwanese Political History:
Reconsidering the Postwar Japanese Historiography of Japanese Colonial Rule in Taiwan,”
in Liao Ping-hui and David Der-wei Wang, eds., Taiwan under Japanese Colonial Rule,
1895–1945 (New York: Columbia University Press, 2006), 19–36.
20.  Nicholas Thomas, Entangled Objects: Exchange, Material Culture, and Colonialism
in the Pacific (Cambridge: Harvard University Press, 1991).
21.  John L. Comaroff and Jean Comaroff, Ethnicity, Inc. (Chicago: University of Chicago
Press, 2009).
22.  Kamimura Tōru, “‘Kōshun shitaban’ shuchōsei no seikaku: mae—Nihon ryōyūki
Taiwan nanbu sanchi no kosumorojii no henbyō,” Nanpō bunka 24 (1997): 101–22.
23.  Henrietta Harrison, “Clothing and Power on the Periphery of Empire: The Cos-
tumes of the Indigenous People of Taiwan,” Positions 11, no. 2 (2003): 337.
24.  Mami Yoshimura, “Weaving and Identity of the Atayal in Wulai, Taiwan” (mas-
ter’s thesis, University of Waterloo, 2007), 166–67; Tamoto Haruna, “ ‘Genjūmin kōgei’ no
hyōshō to seisaku o meguru ikkōsatsu: Taiwan Genjūmin no orimono fukkō o jirei ni,”
Shigaku (Mita shi gakkai) 81, no. 3 (2012): 97, 104; Yamaji, Taiwan Taiyaru-zoku no hyaku-
nen, 402–5.
25.  Hu, “Taiwanese,” 207–11.
26.  P. Kerim Friedman, “The Hegemony of the Local: Taiwanese Multiculturalism and
Indigenous Identity Politics,” boundary 2 (forthcoming); Haidy Geismar, Treasured Posses-
sions: Indigenous Interventions into Cultural and Intellectual Property (Durham and London:
Duke University Press), 2013; Tomoto, “Genjūmin kōgei,” 92; Hu, “Taiwanese,” 210–11.
27.  Hu Chia-yu, “Embodied Memories and Enacted Ritual Materials: Possessing the
Past in Making and Remaking Saisiyat Identity in Taiwan,” (PhD diss., University College
London, 2006), 233–238.
 Notes    281

28.  Laurence G. Thompson, “The Earliest Chinese Eyewitness Accounts of the Formo-
san Aborigines,” Monumenta Serica 23 (1964): 177–78.
29.  Leonard Blussé, Natalie Everts, and Evelien French, eds., The Formosan Encounter:
Notes on Formosa’s Aboriginal Society: A Selection of Documents from Dutch Archival Sources
(Taipei: Shung Ye Museum of Formosan Aborigines, 1999), 1.
30.  Blussé, Everts, and French, Formosan Encounter, 29 (emphasis added).
31.  Wayne C. McWilliams, “East Meets East: The Soejima Mission to China, 1873,” Mon-
umentica Nipponica 30, no. 3 (1975): 237–75; Mizuno Jun, “Seiban Shiki,” in Tairo Mizuno
Jun Sensei, ed. Tairokai (Nagoya: Tairokai, 1930), 188.
32.  Mizuno, “Seibanki,” 188–93.
33.  George Taylor, “A Ramble through Formosa,” in Aborigines of South Taiwan in the
1880s: Papers by George Taylor, ed. Glen Dudbridge (Taipei: Shung Ye Museum of Formosan
Aborigines and Institute of Taiwan History, Academia Sinica, 1999), 122–23.
34.  Ueno Sen’ichi, “Taiwan shima jissen roku,” Tokyo chigaku kyōkai hōkoku 13, no. 11
(1892): 21–48; Sanbō Honbu, ed., Taiwan shi (Hiroshima: Sanbō honbu, 1895).
35.  Tokyo Asahi shinbun, September 28, 1895; Fūzoku gahō, November 28, 1895.
36.  Hashiguchi Bunzō, “Dakekan chihō seiban kaiken no tame shucchō jihōkoku,” Tai-
wan sōtokufu banzoku jijō kōbun ruisan genbun, vol. 1, ed. Huang Liyun (Taipei: Museum
of Ethnology, Academia Sinica); Hashiguchi Bunzō, “Taiwan jijō,” Tokyo chigaku kyōkai
hōkoku 17, no. 3 (1895): 313–18.
37.  Hashiguchi, “Taiwan jijō,” 313–14.
38.  Hashiguchi Bunzō, “Banmin ni shinamono keiyo no gi hōkoku,” in Taiwan sōtokufu
banzoku jijō kōbun ruisan genbun, vol. 1, ed. Huang Liyun (Taipei: Museum of Ethnology,
Academia Sinica).
39.  Inō Kanori, ed., Riban shikō dai ikkan (Taipei: Taiwan sōtokufu keisatsu honsho,
1918; repr., Taipei: Southern Materials Center, 1995), 14–15.
40.  “Taiwan tsūshin,” Yomiuri shinbun, September 30, 1895.
41.  Taiwan sangyō ryakushi (Tokyo: Nōshōmu daijin kanbō bunsho-ka, 1895), 40–41,
52–53.
42.  Taiwan sōtokufu minseikyoku shokusanbu hōbun dai ikken dai ni satsu (Tokyo: Tai-
wan sōtokufu minseikyoku shokusanbu, 1896), 15.
43. Taihoku-shū, ed., Taihoku-shū ribanshi (Taipei: Taihoku-shū keimubu, 1924), 141.
44. Taihoku-shū, Taihoku-shū ribanshi, 172.
45.  Iriye Takeshi and Hashimoto Shigeru, “Taiwan banchi zatsuzoku,” Fūzoku gahō 130
(1896): 29–30.
46.  “Hokuban no sensō,” Taiwan nichinichi shinpō, May 10, 1898, 5.
47.  Paul D. Barclay, “Bansan kōekijo ni okeru ‘banchi’ no shōgyōka to chitsujoka,” Tai-
wan Genjūmin kenkyū 9 (2005): 93–96.
48. Inō, Riban shikō, 102.
49. Taihoku-shū, Taihoku-shū ribanshi, 304–10.
50.  Tavares, “Japanese Colonial State,” 361–85.
51. Inō, Riban shikō, 163.
52. Taihoku-shū, Taihoku-shū ribanshi, 248.
53.  Tavares, “Japanese Colonial State,” 361–85.
54.  Kondō, “Taiwan,” 38.
282     Notes

55. Inō, Riban shikō, 408–10.

56. Inō, Riban shikō, 457, 465.
57. Inō, Riban shikō, 723–25.
58.  Inoguchi Yasuyoshi, ed., Riban shikō dai nikan (Taipei: Taiwan Sōtokufu Keimu-
kyoku, 1921), 63; Uesugi Mitsuhiko, “‘Takasagozoku’ e no jusan seisaku: kōeki seisaku o
chūshin to shite,” Takachihō ronsō 26, no. 4 (1992): 59.
59.  Taiwan sōtokufu, ed., Riban gaiyō (Taihoku: Taiwan sōtokufu minseibu banmu hon-
sho, 1912), 55–56.
60.  Marui Keijirō, Buban ni kansuru ikensho, Bando Kyōiku Ikensho (Taipei: Banmu
honsho, 1914), 71.
61.  Inoguchi, 531.
62.  Mori Ushinosuke, “Taiwan shinrin to banjin no kankei ni tsuite,” Taiwan jihō 89
(1917): 9–19.
63.  Masuya Sei, “Banjin no keizai kannen to kanshū,” Riban no tomo, October 1, 1935,
8–9.
64.  From 1998 through 2015, I visited roughly fifteen museums or theme parks that
feature biographical information about these men, or feature the artifacts they collected or
photographs they produced, in Pingdong, Tainan, Fuxing, Nantou, Hengchun, Wulai, and
several sites in Taipei.
65.  Inō Kanori, “Taiwan tsūshin dai rokkai,” Tokyo jinruigakkai zasshi 11, no. 121 (1896):
274–76.
66.  Inō, “Taiwan tsūshin dai rokkai,” 301–13.
67.  Hu, “Taiwanese,” 200–211.
68.  Jiang Guizhen, Zaixian chuangtong de shijian, Wulai Taiyazu de wenhua tuxiang
(Taipei: National Museum of History, 2010), 96–106.
69.  Li Tsu-ning et al., The Story of Collection in a Century: National Taiwan Museum
Centennial Edition, trans. Christopher Findler and Ruby Li (Taipei: National Taiwan Mu-
seum, 2009), 49–53.
70.  Monbushō, ed., Jinjō shōgaku chiri ken ni: jidō yō (Tokyo: Monbushō, 1910), 21;
Tanaka Hirokichi, Shin kyōkasho kiga no kaisetu oyobi toriatsukai hō: shūshin chiri rekishi
no bu (Tokyo: Kōbundō, 1911), 102.
71.  Satō Haruo, “Wushe,” in Nihon tōchiki Taiwan bunka: Nihonjin sakka hinshū, bekkan
(Naichi sakka), ed. Kawahara Iwao (Tokyo: Rokuin shobō, 1998), 29.
72.  “Chinki na bussan ni gokyō mo fukashi,” Tokyo Asahi shinbun, April 18, 1923, 2.
73.  “Chichibunomiya denka no Taiwan Onari,” Taiwan jihō (July 1925): 12; “Kappanzan
bansha ni onari no Asakanomiya denka,” Taiwan jihō (December 1927): frontispiece.
74.  Ozaki Hotsuma, “Jūyō bijutsu no hozon to Taiwan no Banzokuhin,” Riban no tomo,
April 1, 1933, 2.
75.  E. Taylor Atkins, Primitive Selves: Koreana in the Japanese Colonial Gaze, 1910–1945
(Berkeley: University of California Press, 2010), 124.
76.  “Banjin banchi ni kansuru jimu oyobi jōkyō,” Taiwan zongdu fu gongwen leizuan
shuwei hua dang’an, July 1, 1900, 4625–5–3.
77.  Harrison Forman, “Diary: Japan, Taiwan, Korea [1938],” Harrison Forman Papers,
box 1, folder 7, item 6. Special Collections and University Archives, University of Oregon
Libraries.
 Notes    283

78.  Harrison Forman, letter to the editor, Natural History 47, no. 4 (1941): 182–83.
79.  Michael Rudolph, Ritual Performances as Authenticating Practices: Cultural Repre-
sentations of Taiwan’s Aborigines in Times of Political Change (New Brunswick, NJ: Transac-
tion Publishers, 2008), 7.
80.  Scott Simon, Sadyaq Balae!: L’Autochthonie Formosan dans tous ses états (Quebec
City: Les Presses de L’Université Laval, 2012), 89–90; Rudolph, Ritual Performances, 107–08,
117.
81. Rudolph, Ritual Performances, 6–7, 106; Simon, Sadyaq Balae!, 82–84; Ku Kun-hui,
“Rights to Recognition: Minorities and Indigenous Politics in Emerging Taiwan National-
ism,” in Taiwan Since Martial Law: Society, Culture, Politics, Economy, ed. David Blundell
(Taipei: Shung Ye Museum of Formosan Aborigines, 2012), 93.
82. Rudolph, Ritual Performances, 114–27; Simon, Sadyaq Balae!, 14–16, 228–33; Mit-
suda Yayoi, “First Case of the New Recognition System: The Survival Strategies of the Thao,”
in Taiwan Since Martial Law: Society, Culture, Politics, Economy, 169–70.
83.  Hu, “Embodied Memories,” 248.
84.  Shahid Amin, Event, Metaphor, Memory: Chauri Chaura, 1922–1992 (Berkeley: Uni-
versity of California Press, 1995); Gyan Prakash, Mumbai Fables (Princeton: Princeton Uni-
versity Press, 2010).
85. Simon, Sadyaq Balae!, 28–33; “Gov’t officially recognizes two more aboriginal
tribes,” China Post, June 27, 2014, accessed January 8, 2015, http://www.chinapost.com.tw/
taiwan/national/national-news/2014/06/27/411066/Govt-officially.htm.
86. Rudolph, Ritual Performances, 107–15; Mitsuda, “First Case,” 170.
87.  Arif Dirlik, The Postcolonial Aura: Third World Criticism in the Age of Global Capital-
ism (Boulder, CO: Westview Press, 1997), 227.
88.  James Clifford, “Identity in Mashpee,” in The Predicament of Culture: Twentieth-
Century Ethnography, Literature, and Art, ed. James Clifford (Cambridge: Harvard Univer-
sity Press, 1988), 277–348.
89.  Hsiau A-chin, Contemporary Taiwanese Cultural Nationalism (London: Routledge,
2000), 162–64; Rudolph, Ritual Performances, 55–59; Ku, “Rights to Recognition,” 104, 122.
90. Simon, Sadyaq Balae!, 220–25.
91. Rudolph, Ritual Performances, 60–62.
92.  Hu, “Embodied Memories,” 10.

C HA P T E R 4 .   T H E G E O B O D I E S W I T H I N A G E O B O DY

1. Achille Mbembe, “Necropolitics,” trans. Libby Meintjes, Public Culture 15, no. 1
(2003): 26.
2. Nicholas Thomas, Colonialism’s Culture: Anthropology, Travel, and Government
(Princeton: Princeton University Press, 1994), 71–72.
3.  Inō Kanori, “Shinajin no Taiwan doban ni kansuru jinshu teki kansatsu,” Taiwan
kanshū kiji 5, no. 8 (August 1905): 52.
4.  Inō, “Shinajin no Taiwan,” 54.
5. Thomas, Colonialism’s Culture, 71–72.
6.  The dotted dividing line on the map (published in 1895) in figure 32 bifurcates Taiwan
into the banchi (savage lands) and the unmarked rest of Taiwan; Akishika Kenkitsu, Taiwan
284     Notes

shiyō (Tokyo: Seibidō, 1895), n.p. Charles LeGendre’s 1870 map has a similar bifurcating
line; it is annotated with the characters for “savage border lands” (doban chikai), dividing
the island between zones occupied by “Aborigines” and “the portion of the island occupied
by Chinese”; LeGendre, Notes, 385. Both maps allocate more than half of Taiwan’s territory
to the non-Chinese, or aborigines.
7.  Takeuchi Sadayoshi, Taiwan (Taipei: Taiwan nichinichi shinpō, 1914), 620.
8.  Sakano Tōru, Teikoku Nihon to jinruigakusha, 1884–1952 (Tokyo: Keishobo, 2005),
230; Paul D. Barclay, “An Historian Among the Anthropologists: The Inō Kanori Revival
and the Legacy of Japanese Colonial Ethnography in Taiwan,” Japanese Studies 21, no. 2
(September 2001): 117–36.
9.  Peter Duus, The Abacus and the Sword: The Japanese Penetration of Korea, 1895–1910
(Berkeley and Los Angeles: University of California Press, 1995), 424–34; and Robert Es-
kildsen, “Of Civilization and Savages: The Mimetic Imperialism of Japan’s 1874 Expedition
to Taiwan,” American Historical Review 107, no. 2 (2002): 388–418.
10.  Patrick Wolfe has recently reenergized the notion that racism, as opposed to racial-
ization, is a particularly European construct and is ultimately bound up with white impe-
rialism and shifting configurations of white supremacy. See Traces of History: Elementary
Structures of Race (London and New York: Verso, 2016), 7, 18.
11.  W. J. T. Mitchell, Seeing Through Race (Cambridge: Harvard University Press, 2012),
34–35.
12. Mitchell, Seeing Through Race, 17; Wolfe, Traces of History, 101.
13.  Ke Zhiming, Fantou jia: Qing dai Taiwan zuqun zhengzhi yu shufan diquan (Taipei:
Zhongyang yanjiuyuan shehuixue yanjiusuo, 2001); Emma Jinhua Teng, Taiwan’s Imagined
Geography: Chinese Colonial Travel Writing and Pictures, 1683–1895 (Cambridge: Harvard
University Press, 2004).
14.  Michel Foucault, “Two Lectures,” in Culture/Power/History: A Reader in Contem-
porary Social Theory, ed. Nicholas B. Dirks, Geoff Eley, and Sherry B. Ortner (Princeton:
Princeton University Press, 1994), 213.
15.  These terms have extremely pejorative connotations and have been replaced in of-
ficial and popular discourse with the general term Yuanzhumin/Genjūmin.
16. Takigawa Miyotarō, ed., Shinryōchi Taiwantō (Tokyo: Kokondō, June 12, 1895),
160–202.
17.  Ueno Sen’ichi, “Taiwantō jissen roku,” Tokyo chigaku kyōkai hōkoku 13, no. 11 (Febru-
ary 1892): 21–48; Sanbō honbu, ed., Taiwan shi (Tokyo: January 1895).
18.  Sanbō honbu, Taiwan shi, 71.
19.  Inō Kanori, “Yo no sekishi o nobete sendatsu no kunshi ni uttau,” in Inō Kanori no
Taiwan tōsa nikki, ed. Moriguchi Kazunari (Taipei: Taiwan fūbutsu zasshisha, 1992). Com-
pare Sanbō honbu, Taiwan shi, 72–79, and George Taylor, “Formosa: Characteristic Traits
of the Island and Its Aboriginal Inhabitants,” in Aborigines of South Taiwan in the 1880s:
Papers by George Taylor, ed. Glen Dudbridge (Taipei: Shung Ye Museum of Formosan
Aborigines and Institute of Taiwan History, Academia Sinica, 1999), 154–59. Torii Ryūzō
himself, in a short retrospective on this classificatory project, identified Taylor’s work as
the first widely accepted taxonomy of the island’s aborigines; see “Jinruigaku kenkyū, Tai-
wan no Genjūmin (ichi) joron,” in Torii Ryūzō zenshū, vol. 5 (Tokyo: Asahi shinbunsha,
1976), 9.
 Notes    285

20. Mori Ushinosuke, Taiwan banzokushi dai ikkan (Taipei: Rinji Taiwan kyūkan
chōsakai, 1917; repr., Taipei: Nanten Shokyoku, 1996), 49–52. According to Mori’s account-
ing, there were 69 households and 389 inhabitants of Jiaobanshan-sha, ca. 1915.
21. Thomas, Colonialism’s Culture, 71.
22. Hirano Akio, “Taiwan seiban kisatsu no gaikyō,” Taiwan jinruigakkai zasshi 115
(1895): 6–8.
23.  “Seiban no junfuku,” Fūzoku gahō, November 28, 1895, 17–21.
24.  Eskildsen, “Civilization and Savages,” 388–418.
25. Satō Denzō, Chirigaku kyōkasho: chūtō kyōiku, Nihon chizu furoku (Tokyo:
Hakubunkan, 1898), n.p.
26.  Taipics.com, “Aboriginal Daily Life—Set 6,” accessed January 27, 2017, http://www.
taipics.com/abo_daily_life6.php.
27.  Kamei Tadaichi, Teikoku chirigaku kyōkasho (Tokyo: Sanseidō, 1900), 152.
28.  Shōgaku chiri kansan (Tokyo: Shūeidō, 1900), 29.
29.  Gotō Shinpei, “Formosa Under Japanese Administration,” The Independent 54, no.
2,796 (July 3, 1902): 1,582.
30.  Monbushō, ed., Shōgaku chirisho kan ni (Tokyo: Monbushō, 1903), 69.
31.  Inō Kanori, “Juntai nichijō,” in Inō Kanori no Taiwan tōsa nikki, ed. Moriguchi Ka-
zunari (Taipei: Taiwan fūbutsu zasshisha, 1992), 70–73.
32.  Karl Theodor Stöpel, Eine Reise in das Innere der Insel Formosa und die erste Beste-
igung des Niitakayama (Mount Morrison) (Buenos Aires: Companñía sud-americana de bil-
letes de banco, 1905), 42.
33.  Kamei Tadaichi, Teikoku shin chiri (Tokyo: Sanseidō, 1908), 138.
34.  James Davidson, The Island of Formosa Past and Present (Yokohama: Japan Times
Newspaper Company, 1903), 568; Jun’eki Taiwan Genjūmin kenkyūkai, ed., Inō Kanori shozō
Taiwan Genjūmin shashinshū (Taipei: Jun’eki Taiwan Genjūmin hakubutsukan, 1999), 141.
35.  See Lee Ju-ling, “Constructing an Imaginary of Taiwanese Aborigines through Post-
cards (1895–1945),” in Translation, History and Arts: New Horizons in Asian Interdisciplinary
Humanities Research, ed. Meng Ji and Ukai Atsuko (London: Cambridge Scholars, 2013),
111–35.
36. Kamei, Teikoku shin chiri, 138.
37.  Bai In [Inō Kanori], “Kokutei chiri kyōkasho chū Taiwan no bu no sakusen,” Taiwan
kanshū kiji 4, no. 9 (1904): 816–17.
38.  Tōno Municipal Museum, ed., Inō Kanori: kyōdo to Taiwan kenkyū no shōgai (Tōno,
Iwate-ken: Tōno shiritsu hakubutsukan, 1995), 68.
39.  Inō, “Yo no sekishi.”
40.  At the time, Inō was an employee of the Records Office at a salary of twenty yen
per month.
41. Moriguchi, Inō Kanori, 357; Inō Kanori, “Taiwan tsūshin dai nikai,” Tokyo jinruigak-
kai zasshi 118 (1896):149–51.
42.  Inō, “Taiwan tsūshin dai nikai,” 149–54.
43.  Inō Kanori, “Taiwan tsūshin dai ikkai,” Tokyo jinrui gakkai zasshi 117 (1895): 97.
44.  Vicente L. Rafael, “White Love: Surveillance and Nationalist Resistance in the U.S.
Colonization of the Philippines,” in Cultures of United States Imperialism, ed. Amy Kaplan
and Donald E. Pease (Durham, NC: Duke University Press, 1993), 185–218.
286     Notes

45.  Hirano, “Taiwan seiban,” 6–8.

46.  Inō, “Taiwan tsūshin dai nikai,” 153.
47.  Inō, “Taiwan tsūshin dai nikai,” 152.
48.  Inō Kanori, ed., Taiwan shi, vol. 1 (1902), 8–14, contains an extensive bibliography
of Chinese-language materials on Taiwan known to Japanese officials. In 1904, Inō’s transla-
tion and compilation of Qing records pertinent to the non-Han areas of Taiwan was pub-
lished; it continues to be a primary source for historians. Inō Kanori, ed., Taiwan banseishi
(Taipei: Taiwan sōtokufu minseibu shokusankyoku, 1904).
49.  Timothy Y. Tsu, “Japanese Colonialism and the Investigation of Taiwanese ‘Old
Customs,’ ” in Anthropology and Colonialism in Asia and Oceania, ed. Jan van Bremen and
Akitoshi Shimizu (Richmond, Surrey: Curzon Press, 1999), 197–218; Peter Pels, “From Texts
to Bodies: Brian Houghton Hodgson and the Emergence of Ethnology in India,” in An-
thropology and Colonialism in Asia and Oceania, ed. Jan van Bremen and Akitoshi Shimizu
(Richmond, Surrey: Curzon Press, 1999), 65–92.
50. Inō Kanori, “Seibanfu Ai o itamu,” Tokyo jinruigakkai zasshi 12, no. 138 (1897):
499–500.
51.  Melissa L. Meyer, “ ‘We Can Not Get a Living as We Used To’: Dispossession
and the White Earth Anishinaabeg, 1889–1920,” American Historical Review 96, no. 2
(1991): 387.
52.  Mori Ushinosuke, Chūō sanmyaku ōdan tanken hōbun, 1914, MS 0762/14 (vol. 1),
Taipei: National Taiwan Library, Taipei. This manuscript copy of Mori’s journal from the
Sakuma-campaign years contains several standardized Notes and Queries forms for record-
ing face, eye, nose and teeth shapes, and other physiological traits of individuals. These
forms are filled in, without names.
53. Inō Kanori, “Taiwan tsūshin dai yonkai,” Taiwan jinruigakkai zasshi 11, no. 120
(1896): 224–30.
54. Inō Kanori, “Taiwan tsūshin dai rokkai,” Taiwan jinrui gakkai zasshi 11, no. 121
(1896): 274.
55. “Jiku-shi no banjin buiku to Koore bansha no kijun,” Tokyo Asahi shinbun, Novem-
ber 22, 1897, 6.
56.  “Hokuban daishū-chō no shukushi,” Taiwan kyōkai kaihō 1 (1898): 102; Harada Zen-
shichi [原田善七], “Hokuban ko’ō kettō no zu,” Taiwan kyōkai kaihō 6 (1899): 76–77.
57.  Inō, “Juntai,” 8, 13.
58.  Inō, “Juntai,” 9.
59.  Hu Chia-yu, “Taiwanese Aboriginal Art and Artifacts: Entangled Images of Coloni-
zation and Modernization,” in Refracted Modernity: Visual Culture and Identity in Colonial
Taiwan, ed. Yuko Kikuchi (Honolulu: University of Hawai‘i Press, 2007), 198; Hu Jiayu and
Yilan Cui, eds., Taida renlei xuexi yi neng cangpin yanjiu (Taipei: Guoli Taiwan daxue chu-
ban zhongxin, 1998), 263.
60.  Banjō kenyūkaishi 1 (August 16, 1898): front matter. Thanks to Hu Chia-yu for send-
ing me an intact copy of this issue.
61. Ishikawa Gen’ichirō, ed., Taiwan meisho shashinchō (Taipei: Taiwan shōhōsha,
1899), n.p.
62.  Torii Ryūzō used the terms Yūgeiban and Yūgeimenban to designate Atayal peoples,
while Inō Kanori used the term Atayal.
 Notes    287

63.  “Typical head hunter, Formosa,” George Grantham Bain Collection, Library of Con-
gress Prints and Photographs Division, Washington, DC, Prints and Photographs Online
Catalog, accessed April 3, 2013, http://www.loc.gov/pictures/item/ggb2004002272/.
64.  The two grisly Bain News Service photos were used to illustrate the short piece, sug-
gesting that they had been made available to American publishers before that time. “Elec-
tricity Used to Capture Head Hunters,” Popular Mechanics, May 1909, 444, Google Books,
accessed on January 9, 2015, https://books.google.co.jp/books?id = mt8DAAAAMBAJ&lpg
= PA444&ots = k1snqUUX5J.
65.  “Head Hunting Natives and Japanese Police in Group around a Head,” “Head Hunt-
ers Seated With Their Head, Formosa,” and “Skull Shelf of Head Hunter, Formosa,” George
Grantham Bain Collection, Library of Congress Prints and Photographs Division, Wash-
ington, DC, Prints and Photographs Online Catalog, accessed July 23, 2017, http://www.loc.
gov/pictures/collection/ggbain/.
66.  Nihon Jun’eki Taiwan Genjūmin kenkyūkai, Inō Kanori, 108–11.
67.  A high-resolution scan of this cabinet card and its reverse side was posted on eBay
by the seller eby071 on December 14, 2010.
68.  Inō Kanori and Awano Dennosuke, Taiwan banjin jijō (Taipei: Taiwan sōtokufu
minseibu bunshoka, 1900), 15, stipulates that some Atayal wore ornaments, while others
did not. The popularized versions of Inō’s sketch that appeared in Davidson, Island of For-
mosa, 565, and Takekoshi Yosaburō, Japanese Rule in Formosa, trans. George Braithwaite
(London, New York, Bombay, and Calcutta: Longmans, Green, 1907), 221, are less subtle,
like the placards at the exhibitions and the captions for postcards.
69.  Hu Jiayu and Yilan Cui, Taida renlei xuexi, facing 263.
70.  Hu Chia-yu, “Taiwanese,” 198–99.
71.  Ishizaka Tonan, Photographs of Formosa: Collected in Commemoration of the T ­ okyo
Industrial Exhibition in 1907, 13, in Sōtokufu archives, National Taiwan Library, Taipei. This
drawing was reproduced in an early travel guide, before tourists could actually visit
­indigenous settlements; see untitled railway guide, ed. Taiwan sōtokufu tetsudō bu (Tokyo:
Eriguchi shōkai, 1908), 57, in Sōtokufu archives, National Taiwan Library, Taipei.
72. Davidson, Island of Formosa, 563.
73. Ishikawa, Taiwan meisho.
74.  Benedict Anderson, Imagined Communities (London: Verso, 1983), 163–85.
75.  James C. Scott, The Art of Not Being Governed: An Anarchist History of Upland
Southeast Asia (New Haven: Yale University Press, 2009).
76.  Antonio C. Tavares, “The Japanese Colonial State and the Dissolution of the Late
Imperial Frontier Economy in Taiwan, 1886–1909,” Journal of Asian Studies 64, no. 2 (2005):
361–85.
77.  “Mori Ushinosuke to Tsuboi Shōgorō, September 2, 1902,” Tokyo jinruigakkai zasshi
199, October 20, 1902, 39.
78.  “Taiwan banjin fūzoku shashin zusetu,” Tokyo jinruigakkai zasshi 213 (December
20, 1903): 109.
79.  Paul D. Barclay, “Playing the Race Card in Japanese Governed Taiwan, or: Anthro-
pometric Photographs as ‘Shape-Shifting Jokers’ ” in The Affect of Difference: Representations
of Race in East Asian Empire, ed. Christopher Hanscom and Dennis Washburn (Honolulu:
University of Hawai‘i Press, 2016), 38–80.
288     Notes

80. “Seiban no raisha,” Taiwan nichinichi shinpō, January 30, 1903; “Taiwan banjin
fūzoku,” Tokyo jinruigakkai zasshi, 108.
81.  Taiwan Government Bureau of Aborigine Affairs, ed., Report on the Control of the
Aborigines in Formosa (Taipei: Taiwan Government General, 1911), facing 3; Ishii Shinji, The
Island of Formosa and Its Primitive Inhabitants (London: China Society, 1916), plate VIII; Fu-
jisaki Seinosuke, Taiwan no banzoku (Tokyo: Kokushi Kankōkai, 1930), 248; “Taroko Tribe
Husband and Wife,” ip1225, East Asia Image Collection, Lafayette College, Easton, PA, ac-
cessed March 30, 2013, http://digital.lafayette.edu/collections/eastasia/imperial-postcards/
ip1225.
82. “Aborigines [sic] Woman, Formosa (33),” ip1020, East Asia Image Collection, Lafay-
ette College, Easton, PA, accessed March 30, 2013, http://digital.lafayette.edu/collections/
eastasia/imperial-postcards/ip1020.
83.  See the postcard reproduced in Narita Takeshi, Taiwan seiban shuzoku shashinchō
(Taipei: Narita shashin seihanjo, 1912), 31; Koizumi Tetsu, Bankyō fūzokuki (Tokyo kenset-
susha, 1932), plate 13; “All about the Aborigine Territories postcard collection,” lw0522,
East Asia Image Collection, Lafayette College, Easton, PA, accessed March 30, 2013, http://
digital.lafayette.edu/collections/eastasia/lewis-postcards/lw0522; Richard Goldschmidt,
Neu-Japan; Reisebilder aus Formosa, den Ryukyuinseln, Bonininseln, Korea and dem
südmandschurischen Pachtgebiet (Berlin: J. Springer, 1927), 65.
84.  Matsuda Kyōko, Teikoku no shisen: hakurankai to ibunka hyōshō (Tokyo: Yoshikawa
kōbunkan, 2003),15–19.
85.  Kuni Takeyuki, Hakurankai no jidai (Tokyo: Iwata shoin, 2005), 176.
86.  Barclay, “Senzen no media ni shiyō sareta Mori Ushinosuke no Taiwan Genjūmin
‘Kusshaku-ban’ shashin: hakurankai, shinbun, kyōkasho, ehagaki, zasshi de no shiyō rei no
kentō,” Taiwan Genjūmin kenkyū 19 (2015): 53.
87.  Taiwan Government Bureau of Aborigine Affairs, ed., Report, facing 3; Taiyō 23,
no. 7 (June 15, 1917); Mori Ushinosuke, ed., Taiwan banzoku zufu, vol. 1 (Taipei: Rinji Tai-
wan kyūkan chōsakai, 1915), plate 23; “two Quchi savages of Taiyaru tribe, Farmosa [sic],”
lw0002, East Asia Image Collection, Lafayette College, Easton, PA, accessed July 28, 2013,
http://digital.lafayette.edu/collections/eastasia/lewis-postcards/lw0002.
88.  “Hunting Head Hunters with Live Electric Wires,” Detroit Free Press, April 4, 1909.
89.  Elizabeth Edwards, introduction to Anthropology and Photography: 1860–1920, ed.
Elizabeth Edwards (New Haven: Yale University Press, 1992), 11.
90.  [Shinkō, Urai-sha Woman, Taiyal Tribe], ip1448, East Asia Image Collection, La-
fayette College, Easton, PA, accessed July 23, 2017, http://digital.lafayette.edu/collections/
eastasia/imperial-postcards/ip1448.
91.  “Wiping Out the Formosans: Japan’s War of Subjugation on Their Newly Taken Is-
land,” Washington Post, June 30, 1907, E14.
92.  Inō Kanori, ed., Riban shikō dai ikkan (Taipei: Taiwan sōtokufu keisatsu honsho,
1918; repr., Taipei: Southern Materials Center, 1995), 182.
93. Takekoshi, Japanese Rule in Formosa, 230.
94. See figure 3 and “THE BARBARIC WOMAN OF FORMOSA,” lw0384, East Asia
Image Collection, Lafayette College, Easton, PA, accessed March 30, 2013, http://digital.
lafayette.edu/collections/eastasia/lewis-postcards/lw0384; “Woman of Savage,” lw0391,
East Asia Image Collection, Lafayette College, Easton, PA, accessed March 30, 2013,
 Notes    289

http://digital.lafayette.edu/collections/eastasia/lewis-postcards/lw0391. For details regard-

ing the dozens of publications that included Pazzeh’s portrait, see Barclay, “Senzen,” 47–75.
95.  “Head hunters in Formosa,” lw0264, East Asia Image Collection, Lafayette Col-
lege, Easton, PA, accessed March 30, 2013, http://digital.lafayette.edu/collections/eastasia/
lewis-postcards/lw0264.
96.  See Paul D. Barclay, “The Relics of Modern Japan’s First Foreign War in Colonial
and Postcolonial Taiwan, 1874–2015,” Cross Currents: East Asian History and Culture Re-
view, no. 17 (December 2015): frame 3, accessed July 23, 2017, https://cross-currents.berkeley.
edu/e-journal/issue-17/barclay.
97.  [Under Prince Kan’in at the Taiwan Governor-General’s Residence], ip0192, East
Asia Image Collection, Lafayette College, Easton, PA, accessed May 13, 2016, http://digital.
lafayette.edu/collections/eastasia/imperial-postcards/ip0192.
98. Mori, Banzoku, vol. 1, plate 15.
99.  This village on the South Cape was the combatant tribe during the 1867 Rover inci-
dent (see chapter 1); the men in the picture are wearing Qing hairstyles; Koalut was moved
to normal administration in 1904.
100. Inō, Riban shikō, 349; Fujii Shizue, Riju shiqi Taiwan zongdufu de lifan zhengce (Tai-
pei: Guoli Taiwan shifan daxue lishi yanjiusuo, 1989), 190–91.
101. For example, Saishin Nihon chiri kyōkasho (Tokyo: Sanseidō, 1925), 95; Joshi
kyōiku: saishin nihonchiri (Tokyo: Sanseidō, 1928), 96; Saishin Nihon chiri kyōkasho (Tokyo:
Sanseidō, 1933), 131.
102.  Barclay, “Senzen,” 54.
103.  Three photos of Watan Yūra’s extended family, at least five different poses for por-
traits of Marai Watan, three poses for Yūgai Watan, two poses for Pazzeh Watan, and a
panoramic shot of an Atayal house and granary set in the mountains were the subjects of
these photos. See Barclay, “Senzen.”
104. Inō, Riban shikō, 375–76.
105.  Xia Shengli, et al., Hyakunen Cangsang: Dogura Ryūjirō to Taihoku Kameyama sui-
ryoku hatsudenjo, trans. and ed. Dogura Masao and Dogura Nobuko (Sakurai-shi: Dogura
Masao, 2006).
106.  Michele Lang, An Adventurous Woman Abroad: The Selected Lantern Slides of Mary
T. S. Schaffer (Vancouver: Rocky Mountain Books, 2011), 258–63.
107.  The colorized bust and profile of Marai, as well as several anthropometric shots
included in Takekoshi Yosaburō’s 1905 book, are preserved in the James Wheeler Davidson
Collection at the University of Calgary, accessed March 4, 2017, http://contentdm.ucalgary.
ca/cdm/search/collection/jwd/.
108.  For a detailed analysis of an exemplary album from this period, see Okada Shige-
hiro, “Takamatsu-miya kashi no ‘Banhi tōbatsu kinen shashin-chō’ ni tuite,” Gakushuin
Daigaku shiryōkan kiyō 14 (March 2007): 77–154.
109.  The term “encampment” is taken from Joseph Allen’s essay on Meiji-period pho-
tographic albums of Taiwan, “Colonial Itineraries: Japanese Photography in Taiwan,” in
Japanese Taiwan: Colonial Rule and Its Contested Legacy, ed. Andrew D. Morris (London:
Bloomsbury Press, 2015), 25–48.
110.  Taiwan sōtokufu keimukyoku, ed., Takasagozoku chōsasho dai san-hen: shinka (Tai-
pei: Taiwan sōtokufu keimukyoku, 1937).
290     Notes

111.  Taiwan sōtokufu minseibu banmu honsho, ed., Riban gaiyō (Taipei: Taiwan sōtokufu
minseibu banmu honsho, 1912), 92–96.
112. Untitled railway guide, ed. Taiwan sōtokufu tetsudōbu (Tokyo: Eiguchi shōkai,
1908), 57–58, in Sōtokufu archives, National Taiwan Library, Taipei.
113.  Yamabe Kentarō, ed., Taiwan (II), vol. 22, Gendaishi shiryō (Tokyo: Misuzu shobō,
1971), 416.
114.  Endō Hiroya, Banhi tōbatsu kinen shashinchō (Taipei: Endō shashinkan, 1911), in
Sōtokufu archives, National Taiwan Library, Taipei.
115.  Ide Kiwata, Taiwan chisekishi (Taipei: Taiwan nichinichi shinpōsha, 1937), 707.
116.  Hashimoto Hakusui, ed., Taiwan ryokō annai (Taipei: Fujisaki Tomisō, 1915), 78.
117.  A. B. Kirjassoff, “Formosa the Beautiful,” National Geographic Magazine 37, no. 3
(1920): 246–92.
118. Taiwan sōtokufu tetsudōbu, ed., Tetsudō ryokō annai (Taipei: Taiwan sōtokufu
tetsudōbu, 1921), 32.
119. Sōyama Takeshi, Shokuminchi Taiwan to kindai tsūrizumu (Tokyo: Seikyōsha,
2003), 243–46.
120. Taiwan sōtokufu tetsudōbu, ed., Tetsudō ryokō annai (Taipei: Taiwan sōtokufu
tetsudōbu, 1924), 69.
121.  H. L. Bagallay, “The Savage Territory of Formosa” [May 1925], in Taiwan: Political
and Economic Reports 1861–1960, vol. 7: 1924–1941, ed. Robert L. Jarman (Slough: Archive
Editions, 1997), 24.
122.  Taiwan sōtokufu kōtsūkyoku tetsudōbu, ed., Taiwan tetsudō ryokō annai (Taipei:
Taiwan sōtokufu tetsudōbu, 1927), 88; Taiwan sōtokufu kōtsūkyoku tetsudōbu, ed., Tai-
wan tetsudō ryokō annai (Taipei: Taiwan sōtokufu tetsudōbu, 1930), 97; Taiwan sōtokufu
kōtsūkyoku tetsudōbu, ed., Taiwan tetsudō ryokō annai (Taipei: Taiwan sōtokufu tetsudōbu,
1934), 67.
123.  Abe Akiyoshi, Taiwan chimei kenkyū (Taipei: Sugita shoten, 1937), 8–9.
124.  Taiwan sōtokufu kōtsūkyoku tetsudōbu, ed., Taiwan tetsudō ryokō annai (1934), 9.
125. Mori, Taiwan banzoku zufu, vol. 1, plate 33; a common postcard displaying Atayal
women in Chinese, Japanese, and Atayal clothing, ca. 1920, “Chest-nuts by female savages
of Ataiya’ trive [sic] (5),” lw0189, East Asia Image Collection, accessed August 12, 2016, La-
fayette College, Easton, PA, http://digital.lafayette.edu/collections/eastasia/lewis-postcards/
lw0189.
126.  Paul D. Barclay, “Peddling Postcards and Selling Empire: Image-Making in Taiwan
under Japanese Colonial Rule,” Japanese Studies 30, no. 1 (May 2010): 81–110.
127.  Barclay, “Senzen,” 58–59.
128. Ming-cheng M. Lo, Doctors within Borders: Profession, Ethnicity, and Modernity in
Colonial Taiwan (Berkeley: University of California Press, 2002); Ka Chih-ming, Japanese
Colonialism in Taiwan: Land Tenure, Development, and Dependency, 1895–1945 (Boulder,
CO: Westview Press, 1995); Wang Taysheng, Legal Reform in Taiwan under Japanese Co-
lonial Rule, 1895–1945: The Reception of Western Law (Seattle and London: University of
Washington Press, 2000).
129.  Yao Jen-to, “Governing the Colonised: Governmentality in the Japanese Colonisa-
tion of Taiwan, 1895–1945” (PhD diss., University of Essex, 2002).
130. Inō, Riban shikō, 326–33, 474–76.
 Notes    291

131.  Matsuoka Tadasu, Taiwan Genjūmin shakai no chihōka: Mainoriti no nijū seiki (To-
kyo: Kenbun shuppan, 2012), 39–40.
132. Matsuoka, Taiwan Genjūmin, 101–17.
133.  Matsuda Kyōko, Teikoku no shisō: Nihon “Teikoku” to Taiwan Genjūmin (Tokyo:
Yūshisha, 2014), 190–92.
134.  Hashimoto Mitsuru, “Yanagita Kunio’s ‘Japan,’ ” in Stephen Vlastos, ed., Mirror of
Modernity: Invented Traditions of Modern Japan (Berkeley: University of California Press,
1998), 133–43; Marilyn Ivy, Discourses of the Vanishing: Modernity, Phantasm, Japan (Chi-
cago: University of Chicago Press, 1995), 68–73.
135.  Matsuda Kyōko, Teikoku, 190–92.
136. Yih-ren Lin, Lahuy Icyeh, and Da-Wei Kuan, “Indigenous Language-Informed
Participatory Policy in Taiwan: A Socio-Political Perspective,” in Documenting and Revital-
izing Austronesian Languages, ed. Victoria Rau and Margaret Florey (Honolulu: University
of Hawai‘i Press, 2007), 138.
 Gl os s ary

Terms used in text Characters or kana

aiyū 隘勇
Alishan 阿里山
Amis 阿美
Atayal 泰雅
Awai 亞歪
banba 蕃婆
bandō kyōikujo 蕃童教育所
bango 蕃語
banjin 蕃人
banjin banchi mondai 蕃人蕃地問題
banjo 蕃女
Bankai no hikyō 蕃界の秘境
banpu 蕃婦
banzoku 蕃族
baojia 保甲
Baoli 保力
Bayao Bay 八瑤湾
Beipu 北埔
bentuhua 本土化
bunka seiji 文化政治
293
294     Glossary

Bunun 布農
Chen Ansheng 陳安生
Chen Xilai 陳溪瀨
Chiankoey 雙渓
Dakekan 大嵙崁
dogo 土語
domoku 土目
Dongpushe 東埔社
Du Tingdong 杜庭動
Ernai 爾乃
fan 番
fanren 番人
Fengshan 風山
Fukahori Yasuichirō 深堀安一郎
fukenju 撫墾局
fukutsūyaku 副通訳
Fūzoku gahō 風俗画報
Gantaban 干卓萬
genjin 原人
gishu 技手
gokanen keikaku riban jigyō 五か年計画理蕃事業
Gotō Shinpei 後藤新平
guihuashengfan 帰化生蕃
Hakka 客家
hannin 判人
Hashiguchi Bunzō 橋口文蔵
Hengchun 恒春
Hewen (Hha-hoi) 合脗
hitsudan 筆談
Hiyama Tetsusaburō 檜山鉄三郎
hontōjin 本島人
Huang Qiying 黄祈英
Huozhang 夥長
Igarashi Ishimatsu 五十嵐石松
Inō Kanori 伊能嘉矩
inshu no reigi 飲酒の礼儀
 Glossary    295

Ishii Shinji 石井眞二

Itō Shūkichi 伊藤修吉
Iwan Robao 伊娃莉・羅拔俄
Jiaobanshan 角板山
Jiayi 嘉義
Jiku Shō Min 竺紹珉
jimu shokutaku 事務嘱託
Joseon 朝鮮
jukuban 熟蕃
junho 巡補
junsaho 巡査補
Kagoshima 鹿児島
Kaku Kurata 賀来倉田
Kawano Shuichirō 河野主一郎
kegai 化外
keimenban 黥面蕃
kekkon o yaku 結婚を約
kijun 帰順
Koalut 亀好児
koin 傭員
Kondō Gisaburō 近藤儀三郎
Kondō Katsusaburō 近藤勝三郎
Kondō Yoshio 近藤義雄
koshu 戸主
koyū 固有
Kuomintang / Guomindang 国民党
Kuskus 高士佛
Lan Yinding 藍蔭鼎
Langqiao 瑯｛王＋喬}
Li A’long 李阿龍
lifan tongzhi 理番同知
Linyipu 林圮埔
Liu Mingchuan 劉銘傳
mae-Taikokan-ban 前大嵙崁蕃
Mao Jian 毛奸
Marai 馬来
296     Glossary

Mehebu 馬赫坡
Mei-hai 梅海
mekake 妾
metoru 娶る
Miyakojima 宮古島
Mochiji Rokusaburō 持地六三郎
Mona Ludao 菓那・魯道
Mori Ushinosuke 森丑之助
Mosu 墨宿
Mudan 牡丹
mukōyoshi 婿養子
Murakami Yoshio 村上義雄
Myōzai 名西
Nagakura Yoshitsugu 長倉吉次
naien no otto 内縁の夫
Nakamura Yūsuke 中村雄助
Nan’ao 南澳
Neng’gao 能高郡
Order of the Sacred Treasure, Sixth Rank 叙勲六等, 授瑞寶章
Ortai 鄂爾泰
Ōtsu Rinpei 大津麟平
Paiwan 排湾
Paalan 芭蘭
Pan Bunkiet 潘文杰
Pan Laolong 藩老龍
Papau Iron パパウ・イロン
Pixo Doleh 貝克・道雷
Poliac 保力
Pu Chin 蒲靖
Puli 埔里
Pushige 璞石閣
Puyuma 卑南
Quchi 屈尺
Riban no tomo 理蕃の友
Rukai 魯凱
Sabaree 射麻里
 Glossary    297

Sagara Nagatsuna 相良長綱

Saisiyat 賽夏
Sakuma Samata 佐久間左馬太
sancengzhi zuqun fenbu jiagou 三层制族群分布架构
Sandiaoshe 三貂社
Sanjiaoyong 三角湧
Saprek 射不力
Sasaki Yasugorō 佐々木安五郎
Satō Haruo 佐藤春夫
Satomi Yoshimasa 里見義正
Sazuka Aisuke 佐塚愛祐
Sediq 賽徳克
seiban kyōhi 生蕃饗費
seibanjin 生蕃人
seinendan 青年団
seiryokusha 勢力者
Seiyūkai 政友会
sha 社
Shang Tai 尚泰
sheding 社丁
shegun 社棍
shengfan 生番
sheshang 社商
Shimaigahara 姉妹ヶ原
Shina kanwa 支那官話
Shō (dynasty) 尚
Shōjiro Kunitei 彰城邦貞
Shoucheng Dashan 守城大山
Shuangxikou 双渓口
shūchō 酋長
shujiho 主事補
shuryō 首領
shutchōjo 出張所
shuzoku 種族
Suzuki Hideo 鈴木秀夫
Taidong 台東
298     Glossary

tairiku rōnin 大陸浪人

Taiwan bankai tenbō 台湾蕃界展望
Taiwan banzoku zufu 台湾蕃族図譜
Taiwan doban 台湾土蛮
Taiwan dogo 台湾土語
Taiwan gahō 台湾画報
Taiwan jihō 台湾時報
Taiwan Yuanzhuminzu 台湾原住民族
Takasagozoku 高砂族
Takekoshi Yosaburō 竹越与三郎
Tanaka Tsunatoku 田中綱常
Taroko 太魯閣
Tata 達達
Tgdaya 徳固達雅
Thao 邵
Toda 都達/道澤
tōmoku 頭目
tongshi 通事
Torii Ryūzō 鳥居龍蔵
Truku 徳路固
Tsou 鄒
tsūben 通辯
tsūyakukan 通訳官
tsūyakusei 通訳生
Tuilasok 猪朥束
Urashō 浦庄
Wanda 萬大
Wangsha 網紗
Watan Yūro 瓦丹有洛
wu-pin 五品
Wu Sha 呉沙
Wushe 霧社
Xalut 白狗
Xincheng 新城
Xinzhu 新竹
Xitou 溪頭
 Glossary    299

Xiyanlaowa 四煙老瓦
Yami 雅美
Yang Youwang 楊友旺
Yawai Taimu 亜娃伊・泰目
Yilan 宜蘭
Yūkōmaru 有功丸
Zhang Shichang 張世昌
Zhang Xinzhang 張新丈
Zhang Yichun 張義春
Zheng Yan’ning 鄭延寧
 I nde x

aborigine administration (riban), 4, 90–92, colonial Taiwan, 212; race science and,
119–120, 271n20 35, 215
aborigine common schools (banjin kōgakkō), Arao Kiyoshi, 115
148–150, 153 Atayal (ethnic group), 7map, 8map; alcohol
aborigine territory, 2, 32–33, 58map, 80map, 102, consumption and, 44, 45fig., 85–86; as
193map, 194 combatants, 28–32, 39, 46, 48fig., 49fig., 70,
aborigine youth schools (bandō kyōikujo), 90, 94, 102, 109, 136–137, 178; as “innocent
149, 152 aborigines,” 173–175; colonial education
aiyū-sen. See guardline and, 149–155; cultural renaissance, 188–189;
Akamatsu Noriyoshi, 68–69 ethnogenesis of, 6, 9, 167, 183, 198–228,
Aki (the interpreter), 122–123, 124fig., 217 236–244; interpreters and, 79, 85, 122–128;
Akinaga Nagakichi, 139 land rights, 249; language, 140; textiles, 40,
alcoholic beverages, 44, 74, 81, 85–87, 97, 165, 174 161–162, 167, 169–171, 174, 182–187; trade, 79;
Alliance of Taiwan Aborigines, 5 tourism and, 184, 233–236; wet diplomacy
Alok (Paiwan warrior), 69 and, 82–85, 87, 91, 125–128
Amis (ethnic group), 7map, 8map, 59; as Atkins, E. Taylor, 252n17
attendees of aborigine common schools, 149; Aui Heppa, 96
as dancers, 167; as minority in Langqiao, Aui Nukan, 137
59; in Ueno Sen’ichi’s ethnic classification, authenticity, 3, 36–37, 188, 196
197; language, 144; under “normal Awai, 124–126, 213–217
administration,” 144, 230–232, 246
annex schools (bunkyōba), 142–148 bandō kyōikujo. See aborigine youth schools
anthropologists: as preservationists, 163, 171, banjin kōgakkō. See aborigine common schools
182–183; 188–189, 248; as taxonomists, banpu (female indigene), 123–128, 133, 136
6, 35, 198; as theorists of indigenism, 38; baojia, 31, 119, 146, 157, 244
cultural pluralism and, 184, 191–192, 213, 225; Bassau Bōran, 131
ethnogenesis and, 194–195, 213; photography Bayly, C.A., 11, 15–17, 36
and, 216–222; wet diplomacy and, 86 Beasley, W.G., 26–27
anthropology: authenticity and, 120; cultural Beinan, 9fig., 143
pluralism and, 35–36, 167, 182, 191, 196; in Beipu incident, 156–157, 233

301
302     Index

Bell, Admiral H. H., 56–57 dynastic state, 11–21, 31, 38–39, 75–77, 114, 192,
benmusho (District Administrative Office), 85, 134 264n99
Benton, Lauren, 12, 13, 251, 252n8, 255n45
biculturalism, 60, 121–122, 133 economic competence, 3, 6, 147, 167, 180–181,
bifurcated sovereignty, 3, 13, 17–18, 119, 149, 192, 279n12
248, 283n6 education, 77, 120, 146–157
big bang (of government outlays), 13, 29, 32, 102, Eighteen-Tribe Confederation (of Langqiao),
244–245 59–65, 163, 169, 197
biopolitics/biopower, 29–32, 258n104 Em, Henry, 4
Bukonsho. See Pacification Office embargoes, 59, 95–96, 135, 169, 176–181
Bunkiet, Jagarushi Guri (Pan Wenjie), 70, 142, entangled objects, 168–178
144, 145fig., 146–148, 245 Eskildsen, Robert, 77, 201
Bunun, 7map, 8map, 9fig., 94, 96–97, 102, 11, 167 ethnic enclave, 31, 119
Burbank, Jane, 12–13 ethnic classification, 135, 147, 183, 195–199,
Bureau of Aboriginal Affairs (banmu honsho), 213–217, 230, 259n122, 284n19
103, 139 ethnogenesis, 219–222
Buzan, Barry, 11, 13 ethnography. See anthropology
ethnologists. See anthropologists
camphor: industry, 87, 94, 98, 124, 132, 162, 174, ethnology. See anthropology
178–179; monopoly, 32, 98–101, 118, 188, 178;
wars, 28, 32, 70, 95, 98–106, 175, 225–228 Fangliao, 58map, 62, 75
Cassel, Douglas, 61, 68–70, 75, 125, 164 favoritism, 46, 113, 140
Cassel, Pars, 20 feasts, 52, 64–74, 79, 83, 86, 146, 176; aborigine
casualties, 98–100, 103, 105–106, 110, 176 administration (riban) and, 89–91, 96,
census: categories, 135, 150, 213; ethnogenesis 111–112, 178
and, 213; of Taiwan, 25, 33, 47, 48, 51, Fenggang, 53, 53map, 60, 62, 75
193fig.,196, 197 five-year plan to manage the Aborigines
Checheng, 53map, 59, 60, 142 (gokanen keikaku riban jigyō), 29, 49fig.,
Ching, Leo, 30 104–110, 138–139, 179, 245–246
Chiwas Ludao, 119, 137, 138 flogging, 3, 156–157
Chou Wan-yao, xvii, 50, 75 Foucault, Michel, 21–23, 28–29
citizens (kokumin), 21, 23, 33, 78, 92, 149, 152, 157 Friedman, P. Kerim, 275n90
competitive imperialism, 2, 17, 27, 38, 96, 194 Frühstück, Sabine, 29
Cooper, Frederick, 12–13 Fujii Shizue, 104
cultural rule: in Korea, 3–4, 9, 37; in Taiwan, 6, 9 Fujisaki Seinosuke, 75, 112
cultural survival, 2, 37, 248, 253n21 Fujitani, Takashi, 23, 29–30, 77, 254n27
Fukahori Yasuichirō, 95–96, 130–131
Dakekan (Atayal subgroup), 9fig., 71, 198; as Fukushima Kunari, 68
combattants, 178–179, 223, 233; as culture- Fukushima Yasumasa, 116
bearers, 183, 216, 236; at parleys, 83, 125;
dialect, 125, 215 Gao Jiaxin (Lianes Punanang), 50, 52
Dakekan (place), 7map, 58map, 80map; Gaoshifo. See Kuskus
administrative seat, 83–84, 89–91, 122–125, geobody, 18–19, 33, 65, 162, 194. See also
204–205, 214, 233; junction town, 82, 84fig., second-order geobody
89–91, 125, 172–174, 198, 204–205, 214, 233 gifts, 39–40, 162–163, 261n27; aborigine
Deng Tianbao, 52, 55 administration (riban) and, 82–96, 175–179;
Deng Xiangyang, 5, 96, 154 frontier diplomacy and, 46, 60, 64, 70–71,
dialects, 90, 117, 133, 138 81–82, 123, 173–174, 217; in redistributive
Dirlik, Arif, 37 political systems, 65, 128; in ritual exchange,
discipline: regime of, 13, 21–33, 99, 118–119, 61, 165–166; Yoshimura beating incident
156–157, 244–248 and, 43
Duara, Prasenjit, 36 Gantaban, 9fig., 96–97, 97fig., 112–113
 Index    303

global transformation, 11–17, 36–38, 191–195 as critic, 191, 211; as ethnologist, 85–86,
Gotō Shinpei, 25–26, 29, 99, 111, 178–179, 221 131, 133, 146, 182, 202, 212, 215–216; ethnic
governmentality, 23–29, 32–33, 38, 99 classification and, 10fig., 35, 135; ethnogenesis
guardline (aiyū-sen), 39, 46–47, 47fig., 97–106, and, 188, 219–221; ethnonym “Atayal” and,
108fig., 109fig., 111–113, 135–139, 223–233, 6, 217
226fig., 227fig., 233 interiority, 6, 112
gunboat diplomacy, 56, 88 intermarriage, 39, 46, 121–122, 127, 146–147
Gungu. See Hōgō intermediaries, 12, 20, 79, 88, 95, 111–113, 121–122,
Guomindang (Nationalist Party), 3,5, 187, 128, 165
248–249, 252n10, 253n20, 252n122 international law, 17, 28, 77–78, 192, 222
international system, 2, 11–15, 23, 33, 77, 194–196
Habairon (Papau Iron), 199–200, 202fig., 203fig., interpreter. See tongshi
204, 204fig., 205–206, 206fig., 207, 213 Ira Watan, 202fig., 203fig., 204fig., 205–208,
Hakka (Kejia), 52, 56–62, 85, 121, 134, 178 206fig., 209fig.
Hamashita, Takeshi, 54–55, 166 Iriye Takeshi, 98, 127–128
Hanaoka Ichirō (Dakis Nobing), 153–154 Isa (of Sabaree), 59, 64–74, 88, 164–166
Hanaoka Jirō (Dakis Nawi), 153–154 Ishizuka Eizō, 4
Hancock, William, 79 isshi dōjin (impartiality, non-discrimination),
Hannah, Matthew, 21–27 46, 113
Haruyama Meitetsu, 167–168 Iwan Robao, 96, 124, 130, 133–137, 209
Hashiguchi Bunzō, 82–88, 125–126, 174, 199–200,
202fig., 205 Jiaobanshan (Atayal subgroup), 9fig., 198
headmen. See tōmoku Jiaobanshan (place), 7map, 9fig., 80map; as
Hengchun: lower, (Paiwan subgroup), 9fig., 59, battlefield, 48fig., 109, 232, 234fig.; emissaries,
71, 74fig., 142, 146–148, 151, 164, 181, 223, 230, 84–88, 125, 174, 182, 198–202, 202fig., 205–212;
245–246; Peninsula, 39, 72, 121, 143, 197, 223; school, 152, 153fig., tourism and, 184, 229,
upper, (Paiwan subgroup), 147, 181; walled 232–242; trading post, 177fig.
city, 53map 122, 142, 148 Jiku Shō Min, 131–133, 216–217, 228–229
Hevia, James, 252n11 Jilong Harbor, 103
high velocity capitalism, 18–22, 33, 103, 166 Johnson, James, 63–64, 68–69, 72fig.
Hirschman, Charles, 33
Hiyama Tetsusaburō, 88–89, 93, 129–130, 176, Ka Chih-ming, 29
209, 273n54 Kabayama Sukenori, 83, 85, 90, 200–201,
Hōgō (Gungu), 9fig., 97fig., 135, 137, 153–154 202fig., 223
Hok-lo, 52, 60, 134, 211 Kagoshima. See Satsuma
horizontal integration, 21, 39, 48, 75, 77–78, kaishan fufan (open the mountains, pacify the
119, 146 barbarians), 122, 142–144, 197
House, Edward, 68–71, 163–166 Kamimura Tōru, 165, 263n55
Hu Chia-yu, 37, 167, 187, 189 Kasahara Masaharu, 259n121
Huang Chih-huei, 30 Katz, Paul, 156–157
Hyung Il Pai, 252n17 Kawano Shuichirō, 87–88, 126
kensho. See stipend
identity: aborigine territory police and, 156; Kido Takayoshi, 76–77
Atayal, 167, 206–213, 218; censuses and, 33–34, Kitamura Kae, 141, 146, 152, 154
213; Korean, 4,6; modern indigenous, 2, 37, Koalut, 9fig., 53map, 56–61, 142, 147–148, 151,
188, 219; Saisiyat, 37, 89; Taiwan Indigenous 223, 232
Peoples, 6, 9, 154, 187, 189; Taiwanese, 187 Kōan (the interpreter), 123, 124fig., 217
imperial subjectification (kōminka), 30–31, Kodama Gentarō, 96, 99, 102, 118, 268n181
119–120, 154 kokugo denshūjo (Japanese language institutes),
indigenism, 2, 13, 36–40 141–142
indigenous youth corps. See seinendan kokumin. See citizens
Inō Kanori: as colonial collector, 183, 189; kōminka. See imperial subjectification
304     Index

Kondō Gisaburō, 112–113, 137fig., 138 (interpreters), 64, 142; for tōmoku
Kondō Katsusaburō, 96, 129–133, 130fig., 135–138 (headmen), 111; on the guardline, 106, 117, 136
Korean Government-General, 3–4 Mia. See Miya.
koseki (Japanese household registration), middle ground, 18, 32, 111
138, 140 Ministry of Education (Monbusho): as textbook
kula ring, 165 publisher, 150, 211
Kuskus, 9fig., 50, 52, 53map, 59, 70–71, 147, Minseitō (Popular Government Party), 4
277n148 Mio Yūko, 134
Mitchell, W.J.T., 195
LaFeber, Walter, 18 Mitsuda Yayoi, 187
land tax reforms: in Japan, 24–25, 29; in Taiwan, Miya, 60–64, 68–69, 121
25–27 Miyanohara Tōhachi, 89–91
Langqiao: Peninsula, 14, 21, 51map, 53map, 55–63, Mizuno Jun: as Kabayama Sukenori’s translator,
67–72, 75–79, 90, 121, 165, 256n49; village of, 72, 72fig., 83; as Taiwan Government-General
53map, 61 civil administrator, 129, 166, 202fig.; mission
Lawson, George, 11–13 to Dakekan (1873), 172–173
legal centralism, 17–19, 33, 38 Mizuno, Norihito, 55
LeGendre, Charles, 56–69, 71, 79, 88, 149, Mochiji Rokusaburō, 28, 102, 192, 222
163–166, 192 Mona Ludao, 1, 6, 9, 43, 47, 112
Lemoi Maton, 84fig., 199 Mongolian (race), 33
Lenin, V.I., 11, 17–18, 21 Mori Ushinosuke: as Chinese-language
Leung, Edwin Pak-wah, 54, 75 translator, 116, 182; as critic, 180–182; as
Li Hongzhang, 67, 76 curator, 182, 184; as ethnic taxonomist, 35–36,
Lianes Punanang. See Gao Jiaxin 135, 198; as ethnographer, 86, 128–129, 182,
liquor. See alcoholic beverages 215, 225; as photographer, 182–183, 220–223,
Liu Deshao (Liu Tek-chok), 142–144 228–230, 236, 239, 242–243
Liu Mingchuan, 20, 122–124, 215 Motonaiban, 202fig., 203fig., 204, 204fig.,
Liu Mingdeng, 57–64 205–206, 206fig., 208
Lo, Ming-cheng M., 158 Mudan (Paiwan subgroup), 9fig., 63, 72, 135,
Ludao Bai, 112, 137 147–148
Mudan (place), 9fig., 51map, 53map, 58map
Mackenzie, Lieutenant Captain, 56–61 Mudan village incident, 19–21, 50–75
Malay (race), 33–36, 150–152, 216 Murakami Yoshio, 178
Malepa, 7map, 112, 139–140, 179
Mamdani, Mahmood, 47, 111 Nagano Yoshitora, 93–94, 176
Marai (eight-village settlement), 79 Nakamura Yūsuke, 143, 143fig.
Marai (of Jiaobanshan), 202fig., 205–208, 206fig., Nan’ao (Atayal subgroup), 7map, 9fig., 83,
209fig. 87, 175
Marai (paramount chieftain near Wulai), 122 National Taiwan Museum, 161, 182
Marai Watan (of Rimogan), 220–222, 242fig., nationalism: as analog to isshi dōjin, 46; as
243fig., 244 nineteenth century phenomenon, 2, 38;
March 1 Movement, 3–4 geobodies and, 192–194; Korean, 4, 9;
Masitoban, 7map, 9fig., 136, 140 postcolonial, 5, 188; Satsuma domain, 76;
Matsuda Kyōko, 247 Taiwanese, 5
Matsuda Yoshirō, 154 Native Americans, 16, 30, 78, 111, 215
Matsuoka Tadasu, 32, 151, 245–246 Native Authority, 47–48; 111–113; 155
Mbembe, Achille, 29 native peoples: as dojin (Taiwanese), 117, 171, 208;
Mehebu, 1, 6, 7map, 9fig., 43, 137 as objects of discipline, 29; empires and, 16;
Meiji Emperor, 23–24, 65, 104 in Qing dynasty, 20, 76, 191–192; vs. settlers,
Meiji Restoration, 54, 65–68, 114 16–17; Wilsonianism and, 36
meritorious service awards: for compliant necropolitics, 29
indigenes, 177; for Qing-period tongshi Niezen, Ronald, 2, 36, 38
 Index    305

normal administration, 110, 147–151, 156–157, 230, as teachers, 149–152; attacks on, 1, 112, 158;
235, 289n99 camphor production and, 28; Foucault and,
Northern Tribes, 9fig. 23; in colonial Korea, 3; in guardline
(aiyū-sen) movements, 66, 99–110, 233;
Obin Nukan, 137 military (kenpeitai), 117, 125; Pacification
Okada Shinkō, 87 Office and, 90, 93; political marriages and,
Ōkubo Toshimichi, 21, 23, 68, 76–78 136, 139–141; trading posts and, 180
Ōkuma Shigenobu, 67–68 police stations: as language schools, 120, 153, 155;
open the mountains, pacify the barbarians. Jiaobanshan, 235; Kasumigaseki (Taiwan),
See kaishan fufan 136; Wushe, 1, 112
Osaka Exposition (1903), 221–222 primitivism, 40, 163, 169
Oshikawa Noriyoshi, 90–93 Pu Chin, 85, 124–126, 200, 202fig., 205–208,
Osterhammel, Jürgen, 11–13, 77 206fig., 235
Ōtsu Rinpei, 106, 136–138, 141 public power, 22–23, 78, 82, 93, 103, 112, 155
Ōyama Tsunayoshi, 66 Puli, 7map, 58map, 88; as administrative center,
97, 97fig., 139; as contact zone, 88–89, 96,
Paalan, 7map, 9fig., 88, 96–97, 97fig., 113, 135–137, 130–131; elementary school (shōgakkō),
176, 208, 210fig., 239–244, 244fig. 153–154; Fukahori mission and, 95;
Pacification Office: Bukonsho (Japanese period), Pacification Office, 91–93, 129–131;
90–95, 92fig., 103, 129–134; Fukenju Qing-period guardline and, 98
(Qing-period), 87, 89, 215 punishment: regime of, 21–23; 26–32, 118, 248
Paiwan (ethnic group), 8map, 9fig.; as punitive expeditions (tōbatsu): Qing-period, 20,
culture-bearers, 162, 169, 185, 224; as noble 28; Japanese-period, 30–31, 96, 99, 103–105,
savages, 164, 167; as tribute-takers, 59–63; 131, 169, 224, 229–233. See also guardline.
as wreckers, 62–63; during kaishan fufan Puyuma (ethnic group), 8map, 9fig., 59, 122, 135,
(open the mountains, pacify the barbarians), 142–144, 143fig., 144fig., 165–166, 197, 205,
146–147; in Mudan village incident, 20, 50, 230, 246
52, 68–72, 72fig., 75, 202; Rover incident and,
56; social structure during Qing period, Qiu Ruolong, 5
56, 59, 165–166; Qing views of, 76; tōmoku quantification (calculation/rationality), 27, 76, 179
(headmen), 111; under Japanese colonial quasi-sovereignty, 13, 196–197, 255n45
rule, 142–149, 181, 232, 245–246; U.S. Quchi (Atayal subgroup), 9fig., 79, 122–123,
Consul LeGendre’s views of, 78. See also 216–222, 225–228, 226fig., 227fig., 230
Eighteen-Tribe Confederation (of Langqiao) Quchi (place), 7map, 79–81, 80map, 122–123, 131,
Pan Wenjie. See Bunkiet, Jagarushi Guri 132fig., 149, 150fig., 197, 216, 228
panopticism, 22, 31
Parijarijao, 165 race: administrative bifurcation and, 34fig.,
Paris Universal Exposition (1900), 183, 218, 221 150–152; biopolitics and, 30; ethnic
Pazzeh Watan, 124fig., 219, 221–223, 229, 242, classification and, 35, 191–192; expendable,
289n94, 289n103 223; in settler colonialism, 29–30; mimetic
Perry, Commodore Matthew, 55, 114 colonialism and, 194–195; second-order
photography, 161, 182, 225–229, 236, 241 geobodies and, 194; social Darwinism and,
Pickering, William, 59–61, 121 102; W.J.T. Mitchell’s definition of, 195–196;
Pixo Doleh, 139–140 science of, 212–216. See also racism.
Pixo Sappo, 129, 209–211, 266n122, 273n53 racism: against indigenes, 37; as logical and
Pixo Walis (Gao Yongqing), 96 temporal antecdent of “race,” 195; as Western
pluralism: cultural, 36, 191–195, 212–213, 217–229, invention, 284n10; “polite” vs. “vulgar,”
248; legal, 17–22, 251–252n8 29–30; scientific, 35
Poliac (Baoli), 52–53, 53map, 59, 61, 72, 121, 142 Raga, 9fig., 133
police: aborigine territory 32, 43–44, 47,119, Rainbow and Dragonfly exhibit, 161–162, 182
133, 138–139, 151–152, 156, 179, 232, 246–248; renaissance: indigenous cultural, 5–9, 13, 40, 168,
as feature of modern sovereignty, 21, 31; 187–188, 190
306     Index

Ri Aguai, 32, 98, 178 Shepherd, John R., 16

Richter, Daniel, 256n51 Shimaigaseki (Bukai) incident, 96–97
Rimogan, 7map, 9fig., 79, 80map, 132, 183, 221, Shimoyama Hajime, 139–141
228, 242–244 Shimoyama Jihei, 139–141
Rover incident, 56–57, 62 shipwrecks, 18–19, 48–50, 54–56, 60–67, 75,
Rowe, William T., 20 260n8
Rudolph, Michael, 187 Shiron (of Quchi), 123
Rukai (ethnic group), 8map, 120–121, 135, 169 Siaoliao, 53map, 56, 59–60, 63–65, 68
Ryūkyūans: massacre of fifty-four, 19, 48–55, sightseeing tours (for indigenes), 94, 152, 247
62–68, 75, 172 Simon, Scott, 167, 187, 251n6
Sino-Japanese War, 116, 129
Sabaree, 9fig., 53map, 146–148, 151, 163, 232 social Darwinism, 17, 102–104
Sagara Nagatsuna, 142–144, 148 Soejima Taneomi, 66–67, 172
Saigō Tsugumichi, 19, 55, 67–76, 72fig., 142, sovereignty: along the “savage border,”
146, 223 94–95, 202fig.; as achieved status, 212,
Saisiyat, 7map, 8map, 32, 37, 135, 169, 189, 219 248; bifurcated, 13, 17–18, 119; denial of, 15;
Saitō Makoto, 3–4 dynastic-state, 12, 17–18, 38–39, 55; global
Saitō Otosaku, 91, 128, 176–177 transformation and, 11; high-velocity
Sakano Tōru, 151 capitalism and, 22; horizontal integration
Sakuma Samata, 29–30, 72, 103–106, 136, 223, 245 and, 22; indigenous conceptions of, 6, 9,
Salamao incident, 112–113, 138 136, 249; international system and, 11, 15,
salt: as inducement, 89, 174, 176; as trade item, 18, 21, 33; Meiji period Japan and, 19, 23, 55;
172, 177, 180; as weapon, 93, 172, 178–180; nationalist conceptions of, 9, 11–12, 16–17,
embargoes of, 93–96; Taiwan 33, 77, 192; post-World War I conceptions of,
Government-General monopoly, 99 37, 196, 248; public power and, 78; Taiwan
Sasaki Yasugorō, 104 Government-General and, 14, 48, 82, 94. See
Satsuma, 51map, 56, 65–67, 76 also quasi-sovereignty
Sayon, 156, 278n178 statistics. See quantification
Sazuka Aisuke, 140–141 stipends (kensho), 118, 134, 155
Scott, James C., 219 Stone Gate Battle, 53map, 69–71, 223
second-order geobody: defined, 31–33; aborigine surrenders: to Saigō Tsugumichi (1874), 73fig.; to
territory as, 36, 135, 156, 247; as product of Taiwan Government-General, 32, 46, 48fig.,
colonial visual culture, 232–241; indigenous 104, 112, 135–136, 144, 179, 232, 272n41
ethnic groups and, 40, 194, 229–230, 248. Suzuki Hideo, 239
See also geobody Swinhoe, Robert, 120–121
Sediq, 5–9, 7map, 8map, 9fig., 43, 49, 88, 93, 96,
138, 140–141, 154, 253n26 Tadao Mona, 43–44, 49
Seediq Bale (film), 5, 43 Tai, Eika, 149
Segawa Kōkichi, 239 tairiku rōnin (continental adventurer), 104, 115
seinendan (youth corps), 9, 31, 119, 155, 247–248, Taiwan hakurankai (Taiwan Expo, 1935), 239,
278n175 244, 247
seiryokusha (men of influence), 111–112, 119, 155 Taiwan Indigenous Peoples (Taiwan
Seiyūkai (Friends of the Constitution), 4 Yuanzhuminzu): 169, 188, 248, 251n7
self-determination, 36 Takekoshi Yosaburō, 28, 77, 222
senkusha (pioneers), 115, 119 Tanaka Tsunatoku, 82–84, 125–126, 202fig.
settlers: as “brokers of empire,” 256n50; as ideal Ta-pa-ni incident, 156–157, 244
type in world history, 15–17; Han, in Tappas Kuras, 131
Qing-period Taiwan, 59, 79–82, 123, 165 Tata Rara, 142–144, 143fig., 144fig.
sha/she (administrative units), 6, 9fig., 147, 152, Tavares, Antonio C., 166, 178, 219
254n28 tax base: criteria for inclusion, 6, 47, 230;
shatei (local men), 127–128 expansion of, 17, 21, 63, 110; incomplete
Shen Baozhen, 20, 72 extension of, 225, 245; Qing, 20
 Index    307

tax revenue: Japanese government, 244; Washiiga, 125, 199–206, 202fig., 203fig., 204fig.,
Taiwan Government-General, 101table 206fig., 213, 235–236
taxation: for schools in Taiwan, 278n172; on land, Watan Nawi, 199, 200fig., 213
24, 26; on Qing frontier, 76; reform of, Watan Yū. See Watan Yūra
24–25 Watan Yūra, 86, 122–124, 174, 183, 196, 217–225
Taylor, George, 173, 197–198, 284n19 Watanabe (garrison commander), 82, 84fig., 125,
textiles. See Atayal:textiles 174, 198
Tgdaya (Wushe Sediq), 9fig., 88–89, 95–96, Watosinai, 86, 217
112–113, 129–131, 135–138, 140, 179 Weber, Max, 27–28
Thomas, Nicholas, 168, 191 whalers, 19, 55
Toda, 7map, 9fig.; as Sediq settlement, 131, 180; White, Richard, 16, 111
as Tgdaya enemies, 96, 112–113, 137–138, 176; Wilsonianism, 36–37, 151, 248
guardline and, 136–138, 179 Winichakul, Thongchai, 12, 18
Toketok, 56, 59–68, 71, 142, 162–165, 263n55 Wolf, Eric, 16, 36
Tokugawa shogunate, 18–22, 24, 55–56, 65, 214 Wolfe, Patrick, 29, 195, 284n10
tōmoku (headmen), 59, 81, 111–113, 119, 154–155, World War I. See Wilsonianism
165, 216 wreckers, 19, 67, 114
tongshi (interpreters/translators), 121–129, 162, Wu Micha, 278n182
166, 245 Wulai, 7map, 9fig., 58map, 80map. See also
Torii Ryūzō, 35–36, 133, 194, 198, 284n19, 182, 228, Quchi (place)
286n62 Wushe (Sediq subgroup). See Tgdaya
tourism, 3, 36–37, 156, 184–190, 229–236, Wushe (place), 1, 7map, 53map, 58map; 112,
278n178 135–140, 153–154, 176, 180, 209, 239
trading posts, 79, 87, 96, 139–140, 176–184, 245 Wushe Rebellion, 1, 43; causes of, 39, 49, 113, 119,
translation, 69–70, 114, 163 129, 140–141, 153–154; in popular culture, 5,
translators. See tongshi 43–44; impact on aborigine administration
Treaty of Peace and Amity (US-Japan 1854), 55 (riban), 30, 119, 155, 246–247; ramifications
Treaty of Shimonoseki (1895), 28, 255n42 in Japan, 4
Treaty of Tianjin (1858), 57, 98, 120, 162,
262n38 Xilai’an incident. See Ta-pa-ni incident
treaty-port: era, 98, 122, 127, 165–166, 172, 192; Xitou (Atayal subgroup), 7map, 9fig., 87–88, 126,
powers, 22; system, 54, 57, 61–65, 78, 88, 103 175, 213
tributary system, 48, 54, 62, 65–67, 122
Truku (ethnic group), 9fig., 110, 137fig., 220fig., Yamagata Aritomo, 29, 104
254n30 Yamaji Katsuhiko, 167
Truku (Sediq subgroup, prewar), 7map, 9fig., Yanaihara Sakimitsu, 66–67
112–113, 131, 136–137, 176 Yang Youwang, 52, 121
Ts’ai, Caroline Hui-yu, 23, 106, 157 Yao Jen-to, 21, 25–26, 245
Tsou (ethnic group), 8map, 86–87, 121, 181, Yawai Taimu, 119, 140
205, 235 Yayutz Beriya, 124
Tuilasok, 9fig., 53map, 59, 61–65, 68, 70–74, Yilan, 7map, 58map, 80map, 83, 87–88, 122–126,
142–148, 145fig., 151, 232, 277n148 135, 175–178, 213
Ying Gui, 57, 62
Uchida, Jun, 14, 256n50 Yoshimura Katsumi, 43–44, 49, 113
Ueno Sen’ichi, 78–82, 166, 173–175, 197, 214 Yūgai Watan, 220–222, 242fig., 289n103

Valjeluk Mavaliu, 50, 52 Zongli Yamen, 21, 23, 76

HISTORY | ASIAN STUDIES

Outcasts of Empire unveils the causes and consequences of capitalism’s failure to

“batter down all Chinese walls” in modern Taiwan. Adopting micro- and macrohistori-
cal perspectives, Paul D. Barclay argues that the interpreters, chiefs, and trading-post
operators who mediated state-society relations on Taiwan’s “savage border” during
successive Qing and Japanese regimes rose to prominence and faded to obscurity in
concert with a series of “long nineteenth century” global transformations.

Superior firepower and large economic reserves ultimately enabled Japanese states-
men to discard mediators on the border and sideline a cohort of indigenous headmen
who played both sides of the fence to maintain their chiefly status. Even with reluctant
“allies” marginalized, however, the colonial state lacked sufficient resources to inte-
grate Taiwan’s indigenes into its disciplinary apparatus. The colonial state therefore
created the Indigenous Territory, which exists to this day as a legacy of Japanese
imperialism, local initiatives, and the global commodification of culture.

“Sophisticated and engaging. This highly original narrative of a formative period will be
of great interest to all those concerned with comparative colonial history.”  NICHOLAS
THOMAS, Professor of Historical Anthropology, University of Cambridge

“A multisided and multiscale analysis—incorporating global, regional, and local scales—

embedded in a coherent and compelling narrative.”  PRASENJIT DUARA, Oscar Tang
Professor of East Asian Studies, Duke University

“Analytically precise and theoretically ambitious. A must-read for anyone interested in

the fate of indigenous peoples under modern colonialism.”  LOUISE YOUNG, author
of Japan’s Total Empire: Manchuria and the Culture of Wartime Imperialism

PAUL D. BARCLAY is Professor of History at Lafayette College. He is also general

editor of the East Asia Image Collection, an open-access online digital repository of
historical materials.

Asia Pacific Modern, 16

UNIVERSITY OF CALIFORNIA PRESS

www.ucpress.edu | www.luminosoa.org
A free ebook version of this title is available through Luminos, University of
California Press’s Open Access publishing program for monographs.
Visit www.luminosoa.org to learn more.

Cover illustration: Picture postcard. The photo’s caption reads, “Interpreters

standing behind submitted Aborigine headmen; 1910 Aborigine Punitive
Expedition.” Courtesy of the Rupnow Collection.