The Invention of Religion in Japan

By Jason Ananda Josephson

A study of how Japan once had no concept of "religion," and what happened when officials were confronted by American Commodore Perry in 1853.

Throughout its long history, Japan had no concept of what we call "religion." There was no corresponding Japanese word, nor anything close to its meaning. But when American warships appeared off the coast of Japan in 1853 and forced the Japanese government to sign treaties demanding, among other things, freedom of religion, the country had to contend with this Western idea. In this book, Jason Ananda Josephson reveals how Japanese officials invented religion in Japan and traces the sweeping intellectual, legal, and cultural changes that followed.


More than a tale of oppression or hegemony, Josephson's account demonstrates that the process of articulating religion offered the Japanese state a valuable opportunity. In addition to carving out space for belief in Christianity and certain forms of Buddhism, Japanese officials excluded Shinto from the category. Instead, they enshrined it as a national ideology while relegating the popular practices of indigenous shamans and female mediums to the category of "superstitions" —and thus beyond the sphere of tolerance. Josephson argues that the invention of religion in Japan was a politically charged, boundary-drawing exercise that not only extensively reclassified the inherited materials of Buddhism, Confucianism, and Shinto to lasting effect, but also reshaped, in subtle but significant ways, our own formulation of the concept of religion today. This ambitious and wide-ranging book contributes an important perspective to broader debates on the nature of religion, the secular, science, and superstition.

Praise for The Invention of Religion in Japan

"The Invention of Religion in Japan is truly revolutionary. Original, well researched, and engrossing, it overturns basic assumptions in the study of Japanese thought, religion, science, and history. . . . This book will absolutely reshape the field." —Sarah Thal, University of Wisconsin-Madison


"Written with remarkable clarity, this book makes an excellent contribution to the study of the interface of traditional Japanese religions and politics. Highly recommended." —Choice

"The range of Japanese primary sources consulted in his book is prodigious, as is his familiarity and usage of multidisciplinary theoretical works. . . . Josephson's book is erudite, informative, and interesting. It should be a worthwhile read for Japan scholars as well as scholars and students interested in religious studies theory and history." —H-Shukyo

• Language: English
• Publisher: Open Road Integrated Media
• Release date: Dec 22, 2022
• ISBN: 9780226412351

Book preview

JASON ĀNANDA JOSEPHSON is assistant professor of religion at Williams College.

The University of Chicago Press, Chicago 60637

The University of Chicago Press, Ltd., London

© 2012 by The University of Chicago

All rights reserved. Published 2012.

Printed in the United States of America

21  20  19  18  17  16  15  14  13  12     1  2  3  4  5

ISBN-13: 978-0-226-41233-7 (cloth)

ISBN-13: 978-0-226-41234-4 (paper)

ISBN-13: 978-0-226-41235-1 (e-book)

ISBN-10: 0-226-41233-4 (cloth)

ISBN-10: 0-226-41234-2 (paper)

ISBN-10: 0-226-41235-0 (e-book)

The University of Chicago Press gratefully acknowledges the generous support of the Office of the Dean of Faculty at Williams College toward the publication of this book.

Library of Congress Cataloging-in-Publication Data

Josephson, Jason Ananda.

The invention of religion in Japan / Jason Ananda Josephson.

pages ; cm

Includes bibliographical references and index.

ISBN-13: 978-0-226-41233-7 (cloth : alkaline paper)

ISBN-10: 0-226-41233-4 (cloth : alkaline paper)

ISBN-13: 978-0-226-41234-4 (paperback : alkaline paper)

ISBN-10: 0-226-41234-2 (paperback : alkaline paper) 1. Religion and state—Japan—History—19th century. 2. Japan—Religion—History—19th century. 3. Japan—History—Meiji period, 1868—1912. I. Title.

BL2207.3.J67 2012

322'. 1095209034—dc23

2012003656

This paper meets the requirements of ANSI/NISO Z39.48–1992 (Permanence of Paper).

The Invention of Religion in Japan

JASON ĀNANDA JOSEPHSON

The University of Chicago Press

Chicago and London

Contents

Preface and Acknowledgments

A Note on Texts and Translations

Introduction

The Advent of Religion in Japan

Obscure Obstacles

Unlearning Shūkyō

Unlearning Religion

Overview of the Work

1. The Marks of Heresy: Organizing Difference in Premodern Japan

Difference Denied: Hierarchical Inclusion

Strange Aberrations: Exclusive Similarity

Hunting Heretics

2. Heretical Anthropology

Contested Silences: Two Versions of the Acts of the Saints

Demonic Dharma

Japanese Heretics and Pagans

3. The Arrival of Religion

Negotiating Religion

Taxonomy and Translation: Category in the Webs of Meaning

Unreasonable Demands

4. The Science of the Gods

Shinto as a Nonreligion

The Way of the Gods

Celestial Archeology: The Advent of European Science in Japan

The Science of the Gods: Philology and Cosmology

Ritual Therapeutics for the Body of the Nation

The Gods of Science

From Miraculous Revolution to Mechanistic Cosmos

5. Formations of the Shinto Secular

Secularism Revisited

Hygienic Modernity and the World of Reality

Secular Apotheosis

6. Taming Demons

The Demons of Modernity

Restraining the Wild

Monstrous Gods

Evil Cults

Disciplining Buddhism, Expelling Christianity

7. Inventing Japanese Religion

Religion in Japanese International Missions

Controlling the Heart: Debating the Role of Religion in the Modern State

Inventing Japanese Religions

8. Religion within the Limits

Internal Convictions

External Controls

The Birth of Religious Studies in Japan

Conclusion

The Invention of Superstition

The Invention of the Secular

The Invention of Religion

The Third Term

Postscript

Appendix: Religion Explained

Notes

Character Glossary

References

Index

Preface and Acknowledgments

This monograph was born in Oxford in the depths of the Bodleian Japanese Library, where I came across a book by a Meiji-era Japanese Buddhist reformer, popularly nicknamed Doctor Monster (Yōkai hakase) for his public ghostbusting efforts. In this text, Doctor Monster repeatedly makes the argument that Buddhism is a religion. This troubled me; it seemed odd that an important Japanese thinker would have bothered going to such lengths to justify a truism. At first I thought I was making some basic error in translation. Then I began to wonder how representative this argument was and whether it might merely be the intellectual quirk of a noted eccentric. But anyone who felt compelled to argue, step by step, that Buddhism was a religion presumably thought that some of his readers needed convincing. Had Japanese people thought that Buddhism was something other than a religion? I started trying to answer that question.

Further browsing in Meiji-era Japanese texts raised more problems than it solved. Some thinkers argued that there were no proper religions in Japan, others that there were hundreds of religions in Japan. Some argued that Shinto was not religion but statecraft, that Confucianism was not religion but ethics, or that Buddhism in toto was nothing more than a failed politics or a backward superstition. I knew that there had been no consensus in the nineteenth century about how to translate the English term religion into Japanese, and I wondered if this had not complicated the matter. The Japanese government ultimately intervened to help define religion and separate it out from superstition, and also, oddly, to distinguish it from certain forms of Shinto. It struck me that the category religion was in play in the period and that its relation to indigenous traditions was far from fixed.

I became convinced that figuring out how this term religion worked in the Meiji period and how it was distinguished from science and superstition was essential to understanding Japanese Buddhism, Shinto, and Confucianism. This was true not just in the nineteenth and twentieth centuries. It seemed necessary to work out how religion had been formulated in Meiji Japan to avoid anachronistically projecting contemporary intellectual structures back into Japanese history.

The problem metastasized when I discovered that the predominant Japanese translation for religion, shūkyō (a functional neologism), had been exported to China and Korea, becoming the dominant term for religion in those places as well. In the process, Japanese interpretations of religion influenced the conceptual reorganization of national traditions across the region. Untangling the conceptual knot around Meiji shūkyō seemed necessary for understanding any East Asian religion. Perhaps not surprisingly, this trail led me to the complex intersections of the various threads of modernity—not simply between religion, science, superstition, and the politics of the nation-state, but less obvious connections such as that between science and the birth of the novel.

I have benefited from conversations about the book with people all over the globe. Lacking space to thank everyone by name, I will try and gesture at the institutions where I have spent time. I would like to thank friends, colleagues, librarians, and teachers at Oxford (St. Antony’s College); Harvard, Stanford, and Princeton; École Française d’Extrême-Orient (Paris and Kyoto); Ruhr Universität; the National Diet Library in Tokyo; and Williams College.

I also benefited from feedback by participants at talks I gave at the International Association of Buddhist Studies (London, 2005), the American Academy of Religion (Washington DC, 2006), Boston University (2007), Indiana University–Bloomington (2007), Asian Studies Association Japan (Tokyo, 2008), the Association for Asian Studies (Chicago, 2009), Smith College (2009), University of Wisconsin–Madison (2010), and Columbia University (2010).

Research in Japan and Europe was made possible by support from the Benson Fellowship, Stanford Center for East Asian Studies, Hellman Family, World Travel Fellowship, Käte Hamburger Kolleg, Dynamiken der Religionsgeschichte, and the generous support of Williams College.

I would like to thank my mentors—Carl Bielefeldt, Bernard Faure, and Helen Hardacre—who have been my guiding lights for many years and without whose support the book would not exist. For the tedious work of written feedback on early drafts or individual chapters or providing key sources, thanks go to Micah Auerback, Christopher Bolton, John Breen, Candice Cusack, Rebecca Davis, Peter Duus, Healan Gaston, Will Hansen, Jackie Hidalgo, Andrew Jewett, Neil Kubler, Levi McLaughlin, Keith McPartland, Okada Masahiko, Okajima Hidetaka, Makino Yasuko, Jon Pahl, Anne Reinhardt, Sango Asuka, Lee Schmidt, Eiko Siniawer, Jackie Stone, Jang Sukman, Stephen Teiser, and Michael Zimmerman. My appreciation also goes to Alan Thomas and the staff at the University of Chicago Press for their encouragement and help seeing this book into print and to Emily Ho for preparing the index.

The book would have been very different without the help of detailed comments from my anonymous reader, my formerly anonymous reader—Sarah Thal—and a small group of folks who read and commented on the penultimate version: Denise Buell, Eleanor Goodman, Hans Martin Krämer, and Christian Thorne. I will always be in their debt.

Thanks are due to Dalena Frost for her compassion, support and constant inspiration.

I would also like to thank the late Jacques Derrida, who suggested that I look to Japanese law codes to answer my research question.

This book is dedicated to my parents. Fubo no on wa yama yori mo takaku.

A Note on Texts and Translations

All translations from Japanese, Chinese, French, German, Dutch, Norwegian, Spanish, Portuguese, and Italian are my own unless otherwise noted. For transliteration of Japanese I have followed the Hepburn system, for Russian I have used the ALA-LC Romanization system, for Chinese I have used Pinyin, and for Korean McCune-Reischauer (excepting terms that have come into common English usage, such as Tokyo, for example).

Japanese names in the text are presented in Japanese order—surname followed by given name. The norm when writing about premodern Japanese figures is often to refer to them by their given name. However, since the project bridges modern and premodern periods, for the sake of consistency the family name will be used instead for all.

Introduction

On July 8, 1853, American warships appeared off the coast of Japan. As they were well aware, this was a direct violation of Japanese law. Still, the Americans ignored warning shots fired by cannons mounted in the bay and repulsed Japanese boarding parties with threats and brandished weapons.¹ Ready to cause an international incident if necessary, they pointed their new Paixhans canons at targets in Uraga Harbor and made signs that they wanted to parley with the local officials.² As Toda Ujiyoshi, the Japanese magistrate of the city of Uraga, soon discovered, the Americans were refusing to leave unless a pair of letters were delivered to the Japanese emperor. Religion appears twice in the official communications brought by the Americans, both times to assure the Japanese that the United States had no interest in violating the country’s centuries-old prohibition of Christianity.³ But on this point the Americans were not being truthful.⁴

When Japanese translators encountered the term religion, they had no idea what it meant.⁵ They produced multiple versions of the American letters, rendering religion with a range of terms, each of which implied something radically different.⁶ No word then existed in the Japanese language equivalent to the English term or covering anything close to the same range of meanings.⁷ This book will explore that conceptual disjunction, and the result: the process by which Japanese officials, when confronted with the Western concept in a moment of political crisis, invented religion in Japan.⁸

The Advent of Religion in Japan

Religion is based on the essential difference between man and beast—the beasts have no religion.

LUDWIG FEUERBACH, DAS WESEN DES CHRISTENTHUMS, 1841

Religion is intimate prayer and deliverance. It is so inherent in man that he could not rip it from his heart, without being condemned to separate himself from himself and to kill that which constitutes in him, his humanity.

AUGUSTE SABATIER, ESQUISSE D’UNE PHILOSOPHIE DE LA RELIGION D’APRÈS LA PSYCHOLOGIE ET L’HISTOIRE, 1897

Religion has generally been considered to be a universal aspect of human experience found in all cultures. Indeed, the most influential philosophers (Kant, Heidegger, Rorty), sociologists (Durkheim, Weber, Bourdieu, Habermas), and anthropologists (Geertz, Lévi-Strauss, Murdock) of the last two hundred years seem to have largely taken religion for granted as a cultural universal, individual atheists or nonbelievers notwithstanding.⁹ First and foremost, I want to challenge this presuppostion, which still permeates the modern academy, by demonstrating that in Japan religion had to be invented.

For these universalists, as well as scholars of Japanese religion and Japanese history, the chapters that follow rethink the very concept of Japanese religions by demonstrating that the category is historically conditioned and was consciously formulated to meet political ends. My point is not that words have etymologies but that the emplacement of certain categories transforms their members in demonstrable ways. As I show, defining religion in Japan was a politically charged, boundary-drawing exercise that extensively reclassified the inherited materials of Buddhism, Confucianism, and Shinto. All three traditions were radically changed in a way that has only recently begun attracting scholarly attention. We cannot presume a stable content for Japanese religions in this time period, let alone continuity over time. Thus, the book works to call into question concept of Japanese religions. In so doing, it will undermine the contention that religion is a natural category or a cultural universal.

While the universality of religion tends to be taken as a given in most academic departments, almost fifty years ago, Wilfred Cantwell Smith criticized essentialist approaches to religious studies, arguing that the term religion was of dubious usefulness.¹⁰ Although it took quite a while for these criticisms to gain traction, beginning particularly in the last twenty years, a diverse array of scholars from Talal Asad and Jacques Derrida to Daniel Dubuisson, Timothy Fitzgerald, Jonathan Z. Smith, and Tomoko Masuzawa have all repeatedly demonstrated that religion is not a universal entity but a culturally specific category that took shape among Christian-influenced Euro-American intellectuals and missionaries.¹¹ Building on this foundation, a further group of scholars have traced the European construction of various religions, from the British Discovery of Buddhism to the Jesuits’ Manufacturing Confucianism.¹² In so doing, these scholars have called into question the very field of religious studies by showing that religion masks the globalization of particular Euro-American concerns, which have been presented as universal aspects of human experience.

Although these genealogies of religion have been valuable for the discipline, they have in general focused almost exclusively on the writing of Europeans and Americans. Taken together, they have tended to foster the conception that religion attained international hegemony unilaterally and without resistance. Additionally, they have tended to ignore the other part of the equation—the other terms that were overwritten or produced in response to the globalization of the Euro-American religion. In a sense, they have been preoccupied with European etymology without tracing the other languages with which it has intersected. A side effect of this scholarship has often been to deny agency to very colonial subjects it was trying to empower.

This book contests the narrative that understands modernity to be simply the product of Euro-American culture exported to an imitative or passive Asia. It will recover the voices of non-Western cultures that were the object of that discourse. Compared to studies of, for instance, how the British invented Hinduism, this work is different in the sense that it is about how the Japanese invented religion in Japan. I will emphasize foreign pressures and international networks, but the main actors examined will be Japanese. Japanese intellectuals and policymakers were active participants in the formation of religion according to their own ideological agendas and programs of reform, agendas that could not help but shape European understandings of the subject.

If religion is supposed to be hegemonic imposition, it seems an oversight to sidestep a larger body of postcolonial theory and ignore the manner in which religion took dominance. The new theorists of religion have generally seen the origins and the function of religion globally, in terms of the anthropological enterprise, or as a scholarly or theological category, often imposed colonially in some ill-defined sense.¹³ This is not wrong, but it misses that religion in its modern imposition is largely a diplomatic and legal category and not just an academic or ethnographic one.¹⁴ Japan got religion because it was imposed upon the nation in a treaty, not because European scholars cavalierly applied the term to a foreign context. While studying the intellectual history of religion is valuable, it is a mistake to ignore the way the category has been articulated as a legal and diplomatic term. Considered only as an academic descriptor favored by anthropologists and sociologists, the history of the category religion appears secular, but looked at internationally and diplomatically, religion and calls for religious freedom in non-Western societies were a cover for Christian missionary activity. It was simultaneously secular and Christianizing in different registers.

Furthermore, in legally defining a cultural system as religion, that cultural system is then placed in free competition with Christianity. Equally importantly, defining something as not a religion—as a superstition—means that it can be purged or that the culture that sponsors it can be colonized. Asad has argued that the modern formation of religion produces a subjectivity distinguished from politics—a profound and private inwardness—but the matter is even worse than he thought, because religion turns out to be a rather odd kind of subjectivity. Even the most radical guarantees of freedom of conscience only demarcate a narrow and bounded kind of belief: a private inwardness, but not all that broad in scope.

One of the aims of this book is to extend this critical turn in religious studies by demonstrating how one non-Western nation made the discourse of religion its own. In Japan, the concept of religion had to be actively indigenized, and furthermore, this indigenization was undertaken largely by a centralizing government determined to reconfigure the internal constitution of the nation and to shore up its standing in the world. Looking at the formation of religion in Japan not only reveals its global diplomatic contours but also shows how the discourse of religion is woven into the fabric of modernity and how guarantees of religious toleration function to increase state power and to reconfigure entire cultural systems.

The European concept of religion expanded in part as a Christian universalization, but in so doing, it assimilated diverse cultural systems, which had little to do with religion’s Christian formulation. The ecumenical assertion that all cultures have religions has caused the category to be expanded beyond bounds that could reasonably be defended.¹⁵ The importation of Japanese and other non-Western traditions and interpretations into religion have already begun to undermine the category and produce the conditions by which a genealogical critique could be formulated. Put differently, the genealogical enterprise of Asad and company became thinkable, (in Bourdieu’s sense of the term), because the category of religion had already begun to crumble.

In investigating the formation of the category of religion, this book will also make a contribution to a set of debates about its entangled categories. Asad has argued that the genealogy of religion must be considered in light of the formation the secular.¹⁶ Serge Margel and others have emphasized the emergence of a bifurcation between religion and superstition.¹⁷ On top of this, Michel de Certeau and others have noted that science defines itself in opposition to superstition.¹⁸ To some extent, I think all of these scholars are right, but they have each only captured part of the system. Where they see binary operations, I see a trinary formation in which the real (or in its political form, the secular) is negated by superstition, which is in turn negated by religion. I will explore not only the genealogy of religion in Japan, but also the entangled genealogies of the secular and superstition.

Drawing on a wide range of historical materials, I trace the sweeping changes—intellectual, legal, and cultural—brought about by the construction of the category of religion in nineteenth-century Japan. These developments reflect back onto the global discourse of religion, such that the political goals of Japanese state officials have shaped, in subtle but significant ways, our own understanding of that concept today. I demonstrate that while the discourse of religion emerged in the context of Western Christendom, it is no longer exclusively Western in its current formulations. Rather, the concept was reformulated in the interstices, the international—in the spaces between nations and cultures. Today religion is no straightforward reflection of Western dominance, though certainly responsive to its imperatives; it is instead a transnational product of contested asymmetries of power.

Obscure Obstacles

Everybody who has visited Japan since Xavier first set his foot upon it agrees up to a certain point, and then all becomes vague and obscure.

WILLIAM BLACKWOOD, POLITICAL TRAGEDIES IN JAPAN, 1862

Sometimes in order to advance, to gain knowledge of a subject, we have to forget what we think we know. Thomas Kuhn writes, scientific progress requires us to unlearn the thought patterns induced by experience and prior training.¹⁹ In other words, we must recognize and overcome what Gaston Bachelard refers to as "obstacle épistémologique, that is, those mental structures and assumed truths that unconsciously shape our understanding of a subject.²⁰ As Paolo Rossi argues, these obstacles obscure breaks and discontinuities in our historiography.²¹ Particularly damning in both Bachelard and Rossi’s assessments is the false notion of continuity derived from language itself.²² We tend to assume when we see the same term over different periods of history that the same conceptual structures are implied.²³ But a word can have radically different meanings in different historical moments, and its reoccurrence in different periods can mask fractures, ruptures, and important shifts. As articulated by Nietzsche and subsequently by Foucault, we sometimes need to adopt a genealogical method, treating a concept not as an essence" preserved in perpetuity, but instead tracking discontinuities and transformations.

Let us begin in this chapter by unlearning the terms shūkyō and religion. For readers who are still unconvinced—for anyone who thinks—that religion is a universal, found in every society or among all peoples, the following genealogies will expose some of the term’s discontinuities and aporias. For those already conversant with this genealogical turn, it will locate the discursive object—Japanese religions—in this narrative.

Unlearning Shūkyō

A casual reader of the history of Japanese religion might believe that a version of this book’s argument has already been proposed and rejected by the discipline.²⁴ Three chapters of The Ideology of Religious Studies (2000) by Timothy Fitzgerald were devoted to a discussion of the construction of religion in Japan. Isomae Jun’ichi, in Kindai Nihon no shūkyō gensetsu to sono keifu (2003), also addresses the construction of religion in Japan, focusing particularly on the history of religious studies (shūkyōgaku) as a Japanese discipline.²⁵ These books were not well received, but I think they were basically moves in the right direction, even though neither scholar had the whole argument in place.²⁶ This book remedies their methodological deficiencies, provides new evidence from both a broader swath of Japanese history (1550–1945) and a broader set of sources (including Japanese, Chinese, and a host of European languages), and it extends the argument beyond religion to its entangled categories.²⁷ It also addresses what critics have described as Isomae’s and Fitzgerald’s greatest weakness.

Responding to both Fitzgerald and Isomae, Ian Reader has argued that the Japanese already had something analogous to a concept of religion ready at hand in the early nineteenth century. The crux of the debate is about continuities in the Sino-Japanese term shūkyō (Chinese [Ch.], zongjiao; Korean [Ko.], chonggyo), which was popularized in Japan in the 1870s as a translation for the Euro-American religion and was then exported throughout East Asia. Shūkyō is a compound of two characters, shū and kyō.²⁸ The character shū (Ch., zong) had long been used to mean sect, lineage, or principle, while the character kyō (also pronounced oshie; Ch., jiao) refers to teaching or teachings. In Confucian discourse the paradigmatic case of oshie was from ruler to subjects.²⁹ By the time the word reached Japan, oshie was used in a broad sense to cover not only pedagogy but also systematic knowledge in general. Shūkyō was a qualification and transformation of part of the old meaning of oshie. Thus, shūkyō could mean the teachings of a sect or the principles of the teachings. These characters had occurred together in the premodern period but only rarely and in reference to textualized Buddhist traditions; much of the current debate concerns the connection between modern and early usages of this term.³⁰

According to Reader and his associate Michael Pye, there is a significant continuity of usage between this premodern Sino-Japanese terminology and the modern concept of religion.³¹ They argued that the Japanese term shūkyō was a natural translation for the Euro-American concept of religion. This indigenous concept of shūkyō was supposed to tie "together the notions of an organisation or institutional identity (shū, sect/school) and a set of teachings (kyō) specific to it, and was supposedly equivalent to the meaning of religion."³²

Reader’s main evidence for this terminological continuity and its meaning rests on Michael Pye’s assertion that not only did Japan have an indigenous concept of religion, it also had independently developed its own Religionswissenschaft in the early modern period.³³ Michael Pye has argued that the academic study of religion in Japan was initiated by an Osaka-based intellectual Tominaga Nakamoto (1715–1746).³⁴ Although I would also agree with Pye that there were indigenous precedents for studying Buddhism that later came to inform the Japanese conversation about religion, I think his conclusions are fundamentally incorrect. The lynchpin in Pye’s argument is a single sentence in Tominaga’s critical appraisal of Buddhism, Discourse after Emerging from Meditation (Shutsujō kōgo), in which the characters nishūkyō make an appearance. Although provocative, Pye’s translation of this as two religions seems far-fetched upon inspection of the sources, which clearly indicate that Tominaga is discussing Manichean dualism. Hence, a more likely translation would be the teachings of two principles, basically, instead of ni shūkyō we have nishū kyō.³⁵ Shūkyō does not appear as an independent noun. Behind Pye’s project is an assumption that anyone discussing the three teachings as independent objects must be discussing three religions.³⁶ The evidence I will provide in this book goes against this claim.

For more than thirty years after Perry’s arrival, Japanese intellectuals grappled with the concept of religion. As will be discussed in chapters 3 and 7, they produced different translation terms, each of which seemed to imply a different object: a type of education, something unknowable, a set of practices, a description of foreign customs, a form of politics, a kind of Shinto, a near synonym for Christianity, a basic human impulse. That other terms with very different meaning were equally in play suggests it was far from clear what analogs existed for the Western concept. Not only did Japanese intellectuals and translators produce different terms for religion, they also debated which indigenous traditions and practices fit into the category. If religion was the genus, there were no native species on which taxonomists could agree. This is clear evidence that it is glib to talk of Japanese religion projected back through the centuries.

Finally and most crucially, even if shūkyō referred to the combination of an organization and a set of teachings or doctrines in the premodern period (which I dispute), this was not the Western definition of religion in the nineteenth century and it is not the definition of religion today. The following section will clarify what I mean.

Unlearning Religion

If you take dictionaries seriously, the meaning of the word religion has been undergoing a sea change over roughly the last thirty years. Many dictionaries written in this period in English, French, Italian, and German show a similar pattern: older editions of these works, such as the Le Petit Larousse that I had as a child, define religion as the worship of God or gods, while the more recent versions of these dictionaries have terms such as the supernatural or the sacred supplementing or standing in for the deity.³⁷ This global cultural shift seems to originate in the very gradual recognition that not all the cultural systems conventionally grouped under the category religion worship God. While dictionaries only conservatively track language change, this recent transformation in the meaning of religion is nontrivial and suggests both a secularization and radical broadening of the definition. The presence or absence of God is a critical difference. This shift, which I call the transition from a theocentric to a hierocentric definition of religion, is utterly minor compared to the changes that the term religion has undergone in a somewhat longer span, say, the last few hundred years.

The word religion is a fundamentally Eurocentric term that always functions, no matter how well disguised, to describe a perceived similarity to European Christianity. The hierocentric definition is merely a displaced theocentrism. This is conceptually dangerous, because European Christianity has been in a constant state of flux and its features at any given moment have been projected onto other cultural systems in a distorting manner. Further, I am not arguing that the phenomena generally described as religious did not exist before the modern period, but instead (in a Foucauldian vein) that they did not exist in the same aggregate form that has come to define current uses of the term.

On even a cursory analysis, the similarity of the words for religion in contemporary European languages is striking (German [Ger.], Religion; Dutch [Du.], religie; French [Fr.], religion; Italian [It.], religione; Spanish [Sp.], religion; Portuguese [Pg.], religião; Romanian, religia; Polish, religia; Russian, peлиΓия religiia). This parallel vocabulary demonstrates the terms’ shared philology.

Still, I will argue that at present there are at least two fundamentally different ways of thinking about religion in the Euro-American world. One is a secularized and globalized concept rooted in the sense that religion represents a discrete aspect of human experience. The other is an older but still modern definition that is implicitly theological and rooted in the assumption that a monotheistic divinity has revealed religion to different cultures. Although I do not wish to present a false sense of uniformity, for present purposes what I call the hierocentric definition of religion is best summed up in the current edition of Larousse as an established set of beliefs and tenets defining the relationship between man and the sacred.³⁸ This definition suggests that the special domain of religion is the sacred and that religion is rooted in propositions as an organized set of doctrines. While I think this definition is specious, not least because it rests on a false sacred-profane binary, many scholars would accept this definition and see the sacred (or the transcendent) as the core of religion and as an irreducible dimension of human experience.³⁹ This concept was not born in the last thirty years, but represents the culmination of two hundred years of sedimentation. It represents a partial de-Christianization of the category of religion, resulting from the importation of non-Western cultural systems into the category. Instead of buttressing the category, this definition is dangerously porous and does not provide sufficient diagnostic criteria for determining if something is a religion. In the case of Japan, this definition falls short. There is no native sacred-profane binary, and many of the cultural systems generally understood as Japanese religions are not founded on established beliefs for defining the relationship between man and the sacred.⁴⁰

Against this definition we can set what I call the theocentric concept of religion. For our purposes it is best exemplified in the famous Encyclopédie (1751–1772), which stated that the foundation of all religion is that there is a God who has dealings with his creatures and who requires them to worship him.⁴¹ This text, often understood as helping birth the secular study of religion, nevertheless defines religion in theological terms. This theocentric definition of religion has held sway in some form from roughly the seventeenth until the twentieth century. We do not have to go back to the eighteenth century to find versions of this model; the German scholar Ernst Troeltsch (1865–1923) described religion as flowing and alive, ever drawing on God through direct contact with him, utterly interior, personal, individual and abrupt.⁴² Built into the theocentric understandings of religion is an explicit monotheism, which understands religion to be identical to the worship of God, personified in different cultural forms. Ernst Feil notes that few scholars today in the study of religion would accept Troeltsch’s definition.⁴³ This definition is flawed in part because it so clearly parallels a Protestant understanding of religion as a personal relationship to God. According to that view, a monotheistic God, even if obscured or misunderstood, is at the center of religion. While it sounds generously ecumenical to suggest that the Japanese worship God in their own way, this does not accurately describe Japanese religions.⁴⁴ Disguised in the language of tolerance, this definition is condescending and distorting because it assimilates diverse entities (like the Confucian heaven and the historical Buddha) under the banner of a Christian God.⁴⁵ This was the definition that Perry and company literally carried with them to Japan, and we can see already why the Japanese had trouble finding an indigenous analog.⁴⁶

Both modern theocentric and hierocentric concepts encompass the same traditions (Christianity, Islam, Judaism, Buddhism, etc.). This is the case because the theocentric concept was central in producing the commonly accepted membership in the category. The identification of something as a religion was determined by the Euro-American attempt to recognize God’s role in different cultures. Therefore, it was precisely the idea that various traditions were modes of worshiping God that caused their inclusion in the category religion in the first place. While the hierocentric definition seems to represent a rejection of divine revelation in favor of the anthropological concept of the sacred, it inherited a cluster of religions based on the very principle that it ultimately rejected. Even once it discarded the old principle, the new definition was stuck with the preestablished list of religions, which was then used to generate a variety of vague commonalities necessary to formulate a new definition.⁴⁷ Therefore, the contemporary hierocentric concept only makes sense because of its reliance on an older concept to which it is no longer officially committed. Similarly, the theocentric concept also relies on older layers. This book will trace these layers in reverse chronological order.

In Japan: An Account (1852), a compendium of Western knowledge on Japan used by members of the Perry expedition, Charles MacFarlane paraphrases the remarks of a Dutch sailor who described the Japanese religious situation thusly: "There are twelve several [sic] religions in Japan, and eleven of them are forbidden to eat meat."⁴⁸ MacFarlane goes on concluded that the Japanese religious landscape consists of Buddhism and as many as thirty-four other religions or sects.⁴⁹ While other accounts of the period were more succinct, the taxonomy of Japanese religions was a problem not only for MacFarlane but for scholars over the whole course of the Euro-American engagement with Japan. At stake was more than the enumeration of Japanese cultural institutions, but the whole enterprise of scholarly study of religion in general.

This is because while the nineteenth century was particularly crucial in establishing the modern concept of religion, it did so in the face of vast classificatory enterprise as the modern conceptions of science, politics, and religion were only just being formulated and distinguished from each other.⁵⁰ Central in the period’s conception of religion was the work of German scholar Friedrich Max Müller (1823–1900) who was instrumental in formulating a new discipline for the academic study of religion (Ger., Religionswissenschaft).⁵¹ While contemporary scholars tend to regard this as a secular approach to the topic, the theocentric definition of religion was still very much in place. Müller had introduced the science of religion by stating that we can hear in all religions . . . a longing after the Infinite, a love of God.⁵² This was not lost on Müller’s colleagues. William Elliot Griffis, one of the major historians of Japan and Japanese religions, argued that the science of Comparative Religion is the direct offspring of the religion of Jesus. It is a distinctively Christian science.⁵³ As the scholarly objects of comparative religions were being formulated, Griffis and others like him were looking for discrete Japanese religions, understood as the worship of God in bounded systems of belief and practice. Given that the Japanese were strident in asserting that they did not worship a Christian God, this posed a large taxonomical problem, because it was difficult to identify how many different religions (understood in those terms) existed in Japan.⁵⁴

Both Müller and Griffis had inherited a classificatory schema from the previous generation. The early nineteenth century saw the emergence of much of this terminology, including the formation of the terms Boudhism (1801), Hindooism (1829), Taouism (1839), Zoroastrianism (1854), and Confucianism (1862).⁵⁵ This construction of religions was not merely the production of European translation terms, but the reification of systems of thought in a way strikingly divorced from their original cultural milieu. The original discovery of religions in different cultures was rooted in the assumption that each people had its own divine revelation, or at least its own parallel to Christianity.⁵⁶ In the same period, however, European and American explorers often suggested that specific African or Native American tribes lacked religion altogether. Instead these groups were reputed to have only superstitions and as such they were seen as less than human.⁵⁷ According to these explorers, superstitious peoples were uncivilized and thus could be colonized. Therefore, a lot was at stake in the application of this superstition-religion dichotomy.

Only a handful of Western scholars had been to Japan before 1853, and the early-nineteenth-century scholarly conception of Japan was established by four men with unusual experiences: Vasiliĭ Mikhaĭlovich Golovnin (1776–1831), a Russian explorer who had been caught by the Japanese while exploring the island of Kunashiri and held captive for two years;⁵⁸ Philipp Franz von Siebold (1796–1866), a German naturalist who had served as the resident physician at the Dutch East India Company outpost in Japan;⁵⁹ Germain Meylan (1785–1831), a Dutch merchant who had been Siebold’s former superior in Japan;⁶⁰ and Julius Heinrich (Jules Henri) Klaproth (1783–1835), a German expatriate based in France who had traveled in China and studied with Japanese castaways.⁶¹ These four individuals, publishing in Russian, German, Dutch, and French, provided the main sources for a whole generation of European and American writings on Japan. Their interpretations of Japanese religion varied considerably.

While all four could agree on the existence of Buddhism, their conceptions were quite different. According to Golovnin, for example, "the prevailing religion of Japan is derived from India, as the Japanese themselves attest, and is a branch of the religion of the Bramins [sic]."⁶² In effect, Golovnin was arguing that the Japanese were basically Hindus, a view that was widely held into the nineteenth century.⁶³ By contrast, Meylan argued that Buddhism (Boedsdo) was the name the Japanese gave to "to all religions [Du., Godsiensten] when they are imported from foreign countries.⁶⁴ In other words, Buddhism was not a specific bounded religion but a way to mark the cult of imported gods. While Siebold distinguished Buddhism from Hinduism, he believed that Buddhism was a form of idolatrous" monotheism based in the worship of a supreme god who appeared in different manifestations.⁶⁵ All of these representations of a theological Buddhism are important because they locate Japanese religion solidly in the theocentric concept of religion as the worship of God or gods.

Similarly, while all four referred to Shinto, all but Klaproth argued that Shinto was Japan’s primordial religion and accordingly Japan’s original monotheism. As Meylan put it, Shinto was based in the worship of a supreme deity (Du., Oppermatig Wezen) who was not depicted directly but was implicit in mirrors that were supposed to reflect God’s all-seeing eye.⁶⁶ This did not stop Klaproth and Siebold from having a well-publicized debate about whether Shinto was unique to Japan or whether it was merely Chinese Daoism. Nevertheless, this conception of Shinto emerged more from European preoccupations with urmonotheism than anything else.

All four recognized what we would call Buddhism, Shinto, and Confucianism as the three main Japanese religions.⁶⁷ However, in identifying these religions, they all falsely identified monotheisms or idolatries that reflect Christian concerns built into the category religion not endemic to Japan. Bizarrely, Meylan and Golovnin argued for the existence of a fourth religion. According to Meylan, there was a lost form of Japanese indigenous Christianity that predicted that a savior born of a virgin would redeem the world.⁶⁸ This fit the popular idea that there was indeed a Christianity indigenous to Japan.⁶⁹ This near consensus on the number of Japanese religions was less the result of personal observations than an inherited tripartite schema that structured their perceptions.

The idea that there was a diversity of Japanese religions emerged from the writings of German naturalist Engelbert Kaempfer (1651–1716).⁷⁰ After spending two years in Japan, Kaempfer passed away before the fruits of his efforts made print. Fortunately, some of his notes were published postmortem in 1727.⁷¹ Appearing in the midst of an Enlightenment fascination with Asia, Kaempfer’s work went through multiple translations and editions, ending up in the hands of scholars such as Kant, Goethe, and Voltaire.⁷² For these thinkers, Japan came to represent an unusual site of religious pluralism, in part because of Kaempfer’s account, which articulated a model of Japanese religions that ultimately would determine the shape of the field.

Kaempfer argued that the Japanese were unique in Asia, unrelated to the Chinese and instead descended from the first Inhabitants of Babylon.⁷³ As befitting a group whose ancestors he believed had constructed the original Tower of Babel and given birth to the profusion of languages, the Japanese had also produced a plurality of cults. Accordingly, Kaempfer divided Japanese religion into four types: the original Babylonian idolatry known as Sinto the old Religion or Idol worship, Budsdo The worship of foreign Idols, Siuto The Doctrine of their Moralists and Philosophers, and a lost Deivus or Kiristando is as much as to say the way of God and Christ whereby must be understood the Christian Religion.⁷⁴ In this language, we can see a transitional concept of religion emerging from a concept of idolatry. As will be discussed, to call something idolatry in the preceding centuries was to accuse it of being nothing more than an imitation or false religion. In Kaempfer’s writings, which were criticized by his contemporaries for being too positive, he had the temerity to suggest that forms of Japanese idolatry had some good qualities.⁷⁵ His calls to take Japanese religion seriously were unique, but he was not the first person—even in English—to suggest that Japan had different forms of idolatry.

One of the first popular English presentations of the tripartite schema of Japanese religions can be found in An Historical and Geographical Description of Formosa, an Island subject to the Emperor of Japan (1704).⁷⁶ This work, which was reputed to have been produced by a native of Japanese-occupied Taiwan (Formosa) then living in England, dedicates a chapter to a comparison of Taiwanese and Japanese religions.⁷⁷ Supposedly, there were at that time three kinds of Japanese religions: the first was dedicated to the worship of idols to which Japanese people sacrificed animals and even babies; the second was focused on the worship of one god in the form of the sun and moon; and a third was a form of atheism whose followers believed in no god but made sacrifices to propitiate various spirits and demons.⁷⁸ Setting aside for the moment the role of sacrifice in the above example and the fact that Japan would not actually colonize Taiwan until 1895, it looks quite a bit like the model that Kaempfer would inherit.

Unfortunately, An Historical and Geographical Description of Formosa turns out to have been written by George Psalmanazar (1679?–1763), who was not from Taiwan but was rather a European imposter and fraud.⁷⁹ This did not stop his work from being popular, however, and it went through multiple editions and seems to have been widely regarded as factual for a time.⁸⁰ It represented the beginnings of a popular tripartite schema in English, which Psalmanazar likely produced by lifting from descriptions of China rather than Japan. In this sense, the text is not radically different from the Latin account provided by Athanasius Kircher in China Illustrata (1667).⁸¹ To depart from European scholarship for a moment, the ultimate origin for this schema is likely the Chinese three teachings (sanjiao; Japanese [Jpn.], sankyō). This division of the Chinese cultural landscape into Buddhism, Confucianism, and Daoism was originally formulated in the Tang Dynasty (618–907) and then exported to Japan, where it was much later transformed into a reference to Japanese Buddhism, Confucianism, and later Shinto. Psalmanazar cavalierly applied a Chinese category to Japan that some Japanese were beginning to use domestically. With a crucial difference, in Chinese the three teachings did not refer to three religions, or three forms of idolatry for that matter, but instead represented the claim that different modes of discourse could describe